BULLETIN OF THE UNITED STATES NATIONAL MUSEUM CONTRIBUTIONS FROM THE UNITED STATES NATIONAL HERBARIUM VoLuME 38 [final volume of series] FERNS AND GESNERIADS C. V. Morton SMITHSONIAN INSTITUTION PRESS e WASHINGTON, D.C. e 1974 Publications of the United States N ational Herbarium The United States National Herbarium, which was founded by the Smithsonian Institution, was transferred in the year 1868 to the Department of Agriculture and continued to be maintained by that department until July 1, 1896, when it was returned to the official custody of the Smithsonian Institution. The Department of Agri- culture, however, continued to publish the series of botanical reports entitled “Contributions from the United States National Herbarium,” which it had begun in the year 1890, until, on July 1, 1902, the Na- tional Museum, in pursuance of an act of Congress, assumed responsibility for the publication. The first seven volumes of the series were issued by the Department of Agriculture. ii Part 1. Part 2. Part 3. Part 4. Parr 5. Part 6. Part 7. CONTENTS A revision of 7richantha (Gesneriaceae). By Con- vows rad V. Morton. (Published October 9, 1963). . . 1-27 Studies of fern types, I. By C. V. Morton. (Pub- lished August 31, 1967). ... 6.0.02 ee eee 29-83 The genus Grammitis in Ecuador. By Conrad V. Morton. (Published October 17, 1967). ..... 85-123 The Peruvian species of Besleria (Gesneriaceae). By C. V. Morton. (Published March 28, 1968) . . 125-151 The genera, subgenera, and sections of the Hymeno- phyllaceae. By C. V. Morton. (Published June 12, 1968) 2. eee ee ee es 153-214 Studies of fern types, II. By C. V. Morton. (Pub- lished December 30, 1978)... 666. ee ee eee 215-281 William Roxburgh’s fern types. By C. V. Morton. (Published September 20, 1974)........--. 283-396 [Part 8, The American species of Gleicheniaceae, by C. V. Morton, reported in the literature as “in preparation,” was never completed by the author. | ili Conrad Vernon Morton (1905-1972)' Davip B. LELLINGER ’? Conrad Vernon Morton, named Walter Vernon Morton at the time of his birth on 24 October 1905, was the son of Walter Crow Morton and Nioma (later Noma) Bartholomew Morton. He was born in Fresno, California, where his parents had been life-long residents and his father had a roofing and building contractor’s business. Walter Crow Morton died when Walter was still young, and in about 1917 his mother married Alva B. McCray, a friend of the family who also had been a resident of Fresno for many years. Mr. McCray was employed by the Santa Fe Railroad. The family home at Fresno included a vineyard and various gardens on the property, and so young Walter was introduced to plants at an early age. As a child, he planted and tended a small garden of his own. His mother was well known in the area for her rose garden; she grew about 130 varieties abundantly. During his junior and senior high school years, the family lived in Berkeley, where Mrs. McCray’s sister and family also resided. The climate there, more moderate than in Fresno, was better for Mrs. McCray’s health. When he was in high school, Walter won a fellow- ship to the local Art Institute. Of all the students in the class, he was the only boy and, for this reason, he did not pursue his study of art very far at that time. After he began his professional career, how- ever, he studied both painting and the piano diligently and mastered and practiced both arts with great pleasure for himself and his friends for the remainder of his life. He was also interested in the history of the cinema, in dramatics, and in philately. (For a more detailed account of Morton’s nonbotanical interests, see J. A. Ewan, 1973, Taxon, 22: 271-274.) Morton entered the University of California at Berkeley in 1924. 1An earlier version of this essay was issued 22 June 1973 in the American Fern Journal, 63(2) : 25-48. The author is indebted to Mr. Morton’s sister, Mrs. Howard King, for most of the details of his early years. 2 Department of Botany, National Museum of Natural History, Smithsonian Institution, Washington, D.C. 20560. At that time he changed his first name to Conrad (the legal change was made on 14 April 1926). His interests at first included the physical sciences, mathematics, astronomy, and slavic languages and literature. During the latter part of his freshman year he began a course in general botany, and in his sophomore year he studied elementary taxonomy under Prof. Willis L. Jepson. Apparently this course trig- gered a strong interest in taxonomic botany, for he took several more courses from Prof. Jepson, algology and mycology from Prof. Gardner, plant physiology from Prof. Holman, and cytology and genetics from Prof. Goodspeed. Morton’s first. botanical collections were made around Berkeley in January 1926. He held tuition scholar- ships at the university for the last three of his undergraduate years: a Levi Strauss Scholar in 1925-26, a Henry Morgan Holbrook Scholar in 1926-27, and a Carrie M. Jones Scholar in 1927-28. He was elected to membership in the scientific fraternity Sigma Xi and in the biological sciences fraternity Phi Sigma. In October 1926 he was inducted into the Alpha chapter (California) of Phi Beta Kappa. He received a Bachelor of Arts degree cum laude from the University of California in May 1928, After he graduated, he was appointed a Teaching Fellow in Botany for the summer session, but he resigned to take a position as Phanero- gamic Aid in the Division of Plants, United States National Museum. Smithsonian Institution. Prof. Jepson wrote a highly favorable recommendation in his behalf, In his early years at the Smithsonian, Morton worked on flowering plants in general. His first assignment, he once told the author, was to identify some specimens from North Africa. This, in the face of scant literature and fewer specimens, was an initiation—which he surmounted. Much of his time was taken up with identification of phanerogams from the western United States, the Caribbean, Mexico, and Central America, He soon took up the Gesneriaceae and Solanaceae as specialties: both became life-long interests. He spent the spring of 1933 making general collections in Oaxaca, Mexico. On this trip he fell ill with malaria. which recurred from time to time in succeeding years. In 1936 and 1941 he took collecting trips to Cuba. In addition to his research interests, he made thousands of identifica- tions, helped to curate the herbarium, began to compile an index of the genera to the flowering plants (which was not completed), and made indices to some of the larger fern genera in the New World based on the “Index Filicum” and the “Gray Herbarium Card Index.” In 1939 he was appointed Assistant Curator of Phanerogams in the Division of Plants. He also began to take post-graduate courses in botany at nearby George Washington University. Although he did not hold a master’s degree, he was admitted to the doctoral program 2 on the strength of several lengthy scientific papers that he had published. Having published on ferns occasionally since 1932, by the middle 1930s he began to devote more and more of his time to research on, and identification of, the ferns. This was done under the guidance of Dr. William R. Maxon, who was a fern specialist and the Curator of the Division of Plants. After Maxon’s retirement in 1946, the Division of Plants was reorganized as the Department of Botany. Morton was appointed Associate Curator of the Division of Ferns. At this time, owing to the press of official duties, he withdrew from the doctoral program at George Washington University. The following year he made extensive general collections on the island of St. Vincent and explored a newly discovered valley in the mountainous interior of the island. In 1948 he was appointed Curator of the Division of Ferns, a position he held until 1970. From 1948 to 1959 he also held the title of Acting Curator of the Division of Cryptogams. This was a period of severe staff shortages, and his role in that division did not extend beyond maintaining the collections. From 1950 to 1954 Morton was a member of the International Botanical Congress Committee on Nomenclature. In 1951 he accom- panied Dr. L. O. Williams on a field trip to Honduras. The fern specimens from this trip were to help form the basis for a fern Flora of that country, but the project remained in rough manuscript, as did a similar project on the ferns of Guatemala. Morton realized that the greatest obstacle to preparing tropical fern floristic works was inadequate nomenclatural research. He knew that long trips to study in European herbaria would be necessary to produce adequate treatments of the ferns. He applied for, and in 1954 received, a Simon Guggenheim Foundation fellowship to study type specimens of ferns in European herbaria. The National Science Foundation awarded him grants for the same purpose in 1957 and 1963. He searched portions of the herbaria at Berlin, Brussels, Copen- hagen, Edinburgh, Florence, Geneva, Hamburg, Kew, Leiden, Liver- pool, London (British Museum), Oslo, Oxford, Munich, Paris, Stockholm, Utrecht, Vienna, and Zurich, and took more than 22,000 photographs, including some in his phanerogamic specialties. The extensive labels that accompany the photographs are valuable because they convey the results of Morton’s painstaking research into the type status and taxonomic disposition of each specimen. In 1956 he took a third and last field trip to Cuba. From that time on he was fully occupied with administrative, research, and identifi- cation duties. Occasionally he attended field forays sponsored by the American Fern Society. He also took several pack trips in the western United States, but these were an extension of his interest in wilderness and its conservation and were not undertaken from a botanical stand- point. In 1961 he suffered the first of a series of heart attacks, For a decade he felt that these precluded field work, Despite his afflictions, which included the almost total loss of sight in one eye because of a detached retina, he continued to work in the herbarium and to take frequent trips to other herbaria. From the early 1960s he worked mostly on ferns and, for a time, did little with his phanerogamic specialties. He headed the Inter- national Botanical Congress Special Committee for the Study of Superfluous Names during its period of activity in 1968 and 1969. In June 1970 he represented the Smithsonian Institution at the tercentenary celebration of the Royal Botanic Garden, Edinburgh. He greatly enjoyed a tour on which he visited gardens all over Scotland. In 1970 he was made a Senior Botanist at the Smithsonian, a position that freed him from administrative duties. He returned to his interest in the Solanaceae and began to complete a manuscript on Argentine Solanum. He was awarded a Smithsonian Research Foundation grant for field work and herbarium study in Argentina and spent three months there in early 1972. On his return from the field, he continued to work hard on his Solanum manuscript and seemed to be in better health than usual, but a few months later, on 29 July 1972, he passed away unexpectedly at his home in Washington. Morton was an honorary life member of the American Gesneria Society and an honorary member of the American Fern Society. He served as associate editor of the American Fern Journal from 1941 to 1947 and from 1962 to 1972, and he was the editor-in-chief from 1948 to 1961. He is commemorated by the genera Mortoniella Woodson and Mortontodendron Standl., and by the following botanical names: Asplenium mortonii Duek, Banisteria spruce? var. mortoniana Macbr., Besleria mortoniana Steyerm., Capsicum mortonii Raymond, Clidemia mortoniana Standl., Columnea mortonii Raymond, Cyphomandra mortoniana Smith & Downs, Erigeron mortonianum Reveal, Hupa- torium X mortonianum Alain, Gesneria mortonii Wiehler, iraea spruce? var, mortoniana Machr., Inga mortoniana Leon, Milla mor- toniana Moore, Niphidium mortonianum Lellinger, Schoenocaulon mortoni Brinker, Selaginella mortoniana Crabbe & Jermy, Tetrapteris jamesonii var. mortonii Macbr. [=7. mortonii (Macbr.) Cuatr.], Phelypteris mortonti A. Reid Smith, Tigridia mortonit Molseed, Topobea mortoniana Wurdack. Viburnum mortonianum Standl. & Steyerm., Witheringia mortonii Hunziker, Xiphopteris mortonii Copel., and Microgramma mortoniana de la Sota. UNITED STATES NATIONAL MUSEUM CONTRIBUTIONS FROM THE UNITED STATES NATIONAL HERBARIUM VoLuME 38, Part 1 A REVISION OF TRICHANTHA (GESNERIACEAE) By Conrap V. Morton BULLETIN OF THE UNITED STATES NaTIonAL MusEuM SMITHSONIAN INSTITUTION e WASHINGTON, D.C. e 1963 A REVISION OF TRICHANTHA (GESNERIACEAE) Conrad V. Morton The small genus Trichantha Hooker, composed of small, epiphytic herbs or subshrubs with red or red and yellow striped, conspicuous flowers, grows in the mountains of Panama, Colombia, and Ecuador. A member of the African violet family, it was described in 1844 based on two new species, each known from single collections made by Thomas Lobb from unspecified localities in Colombia. The genus has remained one of the least known of the Gesneriaceae. It was originally distinguished from Columnea and Alloplectus primarily by the presence of small, clavate appendages borne on the corolla tube at the base of the sinuses of the lobes of the limb. These appendages, conspicuous in the living plants but not always obvious in dried material, are of unknown function. Hanstein! reduced Trichantha to a synonym of Columnea sect. Stenanthus. Bentham and Hooker retained it as a distinct genus, placed between Columnea and Alloplectus, noting that in some char- acters it was closer to Alloplectus than Columnea. Fritsch in 1897 (in Engler and Prantl, Die Natiirlichen Pflanzenfamilien) expressed the same opinion. No other species have ever been described in Trichantha. However, in 1959 I described a cultivated plant as Columnea moorei,’ placing it in the section Systelostema rather than Stenanthus because of the leaves being in equal rather than unequal pairs. I did not observe any appendages on the corolla tube, although there is a tuft of hairs in the sinuses where the appendages would be, but Dr. H. E. Moore, who has material growing in the Bailey Hortorium did observe minute appendages in living plants. He subsequently sent me some fresh flowers and I was able to see them too, although it must be admitted that they are extremely minute, almost aborted, and might be overlooked by anyone not definitely searching for them. The presence of appendages indicated a probable relationship be- tween C’. moorei and Trichantha. 1 Linnaea 34:387. 1865. 3 Baileya 7:55. 1959. 2 CONTRIBUTIONS OF THE NATIONAL HERBARIUM Subsequently, Dr. Moore described another cultivated plant as Columnea illepida Moore.* This species, quite unlike any previously known species of Trichantha, was also placed in sect. Trichantha on the basis of the presence of appendages on the corolla tube. A search through the vast quantity of unnamed Gesneriaceae in the US. National Herbarium brought to light eight additional species of the most varied nature in regard to the habit, leaves, calyx, and corolla but all with appendages on the corolla. My opinion as to whether all these plants are properly associated by this one character has regularly varied. It is conceivable that these appendages have originated independently in several unrelated groups. On the other hand after now having written careful descriptions of all of them and having thought deeply about them, I believe that they do belong together. Although they are varied I cannot find that any one of them seems to be more closely related to any species of another group of Columnea or Alloplectus than to each other. I have veered from this way to that also in regard to the question of including Trichantha in Columnea. The difficulty is that the sec- tions of Columnea are based partly on the character of whether the leaves of a pair are equal or strongly unequal. But in my species of Trichantha both kinds of leaves occur, and its species would therefore have to be distributed between two sections, of Columnea (sect. Stenanthus and Systelostoma) or else Trichantha retained as a section based on the single character of corolla appendages with leaf characters varying as they do not in other sections of Columnea, an uncomfort- able situation. It may be that the leaf characters used in Columnea are not fundamental and should be abandoned as sectional characters, but I am not as yet prepared to go that far. Another objection to retaining Trichantha as a section of Columnea is that it really does have relationships with Alloplectus, especially in the character of the quadrate or broader than long anthers. The anthers of Colwmnea are, at least typically, oblong. Some species of Alloplectus sens. lat. to be sure do also have oblong anthers, but in- vestigation may show that these species are referable to Drymonia or perhaps represent a distinct genus Erythranthus. Inasmuch as the question of generic limits in this group of Gesneri- aceae is still controversial, it seems to me wisest to retain Trichantha as a genus, with the full realization that further research will very likely bring some realigznments. The number of essentially unknown described species and the number of undescribed species, particularly from Colombia, is so large that any attempt at generic or sectional readjustment would be premature. 'Balleya 8:56. 1960. REVISION OF TRICHANTHA—MORTON 3 Genus Trichantha Trichantha Hook. Icon. Plant. 3: t.666. 1844. Columnea sect. Trichantha Kuntze in Post & Kuntze, Lex. Pl. Phan. 136. 1904. Small more or less scandent epiphytic subshrubs or succulent, shrublike herbs, with slender, usually hirsute stems, these often radicant, unbranched above the base; leaves of a pair equal or un- usually very unequal, the smaller often stipule-like and often soon deciduous, the larger rather short-petiolate, the blades lanceolate to ovate or even suborbicular, acuminate or cuspidate or rarely obtuse, usually cuneate at the unequal base, rarely rounded, herbaceous in texture, remotely and minutely denticulate above the middle, or (in T. moorei) just slightly crenulate, dull green above, rarely shining (in J. moorei), green or usually tinged with red beneath, hirsute above with many-celled, yellowish or reddish elongate hairs, hirsute beneath (or rarely strigose), often also sparsely strigillose with minute, straight, appressed, 2-celled, “bayonet’’-shaped hairs, the basal cell minute, yellowish or dark, the terminal cell milky-white, very sharp at apex, pinnately veined, the lateral veins mostly 4-7 pairs, obscure above and slightly raised beneath; bracts at base of peduncles minute, hirsute, usually subulate, deciduous; flowers solitary, axillary, or sometimes geminate or ternate, the peduncles elongate, often thickened toward apex, hirsute, often pendulous; calyx 5—lobed to base or nearly, the lobes very slightly unequal, the posterior a little shorter and narrower, usually red but sometimes green or green at base and red at apex, entire (in 7’. aliena), merely deeply toothed (in T. illemda), or deeply, pectinate-dissected with narrow central portion and 2-4 pairs of elongate, linear-filiform lateral lobes, hirsute externally with multicellular, usually red-walled hairs and often also with minute appressed 2-celled hairs, often with sessile glands also, inside glabrous, strigose, hirsute, or densely sericeous, often with numerous sessile glands toward base; corolla slightly oblique in calyx, essentially tubular but sometimes a little ventricose in the middle and slightly contracted in the throat, saccate at the posterior base, hirsute with multicellular hairs externally or sometimes densely long-tomentose (in 7. rosea and T. aliena), glabrous within except for the glandular- pilosulous throat, the limb usually narrow, the lobes erect or rarely spreading, slightly or obviously bilabiate, the two upper lobes partly or almost wholly united into an erect galea, the two lateral lobes mostly deltoid, the lower lobe shorter and narrower, hirsute or tomen- tose externally, glabrous within except for the glandular bases, the tube clear red or usually striped, the base color yellow with dark maroon stripes beginning at the base of the corolla or near the middle and running to the sinus of the lobes or occasionally stopping short 4 CONTRIBUTIONS OF THE NATIONAL HERBARIUM of the limb, the lobes all yellow, yellow with dark margins, red with yellowish margins, or the galea all dark red-maroon, the sinuses of the limb provided with elongate, linear or clavate appendages, these spreading horizontally, yellow or red, hirsute (especially at apex) or long-red-tomentose (in T. rosea), or the appendages obsolete and hardly to be observed in dried material (7. moorez); stamens 4, connate at base into a sheath up to 1 em. long, adnate to the very base of the corolla tube, the filaments glabrous, slender, elongate; anthers glabrous, quadrate, or broader than long, at first lightly connate into a square, briefly exserted, later withdrawn into the corolla tube by contortion of the filaments, becoming free; stami- nodium none; ovsry sericeous; style elongate, glabrous or usually pilosulous; stigma bilobed, exserted after the anthers are withdrawn; disk reduced to a large or small, bilobed or entire, glabrous, fleshy, posterior gland, or occasionally (in 7. aliena) 2 small free, posterior glands; fruit a white spherical berry (in T. teuscheri, the only fruit known). Lectotype: Trichantha minor Hook. Of the two original species, T. minor is the first one listed and the one combined by Hooker with the generic description. Key to the Species Corolla tube densely hirsute-tomentose, the surface completely concealed; corolla lobes fleshy and thick, erect, small, subequal; corolla completely glabrous within throat; style glabrous. Leaves of a pair subequal, green beneath; calyx lobes entire, ca. 2.8 em. long, green at base and red at tips; appendages of the corolla 4—6.5 mm. long, red-hirsute. 2 2... Te aliena Leaves of a pair strongly unequal, the smaller stipule-like and soon deciduous, strongly tinged with rose beneath, or rarely green; calyx lobes not more than 1.5 cm. long, deeply pectinately dissected, with 2—4 pairs of linear lateral lobes, red throughout; appendages of the corolla extremely con- spicuous, 15-20 mm. long, densely rose-tomentose. . . . . . 2. T. rosea Corolla tube merely hirsute, much of the surface visible; corolla lobes, at least the lower three, thinner, spreading, the upper two lobes at least partially united into an erect galea; corolla glandular-pilosulous within throat, at least at the base of the galea; style pilosulous. Calyx lobes spathulate, entire; leaf-blades lanceolate, 5-6 times as long as wide. Leaves of a pairsubequal . . . ...... . . . . . 3. T. formosa Calyx lobes toothed or dissected; leaf-blades oblanceolate, lanceolate, elliptic, or suborbicular, relatively broader. Leaves of a pair equal, very small, 14-16 mm. long; internodes very short only about 1 em. long; appendages minute, hardly observable in dried material; calyx lobes green, with conspicuous red hairs , . 4. T. moorei Leaves of a pair strongly unequal, the small less than half as long as the larger, often stipule-like and deciduous, the larger at least 30 mm. long or usu- ally much more; internodes at least 2.5 em. long (except in 7’. illepida) ; appendages well developed; calyx lobes red (green in 7’. tllepida). REVISION OF TRICHANTHA—MORTON 5 Calyx lobes green, pectinate-toothed but not dissected, the teeth 4-6 pairs on each side of the lobes, not more than 2 mm. long; internodes short (1-2.5 em. long) and stout (the upper ca. 3 mm. in diameter); leaf- blades red-blotched beneath or red allover... ... 5. T. illepida Calyx lobes red, deeply dissected, the linear-filiform lateral lobes 2-4 pairs, 4-15 mm. long; internodes longer and more slender; leaf-blades green beneath or sometimes tinged with reddish but not red-blotched or deep red all over. Leaf-blades elliptic or broadly elliptic, short-acuminate or cuspidate, large, the larger 14-15 cm. long and 5 em. wide or more, hirsute; internodes elongate, mostly 7-12 cm. long (or shorter in T. tropicalis) ; appendages yellow with maroon apex (or clear yellow in T’. tropicalis). Corolla clear red, not striped except for 2 inconspicuous dark stripes near the apex on the upper side, these running to the margins of the galea 6 6 ee ee 6. T. tropicalis Corolla yellow or ‘citron green,” with broad vertical stripes of dark maroon. Leaf-blades foveolate above, not bullate, beneath without minute, white, 2-celled, appressed, ‘‘hbayonet’’-shaped hairs mixed with the long, pluricellular hairs; galea and lateral and lower lobes apparently alldark maroon. . . . - 1. ++ = 7, T. major Leaf-blades bullate above, the long hairs borne on tubercles, not foveolate, beneath with some minute “bayonet’’-shaped hairs among the long ones; galea yellow with maroon margins, appar- ently the lateral and lower lobes all yellow. . . 8. T. bullata Leaf-blades lanceolate or elliptic-lanceolate, long-acuminate, the larger not more than 5.5-9 em. long and 3.7 em. wide; internodes less than 7 cm. long; appendages clear yellow (red in T. clara). Leaf-blades glabrous above except near margin, not at all hirsute beneath but sparsely strigillose with minute, white, appressed, 2-celled, ‘“bayonet’’-shaped hairs. Galea dark maroon, the lateral and lower lobes yellow. . . . . + - + + + 9 T. minor Leaf-blades hirsute above, hirsute beneath or strigillose. Corolla clear light red throughout, not at all striped, the galea darker red, the lateral and lower lobes red with very narrow yellowish margins; appendagesred. . . - - +: 10. T. clara Corolla obviously striped with dark maroon, with narrow yellow stripes, the galea maroon and the other lobes yellow or all lobes yellow; appendages yellow. Corolla tube with broad maroon stripes running to the sinuses of the lobes or up into the galea; galea dark maroon, the other lobes yellow . 2 2. ee eee ee ee OD T. teuscheri Corolla tube with broad maroon stripes and narrow yellow stripes, the maroon stripes fading out before reaching the sinuses of the corolla; galea and other lobes all yellow. . 12. T. elegans 1. Trichantha aliena Morton, sp. nov. Caules rubri internodiis elongatis crassis rubro-hirsutis; folia de- cussata, per paria subaequalia, breviter petiolata, laminis oblongis vel oblanceolatis, utrinque viridibus, apice breviter acuminatis, basi late cuneatis obliquis, utrinque hirsutis subtus non strigillosis; flore 6 CONTRIBUTIONS OF THE NATIONAL HERBARIUM solitarii longe pedunculati; calycis lobi magni integri lanceolati basi angustati apice longe et acriter acuminati, deorsum virides sursum rubri externe hirsuti intus albo-sericei et glandulosi; corolla magna rubra tubulosa basi saccata, tubo externe valde ubique rubro-hirsuto, intus glabro et eglanduloso, limbo vix bilabiato, lobis subliberis sub- aequalibus crassis erectis posterioribus marginibus inflexis, intus glabris, appendicibus linearibus magnis hirsutis; stylus glaber, sursum ampliatus. Stems not radicant at the nodes, the internodes elongate, 5-7 cm. long, red, thick, 4 mm. in diameter, hirsute, the hairs spreading, red- dish, many-celled, the cross walls dark red, ca. 1.5 mm. long; leaves of a pair subequal, decussate, short-petiolate, the petiole 6-7 mm. long, densely yellow-hirsute, 1-1.5 mm. in diameter, the blades oblong to oblanceolate, 5X2 em. to 8.53.4 em., short-acuminate, broadly cuneate and slightly unequal at base, remotely denticulate, green on both sides, hirsute above with yellowish, several-celled hairs, whitish hirsute beneath, the hairs many-celled, ca. 3-4 mm. long, smal] “bayonet”-shaped hairs absent, the lateral veins 6 or 7 pairs, obscure above, slightly raised beneath; flowers solitary in the leaf-axils, long- pedunculate; bracts minute, subulate, ca. 2 mm. long, hirsute and also sparsely yellow-glandular; peduncles 1.7-2 cm. long, ca. 1.2 mm. in diameter, green, reddish-hirsute; calyx lobes free, very slightly un- equal, lanceolate, ca. 2.8 cm. long, ca. 2.5 mm. wide at base, 6.5-7 mm. wide above base, gradually long-sharp-acuminate, entire, herbaceous, green toward base, the upper third conspicuously red, sparsely red- septate-hirsute without, lacking appressed “bayonet’’-shaped hairs, thin-white-sericeous within and also glandular; corolla red, ca. 3.8 em, long, 4 mm. wide at base, 7 mm. wide in the saccate portion above base, then contracted to 4.5 mm., the tube without visible stripes due to the dense pubescence, cylindric, ca. 7 mm. wide, densely hirsute with long, red, multicellular hairs covering the whole surface, glabrous and eglandular within even in throat, the limb small, only slightly irregular, the lobes fleshy, all free, erect, dark red with narrow lighter margins, the two posterior the largest, 4.5-5 mm. long, thick, with incurved margins, densely septate-hirsute externally, glabrous within, the lateral and anterior smaller, ca. 3.5 mm. long, obtusish with in- curved apex, the appendages conspicuous, 4~6.5 mm. long, very slender, dark red, conspicuously red-hirsute; stamen sheath ca. 3 mm. long, the filaments very slender, glabrous; anthers included; ovary white-sericeous at apex; style glabrous, strongly thickened upwardly; disk composed of 2 small (1 mm. long), free, posterior glands. Type in the herbarium of the Royal Botanic Garden, Kew, collected at Altaquer, Department of Narifio, Colombia, May 1876, by E. André (no. 3311). REVISION OF TRICHANTHA—MORTON 7 2. Trichantha rosea Morton, sp. nov. Caules radicantes, internodiis brevibus gracilibus hirsutis; folia per paria inaequalia, minora stipuliformia; folia majora breviter petiolata, lamina elliptica, longe acuminata, basi obliqua, inconspicue dentic- ulata, supra hirsuta, subtus rosea, hirsuta et strigosa; flores solitari, bini vel terni; calyx roseus, lobis dissectis, lobis lateralibus 2—4-jugis, linearibus, externe roseo-hirsutus, intus albo-sericeus; corolla pulchella, tubo cylindrico externe dense et perspicue ubique roseo-hirsuto, superficie non visibili, intus in fauce glabro et eglanduloso, limbo regulari, lobis crassis atrorubris parvis erectis, appendicibus valde valde elongatis, 1.5-2 cm. longis, densissime roseo-hirsutis; stylus glaber. Stems radicant at the nodes, the internodes short, 2-4 cm. long, somewhat zigzag, 3-4 mm. in diameter, shining, somewhat ridged in drying, hirsute, the hairs reddish, spreading, many-celled, borne on slight tubercles; leaves of a pair strongly unequal; larger leaves short- petiolate, the petiole 5-10 mm. long, hirsute, 1 mm. in diameter, the blades narrowly elliptic, 72.4 cm. to 93.4 cm., herbaceous, gradu- ally long-acuminate at apex, rounded and strongly unequal at base, inconspicuously denticulate, green and sparsely hirsute above, the hairs 5-7 celled, the basal cells paler yellowish and enlarged, the others pale reddish with dark red cross-walls, beneath rose-red all over, hirsute with spreading multicellular reddish hairs and also sparsely strigose with closely appressed, white 2-celled, “bayonet’’-shaped hairs about 0.4 mm. long, the lateral veins about 6 pairs, obscure above, elevated beneath; smaller leaves of a pair stipule-like, not over 2.2 cm. long and 1 cm. wide; flowers 1-3 in an axil; bracts minute, linear, scarcely 2 mm. long, hirsute, eglandular; peduncles 7-9 mm. long, hirsute; calyx rose-red, the lobes free, subequal, 13-15 mm. long, the central portion lanceolate-attenuate, ca, 2mm. wide at base, ca. 3 mm. wide above base, the apical half filiform and scarcely 0.5 mm. wide, dissected, with 2—4 pairs of linear lateral lobes, these up to 5 mm. long and 0.25 mm. wide, conspicuously red hirsute without and on lobes, densely white-sericeous within and also with sessile yellow glands; corolla brilliant rose-red, without visible stripes due to the dense pubescence, ca. 4 cm. long, 4 mm. wide at base, 6.5 mm. wide at the saccate portion above base, above base 3 mm. wide gradually increas- ing to 6 mm. wide at middle, scarcely if at all contracted in the throat, conspicuously and densely hirsute externally, the hairs red, long, multicellular, completely covering the corolla surface, glabrous within, the limb scarcely bilabiate, narrow, not wider than the tube, the lobes fleshy, dark red within with very narrow yellowish thinner margins, erect, subequal, all free, subtriangular, ca. 4 mm. long and 3 mm. wide at base, acutish, red-hirsute at base externally, white-hirsute at apex, 680-018—63——2 8 CONTRIBUTIONS OF THE NATIONAL HERBARIUM glabrous and eglandular within, even in throat, the appendages red, extremely conspicuous, 1.5-2 em. long, very densely hirsute with long red, multicellular hairs; filaments united for 6 mm. at base, extremely slender, ca. 0.25 mm. thick, glabrous; anthers not exserted, connate quadrately at first, broader than long, ca. 1.5 mm. long, 1.7-2 mm. wide, glabrous; ovary small, glabrous at base, sparsely white-pilose at apex; style stoutish, ca. 3.5 em. long, glabrous; disk reduced to a small, bilobed, thick, glabrous, posterior gland. Type in the U.S. National Herbarium (no. 2370204), collected at Rio Catripe, Intendencia del Chocd, Colombia, at 100 m. elevation, in wet forest, Jan. 17, 1947, by Oscar Haueht (no. 5437). Paratype: Colombia: Between La Oreja and Quibdd, Choeéd, Apr. 1, 2, 1931, W. A. Archer 1677 (US), Additional specimen examined: Quibdé, January, 1872, R. B. White (K). A note by White reads: “Of the enclosed fragment I can only say that it is an herbaceous plant 9 to 12 inches high, alternate lanceolate velvety leaves. Seed? Habitat, Quibdé woods of Atrato Valley. Most lovely colour. It would seem as if all these hairy plants were placed in wet regions to intake the most of the moisture. Nothing can be more lovely than these bedewed flowers. Illness prevented me from attending to the drying of my specimens collected in some parts of my Quibdé journey and thus many were spoilt and lost.” Certainly referable to 7. rosea, but leaves are absent and so the forma is uncertain. 2a. Trichantha rosea forma viridis Morton, f. nov. A f. typica foliis subtus viridibus costa basi excepta differt, foliis forsan majoribus usque ad 176.5 cm. longis. Type in the U.S. National Herbarium (no. 1771980), collected on the ridge along Yeractii Valley, upper Rio San J uan, Corcovado region, Intendencia del Chocé, Colombia, at 200-275 m. elevation, Apr. 24, 25, 1939, by E. P. Killip (no. 35239), Paratype: Colombia: Cérdoba, Dept. El Valle, 50-100 m., Feb. 17, 1939, Killip & Garcia 33418 (US). 2b. Trichantha rosea forma latifolia Morton, f. nov. A f. typica foliis brevioribus et latioribus, maximis 6X3 em., subtus viridibus vel vix roseo-tinctis differt. Type in the herbarium of the Royal Botanic Garden, Kew, collected in the Department of Antioquia, Colombia, 7700-8800 ft. elevation, June 16, 1880, by W. Kalbreyer (no. 1793). Paratype: Colombia: Cultivated by Veitch, from seeds sent by Kalbreyer, probably collected at the same time as the type cited above, with which this specimen is mounted (IX). REVISION OF TRICHANTHA—MORTON 9 The two leaf variations described above are given only the rank of “forma,” since it is uncertain just how taxonomically significant they are. The brilliant rose-colored leaves of f. rosea certainly contrast strongly with the green leaves of f. viridis, but these two grow in the same geographic and climatic area, the Pacific lowlands (50-275 m.). The small-leaved f. latifolia may be more distinct since it comes from the highlands (ca. 2400 m.) of a different department. This species is one of the most distinctive by its regular (not bilabiate), densely rose-tomentose corollas, with extraordinarily long and conspicuous appendages. 3. Trichantha formosa Morton, sp. nov, Caules graciles, internodiis brevibus roseo-hirsutis; folia non valde inaequalia, petiolata, lamina lineari-oblonga, acuminata, basi cuneata, remote denticulata, utrinque hirsuta, subtus viridi, pilis minutis appressis nullis; flores 1-3 axillares, longe pedunculati, pedunculo roseo-hirsuto; calycis lobi subaequali spathulati integri virides vel rubri, basi angustati, apice acuti, utrinque hirsuti; corolla tubulosa, tubo luteo, perspicue atrorubro-lineato, externe glanduloso-hirsuto et minute pilosulo, limbo parvo bilabiato, lobis rubro-marginatis, intus glabris. Stems radicant at the nodes, the internodes normally ca. 3 cm. long, 2.5-3 mm. in diameter, densely rosy-hirsute, the hairs many-celled, with reddish lateral walls and dark septae, 2-2.5 mm. long, widely spreading; leaves of a pair slightly but not conspicuously unequal, the smaller more than half as long as the larger, the upper short- petiolate (ca. 7 mm.), the lower longer petiolate (18 mm.), the petiole densely red-hirsute, slender, 0.5-0.8 mm, thick, the leaf-blades linear- oblong, 6X1.2 cm. to 101.6 cm., broadest near middle, acuminate, equally cuneate at base, herbaceous, remotely and inconspicuously denticulate, green above, hirsute, the hairs several-celled, pale, ca. 1.5 mm. long, green beneath, the midrib red, densely pale-hirsute or the hairs along the midrib reddish, minute, white, ‘“‘bayonet’’-shaped hairs lacking, the lateral veins ca. 6 pairs, obscure above, slightly elevated beneath; flowers 1 to 3 in an axil, long-pedunculate; peduncles up to 4.2 cm. long, densely rosy-hirsute; calyx lobes green or red, subequal, 9-11 mm. long, spathulate, narrowed at base and ca. 1.5 mm. wide, broadened upwardly, 2.5-3.3 mm. wide, acute at apex, entire, densely hirsute externally, the hairs pale or reddish, several- celled, 1.5-2 mm. long, and also sparingly strigillose with appressed, minute, white, 2-celled, ‘“bayonet’-shaped hairs, densely hirsute within and also with a few, sessile, yellow glands toward the base; corolla ca. 4.3 cm. long, ca. 2 mm. wide at base, ca. 4 mm. wide in saccate portion above base, the tube cylindric, becoming 6 mm. wide 10 CONTRIBUTIONS OF THE NATIONAL HERBARIUM above the middle, very slightly contracted in throat, dull yellow with conspicuous maroon stripes running from near the base to the sinuses, sparsely hirsute externally, the hairs spreading, ca. 2 mm. long, many-celled, the terminal cell glandular, and also with numerous spreading unicellular eglandular hairs 0.25 mm. long or less, glabrous within, except for the obviously glandular-pilosulous throat at the base of the galea, the limb obviously bilabiate, the upper 2 lobes united into a slightly bilobed galea 4 mm. long, this almost entirely maroon, the 2 lateral lobes broadly deltoid, ca. 3 mm. long and wide, broadly margined with maroon, the lower lobe narrower and rela- tively longer, all lobes sparsely hirsute without, glabrous within, the appendages inconspicuous, 1-2 mm. long, yellow throughout or dark maroon at base and yellow at apex, long-hirsute at apex; filaments united in a sheath ca. 2 mm. long at base, very slender, glabrous; anthers included, lightly connate but soon free, ca. 2.5 mm. long and broad, the cells somewhat separated by the thickened connective; ovary strongly white-villous; style slender, densely hirsutulous up- wardly; stigma bilobed; disk reduced to a minute, dark, scarcely emar- ginate posterior gland. Type in the U.S. National Herbarium (no. 1518165), collected at Umbria, Comisaria del Putumayo, Colombia, at 325 m. elevation, January-February 1931, by G. Klug (no. 2114). Paratype: Colombia: Umbria, October-November 1930, Klug 1822 (US). This species, the only one known from the Amazonian watershed, nay be related to 7. allepida, which has a corolla quite similar in shape and color, but lacks the numerous, small unicellular hairs and gland- tipped longer ones of 7. formosa. The spathulate, entire calyx lobes are unique in this section. 4. Trichantha moorei (Morton) Morton, comb. nov. Columnea moorei Morton, Baileya 7:55, fig. 15. 1959. Stems succulent, scandent, at least 30 em. long and probably much longer, unbranched (except probably at base), probably not radicant at the nodes, the internodes very short, about 1 cm. long, fleshy, glabrescent, minutely strigillose with 2-celled “bayonet’’-shaped hairs and also with a few multicellular hairs toward the apex, bearing 2 pairs of conspicuous glands (these often coalescent in pairs) at each node between the leaves and just below the stipular line, the inter- nodes short and regular, 10-12 mm. long; leaves thick-fleshy, dark green and shining above, light green beneath, obviously decussate, those of a pair equal, short-petiolate (the petiole 3 mm. long, 1 mm. thick, glabrate), the blades broadly elliptic,” very uniform in size, 14-16 mm. long, 10-12 mm. wide, obtuse at base and apex, almost REVISION OF TRICHANTHA—MORTON 11 entire but with one or two low broad inconspicuous crenations on each side, almost glabrous, bearing a few minute appressed hairs beneath especially on the midrib, the margins obviously ciliolate with several- celled hairs; flowers solitary in an axil, bibracteate, the bracts minute, linear, 1-2 mm. long, deciduous, pilosulous; peduncles arching, curved at apex, 3-4.5 em. long, slender, 1 mm. thick at base, becoming en- larged and 2-3 mm. thick at apex, red, conspicuously long-setose- pilose, the hairs red, 2-3 mm. long, many-celled, spreading at right angles; calyx green, 5-patted, the lobes erect, equal, 12-15 mm. long, 8— 11 mm. wide including the teeth, conspicuously and deeply pectinately parted, the central portion of the lobe lanceolate, about 3 mm. wide, the teeth 4-6 pairs, linear, spreading horizontally, the basal ones about 3-4 mm. long, the uppermost about 2 mm. long, all 0.8-1 mm. wide just above the base, conspicuously red-gland-tipped, the body and teeth externally conspicuously long-red-hirsute, the hairs 2-4 mm. long, many-celled, sharp-pointed, and also with a few, appressed, white, 2-celled, ‘‘bayonet’’-shaped hairs, internally nearly glabrous but with a few stiff red hairs and also slightly glandular; corolla suberect in calyx, red (except the lobes), 5-5.5 cm. long, gibbous at the posterior base, the gibbosity 3 mm. wide, the tube 3-4 mm. wide just above the gibbosity, gradually enlarged upwardly but only slightly ventricose, 10-11 mm. wide near the apex, sparsely but conspicuously hirsute externally, the hairs 4-6 mm. long, red, many-celled, sharp-pointed, horizontally spreading, and also with minute spreading hairs, glabrous within except toward the throat, where conspicuously glandular- pilosulous, the throat only slightly contracted, the limb slightly bilabiate, the two upper lobes red, with narrow yellow margins, erect, partly connate, rounded, 4 mm. long, connate for about 2 mm., sparsely ciliolate, glabrous within at apex but strongly capitate- glandular lower down, the three lateral lobes clear yellow, not red- margined, slightly plicate at the angles between the lobes, erect, subequal, subdeltoid, about 5-6 mm. wide at the base and 5 mm. long, sparsely hirsute externally, more or less ciliate, glabrous within and not capitate-glandular, the appendages yellow, extremely minute and hardly observable in dried specimens; stamens attached to the corolla at the very base, the filaments pale yellowish-white, connate at the base for 4-5 mm., free upwardly, somewhat curved but not contorted, glabrous, the anthers slightly exserted, all four permanently connate, subquadrate, about 1 mm. long and wide, the cells oblong, fully dehiscent longitudinally, glabrous; staminodium none; ovary oblong in outline, 4-4.5 mm. long, densely white-sericeous; style white, straight, 5-5.5 cm. long, exserted, pilosulous; stigma bilobed; disk reduced to a solitary posterior gland, this white, thick, fleshy, not bilobed, ca. 1-1.5 mm. wide, 1.5-2 mm. long. 12 CONTRIBUTIONS OF THE NATIONAL HERBARIUM Type in the U.S. National Herbarium (no. 2216847), from material cultivated in the greenhouses of Bailey Hortorium, Ithaca, N.Y., collected by H. E. Moore, Jr., in 1958, originally received from an unspecified locality in Panama through Mr. Paul Arnold of the House of Plants, Binghamton, N.Y. This plant had been sent to Mr. Arnold by Mrs. Catherine B. Shepherd of Arlington, Va., who in turn received it from Mr. Martin S. Sawyer, of Gatun, Canal Zone. Another specimen has been received directly from Mr. Arnold. 5. Trichantha illepida (Moore) Morton, comb. nov. Columnea illepida Moore, Baileya 8:56, fig. 19. 1960. Stems stout, probably not radicant at the nodes, the internodes short, 1-2.5 cm. long, very stout, the upper ones ca. 3 mm. in diameter, somewhat zigzag, strongly ridged when dry, hirsute, the hairs spread- ing, yellow, multicellular, eglandular, 2-3 mm. long, borne on tubercles; leaves subdistichous, not bullate, those of a pair strongly unequal; larger leaves short-petiolate, the petiole 0.3-2 em. long, hirsute, the blades variable in shape from ovate-lanceolate to oblanceolate, up to 13 em. by 5 em., acuminate or subcuspidate at apex, broadly cuneate and strongly unequal at base, minutely and remotely denticulate, green above, hirsute, the hairs colorless, several- celled, 1.5-2 mm. long, beneath green with conspicuous red blotches or red all over, septate-hirsutulous all over, with also a few, appressed, 2-celled, “‘bayonet’’-shaped hairs, the lateral veins 5 or 6 (to 9?) pairs, slightly elevated on both sides; smaller leaves of a pair early deciduous, like the larger but subsessile, not over 2.5 cm. long; flowers several in an axil, slender pedicellate; bracts minute, ca. 40.75 mm., hirsute and with sessile yellow glands; peduncles red, 1 mm. thick or less, hirsute; calyx green, herbaceous, 1.5-2 em. long, the lobes free, slightly unequal, the posterior shorter and narrower, the central portion 2.5-3.5 mm. wide, strongly pectinate-toothed, the teeth 4 or 5 (6) pairs on each side, linear, the larger 2 mm. long and 0.5 mm. wide, hirsutulous on both sides with colorless hairs, some long and many- celled, some short and 2-celled, both surfaces also with sessile yellow globular glands; corolla ca. 5 em. long, slightly oblique in the calyx, slightly spurred at the posterior base, the tube dull, clear yellow, conspicuously striped with maroon from just below the middle to the base of the sinuses between the lobes, the stripes 0.5-1 mm. wide, the tube 7.5-10 mm. in diameter near the middle, slightly contracted in the throat, externally sparsely hirsute and provided also with some small spreading unicellular hairs, glabrous within except for the glandular-pilosulous throat, the limb bilabiate, ca. 15 mm. wide, the galea (posterior lip) 5.5-6 mm. high, bilobed, conspicuously spotted with maroon, the two lateral lobes deltoid, ca. 5 mm. long, REVISION OF TRICHANTHA—MORTON 13 5-6 mm. wide at base, margined with maroon, the anterior lobe ca. 4.5 mm. long and 4 mm. wide at base, margined with maroon, all the lobes hirsute externally and with short, white, thick-based unicellular hairs also, the appendages yellow, inconspicuous when dry and not over 1 mm. long; stamens included; anthers quadrately connate; ovary green, pilose; style puberulous; stigma bilobed; disk reduced to a whitish, bilobed posterior gland. Type: A cultivated specimen in the Bailey Hortorium, of unknown provenience. Additional specimen examined: Cultivated: Fantastic Gardens, South Miami, Fla., Feb. 25, 1954, R. G. Wilson. The specimen from Fantastic Gardens was determined as a new species by me in 1954. It was guessed to be either from Tingo Maria, Peru, or from the Panama Canal Zone. Later inquiries for more and better material brought the news that the plant had been lost. The plant that later turned up in the Bailey Hortorium without accession data was supposed to be from either Tingo Maria, Peru, or Ecuador. It is almost certainly from seeds or a plant sent from Fantastic Gar- dens to the Bailey Hortorium sometime around 1954, rather than a new introduction, as shown by the mention of Tingo Maria, Peru, in both cases. The determination of the real locality will have to await a collection of the species in the wild. The locality Tingo Maria is not likely, because no other species of Trichantha is known from Peru and because the climatic conditions in Tingo Maria are different from those where the other species grow. The Panama Canal Zone does not seem likely either, because this area has been well collected and it does not seem likely that this conspicuous plant would have been overlooked by others. The locality Ecuador is much more likely, but there seems no evidence in its favor. This species does not show any close relationship with the others. It differs especially in its broader, toothed rather than dissected calyx lobes from T. elegans, T. teuscheri, and others. 6. Trichantha tropicalis Morton, sp. nov. Caules elongati, radicantes, internodiis elongatis gracilibus hirsutis; folia per paria inaequalia, minora stipuliformia, majora magna, elliptica acriter acuminata basi rotundata remote denticulata, supra hirsuta, subtus viridia hirsuta et appresso-strigillosa ; pedunculi pen- duli apice valde incrassati; calyx ruber dissectus, lobis lateralibus 2-vel 3-paribus linearibus rubro-hirsutis; corolla basi gibbosa, tubo rubro non lineato, apice superiore excepto, hirsuto, limbo bilabiato, galea atrorubra pallide marginata, lobis lateralibus flavis rubro-marginatis, lobo inferiore flavo basi rubro-maculato. 14 CONTRIBUTIONS OF THE NATIONAL HERBARIUM Stems ca. 40 cm. long, radicant at the nodes, the internodes 4-5 em. long, reddish, rugose, slender, only 2.5 mm. in diameter near base, coarsely hirsute, the hairs persistent, borne on small tubercles, 7-9-celled, red, 4-5 mm. long; leaves of a pair strongly unequal; larger leaves short-petiolate, the petiole 8-10 mm. long, stout, 1-1.5 mm. in diameter, hirsute, the blades elliptic, 104.8 em. to 14 <5.2fem., the apex sharply and rather abruptly long-acuminate, rounded and slightly unequal at base, herbaceous in texture, incon- spicuously and remotely denticulate, uniformly hirsute above, the hairs ca. 5 mm. long, many-celled, hyaline, beneath paler green, similarly hirsute with elongate, spreading, dark, many-celled hairs with dark cross-partitions, and also with small (0.6 mm. long), appressed, 2-celled hairs, the basal cell small, yellow, the terminal cell elongate, white, sharp; smaller leaves of a pair early deciduous, short-petiolate, the blades lanceolate, not over ca. 1 cm. long and 8 mm. wide, hirsute like the larger leaves; flowers solitary or geminate; bracts deciduous; peduncles pendulous, 2.5-3.5 em. long, ca. 1 mm, thick at base, strongly enlarged and 3 mm. thick at apex, densely red-hirsute; calyx red, the lobes free, slightly unequal, the 4 dorsal and lateral about 16 mm. long in flower, the posterior one shorter, ca. 14 mm. long, all deeply dissected, the central portion ca. 3 mm. wide at base, 0.3 mm. wide upwardly, with 2 or 3 pairs of linear lateral lobes, these 8-10 mm. long, 0.8 mm. wide, the central portion red-hirsute without and also with some inconspicuous, “bayonet”’- shaped, white, appressed hairs, slightly strigillose within, hirsute at apex, bearing sessile yellow glands toward base, the lobes long- ciliate with red, several-celled spreading hairs; corolla ca. 5.5 em. long, oblique in calyx, ca. 6 mm. wide at the conspicuously saccate base, the tube narrowed above base to 3 mm., the tube clear red nearly throughout, not striped except for 2 inconspicuous dark stripes near the apex, these running to the margins of the galea, becoming 7-8 mm. wide at the middle, sparingly reddish-septate-pilose ex- ternally, lacking small whitish unicellular hairs, glabrous within except in the glandular-pilosulous throat, slightly contracted in throat and 5.5-6 mm. wide, the limb bilabiate, the galea ca. 5 mm. long and 7 mm. wide, rounded and bilobed at apex, dark red with a paler border, the lateral lobes broadly deltoid, ca. 4 mm. long and 4 mm. wide at base, yellow in middle, with dark red margins, the posterior lobe 3.5 mm. long and 2.5 min. wide, yellow with a dark red basal spot, the appendages conspicuous, yellow, 3-3.5 mm. long, 0.5 mm. thick, clavate, glabrous except the conspicuously red- pilose apex; filaments united at base into a sheath 1 cm. long, gla- brous, extremely slender; anthers not exserted, borne near corolla throat, united into a square about 4 mm. long and wide; ovary REVISION OF TRICHANTHA—MORTON 15 conic, sparsely white-sericeous; style slender, slightly pilosulous; disk reduced to a coarse, slightly bilobed, posterior gland. Type in the U.S. National Herbarium (no. 2370203), collected in the San Juan de Micay Valley, Department of El Cauca, Colombia, at 100 m. elevation, Dec. 28, 1946, by Oscar Haught (no. 5406). By its clear red, unlined corollas, and red, deeply dissected calyces related only to 7. clara, which differs in the size, shape, and pubes- cence of the leaves. 7. Trichantha major Hook. Icon. Plant. 7: 1.667. 1844. Columnea major Hanst. Linnaea 36:388. 1865-6. Stems 15-20 em. back from the apex 3-4 mm. thick, the internodes elongate (9-12 cm.) or shorter (4.5 cm.), obviously hirsute with long (5 mm. long or more), many-celled pale-brown hairs, these numerous but still distant and not obscuring the surface; stems radicant, the adventitious roots thread-like, borne near the nodes; leaves of a pair strongly unequal; longer leaves petiolate, the petioles 6-13 mm. long, hirsute, the blades broadly elliptic, 6.5 4.5 cm. to 14x7cm., abruptly cuspidate at apex (ca. 1 em.) rounded and strongly unequal at base, inconspicuously and remotely toothed upwardly, thin-herbaceous the upper surface not bullate, at maturity obviously foveolate, slightly paler beneath, strongly hirsute on both sides, the hairs 4.5 mm. long, many-celled, distant and not obscuring the leaf surface, “bayonet’’- shaped hairs absent, the principal veins 4 to 6 pairs, strongly curved, obscure above, apparent and slightly raised beneath, the veinlets obscure; smaller leaves short-petiolate, the blades subrotund, about as wide as long, slightly cuspidate, mostly 1.5-2.5 cm. long, hirsute; flowers axillary, geminate (always?); bracts subulate, 5 mm. long, 0.5 mm. wide, hirsute, eglandular; peduncles 3-3.5 cm. long, apically thickened, long-hirsute, the hairs many-celled, with dark reddish cross-walls; calyx red, 2.5-3 cm. long, the lobes free, subequal, deeply laciniate, with the lobes ca. 3 em. wide but the central lanceolate portion only 2-2.5 mm. wide, the lateral lobes 2 pairs, subequal, mostly ca. 1.5 cm. long, ca. 1 mm. wide at base, long-attenuate, the central lobe ca. 2 cm. long, strongly long-hirsute externally, and also strigillose with “bayonet’’-shaped hairs, within hirsute toward apex and with sessile glands toward base; corolla erect in calyx, ca. 4.5 cm. long, yellow with conspicuous dark red or maroon vertical stripes running to the sinuses of the limb, tubular, scarcely ventricose, the tube becoming ca. 5.5 mm. wide about the middle, slightly long- hirsute externally, with also a few short, colorless, 1-celled, spreading hairs, hardly contracted in the throat, glandular-pilosulous within throat, the limb just slightly irregular, the lobes hirsute externally and with sessile glands, all dark colored, subrotund, 3-4 mm. long, entire, 16 CONTRIBUTIONS OF THE NATIONAL HERBARIUM not ciliate, the appendages yellow with dark apices, clavate, spreading, ca. 4 mm. long, long-hirsute at apex only; anthers barely exserted from the corolla tube, persistently connate, broadly oblong, ca. 1.5 mm. long, 0.9 mm. wide; ovary white-sericeous; style pilosulous; disk reduced to a bidentate, fleshy, glabrous posterior gland ca, 2mm. high. Type in the herbarium of the Royal Botanic Gardens, Kew, collected in Colombia by William Lobb (no. 117). No definite locality is given, but since the species has been twice collected in the region of Barbacoas and since Lobb is known to have collected there, it is likely that the type came from there. Additional specimens examined: Colombia: Province of Barbacoas, Dept. of Narifio, 800 m. alt., J. Triana (US), Corregimiento Santander (Buenavista) to Barbacoas, vertiente del Rio Telembi, Narifio, 840-200 m., Aug. 3-5, 1948, H. Garcia-Barriga 13176 (HNC), 8. Trichantha bullata Morton, sp. nov. Caules elongati, internodiis clongatis; folia per paria inaequalia, minora lanceolata, majora elliptica, remote denticulata, supra bullata, valde hirsuta, subtus rubescentia, hirsuta et strigosa, basi inaequalia, apice acuminata; calycis lobi rubri, externe perspicue hirsuti, intus apice hirsuti, deorsum parce strigillosi, lobis valde pinnatisectis, lobulis lateralibus 4 vel 5 utrinque latere; corolla in tubo flava et atrobrunneo-rubro-lineata, lineis faucem non attingentibus, lobis paullo inaequalibus, flavis, rubro-marginatis, appendicibus validis, hirsutis. Stems elongate, up to 2 m. long (see. Lehmann), the internodes elongate, normally 7-12 em. long, very slender, 2-3 mm. in diameter, dark red, shining, radicant at the nodes, usually somewhat zigzag, hollow, densely yellow-hirsute, the hairs 3-4 mm. long, spreading, coarse, borne on prominent tubercles, multiseptate (7—-9-celled), the septae dark; leaves of a pair strongly unequal, the smaller ones often deciduous, thus making the leaves appear alternate; internodes elon- gate, those below the apex 7-9 cm. long; larger leaf-blades elliptic, 15xX6 em. to 6X2.3 cm. (uppermost), gradually and sharply acuminate, remotely denticulate, the base rounded, slightly unequal, above green, not foveolate, slightly bullate, strongly hirsute, the hairs borne on prominences, elongate, 3-4 mim. long, spreading, yellow with dark cross-walls, 7—9-celled, beneath green or usually reddish or deep dark red, not bullate or foveolate, strongly hirsute as upper surface, with some short ‘bayonet-hairs” intermixed, short- petiolate, the petiole red, 7-14 mm. long, thick, hirsute; smaller leaf of a pair narrowly lanceolate, not more than 3.5X1 cm., long- attenuate, the base cuneate, hirsute; flowers solitary or 2 to 4 in an axil; bracts subulate, minute, scarcely 5 mm. long, hirsute, deciduous; REVISION OF TRICHANTHA—MORTON 17 peduncles 2.5-3.5 cm. long, exceptionally thick, long-red-hirsute; calyx 5-lobed to base, the lobes approximately equal, ca. 2.5 cm. long, deeply pinnately parted, the central portion linear-lanceolate, ca. 3 mm. wide at base, tapering gradually to the acute apex, the lateral lobes 4 or 5 on each side, linear, the lower about 13 mm. long, about 1 mm. wide at the base, gradually tapering to the acuminate apex, conspicuously hirsute without, the hairs very long (4-5 mm.). they (as well as the calyx lobes themselves) Bordeaux red when fresh (sec. Killip), violet-carmine when dry, simple, pluricellular, the cell wall obvious even to the naked eye, very numerous on the mar- gins, and also strigose with “‘bayonet’’-shaped hairs, within hirsute at apex, sparsely strigillose below; corolla ca. 6 cm. long, narrowly tubular, conspicuously saccate at the dorsal base, about 7 mm. wide at the base, 4.5 mm. wide above the saccate portion, gradually ex- panding clavately toward the apex and slightly curved toward the ventral side, reaching 10 mm. in diameter, slightly contracted at the throat (9 mm. wide), the total length of the corolla tube approxi- mately 5.5 cm., the limb scarcely expanded, glandular-pilosulous within throat, not much wider than the tube, the three ventral lobes ovate or almost semicircular, obtuse, 4 mm. wide at base, 2.6 mm. long, the two dorsal united at base, bearing clavate or oblong append- ages in the sinuses, these yellow with maroon apices, reflexed, about 2 mm. long and 1 mm. wide (the lowest ventral one narrower and longer, about 3 mm. long, when dry), the dorsal one borne on the back of the two united dorsal lobes, the tube externally densely pubescent with two kinds of hairs, some very short, few-celled, colorless, sharp-pointed, numerous all over the surface although not prominent on the saccate base, the others similar to the red calyx- hairs, confined to the upper half of the tube, extremely long and conspicuous toward the apex and especially on the appendages, the tube ‘“citron-green” (sec. Killip) with broad stripes of burnt-lake, these often about 3 mm. wide or even more, stopping about 5 mm. below the throat except the broad one running to the middle of the galea, the lobes light cadmium-yellow margined with red, the append- ages conspicuous clavate, ca. 2 mm. long, yellow with maroon apex, long-hirsute and also with numerous, sessile, yellow glands; filaments longer than the corolla but twisted and curved so as not to be exserted, very slender, ca. 0.2 mm. wide, didynamous, broadening toward the base, finally uniting with each other into a tube (broken on the dorsal side) about 6 mm. above the base, this tube becoming adnate to the corolla tube at about 1 mm. above the base; anthers borne about 5 mm. below corolla throat, glabrous, quadrately united, the four forming a rectangle 5 mm. high and 4 mm. wide, the cells par- allel, not confluent, openiug from base to apex, the connective square, 18 CONTRIBUTIONS OF THE NATIONAL HERBARIUM fleshy, glabrous; ovary oblong, 5 mm. long, conspicuously villous at apex, the hairs colorless, pluricellular; style ca. 5 cm. long, slightly exserted, flattened and about 1.5 mm. wide at the almost glabrous base, becoming narrower above and densely glandular-puberulous; stigma bilobed, bearing a few long, white, villous hairs; disk reduced to one large, transversely oblong, glabrous, unlobed, dorsal gland 1 mm. long and 1.5 mm. wide. Type in the U.S. National Herbarium (no." 1142440), collected at “La Gallera,” Micay Valley, Department of El Cauca, Colombia, at 1800-1900 m. elevation, July 1, 1922, by E. P. Killip (no. 7928). Isotype in the New York Botanical Garden. Paratypes: Colombia: Near Rio San Joaquin, “La Gallera,” Micay Valley, Department of El Cauca, 1100-1300 m., Killip 7803 (NY). West Andes of Popayan, Department of El] Cauca, 1200-1600 m., Lehmann 6000 (K, US). La Costa, near El Tambo, Department of El Cauca, 900-1200 m., von Sneidern 467, 622, 704 (all S). This species, evidently confined to moist virgin forest of the Western Cordillera of the Department of El Cauca at middle elevations, is obviously related to 7. major, which is still surely different in its minutely foveolate upper leaf-surfaces, its cuspidate rather than acuminate leaf-blades, its rotund rather than lanceloate smaller leaves of a pair, and its longer and more slender lobes of the calyx lobes. 9. Trichantha minor Hook. Icon. Plant. 7: t. 666. 1844. (But not Bot. Mag. Curtis: t. 5428. Cfr. T. elegans.) Columnea minor Hanst. Linnaea 34:387. 1865-6. Stems slender, not radicant (?), the internodes 3-3.5 cm. long, 2.5-3 mm. thick 20 cm. back from the tips, densely hirsute, the hairs more numerous, closer, shorter (ca. 3 mm.) than in 7. major, mostly clearly antrorse rather than spreading; leaves of a pair unequal (?); larger leaves petiolate, the petiole 5—7 mm. long, antrorsely hirsute, the blades lanceolate, 3.51.2 to 5.52 cm., gradually acuminate at apex, broadly cuneate and oblique at base, inconspicuously and remotely denticulate, thick-herbaceous, glabrous above at maturity, except near the margins, these with numerous, conspicuous, 4—6-celled brownish ascending hairs 3 mm. long or more, dark-green above, paler beneath, reddish, inconspicuously strigillose with straight sharp 2-celled white hairs on the surface, antrorsely hirsute on the midrib and especially on the margins with straight, many-celled hairs, the veins about 6 pairs, inconspicuous above, slightly elevated beneath, the veinlets obscure on both sides; flowers axillary, solitary or gem- inate; bracts inconspicuous, linear, ca. 5 mm. long, hirsute and also with sessile glands; peduncles 1.5-2 em. long, thickened toward apex, REVISION OF TRICHANTHA—MORTON 19 strongly hirsute, the hairs spreading, slender, 3-4.5 mm. long, with red cross-walls; calyx red, 2-2.5 cm. long, deeply laciniate, the central undivided portion only 2-2.5 mm. wide, the lateral lobes 2 pairs, linear, 1.3-1.6 cm. long, 0.6-0.8 mm. wide, long-attenuate, hirsute on both sides and margins and also bearing some appressed minute sharp white 2-celled hairs; corolla erect in calyx, ca. 4 cm. long, tubular, hirsute externally with long many-celled hairs and also short-pilosulous with spreading, unicellular hairs, the tube somewhat ventricose, becoming 8 mm. wide at the middle, slightly contracted in the throat, probably glandular-pilosulous within at base of galea, the limb slightly bilabiate, the galea 5 mm. long, dark red, the other 3 lobes 3 mm. long, probably yellow, the appendages clavate, ca. 3.5 mm. long, ca. 0.8 mm. thick, long-hirsute at apex. Type in the herbarium of the Royal Botanic Garden, Kew, col- lected by William Lobb (no. 175). According to Killip (Smiths Mise. Coll. 87(1): 5. 1932) no. 175 was collected at ‘‘Barbacosa,” Colombia, which is undoubtedly a slip for Barbacoas, in the Depart- ment of Narifio, Colombia. Additional collections examined: Colombia: Alto de Armada, Department of Narifio, 1458 m. alt., May 22, 1876, H. André K 36 (K, NY, US). The ability to place this locality in Narifio is due to the kindness of Dr. L. B. Smith in making available his unpublished paper on André’s itinerary. The character of having the leaves entirely glabrous on the upper surface except at the margins easily separates 7. minor from other species with which it might be confused, e.g. 7. elegans or T. teuscherv. A confusion is introduced by the plant illustrated in Curtis’ Botanical Magazine plate 5428 as T. minor. A specimen authenti- cating the plate is preserved in the herbarium at Kew, mounted on the same sheet as the type of 7. minor; it was received from the Veitch Nursery in April 1864, just at the time the plant was illustrated. Although this plant is similar to 7. minor, it cannot really be the same, differing in the size and pubescence of the calyx, the coloration of the corolla limb and the pubescence of the upper leaf surface. This horticultural specimen, of unknown provenience, belongs to the species here described as TY. elegans. 10. Trichantha clara Morton, sp. nov. Caules radicantes, penduli, tenues, parce hirsuti; folia per paria inaequalia, minora stipuliformia, majora lanceolata vel oblongo- lanceolata, supra medium denticulata, supra parce hirsuta, subtus strigosa; calyces rubri, lobis alte laciniatis, lobulis lateralibus 2- paribus, externe rubro-hirsuti et etiam strigosi, intus apice parce hirsuti basi glabri parce glandulosi; corolla rubra, non lineata, limbo 20 CONTRIBUTIONS OF THE NATIONAL HERBARIUM paullo bilabiato, lobis 3 inferioribus rubris marginibus angustis flavescentibus, 2 superioribus basi maculam atrem externe gerentibus, appendicibus perspicuis rubris. Stems radicant at the nodes, the terminal branchlets pendulous, slender, sometimes flexuous, the internodes short (1.5 cm.) or usually longer (up to 6 cm.), ca. 2mm. thick 20 cm. below tip, reddish, sparsely hirsute, the hairs stiff, antrorse, ca. 4.5 mm. long, many-celled; leaves of a pair strongly unequal; larger leaves short-petiolate (ca. 5 mm.), the blades lanceolate to oblong-lanceolate, 4.51.5 cm. to 5.7X 2.7 cm., gradually and sharply acuminate at apex, oblique at base and broadly cuneate to subrounded, remotely but sharply toothed above the middle, the minute teeth recurved in drying and visible only from the lower side of the leaves, light green above, the surface not foveolate, sparsely long-hirsute all over, paler and sometimes reddish beneath, sparsely strigose, the hairs stiff, appressed, 2-celled, the basal cell short, dark-colored, the terminal cell white, bayonet- shaped, the veins 4 or 5 pairs, arcuate, obscure above, very slightly prominulous beneath, the veinlets obscure on both sides; smaller leaves stipule-like, 5-14 mm. long, ovate to linear-lanceolate, with pubescence above and beneath like the larger leaves; flowers gemi- nate, borne in the axils of the larger leaves; bracts 2, minute, hirsute, eglandular, deciduous, linear, ca. 50.5 mm.; peduncles slender, 2-2.5 cm. long, thickened toward apex, long-red-hirsute, the hairs slender, widely spreading, many-celled, 3.54.5 mm. long; calyx red, ca. 15 mm. long, the whole calyx with the lobes ca. 2 cm. wide, the 5 lobes free, deeply laciniate, the central portion ca. 1.5 mm. wide, the lateral lobes (2 pairs) subulate, 9-10 mm. long, 0.4-0.5 mm. wide, conspicuously long-red-hirsute, and also sparingly strigose like the lower surface of the leaves with white, bayonet-shaped hairs, glabrous within (except at the hirsute apex), and with a few, sessile glands at base; corolla all clear red, without stripes, erect in calyx, 4.5-5.5 cm. long, slightly saccate at posterior base, 3 mm. wide just above base, ventricose upwardly, becoming 8 mm. wide above middle, slightly contracted in throat, where 6 mm. wide, externally long- hirsute with many-celled, red, flaccid hairs with dark cross walls, also a few small, spreading sharp-pointed, 2-celled hairs, the limb slightly bilabiate, the two upper lobes partly connate, 2-2.5 mm. long, red with a dark spot at base externally, glandular-pilosulous at base, the three lower ca. 1.5 mm. long, clear red, with narrow, yellowish margins, the appendages conspicuous, clavate, red, 3-4 mm. long, 1 mm. thick at apex, apically long-hirsute; filaments connate at base for 5 mm. or more, glabrous; anthers perhaps exserted briefly, but normally borne about medially in the tube by curling of the filaments, connate at first but finally free, broader than long, ca. 1.5 mm. long REVISION OF TRICHANTHA—MORTON 21 and 2 mm. wide, the cells contiguous but separated by a fleshy con- nective; style minutely glandular-pilosulous; stigma 2-parted, pilosul- ous; ovary white-sericeous; disk reduced to a small posterior gland, this dark colored, inconspicuously 2-lobed, ca. 1 mm. long. Type in the U.S. National Herbarium (no. 1853860), collected at Piedra de Moler, Hoya del Rio Digua, Cordillera Occidental, Depart- ment of El Valle, Colombia, at 900-1180 m. elevation, Aug. 19-28, 1943, by J. Cuatrecasas (no. 14922). Paratypes: Colombia: La Elsa, Hoya del Rio Digua, Dept. of El Valle, 1000-1200 m. alt., Nov. 9, 1943, “enredadera herbacea; ramas purpureas; hoja hérbaceo crasa, verde en el haz, gris clara en el envés; pedtinculo cArdeno; caliz con pelos cirdenos; corola cirdena,” Cuatre- casas 15312 (US). Between La Oreja and Quibdé6, Intendencia del Chocé, Apr. 1, 2, 1931, Archer 1687, 1688 (US). La Costa, near El Tambo, Dept. of El Cauca, 1000-1400 m. alt., 1935-1936, von Sneidern 469, 497, 498 (S). A species of wider distribution than most, probably being not uncommon in its special, undisturbed habitats. 11. Trichantha teuscheri Morton, sp. nov. Planta epiphytica, caulibus elongatis vix ramosis, leviter rubro- hirsutis; folia per paria valde inaequalia, minora saepe stipuliformia, majora lamina lanceolato-elliptica, longe acuminata, basi vix obliqua, remote denticulata, supra hirsuta, subtus viridi vel rubro-tincta, septato-hirsuta, pilis unicellularibus nullis; flores penduli, longe pedunculati, pedunculo valde rubro-hirsuto; calycis lobi dissecti, parte centrali lineari-lanceolati externe rubro-hirsuti et parce strigillosi, intus basi glandulosi, lobis lateralibus 2-vel 3-paribus linearibus utrinque hirsutis; corolla tubulosa late rubro-lineata, vix flavo-lineata, lineis flavis angustissimis, externe hirsuta, intus in fauce glanduloso- pilosula, limbo bilabiato, lobis superioribus in galea rubra connatis, lobis lateralibus et inferioribus flavis, appendicibus flavis conspicuis apice hirsutis; stylus apice glanduloso-pilosulus. Stems elongate, 70 cm. long or more, the internodes elongate, 3-6.5 em. long, very slender, ca. 1.5 mm. in diameter, red, sparsely but conspicuously red-hirsute, the hairs spreading, 1-2 mm. long, several- celled, the lateral walls pale rosy, the cross-partitions dark red; leaves of a pair strongly unequal; larger leaves petiolate, the petiole 6-13 mm. long, slender, ca. 0.8 mm. in diameter, hirsute like the stems, the blades lance-elliptic, mostly ca. 5X2 em., the largest 8.53.7 cm. varying to 92.7 cm., the apex gradually or rather abruptly long- sharp-acuminate, the base broadly cuneate or slightly rounded, only slightly unequal at base, herbaceous in texture, the margins remotely and minutely denticulate, above dark green, not bullate or foveolate, 92 CONTRIBUTIONS OF THE NATIONAL HERBARIUM conspicuously hirsute, the hairs spreading, yellowish, 2-5 celled, 1-2 mm. long, beneath green with red veins or more or less suffused with red all over, deciduously hirsute, the hairs yellowish (or reddish along the veins), several-celled, 1-2 mm. long, and also with a few small, white appressed ‘‘bayonct’’-shaped hairs, the lateral veins 5 or 6 pairs, obscure above, slightly raised beneath; smaller leaves of a pair often minute and stipule-like, 5 mm. long or less or sometimes larger and up to 3 cm. long by 11 mm. wide, on petioles ca. 5 mm. long; flowers solitary or commonly 2 or 3 in an axil; bracts minute, subulate, 1.5-2 mm. long, hirsute, and also yellow-glandular; peduncles 2-3 cm. long, curved, red-hirsute, conspicuously thickened toward apex; calyx 5-lobed nearly to base, the lobes 16-18 mm. long, red, deeply dissected, the central portion linear-lanceolate, 2.5-3 mm. broad at base, attenuate and only 0.5 mm. wide at apex, conspicuously red- hirsute externally, the hairs spreading, many-celled, 2-3 mm. long, with red lateral walls and dark red cross-partitions, and also with a few appressed “bayonet’’-shaped hairs, not hairy within but at base densely glandular, the glands globular, yellow, sessile, the lateral lobes 2 or 3 pairs, linear, 4.5-6.5 mm. long, ca. 0.5 mm. wide, con- spicuously red-hirsute on both sides; corolla pendulous, 4-4.5 em. long, 4 mm. wide at base, 7 mm. wide in saccate portion above base, then narrowed to 3.5 mm., the tube slightly ventricose, becoming 7-8 mm. wide about the middle, gradually narrowed to 5 mm. in throat, dark red with very slender and inconspicuous yellow lines running to the base of the middle of the lobes, the yellow areas broader toward base of corolla tube, hirsute externally, the hairs spreading, rather sparse, ca. 2 mm. long, several-celled, colorless, with colorless cross-partitions, eglandular, small unicellular hairs present but few, glabrous within except for the densely glandular-pilosulous throat, the limb bilabiate, the 2 upper lobes united into an emarginate galea, this ca. 4 mm. long and 6-7 mm. wide, dark red, glandular-pilosulous at base, the lateral lobes all yellow, deltoid, ca. 3mm. long and 4 mm. wide at base, the lower lobe yellow elliptic, ca. 4 mm. long and 2.5 mm. wide, the appendages yellow, conspicuous, ca. 4 mm. long, clavate, ca. 1 mm. wide at apex, conspicuously hirsute at apex, the hairs colorless or slightly tinged with red; filaments united into a sheath ca. 7 mm. high at base, slender, glabrous; anthers lightly connate at first, soon free, 1.5 mm. long, 2 mm. wide, the connective prominent, dark; ovary conic, puberulous at base, lightly white- sericeous at apex; style ca. 3.8 em. long, sparsely pilosulous at base with unicellular hairs, at apex glandular-pilosulous, the hairs several- celled; stigma conspicuously bilobed; disk reduced to a bilobed thick fleshy glabrous posterior gland ca. 2 mm. high and wide; fruit a globular, white berry 1 cm. in diameter (sec. Teuscher). REVISION OF TRICHANTHA—MORTON 293 Type in the U.S. National Herbarium (no. 2370201), collected in the Montreal Botanical Garden, by M. Raymond and M. Gougeon (no. 2135-56), from material originally collected by Henry Teuscher in Ecuador in a wet moss-covered forest at an altitude of ca. 3400 ft. near the road from Cuenca to Quayaquil, on the Pacific slope of the Andes, about 20 to 25 road-miles beyond Ducur coming from Cuenca in the spring of 1956. Paratypes: All grown from cuttings of the type plant in the Mon- treal Botanical Garden: Bailey Hortorium, June 1959, Moore 7745 (US); University of California Botanical Garden, July 27, 1959, Hutchison (U.C. 58. 263). This plant has been identified erroneously with Trichantha minor, under which name it is well illustrated in the Delectus Seminum et Sporarum, no. 22, 1957-58, of the Montreal Botanical Garden (frontispiece). An article concerning its discovery and character- istics was published by Henry Teuscher in the National Horticultural Magazine (vol. 37, no. 3, pp. 172-174. July 1958), along with a full- page illustration of the plant in fruit. Another mention is by Paul Arnold in an article entitled ‘American Gesneriad Reintroduced from Canada” (Gesneriad Journal, vol. 7, no. 2, p. 3. 1959). Mr. Raymond submitted a specimen to me in 1957 of Mr. ‘Teuscher’s plant, asking about the determination as Trichantha minor, but I was unable to verify it, in the absence of the type or any authentically named material or even of any detailed and adequate description. I now have the holotype of 7. minor before me, kindly sent on loan from Kew, and it shows a related but obviously different species. The two can be distinguished as follows: Leaf-blades glabrous above except at the margins, beneath sparingly strigillose on the surface with appressed minute (0.5-0.6 mm. long), 2-celled, white sharp hairs, the midrib appressed-pilose with long, many-celled hairs; calyx large, more than 2 cm. long, the lobes externally with some minute appressed white hairs as well as long, many-celled hairs . .......-+--+:+5 T. minor Leaf-blades hirsute all over above, hirsute beneath, minute, white, 2-celled hairs few or absent; calyx 1.6-1.8 cm. long, or sometimes smaller, hirsute, lacking minute, white hairs . 2. 1. 1 1 ee ee ee ee es T. teuscheri 12. Trichantha elegans Rose, sp. nov. Caules graciles, rubescentes, internodiis brevibus dense rubro- hirsutis, pilis tenuibus; folia per paria inaequalia, minora saepe stipuliformia, majora elliptica vel ovato-elliptica, acuminata, basi rotundata et obliqua, supra hirsuta, subtus parce strigillosa; calyces mediocres, roseo-rubri, lobis valde laciniatis, laciniis lateralibus 2- paribus, linearibus, externe rubro-hirsutis; corolla in tubo late pur- pureo-rubro-lineata, anguste flavo-lineata, faucem versus flava, limbo 24 CONTRIBUTIONS OF THE NATIONAL HERBARIUM flavo, non lineato, viridi-marginato, appendicibus magnis_flavis clavatis hirsutis. Small succulent epiphytic herb, the stems radicant at the nodes, little branched, the internodes short (1.5 em.) or moderately elongate (to 4 cm.), somewhat zigzag, 2 mm. thick or less 10 em. back of the apex, densely hirsute, the hairs 3-4 mm. long, red, flaccid, many-celled, and also strigose with appressed, bayonet-shaped 2-celled hairs, the basal cell small, red, the apical white, sharp-pointed; leaves of a pair strongly unequal; larger leaves short-petiolate, the petiole 3-4 mm. long, densely hirsute, the blades eltiptic, or ovate-elliptic, 2.81.2 to 8X3.5 cm., gradually and sharply acuminate, rounded and oblique at base, inconspicuously and remotely toothed, light green above, the surface not foveolate, long-hirsute, the hairs colorless, flaccid, many- celled, beneath paler and reddish, sparingly and inconspicuously strigose, the hairs bayonet-shaped, appressed, 2-celled, the basal cell short, the apical cell elongate, white, sharp-pointed, with longer, several-celled, straight, yellowish hairs along the midrib and some- times principal veins, the veins 4 or 5 pairs, not prominent above, obvious and very slightly elevated beneath, strongly red-colored along the veins, the veinlets obscure on both surfaces; smaller leaves mostly stipule-like and not over 8 or 9 mm, long, but occasionally better developed and almost half as long as the larger, similarly pubescent; flowers geminate in the axils of the larger leaves; bracts minute, linear, densely hirsute; peduncles 1.8-2.3 cm. long, densely hirsute, and also antrorsely strigose with bayonet-shaped hairs like those of the lower leaf surfaces; calyx all dark rose-red, ca. 2 em. long and wide (incl. the lobes), 5-parted to the base, the sepals deeply laciniate, the central portion 2.5-3 mm. wide, the lateral lobes 2 pairs, linear, 8-11 mm. long, 0.5-0.6 mm. wide, externally strongly red- hirsute, and also with white, appressed, “bayonet’’-shaped hairs, within sparsely hirsute and also slightly strigose with “bayonet’’- shaped hairs; corolla 3-4 cm. long, erect in calyx, rose-red at base, with broad dark purplish-red stripes upwardly running toward the sinuses of the lobes but stopping short of them, and narrow yellow lines, the yellow becoming general toward the throat, the tube 4 mm. wide at base, ventricose upwardly and becoming 7-8 mm. wide at middle, obviously long-hirsute, and bearing also minute, spreading, acute, 1-celled hairs and globular sessile yellow glands, slightly con- tracted to throat, this with numerous glandular hairs within at the base of the galea, this ca. 6 mm. wide, the limb yellow with greenish margins, slightly irregular, 2 lobes ca. 3 mm. long, 3 lobes ca. 2 mm, long, the appendages yellow, obvious, clavate, 2.5-3 mm. long, long- hirsute apically; filaments long-connate at base, glabrous; anthers barely exserted from the corolla tube at anthesis, connate, eventually REVISION OF TRICHANTHA—MORTON 25 free, quadrate, ca. 1.6 mm. long and 2 mm. wide, the cells not contig- uous, separated by the enlarged connective; style glandular-pilo- sulous; stigma obviously bilobed; ovary white-sericeous; disk reduced to a dark-colored, glabrous, truncate, entire posterior gland ca. 1.5 mm. high. Type in the U.S. National Herbarium (no. 2370202), collected on the road between Quito and Santo Domingo, Province of Pichincha, Ecuador, at 1300 m. elevation, Apr. 3, 1942, by Oscar Haught (no. 3215). Portion of holotype deposited in the herbarium of the New York Botanical Garden. Paratypes: Ecuador: Vicinity of Ayabamba,* Province of El Oro, Oct. 15, 16, 1918, Rose & Rose 23462 (NY, US). Rich, moist luxuriant forest along trail from Sambotambo, following headwaters of Rio Moro Moro, south to Buenaventura at and along highway to Portovelo, Province of El Oro, 1035-1800 m. alt., Aug. 29, 1943, “pendent epiphyte from branches; leaves subcoriaceous, dark dull green above with sulcate midrib, pale gray green below with lavender nerves; pedicels and calyx deep rose; corolla purple, the lobes with yellow appendages and green tips, their margins covered with rose- colored hairs,” Steyermark 54202 (US). Dense rich jungle between Rio Blanco and Rio Norcay, on road between Chacanceo and Molle- turo, Province of Azuay, at 1520 m. alt., June 4, 1948, “pendent epiphyte; calyx rose-orchid; corolla tube purple for % part, then greenish-yellow for upper 4, the appendages on corolla lobes yellow; leaves silvery green, below suffused with lavender,” Steyermark 52820 (F, US). Another, slightly doubtful, collection is Lehmann K 273 (K), without locality; the sheet bears a penciled word which may be “Guechi,” “Gueshi,” or “Guesbi,” but it is uncertain if this is a com- mon name or a locality. As indicated above in the discussion under 7. minor, this is the species illustrated in Curtis’ Botanical Magazine t. 4428 (1864) as T. minor. ‘ Ayapamba on the label, but this is an error, as shown by Rose’s field notebook. Index [Synonyms in italics. New species, new names, and combinations in boldface. Page numbers of principal entries in italics.] Alloplectus, 1, 2 Trichantha —Continued Columnea, 1, 2 clara, 5, 15, 19 illepida, 2, 12 elegans, 5, 13, 18, 19, 23 major, 15 formosa, 4, 9, 10 minor, 18 illepida, 3, 4, 5, 10, 12 mooret, 1, 10 major, 5, 16, 18 sect. Stenanthus, 1, 2 minor, 4, 5, 18, 19, 23, 25 sect. Systelostema, 1, 2 moorei, 3, 4, 10 sect. Trichantha, 2, 3 rosea, 3, 4, 7,8 Drymonia, 2 f. latifolia, 8, 9 Erythranthus, 2 f. rosea, 9 Trichantha, 1, 2, 3, 13 f. viridis, 8, 9 aliena, 3, 4, 6 teuscheri, 4, 5, 13, 19, 21, 23 bullata, 5, 16 tropicalis, 5, 13 27 U.S. GOVERNMENT PRINTING OFFICE:1963 BULLETIN OF THE UNITED STATES NATIONAL MUSEUM CONTRIBUTIONS FROM THE UNITED STATES NATIONAL HERBARIUM Vo.umeE 38, Part 2 STUDIES OF FERN TYPES, I By C. V. Morton SMITHSONIAN INSTITUTION PRESS ¢ WASHINGTON, D.C. © 1967 STUDIES OF FERN TYPES, I C. V. Morton In 1934 Carl Christensen, in his Index Filicum, Supplement III, gave the number of recognized species of ferns as 9,387, which today must be considered a conservative estimate. The new fourth supple- ment to the Index Filicum (1965) does not indicate the number of new species cited, but it must be well in excess of 1,000. Most of the well- known older species have numerous specific synonyms or reputed synonyms. I would guess at present that the number of species of ferns is in excess of 15,000. The Index does not list the subspecies, varieties, or forms, and not all the hybrids. I should thus estimate that more than 50,000 entities have been described in the ferns, and there are therefore a corresponding number of types. Needless to say, the earlier botanists did not work with a concept of types in mind. A number of American workers began using the terms “type” and “cotype” (which usually meant a duplicate of the type, 1.e., the present isotype) about 1900, and German fern students used “Typus” or “Original” similarly, but the formal recognition of types in the International Rules of Nomenclature did not begin until the International Botanical Congress in Cambridge in 1930. Since that time, succeeding Congresses have clarified the concepts of types, and the designation of a nomenclatural type has been mandatory since January 1, 1958. When original descriptions are based on a single collection, natu- rally there is no problem as to the type, but when descriptions are based on several collections, i.e., syntypes, as is of course frequently the case, it is necessary according to the Code to designate one of the syntypes as the lectotype. This should not be done routinely, but only after a study of the taxonomy and nomenclature of the group involved. However, it is sometimes not practicable to study all of the syntypes involved; in some such cases one may be obliged to designate a lectotype, making sure that the specimen chosen agrees with the original description. As is well recognized, collections bear- ing the same number may be mixtures, and so each specimen must be carefully scrutinized. Duplicates of the holotype are officially designated as isotypes by the Code. Nonameis ascribed to duplicates 29 30 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM of the syntypes; the term “isosyntype” is here used;? although the meaning is readily evident by analogy, this term could appropriately be added to the section of the Code dealing with the nomenclature of types. The majority of the estimated 50,000 fern names have never been properly typified by the current rules, and in fact many of the types of the older species have never been studied. Even the fern specimens in the Linnaean Herbarium in London have never been studied by an expert pteridologist, identified with current collections, or properly annotated. The types at the Royal Botanic Gardens, Kew, are partially indicated and segregated, especially those of Hooker and Baker, and some of the types in Berlin were indicated by Hieronymus, but in most herbaria the types remain unmarked, except in the case of recently described species. In 1954, I was privileged to receive a grant from the John Simon Guggenheim Memorial Foundation to study fern types in various herbaria in Europe for a period of seven months. During this time through the courtesy of the various directors and curators I studied and photographed 8,011 fern types, mostly in Muséum National d’Histoire Naturelle, Paris, the Rijksherbarium, Leiden, and the British Museum (Natural History), London, although I visited briefly the Jardin Botanique de |’Etat, Brussels, the Conservatoire et Jardin botaniques, Genéve, the Staatsinstitut fiir allgemeine Botanik, Hamburg, the Botanical Museum, Copenhagen, the Riksmuseet, Stockholm, the Botanisk Museum, Oslo, and the Royal Botanic Gardens, Kew. In subsequent years, under the auspices of the Smithsonian Institution, I visited Europe several times and photo- graphed additional types in some of the herbaria mentioned above, and also in the Botanischer Garten, Zurich, the Botanisches Museum, Berlin-Dahlem, and the Royal Botanic Garden, Edinburgh. The total number of type photographs is now 13,011, but not all of these have been labeled and studied. In 1957, I applied for and received a grant from the National Science Foundation (Grant G4080) amounting to $14,900 for the purpose of having duplicate prints made of the original 8,011 photographs that I had taken in 1954, and to have labels typed for these. Altogether more than 40,000 prints were made, which are being distributed to various institutions throughout the world as they are worked over and labeled. The money from this grant having been exhausted, I applied for and received a continuation from the National Science Foundation 1 Because of the numerous mixed collections some botanists have had so little faith in the authenticity of duplicate syntypes that they have referred to them jokingly as “‘arithmotypes.” STUDIES OF FERN TYPES, I—MORTON 31 (Grant GB-1243) for the amount of $5,750. The work of label typing is still going on. The study of the original types, on which I made many notes, and of the photographs has served to clarify the status of many names. ‘The types show that some species have been wrongly interpreted, having been recognized as distinct when they are actually synonyms, or vice versa. Others have been essentially relegated to the status of “species dubiae,” although the types prove to be readily identifiable. Ultimately the number of described species that will permanently remain dubious is bound to be very small after all the herbaria have been combed for types, a task that will occupy fern students for a great many years. Identifying some of the old types will require experts to decipher cryptic marks on the labels and sheets, to identify old handwritings, watermarks of the old paper, and so forth. How- ever, the locating of most types does not require quite these fine details. The following notes are the first of a projected series discussing some of the more interesting types, especially those that involve some change in the current interpretations of the species involved. It is regrettable that the study of types should result in some changes in the specific epithets of some well-known plants, but this is inevitable. It is caused partly by the brief and inadequate descriptions of many early (especially but not exclusively) writers, and by the unwillingness of some later workers to attempt to locate and place the types of earlier writers. Although such changes are bound to be numerous they are not limitless. Ultimately a more stable nomenclature will result, although there will always be plenty of divergence of opinion when it comes to such (essentially subjective) taxonomic matters as the delimitation of genera and broad or narrow specific concepts. The present work follows the precedent-setting work of the lateC. A. Weatherby on the fern types described by Desvaux.? The original names are listed alphabetically, each followed by the correct name, according to the present Code of Botanical Nomenclature and according to the taxonomic system that I follow, which is essentially that of Christensen.® 1. AcrosticHUM acUMINATUM Willd. in L. Sp. Pl. ed. 4, 5:116. 1810=Photi- nopteris acuminata (Willd.) Morton, comb. nov. Lomaria speciosa Blume, Enum. Pl. Jav. 202. 1828. Type: Java, Blume. Photinopteris simplex J. Smith, London Journ. Bot. 3:403. 1841. Nomen nudum. Based on Cuming 64, from Luzon, Philippine Islands. 2“On the types of Desvaux’s American species of ferns,’ Contr. Gray Herb. CXIV:13-35. 1936; and “On certain type specimens in ferns,” Contr. Gray Herb. CX XIV:13-22. 1939. 3 In the Index Filicum, Suppl. III, and in Verdoorn, Manual of Pteridology. 32 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM Photinopteris hors fieldii J. Smith [London Journ. Bot. 3:403. 1841, nomen nudum] ex Hook. & Bauer, Gen. Fil. ¢. 92. 1842. Syntypes: Singapore, Wallich; Java, Horsfield; Luzon, Cuming 362.‘ In consideration of the specific epithet chosen, I select the Horsfield specimen as lectotype; it is presumably at Kew. ?Photinopteris cumingit Presl, Epim. Bot. 192. 1849 [1851]. Type: Min- doro, Philippine Islands, Cuming 362 p.p., presumably in Prague. Presl considered Cuming 362 a mixture, the other part being P. horsfieldit. Photinopteris humboldtii Presl, Epim. Bot. 192. 1849 [1851]. Based on Acrostichum acuminatum Willd. An illegitimate change of specific epithet. Photinopteris speciosa Presl, Epim. Bot. 264. 1849 [1851]. Acrostichum rigidum Wallich ex Hook. Sp. Fil. 5:281. 1864. An illegitimate change of specific epithet. Cited as synonyms are Lomaria speciosa Blume, Photinopteris simplex J. Smith, and P. horsfieldii J. Smith. It is to be considered as a renaming of L. speciosa Blume, with the same type. The Wallich name presumably appears on the sheet that was one of the syn- types of P. hors fieldiz. Type: Herb, Willdenow 19539 (B), photograph (by R. M. Tryon) US, labeled as Peru, Malaspina Expedition. Willdenow based his Acrostichum acuminatum on a sterile specimen collected by the Malaspina Expedition and on the Lingua cervina scandens citri foliis major of Plumier (t. 115), from Martinique. The actual specimen studied and described by Willdenow should be given precedence as the type over the citation of a pre-Linnaean plate, just as most Swartz species of ferns were based on actual specimens, some of which were identified with plates by Plumier. The specimen in the Willdenow Herbarium is clearly Photinopteris speciosa. Presl re- named it Photinopteris humboldtit, probably going chiefly on the very different locality “Peru,” stated on the sheet, but it is clear that this is a wrong locality. Many of the plants of the Malaspina Expedition (collected by Nee or Haenke) were wrongly localized as Peru or Ecuador when they really came from the Philippine Islands, as this one doubtless did. The characteristic genus Photinopteris is not un- common in the Philippine Islands, but has never been found in Peru, or indeed in any place in the New World. The earliest name for this species is Acrostichum acuminatum Willd., and so the above new com- bination P. acuminata is needed. The different species Acrostichum acuminatum Poiret is also 1810, but Poiret’s publication is later than Willdenow’s, according to recent studies by Dr. William Stearn. 2. ACROSTICHUM LANCIFOLIUM Desvaux, Gesell. Naturf. Freund. Berlin Mag. 5:310. 1811=Elaphoglossum lancifolium (Desvaux) Morton, comb. nov. ‘A specimen at Geneva (Morton photographs 3873, 3874) labeled “Cuming 352” is an error of transcription, for Cuming collected this species only once, his number 362. This sheet gives the locality as Mindoro, although Hooker and Bauer give the locality as Luzon. STUDIES OF FERN TYPES, I—MORTON 30 Acrostichum salicifolium Willd. ex Kaulf. Enum. Fil. 58. 1824. Type: Bourbon [Réunion], Desfontaines. Elaphoglossum salicifolium (Willd. ex Kaulf.) Alston, in Exell, Cat. Vasc. Pl. San Thomé 92. 1944. The type of A. lancifolium Desv. came from Mauritius according to the original description, but the type specimen (Herb. Jussieu Cat. 1004, Morton photograph 2887) is marked “Ile de France [Le., Mauritius] et du Bourbon [i.e., Réunion].” It was received from du Petit Thouars. The species does occur both in Mauritius and Ré- union, and also in Madagascar. In the recent treatment of the ferns of Madagascar, Madame Tardieu-Blot continues to use the name E. salicifolium, but the epithet lancifoliwm has priority, and there appears to be no obstacle to its use. 3. ADIANTUM acumINATUM Desv. Ges. Naturf. Freund. Mag. Berlin 5:327. 1811=Adiantum tetraphyllum Humb. & Bonpl. ex Willd. in L. Sp. PI., ed. 4, 5:441. 1810. Type: Puerto Rico, Ledru (P, photograph by Weatherby, US). Weatherby in his study of the fern types of Desvaux did not place A. acuminatum Desv. In the Index Filicum it is placed as a synonym of A. villosum L., but the type shows that it is a rather common form of A. tetraphyllum Humb. & Bonpl. as it occurs in Puerto Rico. The segments are obtuse, and so it may be presumed to be the form described as A. tetraphyllum f. obtusa Kuhn.’ However, since Kubn did not cite a type or any specimens of his forma, a lectotype will have to be selected from material in Berlin that has been identified as f. obtusa by Kuhn, which is presumably not the same as A. ¢etra- phyllum var. obtusum Mett. ex Fourn. Mex. Pl. 129. 1872, a synonym of A. pulverulentum. 4. ADIANTUM PoLIruM Humb. & Bonpl. ex Willd. in L. Sp. Pl. ed. 4, 5:442. 1810=—Adiantum polyphyllum Willd. var. politum (Humb. & Bonpl.) Morton, comb. nov. Type: Cumana, Venezuela, Humboldt & Bonpland (Herb. Willd. no. 20083-1, photograph Tryon, US). A second sheet in Herb. Willdenow (20083-2) was originally in folder 20085 (A. villosum L.) but has been reidentified as A. politum; it appears identical with the holotype of A. politum, and is doubtless a part of the same collection. In the Index Filicum A. politum is listed as a dubious species. A study of the photograph of the type shows that it is closely allied to A. polyphyllum Willd. (type: Caracas, Bredemeyer, Herb. Willd. 20104-1, photograph Tryon, US), which is fully quadripinnate, or in some collections even 5-pinnate at base. The frond is merely tripin- nate in A. politum; this may not be a fundamental distinction, but 5 Jahrb. Bot. Gart. Berlin 1:342. 1881. 34 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM still I have not seen any specimens to match it, and so it may rank asa variety. Dr. Maxon studied the type in Berlin in 1930 and wrote: ‘Tt is a delicate plant of the polyphyllum alliance and looks like A. pectinatum Fée, with very small sori.”’ However, it is doubtful if there is any relationship at all with A. pectinatum, which is somewhat similar in the division of the blade but which differs strongly in having pubescent rhachises and rhachillas; in A. politum (and A. polyphyllum) these are wholly glabrous, dark purple, smooth, and shining. A synonym of A. polyphyllum var. polyphyllum is A. cardiochlaena Kunze,° judging from the description and the comments by Kunze. 5. ADIANTUM SCHAFFNERI Fourn. Bull. Soc. Bot. France 27: 328. 1880= A. capillus-veneris L. (forma). Typz: “In montibus ca. San Luis de Potosf,’’ Mexico, October 1877, Schaffner 64 (holotype P, Morton photograph 2623). The original description is as follows: L’envoi de M. Schaffner contient encore plusieurs espéces que je crois nouvelles, et sur lesquelles j’aurai l’occasion de revenir. Ce sont: 1. L’Adiantum Schaffneri (no. 64).—J’ai envoyé, il y a plusieurs mois, cette plante 4 M. Bommer, qui prépare une monographie du genre Adiantum: je regrette de n’avoir pas regu de réponse de ce savant; mais je ne voudrais pas la déecrire, de crainte de surcharger la nomenclature d’un double emploi inutile, M. Bommer pouvant lui avoir donné un autre nom dans son mémoire en cours dimpression. L’ Adiantum Schaffneri a les pinnules de IA. Capillus Veneris, mais les frondes courtes, 4 peine ramifiées et souvent méme simplement pinnées, naissant trés serrées sur un rhizome horizontal. As is to be seen from the above, Fournier says that he does not want to describe a new species, since Bommer may be describing the plant also, but then he goes ahead and assigns a specific epithet Schaffnert and gives enough description to validate a species ordi- narily. Christensen accepted A. schaffnert as validly published, but I am inclined to reject it as invalid, under Art. 34 of the Code (1961 edition), which states that a name is not validly published when it is not accepted by the author who published it. The matter is not of great importance, since the type seems to be only a depauperate form of the common A. capillus-veneris with the fronds simply pinnate or slightly bipinnate only at the very base. 6. AspipIum aToMARIUuM Muhl. ex Willd. in L. Sp. Pl. ed. 4, 5:279. 1810= Cystopteris fragilis (L.) Bernh. var. mackayi Lawson, Fern Flora of Can- ada 233. 1889. In the Index Filicum and also in Broun’s Index to North American Ferns this species is listed as a synonym of Cystopteris bulbifera 6 Linnaea 17:569. 1843; type: Caracas, Jan.—Apr. 1842, J. Linden, which I have not seen. STUDIES OF FERN TYPES, I—MORTON 35 (L.) Bernh., and it is retained there by Blasdell in his recent ‘‘A Monographic Study of the Fern Genus Cystopteris,” although there is no indication that he ever saw the type or made any effort to do so. The holotype is in the Willdenow Herbarium (B), sheet no. 19822. It is a sterile plant, which is referable to C. fragilis rather than to C. bulbifera. From its locality, and general appearance, it belongs to the var. mackayi, which Blasdell considers a hybrid between C. diaphana and C. fragilis, which is unbelievable since one of the parents (diaphana) does not occur within thousands of miles of Pennsylvania. 7. AspIpIUM ATTENUATUM Kunze ex Mettenius, Abh. Senckenb. Ges. Frankfurt 2:380. 18587=Thelypteris attenuata (Kuntze) Morton, comb. nov. Based on Dryopteris attenuata Kuntze, Rev. Gen. Pl. 2:812. 1891. Lastrea attenuata J. Smith in Hook. Journ. Bot. 3:412. 1841, nomen nudum. Nephrodium attenuatum (Kunze) Baker in Hook. & Bak. Syn. Fil. 263. 1867, non Moore, 1858. Dryopteris attenuata Kuntze, Rev. Gen. Pl. 2:812. 1891. New name for Aspidium attenuatum Kunze, non Swartz. Dryopteris stenobasis C. Chr. Ind. Fil. 294. 1905. Based on Aspidium attenuatum Kunze, non Swartz. Since the name D. attenuata Kuntze was available and correct, D. stenobasis was superfluous when published and therefore illegitimate. Typs: Samar, Philippine Islands, Cuming 327 (isotype BM, Morton photo- graph 6451). The basis for the validity of Dryopteris attenuata Kuntze is Art. 72 Note: “When a new epithet is required, an author may, if he wishes, adopt an epithet given to the taxon in an illegitimate name [here Aspidium attenuatum Kunze, non Swartz], if there is no obstacle to its employment in the new position or sense [here there was no available prior epithet and no previous use of the combination Dryopteris attenuata]; the epithet in the resultant combination is treated as new [thus here D. attenuata Kuntze, a new name, and not D. attenuata (Kunze) Kuntze].” Copeland® adopts for this species the name Cyclosorus alatellus (Christ) Copel., based on Nephrodium alatellum Christ.? He may be right that these species are the same, but the valid use of the epithet attenuata, dating from 1891, has priority over N. alatellum, dating from 1901. 7 Non Aspidium attenuatum Swartz in Journ. Bot. Schrad. 18001:34. 1801. 8 Gen. Fil. 142. 1947. ° In Schum. & Lauterb. Fl. Deut. Siidsee 112. 1901, based on a type from New Guinea. 36 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM 8. ASPIDIUM CHONTALENSE Fourn. Bull. Soc. Bot. France 19:254. 1872=Lastre- opsis exculta (Mett.) Tindale, Vict. Nat. 73:185. 1957 subsp. guate- malensis (Baker) Tindale, Contr. New So. Wales Nat. Herb. 3:245. 1963. Nephrodium guatemalense Baker in Hook. & Bak. Syn. Fil. ed. 2, 498. 1874. Syntypes: Guatemala, Salvin « Godman ; Chiapas, Ghiesbreght 422. Dryopteris exculta (Mett.) C. Chr. var. guatemalensis (Baker) C. Chr. Dansk. Vid. Selsk. Skrift. VIII, 6:96. 1920. Tyre (of A. chontalense): Chontales, Nicaragua, Levy 516 (holotype P, Morton photograph 4657). In his monograph of Dryopteris, Christensen regarded Aspidium chontalense Fourn. as a dubious species, thinking it perhaps the same as D. hemsleyana or D. chiriquiana. An examination of the holotype shows that it goes in a different section from these species, and is in fact identical with the rather common Central American plant usually known as D. ezeulta var. guatemalensis. This species, along with the closely related D. effusa, has been placed in a distinct genus Para- polystichum, which is not sufficiently distinct from Lastreopsis, according to the recent treatment by Tindale.’ 9. ASPIDIUM DIPLAZIOIDES Moritz ex Mett. Abh. Senckenb. Ges. Frankfurt 2:367. 1858=Thelypteris diplazioides (Moritz) Ching, Bull. Fan. Mem. Inst. Biol., Bot. 10:251. 1941. Dryopteris diplazioides (Moritz) Kuntze, Rev. Gen. Pl. 2:812. 1891. Dryopteris moritziana Urban, Symb. Antill. 4:21. 1903 (Illegit.). Type (of Aspidium diplazioides Moritz): Moritz 408, from Tovar, Venezuela ; the original was probably lost in Leipzig, but I have seen isotypes at Hamburg, Leiden, and Paris (Morton photographs 1211, 4712, 5277, 5278, 5279). The new combination was published by Ching without comment, along with a large number of other new combinations, with some errors as follows: “T. diplazioides (Maritz) comb. nov. Dryopteris dipalzioides (Maritz) Urban, C. Chr. Ind. 278” Aside from the misspellings of Moritz and diplaziowes, the authority stated is incorrect—it should be (Moritz) Kuntze. Urban’s Dryopteris diplazioides (Desv.) Urban was based on Gymnogramma diplazioides Desv., an altogether different species (C. Chr. Ind. 262). However, this wrong citation of basionym will not invalidate the combination, which is clear from the parenthetical citation of Moritz and from the page reference to Christensen’s Index Filicum. 10. AspipIuM GERMANII L’Herminier ex Fée, Mém. Foug. 11:82, t. 22, f. 3. 1866=Thelypteris kunthii (Desv.) Morton (see p. 53). 1 Monograph of the genus Lastreopsis Ching. Contr. New So. Wales Nat. Herb. 3:249-339, pl. I-X XIII. 1965. STUDIES OF FERN TYPES, I—MORTON 37 Typu: Rivitre Noire, Ravine-Paradis, Guadeloupe, L’ Herminier in 1861. There are two specimens collected in Guadeloupe by L’Herminier at Paris that agree with Fée’s description; they are labeled A. germanz but can hardly be holotypes since they lack the detailed locality data cited by Fée; however, in the absence of the Fée Herbarium, perhaps lost, the one with rhizome (Morton photograph 4677) can be designated lectotype; the other one, obviously a part of the same collection is an isolectotype (Morton photograph 4426). It has always seemed rather odd that a species as widespread and relatively distinct as Dryopteris normalis C. Chr. never had a name before 1910. It appears now that there are two earlier names, at least, one of which forms the basis of Thelypteris kunthii (see below) and the other Aspidium germanii Fée (originally spelled “germanv’). The latter was considered as a dubious species by Christensen in the Index Filicum, and as dubious also in his monograph of Dryopteris, where it is mentioned under J. patens (Swartz) Kuntze but still considered dubious. ‘The specimens selected above as lectotypes are surely normalis. The peculiar thing is that there does not seem to be any other material of normalis from Guadeloupe. However, the species has been collected on Dominica, and there is every reason to expect it to occur in Guadeloupe and the other Windward Islands. 11. ASPIDIUM INTERMEDIUM Muhl. ex Willd. in L. Sp. Pl. ed. 4, 5:262. 1810 = Dryopteris intermedia (Muhl.) A. Gray, Manual ed. 1, 630. 1848. I have studied the holotype (Pennsylvania, Muhlenberg, in Herb. Willd. no. 19788, photograph by Tryon, US). It consists of three portions of fronds. The frond at the left is Dryopteris spinulosa (O. F. Muell.) Watt in its usual sense for the eastern United States. The two fragments at the right are D. intermedia in the usual sense. The two latter are hereby designated as lectotypes, and the different frond at the left excluded, although it undoubtedly formed a part of the original concept of the species. The current nomenclature is thus preserved. 12. Asprpium LEvy1 Fournier, Bull. Soc. Bot. France 19:255. 1872= Thelypteris levyi (Fourn.) Morton, comb. nov. Dryopteris levyi Kuntze, Rev. Gen. Pl. 2:813. 1891. Type: Chontales, Nicaragua, Levy 463 (Holotypus P, Morton photograph 4690). For comments regarding the relationships of this rare species of the section Goniopteris see Christensen, Monograph of the Genus Dryopteris 1:211. 1913. 13. Asprprum opposirum Kaulf. ex Spreng. in L. Syst. Veg. ed. 16, 4:108. 1827 =Ctenitis opposita (Kaulf.) Copel. Gen. Fil. 124. 1947 [as “oposita’’]. Dryopteris opposita (Kaulf.) Kuntze, Rev. Gen. Pl. 2:813. 1891. 38 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM Dryopteris mascarenarum Urban, Symb. Antill. 4:14. 1903. Based on Aspidium oppositum Kaulf., 1827, non A. oppositum Swartz, 1829. Ille- gitimate because superfluous when published. The species A. oppositum Kaulf. had priority over the Swartz name, and it was properly transferred to Dryopteris by Kuntze.'! Ctenitis mascarenarum Tardieu-Blot, Notul. Syst. 15:90. 1954. Ilegiti- mate because the earliest available specific epithet was not adopted. The locality given by Sprengel was “C.B.S.,” ice., Cape of Good Hope, but this was an error, for this species is not known from South Africa. Most of the new Kaulfuss’ species published by Sprengel were based on the collections of Sieber, and this one was evidently based on Sieber 36 from Mauritius (Morton photograph 5274). The species is restricted to Mauritius, Réunion, and Madagascar. In her latest work on the ferns of Madagascar, Madame Tardieu has continued to use the name Ctenitis mascarenarum (1958, p. 342), apparently overlooking entirely the publication of C. opposita (Kaulf.) Copel. 14. ASPIDIUM SCLEROPHYLLUM Poeppig ex Spreng. in L. Syst. Veg. ed. 16, 4:99. 1827=Thelypteris sclerophylla (Pocppig ex Spreng.) Morton, Amer. Fert Journ. 41:87. 1951. In making the new combination I gave the basionym as Aspidium sclerophyllum Kunze ex Spreng. Syst. Veg. 4:99. 1827, following the usage in the Index Filicum. I remarked in a footnote that Sprengel actually attributed the species to Poeppig, but that this was pre- sumably an error. It is true that later, in 1834, Kunze attributed the species to himself, and he may very well have provided the name for Poeppig. Still, the name is attributed to Poeppig, and there is absolutely no mention of Kunze in the original description. I have now seen two isotypes (in Leiden and Hamburg, Morton photographs 1052 and 5238, respectively) bearing original labels reading “Aspidium sclerophyllum En. Pl. Cub. Ms.” referring to Poeppig’s own unpub- lished manuscript on his Cuban collections. Therefore, it seems necessary to regard Poeppig as the author. The description itself was presumably provided by Sprengel. Although primarily a zoolo- gist, Poeppig did work on his own botanical collections and published some taxonomic botanical papers. The same arguments apply to the authorities of the names Anemia cicutaria Poeppig, A. euneata Poeppig, Polypodium barbatum Poeppig, and Adiantum fructuosum Poeppig, all of which have generally been attributed erroneously to Kunze. "Of course the name Dryopteris opposita (Vahl) Urban, Symb. Antill. 4:14. 1903, is illegitimate, being a later homonym of D, opposita (Kaulf.) Kuntze, 1891. STUDIES OF FERN TYPES, I—MORTON 39 15. ASPLENIUM ADIANTOIDES Lam. Encycl. Méth. 2:309. 1786. In the Species Plantarum (1753), Linneaus proposed Trichomanes adiantoides as follows: Trichomanes frondibus pinnatis: pinnis ensiformibus acuminatis inciso-serratis: serraturis bifidis. Fl. zeyl. 385. Adianthum africanum rutae murariae aemulum, segmentis longioribus acutis. Pluk. alm. 10. t. 123, f. 6. Filix non ramosa zeylanica, foliis adianthi in modum serratis. Burm. zeyl. 97, t. 43. Habitat in India, Africa. Thus, Trichomanes adiantoides L. was based on two elements, one from Ceylon, illustrated by Burmann from a Hermann specimen, and one from Africa, illustrated by Plukenet. These two elements are certainly not the same species, and are probably not even closely allied; the former, from Ceylon, has been known generally as Asplenium faleatum Lam. and the latter, from Africa, as Asplenium praemorsum Swartz or A. furcatum Thunb. Nomenclaturally, there is an intricate problem brought on by the treatment of Lamarck, who was the first to realize that two species were involved. In 1786, Lamarck described the two species, the Ceylon plant as Asplenium faleatum Lam., citing the Burmann t. 43, and the African plant as Asplenium adiantoides Lam., treated as though it were a new species, although the name appears to have been adopted from Linnaeus, and Plukenet ¢. 123, f. 6, is cited as a synonym, and mention made of the fact that Trichomanes advantoides L. was a mixture. If would seem that the type of 7. adiantoides L. must be the Ceylon plant, for Linnaeus diagnosed his species in his own Flora Zeylanica in 1747 on the basis of Burmann’s plate and very likely also on an actual specimen collected in Ceylon by Hermann. Lamarck was therefore unjustified in typifying the Linnaean species on the basis of the African plant illustrated by Plukenet. Instead of describing the Ceylon plant as a new species, A. faleatum, Lamarck should have called it A. adiantoides (L.) Lam., and the African plant should have been the new species (from Lamarck’s information; actually, the African species had already been described unknown to Lamarck three times: as Trichomanes aethiopicum Burm. (1768), Asplenium lanceolatum Forsk. (1775), and Acrostichum filare Forsk. (1775)). Therefore, it appears that A. faleatum Lam. was an illegiti- mate name, since the type of J. adiantoides was cited in synonymy, and that A. adiantoides Lam. must be treated as a new species rather than a new combination, legitimate, but with legitimate prior taxo- nomic synonyms. The type of A. faleatum Lam. must be considered the same as that of Trichomanes adiantoides L., since it is a renaming of the Linnaean species and not a new species, and therefore it is the 40 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM t. 43 of Burmann and the Hermann specimen from Ceylon, if that is in existence. Christensen came to this same conclusion in his Index Filicum, and attempted to remedy it by proposing for the Ceylon species (A. faleatum) the new combination Asplenium adiantoides (L.) C. Chr. (Ind. Fil. 99. 1905) (not A. adiantoides Lam., 1786, not Raddi, 1819, not Raoul, 1844), but of course this is illegitimate, being a later homonym thrice over, although it has still remained in use somewhat, e.g., by Tardieu in the new Flora of Madagascar. The next oldest name that is usable, and which refers to this species in a broad sense is Asplenium polyodon Forster (Prodr. 80. 1786). The type of A. adiantoides Lam. was not indicated. Two specimens were cited, one from the Cape of Good Hope, Sonnerat and one from the Ile de France, Commerson. Finally there was also cited Peru, Jussieu (Herb. Jussieu Cat. 1252, P, Morton photograph 3030), but the comment “mais cette derniére est 4 pinnules un peu plus grandes,” indicates that the Peruvian plant was not considered typical; the speci- men represents a form of the American A. praemorsum Swartz. The Sonnerat plant is presumably the one now in the Lamarck Herbarium at Paris (Morton photograph 2752), which lacks locality data on the sheet; and it may be also the very young plant, just partially devel- oped, which is annotated ‘Asplenium adiantoides, dict.’ (Morton photograph 2753). The Commerson specimen from Ile de France, now in the Jussieu Herbarium at Paris, is a fine, mature specimen (Morton photograph 3029); it must have been the one chiefly used in drawing up the original description; this latter specimen is here desig- nated as lectotype. This species is widespread in both the Old and New World; it should be known at present as Asplenium aethiopicum (Burm.) Becherer. It is extremely variable and may be a collective species. The American plants, if they can be separated, would continue to be known as A. praemorsum Swartz. The synonymy of these two species may be summarized as follows: Asplentum aethiopicum (Burm.) Becherer, Candollea 6:23. 1935. Trichomanes aethiopicum Burm. Fl. Cap. Prodr. 28 bis. 1768. Asplenium lanceolatum Forsk. Fl. Aeg. Cent. VII. 185. 1775, non Huds. 1762. Acrostichum filare Forsk. Fl. Aeg. Cent. VII. 184. 1775. Asplenium adiantoides Lam. Encycl. Méth. 2:309. 1786. Asplenium praemorsum Swartz, Prodr. Veg. Ind. Oce. 130. 1788. Asplenium falsum Retz. Obs. 6:38. 1791. Asplenium furcatum Thunb. Prodr. Fl. Cap. 172. 1800. Asplenium filare Alston, Journ. Bot. 72:4. 1934. Asplenium polyodon Forst. Fl. Ins. Austr. Prodr. 80. 1786. Trichomanes adiantoides L. Sp. Pl. 2: 1098. 1753. STUDIES OF FERN TYPES, I—MORTON 4] Asplenium falcatum Lam. Encycl. Méth. 2:306. 1786. Nom. abort. Asplenium cultratum Gaud. Frey. Voy. Bot. 317. 1827.” Asplenium intermedium Kaulf. ex Spreng. in L., Syst. Veg. ed. 16, 4:84. 1827, non Pres] 1822. Asplenium kaulfussii Presl, Tent. Pterid. 106. 1836, non Schlecht. 1825. Asplenium cumingii Mett. Fil. Hort. Lips. 74, ¢. 12, f. 8. 1856. Asplenium forsterianum Col. Tasm. Journ. 2:171. 1845.” Tarachia haenkeana Presl, Epim. Bot. 76. 1849 (1851)." Asplenium adiantoides (L.) C. Chr. Ind. Fil. 99. 1905, non Lam., 1786, non Raddi, 1819, non Raoul, 1844. 16. Asplenium affine Swartz var. tanalense Baker, Journ. Bot. Brit. & For. 18:329. 1880. Asplenium gilpinae Baker, Journ. Linn. Soc. [London] 16:200. 1877. Asplenium affine var. gilpinae Tardieu, Mém. Inst. Sci. Madagasce. 7:48. 1956. Tyre: Tanala, Madagascar, L. Kitching (not seen, presumably at K). Madame Tardieu has used the name var. gilpinae for this plant in her account of the ferns of Madagascar."® However, although the epithet gilpinae has priority (1877) as a species, the name var. tanalense (1880) has priority as a variety over gilpinae (1956), and consequently the name var. tanalense is correct, since names have priority only within their own rank." 17. ASPLENIUM MACDONELLII Beddome, Journ. Bot. Brit. & For. 27:73. 1889= Dryoathyrium macdonellii (Beddome) Morton, comb. nov. Cornopteris macdonellii Tardieu, Amer. Fern Journ. 48:32. 1958. Parathyrium macdonellii Holttum, Kew Bull. 1958:449. 19959. Type: Chumba Valley, Himalaya Mountains, India, 5,000 feet alt., Macdonell (not seen, presumably K). For a comment on Dryoathyrium see under Dryopteris forsythi- majoris (p. 43). 18. ATHYRIUM PRAESTANS Copel. Amer. Fern Journ. 38:132. 1948=Diplazium praestans (Copel.) Maxon ex Morton, comb. nov. Type: Chazuta, Peru, Klug 4002 (holotype US). In describing this species as an Athyrium, Dr. Copeland was follow- ing his own views as outlined in his Genera Filicum (1947), in which the large genus Diplazium was reduced to a synonym of Athyrium. This illustrates very well the inconsistency of Copeland’s views regarding genera. Sometimes, as in the Hymenophyllaceae and some groups of Polypodiaceae, he split the genera so finely that the charac- ters are hardly more than specific, and again large and generally 12 Synonymous fide C. Chr. Ind. Fil. 1905. 18 In Humbert, Fl. Madag. Fam. 5, 1:234. 1958, 4 International Code of Botanical Nomenclature, 1961 ed., Art. 60. 42 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM recognized and generally distinguishable genera like Diplazium and Athyrium are lumped together with a minimum of justification and justice. Most species are readily and easily placed in Diplazium or Athyrium, and detailed study would allow the definite placement of the few species that seem at present dubious or intermediate. It seems likely that this could be done readily if fresh material becomes available, since it has been found by Manton and Sledge, Brownlie, Mehra and Bir, and others that Athyrium has a basic chromosome number of z=40 and Diplazium r=41. Inplazium praestans, originally segregated under this name by Dr. Maxon but left unpublished, has been known only from the Depart- ments of San Martin and Junin, Peru. It may now be reported from the Department of Loreto: San Alejandro River, Department of Loreto, Peru, July 24, 1958, F. Woytkowski 5118 (US). Mr. Woytkow- ski observed that there were 30 to 40 plants, all distinct, crowded in a small area of a wet shady forest at an elevation of about 400 meters. 19. BLECHNUM TREUBII van Alderw. van Rosenb. Bull. Dept. Agric. Indes Neerl. 18:13. 1908=Blechnum lanceola Swartz, Kungl. Svenska Vetens.-akad. Handl. 1817:71, t. 8, f. 2. 1817. Tyre: Cultivated in the Botanical Garden, Bogor. A specimen in the Rijks- herbarium labeled as “Cult. in Hort. Bog. II. K. X. 23” (Morton photograph 728) is probably authentic material, if not a part of the type. This species was said by van Alderwerelt van Rosenburgh to have been brought to the botanical garden in Bogor from Mount Salak, Java, but this is hardly believable. No one else has ever found a similar plant growing wild in Java, but the specimen is identical with the well-known Brazilian species Blechnum lanceola Swartz, the most distinctive species in the subgenus Blechnum, the only species having a simple blade, this absolutely entire and unlobed. It must be presumed that the plant was mislabeled in Bogor, and that it really was introduced from Brazil. 20. Dryopreris arcana Maxon & Morton, Bull. Torr, Bot. Club 65:352, t. 11. 1938=Thelypteris arcana (Maxon & Morton) Morton, comb. nov. Type: Tena to Napo, Ecuador, Mexia 7174 (US). 21. DryorreRIs CONSOBRINA Maxon & Morton, Bull. Torr. Bot. Club 65:356. 1938=Thelypteris consobrina (Maxon & Morton) Tryon, Rhodora 69:5. 1967. Type: La Merced, Peru, Killip & Smith 24087 (holotype US, isotype NY). 22. DryYopreris ENSIFORMIS C. Chr. Dansk. Vid. Selsk. Skrift. VII, Naturv. Afd. 10(2) :269, f. 46. 1913=Thelypteris ensiformis (C. Chr.) Tryon, Rhodora 69:6. 1967. Type: La Palma, Costa Rica, Tonduz (Herb. Inst. Nat. Cost. 12533) (holotype C, isotype US). STUDIES OF FERN TYPES, I—MORTON 43 93. DRYOPTERIS FORSYTHU-MAJoRIS C. Chr. Dansk Bot. Ark. 7:63. 19382= Dryoathyrium forsythii-majoris (C. Chr.) Morton, comb. nov. Parathyrium forsythii-majoris Holttum, Kew Bull. 1958:449. 1959. Type: Madagascar, Forsyth Major 169 (isotype C, Morton photograph 5685). In 1956, Mme. Tardieu-Blot published a paper ‘5 on the genus Cornopteris Nakai, and referred a number of species to the genus that had formerly been called Dryopteris or other genera. The species Dryopteris forsythii-majoris was formally transferred to Cornopterts in the American Fern Journal (48:32. 1958). However, not long there- after R. E. Holttum published ! a paper “Parathyrvum, a new genus of ferns, with comments on Cornopteris Nakai,” in which he showed that the type of Cornopteris, C. decurrenti-alata (Hook.) Nakai, is closely allied to Athyrium, and is essentially an exindusiate Athyrium, whereas C. boryana (Willd.) Tardieu (based on Aspidium boryanum Willd.) and its allies constitute a different genus, allied to Ctenitis, for which he proposed the name Parathyrium. Unfortunately, he overlooked the fact that the same species Aspidium boryanum Willd. had already been made the type of a new genus Dryoathyrium Ching in 1941 as he pointed out in a subsequent paper.'? Ching and Tagawa have transferred several species to Dryoathyrium, but there are several more that belong there. 24. DrYopTERIS LINGULATA C. Chr. Dansk. Vid. Selsk. Skrift. VII. Naturv. Afd. 10(2):271. 1913=Thelypteris lingulata (C. Chr.) Morton, comb. nov. Typre: Rio Hondo, Costa Rica, Pittier 10349 (isotype US). 25. DRYOPTERIS MINUSCULA Maxon, Kew Bull. Mise. Inf. 1932:135. 19382= Thelypteris minuscula (Maxon) Morton, comb. nov. Tyre: Rio Dagua, near Buenaventura, Colombia, Lehmann 4433 (holotype K). 96. DrropreRis NEsIoTicA Maxon & Morton, Bull. Torr. Bot. Club 65:362, ¢. 12. 1938=Thelypteris nesiotica (Maxon & Morton) Morton, comb. nov. Type: Trinidad, Jenman (holotype NY). 27. DRYOPTERIS PARVISORA C. Chr. Ark. fér Bot. 14(19):5. 1916=Dryoathyrium parvisorum (C. Chr.) Morton, comb. nov. 16 “Sur le genre Cornopteris Nakai et les Cornopteris Malgaches,”’ Mém. Inst. Sci. Madagascar, Sér. B, 7:27-32. 1956. The species C. forsythii-majoris is illustrated in figure I, nos. 6-8. 16 In “Notes on Malaysian Ferns, with descriptions of a new genus and a new species,’ Kew Bull. 1958:447-455. 1959. 7 “Vegetative characters distinguishing the various groups of ferns included in Dryopteris of Christensen’s Index Filicum, and other ferns of similar habit and sori,”? Gardens’ Bull., Singapore, 17:361-367. 1960. 232-200—67——2 44 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM Cornopteris parvisora Tardieu, Amer. Fern Journ. 48:32. 1958. Parathyrium parvisorum Holttum, Kew Bull. 1958:449. 1959. Tyre: Moramanga, Madagascar, Oct. 1, 1912, Afzelius& Palm 371 (holotype not seen, presumably 8). 28. DRYOPTERIS sTANDLEYI Maxon & Morton, Bull. Torr. Bot. Club 65:368 1938=Thelypteris standleyi (Maxon & Morton) Tryon, Rhodora 69:8. 1967. Type: Quirigudé, Guatemala, Standley 23126 (holotype US). 29. DRYOPTERIS TURRIALBAE Rosenst. Repert. Sp. Nov. Fedde 22:10. 1925= Thelypteris turrialbae (Rosenst.) Morton, comb. nov, Type: Turrialba, Costa Rica, Brade 357 (isotype NY). 30. ELAPHOGLOSSUM ALATUM Gaudichaud, in Vaillant, Voyage Autour Monde Bonite Bot. Atlas #. 136, 1845-50= Elaphoglossum gorgoneum (Kaulf.) Brack. Acrostichum gorgoneum Kaulf. Enum. Fil. 63. 1824. Type: Oahu, Hawaiian Islands, Chamisso (presumably LE). Acrostichum sessile Fée, Gen. Fil. 43. 1852. Based on Elaphoglossum alatum Gaud. A renaming, because of the unavailability of the epithet alatum under Acrostichum. Not Acrostichum alatum Roxb. (1844) or A. alatum Fée (1845). Elaphoglossum gorgoneum (Kaulf.) Brack. U.S. Expl. Exped. 16:74. 1854. Elaphoglossum sessile (Fée) Moore, Ind. Fil. 14. 1857. Illegitimate, since the name FE. alatum Gaud. was legitimate and available under the genus Elaphoglossum. The species Elaphoglossum sessile (Fée) Moore is listed as dubious in Christensen’s Index Filicum, and apparently has remained unplaced. As shown by the synonymy above, this name is illegitimate, since Moore did not adopt the earliest specific epithet available, namely Elaphoglossum alatum Gaud. Although there is no description, the plate provided by Gaudichaud, which is a fine one, with analyses, is a valid publication by Article 44 of the International Code of Botanical Nomenclature (ed. 1961). The holotype in the Muséum National d’Histoire Naturelle, Paris, has been studied (Morton photograph 4035); it is obviously referable to the distinctive endemic species of the Hawaiian Islands . gorgoneum, which has priority. For addi- tional taxonomic synonyms, see Robinson, Bull. Torr. Bot. Club 39:572. 1912. The date of publication of t. 185 of Gaudichaud is uncertain, but between 1845 and 1850.18 31. Elaphoglossum pellucidum Gaudichaud, in Vaillant, Voyage Autour Monde Bot. Atlas t. 79, f. 6. 1844. Acrostichum micradenium Fée, Mém. Foug. 2:43. 1845. Elaphoglossum nitidum Brack. U.S. Expl. Exped. 16:70. 1854. Type: Hawaiian Islands, Wilkes Expedition (holotype US). Elaphoglossum micradenium (Fée) Moore, Ind. Fil. 12. 1857. 18 See Johnston, Journ. Arn. Arb. 25:481-487. 1944. STUDIES OF FERN TYPES, I—MORTON 45 Elaphoglossum pellucidum Gaud. was rejected in Christensen’s Index Filicum in favor of E. micradenium (Fée) Moore, apparently because it was without a description. However, the plate with analyses provided by Gaudichaud validates the name according to the Code of Nomenclature. There is a further question regarding the date of publication, given as 1846 by Christensen. However, ¢. 79 was published in 1844, according to Johnston.’? A sheet in the herbarium in Paris collected by Gaudichaud (no. 13), from the “Tles Sandwich,” September and October, 1836, on the Voyage of the Bonite, was obviously used for the illustration in the publication; this holotype is Morton photograph 4028. Although it is the holotype of E. pellucidum it does not bear this name. Fée’s Acrostichum micra- denium was also based on a Gaudichaud collection from the Hawaiian Islands, and the holotype is probably the same sheet as the holotype of E. pellucidum, or at least this sheet is an isotype. 32. Gonroprerts MOLLIS Fée, Gen. Fil. 252. 1852=Thelypteris ghiesbreghtii (Hook.) Morton, comb. nov. Polypodium ghiesbreghtii Linden, Cat. 18. 1858 ? (not seen; nom nud.?) From the name, the type would appear to have been cultivated material, received originally from Ghiesbreght, probably from Chiapas, Mexico. Polypodium crenatum var. ghiesbreghtit Hook. Sp. Fil. 5:3. 1864. Type: Tabasco, Mexico, Linden 1499 (as “Gheisbeghtiv’”’). Polypodium ghiesbreghtii Linden ex Baker, in Hook. & Bak. Syn. Fil. 315. 1868. Dryopteris ghiesbreghtit C. Chr. Ind. Fil. 267. 1905. Dryopteris mollis (Fée) Maxon, Contr. U.S. Nat. Herb. 13:18. 1909, non D. mollis (Jacq.) Hieron. (1907). Typr: Tabasco, Mexico, Linden 1499 (isotype P, Morton photograph 4679 Although Goniopteris mollis Fée is the earliest specific name, the epithet mollis may not now be transferred to Thelypteris because of T. mollis (Mett.) Tryon. This is a common species of the section Goniopteris, allied to Thelypteris poiteana (Bory) Proctor. Maxon knew it from Tabasco, Guatemala, and Costa Rica in 1909, but it has since been found to be quite common, and is known also from Chiapas, British Honduras, Honduras, Nicaragua, and Panama. 33. GyMNOGRAMMA poLypopio1pEs Link, Hort. Berol. 2:50. 1833 = Thelypteris linkiana (Presl) Tryon, Rhodora 69:6. 1967. Gymnogramma diplazioides Desv. Mém. Soc. Linn. Paris 6:214. 1827. Type: Hispaniola, without collector (presumably P). Grammitis linkiana Presl, Tent. Pterid. 209. 1838. Based on Gymnogramma polypodioides Link, Hort. Berol. 2:50. 1833, non Sprengel, 1827. Dryopteris diplazioides Urban, Symb. Antill. 4:21. 1903, non Kuntze, 1891. Dryopteris linkiana Maxon, Journ. Washington Acad. Sci. 14:199. 1924. 19 See Journ. Arn. Arb. 25:481-487. 1944. I am indebted to Mr. William T. Stearn for pointing out this Johnston reference to me. 46 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM Thelypteris diplazioides (Desv.) Proctor, Bull. Inst. Jamaica Sci. Ser. 5:59, 1953, non 7. diplazioides (Moritz ex Mett.) Ching, Bull. Fan. Mem. Inst. Biol., Bot. 10:251. 1941. Type: A cultivated specimen (presumably B). This species has had a regrettable nomenclatural history, because of the existence of another specific epithet diplaziordes for an entirely different species which Kuntze transferred to Dryopteris in 1891 as D. diplazioides (Moritz) Kuntze. N evertheless, in spite of the existence of this earlier Dryopteris diplazioides, Urban in 1903 proposed the new combination D. diplazioides (Desv.) Urban, which of course was a later homonym and _ therefore illegitimate, but was nevertheless erroneously adopted in the Index Filicum. At the same time the perfectly valid Dryopteris diplazioides (Moritz) Kuntze was renamed D. moritziana Urban, another illegitimate name, superfluous when published, which was also recognized by Christensen. This nomen- claturally impossible treatment has inevitably created some con- fusion, which was continued by the publication by Mr. Proctor of yet another illegitimate combination Thelypteris diplazioides (Desv.) Proctor. The earliest available specific name remains linkiana. 34, HEMICARDION CUMINGIANUM Fée, Gen. Fil. 283. 1852== Cyclopeltis cumingi- ana (Fée) Morton, comb. nov. Lastrea presliana J. Smith in Hook. Journ. Bot. 3:412. 1841, nom. nud, Cyclopeltis presliana Berkeley, Introd. Crypt. Bot. 517. 1857. Polystichum preslianum Moore, Ind. Fil. 84. 1857. Type: Luzon, Philippine Islands, Cuming 68 (isotype BM, Morton photograph 6623). Lastrea prestiana J. Smith was based on “Nephrodium semicordatum, Presl (exclus. syn. Sw. Willd.),” evidently referring to Presl’s treat- ment in Reliquiae Haenkeanae 32. 1825. Presl’s Nephrodium semicordatum is a new combination based on Polypodium semicor- datum Swartz (1788) and must have the same type, namely a plant from the West Indies. Presl did refer to the species a specimen from Luzon, but he gave no description of it apart from a general descrip- tion including the American type. Therefore, there is no description in Presl to base a new species on. Smith himself gave no description, except the comment that the sori are in the middle of the venules in his LZ. presliana and are terminal in the West Indian species. In connection with his original description of the genus Cyclopeltis ?° Smith remarked that these characters were inadvertently reversed and that he meant to say that the sori of the West Indian species are lateral and of his L. presliana terminal. This cannot be taken as an adequate description of a new species, especially as the character is untrue. The sori are dorsal on the venules of both species. 70 Bot. Mag. Curtis 72, Compend. 36. 1846. STUDIES OF FERN TYPES, I—MORTON 47 Although the species remained a nomen nudum, the name Cyclo- peltis presliana (J. Smith) Berkeley has remained in general use, and if there were no competing name it could be accepted and dated from a later time. However, Hemicardion cumingianum Fée (1852) was validly described and has priority over any description of the species under the epithet presliana. This was recognized inferentially by Holttum2! in a comment under Cyclopeltis crenata (Fée) C. Chr. to the effect that it [C. crenata] “differs from both these species in its strongly toothed scales, and from C. presliana (which should more properly be called C. cumingiana Fée) in the lower pinnae not being gradually and evenly reduced to a small size.” However, this is an error, for there is no “C. cumingiana Fée,” i.e., “Oyclopeltis cumingiana Fée” but only Hemicardion cumingianum Fée. This is an inadvertent new combination, but it cannot be considered validly published, since there is no citation of the basionym, which is required for valid publication of a new combination after January 1, 1953. Therefore, the above new combination is required. 35. HyMENOPHYLLUM DELICATISSIMUM Fée, Crypt. Vasc. Brés. 2:86, t. 105, f. 1. 1872-73=H. elegans Spreng. in L. Syst. Veg. ed. 16, 4:133. 1827. Typr: Serra dos Orgios, Brazil, Glaziou 3591 (holotype P, Morton photograph 4595). In my revision, through typographical errors, the page of publication was erroneously given as 83 instead of 86, and the type number as 3491 instead of 3591. 36. HyMENOPHYLLUM ELEGANTISSIMUM Fée, Mém. Foug. 11:118, t. 29, f. 2. 1866=H. lineare (Swartz) Swartz, Journ. Bot. Schrad. 1800 (2):100. 1801. Tyre: Guadeloupe, L’Herminier (probable holotype P, 2 sheets, Morton photographs 4572, 4573). In my revision of the American species of the section Sphaerocioni- um 2 I placed this species as a synonym of H. hirsutum (L.) Swartz (H. ciliatum Swartz) going by the original description and figure only. However, the two sheets at Paris that are labeled H. elegantissimum and which agree well enough with the description and illustration of Fée are clearly H. lineare (Swartz) Swartz, or at least the Guadeloupe form of this species, which is unusually large and delicate. The only sheet at Stockholm that could be the type of Trichomanes lineare Swartz does not bear the name in Swartz’ hand, but it presumably is the holotype, from Jamaica, collected by Swartz (Morton photograph 6191). A duplicate agreeing with this specimen is in the Willdenow 21 Ferns of Malaya 527. 1954. 22 Contr. U.S. Nat. Herb. 29:139-201. 1947. 48 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM Herbarium (no. 20221) at Berlin, a fragment sent to Willdenow by Swartz (photograph by Tryon, US). 37. Hymenophyllum fragile var. venustum (Desv.) Morton, Contr. U.S. Nat. Herb. 29:173. 1947. In making this new combination for a rather common south Brazilian variant of the widespread HZ. fragile (Hedwig) Morton, I indicated that the basionym Hymenophyllum venustum Desv. (Mém. Soe. Linn. Paris 6:332. 1827) was a new name for H. hirsutum sensu Raddi, not Swartz, and thus that the type was from Mandiocca, Brazil, Raddi. It is true that Desvaux in describing his new species H. venustum cited “Hymenophyll. hirsutum Raddi, Syn. fil. bras., p. 19. Exel. syn.” but it is extremely unlikely that he saw any specimen collected by Raddi. On the other hand, a specimen in the Desvaux Herbarium at Paris labeled H. venustum in Desvaux’s hand is undoub tedly the actual specimen used in drawing up the description. This is the holotype (Morton photograph 4587). 38. Hymenophyllum producens Fée, Crypt. Vasc. Brés. 1:196, t. 71, f. 4. 1869. Type: Serra dos Orgaos, Brazil, Glaziou 3349 (holotype P, Morton photograph 4625). In my revision * I placed this species as a doubtful synonym of H. valvatum Hook. & Grev., following Christensen, but with a good deal of doubt, correctly as it now appears. The type shows that this species does not belong in the section Sphaerocionium but is a species of the section Mecodium, and is one of the allies of H. polyanthos (Swartz) Swartz. Its proper disposition must await a study of the Brazilian species of this difficult group. 39. Hymenophyllum silveirae Christ, in Schwacke, Pl. Nov. Minciras 2:14. 1900. Synryre: Itacolumi, Brazil, Schwacke 12528 (P, Morton photograph 4619). In my revision of Hymenophyllum section Sphaerocionium I recog- nized this species as distinct, going on Christ’s description of the fronds as caespitose, although I expressed some doubt as to the correctness of this character, which is hardly to be expected in the genus Hymen- ophyllum. This syntype shows that the fronds are definitely not caespitose but are scattered on a filiform rhizome as in related species. The fronds are very small and probably depauperate. I judge that they represent depauperate plants of H. pulchellum Schlecht. & Cham. The only character that might be distinctive is that the plants of FH. silveirae are said to be terrestrial, whereas those of HH. pulchellum 73 Contr. U.S. Nat. Herb. 29:161. 1947. STUDIES OF FERN TYPES, I—MORTON 49 are epiphytic, so far as known. But if this character is important or uniformly true remains to be determined. 40. Loncnrris JAVANICA Desr. in Lam. Encycl. Méth. 3:594. 1789=Blotiella javanica (Desr.) Morton, comb. nov. Lonchitis pubescens Willd. ex Kaulf. Enum, Fil. 195. 1824. The type of L. javanica, as the name indicates, was supposed to have been collected in Java by Commerson. In the Index Filicum, Christensen placed it as doubtfully being the same as the later L. pubescens, a species of Madagascar, Réunion, Mauritius, and the Seychelles Islands, the doubt occasioned by the locality. No species of Lonchitis has been known from Java and is not to be expected there. There is no specimen in Paris of L. javanica from Java, but there is a specimen in the Jussieu Herbarium, Cat. no. 1286 A, labeled L. javanica, in an old hand. The smaller specimen mounted on the sheet agrees in its dimensions and characters with the original de- scription by Desrousseaux and is doubtless the holotype. It is, as might be expected, not from Java but from “Tle de Bourbon,” iLe., Réunion, collected by Commerson. It is a small specimen of the species later described as L. pubescens Willd., as suspected by Christensen. The type of Lonchitis L. has often been considered to be L. aurita L., as for instance by Christensen, but Tryon * has shown that this is not a suitable species, because the application of the name is un- certain. Linnaeus did not have a specimen of L. aurita but based the species on Plumier Tract. Fil. ¢. 17, an illustration that does not correspond with any plant known from Martinique, where Plumier obtained his material. The illustration shows a plant with the vena- tion and general appearance of Ptervs subgenus Litobrochia, but no species from Martinique has such broad and blunt segments. Still, it may be conjectured that Plumier did have a species of Pterts in hand, for Lonchitis as generally delimited does not grow in Martinique. I cannot believe that Tryon is right in thinking that t. /7 represents L. hirsuta L. in part. This plant was also known to Plumier, who illustrated it very well, for him, in ¢. 20. The illustration and de- scription of the veins as reticulate removes L. hirsuta from considera- tion, as well as the illustration and description of the stipe as bearing “soft, black spines,” which are presumably scales with indurated bases. Such scales have been attributed to some Pteris species of the Lesser Antilles, such as P. aculeata Swartz, but L. hirsuta does not have any scales, only hairs. Tryon thinks that the description of L. aurita was based on a mixture of L. hirsuta and some unidentified species of Pteris, but I do not see any decisive evidence of this. For 2 Contr. Gray Herb. 191:93-100. 1962. 50 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM this reason I think that Lonchitis aurita L. must be listed among the dubious species of Pteris. The genus usually called Lonchitis must therefore be called Blotiella Tryon (type: Lonchitis glabra Bory), and the small genus Anisosorus Trev. becomes a synonym of Lonchitis L. (type L. hirsuta L.). 41. Meniscium Arrine Presl ex Ettingsh. Denkschr. Akad. Wiss. Math. Naturw. (Wien) 23:94, t. 13, f. 3. 1864; Farnkr. Jetztw. 170, fig. 73, t. 135, f. 16. 1865= Thelypteris affinis (Presl) Morton, comb. nov. Dryopteris dispar Maxon & Morton, Bull. Torr. Bot. Club 65:364. 1938. Based on Meniscium affine Presl, non Dryopteris affinis Newm. (1854). Type: Brazil, collector unknown (not seen). I have recently * considered Meniscium a section of T, helypteris, which necessitates a number of new combinations, which are proposed in this paper. Several of the most common species have already been transferred to Thelypteris, i.e., T. serrata (Cav.) Alston (1932), T. angustifolia (Willd.) Proctor (1953), 7. reticulata (L.) Proctor (1953), and 7. salzmannii (Fée) Morton (1960). Under Dryopteris several of the old specific epithets were preoccupied and not available, but these may now be used under Thelypteris. 42. MENISCIUM ANDREANUM Sodiro, Recens. 71. 1883; Crypt. Vase. Quit. 392. 1893=Thelypteris andreana (Sodiro) Morton, comb. nov. Dryopteris andreana C. Chr. Ind. Fil. 252. 1905. Typr: Rio Toachf, near Santo Domingo, Ecuador, Sodiro (not seen). 43. MENISCIUM ARBORESCENS Humb. & Bonpl. ex Willd. Sp. Pl. 5:133. 1810= Thelypteris arborescens (Humb. & Bonpl.) Morton, comb. nov. Phegopteris arborescens Mett. Fil. Lechl. 2:24. 1859. Phegopteris mollis Mett. Ann. Sci. Nat. V, Bot. 2:242, 1864. Type: Llano de San Martin, Paraiso, Colombia, alt. 300 m., Triana (isotype US). Nephrodium sorbifolium var. molle Hicron. Bot. Jahrb. Engler 34:449. 1904. Nephrodium sorbifolium f. angustipinnatum Hieron. Bot. Jahrb. Engler 34:449. 1904. Type: Alto de las Cruces, near Cali, Colombia, alt. 1100 m., Lehmann 2927 (isotype US). Dryopteris sorbifolia var. mollis Hieron. Hedwigia 46:351. 1907. Dryopteris reticulata var. arborescens Brause, Verh. Bot. Ver. Brandenb. 51:2. 1910. Dryopteris permollis Maxon & Morton, Bull. Torr. Bot. Club. 65:372. 1938. Based on Phegopteris mollis Mett., non Dryopteris mollis (Jaeq.) Hieron. (1907). Thelypteris mollis Tryon, Rhodora 69:7. 1967. When Dr. Maxon and I prepared our revision of Dryopteris subgenus Meniscium in 1938 we had not seen the type of Meniscium arborescens Humb. & Bonpl. We placed it with a query asa synonym of D. host- mannii (Klotzsch) Maxon & Morton and were not greatly concerned * Amer. Fern Journ. 53:154. 1963 [1964]. STUDIES OF FERN TYPES, I—MORTON 51 regarding its identity, since the specific epithet could not be taken up under Dryopteris, because there already existed a different D. ar- borescens (Baker) Kuntze, an entirely unrelated species from Samoa. However, under the generic name Thelypteris the epithet arborescens is not preoccupied. I have recently seen a photograph by Dr. Tryon of the holotype of M. arborescens from the Willdenow Herbarium in Berlin (sheet no. 19576); the type came from the region of the Mission of Santa Cruz, Venezuela.” The photograph matches closely a col- lection from Vegas del Rio de El Cantafio, State of Aragua, Venezuela, April 25, 1937, made by H. Pittier (no. 14003), which is the species that Maxon and I called Dryopteris permollis, rather than D. host- mannii. One of the chief features rendering Meniscium arborescens a questionable species was the description by Humboldt and Bonpland of the plants being arborescent and having a trunk 6-feet high, which is unlike any known Meniscium. This must have been an error, very likely a confused memory. The type specimen is a detached frond, without a rhizome, but doubtless the rhizome is subterranean and short-creeping like other specimens of D. permollis and other Menisciums. 44. MeNrscium curysopioiwEs Fée, Gen. Fil. 225. 1852=Thelypteris chryso- dioides (Fée) Morton, comb. nov. Dryopteris chrysodioides Maxon & Morton, Bull. Torr. Bot. Club 65:373. 1938. Typs: “Habitat in America australi., Collect. Pamplin., in Herb. cl. Moug., no. 55” (not seen; possibly at Oxford University). The following variety may also be transferred: Dryopteris chryso- dioides var. goyazensis Maxon & Morton, Bull. Torr. Bot. Club 65:374. 1938—Thelypteris chrysodioides var. goyazensis (Maxon & Morton) Morton, comb. nov. Type: Rio Corumba, Goyaz, Brazil, Glaziou 22631 (holotype NY). 45. Meniscrum FALcATUM Liebm. Dansk. Vid. Selsk. Skrift. V, 1:183. 1849 = Thelypteris falcata (Liebm.) Tryon, Rhodora 69:6. 1967. Meniscium jurgensenii Fée, Gen. Fil. 223. 1852 (as “Jungersenii’). Type: Mexico, Jurgensen 917 (not seen). Dryopteris falcata C. Chr. Dansk. Vid. Selsk. Skrift. VII. Naturv. Afd. 10(2):270. 1913, non Kuntze (1891). Dryopteris jurgensenit Maxon & Morton, Bull. Torr. Bot. Club 65:360. 1938. Typr: Lacoba, Chinantla, Puebla, Mexico, Liebmann 2756 (holotype C!, iso- types K, US). 46. MENISCIUM GIGANTEUM Mett. Fil. Lechl. 1:19. 1856=Thelypteris gigantea (Mett.) Morton, comb. nov. %6 See Sandwith on the Venezuelan localities of Humboldt & Bonpland, Kew Bull. Misc. Inf. 1925:301. 1925. 52 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM Dryopteris gigantea C. Chr. Ind. Fil. 267. 1905, non Kuntze (1891). Dryopteris simplicifrons C. Chr. Ind. Fil. 486. 1906. Based on Meniscium giganteum Mett. Type: San Gavan, Peru, Lechler 2292 (isotype K). 47. MENISCIUM LONGIFOLIUM Desv. Mém. Soe. Linn. Paris 6:223. 1827 = Thely- pteris longifolia (Desv.) Tryon, Rhodora 69:7. 1967. Dryopteris desvauxit Maxon & Morton, Bull. Torr. Bot. Club 65:369. 1938. Based on Meniscium longifolium Desv., non Dryopteris longifolia (Fée) Hieron. Tyre: Brazil, without further locality or collector (P, photograph of holotype US). It is not necessary to repeat here the intricate nomenclatural situation of this species when it is referred to the genus Dryopteris. When it is placed in Thelypteris, there is no problem, for the earliest specific epithet longifolia can be adopted without question. The following form may also be transferred: Dryopteris desvauaii f. glandulosa Maxon & Morton, Bull. Torr. Bot. Club 65:372. 1938. Type: Morro das Pedras, Sao Paulo, Brazil, Brade 5753 (holotype NY)=Thelypteris longifolia (Desv.) Tryon f. glandulosa (Maxon & Morton) Morton, comb. nov. 48. MERTENSIA SQUAMULOSA Desv., Journ. de Bot. Appliqué 1:268. 1813= Gleichenia squamulosa (Desv.) Moore, Ind. Fil. 383. 1862. This species has generally been considered dubious, even by Poiret, who probably had material available for study. The original descrip- tion, although not so very brief, does not mention the really distinctive characters, and the locality is stated only generally as South America. It is as follows: “Stipite ramisque angulato squamoso; frondibus pinnatis pinnis lanceolato-acutis, pinnulis linearibus oblongis, extimis confluentibus. Habitat in America australi.” The holotype, in the Muséum National d’Histoire Naturelle, Paris, in the Desvaux Herbarium is a good specimen. It was studied by Weatherby, who reported” that it was apparently the same as G. pedalis (Kaulf.) Sprengel,’ a well-known Chilean species. Weath- erby quoted the locality as “America australi (Termae Chili),”’ but this was a misreading of the label, which really reads ‘America australi (in Peruvia Chilen.).” Although the Desvaux epithet has priority, Weatherby hesitated to displace the name G. pedalis without further study. I can now verify Weatherby’s determination, and so there seems now no reason not to adopt the prior name G. squamulosa (Desvaux) Moore. The holotype has been photographed (Morton 4534). *7 Contr. Gray Herb. 114:27. 1936. 8 In L. Syst. Veg. ed. 16, 4:26. 1827, based on Mertensia pedalis Kaulf. Enum. Fil. 39. 1824, STUDIES OF FERN TYPES, I—MORTON 53 The confusion as to the locality undoubtedly goes back to the original collector. There is another sheet in Paris (Morton photo- graph 4535) which 1s identical with the holotype, and which bears the label Chile, Dombey. Another collection at Paris (Herb. Bory, Morton photograph 4492) is also the same, and this one gives the locality as Peru (no collector stated). This is not the only instance in which Dombey’s localities are confused between Peru and Chile. This collection marked as from “Peru” is identical with Chilean specimens of @. pedalis and unlike any Peruvian species. 49. Neorrorrenis stiritata J. Smith, Cat. Cult. Ferns 49. 1857=Asplenium stipitatum (J. Smith) J. Smith, Ferns Brit. & For. 210. 1866. Asplenium squamulatum Blume var. smithii Hook. Sp. Fil. 3:83. 1860. Based on Neottopteris stipitata J. Smith (1857). Asplenium robinsonii F. von Mucell. Journ. Bot. Brit. & For. 22:289. 1884. Type: Norfolk Island, Robinson in 1884 (holotype BM, Morton photograph 7202, isotype K, Morton photograph 8010). Type: Cultivated in the Royal Botanic Gardens, Kew (holotype from the J. Smith Herbarium, BM, Morton photograph 7203). The original Neottopteris stipitata J. Smith in Hook. Journ. Bot. 3-409. 1841, was a nomen nudum, based on Cuming 195 from Cama- rines Sur (Luzon, Philippine Islands]. A plant at the Royal Botanic Gardens, Kew, was cultivated under this name for a long time, and was supposed to have been grown from spores from the Cuming collection. However, the Cuming collection is Asplenium squamu- latum Blume whereas the cultivated plant is different. Although much less extreme in its irregularity it appears to agree with the type of Asplenium robinson F. von Muell., and it was therefore very likely introduced into Kew from Norfolk Island, and not from the Philippine Islands. 50. Nepuroprum xuntum Desvaux, Mém, Soc. Linn. Paris 6:258. 1827=The- lypteris kunthii (Desv.) Morton, comb. nov. Dryopteris normalis C. Chr, Ark. for Bot. 9" 31. 1910. Thelypteris normalis Moxley, Bull. So. Calif. Acad. Sci. 19:57. 1920. Type: Cumanacoa, Venezuela, Humboldt & Bonpland (lectotype P). The complete original description of Desvaux was: “59, N[ephrodium]. kunthii N. Aspid{ium]. patens Kunth, in Humb. et Bonpl., Nov. gen. 1, p. 18. Exel. syn.” This is not a good way to publish a new species, but it is a possible and valid way, by reference to a description under another name previously, effectively, and validly published. There is a description in H.B.K., Nov. Gen. et Sp. 1:13. 1815, and this work is certainly effectively and validly published. Therefore the description of Nephrodium kunthii is the one given in H.B.K. and the type is the 54 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM specimen or specimens on which that was based, namely Humboldt and Bonpland specimens collected “prope Cumanacoa, Guanaguana et Caripe” [Venezuela]. The synonyms to be excluded were Aspidium patens Swartz and Polypodium patens Swartz. Desvaux properly transferred the latter to Nephrodium as N. patens (Swartz) Desy. (Mém. Soc. Linn. Paris 6:258. 1827). A specimen in the Desvaux Herbarium at Paris (Morton photo- graph 4424) has the name N. kunthti Desv. and the synonym Aspidium patens Kunth in the hand of Desvaux. The locality is given as “Habitat ad Caracas.” It is a single pinna, which was very likely removed by Desvaux from a Humboldt and Bonpland collection. The locality “Caracas” to older authors often signified the whole of modern Venezuela. However, this fragment cannot be considered the type of the species, for the description of Kunth was not based on it. In the Humboldt Herbarium at Paris there is a specimen collected by Humboldt and Bonpland labeled Aspidium patens and which was collected at Cumanacoa, Caripe. This is undoubtedly the plant or one of the plants on which the description of Aspidium patens was drawn by Kunth, and thus surely a syntype. J designate this speci- men as lectotype of Nephrodium kunthti Desy. This is a fairly good specimen; I did not photograph it but I studied it and made a note about it. Another specimen in the general herbarium at Paris from Caripe, Herb. Bonpland, and labeled Aspidium patens (Morton photograph 4425) is another syntype. These three specimens, the fragment in the Desvaux Herbarium, the one in the Humboldt Herbarium, and the one in the general herbarium are all the same. Desvaux was right; they are not Aspidium patens Swartz [Thelypteris patens (Swartz) Small] but the nearly related species recognized and described by Christensen as Dryopteris normalis, which is common in the southern United States and the West Indies and which occurs also scattered on the Continent from Mexico south probably to Bolivia and Brazil. It is unfortunate that the well-known name normalis should disappear, but there is no help for it. For another synonym, see under Aspidium germanii Fée (p. 36). 51. PHEGOPTERIS CANESCENS Mett. Abh. Senck. Naturf. Ges, 2:314. 1858= Ctenitis canescens (Mett.) Morton, comb. nov. Polypodium canescens Kunze ex Mett. Abh. Senck. Naturf. Ges, 2:314. 1858, pro syn.; Hook. Sp. Fil. 4:262. 1862 (non Blume, 1828), Polypodium blanchetianum Kunze ex Mett. Abh. Senck. Naturf, Ges. 2:314. 1858, pro syn. Dryopteris blanchetiana Hieron. Hedwigia 46:344. 1907, Ctenitis blanchetiana Copel. Gen. Fil. 124, 1947. Illegit. The type and only specimen originally cited by Mettenius was Morieand 2454, from Bahia, Brazil, which actually should be Blanchet STUDIES OF FERN TYPES, I—MORTON 55 2454 (Herb. Moricand). The original Phegopteris canescens Mett. is entirely legitimate but it could not be transferred to Dryopteris, because of the existence of the prior and different Dryopteris canescens (Blume) C. Chr. However, there is no obstacle to the use of the epithet canescens under the genus Ctenitis. In transferrmg the epithet blanchetiana in preference, Copeland was routinely trans- ferring names as he usually did without checking the synonymy or the availability of prior epithets. 52. PHEGOPTERIS MEMBRANACEA Mett. Fil. Lechl. 2:22. 1859=Thelypteris membranacea (Mett.) Tryon, Rhodora 69:7. 1967. Nephrodium lechleri Hieron, Bot. Jahrb. Engler 34:448. 1904. Syntypes: Azangaro, Peru, Lechler 1785 and San Gavan, Peru, Lechler 2321. The former is here chosen as lectotype. Hieronymus based his new species on the same two specimens that were the original syntypes of Phegopteris membranacea Mett., and thus his species is identical with P. membranacea and the epithet lechleri superfluous. As pointed out by Maxon & Morton in 1938, Hieronymus misapplied the name membranacea to an entirely different plant that formed no part of the original concept of the species. Dryopteris membranacea C. Chr. Ind. Fil. Suppl. 1:35. 1913. Lecrorype: Azangaro, Peru, Lechler 1785 (K) (chosen by Maxon & Morton, Bull. Torr. Bot. Club 65:366. 1938). 53. PHEGOPTERIS NICARAGUENSIS Fournier, Bull. Soc. Bot, France 19:252. 1872=Thelypteris nicaraguensis (Fournier) Morton, comb. nov. Dryopteris nicaraguensis C. Chr. Ind, Fil. 279. 1905. The type came from Chontales, Nicaragua, Levy 460 bis. A frag- ment is in the Christensen Herbarium at the British Museum (Morton photograph 6374). This is one of the commonest species of the section Goniopteris in Central America. 54. PoLtyrpopIuM ADIANTOIDES Aublet, Hist. Pl. Guiane 2:962. 1775=Polybotrya caudata Kunze. This species has remained unknown. The excessively brief original description consisted only of “pinnis auriculatis.” The type specimen is in the British Museum (Morton photograph 6626). It is Polybotrya caudata Kunze. Fortunately, the specific epithet of Aublet, which is not really especially appropriate, cannot be taken up, since Aublet’s species is a later homonym of the different Polypodium adianthoides Burm. (1768), which appears to be also a dubious species at the present time. Very likely Burmann’s type is in Geneva. 55. Pouypopium compronirotium Desv. Mag. Naturf. Freund. Berlin 5:316. 1811 [as “comptoniaefolium’’]|=Grammitis trifurcata (L.) Copel. Gen. Fil. 211. 1947 (Polypodium trifurcatum L.). Polypodium comptonioides Desv. Mém. Soc. Linn, Paris 6:231. 1827. An illegitimate change of name for P. comptontifolium. Tyrer: Bourbon [Réunion], Commerson (holotype P, Jussieu Herb. Cat. 1095, Morton photograph 2947). 56 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM Although Desvaux cited the locality of his P. comptoniifolium as Bourbon, and the label gives the same data, this must be an error, for no species remotely like this has been found since in Bourbon, the present-day Réunion Island, but the type is quite typical of the West Indian Grammitis trifurcata. Madame Tardieu must have come to the same conclusion, for she does not mention P. comptoniifolium in her treatment of the species of Grammitis in Réunion and the other islands of the Madagascarian region.” 56, Potyroprum consueatum Poir. in Lam. Encycl. Méth. 5:516. 1804=Poly- podium phyllitidis L. (forma). Typz: ‘Cette plante est originaire des Indes. (V. 8. in Herb. Jussieu).”” The specimen mentioned in the Jussieu Herbarium (Cat. 1071) (Morton photograph 2934) is indicated as from ‘‘ Amerique meridionale donné par M. Houston.” The original ticket with the name Polypodium conjugatum is in the hand of Poiret, and the specimen agrees with the description. It has apparently been assumed that this is an Old World species, and although it was based on a single specimen it was inexplicably referred by Christensen in the Index Filicum to both Polypodium phymatodes and Drynaria quercifolia. But as the label indicates, this is actually an American plant received from William Houstoun, very likely one of his own collections from Mexico, Cuba, or Jamaica. The specimen is a teratological (forked and variously lobed) specimen of the common tropical American species Polypodium phyllitidis 1. 57. PoLypoprumM corpIroLium Mart. & Gal. Mem. Acad. Brux. 15:31, t. 4, f. 2. 1842, non L., 1753=Tectaria heracleifolia (Willd.) Underw. Bull. Torr. Bot. Club 33:200. 1906. Type: Antigua, Veracruz, Mexico, June—October 1840, Galeotts 6313 (BR, holotype, Morton photograph 5175), In the Index Filicum, this species is referred to Aspidium trifoliatum L., ie., Tectaria trifoliata (L.) Cav. The type shows that it is a juvenile specimen of T. heracleifolia with simple unlobed blades. Although juvenile, it is fertile, a fact that doubtless persuaded Martens and Galeotti to consider it different. However, such juveniles are not at all uncommon in this species. Polypodium cordifolium L. is an entirely different thing, i.e., Nephrolepis cordifolia (L.) Presl. 58. PoLypopiuM ExPANSUM Poir. in Lam. Encyel. Méth. 5:523. 1804=Tectaria incisa Cav. Type: Based on a specimen from America in the Lamarck Herbarium, Paris. This was referred by Christensen in the Index Filicum to Aspidium martinicense Spreng., with a query. I have seen the type, which is marked “No. 127. Poly- podium, unique. Polypodium expansum Dict.” (Morton photograph 2664). 20 “Les ‘Grammitis’ de la région Malgache,” Notul. Syst. [Paris] 15:421-425, 1959. STUDIES OF FERN TYPES, I—MORTON 57 The type is the upper part of a frond of typical (glabrous) Tectaria incisa Cav., of which Aspidium martinicense is a synonym (Cf. C. Christensen, Dansk Bot. Ark. 9(3):14. 1937). There is another specimen in the Lamarck Herbarium (Morton photograph 2665) which is also labeled “Polypodium expansum Dict.” It was collected (or received) from Sonnerat. There is some error here, because this is an utterly different plant, which does not agree at all with the description of Polypodium expansum Poir. Although I have not studied it closely, it is, I believe, Dryopteris varia (L.) Kuntze, and must have come from Asia. 59. PoLYPoODIUM FLABELLIFORME Poir. in Lam. Encycl. Méth. 5:519. 1804=Gram- mitis flabelliformis (Poir.) Morton, comb. nov. Polypodium rigescens Bory ex Willd. in L. Sp. Pl. ed. 4, 5:183. 1810. Type: Bourbon (Réunion), Bory (presumably in Herb. Willd., B). This was originally published with the diagnosis ‘“Polypodium fronde angustissima, elongata, pendula; lobis alternis, obtusis. (N)” and the citation of the illustrations “Polypodium aliud pendulum, minimum. Plum. Fil. pag. 68, tab. 87,” and ‘“Polypodium lonchitidis folio, angustissima, pendulum. Petiv. Fil. tab. 10, fig. 1.” A variety is described: “A. Idem, fronde breviori, lobis longioribus.N” The citation regarding the type is: “Cette plante croft dans la Martinique, sur le trone des vieux arbres, d’ow ses feuilles pendent vers la terre. (V. s. in herb. Juss. & Desfont.).” Thus, it is seen that the species was based on literature references to Plumier and Petiver and on dried specimens in the Jussieu and Desfontaines Herbaria. In such cases there is no doubt that the actual specimens studied should be regarded as the types rather than the figures cited, which were based on material not seen by Poiret.*° In the Jussieu Herbarium at Paris there is only one specimen that could be the type. It is labeled ‘“Polyp. flabelliforme Poir. Dict.” in the hand of Poiret; it also bears the references to the illustrations of Plumier and Petiver (Morton photograph 2951). This is surely the specimen mentioned in the original description from the Jussieu Herbarium. This sheet is from the island of Bourbon (Réunion) in the Indian Ocean, from the Herbarium of Commerson (which usually means actually collected by Commerson). It is surely the species subsequently described as Polypodium rigescens Bory. Very likely the locality “Martinique” in the original description came only from Plumier. Poiret was in error in identifying this Commerson collection with the Plumier plate, for the two are evidently different. 30 The Plumier illustration quoted is the left-hand plant on the plate, the identity of which is uncertain, but it is one of the small species of the “« Xiphopteris” group P, hartii, P. knowltoniorum, P. serricula, or P. taentfolium, perhaps the latter. 58 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM The name Polypodium flabelliforme Poir. was erroneously applied in the Index Filicum to a small, delicate species of the Lesser Antilles, especially Guadeloupe, the proper name of which is still uncertain. Mr. George Proctor gave me a manuscript for criticism some time ago (about 1960) in which he attempted to show that the name Polypodium suspensum L. applied to this species. The basis of P. suspensum is the larger, right-hand plant of Plumier’s plate 87. I could not agree and wrote to Mr. Proctor (Jan. 30, 1961) as follows: . L agree that suspensum cannot be a synonym of asplenifolium or a near ally. As you say, it must be searched for among the essentially glabrous species. How- ever, it does not seem to me that it really can be flabelliforme, and I don’t think that the Grenada specimen labelled tovarense, of which you have a photograph, can be “flabelliforme’”’ either; the segments are of a different shape, acute rather than rounded and fan-shaped (jlabelliforme). I won’t presume to identify the Grenada specimen from the photograph. But it does seem to me that pl. 87 of Plumier, the type of suspensum, can be jubiforme. You eliminate this from con- sideration by saying that it has “narrower, nearly exstipitate, downwardly attenuate fronds and relatively narrow, oblong segments,” but I can’t quite agree. Sometimes the segments are narrow and oblong in jubiforme, but this is a variable species and some specimens have the segments relatively broad at base. The illustration does show the blade somewhat reduced at base and with only short stipes. Plumier’s illustrations are notoriously inaccurate, and I think that this is as good as might be expected for jubiforme. In any case, “flabelliforme’’ is also downwardly reduced and nearly exstipitate. Plumier’s description and also his illustration says that the segments have the shape of the front of ashoe or a foot, i.e., a high instep curving down to a narrow toe, in other words the segments are broad at base, acute at apex and curved on the proximal side. This is not at all true of “flabelliforme’”’ but is more or less true of jubiforme—not all specimens and not all segments, but some. You disregard the disposition of the sori near the apex of the segments, although this is particularly mentioned in the Plumier description and is shown in the illustration. This is characteristic of jubiforme and not of ‘‘flabelliforme.” Finally, jubiforme is common in Martinique; all collectors have found it, and it would be odd if Plumier had not. The only illustration of Plumier that could be it is this pl. 87. To sum up, it appears to me that there is good reason to believe that suspensum is the same as jubiforme and the earliest name. The alternative would be to regard suspensum as a “sp. dub.” and allow it to remain unused for anything. The following year Mr. Proctor wrote* on the identification of Polypodium suspensum L.., claiming the identification with P. qubiforme as his own, and stating that his previous manuscript identifying the species with P. flabelliforme had been written with “private misgivings,” although certainly none were expressed to me at the time. He further misquotes me as saying “that if the identity of the Linnaean P. suspensum could not be clearly settled, it might be better to treat it as a ‘nomen ambiguum’ than to displace a well-established name,” but as indicated above in the quotation from my letter I said nothing of the sort. On the contrary I said that there is good reason to believe 31 Brit. Fern Gazette 9:77. 1962. STUDIES OF FERN TYPES, I—MORTON 59 that suspensum is the same as jubiforme and the earliest name for that species. Mr. Proctor also stated in his note that the type of P. flabelliforme represented a different species from Mauritius, without indicating me as the source of his information. I did write to him concerning the type of this species, identifying it with P. rigescens Bory, but I certainly did not say that it came from Mauritius; I told him that it came from Bourbon (Réunion), as it does. 60. PotyrpopiuM HOSTMANNII Klotzsch, Linnaea 20:397. 1847 = Thelypteris hostmannii (Klotzsch) Morton, comb. nov. Dryopteris hostmannit Maxon & Morton, Bull. Torr. Bot. Club 65:369, t. 14. 1938. Type: Surinam, Hostmann & Kappler 828 (holotype B!, isotypes K, NY). This is one of the species of Meniscitum, which I now treat as a section of Thelypteris. 61. PotypopiuM INvIsuM Swartz, Prodr. Veg. Ind. Occ. 133. 1788, non Forster, 1786= Thelypteris invisa (Swartz) Proctor. Aspidium invisum Swartz, Journ. Bot. Schrad. 1800?7:34. 1801. Considered to be a legitimate new name dating from 1801, and not a transfer of Poly- podium invisum Swartz, Based on the same type. Nephrodium invisum (Swartz) Desv. Mém. Soc. Linn. Paris 6:257. 1827. Based on Aspidium invisum Swartz. Lastrea invisa (Swartz) Presl, Tent. Pterid. 75. 1836. Based on Aspidium invisum Swartz. Nephrodium sloanei Baker, in Hook. & Bak. Syn. Fil., ed. 2.263. 1874. Based on Polypodium invisum Swartz, non Forster, but illegitimate, since the combination N. invisum (Swartz) Desv. was prior and correct. Also illegitimate, since a later homonym of N. sloanez Presl, 1825. Dryopteris sloanei Kuntze, Rev. Gen, Plant, 2:813. 1891. A legitimate name, considered not as a transfer of the illegitimate N. sloanet Baker, non Presl, but a new name, based on the same type, necessitated by the unavail- ability of the epithet invisa under Dryopteris because of the different species Dryopteris invisa (Forster) Kuntze. Dryopteris oligophylla Maxon, Contr. U.S, Nat. Herb. 10:489. 1908. Based on Polypodium invisum Swartz, non Forster. Iegitimate, since superflu- ous, the name Dryopteris sloanet Kuntze being nomenclaturally synony- mous, legitimate, and available. Thelypteris oligophylla (Maxon) Proctor, Bull. Inst. Jam, Sci. Ser., 5:62. 1953. Based on Dryopteris oligophylla Maxon. Illegitimate, since the earliest available specific epithet was not adopted. All the epithets Aspidium invisum Swartz, Lastrea abrupta Presl, and Dryopteris sloanet Kuntze were available and prior to Dryopteris oligophylla. Thelypteris invisa (Swartz) Proctor, Rhodora 61:306. 1959. To be con- sidered as based on Aspidium invisum Swartz. Typr: Jamaica, Swartz. In 1786, Forster described one of his collections from Tahiti as Polypodium invisum Forster, and two years later Swartz, doubtless unaware of Forster’s publication, described one of his own collections 232-200—67——3 60 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM from Jamaica as Polypodium invisum Swartz. Strangely enough, by a coincidence these two species from widely separated regions are rather similar, both habitally and taxonomically, belonging to Dryop- teris subgenus Cyclosorus of Christensen. Exactly why both Forster and Swartz considered these innocuous ferns “invisum,” i.e., hateful or detested, is debatable. Still, later pteridologists who have found it necessary to classify the numerous, close, and impossibly variable species of this group, such as D. patens, D. normalis, D. feei, D. auges- cens, D, wnita, D. arida, D. feroz, and so on can perhaps make a cuess. Forster’s species, being prior, has properly continued to retain the epithet invisum, being usually known as Dryopteris invisa (Forster) Kuntze. I have examined the type in the British Museum (Natural History) and found that it agrees with such Tahiti specimens as Setchell & Parks 9 and 16, and Wilkes Expedition 7 in the U.S. National Herbarium. Only the lowest pair of veins is truly united in the leaf tissue, the second pair connivent to the sinus. Indusia are present and are long-pilose. The sporangia bear one or occasionally two hairs on the lateral faces. The upper surface is glabrous except along the midrib, but the lower is pilosulous on the costules and veins. The pinnae are only slightly cut, about one-third the way to the mid- rib or less. By these characters this species may be distinguished from unita and allied species. It is found also in the Fiji Islands (Wilkes Expedition) and elsewhere in Polynesia and Melanesia, but its exact range remains to be determined. According to my present generic concepts in this group, Polypodium invisum Forster belongs in Thelypteris, subgenus Cyclosorus.2 Un- fortunately, it cannot now be transferred to Thelypteris, because of the recent publication of the combination Thelypteris invisa (Swartz) Proctor for the West Indian species, an unwise action on the part of Mr. Proctor, for now the well-known Old World invisa must be re- named, and the name invisa transferred to a West Indian and tropical American species, which can only result in confusion. However, there is no help for it. Thelypteris forsteri Morton, nom. nov. Polypodium invisum Forster, Fl. Ins. Austr. Prodr, 81. 1786. Type: Tahiti, Forster (BM). Not Thelypteris invisa (Swartz) Proctor, 1959. Nephrodium invisum (Forster) Carruthers in Seem. Fl. Vit. 362. 1873, non Desv., 1827. Dryopteris invisa (Forster) Kuntze, Rev. Gen. Plant. 2:813. 1891. Aspidium invisum (Forster) Christ, Bot. Jahrb. Engler 23:350. 1896, non Swartz, 1801. Cyclosorus invisus (Forster) Copel. Gen. Fil. 142. 1947. 84 Amer, Fern. Journ. 53:153. 1963 [1 964]. STUDIES OF FERN TYPES, I—MORTON 61 The American species described originally by Swartz has had an unhappy nomenclatural history. In recent times commonly known as Dryopteris oligophylla Maxon, it is allied to D. normalis, D. augescens, D. serra, D. patens, and others; it is the giant of the group, well-grown mature plants often being two meters high, with pinnae 30 cm. long or more and 3 cm. wide. The lowest veins are connivent to the sinus but not actually united, for which reason Copeland placed the species in “Tastrea,” where it is wholly out of place. In its sum total of characters (aspect, texture, basally abrupt leaf-blades, persistent hairy indusia, and so forth), it is closely allied to D. dentata and related species, and thus belongs in subgenus Cyclosorus. This West Indian plant can be known as Thelypteris invisa, but not with the authority proposed by Proctor, who gave the basionym as Nephrodium invisum Desv. (1827). However, the first validation of the epithet invisum, by Art. 72 (Note) of the International Code of Botanical Nomenclature, was as Aspidiwm invisum Swartz (1801). Inasmuch as Desvaux cited Swartz’ name as the basis for his Nephro- dium invisum, Proctor’s new combination can be considered as valid, with the parenthetical authority changed from Desvaux to Swartz. The var. invisa, with the sori strictly medial, is restricted to the Greater Antilles. The closely related forms of the Continent, ranging from Mexico to Ecuador are referable to: Thelypteris invisa var. aequatorialis (C. Chr.) Morton, comb. nov. Dryopteris oligophylla Maxon var. aequatorialis C. Chr. Dansk. Vid. Selsk. Skrift. [Monogr. Dryopteris] VII, 10(2):189. 1913. Synrypes: Andes, Ecuador, Sodiro (C); Santa Jues, Rio Pastaza, Ecuador, Stuebel 871 (B); Niebli, Lehmann 5053 (B); Banos, Rio Pastaza, Ecuador, Spruce 5296 (Bonaparte Herb., P); Peru, Schenke in 1909 (Rosenst. Herb., 8); Bolivia, Bang 2312 (B, US). As lectotype I choose Spruce 5296 (P). Although similar of course to typical T. invisa, this seems to be rather distinct. The distinguishing character of T. invisa var. invisa is that the basal two or three segments of the lower pinnae are abortive, as well illustrated by Christensen (1913, f. 25). Variety aequatorialis has the basal segments smaller, but not aborted; in addition, it is sometimes a smaller plant, with fewer pairs of veins in the segments (10-12 pairs, as contrasted with 15 pairs or more in var. invisa), and it is much hairier; the veins, costules, and indusia are densely pilose, and some hairs are borne on the costules and veins of the upper surface, which are glabrous in var. invisa. In size and some other respects, var. aequatorialis resembles T. kunthw, which is lighter green in color, softer in texture, with fewer pairs of veins (5-9 pairs), and which has the bases of the lower pinnae somewhat enlarged rather than contracted. 232-200—67—4 62 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM Thelypteris invisa var. pallescens (C. Chr.) Morton, comb. nov. Dryopteris oligophylla var. pallescens C. Chr. Dansk. Vid. Selsk. Skrift VII, 10(2):188. 1913. A large number (19) of syntypes are cited. I do not have all these at hand. I choose Eggers 15037 from El Recreo, Ecuador, as lectotype (U.S. 831342, annotated by Christensen) The nomenclature has been complicated by a taxonomic problem caused by the fact that a closely allied plant was described from Peru as Aspidium abruptum Kunze, i.e., Dryopteris kunzeana (Hooker) C. Chr. Christensen stated that this certainly was not specifically distinct from D, oligophylla, and named it D. oligophylla var. kunzeana (Hooker) C. Chr. I have looked into the matter and I must say that T agree. Still, this South American plant can be recognized as va- rietally distinct. It has the sori slightly but perceptibly supramedial. Thelypteris invisa var. kunzeana (Hook.) Morton, comb. nov. Aspidium abruptum Kunze, Linnaea 9:93. 1834, non Blume, 1828. Type: Pampayacu, Peru, Poeppig in 1829. Lastrea abrupta Presl, Tent. Pterid. 75. 1836. By Art. 72 (Note) to be considered not as a transfer but as a new name, dating from 1836, for Aspidium abruptum Kunze, non Blume. Nephrodium kunzeanum Hooker, Spec. Fil. 4: 102. 1862. Based on Aspidium abruptum Kunze, non Blume, but illegitimate, since Hooker should have adopted the epithet abrupta Presl, which was available under Nephrodium. Nephrodium abruptum (Presl) Baker, Syn. Fil. 263. 1868. To be considered as based on Lastrea abrupta Presl. Dryopteris abrupta (Presl) Kuntze, Rev. Gen. Plant. 2:812. 1891. To be considered based on Lastrea abrupta Presl, and not a transfer of Aspidium abruptum Kunze. A correct name under the genus Dry- opteris. Dryopteris kunzeana (Hooker) C. Chr. Ind. Fil. 273. 1905. IRlegitimate. Dryopteris oligophylla Maxon var. kunzeana (Hook.) C. Chr. Dansk. Vid. Selsk. Skrift. VII, 10(2):189. 1913. I have not seen the type, collected by Poeppig, but the following collections agree with the original description and with Christensen’s characterization: Peru: Rio Marafion, below Rancho Indiana, Distr. Iquitos, Dept. of Loreto, Jan, 28, 1932, Mexia 6462. Lower Rio Nanay, Dept. Loreto, May-June 1929, Ll, Williams 379. Muna, May 23-June 4, 1923, Macbride 3991. La Merced, Dept. Junin, May 29-June 4, 1929, Killip & Smith 23542. Christensen cites also Spruce 4066, from Tarapoto, and Schenke 47, as well as a collection from Rio Balao, Ecuador (Eggers 14523). The South Brazilian Dryopteris oligophylla var. lutescens C. Chr.33 seems to me to be doubtfully distinguishable from var. kunzeana. It was based on five specimens: Minas Gerais, Mosen 2144, 2145; Sao 33 Dansk. Vid. Selsk. Skrift VII, 10(2):188. 1913, STUDIES OF FERN TYPES, I—MORTON 63 Paulo, Regnell III, 1448, Widgren; and Rio Grande do Sul, Juergens & Stier (Rosenst. Exs. 182). I do not designate a lectotype. 62. Potypopium NiGcRIPEs Hasskarl, Catalogus Plantarum in Horto Botanico Bogoriensi Cultarum Alter 4, 1844=Tectaria melanocaulis (Blume) Copeland. In Christensen’s Index Filicum, Polypodium nigripes Hassakarl is unplaced and considered a dubious species. In the Rijksherbarium, Leiden, there is a sheet labeled P. nigripes, with the number 1718 H.B., the “H.B.” standing for Hortus Bogoriensis, which I take to be a Hasskarl collection (Morton photograph 2324). Very likely the holotype is in the herbarium of the Hortus Bogoriensis, but this is presumably an isotype. It is a Tectaria, and bears an identification as Tectaria melanocaulis (Blume) Copeland by Rosenstock, apparently a correct determination. Another sheet in Leiden, also labeled P. nigripes, is also from the Hortus Bogoriensis, and bears the name “Pakoe tjaga’”’; it is a juvenile leaf, representing the same species Tectaria melanocaulis (Morton photograph 2327). Another specimen in Leiden is Zollinger 1626 (Morton photograph 2325), from “vom Wasserfall Tjikapundung bei Bandong in der Erde,’’ collected March 17, 1844. Since this is not from the garden at Bogor it cannot be a type, but may very well have been studied by Hasskarl and thus also be authentic; it represents the same species 7. melanocaulis. 63. PotypopIuM LEucAaTomMos Poir. in Lam. Encyel. Méth. 5:516. 1804= Polypodium aureum L. var. Types: Cayenne [French Guiana], LeBlond (holotype, Lamarck Herbarium P, Morton photograph 2684). There has been some doubt about this species, despite the fact that the type is readily available in the herbarium in Paris. In the Index Filicum, Christensen referred it to Polypodiwm aureum L. var., correctly as it turns out. The type is a large frond, lacking stipe and rhizome; the venation is regularly phlebodioid and the sori in two ranks on each side of the costae. In his ‘Plantae Stuebelianae” in 1909, Hieronymus * took up the name P. leucatomos Poiret [misspelled “leucotomos’’] as the correct name for the species long known as P. decumanum Willd., without explanation. There is no indication that he saw Poiret’s type, and it is unlikely that he did so. Polypodiwm decumanum is allied with P. aureum, but has the sori in four to seven rows on each side of the costae. Christensen * in the first supplement to the Index Filicum followed Hieronymus in taking up the name P. leucatomos in place of P. decumanum, erroneously as it now appears. 3 Hedwigia 48:267. 1909. 35 Ind. Fil. Suppl. 1:125. 1913. 64 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM It has appeared on herbarium labels and occasionally in publications, e.g., by Kramer, Meded. Bot. Mus. Utrecht 124:489. 1954. 64. PoLypopruM oprusmosum Desv. Ges. Naturf. Freund. Mag. Berlin 5:317. 1811=Ctenitis desvauxii Tardieu-Blot. Aspidium desvauzit Mett. ex Kuhn, Fil. Afr. 231. 1868. Renaming of Polypodium obtusilobum Desv. under Aspidium, the epithet obtusilobum not being available because of the prior and different Aspidium obtusilobum Fée (1857). Dryopteris obtusiloba (Desv.) C. Chr. Ind. Fil. 280. 1905. Ctenitis desvauxit Tardieu-Blot, Notul. Syst. 15:82. 1954. Proposed as a new species, with a Latin diagnosis. There is no mention of Polypodium obtusilobum Desv. or Aspidium desvauxii Mett., nor any mention of the Desvaux type. No type was indicated. Four specimens, all from Mauritius, were cited, of which the first may stand as lectotype: “Bois de lanse Courtois de la montagne du Corps de Garde, Boivin, aott, 1849.” The identification ‘“Polypodium thelypteroides Desv.?” appears on the label, evidently the reason for the choice of the specific epithet “desvauzit.” Ctenitis desvauzii (Mett. ex Kuhn) Tardieu-Blot, Notul. Syst. 16:181. 1960. Based on Aspidium desvauxii Kuhn and Polypodium obtusilobum Desv. Not validly published, because the basionym, although stated, is not fully cited with the place of publication (Code, Art. 33), and also illegitimate, since a later homonym of C, desvauzii Tardicu-Blot (1954). The type was said by Desvaux to be from Madagascar, but this was an error. The holotype is in the Jussieu Herbarium (P), Cat. no. 1115 (Morton photograph 2959), and it definitely says “Ile de France” ie., Mauritius, Herb. Commerson. This species is apparently known only from Mauritius and not from Madagascar. The epithet obtusilobum, although entirely validly published and the oldest, cannot be trans- ferred to the genus Ctenitis because of the different species Clenitis obtusiloba (Baker) Ching (Bull. Fan Mem. Inst. Biol. Bot. 8:296. 1938), based on a type from Ceylon. 65. PoLypopruM PELLITUM Willd. ex Kaulf. Enum. Fil. 89. 1824= Polypodium lycopodioides L. Sp. Pl. 1082. 1753. In the Index Filicum (1906) P. pellitum is left as a dubious species, and apparently no one has examined the type and published on its identity since that time. I have studied the type in the Willdenow Herbarium, Berlin; it is sheet no. 19604, collected in Brazil, by Commerson. It represents P. lycopodioides L. The blades lack scales, as they should in this species. 66. PoLypopiuM PENNATUM Poir. in Lam, Encycl. Méth. 5:535. 1804= Thelyp- teris pennata (Poir.) Morton, comb. nov. Polypodium megalodus Schkuhr, Kr. Gew. 1:24, t. 19b. 1806. Dryopteris megalodus Urban, Symb. Antill. 4:21. 1903. Thelypteris megalodus Proctor, Bull. Inst. Jamaica, Sci. Ser. 5:61. 1953. STUDIES OF FERN TYPES, I—MORTON 65 Type: “Amer. Merid.,”’ without collector (holotype P, Morton photograph 4704). Christensen, in his Monograph of the Genus Dryopteris, suggested that Polypodium pennatum Poir. was the earliest name for the species that he called D. megalodus (Schkuhr) Urban, but he hesitated to take up the name without studying the type. I have seen the type in Paris and it is definitely the same as megalodus, which is a distinctive species. There is an isotype in the Persoon Herbarium, Rijksherbarium, Leiden, with the label probably in the hand of Poiret (Morton, photograph 1195). The type locality, other than “Amer. merid.,” is unknown, but it is very likely West Indian. 67. Potypopium sERRATUM Aublet, Hist. Pl. Guian. 2:962. 1775=Bolbitis guianensis (Aubl.) Kramer, Acta Bot. Neerl. 3:486. 1954. Asplenium auritum Swartz var. acutum Mett. f. serratum (Aubl.) Mett. Abh. Senckenb, Naturf. Gesell. Frankfurt 3:147. 1858. Type: French Guiana, Aublet (holotype BM, Morton photograph 7464). Polypodium serratum Aublet (non alior.) has always been a dubious plant. Aublet’s original description is too brief: . . . fronde simplici; pinnis alternis, serratis. Lonchitis minor, pinnis latioribus, leviter denticulatis, superiori latere auricula- tis. Sloan. Hist. Jam. vol. 1. pag. 78. Cat. p. 16. tab. 33. fig. 1. The description “simplici” is an error, because simple fronds do not have pinnae, and Aublet describes the pinnae. Aublet tried to identify his specimens, and new species, with previously published descriptions and illustrations by Plumier, Sloane, and others, often incorrectly since his specimens came from French Guiana, and the illustrations cited usually were of West Indian species, mostly from Jamaica or Hispaniola. In this case, the reference to Sloane’s illustration of a Jamaica plant is such an error, which has caused a continuing confu- sion. Sloane’s drawing (very poor) is of some species of Asplenium, and Mettenius went so far as to identify it with a form of Aspleniwm auritum Swartz, and Posthumus, in his Ferns of Surinam, listed Polypodium serratum as a synonym of Asplenium sulcatum Lam. Christensen, in his Index Filicum, was much closer when he indicated that P. serratum was “Asplenium sp. vel Leptochilus guianensis.” An examination of the holotype in the British Museum shows that it is indeed Leptochilus guianensis (Aubl.) C. Chr., which must now be known as Bolbitis guianensis (Aubl.) Kramer, incidentally a name omitted from the new Index Filicum, Supplement IV. 68. Po.ypoprum Triste Kunze, Linnaca 9:47. 1834=Thelypteris tristis (Kunze) Tryon, Rhodora 69:8. 1967. Dryopteris tristis Kuntze, Rev. Gen, Pl. 2:814. 1891, 66 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM Type: Mission Tocache, Huallaga, Peru, Poeppig 1959. The original was doubtless destroyed in Leipzig. I did not find any isotypes in the British Museum, or in the herbaria in Paris and Leiden. This species was not known from Venezuela when Christensen pub- lished his monograph of Dryopteris in 1913. Jnuth listed it in his Initia Florae Venezuelensis, but without noting any specimens; like many of Knuth’s records in this work, the inclusion was very likely based on probabilities. A definite Venezuelan record is as follows: VENEZUELA: Forest 3-4 km. southeast of ‘‘Los Patos,” north of Rfo Hacha and north of Rfo Supamo, 30 km. south of El Manteco, State of Bolfvar, alt. 365 m., Aug. 9, 1960, Steyermark 98027 (VEN). 69. PoLypopium vaRroLaTuM Willd. in L. Sp. Pl. ed. 4, 5:192. 1810=Poly- podium triseriale Swartz, Journ. Bot. Schrad. 1800 (2):26. 1801, var. Type: Brazil, communicated by Hoffmannsegg (perhaps collected by Sieber ?) (holotype Willd. Herb., B, sheets 19685 [1-3], photograph of one sheet by Tryon, US). Mettenius, in his monograph of Aspidium,** placed “Polypodium variolatum Willd. Herb. Spreng.” as a synonym of Aspidium macro- phyllum Swartz, ie., the present Tectaria incisa Cav., and Christensen in the Index Filicum followed him and placed P. variolatum Willd. as an undoubted synonym of Aspidium martinicense Spreng., another name for the same species of Tectaria. Mettenius would hardly have made a mistake of this sort, and so it may be presumed that the specimen labeled P. variolatum in the Sprengel Herbarium (present location unknown) really is a Tectaria. However, the holotype, which I saw in the Willdenow Herbarium in Berlin, is by no means a Tectaria, but a typical Polypodium of the section Goniophlebium. It has been identified by Hieronymus, probably correctly, as P. menisciifolium Langsd. & Fisch., but so far as I can tell at present this species is only a variety or form of the common and widespread P. triseriale Swartz (much better known under the later name P. brasiliense Poir.). This group of species needs to be critically revised. 70, PoLysTiIcHUM CYPHOCHLAMYS Fée, Gen. Fil. 279. 1852=P. echinatum (Gmelin) C. Chr. Type: Cuba, Linden 2175 (holotype P, Morton photograph 4300). In 1909, Maxon * listed P. cyphochlamys Fée as a doubtful synonym of P. triangulum (L.) Fée, without having seen the type. At that 36 Abh. Senckenb. Ges. Frankfurt 2:406. 1858. 37 Contr. U.S. Nat. Herb. 13:28. 1909. STUDIES OF FERN TYPES, I—MORTON 67 time he was misidentifying Polypodium triangulum L. with a common and widely distributed species of the Greater Antilles with spiny, holly-like pinnae. Later, Maxon ** looked into the question further and came up with a different answer. Since there is no specimen in the Linnaean Herbarium of Polypodium triangulum, the illustration cited by Linnaeus must be taken as the type; this illustration by Petiver was not original but was redrawn from Plumier t. 72, which thus becomes the type of P. triangulum. The Plumier description and drawing, although the latter is stylized, best represent a species of Hispaniola collected several times by Ekman, Leonard, and Picarda, quite different from the species previously called Polystichum triangu- lum, which now becomes P. echinatum (Gmelin) C. Chr. The holotype of P. cyphochlamys Fée shows that this species is a synonym of typical P. echinatum. 71. Potysticnum KitLipu Maxon, Contr. U.S. Nat. Herb. 24:53, t. 20. 1922= Polystichum trapezoides (Swartz) Presl, Tent. Pterid. 83. 1836. Aspidium trapezoides Swartz, Journ. Bot. Schrad, 1800(2):31. 1801. Type: Jamaica, Swartz (Isotype B, Willd. Herb, 19749, photograph by Tryon US). Typg: One mile below Ipswich, Parish of St. Elizabeth, Jamaica, Apr. 1, 1920, Mazon & Killip 1520 (Holotype US). In his treatment of the West Indian species of Polystichum in 1909, Maxon? mentioned a collection from Troy, Jamaica (Underwood 2837) as possibly representing an undescribed species; this he later described in 1922 as P. killipii. In 1909 he considered Aspidium trapezoides Swartz as a doubtful synonym of his P. “triangulum,” i.e., the present P. echinatum, and in describing P. killipu he appar- ently did not again consider the identification of A. trapezoides. A study of a photograph of an isotype of A. trapezoides shows that Maxon was probably right in considering it (in herbarium) as the same as his P. killipii, but a detailed study of the holotype in Stockholm is necessary for a definite decision. A plant from Cuba (El Yunque, Baracoa, Oriente, Ekman 3918, US) is rather poor, but apparently represents P. trapezoides, which thus occurs in Cuba as well as Jamaica. The Cuban species of Polystichum, not so numerous as in Hispaniola or Jamaica, may be distinguished by the following key: 8 The identification of Polypodium triangulum L. Journ. Washington Acad. Sci. 18:582-586. 1928. 39 Contr. U.S. Nat. Herb. 13:29. 1909. 68 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM Key to the Cuban Species of Polystichum Blades pinnate below the middle only, the apex long-attenuate, radicant. 71a. P. rhizophyllum Blades fully pinnate or twice-pinnate. Blades once-pinnate only, Apex of blades attenuate or flagelliform, proliferous. Pinnae auriculate at both upper and lower bases. Pinnae holly-like, with several marginal spines . . 71b. P. ilicifolium Pinnae with only the auricles and apex spinescent. Pinnae coriaceous, with obscure veins, strongly hastate, the auricles and apex strongly spinose ... . . 7le. P. machaerophyllum Pinnae herbaceous, with evident veins, broadly subhastate but not spinescent ..... ..... . Td. P. deminuens Pinnae not auriculate at the lower base ..... . Tle. P. decoratum Apex of blades not proliferous. Stipes and rhachises densely and persistently paleaceous. 71f. P. triangulum Stipes and rhachises with few scales, these mostly basal. Blades short, ovate to broadly ovate, long-stipitate . 71g. P. wrightii Blades elongate, lanceolate to linear-lanceolate, relatively shorter stipitate. Superior auricle spinescent, never free . . . . . 71h. P. echinatum Superior auricle merely mucronate, not spinescent, often nearly free in mature plants .... ... .. Ti. P. trapezoides Blades partly or fully twice-pinnate, prolifer rous. Apex of blade acute, the proliferous bud borne on the rhachis below the apex... . .. .- . T1j. P. polystichiforme Apex of blade truncate, ‘the bud terminal . » 2... . Tk. P. viviparum 71a. Polystichum rhizophyllum (Swartz) Presl, Tent. Pterid. 82. 1836. Polypodium rhizophyllum Swartz, Veg. Ind. Occ. Prodr, 132. 1788. Type: Jamaica, Swartz (Isotype B, Herb. Willd. 19739, photograph by Tryon, US). MWlustrated by Hook. & Grev., Icon. Fil. 1:t. §9. 1829. Polystichum krugit Maxon, Proc. Biol. Soc. Washington 18:215. 1905. Type: Cayey, Puerto Rico, Sintenis 2240 (holotype US). The reduction to P. rhizophyllum was made by Maxon himself (Contr. U.S. Nat. Herb. 13:36. 1909). Although common in Puerto Rico, this species is rare in Jamaica and Cuba, and has apparently not been reported from Hispaniola, but it is known from that island by one collection: Morne de La Selle, Holdridge 1879 (US). 71b. Polystichum ilicifolium Fée, Gen. Fil. 279. 1852. Polystichum aquifolium Underw. & Maxon, Bull. Torr. Bot. Club 29:584. 1902 (nom. abort.). A renaming of P. ilicifolium Fée from the mistaken idea that this name was invalidated by P. ilicifolium (Don) Moore, but the latter combination was not published until 1858, and it is thus the illegitimate later homonym and not P. ilicifolium Fée (1852). STUDIES OF FERN TYPES, I—MORTON 69 Type: Santiago, Cuba, Linden 2193 (holotype P, Morton photograph 4301). The type is a single frond, without rhizome, quite typical of the species as usually interpreted. The species was illustrated by Fée, Mém. Foug. 6:t. 6, f. 4. 1853. 71c. Polystichum machaerophyllum Slosson, Bull. Torr. Bot. Club 40:688, t. 26. 1914. Type: Arroyo del Medio, Sierra de Nipe, Oriente, Cuba, Shafer 3262 (isotype US). 71d. Polystichum deminuens Maxon, Contr. U.S. Nat. Herb. 24:53, ¢. 79. 1922. Tyre: Near Josephina, Oriente, Cuba, Nov. 4, 1859, Wright 1057 (holotype YU, fragment US). Known only from the type. 7le. Polystichum decoratum Maxon, Contr. U.S. Nat. Herb. 13:30, é 3. 1909. Tyre: Farallones de La Perla, near Monte Verde, Oriente, Cuba, Mazon 4408 (holotype, US). Endemic in the Province of Oriente, Cuba, where it is not rare. 71f. Polystichum triangulum (L.) Fée, Gen. Fil. 279. 1852. Polypodium triangulum L. Sp. Pl. 2:1088. 1753. Type: Based on T'ri- chomanes folio triangulo dentato. Pet. fil. 76. @. 1, f. 10, an illustration which is a copy of Plumier, Tract. Fil. Amer. ¢. 72. 1705. As mentioned above under P. cyphochlamys, Maxon identified the Plumier description and figure with a rather rare species of Hispaniola. From the description, he is probably right, although it would be hard to be sure from the figure alone. An illustration of P. triangulum in this accepted sense of Maxon is given by Maxon, Journ. Washington Acad. Sci. 18:583, f. 1. 1928. Although this species is known only from Hispaniola and not from Cuba, it is included here because the epithet triangulum has been widely applied to Cuban plants erroneously. 71g. Polystichum wrightii (Baker) C. Chr. ex Maxon, Contr. U.S. Nat. Herb. 16:50. 1912. Polypodium wrightti Baker in Hook. & Bak. Syn. Fil. 304. 1867. Type: Cuba, Wright 3924 (holotype presumably K). Dryopteris sauvallet C. Chr. Ind. Fil. 291. 1905. Based on Polypodium wrightit Baker. Polystichum longipes Maxon, Contr. U.S. Nat. Herb. 13:34, t. 6. 1909. Type: Cuba, Wright 3924 (holotype US). Although this species is based on the same collection number as Polypodium wrightii, and Maxon at the time was ignorant of the existence of P. wrightii, his species cannot be considered as nomenclaturally synonymous, since it is based on a different sheet of this collection. Wright’s fern collections are notoriously mixed, and it is theoretically possible that Maxon’s species is different from Baker’s, although in this instance that does not appear to be the case. 70 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM An extremely rare endemic, probably still known only from the original collections by Wright. Maxon cites duplicates at GH, NY, YU, and the Sauvalle Herbarium, Havana. 71h. Polystichum echinatum (Gmelin) C. Chr. Ind. Fil. 581. 1906. Polypodium echinatum Gmelin in L. Syst. Nat. ed. 13, 2(2):13809. 1791. Type: Sloane, Voy. Jam, Nat. Hist. 2:. 86, f. 4, 6. 1707. Although Gmelin probably saw only the illustrations cited, the specimens on which the illustrations were based are preserved in the Sloane Herbarium at the British Museum (Natural History), and were seen by Maxon (Journ. Washington Acad. Sci. 18:584. 1928). These can be designated as lecto- types. Although they represent somewhat different forms, Maxon con- sidered them conspecific and that they represented the common species of the Greater Antilles that long passed as P. triangulum. Swartz cited these same two figures in his description of his Aspidium mucronatum, but his species was based not on these figures but on Jamaican specimens that he had collected and which are preserved in the herbarium at Stockholm; these represent a different species, Polystichum mucronatum (Swartz) Presl, which is endemic in Jamaica. A synonym of P. mucronatum is Polystichum struthionis Maxon, Contr. U.S. Nat. Herb. 13:37, t. 8, fig. A, B. 1909, which was based on ‘‘ Aspidium mucronatum Hook. Sp. Fil. 4:9, t. 216. 1862, not Sw. 1801’ in the mistaken belief that Swartz had based his species on the Sloane illustrations rather than specimens. Maxon’s P. struthtonis should be considered a superfluous name. Even so, it should be typified, something that Maxon did not do. The only speci- men seen by Hooker and also cited by Maxon is Jamaica, Wilson (KX), and this is herewith designated as lectotype of P. struthionis Maxon. Polystichum falcatum Fée, Gen. Fil. 279. 1852. Type: Port-au-Prince, Haiti, L’Epagnier (not seen). Referred here on the authority of Maxon. Polystichum cyphochlamys Fée, Gen. Fil. 279. 1852. See above for a dis- cussion of the type. 71i. Polystichum trapezoides (Swartz) Presl, Tent. Pterid. 83. 1836. See above for a discussion of the type and synonymy. 71j. Polystichum polystichiforme (Fée) Maxon, Contr. U.S. Nat. Herb. 13:35. 1909. Phegopteris polystichiformis Fée, Gen. Fil. 247. 1852. Type: Monte Libano, Oriente, Cuba, Linden 1874 (isotype BM, Morton photograph 6415). This is a rare species of Cuba and Jamaica. Maxon did not see a type, but he correctly identified the species with Wright 832 from Monte Verde, Oriente, Cuba. 71k. Polystichum viviparum Fée, Gen. Fil. 280. 1852. Type: Oriente, Cuba, Linden 1742 p.p. Polystichum heterolepis Fée, Gen. Fil. 279. 1852. Type: Oriente, Cuba, Linden 1742 p.p. (isotype BR, photograph by Weatherby, US). As noted by Maxon,“ P. viviparum Fée and P. heterolepis Fée were founded on portions of the same collection, and the differences noted 0 Contr. U.S. Nat. Herb. 13:33, t. 6. 1909. STUDIES OF FERN TYPES, I—MORTON 71 by Fée are such as ordinarily obtain between different plants from a single collection. In uniting the two species, Maxon chose the name P. heterolepis, without explanation, but doubtless going on “page priority.”” However, this has nothing to do with the choice between names of the same date. In the Index Filicum, Christensen (p. 588, 1906) chose the name P. viviparum, and reduced P. heterolepis to synonymy under it (p. 582), and he must be followed. 72, PreRIS CONFLUENS Thunb. Prodr. Pl. Cap. 171. 1800=Thelypteris con- fluens (Thunb.) Morton, comb. nov. Aspidium thelypteris (L.) Swartz var. sguamigerum Schlecht. Adumbr. 23, t. 11, 1825. Type from Cape Province, South Africa. Nephrodium squamulosum Hook. fil. Fl. N. Zeal, 2:39. 1855. Aspidium squamigerum Fée, Mém. Foug. 8:104. 1857. Thelypteris squamulosa Ching, Bull. Fan Mem. Inst. Biol. Bot. 6:5, 329. 1936. Thelypteris palustris var. sqguamigera Tardieu, Mém. I.F.A.N. 28:119, t. 20, f. 7-9. 1953. Type: Cape of Good Hope, South Africa, Thunberg (UPS, seen by Schelpe). Dr. E. A. C. L. E. Schelpe pointed out recently *' that Pteris con- fluens Thunb. (1800) is not a synonym of Pellaea auriculata (Thunb. Fée, as it is listed in the Index Filicum, but is really the same as the South African Thelypteris palustris Salisb. var. squamigera (Schlecht.) Tardieu. He saw the type of Thunberg’s species in the herbarium at Uppsala. He overlooked the fact that the epithet confluens antedates palustris by many years. Polypodium palustre Salisb., dating from 1796,” is twice illegitimate, first because it is a later homonym of P. palustre Burm. (1768) and second because it was a superfluous name, being an unnecessary change of specific epithet on transferring Acrostichum thelypteris L. (1753) to Polypodium. The epithet palustris validly dates from Thelypteris palustris Schott (1834) .* This is a most distressing circumstance. If the South African plant, which occurs also in southern India and China, and also in New Zealand, is considered only varietally different from 7. palustris Schott, it means that the latter will become a variety of 7. confluens. This plant is one of the best known ferns of western Europe and the United States, where it is commonly called ‘“Marsh Fern.” Until recently it has usually been called Dryopteris thelypteris (L.) A. Gray. It is possible that this northern plant really is specifically distinct from the South 41The identity of some fern types in the Thunberg Herbarium,” Journ. So. Afr. Bot. 29:91. 1963. See also ‘‘A review of the southern African species of Thelypteris,” Journ. So. Afr. Bot. 31:260. 1965. 42 Prodr, 403. 1796. 47 Gen. Fil. adt. 10. 1834, 252-200—67——5. 72 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM African, and I prefer to consider it so for the present rather than displace the well-known name 7. palustris. The chief difference seems to be in the presence of small scales along the costae of the pinnae beneath in the South African T. confluens. If the European and North American plant is considered eventually only varietally different, it will be a difficult problem to ascertain the oldest varietal name applicable to any form of this plant. 73. PTERIS DOLABRIFORMIS Poir. in Lam. Encyel. Méth. 5:722. 1804=Adiantum villosum L. Syst. Nat. ed. 10, 2:1328, 1759. Tyre: Santo Domingo, collector unknown (holotype P, ex Herb. Poiret, Morton photograph 2645, left-hand plant; the right-hand plant mounted on the sheet from the Morne de la Soufrigre, Guadcloupe, October 1827, is not a type, but it also represents A. villosum L.). In the Index Filicum P. dolabriformis Poir. is placed as a doubtful synonym of Adiantum pulverulentum L., but the type, although frag- mentary, obviously represents A. villosum L. in the usual sense. 74. Pteris macroptera Link, Hort. Reg. Bot. Berol. Descr. 2:32. 1833. A specimen (Morton photograph 5446) distributed under this name from the Hortus Berolinensis is undoubtedly authentic material, for it quite agrees with the original description. The species was described from material cultivated in the botanical garden in Berlin from material originally received from Brazil. The species is a synonym of the common and widely distributed Pteris altissima Poiret, which is filed in most herbaria as P. kunzeana Agardh, a later name taken up in the Index Filicum. The venation of this species is characteristic: Along the midribs of the pinnae on both sides are three areoles between adjacent midribs of the segments, two elongate and one short. Allied species such as P. polita (P. propinqua) and P. decurrens normally have just one elongate areole between adjacent costules. The segments are more or less equal at the upper and lower base. Another species, apparently confined to Brazil, has passed as P. macroptera, but it is obviously different. It has distant segments which are very strongly decurrent at the lower base and not surcurrent at the upper. Because of the distance between adjacent midribs of the segments there are several (more than three) areoles along the midrib of the pinna between adjacent segment-midribs. This large species has two available names, the one by Fée being the older. Pteris angustata (Fée) Morton, comb. nov.‘ Litobrochia angustata Fée, Crypt. Vase. Brés. 1:49, ¢. 11, f. 1. 1869. Syntypes from Rio de Janciro, Brazil, Glaziou 2149 and 2310 (not seen; presumably in Paris). “ Pleris angustata Wallich, List. no. 93. 1828, is a nomen nudum. STUDIES OF FERN TYPES, I—MORTON 73 Pteris paulistana Rosenst. Hedwigia 46:89. 1906. Type: Rio Grande, Séo Paulo, Wacket 32 (Rosenst. Fil. Austrobras. 344) (isotype US). RANGE: Rio de Janeiro, SAo Paulo, Minas Gerais, and Paran4, Brazil. Other collections in the National Herbarium are: BRAZIL: Rio pE JANEIRO: Tijuca, Smith & Brade 2208; Rio de Janeiro, Martius (HBG, Morton photograph 2201). So Pauto: Iguape, Brade 8244. Minas Gerais: Areponga to Fazenda de Grama, Meria 4244 (distr. as P. decur- rens). Parand&: Serra do Mar, Dusén 548a. Without specific locality: Bowie & Cunningham 4; Glaziou 7952, 75. Preris POLYPODIOIDES Poir. in Lam. Encycl. Méth. 5:716. 1804=Thelyp- teris totta (Thunb.) Schelpe, Journ. So. Afr. Bot. 29:91. 1963, var. hirsuta (Mett.) Morton, comb. nov. Aspidium unitum var. hirsutum Mett. Ann. Lugd. Bat. 1:230. 1864. Dryopteris gongylodes var. hirsuta C. Chr. Dansk. Vid. Selsk. Skrift. VII, 10(2):198. 1913. Cyclosorus goggilodus var. hirsutus Farwell, Amer. Midl. Nat. 12:259. 1931. Type: Brazil, without collector (Lamarck Herb., P, Morton photograph 2755). In the Index Filicum, Pteris polypodioides Poir. is placed as a synonym of Dryopteris unita (L.) Kuntze. Probably no recent pteriodologist has critically examined the type, which turns out to be the hirsute variety of the common and widespread species usually known as Dryopteris gongylodes (Schkuhr) Kuntze (the “D. goggilodus”’ of some recent authors who believe in blindly following the original spelling, even though that is patently impossible orthographically). When I discovered this, I was intending to abandon reluctantly the well-known epithet gongylodes and to propose a new combination based on Pteris polypodioides Poir., which has priority, but Dr. Schelpe has recently shown that there is still another earlier name for the species— Polypodium tottum Thunb. (1800). He is right in proposing the new combination Thelypteris totta (Thunb.) Schelpe to replace T. gongylodes (Schkuhr) Small, but it is most unfortunate, because the epithet ‘‘totta” has been widely used for an entirely different and also wide- spread species that is correctly known now as Thelypteris pozoi (Lagasca) Morton (correct, that is, according to my opinion of the taxonomy; some authors would prefer to call it Leptogramma pozot or Stegnogram- Ma poz0r). Thelypteris totta occurs in two forms, one with the blades essentially glabrous beneath and aften capitate-glandular (the original Aspidvum gongylodes Schkuhr, the type from British Guiana, usually called Dryopteris gongylodes var. glabra (Mett.) C. Chr), and one with the pinnae conspicuously pilosulous beneath on the midribs and veins, the blades eglandular, and the indusia hairy (characters shown by the type of Pteris polypodioides); the latter is usually called Dryopteris 74 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM gongylodes var. hirsuta (Mett.) C. Chr. These two forms are best considered as varieties provisionally, for although they look very different in extreme specimens, they do seem to be connected by inter- mediates that are only slightly pubescent; this would be a good problem for cytological and genetical study. Dr. Schelpe did not mention any pubescence on the type of Poly podium tottum Thunb., and so it may be assumed that it is essentially glabrous, especially since the hirsute variety apparently does not occur in the region of the Cape of Good Hope, where Thunberg’s type came from. The name “var. glabra” will disappear, and the typical form should be known merely as var. totta. This species has been described many times, and I have assembled over two typewritten pages of synonymy. The more important synonyms (mostly on the authority of the Index Filicum) of var. totta are: Aspidium pohlianum Pres], Delic. Prag. 1:173. 1822. Type from Madeiras, 3razil, Pohl (not seen). Nephrodium venulosum Desy. Mém, Soc. Linn. Paris 5:255. 1827. Type from “‘insulis africanis’’ (not seen). Hypopellis propinquoides Bory, in Bélanger, Voy. Bot. 2:69. 1833. Type from Java (not seen). Aspidium ecklonit Kunze, Linnaea 10:546. 1836. Type from South Africa (not seen). Some of the synonyms of var. hirsuta are presumed to be: Nephrodium propinquum WR. Brown, Prodr, Fl Nov. Holl. 148. 1810, Type from Australia, Banks & Solander, in 1770 (holotype BM, Morton photograph 6621). Aspidium continuum Desv. Ges. Naturf. Freund. Berlin Mag. 5:320. 1811. Based on Pleris polypodioides Poir. Aspidium resiniferum Kaulf. Fnum. Fil. 237. 1824. Type from the Hawaiian Islands, Chamisso (not seen). Aspidium venulosum Blume, Enum. 151. 1828. Type from Java, Blume (not seen). Nephrodium paludosum Liebm. Dansk. Vid. Selsk. Skrift. V, 1:275. 1849. Type from San Antonio, Huatusco, Veracruz, Mexico, Liebmann 2658 (fragment US). Goniopteris cheilocarpa Fée, Gen. Fil. 251. 1852. Syntypes from Brazil, Claussen 112, Gardner 53 (not seen). Nephrodium inaequilaterum Colenso, Trans. N. Zeal. Inst. 20:229. 1888. Type from New Zealand, Colenso (not seen). The very large plant of southern Brazil, with the pinnae up to 40 cm. long or more may be known as: Thelypteris totta var. longipinna (C. Chr.) Morton, comb. nov. Dryopteris gongylodes var. longipinna C, Chr, Dansk. Vid. Selsk. Skrift., VII, 10(2):194. 1918. There are five syntypes from Brazil, Uruguay, and Paraguay. Since I have not seen all of them, I do not choose a lectotype. STUDIES OF FERN TYPES, I—MORTON 75 76. Pteris polita Link, Hort. Reg. Bot. Berol. Deser. 2:30. 1833. Pteris propinqua Agardh, Ree. Pterid. 65. 1839. In his original description of Pteris propinqua Agardh quoted P. polita Link as a synonym with a query, evidently not being sure enough of Link’s species to adopt the name. The syntypes of pro- pinqua came from Jamaica (Bancroft, MacKadyn). Link’s P. poltta has remained in limbo, so to speak, ever since. It was based on plants cultivated in the botanical garden in Berlin from material originally received from Brazil. In the herbarium of the Staatsinstitut fiir allgemeine Botanik, Hamburg, I found two sheets (Morton photo- graphs 5449, 5450) of P. polita that had been collected in the Hortus Berolinensis and which are undoubtedly authentic. They bear the date 1834, which may be the date of collection or the date sent out. The data are similar to those of many other species described by Link in 1833; all have proved to be correctly named and are essentially isotypes. It is evident that Link (probably with Friedrich Otto) col- lected samples from the hothouses in Berlin of his new species de- scribed in his Hortus Regius Botanicus Berolinensis and distributed them widely. ‘The specimens in Hamburg show that P. polita Link is truly the same as the later P. propingua Agardh, and therefore the latter name, which has become rather widely known through its adoption in the Index Filicum, must be displaced. The species is widely distributed in tropical America and does grow both in Brazil, the type locality for P. polita, and in Jamaica, the type locality for P. propinqua. 77. SALVINIA ROTUNDIFOLIA Willd. in L, Sp. PL, ed. 4, 5:537. 1810=S. auriculata Aubl. Hist. Pl. Guian. 2:969, t. 8367. 1775. Type: Brazil, Hoffmannsegg (B, Herb. Willd. no, 20250). The first author to distinguish clearly between two common tropical American species of Salvinia was C. A. Weatherby,“ who pointed out that S. auriculata Aubl. had the clusters of four hairs with the hairs united by their tips, and that S. rotundifolia Willd. had the hairs completely distinct and tapering at the tip. Unfortunately, he did not see the type of S. rotundifolia Willd. Through the courtesy of D. W. Domke, the director of the herbarium at the Botanisches Museum, Berlin-Dahlem, I was privileged to study the holotype. It is a good specimen, which is obviously the same as S. auriculata Aubl., closely matching Samuels 73, from Surinam. It is unfortunate that the name S. rotundifolia must disappear into synonymy, but there is no help for it. As a matter of fact, the name was considered a synonym of S. auriculata by J. G. Baker and by Christensen, a 5A further note on Salvinia,’’ Amer. Fern Journ, 27:98-102. 1937. 76 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM conclusion also agreed with by Miss J. Kopp in her dissertation on Salvinia. According to Kopp, the species that Weatherby and others following him ** have called S. rotundifolia should be called S. minima Baker,” a species that was based on a collection (Fritz Mueller 479) from Santa Catarina, Brazil, which is presumably at the Royal Botanic Gardens, Kew, but may be in the British Museum. 78. ScHIZAEA PENICILLATA Humb. & Bonpl. ex Willd. in L. Sp. Pl. ed. 4, 5:86. 1810= S. pennula Swartz, Syn. Fil. 150, 379. 1806. Recent authors have followed Prantl * in recognizing two species of Schizaea section Digitata [Actinostachys Wallich as a genus] from northern South America, one S. pennula Swartz, with nonstriate spores, and S. penicillata Humb. & Bonpl., with striate spores; aside from the spore difference, the latter species has smaller and fewer sorophores. However, S. penicillata was originally proposed merely as a change of name for S. pennula Swartz, and was unnecessary, and therefore was superfluous and illegitimate. Martius correctly placed S. penicillata as a synonym of S. pennula, and described the second of the species mentioned above as S. subtrijuga Mart.,*® which is a correct name; the types of S. subtrijuga were collected by Martius from Arara- Coara and Cupati, Amazonas, Colombia. 79. TRICHOMANES AERUGINOSUM Poir. in Lam. IEneyel. Méth. 8:76. 1808 =Hymenophyllum aeruginosum (Poir.) Carm. Trans. Linn. Soe. [London] 12:513. 1818. Trichomanes hirsutum sensu DuPetit Thouars, Esq. Fl. Trist. d’Acunha 34. 1804, non L. Hymenophyllum capillare Desv. Mém. Soc. Linn. Paris 6:333. 1827. Type: Tristan d’Acunha Island, DuPetit Thouars (presumably P, not seen). Hymenophyllum fuluum van den Bosch, Nederl. Kruidk. Arch. 5(3) :196. 1863. Type: Tristan d’Acunha, collector unknown [but surely DuPetit Thouars] (holotype in Herb. Berlin [not seen], fragment of holotype L, Morton photographs 1551, 2532). Van den Bosch originally gave the locality as Madagascar, but subsequently wrote on the sheet “Tristan d’Acunha, Thouars”; his plants are thus very likely a part of the same collection as the type of H. capillare. Type: Tristan d’Acunha Island, Bory de St. Vincent (holotype P, Morton photograph 4546). The name Hymenophyllum capillare Desy. has commonly been applied to a species of Madagascar, Réunion, and tropical Africa. “© Dr. Elias de la Sota has published several valuable papers on S. auriculata, S. rotundifolia sensu Weatherby, and other species. See Darwiniana 12:465-520, 1962; 612-623. 1963. 47 Journ. Bot. Brit. & For, 1886:98. ** Untersuchungen zur Morphologie der Gefisskryptogamen 2:132. 1881. * Tcon. Crypt. Vase. 117. 1834. STUDIES OF FERN TYPES, I—MORTON 77 However, the type came from the little island of Tristan d’Acunha in the South Atlantic Ocean, and only one species of this alliance grows there, the previously and generally recognized species H. aeruginosum (Poir.) Carm., of which H. capillare must be a synonym. If the African and Mascarene Island plants should prove to be different,which may very well be true, then the proper name for them will be H. pendulum Bory, which was based on a plant collected on Réunion Island by Bory de St. Vincent. 8). TRICHOMANES MUSCOIDEs Swartz, Journ. Bot. Schrad. 1800 (2):95. 1801= Trichomanes hymenoides Hedwig, Fil. Gen. Sp. t. 3, f. 3. 1799. Swartz gave a description of his 7. muscoides but cited “Hedw. icon. fil.” as a synonym, and later the exact illustration of Hedwig intended is indicated by Swartz in his Synopsis Filicum (p. 142. 1806) as Trichomanes hymenoides Hedwig. Therefore, 7. muscordes Swartz should by the Code be considered as a renaming of T. hymenoides, superfluous and therefore illegitimate. The typification of these names is of some interest. Hedwig indicated his species as “Habitat . . .” thus showing that he did not know the origin of his material. Swartz sent many of his specimens from Jamaica to Hedwig who illustrated them, under the names assigned by Swartz. It is likely that Swartz sent a specimen of a Jamaican species to which he had assigned the name 7. muscordes, and that this label became lost. Hedwig retained the specimen, and described it as a new species, 7'. hymenoides. Swartz, writing just about the same time or a year later, realized what had happened, and considered his own name 7’. muscoides as having priority, and so he adopted it and placed the Hedwig plate in synonymy. This would have seemed right to him, but it is not in accordance with our current Code of Nomenclature. The Swartz specimen from Jamaica at Stockholm would be authentic, technically an isotype rather than a holotype. The holotype of Hedwig could not be located by Wessels Boer in his recent treatment of Trichomanes section Didymoglossum,” and so the plate was considered as typifying the species. It is likely that the holotype is actually in the herbarium in Geneva; it could be identified by comparison of the plants with Hedwig’s illustration. Boer (1962) considered 7’. muscoides Swartz as a heterotypic synonym of T. hymenoides, and based on the Swartz collection at Stockholm, but as indicated above I believe it should be considered a homotypic synonym, a renaming of J. hymenoides rather than a new species independently described. A specimen collected by Swartz in Jamaica in the Rijksherbarium, Leiden (Morton photograph 2416) would probably be an isotype. 59 Acta Bot. Neerl. 11:306. 1962. Index (Synonyms in italics. New species, new names, and combinations in boldface. Page numbers of principal entries in italics.) Acrostichum, 44 Aspidium—Continued acuminatum, 31, 32 levyt, 37 alatum, 44 macrophyllum, 66 filare, 39, 40 martinicense, 56, 57, 66 gorgoneum, 44 mucronatum, 70 lancifolium, 32, 33 oppositum, 37, 38 micradenium, 44, 45 patens, 54 rigidum, 32 pohlianum, 74 salicifolium, 33 resiniferum, 74 sessile, 44 sclerophyllum, 38 thelypteris, 71 squamigerum, 71 Actinostachys, 76 thelypteris Adiantum, 34 var. sguamigerum, 71 acuminatum, 33 trapezoides, 67 capillus-veneris, 34 trifoliatum, 56 cardiochlaena, 34 unitum fructuosum, 38 var. hirsutum, 73 pectinatum, 34 venulosum, 74 politum, 33, 34 Asplenium, 65 polyphyllum, 33, 34 adiantoides, 39-41 var. politum, 33 aethiopicum, 40 pulverulentum, 33, 72 affine schaffneri, 34 var. gilpinae, 41 tetraphyllum, 33 var. tanalense, 41 f. obtusum, 33 auritum, 65 var. obtusum, 33 f. serratum, 65 villosum, 33, 72 cultratum, 41 Anemia cumingit, 41 cicutaria, 38 falcatum, 39-41 cuneata, 38 forsterianum, 41 Anisosorus, 50 furcatum, 39, 40 Aspidium, 64, 66 gilpinae, 41 abruptum, 62, 63 intermedium, 41 alomarium, 34 kaulfussit, 41 attenuatum, 35 lanceolatum, 39, 40 boryanum, 43 macdonellii, 41 chontalense, 36 polyodon, 40 continuum, 74 praemorsum, 39, 40 desvauzit, 64 robinsonii, 53 diplazioides, 36 squamulatum ecklonii, 74 var. smithii, 53 germantt, 36, 37, 54 stipitatum, 63 gongylodes, 73 sulcatum, 65 intermedium, 37 Athyrium, 41-43 invisum, 59-61 praestans, 41 79 80 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM Blechnum, 42 lanceola, 42 treubit, 42 Blotiella, 50 javanica, 49 Bolbitis guianensis, 65 Cornopteris, 43 boryana, 43 decurrenti-alata, 43 forsythit-majoris, 43 macdonellit, 41 parvisora, 44 Ctenitis, 43, 55, 64 blanchetiana, 54 canescens, 54 desvauxii, 64 mascarenarum, 38 opposita, 37, 38 Cyclopeltis, 46 crenata, 47 cumingiana, 46, 47 presliana, 46, 47 Cyclosorus alatellus, 35 goggilodus var. hirsutus, 73 invisus, 60 Cyclosorus, subg., 60, 61 Cystopteris bulbifera, 34, 35 diaphana, 35 fragilis, 35 var. mackayi, 34, 35 Didymoglossum, sect., 77 Digitata, sect., 76 Diplazium praestans, 41, 42 Drynaria quercifolia, 56 Dryoathyrium, 41, 43 forsythii-majoris, 43 macdonellii, 41 parvisorum, 43 Dryopteris, 36-38, 43, 45, 46, 50-52, 55, 59, 62, 65, 66 abrupta, 62 affinis, 50 andreana, 50 arborescens, 51 arcana, 42 arida, 60 attenuata, 35 augescens, 60, 61 blanchetiana, 54 canescens, 55 chiriquiana, 36 Dryopteris—Continued chrysodioides, 51 var. goyazensis, 61 consobrina, 42 dentata, 61 desvauzii, 52 f. glandulosa, 52 diplazioides, 36, 45, 46 dispar, 50 effusa, 36 ensiformis, 42 exculta var. guatemalensis, 36 falcata, 51 feei, 60 ferox, 60 forsythii-majoris, 41, 43 ghiesbreghtii, 45 gigantea, 52 goggilodus, 73 gongylodes, 73 var. glabra, 73, 74 var. hirsuta, 73, 74 var. longipinna, 74 hemsleyana, 36 hostmannii, 50, 51, 59 intermedia, 37 invisa, 60 jurgensenii, 51 kunzeana, 62 levyi, 37 lingulata, 43 linkiana, 45 longifolia, 52 mascarenarum, 38 megalodus, 64, 65 membranacea, 55 minuscula, 43 mollis, 45, 50 moritziana, 36, 46 nesiotica, 43 nicaraguensis, 55 normalis, 37, 538, 54, 60, 61 obtusiloba, 64 oligophylla, 59, 61, 62 var. aequatorialis, 61 var. kunzeana, 62 var. lutescens, 62 var. pallescens, 62 opposita, 37, 38 parvisora, 43 patens, 37, 60, 61 permollis, 50, 51 Dryopteris—Continued reticulata var. arborescens, 50 sauvallei, 69 serra, 61 simplicifrons, 52 sloanet, 59 sorbifolia var. mollis, 50 spinulosa, 37 standleyt, 44 stenobasis, 35 thelypteris, 71 tristis, 65 turrtalbae, 44 unita, 60, 73 varia, 57 subg. Cyclosorus, 60 Elaphoglossum alatum, 44 gorgoneum, 44 lancifolium, 32 micradenium, 44, 45 nitidum, 44 pellucidum, 44, 45 salicifolium, 33 sessile, 44 Gleichenia pedalis, 52, 53 squamulosa, 52 Goniophlebium, sect., 66 Goniopteris cheilocarpa, 74 mollis, 46 Goniopteris, sect., 37, 45, 55 Grammitis, 56 flabelliformis, 57 linkiana, 45 trifurcata, 55, 56 Gymnograma diplazioides, 36, 45 polypodioides, 45 Hemicardion cumingianum, 46, 47 Hymenophyllum, 48 aeruginosum, 76, 77 capillare, 76, 77 ciliatum, 47 delicatissimum, 47 elegans, 47 elegantissimum, 47 fragile, 48 var. venustum, 48 fuluum, 76 INDEX Hymenophyllum—Continued hirsutum, 47, 48 lineare, 47 pendulum, 77 polyanthos, 48 producens, 48 pulchellum, 48 silveirae, 48 valvatum, 48 venustum, 48 sect. Mecodium, 48 sect. Sphaerocionium, 47, 48 Hypopeltis propinquoides, 74 Lastrea, 61 abrupta, 59, 62 attenuata, 35 invisa, 59 presliana, 46 Lastreopsis, 36 exculta subsp. guatemalensis, 36 Leptochilus gutanensis, 65 Leptogramma pozot, 73 Litobrochia angustata, 72 Litobrochia, subg., 49 Lomaria speciosa, 31, 32 Lonchitis, 49, 50 aurita, 49, 50 glabra, 50 hirsuta, 49, 50 javanica, 49 pubescens, 49 Mecodium, sect., 48 Menisctum, 50, 51 affine, 50 andreanum, 50 arborescens, 50, 51 chrysodioides, 51 falcatum, 51 giganteum, 51, 52 jurgensenii, 51 longifolium, 52 Meniscium, sect., 59 Mertensia pedalis, 52 squamulosa, 52 Neottopteris stipitata, 63 Nephrodium, 54 abruptum, 62 alatellum, 35 attenuatum, 35 guatemalense, 36 inaequilaterum, 74 invisum, 59-61 81 82 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM Nephrodium—Continued kunthit, 53, 54 kunzeanum, 62 lechleri, 55 paludosum, 74 propinguum, 74 semicordatum, 46 sloanei, 59 sorbifolium, 50 f. angustipinnatum, 50 var. molle, 50 squamulosum, 71 venulosum, 74 Nephrolepis cordifolia, 56 Parapolystichum, 36 Parathyrium, 43 forsythti-majoris, 43 macdonellit, 41 parvisorum, 44 Pellaea auriculata, 71 Phegopteris arborescens, 50 canescens, 54, 55 membranacea, 54 mollis, 50 nicaraguensis, 55 polystichiformis, 70 Photinopteris, 32 acuminata, 31, 32 cumingit, 32 hors fieldii, 32 humboldtii, 32 simplex, 31, 32 speciosa, 32 Polybotrya caudata, 55 Polypodium, 71 adianthoides, 55 adiantoides, 55 asplenifolium, 58 aureum, 63 barbatum, 38 blanchetianum, 54 brasiliense, 66 canescens, 54 comptonitfolium, 55, 56 comptonioides, 55 conjugatum, 66 cordifolium, 56 crenatum var. ghiesbreghtii, 45 decumanum, 63 echinatum, 70 expansum, 56, 57 flabelliforme, 57-59 ghiesbreghtii, 45 Polypodium—Continued hartii, 57 hostmannii, 59 invisum, 59, 60 jubiforme, 58 knowltoniorum, 57 leucatomos, 63 lycopodioides, 64 megalodus, 64 menisctifolium, 66 nigripes, 63 obtusilobum, 64 palustre, 71 patens, 54 pellitum, 64 pennatum, 64, 63 phyllitidis, 56 phymatodes, 56 rhizophyllum, 68 rigescens, 57-59 semicordatum, 46 serratum, 66 serricula, 57 suspensum, 58 taenifolium, 57 thelypteroides, 64 tottum, 73, 74 tovarense, 58 triangulum, 67-69 trifurcatum, 55 triseriale, 66 triste, 66 variolatum, 66 wrightit, 68, 69 sect. Goniophlebium, 66 Polystichum, 67 aquifolium, 68 cyphochlamys, 66, 67, 69, 70 decoratum, 68, 69 deminuens, 68, 69 echinatum, 66-68, 70 falcatum, 70 heterolepis, 70, 71 ilicifolium, 68 killipti, 67 krugit, 68 longipes, 69 machaerophyllum, 68, 69 mucronatum, 70 polystichiforme, 68, 70 preslianum, 46 rhizophyllum, 68 struthionis, 70 Polystichum—Continued trapezoides, 67, 68, 70 triangulum, 66-68, 69, 70 viviparum, 68, 70, 71 wrightii, 68, 69 Pteris, 49 aculeata, 49 altissima, 72 angustata, 72, 73 confluens, 71 decurrens, 72 dolabriformis, 72 kunzeana, 72 macroptera, 72 paulistana, 73 polita, 72, 76 polypodioides, 73, 74 propinqua, 72, 75 subg. Litobrochia, 49 Salvinia, 75, 76 auriculata, 75, 76 minima, 76 rotundifolia, 75, 76 Schizaea penicillata, 76 pennula, 76 subtrijuga, 76 sect. Digitata, 76 Sphaerocionium, sect., 47, 48 Stegnogramma pozot, 73 Tarachia haenkeana, 41 Tectaria, 63, 66 heracleifolia, 46 incisa, 66, 57, 66 melanocaulis, 63 trifoliata, 56 Thelypteris, 45, 50-52, 59, 60 affinis, 50 andreana, 50 angustifolia, 50 arborescens, 50 arcana, 42 attenuata, 35 chrysodioides, 51 var. goyazensis, 51 confluens, 71, 72 consobrina, 42 diplazioides, 36, 46 faleata, 51 forsteri, 60 ensiformis, 42 ghiesbreghtii, 45 INDEX Thelypteris—Continued gigantea, 51 gongylodes, 73 hostmannii, 59 invisa, 69-61 var. aequatorialis, 61 var. kunzeana, 62 var. pallescens, 62 kunthii, 36, 37, 53, 61 levyi, 37 lingulata, 43 linkiana, 46 longifolia, 52 f. glandulosa, 52 megalodus, 64 membranacea, 54 minuscula, 43 mollis, 45, 50 nesiotica, 43 nicaraguensis, 55 normalis, 53 oligophylla, 59 palustris, 71, 72 var. squamigera, 71 patens, 54 pennata, 64 poiteana, 45 pozoi, 73 reticulata, 50 salzmannii, 50 sclerophylla, 38 serrata, 50 squamulosa, 71 standleyi, 44 totta, 73 var. hirsuta, 73 var. longipinna, 74 tristis, 65 turrialbae, 44 subg. Cyclosorus, 60, 61 sect. Goniopteris, 37, 45, 55 sect. Meniscium, 59 Trichomanes adiantoides, 39, 40 aeruginosum, 76 aethiopicum, 39, 40 hirsutum, 76 hymenoides, 77 lineare, 47 muscoides, 77 sect. Didymoglossum, 77 Niphopteris, 57 U.S. GOVERNMENT PRINTING OFFICE: 1967 83 BULLETIN OF THE UNITED STATES NATIONAL MUSEUM CONTRIBUTIONS FROM THE UNITED STATES NATIONAL HERBARIUM Vo.LumeE 38, Part 3 THE GENUS GRAMMITIS IN ECUADOR By Conrap V. Morton SMITHSONIAN PRESS e WASHINGTON, D.C. ¢ 1967 THE GENUS GRAMMITIS IN ECUADOR Conrap V. Morton Introduction The Ecuadorian collections of the late Dr. Wendell H. Camp con- tained a rather large number of specimens of the fern genus Grammuitis as I currently recognize it. The present paper is the result of the study of this material, which included a number of rarities, among them several undescribed species. For the past 35 years or more, I have been following Dr. William R. Maxon in recognizing an extremely large, very diverse genus Poly- podium. Maxon was, in turn, following Christensen, Hieronymus, and other noted fern students. Evidence, however, has been ac- cumulating in recent years that there are two major groups of species— those centering around the true Polypodium, as typified by P. vulgare L., and those belonging to Grammitis Swartz and some closely related small genera. It seems that there is no alternative now to recognizing Polypodium and Grammitis as distinct genera; in fact some authors go so far as to place these genera in different families, as Polypodiaceae and Grammitidaceae, although this is an extreme view to take of plants that, in some instances, can hardly be distinguished from one another. Copeland in his monograph of Ctenopteris (which I consider to be Grammitis), in at least one instance included specimens of a true Polypodium and a Grammitis under a single name (Ctenopteris trun- corum), and he also noted that Polypodium pectinatum, P. plumula, and their allies were probably referable to Ctenopteris, although more recent studies have shown that these species are truly Polypodium, closely mimicking ‘“‘Ctenopteris” in habit. Polypodium and Grammitis differ in all parts of the plants—vegeta- tive fronds, sporangia, spores, and gametophytes. Possibly the most uniform and easily observable difference is in the spores, those of Polypodium being monolete and those of Grammitis trilete; however, not all species have been investigated from this point of view. If this character really holds, it is fortunate; that it is not necessarily a fundamental character is shown by the fact that both monolete and trilete spores may occur in species unquestionably in the same genus, as in Vittaria, Antrophyum, and Dicranopteris. The spores of Gram- 85 86 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM mitis are said to be green and those of Polypodium not green. A still more easily observable character is found in the type of hairs on the vegetative parts. In Grammitis, these hairs are spreading, rigid, dark, elongate, and one-celled; such hairs seem to be always found on the stipes, and are also frequently found on the blades, especially on the margins or among the sporangia. In Polypodium, such hairs (usually called setae) are absent; hairs, if present, are either short and glandular or flaccid and septate. The rhizome scales in Polypodium are usually of a more “clathrate” type, with large lumina and narrow walls, although such scales are found also in some species of Grammitis. Although the fronds of some species of Polypodium are scaleless (such as P. vulgare), they often have scales on the stipes, midribs, or leaf surfaces. According to Kenneth A. Wilson,’ the sporangial stalk in Polypodium is two-celled at the base and that of Grammitis one-celled. Alma G. Stokey and Lenette R. Atkinson have published a paper ? on the gametophytes of Grammitis in which some differences from Poly- podium have been shown (1958, p. 402): “The bulbous basal cell or cells, or the mass or plate, giving rise to the filament, the bead-like cells of the filament, the growth of the filament by ‘Budding,’ the pro- tracted filamentous stage, and the characteristic fragmentation of the filament” in Grammitis. The gametophytes of Polypodium are cor- date and have the typical “polypodioid” type of development. The genus Grammitis Swartz was described * as follows: “Capsulae in lineolis rectis, sparsis. Jndusia nulla. Origo vocis a ypayun, linea... .’”’ This was intended to be contrasted with Polypodium L., which had the following diagnosis: ‘‘Capsulae in puncta subrotunda sparsa, congestae absque Jndusio.”” Thus, Grammitis included those species with the sporangia scattered in a straight line, whereas Polypodium had them congested in a rotund sorus. Although this character is not a real distinction, for some species properly included in Grammitis have rotund sori and conversely some undoubted Polypodiums, such as P. astrolepis Liebm., have elongate sori, the species included by Swartz did form a reasonably cohesive group. They were G. linearis Swartz [now G. graminea (Swartz) Copel.], G. marginella (Swartz) Swartz, G. lanceolata Swartz [now Loxogramme lanceolata (Swartz) Presl], G. serrulata (Swartz) Swartz [based on Acrostichum serrulatum Swartz, now called Xiphopteris serrulata (Swartz) Kaulf. by some authors], G. graminoides (Swartz) Swartz [Acrostichum graminoides Swartz, now Cochlidium graminoides (Swartz) Baker], and C. myosuroides (Swartz) Swartz [based on Polypodium 1 “The sporangia of three problematic species of Polypodium,” Amer. Fern Journ, 49:147-151. 1959. 2“The gametophyte of the Grammitidaceae, Phytomorphology’’ 8:391-403. 1958. See also, Alma @G. Stokey, ‘‘Polypodium pectinatum and P. plumula—Polypodiaceae or Grammitidaceae?,’? Amer. Fern Journ, 49:142-146. 1959, 8 In Journ. Bot. Schrad. 1800 (2):17. 1801. MORTON—GENUS GRAMMITIS 87 myosuroides Swartz, called Xiphopteris myosuroides (Swartz) Kaulf. by some authors]. Thus, all the species are still retained in Grammitis or referred to the closely allied groups Loxogramme, Cochlidium, and Xiphopteris. There was, of course, no original designation of a type for the genus Grammitis, nor were there any comments by Swartz indicating that he had any particular species in mind as being more “typical” than another. The designation of a lectotype is of importance, since although the original six species are undoubtedly allied, they are sufficiently different to have been referred to different genera. Copeland ‘ merely states: ‘The first species listed is G. linearis, the same au- thor’s Polypodium gramineum, Prod. (1788) 130. This, being its first species, and being retained in it here, as it has been by all authors except J. Smith who have maintained the genus, must be its type species.” As often occurs, however, Copeland’s typifications are based on false premises. Very few authors in the past and probably none at present consider that the first species listed has any more claim to be the lectotype of a new genus than any of the others. In fact, the International Code of Nomenclature (1961 ed., p. 64, in the “Guide for the determination of types’) specifically states: “Designation of a lectotype should be undertaken only in the light of an understanding of the group concerned. Mechanical systems, such as the automatic selection of the first species or specimen cited . . . should be avoided as unscientific. . . .” In the case of Grammitis, John Smith ® excluded this first species, G. linearis, from the genus Grammitis, calling it Lomaphlebia; for the purposes of typification, it does not matter that the character used by Smith in segregating this new genus may not be considered of generic signifi- cance. Smith specifically typified Grammitis on G. australis R. Brown, but since this was not one of the original species of the genus it is not a possible choice; however, Smith did keep one of the original species in the genus, G. marginella. Christensen ® followed this up by definitely selecting G. marginella as the type of the [subsection] Grammitis, recognizing also a [subsection] Lomaphlebia, with type G. graminea. This must, therefore, be considered as a definite choice of G. marginella as lectotype of Grammitis. It is a rather unfortunate fact that none of the six original species of Grammitis are really representative of most of the species of the genus as it is currently recognized. Strangely enough, although these are rather numerous, they were not known in the time of Swartz. The type species G. marginella is no exception; it has the blades with 4 Phil. Journ. Sci. 80:93. 1951. 5 Hist. Fil. 182. 1875. ®Ind. Fil. XLIX. 1906. 88 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM a conspicuous black border, merely a border and not a marginal vein, for the real veins end some distance within the margin. This is a peculiar character, but there are no other differences. Even Copeland, whose generic concepts in this group are certainly narrow, did not separate this group of species from Grammitis, but called it Grammitis subg. Melanoloma.’ No type was specifically designated, and, therefore, G. marginella Swartz is here designated lectotype. This group of about 14 species, of both the Old and New World, does not occur in Ecuador. It must now be called sect. Grammitis, of which subg. Melanoloma is a synonym. Since the name Grammitis must now be applied to the small group of species typified by G. marginella, another sectional name must be found for the bulk of the species that have commonly been called Grammitis. A possible name is Grammitis sect. Chilopteris Presl,° which included originally the three species G. billardieri, G. linearis [i.e. graminea], and G. marginella. Christensen,® by listing this as a complete synonym of [subsect.] Grammitis, effectively chose G. marginella as type, and thus sect. Chilopteris remains a synonym of sect. Grammitis. The next possible name, and the one that is correct, is Grammitis sect. Grammitastrum Fourn.,’° which was based on the two species Polypodiwm pseudoaustrale (Fourn.) Fourn. and P. lasiosorum (Blume) Hook. The former species, which is based on Grammitis pseudoaustralis Fourn.," is here designated lectotype; it is much the more suitable, for it was well known to Fournier, whereas he may have misidentified P. lasiosorum. The late E. B. Copeland published revisions or monographs of Grammitis,” Xiphopteris," and Ctenopteris,* of varying worth. The Grammitis paper is creditable and typical of Copeland’s previous monographic works—fairly good taxonomic judgments combined with unusable keys, no descriptions, and a lack of definiteness in statements of range. The Xiphopteris paper is brief, and actually only an adaptation of Maxon’s previous work on the species allied to Polypodium serrulatum * and P. trichomanoides,' with the addition of several new species, most of which had been previously segregated by Maxon (as species of Polypodium). The Ctenopteris paper, under- taken when Copeland was at an advanced age, is much the poorest of 7 Phil. Journ. Sci. 80:253. 1951. 8’ Tent. Pterid. 208. 1836. * Ind. Fil. LIX. 1906. 10 Ann. Sci. Nat. [Paris] V, 18:282. 1873. ' Bull. Soc. Bot. France 16:423. 1869. ! “Grammitis,’’ Phil. Journ. Sei. 80:93-267, pls. 1-6, figs. 1-108. 1951. 3 ““The American species of Xiphopteris,’’ Amer. Fern Journ. 42:41-52, 93-110, 1952. 4 “Ctenopteris in America,’’ Phil. Journ. Sci. 84:381-470, pls. 1-16. 1955, 18 “Notes upon Polypodium duale and its allies,’ Contr. U.S. Nat. Herb. 17:398-406. 1914. 16 “*Polypodium trichomanoides and its American allies,’ Contr. U.S. Nat. Herb. 17:542-557. 1916. MORTON—GENUS GRAMMITIS 89 any of Copeland’s published works, being replete with errors of all sorts—typographical, certainly, but more importantly in morphological observations and in taxonomic judgments. It is essentially un- usable. The species are distributed among 10 groups, but these groups are not named, not keyed, and only briefly or not at all characterized or described. I have noted below an instance of what is essentially the same species that Copeland has placed under three different names in three different groups. A restudy of the American species is much needed; the present paper is the beginning of such a study. Copeland attempted to maintain Grammitis, Xiphopteris, and Ctenopteris as distinct genera, Grammitis being defined as having the fronds simple or lobed less than halfway to the midrib, Xiphopteris with lobed to pinnatifid fronds with the veinlets simple or with a single branch, and Ctenopteris with pinnatifid fronds, the segments with pinnate veins. Since there are no other differences, these are “one-character” genera, and this one character an essentially insignifi- cant one. The division of the frond (simple as opposed to pinnatifid) has never been considered a generic character in ferns, apart from other differences. And the difference between having one vein branch and several of them is only correlated with having the lobes short or elongate. These ferns vary from having no lateral vein-branches, to having one, two, several, or many. Copeland realized these affinities very well, but chose to recognize the three genera as a matter of convenience. To me, it is more convenient to group them together as a coherent genus. For the purpose of classifying the species it may be convenient to group the species in corresponding “‘sections,”’ with the realization that these sections may not represent natural phylo- genetic units. Systematic Treatment Grammitis Swartz The system of American Grammitis that I adopt is: Grammitis Swartz, in Journ. Bot. Schrad. 1800(2):17. 1801. Sect. Grammitis (Lectotype: Polypodiwm marginellum Swartz). Grammitis subg. Melanoloma Copel. Phil. Journ. Sci. 80:253. 1951 (Lectotype: Polypodium marginellum Swartz, Prodr. Veg. Ind. Occ. 130. 1788). Sect. Grammitastrum Fourn. Ann. Sci. Nat. [Paris] V, 18:282. 1873 (Lectotype: Grammitis pseudoaustralis Fourn. Bull. Soc. Bot. France 16:423. 1869). Grammitis subg. Grammitis sensu Copel. 90 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM Sect. Cryptosorus (Fée) Fourn. Ann. Sci. Nat. [Paris] V, 18:282. 1873 (Lectotype: C. blumei Fée, based on Polypodium obliquatum Blume).” Ctenopteris auctt.'8 Sect. Xiphopteris (Kaulf.) Presl, Tent. Pterid. 208. 1836 (Type: Acrostichum serrulatum Swartz.) Xiphopteris Kaulf. Berlin Jahrb. Pharm. 1820:35. Key to the Species of Grammitis in Ecuador Fronds entire or lightly repand (sect. Grammitastrum). Rhizome not paleaceous. Veins forked ..... .1. G. jungermannioides Rhizome with minute scales at apex, these more or less hidden among the numerous hairs. Veinsforked. .. 2... eee ee ee GR yarumalensis Veinssimple. .. 2... . ee ee eee ee . 8G. spruce Fronds lobed to pinnatisect. Segments with only a costa or this with a single anterior veinlet (sect. X iphopteris). Costa of the segments unbranched. Fertile part of the frond subentire, the sterile part toothed or lobed less than halfway to the costa. Sporangia borne on the costa as well as on the base of the veins, thus confluent; rhizomes elongate; sterile part of the blade merely toothed, the fertile part short and contracted . . . 4. G. serrulata Sporangia borne only on the base of the veins, not confluent; rhizomes short, erect; sterile part of the blade lobed, with deltoid segments, the fertile part elongate and tail-like. ..... .5.G. jamesonii Fertile and sterile parts of the blades alike, deeply pinnatifid, the lobes oblong, narrowed from the base. Stipes stout, erect; rhizome scales ciliate. 2... ee ee ew... . . 6 G. caucana Costa of the segments with a single anterior branch (veinlet), this often subbasal. Rhizome scales not setose on the margins, sometimes with a very few short, several-celled, glandular hairs. Blades deeply lobed, densely setose, long-stipitate . . ....... 2... =. .%. G. aphelolepis Rhizome scales setose on the margins, the setae stiff, spreading, one- celled, eglandular. Blades nearly or quite exstipitate, decurrent as a wing on the stipe, lobed less than halfway to the costa, conspicuously setose throughout... .........2.2.2.. .8. G. truncicola Blades obviously stipitate, deeply lobed nearly to the costa, the lobes oblong, slightly gibbous on the upper margin. 9. G. blepharolepis 17 Fée had only two species in his genus Cryptosorus: C. dionaea and C. blumei. The genus was distin- guished primarily, as indicated by the generic name chosen, by the sunken sori, a character well shown in C. blumei (1.e. Polypodium obliquatum Blume) but not or hardly in C. dionaea, which is considered to be a synonym of Polypodium venulosum Blume. Therefore, C. blumei is the logical choice of lectotype. 48The question of the publication of Ctenopteris has been discussed recently elsewhere (Amer. Fern Journ. 56:65-68, 1966). 1” The term “gibbous,” i.e., hunch-backed, is used to describe a slight hump or lobe on the anterior margin of some of the larger segments of these ferns; it is an indication of the forking of the vein. MORTON—GENUS GRAMMITIS 91 Segments with a costa, this with pinnate veinlets (sect. Cryptosorus). Veinlets forked. Epiphytes. Fronds not setose beneath (except slightly on the midrib beneath in G. crispata); rhizome scales not ciliate; blades lobed to pinnatifid but not pinnatisect. Fronds white-ceraceous beneath, not at all setose. Segments elongate, not fully confluent, auriculate at the upper base . . 10. G. farinosa Fronds not white-ceraceous beneath. Veins free or a few casually anastomosing; blades with a short, entire decurrent base; sori superficial, mostly elongate, without setae intermixed with the sporangia; blades not setose on the midrib beneath, sometimes with a few marginal setae. Blades with an accessory veinlet arising from the costa; basal superior veinlet running toward and nearly reaching the sinus; blades deeply pinnatifid, the larger ones with a costal wing only ca. 1 mm. wide on either side . . . . . 11. G. eminens Blades with no veinlets arising from the costa apart from the main lateral veins; basal superior veinlet not nearly reaching the sinus; blades shallowly lobed to pinnatifid, the costal wing always broad... .. .... . 12. G. trifurcata Veins more or less regularly anastomosing to form a series of aeroles along the midrib; blades with an elongate entire basal portion; sori impressed, mostly round, sparsely setiferous; blades sparsely setiferous on the midrib beneath, copiously ciliate . 13. G. crispata Fronds more or less setose beneath; rhizome scales ciliate; blades pin- natisect, the segments free from each other, not connected by a con- tinuous wing, more surcurrent than decurrent. Rhizome scales eglandular-ciliate; segments acute, mostly twice as long as broad, evidently crenate-undulate, glabrous above, setose beneath. . ... .... .14. G. semiadnata Rhizome scales glandular- -ciliate; segments rounded, mostly 1.5 times as long as broad, entire or lightly undulate, sparsely setose above near the margin and apex, glabrous beneath . . . 15. G. intricata Veinlets simple, not forked. Sporangia pilose or setose. Pendent elongate epiphytes (or rarely saxi- colous), softly hairy (or merely setiferous-margined in G. alfarit) ; rhizome scales ciliate (or absent in G. alfariz). Blades (or some of them) once- or twice-forked. Plants sometimes saxicolous at high elevations. . .. . . . .16. G. heteromorpha Blades not forked. Blades (or some of them) bipinnate. Plants always epiphytic. 17. G. variabilis Blades simply pinnate or only pinnatisect. Rhizome scales absent; segments strongly unequal-sided at base, strongly gibbous or humped at upper base. . . 18. G. alfarii Rhizome scales present; segments more or less equal-sided, not gibbous. Pinnae 3-10 mm. long. Pinnae usually 6-10 mm. long; white hydathodes none or inconspicuous on the upper surface; blades usually more than 12 cm. long. Rhachis without coarse, dark stellate hairs; veins mostly only 2pairs. ........... .19% G. senilis 238-617—67—2 92 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM Rhachis with some dark, coarse, stellate hairs mixed with the soft white ones; veins 3 or 4 pairs. . 20. G. dependens Pinnae usually 3-6 mm. long; white hydathodes present or none; blades usually less than 12 cm. long. White hydathodes present and conspicuous on the upper surface; segments oblong, not much or not at all nar- rowed toward the base. .... . .21. G. subtilis White hydathodes none; segments obviously narrowed at the baseabove. ... .. . . . 22. G. subflabelliformis Pinnae longer, up to 45 mm. long. Marginal hairs all solitary; hairs of the rhachis beneath solitary; hairs all rather coarse and stiff. . . 23. G. lehmanniana Marginal hairs mostly paired; hairs of rhachis beneath often partly fasciculate; hairs mostly soft and silky. Sporangia with only one short hair or many of them hairless; plants sparingly hairy; blades pectinate, the segments close... . 2... . 2. 2... 24, GL laxa Sporangia all with hairs, these usually 2 or more and often elongate; plants densely hairy; blades with rather distant segments. Segments with 9-14 pairs of veins (and usually also sori), elongate, usually over 20 mm. long. Hairs of rhachis beneath all pale, even the fasciculate ones... .. . . . 25. G. lanigera var. lanigera Hairs of the rhachis beneath (or some of them) coarser, darker, and fasciculate, the fascicle often somewhat stalked and thus forming a “stellate” hair. 25a. G. lanigera var. stella Segments with 2-7 pairs of veins (and usually sori), short- er,under18mm.long..... . . . 26. G.cultrata Sporangia glabrous. Blades (or some of them) once- or twice-forked. Epiphytes (or saxicolous at high elevations); plants soft-hairy . . . 16. G. heteromorpha Blades not forked. Blades (or some of them) bipinnate. Soft-hairy epiphytes. 17. G. variabilis Blades pinnate, or rarely pinnate-pinnatifid. Segments ascending at an acute angle, narrow and elongate, long- decurrent at the base. Epiphytes. Blades pinnatifid only, the segments connected by a narrow wing. Rhachis above green and glabrous; segments entire. 27. G. subsessilis Blades fully pinnate. Pinnae subentire. Rhachis more or less viscid-puberulous. 28. G. jamesonioides Pinnae crenulate to pinnatifid. Rhachis above with numerous, short, white, stiff hairs; pinnae merely crenulate . . . 29. G. pseudocapillaris Rhachis above with a few, conspicuous, elongate, dark setae; pinnae pinnatifid. . . . . . 80. G. myriophylla MORTON—GENUS GRAMMITIS 93 Segments spreading horizontally. Stipes densely short-pubescent. Blades not setulose; rhizome scales ciliate; rhizomes erect, short; texture firm; veins obscure; segments fully adnate but not confluent. Middle segments 10-20 mm. long; epiphyte. . 31. G. major Middle segments ca. 6 mm. long, relatively broad; terrestrial. 32. G. tunguraguae Stipes with sparse or numerous setiform hairs, not short-pubescent. Rhachis hirtous or setose above. Plants epiphytic (or occa- sionally apparently saxicolous or terrestrial in G. anfrac- tuosa, G. taxifolia, G. leucosticta, and G. semihirsuta). Plants spreading by delicate, radicant, rootlike, stolon-like rhizomes. Blades small and delicate. 33. G. anfractuosa Plants not stoloniferous. Segments 3 times as long as broad or longer. Rhizome scales not ciliate; hydathodes without lime in- crustations above. Segments elongate; veins 6-15 pairs; stipes stout, 0.7-1 mm. in diameter. Lateral veins dark and visible; segments narrow and elongate, with parallel sides, obtuse at apex, with abundant, pale brown setae on the lower surfaces, the margins usually plane, the texture rather thin. 34. G. taxifolia Lateral veins immersed, hardly visible; segments shorter, acutish, the margins incurved, with dark, numerous or few setae, the texture thick. Stipes up to 3.5 cm. long, glabrate; segments sparingly setose. ........... . 35. G. firma Stipes up to 11 cm. long, densely setose; segments densely dark-setose. . . . . . 36. G. vernicosa Rhizome scales ciliate, sometimes sparingly so; hydathodes with obvious lime-dots on the upper surface. Plants lax, the stipes delicate, 0.2-0.4 mm. in diameter; veins 3-6 pairs, immersed, not visible. Segments with dark setae on the surface; rhizome scales dark, with dark cilia. . . . . . . 37. G. pichinchensis Plants coarser, the stipes (0.5) 0.7-1.6 mm. in diameter; veins often more numerous (few in G. cuencana). Veins only 2-4 pairs; segments short and broad; stipes short, 1-2 cm. long. . . 38 G. cuencana Veins more numerous; segments elongate; stipes usually longer. Veins 4-8 pairs, visible but not obvious. Dark setae present on the surfaces of the segments, usually none on the margins; segments more or less gradually narrowed from a broad base; stipes 2.5-11 mm. long. Rhizome scales blackish, with dark cilia. 39. G. pichinchae Rhizome scales dark brown, with pale brown cilia. 40. G. attenuatissima 94 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM Veins more numerous, (5) 8-20 pairs, usually dark and readily visible. Rhizome scales very sparsely ciliate; stipes elongate, 7-25 cm. long. Veins (5) 8-11 pairs; segments more or less gradually narrowed from a_ broad base. Dark setae very sparse on the surfaces and margins. . .... . . 41. G. leucosticta Veins usually more numerous, 8-20 pairs; seg- ments elongate, with more or less parallel sides. Dark setae numerous and short on the surfaces; cilia of the rhizome scales weak and spread- ing. ....... . . 42.G.alsopteris Dark setae absent on the lower surfaces, very few on the upper margin; cilia of the rhizome scales rigid, appressed-ascend- ing... .... . . 43. G. semihirsuta Rhachis glabrous above or nearly so. Rhizome seales not ciliate. Plants epiphytic. Midrib immersed, not or searcely visible. Rhizomes slender, ca. 1 mm. in diameter. Setae usually sparse, except among the sporangia. . 44. G. rigens Rhizomes thicker, 1.5-2.5 mm. in diameter. Black setae present on the surfaces of the fertile segments; rhizome scales without hairlike tips. 45. G. pseudonutans Black setae present only on the rhachis beneath and among the sporangia; rhizomes scales with elon- gate hairlike tips only one cell thick. 46. G. sodiroi Plants saxicolous or terrestrial, mostly rigid and erect. Segments mostly not more than twice as long as broad. Rhizomes slender, 0.9-1.1 mm. in diameter (excluding the scales). Midribs of the segments dark and prominulous; blades up to 24cm. long. Stipes up to 5 em. long, 0.5 mm. in diameter; segments truncate, obtuse, 5-6 mm. long and 4 mm. broad. . .. . 47. G. wolfii Midribs of the segments immersed in the leaf-tissue, hardly visible; blades usually much less than 24 cm. long. Rhizomes always very widely creeping, with distant fronds. Stipes elongate, 2-6 em. long, slender, and wiry; segments usually about twice as long wide. ........ . 48. G. flabelliformis Rhizomes often shorter, with closer fronds. Segments mostly about 1.5 times as long as wide; stipes very slender (0.2-0.4 mm. in diameter), mostly only 1-2 cm. long. 49. G. peruviana Segments mostly about as long as broad, very thick; stipes stouter, often more elongate. 50. G. moniliformis MORTON—GENUS GRAMMITIS 95 Rhizomes thicker, 1.5-5 mm. in diameter, excluding the scales. Segments triangular from a broad base . 51. G. rosarum Segments broadly oblong to semicircular. Plants large and coarse, mostly 25-35 cm. long; stipes ca. 1mm.indiameter. . 52. G. assurgens Plants smaller, mostly less than 20 cm. long; stipes usually ca. 0.5 mm. in diameter. 50. G. moniliformis 1. Grammitis jungermannioides (Klotzsch) Ching, Bull. Fan Mem. Inst. Biol. Bot. 10:240. 1941. Polypodium jungermannioides Klotzsch, Linnaea 20:373. 1847. Type: Mérida, Venezuela, Moritz 312 (presumably B). Grammitis repanda Kunze ex Mett. Fil. Lechl. 1:9. 1856. Illegitimate renaming of Polypodium jungermannioides Klotzsch. Polypodium sprucei var. furcativenosa Hieron. Bot. Jahrb. Engler 34:499. 1904. Syntypes: Near Putumayo, on Rio Balsayacu, Cordillera de Pasto, Colombia, Lehmann 654 (isosyntype US). Chilcasco, Guatemala, Salvin & Godman. Polypodium sprucei var. costaricensis Christ, Repert. Sp. Nov. Fedde 8:17. 1910. Type: Voledn de Barba, 2,200 m., Costa Rica, Brade 296 (isotype NY). Range: Jamaica; Guatemala to southern Colombia. Habitat: Epiphytic. This species has not yet been found in Ecuador, but it may very well occur there. Its characters were discussed by Maxon in detail.”° 2. Grammitis yarumalensis (Hieron.) Proctor, Bull. Inst. Jam., Sci. Ser. 5:36. 1953. Polypodium yarumalense Hieron. Bot. Jahrb. Engler 34:499. 1904. Type: Yarumal, Colombia, Lehmann 7390 (isotype US). Range: Jamaica, Hispaniola, Panama, and Colombia. Habitat: Epiphytic. No Ecuadorian collections are known, but they are to be expected. These small epiphytes are easily overlooked. 3. Grammitis sprucei (Hook.) J. Smith, Hist. Fil. 181. 1875. Polypodium sprucei Hook. Second Cent., pl. 10. 1860. Type: Tarapoto, Peru, Spruce 4746 (isotype US). Range: Known only from the type. This minute fern should be searched for in Ecuador, among mosses and hepatics on tree trunks in dense moist forests at middle elevations. 4. Grammitis serrulata (Swartz) Swartz, Journ. Bot. Schrad. 1800 (2):18. 1801. Acrostichum serrulatum Swartz, Prodr. Veg. Ind. Oce. 128. 1788. Type: Jamaica, Swartz. % Proc. Biol. Soc. Washington 51:35, 36, 1938. 96 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM Aiphopteris serrulata Kaulf. Enum. Fil. 85. 1824. Polypodium serrulatum Mett. Fil. Hort. Lips. 30. 1856, non P. serrulatum Swartz (1801). Xiphopteris extensa Féc, Mém. Foug. 11:14. 1866. Type from Guadeloupe, L’Herminier. Non Polypodium extensum Forst. (1786). Polypodium duale Maxon, Contr. U.S. Nat. Herb. 16:61. 1912. Based on Acrostichum serrulatum Swartz, non Polypodium serrulatum Swartz. Range: Common throughout the American tropics, and also in Mauritius and tropical Africa. Habitat: Creeping on tree trunks, or occasionally on rocks, mixed with mosses and hepatics, from sea-level to 1,300 m. Keuadorian specimens examined: Between Bafos and Mera, Prov. Napo-Pastaza, Mexia 6966. El Sajado, on Rfo Santiago, Prov. Concepcién, Mexia 8449. Road between Quito and Santo Domingo, Prov. Pichincha, Haught 3219. Eastern slope of the cordillera, between Chontal and Santa Elena, on trail from Sevilla de Oro to Mendez, Prov. Santiago-Zamora, Camp E-809. 5. Grammitis jamesonii (Hook.) Morton, comb. nov. Xiphopteris jamesoni Hook. Second Cent. Ferns pl. 14. 1861. Type: Andes of Quito, Ecuador, Jameson (K!). Polypodium serrulatum var. strictissimum Wook. Sp. Fil. 4:175. 1862. Based on X. jamesonii Hook. Polypodium serrulatum var. major Mett. Ann. Sci. Nat. [Paris] V, 2:249. 1864. Based on X. jamesonii Hook. Polypodium jamesonii Jenman, Bull. Dept. Jam. 4:112, 1897, non Mett. 1883. Polypodium serrulatum var. jamesonii Krug, Bot. Jahrb. Engler 24:125. 1897. Based on X. jamesonii Hook. Polypodium strictissimum Hieron. Bot. Jahrb. Engler 34:501. 1904. Polypodium strictissimum f. minus Hieron. Hedw. 44:85. 1905. [as ‘“‘minor”] Type: Quito, Ecuador, Cuming 5 (presumably B). Range: Venezuela, Colombia, and Ecuador, at elevations from [1,500?] 1,800 to 3,400 m. Habitat: Epiphytic on mossy tree trunks, or sometimes on rocks at the higher elevations. Ecuadorian specimens examined: Along trail between Rio Clava- dero and the ridge 10 miles south of Las Toldadas, east of Cayambe Peak, Prov. Imbabura, Wiggins 10499. Boggy area east of Cayambe Peak, Prov. Imbabura, Wiggins 10422. Péramo and sub-péramo area north and northwest of the Péramo del Castillo, 6-8 km. NNE of Sevilla de Oro, Prov. Azuay, Camp E-5131. 6. Grammitis caucana (Hieron.) Morton, comb. nov. Polypodium caucanum Hieron. Bot. Jahrb. Engler 34:503. 1904. Type: Rio Dagua, Dept. Cauca, Colombia, Lehmann 3257 (holotype B, not seen). A tphopteris caucana Copel. Amer. Fern Journ. 42:98. 1952. MORTON—GENUS GRAMMITIS 97 Range: Reported by Copeland from Nicaragua to British Guiana and Ecuador. Habitat: Terrestrial or epiphytic, at elevations from 700 to 2,300 m. Ecuadorian collections seen: None. The only other species with simple veins that is likely to occur in Ecuador is Grammitis delitesceens (Maxon) Proctor (Polypodium delitescens Maxon, Bull. Torr. Bot. Club 32:74. 1905; Contr. U.S. Nat. Herb. 17: 403, pl. 12, fig. 10. 1914), which is common in Jamaica, and occurs also in Central America; it has been reported from Colom- bia. It is a much more delicate plant, with the lamina not setose as in G. caucana. Maxon did not definitely designate a type; the lectotype chosen by Copeland (Amer. Fern Journ. 42:52. 1952) is Jamaica, Mazon 1513 (US). 7. Grammitis aphelolepis Morton, sp. nov. PuaTe 1 Planta epiphytica; rhizoma breve erectum ca. 2 mm. diam. basibus stipitium exclusis, apice paleaceum, paleis pallide brunneis, anguste vel late lanceolatis, 2-3 mm. longis, 0.6-1.0 mm. latis, apice abrupte an- gustatis et subcuspidatis, margine integris, non setoso-ciliatis hinc inde pilis minutis, ca. 70 » longis et 10 y latis, septatis flaccidis capitato— elandulosis praeditis, basi ca. 14 cellulis latis, cellulis subquadratis vel rectangularibus, 55-130 » longis, 45-55 » latis, parietibus valde incrassatis; frondes caespitosae, 10-20 cm. longae, numerosae, sti- pitatae, stipitibus brunneis, 1.5-4 cm. longis, 0.3-0.5 mm. diam., dense setosis, setis 2-2.5 mm. longis, late patentibus unicellularibus, pilis minutis intermixtis nullis; laminae lineares, 7-14 cm. longae, 7-10 mm. latae, pinnatifidae, rhachi supra immersa dense setosa, subtus elevata, setosa, ala costali angusta 0.25-0.75 mm. lata, segmentis crassis, 18-35-jugis, late oblongis, 4-5 mm. longis, medio ca. 2 mm. latis, ascendentibus, apice rotundatis, margine superiore gibbosis, utrinque dense setosis, setis brunneis rigidis 2-2.5 mm. longis; venae unifurcatae; sori dorsales basi venulae anticae grandes orbiculares, non setosi, non paraphysati; annulus ex ca. 13 cellulis constitutus; sporae eloboso—tetraedricae, 32-43 » diam., sublaeves. Type in the U.S. National Herbarium, no. 2,080,287, collected near the laguna on Péramo del Castillo, on the crest of the Eastern Cordil- lera on the trail between Sevilla de Oro and Mendez, Province of Azuay, Ecuador, Aug. 29, 1945, at 9,000-11,000 {t. elevation, by W. H. Camp (no. E-5107). Paratypes (all Ecuador): Péramo and sub-péramo area north and northwest of the Péramo del Castillo, Prov. Azuay, 10,000-11,200 ft. alt., Aug. 31, 1945, Camp E-5139 (US); between Huagrarancha and Loma de Gal&pagos, Prov. Azuay, 3,140-3,505 m. alt., July 9, 1943, 98 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM Steyermark 53464 (US); near lower margin of pdramo, eastern slopes of Cayambe Peak, Prov. Imbabura, 11,200 ft. alt., July 16, 1944, Wig- gins 10,412 (US); ridge about 10 km. south of Las Toldadas, east of Cayambe Peak, Prov. Imbabura, 10,650 ft. alt., July 15, 1944, Wiggins 10,402 (US). 8. Grammitis truncicola (Klotzsch) Morton, comb. nov. Polypodium truncicola Klotzsch, Linnaea 20:374. 1847. Type: Colonia Tovar, Venezuela, Moritz 252 (isotype US). Polypodium truncicola var. major, Klotzsch, loc. cit. Type: Moritz 252. Klotzsch divided P. truncicola into two varieties: major and minor, the former of which must be considered as the typical variety, as was done by Copeland (Amer. Fern Journ. 42:102. 1952). Polypodium truncicola var. minus Klotzsch, loc. cit. [as “minor’’]. Type: Mérida, Venezuela, Moritz 333 (isotype US). Polypodium andinum Hook. Second Cent., pl. 6. 1847. Syntypes: Banks of Rio Hondacha, Andes of Quito, Ecuador, Jameson 780 K!, and Mount Picote, near Moyobamba, Peru, Nilson (Spruce 4780 K!). The Jame- son specimen, presumably at Kew, is here designated lectotype. Grammitis andina Ching, Bull. Fan Mem. Inst. Biol. Bot. 10:240. 1941. Xiphopteris truncicola Copel., Amer. Fern Journ. 42:101. 1952. Range: Costa Rica, Venezuela, Colombia, Ecuador, and Peru, from sea level (in western Colombia) to 2,300 m. elevation. Habitat: Epiphytic in dense and damp forests. Ecuadorian specimens examined: Apparently known in Ecuador only from the Jameson specimen cited above as lectotype of Poly- podium andinum. Since the species P. truncicola and P. andinum were both published in 1847 and the exact date of publication is doubtful, Copeland’s decision to adopt truncicola must be followed. This species is dis- tinctive because of the fronds being merely shallowly lobed and not deeply pinnatifid. 9. Grammitis blepharolepis (C. Chr.) Morton, comb. nov. Polypodium gracillimum Hieron. Hedw. 48:250, pl. 12, fig. 18. 1909, non Copel. (1905). Between Quito and Mindo, Ecuador, Stuebel 747 (B, not seen). Polypodium blepharolepis C. Chr. Ind. Fil. Suppl. 1:58. 1913. Based on P. gracillimum Hieron., non Copel. Xiphopteris blepharolepis Copel. Amer. Fern Journ. 42:109. 1952. Range: Venezuela, Colombia, Kcudador, and Peru [fide Copeland], at low to middle elevations. Habitat: Epiphytic in dense forests. Ecuadorian specimens examined: Known in Ecuador only from the type. A related species found at low elevations in the Departments of Cauca and El Valle in Colombia that will probably be found in northern Ecuador is Grammitis daguensis(Hieron.) Morton, comb. nov. MORTON—GENUS GRAMMITIS 99 [Polypodium daguense Hieron. Bot. Jahrb. Engler 34:504. 1904; type: Rio Dagua, Dept. Cauca, Colombia, Lehmann 1951 (isotype US)]. It can be distinguished by the segments not being at all gibbous. Still more closely similar in its gibbous segments, but differing in having entire and not setulose-margined rhizome scales is Grammitis trichomanoides (Swartz) Ching, which is primarily West Indian but which occurs scatteringly elsewhere. Polypodium undulatum Fourn. (1872) (non Willd., 1810) was based in part on Jameson 537 from Ecuador, which I have not seen but which may represent G. trichomanovdes. 10. Grammitis farinosa (Hook.) Morton, comb. nov. Polypodium farinosum Hook. Icon. Pl. 10: pl. 947. 1854. Type: “On the trunk of an old tree at the eastern ascent of the Cordillera of Quito [Ecuador], where the forests commence; rare;’”” W. Jameson (KI, fragment US). Ctenopteris farinosa Copel. Phil. Journ. Sci. 84:470. 1955. Range: Colombia and Ecuador. Habitat: Epiphyte, at elevations from 2,500 to 3,500 m. Ecuadorian specimens examined: Eastern side of Mount Tung- uragua, Prov. Tunguragua, Rimbach 25. 11. Grammitis eminens Morton, sp. nov. PLATE 2 Planta epiphytica; rhizoma breve, perspicue paleaceum, paleis pallide flavis, lanceolatis, ca. 3 mm. longis et 0.6 mm. latis, gradatim longe acuminatis, apice non piliformibus, clathratis, basi ca. 7-cellulis latis, margine subundulatis, non ciliatis, cellulis subrectangularibus magnis, ca. 175-250 yu longis, 50-70 yu latis, parietibus crassis pallide luteis; frondes subcaespitosae 4-8, 14-43 cm. longae, verisimiliter subdimorphae, steriles breviter stipitatae, stipitibus 1-5 cm. longis quam frondibus multo brevioribus, ca. 0.7 mm. diam., teretibus, non costatis, longe et molliter setosis, setis 1.5-2 mm. longis, ca. 20 4 latis, unicellularibus, late patentibus, pallide brunneis, setis brevioribus rigidis nullis; lamina sterilis 10-20 cm. longa, 12-24 mm. lata, pinna- tifida, basi cuneata, apice acuminata, rhachi immersa late alata, ala basi 1 mm. lata, supra latiore, segmentis 20-30-jugis, ascendentibus, latis, usque ad 13 mm. longis, 3.5-5 mm. latis, basi adnatis non auriculatis, apice rotundatis, margine integris, textura crassis, supra viridibus glabris, hydathodis nullis, subtus pallidioribus, non ceracels, fere glabris, in costa sparsissime setosis, setis elongatis curvatis paucissimis, margine in sinubus pilis minutis glandulosis sparse dis- sitis; frondes fertiles longe stipitatae, stipite 1/3 longitudinis laminae vel lamina fere aequali vel longiore, 5.5-18 cm. longo, 0.8-1.4 mm. diam., lamina quam sterili latiore, 14-22 cm. longa, 3-6 cm. lata, segmentis lanceatis, 16-23-jugis, maximis 5.5 cm. longis et 5-7 mm. 238-617—67——3 100 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM medio latis; venulae pinnatae, immersis, in segmentis utrinque obscuris, 9-22-jugae, ascendentes, subbasaliter furcatae, ramis margine non attingentibus, ramis venarum basalium non raro apice anasto- mosantibus itaque areolam formantibus, venula basalis superior fere ad sinus attingens, venula altera ex costa excurrens, furcata, ramulis saepe anastomosantibus; sori paullo submediales, dorsales, in ramula antica venularum, subrotundi, receptaculo paullo elongato; sporae sphaerico-tetraedricae, triletae, 34-39 » diam., laeves. Type in the New York Botanical Garden, collected in the péramo and sub-péramo area north and northwest of the Paéramo del Castillo, ca. 6-8 km. NNE of Sevilla de Oro, Province of Azuay, Ecuador, at 10,000-11,200 ft. elevation, Aug. 31, 1945, by W. H. Camp (no. E-5169). Paratypes: Near the laguna at Paramo del Castillo, crest of the eastern cordillera in the trail between Sevilla de Oro and Méndez, Province of Azuay, Ecuador, at 9,000-11,000 ft. elevation, August 29, 1945, by W. H. Camp (no. E-5162) (NY, US). 12. Grammitis trifurcata (L.) Copel. Gen. Fil. 211. 1947. Polypodium trifurcatum L. Sp. Pl. 1084. 1753. Type: Martinique, Plumier, Tract. Fil. pl. 138. Polypodium trifurcatum L. var. brevipes Hieron. Bot. Jahrb. Engler 34:500. 1904. Syntypes: Quebrada de Imbi, Rfo Cuaiquer, Cor- dillera de Pasto, Colombia, Lehmann 77; Farallones de Cali, Colombia, Lehmann 1977; Cordillera west of Cali, Colombia, Lehmann 7664 (isosyntype US). Range: West Indies; Surinam and Venezuela to Bolivia, at elevations from 600 to 2000 m. Habitat: Epiphyte in dense, wet forests. Ecuadorian specimens examined: Rio Suguibi, Rimbach 27. Rio del Cinto, on the southern slopes of Mount Pichincha, Stuebel 735. Although Copeland placed this species in Grammitis in his restricted sense, it is out of place there. The alliance seems to be with the species placed by Copeland in the “eurvata” group of Ctenopteris. 13. Grammitis crispata (J. Smith) Morton, comb. nov. Ctenopteris crispata J. Smith, in Seem. Bot. Voy. Herald 227, pl. 48. 1854. Syntypes: Isla de Cacagual, southern Darien, Panama, and Bay of Chocéd, Colombia, Seemann 995K! Polypodium crispatum Hook. Sp. Fil. 5:1. 1864, non L. (1758). ?Polypodium trichosorum Hook. Second Cent., pl. 12. 1860. Type: Forests of Archedona, Andes of Quito, Ecuador, Jameson 349. Glyphotaenium crispatum J. Smith, Hist. Fil. 77, pl. 2. 1872. ?Polypodium ecostatum Sodiro, Crypt. Vase. Quit. 308. 1893. Type: Forests of Los Colorados, Ecuador, Sodiro. Polypodium goniopteroides C. Chr. Ind. Fil. 188. 1905; 530. 1906. Based on Clenopteris crispata J. Smith, non P. crispatum L. (1753). MORTON—GENUS GRAMMITIS 101 Range: Southern Panama and Colombia, probably south to Ecuador. Habitat: Epiphytic. If the above synonymy is correct, this species is represented in Ecuador by the types of two supposed new species, neither of which was originally associated with Ctenopteris crispata J. Smith. In the “Index Filicum” both are recognized as species. Copeland suggested that Polypodium ecostatum might be the same as his Glyphotaenium erispatum. 14. Grammitis semiadnata (Hook.) Morton, comb. nov. Polypodium semiadnatum Hook. Icon. Pl. 10, pl. 948. 1854. Type: Pilzhum, Andes of Quito, Ecuador, 3,600 m., Jameson (K, not seen). Ctenopteris semiadnata Copel. Phil. Journ. Sci. 84:458. 1955. Range: Known definitely only from Ecuador. The Colombian specimens referred here by Copeland may be different. Habitat: Epiphyte, or pendent on sides of sphagnum hummocks, at elevations from 3,000 to 3,600 m. Ecuadorian specimens examined: Papallacta, Prov. Pichincha, 3,400 m., 1918, Mille. San Gabriel, Prov. Carchi, ‘in silva prim. pr. reg. ‘pdramo’,” Holmgren 902. Boggy ciénagas 1-4 miles south of Las Toldadas, east of Cayambe Peak, Prov. Imbabura, 3,090 m., Wiggins 10511. 15. Grammitis intricata Morton, sp. nov. PLaTE 3 Planta epiphytica; rhizoma erectum, breve, ca. 1 cm. longum, vix 2 mm. diam., paleaceum, paleis fuscis lanceolatis ca. 1.8 mm. longis, 0.4 mm. latis, 9 vel 10 cellulis latis, subobtusis, margine paullo glandu- loso-pilosis, pilis brevibus crassis, obtusis, ca. 50 yp longis, 20 yu latis, clathratis, cellulis anguste rectangularibus, ca. 100 4 longis, 40 u latis, parietibus crassissimis, luminibus angustis; frondes subfascicula- tae, 25-30 cm. longae, laxae pendulae, pinnatae, stipitibus quam laminis multo brevioribus, gracilibus fuscis, 3-7 cm. longis, 0.5-0.7 mm. diam., deorsum teretibus, sursum vix alatis, parce setulosis, setis gracillimis patentibus 1.5-2 mm. longis unicellularibus, pilis 2-vel 3-septatis minutis ca. 0.5 mm. longis apice glanduliferis sparse intermixtis, rhachi gracili basi non alata, apice vix alata, fusca, supra dense brunneo-setosa, setis usque ad 3 mm. longis unicellulari- bus, subtus setosa atque capitato-glandulosa, pilis minutis septatis; laminae lineares, 20-26 cm. longae, 13-20 mm. latae, fere ubique pinnatae, basi paullo decrescentes, apice obtusae non pinnatifidae; pinnae 28-35-jugae, alternae, ovatae, 8-10 mm. longae, 5-7 mm. latae, horizontaliter patentes, basales distantes semiadnatae, superiores ubique adnatae, apice rotundatae, margine integrae vel leviter undula- tae, textura herbaceae, vix crassae, supra setis paucissimis elongatis 102 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM marginem et apicem versus praeditae, subtus glabrae, margine parce longe setoso-ciliatae; venae immersae obscurae, costa flexuosa, venulis 3-jugis, gracilibus, omnibus infra medium furcatis vel basalibus bis furcatis, liberis, marginem non attingentibus; sori superficiales, 3-jugi, apicales in ramulo antico venarum, magni rotundi ca. 1.75 mm. diam., non setosi, non paraphysati; annulus ex ca. 15 cellulis compositus; sporae sphaerico-tetraedricae triletae, ca. 33 » diam., fere laeves. Type in the U.S. National Herbarium, no. 2,204,844, collected in the Cordillera Oriental, at Guamani Pass, east of Pifo, Province of Pichincha, Ecuador, Nov. 10, 1944, at 3,900 m. elevation, by Joseph A. Ewan (no. 16436). Ewan’s note on the label says: “In rather deep shade of bosque thickets, the old fronds persisting to form a heavy mat beneath the current live fronds, serving to hold moisture.” Grammitis semiadnata, definitely known only from Ecuador, is a close relative. The shape and pubescence is distinctive. Copeland places this species in this group “meridensis,” but it is hardly well placed there, for the frond is not truncate at the base, as that group is characterized. 16. Grammitis heteromorpha (Hook. & Grev.) Morton, comb. nov. Polypodium heteromerphum Hook. & Grev. Icon. Fil. 1, pl. 108. 1829. Type: Mount Pichincha, Prov. Pichincha, Ecuador, Jameson K! Ctenopteris heteromorpha Copel. Phil. Journ. Sci. 84:412. 1955. Range: Mexico to Peru. Habitat: In Ecuador rupicolous at higher elevations, 3,500 to 4,500 m., in Colombia epiphytic on trees on the borders of p&ramos, 3,000 to 3,600 m. Eeuadorian collections examined: Voleén Antisana, Prov. Pichin- cha, 4,000-4,500 m., Anthony & Tate 282 (US). Mount Pichincha, Prov. Pichincha, 4,000 m., Mille (US); ibid., 4,500 m., Stuebel 730 (US); ibid., 4,400 m., Firmin 576. Mount Carihuayrazo, Prov, Tunguragua, 4,200 m., Asplund 8479 (US). Cayambe Peak, Prov. Imbabura, 4,050-4,350 m., Little & Parades 6837 (US). Cubillin, Valley of Alao, 3,500 m., Rimbach 76 (US). Without specific locality, Sodiro, Mille, Couthouy (all US). 17. Grammitis variabilis (Mett.) Morton, comb. nov. Polypodium variabile Mett. ex Kuhn, Linnaea 36:133. 1869. Syntypes: New Granada, PurdieK!, and Ecuador, JamesonK! Range: Colombia to Peru. Habitat: On rocky cliffs in pféramos, 3,000 to 4,200 m. elevation. Keuadorian specimens examined: Mount Atacatzo, Prov. Pichin- cha, 3,000 m., Afille (US). Although reduced to a synonym of C. heteromorpha by Copeland, this species appears recognizable to me, differing in having the blades not forked, but rather regularly bipinnate. MORTON—GENUS GRAMMITIS 103 18. Grammitis alfarii (Donn. Smith) Morton, comb. nov. Polypodium alfarit Donn. Smith, Bot. Gaz. 33:262. 1902. Type: Sierra Alta de Navarro, Province of Cartago, Costa Rica, 2,000 m., May, 1901, Alfaro 73 (Donn. Smith 8063) (US). Polypodium oligosorum Mett. ex Kuhn, Linnaea 36:132. 1869, non Klotzsch (1847). Syntypes: Venezuela, Moritz 460, Fendler 208, Karsten 10. Rhizome epiphytic, short, slender, ca. 1 mm. in diameter, scaleless; fronds subfasciculate, crowded, 8-10, 8-13 cm. long; stipes 10-14 mm. long, very slender, only 0.1-0.2 mm. in diameter, blackish, densely long-setose; blades linear, 8-15 mm. broad, attenuate to base and apex, light green, very thin and flaccid in texture, pinnatisect, the rhachis dark above and glabrous, paler beneath and sparingly setose; segments 16-38 pairs, alternate, 3-8 mm. long, 3-5 mm. broad at base, subtriangular, the apex obtuse or rounded, the base fully adnate but twice as long adnate on the superior side as on the inferior, rounded at the upper base, short-decurrent at the lower, the lower margin somewhat sigmoid, the upper gibbous with a conspicuous hump at the base extending about half the length of the pinna, the surfaces glabrous, the margin with sparse long (0.4-0.5 mm.), dark brown setae; veins dark, near the lower surface, easily visible, not nearly reaching the margin, the midvein with 2 branches on the upper side, the lower branch elongate and extending up into the hump, sterile, the upper reduced to a short, fertile spur, on the lower side with only one branch, a short fertile spur; white lime dots not present on hydathodes above; sori 2 to a pinna normally, subcostular, borne on the apices of the short distal and proximal spurs, rotund; sporangia numerous, long-setose, the brown, acicular setae borne near the sides of the annulus. Range: Costa Rica, Venezuela, and Ecuador, at middle elevations, 950-2,700 m. Ecuadorian specimen examined: Eastern slopes of the cordillera, valley of Rio Negro, near the junction of Rio Negro and Rfo Pailas, Prov. Santiago-Zamora, Prieto (Camp E-4904) (NY, US). In its peculiar aspect, with the blade segments much more strongly adnate above than below and with a conspicuous hump on the upper side at the base, thiss pecies is similar to Grammitis phlegmaria (J. Smith) Morton, comb. nov. [Polypodium phlegmaria J. Smith, London Journ. Bot. 1:194. 1842], which species differs strongly, however, in the presence of rhizome scales, and in the nonsetiferous margins of the segments and the glabrous sporangia. 19. Grammitis senilis (Fée) Morton, comb. nov. Polypodium senile Fée, Mém. Foug. 7:60, pl. 25, fig. 1. 1857. Type: Ocafia, Dept. Santander, Colombia, Schlim 364. Cf. Maxon, Contr. U.S. Nat. Herb. 13:43. 1909. 238-617—67——4 104 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM Ctenopteris senilis Copel. Phil. Journ. Sci. 84:398. 1955. Range: Guatemala to Ecuador; also Cuba, and Hispaniola(?). Habitat: Pendent epiphyte, at 1,700 to 3,200 m. elevation, or oc- casionally at lower altitudes. Ecuadorian specimen examined: Valley of Rio Sangarinas, Cordi- llera de Llanganates, Prov. of Tunguragua, 3,000 m., Asplund 9764 (US). 20. Grammitis dependens (Baker) Morton, comb. nov. Polypodium dependens Baker, in Hook & Bak. Syn. Fil. 335. 1868. Type: Mount Pichincha, Ecuador, Spruce 5637 *! (photograph Maxon 146, U8). Range: Known definitely only from Prov. Pichincha, Ecuador. Probably also in Peru and Bolivia. Habitat: Pendent epiphyte, probably from high elevations. Ecuadorian specimens examined: Mount Pichincha, Mille 6 (US). Cerro Corazén, Sodiro 48/20 (fragment US). Although reduced to a synonym of C. heteromorpha by Copeland, it appears to me to be distinct, and probably more closely allied to G. sericeo-lanatum. Apparently, it never has forked fronds like G. heteromorpha, and the segments are different in outline. 21. Grammitis subtilis (Kunze ex Klotzsch) Morton, comb. nov. Polypodium subtile Kunze ex Klotzsch, Linnaea 20:375. 1847. Type: Mérida, Venezuela, Moritz 325 (isotype US). Ctenopteris subtilis J. Smith, Hist. Fil. 184. 1875. Range: Guatemala to Ecuador. Habitat: Epiphyte in forest at elevations from 1,500 to 2,400 m. Ecuadorian specimens examined: Alégria, east of Voledén de Cayambe, Prov. Imbabura, 2,900 m., Drew E-170 (US). 22. Grammitis subflabelliformis (Rosenst.) Morton, comb. nov. Polypodium sub flabelliforme Rosenst. Repert. Sp. Nov. Fedde 7:306. 1909. Type: Mount Abitagua, Ecuador, Spruce 5271 (isotype US). Polypodium subflabelliforme var. minor Rosenst. op. cit. 307. Type: Mount Tunguragua, Ecuador, Spruce 5272 (not seen). Ctenopteris subflabelliformis Copel. Phil. Journ. Sci. 84:400. 1955. Range: Known in Ecuador only from the types; reported from Peru and Bolivia by Copeland. Habitat: Epiphyte. 23. Grammitis lehmanniana (Hieron.) Morton, comb. nov. Polypodium lehmannianum Hieron. Bot. Jahrb. Engler 34:513. 1904. Syntypes: Near Cuaiquer and San Pablo, Cordillera de Pasto, Dept. Narifio, Colombia, 1,000-1,300 m., Lehmann 19 (US); between Bajios and Jivaria de Piutuc, Pastaza Valley (prob. Prov. Tunguragual], 41 Originally published as 563, in error. MORTON—GENUS GRAMMITIS 105 Ecuador, Stuebel 1011. Lectotype: Lehmann 19, inferentially selected by Maxon, Proc. Biol. Soc. Washington 57:18. 1944. Polypodium pastoense C. Chr. Ind. Fil. 551. 1906. Illegitimate, an un- necessary renaming of P. lehmannianum Hieron., non P. lehmannii Mett. Polypodium sublongipes Christ, Bull. Soc. Bot. Genéve II, 1:218. 1909. Type: Costa Rica, Wercklé in 1904 (P). Synonymy sec. Maxon, loc. cit. Range: Guatemala to northern Ecuador. Habitat: Epiphyte at low elevations, 300—1,300 m. Ecuadorian specimens examined: Puyo, Prov. Napo-Pastaza, 300 m., Mexia 6922 (US). Although he saw and annotated the specimens, the North American range was overlooked by Copeland in his published account, as was also the synonym P. sublongipes. 24. Grammitis laxa (Presl) Morton, comb. nov. Polypodium laxum Presl, Rel. Haenk. 1:23, pl. 4, fig. 1. 1825. Type: Mountains of Peru, Haenke (not seen). Range: Peru and Bolivia; not as yet known in Ecuador, but of possible occurrence. Habitat: Epiphyte at high elevations. Although made a synonym of C. lanigera by Copeland, this seems distinguishable to me. I am taking as representative: La Convencidén, Dept. Cuzco, Peru, 4,500-4,750 m., Bies 2130 (US) ; between Hudnuco and Pampayacu, Peru, Kanehira 122 (US); and Yungas, Bolivia, Bang 483 (US); the last named was identified as lazum by Hieronymus. 25. Grammitis lanigera (Desv.) Morton, comb. nov., var. lanigera Polypodium lanigerum Desv. Ges. Naturf. Freund. Berlin Mag. 5:316. 1811. Type: Peru, Dombey (P, photograph Cintract, US). Polypodium sericeo-lanatum Hook. Sp. Fil. 4:221. 1864. Syntypes: Pichincha, Ecuador, 12,000 ft., and woods near Bajios, Jameson 235, 73, 29, 394; mountains of Guayrapata, Ecuador, Spruce 5277 K!; Oca- fia, Dept. Santander, Colombia, 10,000-11,000 ft., Schlim 313 (indi- cated by Hooker as aberrant). Lectotype: Ravines of Pichincha, 11,000-12,000 ft., Jameson 235 (K, fragment US). Polypodium alternifolium Hook. Sp. Fil. 4:222. 1864, non Willd. (1810). Syntypes: Near Esmeraldas and between Cuenca and Guayaquil, Ecuador, 3,000-10,000 ft., Jameson (K, photograph US), Hartweg 1496 (photograph US). Lectotype: the Jameson collection (K, fragment US). Polypodium longum C. Chr. Ind. Fil. 541. 1906. New name for P. alternifolium Hook., non Willd. Ctenopteris lanigera Copel. Phil. Journ. Sci. 84:420. 1955. Ctenopteris sericeo-lanata Copel. op. cit. 453. Ctenopteris longa Copel. loc. cit. Grammitis sericeo-lanata Proctor, Rhodora 63:35. 1961. 106 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM Range: Costa Rica to Bolivia; Hispaniola, Martinique. Habitat: Epiphyte (or perhaps occasionally rupicolous) in cool forests, in Ecuador at elevations from 2,500 to 3,300 m. The altitude ‘3,000 feet”’ given by Hooker in the original description of P. alter- nifolium was surely an error. Ecuadorian specimens examined: Mount Pichincha,Prov. Pichincha, August 1901, Sodiro (US) (this collection bears a note by Weatherby: “Like the type of P. lanigerum’’; it is also a topotype of P. sericeo- lanatum); ibid., Mille 10, 11, 382, s. n. (US). Mount Tunguragua, Prov. Tunguragua, Rimbach 126 (US); ibid., Rosenstock, Fil. Ecuad. 33 (US). Huagrarancha, Prov. Azuay, “Ilashipilla,” Steyermark 53415 (US). Pan-American Highway 40 km. south of Cuenca, Prov. Azuay, Wiggins 10768 (US). Bafios, Prov. Tunguragua (probably), Jameson 29 (K, syntype of P. sericeo-lanatum, photograph and fragment US). Although Copeland recognized all three species C. lanigera, C. sericeo-lanatum, and C. longa as distinct, and even placed the first named in a different one of his “groups,” I am unable to distinguish them. There is certainly a good deal of variation, but not more than to be expected. 25a. Grammitis lanigera (Desv.) Morton var. stella (Copel.) Morton, comb. nov. Clenopteris stella Copel. Phil. Journ. Sci. 84:452. 1955. Type: Valley of Rfo Urubamba, Dept. of Cuzco, Peru, Biies A32 (US). Range: Colombia to Peru. Habitat: Epiphyte, at elevations from 2,750 to 3,500 m. Kcuadorian specimens examined: Pichincha, Prov. Pichincha, Sodiro (US). Paramba, August 1904, Sodiro (US). Las Torres, Cordillera de Llanganates, Prov. Tunguragua, Asplund 9814. North of Pdéramo del Castillo, Prov. Azuay, Camp E-5164. La Floresta, south of Sigsipamba, on Rio Blanco, Prov. Imbabura, Wiggins 10298. For some reason, not stated and not clear, Copeland placed his new species C. stella in a different group from either C. lanigera or C’. sericeo-lanata, thus having essentially the same species in three different groups. In the group in which Copeland placed it [the “Group of Ctenopteris suspensa’’], C. stella would indeed be distinct, but it is hardly different from typical sericeo-lanata, which normally has some of the hairs along the rhachis beneath stellate. The var. stella differs only in having these hairs darker (and thus more obvious) and perhaps often somewhat longer stalked. 26. Grammitis cultrata (Willd.) Proctor, Rhodora 63:35. 1961. Polypodium cultratum Willd. Sp. Pl. 5:187. 1810. Type: Jamaica, Swartz (B, Herb. Willd. 19674, photograph US). MORTON—GENUS GRAMMITIS 107 Polypodium elasticum Bory ex Willd. Sp. Pl. 5:183. 1810. Type: Bourbon [Réunion], Bory. Polypodium cultratum var. minus Tlook. Sp. Fil. 4:190. 1864. Based on P. elasticum Bory. “Polypodium cultratum P. elasticum’’ Baker, in Hook. & Bak. Syn. Fil. 190. 1868. (Invalid form of varital publication in binomial form.) Ctenopteris elastica Copel. Phil. Journ. Sci. 84:426. 1955. Range: Mexico to Peru; Greater Antilles. Habitat: Epiphytes, at elevations from 1,350 to 3,300 m. Ecuadorian specimens examined: Rio Palora, Prov. Napo-Pastaza, 1,400 m., Rimbach 68 (US). Cordillera Cutuct, Prov. Santiago- Zamora, 1,350-1,650 m., Camp E-1322 (US). North of Péramo del Castillo, Prov. Azuay, 3,000-3,300 m., Camp E-5167 (US). Copeland said that if P. cultratum and P. elasticum are the same the “prior and valid’ specific epithet is elastica, but this is not true. Both elasticum and cultratum were published at the same time in the same work; by “‘prior’’ in such instances Copeland meant “page priority,”’ which has nothing to do with real priority. Copeland never followed the rule in the Code for the determination of the correct specific epithet in names of the same date [1961 ed., Art. 57], namely: “The author who first unites taxa bearing epithets of the same date has the right to choose one of them, and his choice must be followed.” Hooker in 1864 in reducing P. elasticum to a variety of P. cultratum (as var. minus) effectively made a choice between the two epithets, and he must be followed, as he has been by Christensen and others. No other author before Copeland (1955) ever made the reverse choice of adopting elasticum and reducing cultratum to synonymy. 27. Grammitis subsessilis (Baker) Morton, comb. nov. Polypodium pteropus Hook. Sp. Fil. 4:192, pl. 275B. 1864, non Blume, 1828. Syntypes: Andes of Quito, Jameson 348, Spruce 5712; Mount Abitagua, Spruce; Mount Roraima, Schomburgk; ‘‘New Granada.” Hartweg 1495. Since I have not seen the original collections, I refrain from selecting a lectotype. Polypodium subsessile Baker, in Hook. & Bak. Syn. Fil. 329. 1868. New name for P. pteropus Hook. non Blume. Ctenopteris pteropus J. Smith, Hist. Fil. 185. 1875. Polypodium euchlorum Kunze ex Hieron. Bot. Jahrb. Engler 34:509. 1904. Type: Must be considered an illegitimate renaming of P. subsessile Baker, cited in synonymy, although nominally based on P. euchlorum Kunze ex Klotzsch, Linnaea 20:375. 1847 (nom. nud. in syn.). Ctenopteris subsessilis Copel. Phil. Journ. Sci. 84:411. 1955. Range: Costa Rica to Bolivia. Habitat: Epiphyte in dense forests at elevations from 1,100 to 2,700 m. (or at low elevations of 200 m. in the Chocé region of Colombia). Ecuadorian specimens examined: Chiguinda, East Andes of Sigsig, Prov. Azuay, 1,800 m., Lehmann 6517 (US). Rio Palora, Prov. Napo- 108 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM Pastaza, 1,500 m., Rimbach 74 (US). Bobonaza, Prov. Napo-Pastaza, 1,900 m., Fuller 116 (US). Rfo Negro, Prov. Santiago-Zamora, Prieto (Camp E-4903) (US). Cordillera Cutuci, Prov. Santiago- Zamora, 1,400—1,500 m., Camp E-1169 (US). Eastern Cordillera, 1-8 km. north of Sevilla de Oro, Prov. Azuay, 2,400-2,700 m., Camp E-4462 (NY). 28. Grammitis jamesonioides (Féc) Morton, comb. nov. Polypodium jamesonioides Fée, Mém. Foug. 7:59, pl. 21, fig. 4. 1857. Type: Ocafia, Dept. Santander, Colombia, Schlim 399 (not seen). ?Polypodium azuayense Sodiro, Crypt. Vasc. Quit. 323. 1893. Type: Quinoas, Prov. Azuay, Ecuador, Rimbach 18 (not seen). Range: Costa Rica to Ecuador; Hispaniola. Habitat: Pendent epiphyte, at elevations from 2,600 to 3,600 m. Ecuadorian specimens examined: Mount Tunguragua, Sodiro (US). Huagrarancha, south of El Pan, Rio Collay, Prov. Azuay, Steyermark 53380 (US). Harta-Naqua, Espinosa 1021 (US). The material from Hispaniola, Costa Rica, and Panama needs to be studied further, for it is not identical with that of South America. 29. Grammitis pseudocapillaris (Rosenst.) Morton, comb. nov. Polypodium pseudocapillare Rosenst. Meded. Rijksherb. Leiden 19:17. 1913. Type: Above Tablas, Herzog 2190a (isotype US). Ctenopteris pseudocapillaris Copel. Phil. Journ. Sci. 84:407. 1955. Range: Colombia to Bolivia. Habitat: Epiphyte, 3,000 to 3,400 m. elevation. Ecuadorian specimens examined: Hoyada de Gal&pagos, between Huagrarancha and Loma de Galdépagos, Prov. Azuay, Steyermark 53475 (US). Péramo area north of Péramo del Castillo, Prov. Azuay, Camp E-5140 (US). 30. Grammitis myriophylla (Mett.) Morton, comb. nov. ?Polypodium longisetosum Hook. Sp. Fil. 4:225. 1864. Type: Andes of Quito, Ecuador, Jameson 79 (not seen). Polypodium myriophyllum Mett. ex Hook. & Bak. Syn. Fil. 338. 1868.2? Type: Tatanara, Peru, Lechler 2567 (fragment US and Morton photo, 1931). Ctenopteris myriophylla Copel. Phil. Journ. Sci. 84:412. 1955. Range: Ecuador to Bolivia. Habitat: Pendent epiphyte, at elevations from 2,800 to 3,400 m. Ecuadorian specimens examined: Trail between Rio Clavadero and ridge 10 miles south of Las Toldadas, east of Cayambe, Prov. Imbabura, 3,400 m., Wiggins 10498 (US). 2 Cited by Copeland as “‘Mettenius, Fil. Lechler. I (1845) 6,’”’ but it is a nom. provis. at this place, the collection being referred to the Javan species P. tenuisectum Blume, with the remark that it might prove distinct. MORTON—GENUS GRAMMITIS 109 31. Grammitis major (Copel.) Morton, comb. nov. Polypodium tenuiculum var. acrosora Hieron., Bot. Jahrb. Engler 34:511 1904. Type: In dense forest around Yerbabuena, western slopes of the West Andes of Cuenca, Prov. Azuay, Ecuador, 2,500-2,900 m., Lehmann 5727 (isotype US). Ctenopteris major Copel. Phil. Journ. Sci. 84:455. 1955. Based on P. tenuiculum var. acrosora Hieron. Range: Known only from the type. Habitat: Epiphyte. Although Copeland’s new species C. major is based solely on P. tenuiculum var. acrosora Hieron., on another page of his work (1955, p. 440) he lists this variety also under C. tenwiewla, and adds the in- explicable comment: ‘“The Herb. Univ. Calif. are: Brade 6465 (iso- type of var. brasiliense Ros.) and Luetzelburg 6726, both from Rio. Veins simple,. . .”’ This statement is inexplicable because this species [or variety] does not occur in Brazil. 32. Grammitis tunguraguae (Rosenst.) Morton, comb. nov. Polypodium tunguraguae Rosenst. Repert. Sp. Nov. Fedde 7:307. 1909. Type: Mount Tunguragua, Prov. Tunguragua, Ecuador, August 1857, Spruce (fragment of holotype US). Ctenopteris tunguraguae (Rosenst.) Copel. Phil. Journ. Sci. 84:394. 1955. Range: Known only from the type. Habitat: Mossy slopes, presumably terrestrial. 33. Grammitis anfractuosa (Kunze ex Klotzsch) Proctor, Rhodora 63:35. 1961. Polypodium anfractuosum Kunze ex Klotzsch, Linnaea 20:375. 1847. Type: Mérida, Venezuela, Moritz 330 (isotype US). Polypodium saxicola Baker, Journ. Bot. Brit. & For. 15:264. 1877, non Swartz (1817). Type: ‘No. 84, Herb. Kew. 1877” [coll. Jenman?]. (Not seen.) Polypodium induens Maxon, Bull. Torrey Bot. Club 32:75. 1905. Based on P. saxicola Baker, non Swartz. Range: West Indies; Guatemala to Peru. Habitat: Normally epiphytic, but occasionally on humus-covered rocks or banks, in Ecuador from 1,225 to 2,195 m. elevation. Ecuadorian specimens examined: At confluence of Rio Mapoto and Rio Pastaza, Prov. Tunguragua, Penland & Summers 207. Junction of Rio Barbara and Rfo Arenillas, Prov. Santiago-Zamora, Steyermark 53612 (US). 34. Grammitis taxifolia (L.) Proctor, Rhodora 63:35. 1961. Polypodium taxifolium L. Sp. Pl. 1086. 1753. Type: There is no specimen in the Linnaean Herbarium, and consequently the species must be considered to be based on the cited synonym Polypodium tenue et pendulum Plum. (Tract. Fil. Amer. 69, pl. 89. 1705), based on a plant from Morne de la Calabasse, Martinique, Plumier. Topotypes in the National Herbarium (Calabasse, Duss 1575, 1674, 4573) agree 110 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM with the description and figure of Plumier. [The distributor of the Duss duplicates united different numbers of the collection representing the same species, and so it is impossible to tell which of the three numbers the collection represents.] Ctenopteris taxifolia Copel. Phil. Journ. Sci. 84:447. 1955. Range: West Indies; continental range uncertain, although reported from Costa Rica to Brazil and Bolivia. Habitat: In Ecuador, ‘‘on rotting logs, bases of trees, and soil,” at 1,500 m. elevation. Normally epiphytic at middle elevations. Ecuadorian specimens examined: Dense jungle, El Topo, south side of Rio Pastaza, between Bafos and Mera, Prov. Tunguragua, Prescott & Wiggins 21 (US). Quite similar in aspect to G@. semihirsuta, differing in the nonciliate rhizome scales and in the setulose lower surface of the segments. 35. Grammitis firma (J. Smith) Morton, comb. nov. Polypodium firmum Klotzsch, Linnaea 20:378. 1847, non Kaulf. (1827).?3 Syntypes: ‘In alpibus Chile, Philippi; Guiana angl. Rich. Schomburgk 1170.” Looser (Revista Universitaria [Chile] 361!:75. 1951) chose the Schomburgk collection as lectotype, as others have before him inferentially (e.g. Maxon, Proc. Biol. Soc. Washington 52:119. 1939), on the basis that the Chilean record is doubtful, perhaps erroneous, since the species has never again been found in Chile. Clenopteris firma J. Smith, Hist. Fil. 184. 1875. Considered a new name for P. firmum Klotzsch, non Kaulf. Polypodium aromaticum Maxon, Proc. U.S. Nat. Mus. 27:743. 1904. Type: Blue Mountain Peak, Jamaica, 1,950-2,225 m., Underwood 1449 (fragment US). Ctenopteris aromatica Copel. Phil. Journ. Sci. 84:418. 1955. (Illegit., because the earliest legitimate epithet firma was not adopted.) Range: Jamaica: Guatemala to Bolivia (and Chile?). Habitat: Epiphyte. Not known as yet from Ecuador, but since it is found both north and south, it is likely to be found there. 36. Grammitis vernicosa (Copel.) Morton, comb. nov. Ctenopleris vernicosa Copel. Phil. Journ. Sci. 84:452. 1955. Syntypes Alto de la Estrella, Costa Rica, Standley 39140 (US). “Canaan,”’ Mount Puracé, Dept. Fl Cauca, 3,100-3,300 m., Pennell & Killip 6638 (US). Western Cordillera, Ecuador, 2,500 m., Rimbach 303 (US). Lectotype: Standley 39140 (although not so indicated in the published work, Copeland indicated the Standley specimen as type on the sheet). Range: Costa Rica to Ecuador. Habitat: Epiphyte, at elevations from 2,500 to 3,300 m. Additional Ecuadorian specimens examined: Outskirts of Pifo, Prov. Pichincha, 2,500 m., Mezia 7353a. 23 P. firmum Kaulf. is wrongly said by Copeland to be a nomen nudum. It is a legitimately published renaming of P. attenuatum R. Brown, non Willd., which was previously renamed P. brownii Wikstr. (1825). MORTON—GENUS GRAMMITIS 111 37. Grammitis pichinchensis (Hieron.) Morton, comb. nov. Polypodium subscabrum sensu Hook. Sp. Fil. 4:183, pl. 274A. 1864, pro parte, non Klotzsch (1847). Polypodium pichinchense Hieron. Bot. Jahrb. Engler 34:506. 1904. Syn- types: Ecuador, in declivibus occasum solis spectantibus ad truncos arborum, alt. s.m.c. 3,000-3,400 m., May 1862, Jameson; in regione urbis Quito, Cuming 32 p.p. Lectotype: Jameson (B, probable iso- syntype US). Polypodium ecuadorense C. Chr. Ind. Fil. 524. 1906. Illegitimate re- naming of P. pichinchense Hieron., non P. pichinchae Sodiro. The epithets, although similar, must be considered different under the Code. Ctenopteris ecuadorensis Copel. Phil. Journ. Sci. 84:434. 1955. (Illegit.) Range: Known only from Ecuador. Habitat: Epiphyte, at elevations from 2,800 to 3,500 m. Ecuadorian specimens examined: Sin. loc., 3,000 m., Sodiro. Pichincha, Prov. Pichincha, March, 1903, Sodiro. Mount Tungu- ragua, Prov. Tunguragua, Rimbach 26, 129; ibid. Rosenst. Fil. Ecuad. 37. Sin. loc., 2,800 m., August 1918, Mille 4; ibid., 3,180 m., March 8, 1921, Popenoe. Volcéin Atacatzo, 3,200 m., 1918, Mille; ibid., 3,400 m., 1919, Mille. “San Miguel,” margin of pdéramo between Atuntiaqui and Hacienda Pifion, Prov. Imbabura, Wiggins 10325. ‘Tipicocha, border of Prov. Chimborazo and Cafiar, Camp E4075. Péramo north of Péramo del Castillo, Prov. Azuay, Camp K-5183. Polypodium ecuadorense is recognized on page 434 of Copeland’s treatment as a valid species (and transferred to Ctenopteris) and on page 455 it is reduced to the synonymy of C. pichinchae, another instance of the extreme carelessness of Copeland’s work. Since it is actually not very close to C. pichinchae, he very likely intended to recognize it as distinct but was misled by the similarity of the names P. pichinchae and P. pichinchense. 38. Grammitis cuencana (Hieron.) Morton, comb. nov. Polypodium cuencanum Hieron. Bot. Jahrb. Engler 34:505. 1904. Type: Near Chagal and Yerbabuena, Western Cordillera of Cuenca, Prov. of Azuay, Ecuador, Lehmann 5728 (isotype US). Ctenopteris cuencana Copel. Phil. Journ. Sci. 84:396. 1955. Range: Known only from the type. Habitat: Epiphytic in dense forest, at 2,400 to 2,800 m. elevation. 39. Grammitis pichinchae (Sodiro) Morton, comb. nov. Polypodium pichinchae Sodiro, Crypt. Vase. Quit. 329. 1893. Type: Mount Pichincha, Prov. Pichincha, Ecuador, Sodiro (not seen). Ctenopteris pichinchae Copel. Phil. Journ. Sci. 84:455. 1955. Range: Known only from northern and eastern Ecuador. Habitat: Epiphyte, at elevations of 3,000 to 3,200 m. 112 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM Ecuadorian specimens examined: Papallacta, Prov. Pichincha, December 1923, Mille. Mount Tunguragua, Prov. Tunguragua, August 1903, Sodiro; ibid., Rimbach 127a, 131, 13la. Between Rio Clavadero and ridge 10 miles south of Las Toldadas, east of Cayambe, Prov. Imbabura, Wiggins 10495. Junction of Rio Itzintza and Rio Chupiasa, Cordillera Cutuct, Prov. Santiago-Zamora, 3,500 ft., Camp E-1394 (NY). 40. Grammitis attenuatissima (Copel.) Morton, comb. nov. Ctenopteris altenuatissima Copel. Phil. Journ. Sci. 84:456. 1955. Type: Mount Tunguragua, Prov. Tunguragua, Ecuador, December 1904, Sodiro (orig. distr. as P. taxifolium L.) (US). Range: Known only from Ecuador. Habitat: Epiphyte, at elevations from 3,200 to 3,300 m. Additional Ecuadorian specimens examined: Western Cordillera of Riobamba, Prov. Chimborazo, Rimbach 22. Rio Luguibi [?], Rim- bach 113a. 41. Grammitis leucosticta (J. Smith) Morton, comb. nov. Polypodium leucosticta Fée, Gen. Fil. 240. 1852, non P. leucosticton Kunze ex Klotzsch (1847). Type: Quito, Ecuador, Jameson (possible iso- type US, Jameson, sin. loc.) Ctenopteris leucosticta J. Smith, Hist. Fil. 184. 1875. To be considered a new name for P. leucosticta Fée, non Kunze, rather than a transfer. Polypodium longiusculum C. Chr. Ind. Fil. 541. 1906. New name for P. leucosticta Fée. Ctenopteris longiuscula Copel. Phil. Journ. Sci. 84:456. 1955. (Illegit., being superfluous, the name C. leucosticta J. Smith being available.) Range: Ecuador and Peru. Habitat: On banks at middle and high elevations. Ecuadorian specimens examined: [Volcén] Pichincha, Prov. Pi- chincha, 1864, Jameson (US); ibid., Sodiro 26/54; ibid., April 1921, Mille 12; ibid., September, 1890, Sodiro. [Cerro] Igualata, Prov. Tunguragua, 2,800 m., July 1915, Mille. Along the Rio Matadero, west of Cuenca, Prov. Azuay, 10,000-10,500 ft., Mar. 3, 1945, Camp E-2001 (NY). 42. Grammitis alsopteris Morton, sp. nov. PLATE 4 Planta epiphytica; rhizoma ascendens, breve, crassum, apice paleaceum, paleis brunneis vix rigidis, anguste lanceolatis 3-5 mm. longis, medio ca. 0.5 mm. latis, acuminatis, apice piliformibus, ca. 13 cellulis latis, cellulis modice elongatis subrectangularibus 80-160 p longis, ca. 38 yw latis, parietibus modo crassis brunneis, margine evidenter ciliatis, ciliis usque ad 0.25 mm. longis patentibus vel superioribus plus minusve adscendentibus, gradatim acuminatissimis, pallide brunneis, unicellularibus; frondes ca. 5, subcaespitosae, usque ad ca. 50 cm. longae et 7 cm. latae; stipites 8-13 cm. longi, crassi, ca. 1.8 mm. diam., fusci, teretes, non alati vel costati, ubique MORTON—GENUS GRAMMITIS 113 dense setosi, setis patentibus rigidis brevioribus ca. 0.25 mm. longis, longioribus paucis ca. 1 mm. longis; laminae lanceolatae pinnatisectae basi vix angustiores vel interdum pinnis 2 vel 3 reductis, apice breviter acuminatae, rhachi crassa basi ca. 1.5 mm. diam., fusca, dense breviter setosa supra et infra, cum setis elongatis tenuibus ca. 2 mm. longis intermixtis; segmenta numerosa plus quam 50-jugata, alterna vel subopposita anguste elongata, 2-4.5 cm. longa, 3-3.5 mm. lata, margine leviter revoluta, non ciliata, apice rotundata vel vix acuta, basi adnata, lateribus subparallelis integris, supra pallide viridia glabra, hydathodis calcareis submarginalibus conspicuis praedita, subtus in costa setosa, in superficiebus minute setulosa, setulis brunneis patentibus rigidis vix 100-150 y longis; venae numerosae 20-30- jugatae, angulo acuto excurrentes, simplices; sori uniseriati sub- marginales rotundi in venulis terminales, non paraphysati; annulus ex 12 vel 13 cellulis compositus; sporae sphaerico-tetraedricae, ca. 36 » diam., superficebus levissime tuberculatae. Type in the U.S. National Herbarium, no. 1,691,374, collected on tree trunk at Hacienda La Mascota, Canton Mera, Province of Napo-Pastaza, Ecuador, Feb. 28, 1935, at 900 m. elevation, by Ynes Mexia (no. 7013) (originally distributed as Polypodium semihirsutum). The following Ecuadorian specimens, all more or less poor or frag- mentary, are probably referable here: Rio Savonilla and Zamora, eastern Andes of Loja, 600-1,200 m. alt., Nov. 1876, Lehmann LXXXV; Eastern Cordillera, between Ona and Rio Yacuambi, Prov. Azuay, 8,000-9,500 ft. alt., Sept. 10-19, 1945, F. Prieto P-217; Carmen, 7,500 ft. alt., Nov. 13, 1953-June 18, 1954, Tate 488. This species was included by Copeland in his Ctenopteris semihirsuta (Klotzsch) Copel., but it seems sufficiently distinct by the characters noted in the key. 43. Grammitis semihirsuta (Klotzsch) Morton, comb. nov. Polypodium semihirsutum Klotzsch, Linnaea 20:379. 1847. Type: Peru, Ruiz & Pavon (isotypes US). Range: Mexico to Bolivia and Brazil; Jamaica, Hispaniola. Habitat: Epiphyte, at elevations from 1,400 to 3,300 m. At the highest elevations, on the margins of péramos, said to be sometimes terrestrial. Ecuadorian specimens examined: Mount Tunguragua, Prov. Tunguragua, December 1904, Sodiro (US); ibid., 3,000 m., Popenoe (US). Mount Pichincha, Prov. Pichincha, 2,500 m., Mille 13 (US). Lloa Valley, near Quito, Mille (US). Ayapamba [prob. Prov. El Oro], Rose 23460 (US). Rio Palora, Prov. Napo-Pastaza, 1,400 m., Rimbach 69 (US). Margin of péramo between Atuntiaqui and Haci- enda Pifion, Prov. Imbabura, 3,300 m., Wiggins 10331 (US). 114 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM 44, Grammitis rigens (Maxon) Proctor, Brit. Fern Gaz. 9:219. 1965. Polypodium rigens Maxon, Proc. U.S. Nat. Mus. 27:741. 1904. Type: John Crow Peak, Jamaica, 1,650-1,800 m., Mazon 1346 (US). Ctenopteris rigens Copel. Phil. Journ. Sci. 84:422. 1955. Range: Greater Antilles; Guatemala to Peru. Habitat: Epiphyte on mossy trunks, at middle elevations, 1,500- 3,000 m. This species is not known definitely from Ecuador, but it is to be expected. It is very close indeed to Grammitis pilosissima (Mart. & Gal.) Morton, comb. nov. (Polypodium pilosissimum Mart. & Gal. Mem. Acad. Brux. 15:39, pl. 9, fig. 2. 1842) described from Mexico, which is, however, saxicolous or terrestrial rather than epiphytic, and which also has darker and more numerous setae on the surfaces and margins; however, its distinctness is by no means certain. 45. Grammitis pseudonutans (Christ & Rosenst.) Morton, comb. nov. Polypodium pseudonutans Christ & Rosenst. Repert. Sp. Nov. Fedde 5:15. 1908. Type: On trees, slopes of Mount Tunguragua, Prov. Tun- guragua, 3,000 m., Rimbach (isotype, US). Clenopteris pseudonutans Copel. Phil. Journ. Sci, 84:389. 1955. Range: Ecuador (Prov. of Carchi, Imbabura, and Tunguragua), and Colombia (Péramo del Tébano, Dept. Narifio, 3,200 m., Cuatre- casas 11920 (US). Habitat: Epiphyte in forest or at lower margins of pdéramos, 3,000 to 3,200 m. elevation. Ecuadorian specimens examined: Epiphytic, near lower margin of piramo, eastern slopes of Cayambe Peak, alt. 10,900 ft., Prov. Imbabura, Wiggins 10,407 (US). Small forms probably referable here are: tree trunk in dense forest, Hacienda La Rinconada, on trail between Moran and Olivos, Canton Espejo, Prov. Carchi, alt. 3,200 m., Mexia 7453 (US); base of tree ferns southeastern slopes of Volein de Chile, near Tufino, Prov. Carchi, alt. 11,200 ft., Wiggins 10,619 (US). Copeland did not clearly understand G. pseudonutans, which can be distinguished from G. sodiroi and other similar species by its thicker rhizome (2-2.5 mm. in diameter), rhizome scales without elongate, hairlike tips, stipe 2-ridged at summit above, lower pinnae much broader than long, setae not confined to sori, and so forth. Cuatre- casas 11920 and Wiggins 10407 were identified by Copeland as sodiroi, Wiggins 10619 as rigescens, and Mexia 7453 as megaloura, a quite different plant known otherwise only from the type from Guatemala. 46. Grammitis sodiroi (Christ & Rosenst.) Morton, comb. nov. Polypodium sodiroi Christ & Rosenst. Repert. Sp. Nov. Fedde 5:14. 1908. Type: On tree trunks, Mount Tunguragua, Prov. Tunguragua, Ecuador, 3,500 m., Rimbach 24 (isotype US). MORTON—GENUS GRAMMITIS 115 Range: Confined to Ecuador and known definitely only from Voledn Tunguragua, but two other somewhat aberrant specimens may be referable here: near Tipococha, Prov. Chimborazo-Cajiar border, Camp E-4076; Hacienda Pifion, Prov. Imbabura, Wiggins 10359. Habitat: Epiphyte in forest, 2,900 to 3,600 m. elevation. 47. Grammitis wolfii (Hieron.) Morton, comb. nov. Polypodium wolfii Hieron. Hedwigia 48:249. 1909. Type: Near Mindo, at base of Mount Pichincha, Prov. Pichincha, Stuebel 745 (not seen). Range: Known only from the type. This species was overlooked by Copeland. From the description, it appears to be a valid species of the group of @. moniliformis. 48. Grammitis flabelliformis (Poir.) Morton, Contr. U.S. Nat. Herb. 38:57. 1967 Polypodium flabelliforme Poir., in Lam. Encycl. Meth. 5:519. 1804 Polypodium rigescens Bory ex Willd. Sp. Pl. 5:183. 1810, Type: Bourbon [Réunion], Bory. Ctenopteris rigescens J. Smith, Hist. Fil. 184. 1875. Range: Réunion; high mountains of Kenya, Tanganyika, Uganda, Nyasaland, and Congo; Hispaniola (one known collection) ; Venezuela along the Andes to Peru (and Bolivia?) ; Brazil. Habitat: Perhaps variable. All the African material is epiphytic, including the type; all the American appears to be terrestrial or saxicolous, growing in péramos at elevations from 2,500 to 4,200 m. Ecuadorian specimens examined: Pichincha: Mount Pichincha, 2,500 m., Mille. Voledn Atacatzo, 3,800 m., Mille; ibid., 4,200 m., Firmin 539. Some of the American collections, including the Ecuadorian ones, appear to be slightly different from the African in having a more slender rhizome, less than 1 mm. in diameter, but this does not seem to hold in all regions. The species is exceedingly close to @. monili- formis, differing chiefly in having the segments more elongate, about twice as long as wide. 49. Grammitis peruviana (Desv.) Morton, comb. nov. Polypodium peruvianum Desv. Mém. Soe. Linn. Paris 6:231. 1827. Type from Peru. Ctenopteris peruviana J. Smith, Hist. Fil. 184. 1875. Range: Ecuador to Bolivia; reported from Brazil and Argentina by Copeland. Habitat: On rocks in péramos, 3,400 to 4,200 meters elevation. Ecuadorian specimens examined: On rocks around Huasi-huaico, in Andes west of Cuenca, Prov. Azuay, 3,000-3,200 m., Lehmann 5006 (US). 50. Grammitis moniliformis (Lagasca ex Swartz) Proctor, Brit. Fern Gaz. 9:219. Polypodium moniliforme Lagasca ex Swartz, Syn. Fil. 33. 1806. Type from Peru [Ruiz & Pavon ?]. 116 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM Polypodium subcrenatum Hook. Icon. Pl. 8: pl. 719. 1848 [1845 teste Ind Fil.]. Andes of Quito, Ecuador, Jameson 215. Ctenopteris moniliformis J. Smith, Hist. Fil. 184. 1875. Range: Mexico to Bolivia; Greater Antilles. Habitat: On mossy banks, at elevations from 2,300 to 3,900 m. Ecuadorian specimens examined: Cerro Villanaco, Prov. Loja, Wiggins 10967. Between Loja and Portovelo, Prov. Loja [?], Rose 23314. Headwaters of Rio Mangén, northeast of Azogues, Prov. Cafar, Mosberg & Prieto 22812. P&éramo in vicinity of Cafar, Prov. Canar, Rose 22757. Between Ona and Cuenca, Prov. Azuay, Hitch- cock 21627. Piramo Tinajillas, 44 km. south of Cuenca, Prov. Azuay, Wiggins 10783 (NY, US). Above Sayaus, east of Cuenca, Correll E360. Along Rio Matadero, west of Cuenca, Prov. Azuay, Camp E-1999. Péramos, vicinity of Toreador, between Molleturo and Quinoas, Prov. Azuay, Steyermark 53013. Nabén, Prov. Azuay, Rose 22999. Along Rio Saloya, between Voledén Atacasp and Volcén Pichincha, Prov. Pichincha, Steyermark 52492. Péramo, 25 km. southwest of Quito, Prov. Pichincha, Wiggins 10253. Mount Pichin- cha, Prov. Pichincha, 1921, Mille. Rio Silente, Voleén Corazén, Prov. Pichincha, Lehmann 408. Sin. loc., March 1921, Mille 5. This is the commonest species of the group in Ecuador, judging from the number of collections, and it is also the most widely dis- tributed. It is the one common truly terrestrial species, most of the others being either saxicolous or epiphytic. 51. Grammitis rosarum (Copel.) Morton, comb. nov. Ctenopteris rosarum Copel. Phil. Journ, Sci. 84:387. 1955. Type: Vicinity of Huigra, Dept. of Chimborazo, Ecuador, Rose 22232 (US). Range: Known from Ecuador from the type and one other collec- tion (Carmen, Dept. of Bolivar, Tate 480) and from one collection from the Department of Antioquia, Colombia (Daniel 615). Habitat: Not definitely known. Presumably terrestrial or saxi- colous, at middle elevations. 52. Grammitis assurgens (Maxon) Morton, comb. nov. Polypodium assurgens Maxon, Contr. Gray Herb. 165:73. 1947. Type: Quito-Santo Domingo road, Provinee of Pichincha, Ecuador, April 5, 1942, Haught 3226 (US). Ctenopteris assurgens Copel. Phil. Journ. Sci. 84:386. 1955. Range: Known from the type collection only in Ecuador, three collections from the Department of Cuzco, Peru, and two from Cerro Tatama, Department of Caldas, Colombia. Habitat: Terrestrial on rocky or mossy banks or in grassy p&ramos, at elevations from 2,300 to 3,700 m. MORTON—GENUS GRAMMITIS 117 Dubious Species Polypodium curvans Mett. Ann. Sci. Nat. [Paris] V, 2:253. 1864. Polypodium curvatum sensu Mett. Abh. Senck. Naturf. Ges. 2:58. 1857, non Swartz (1801). Type: Peru, Lechler. Placed in the synonymy of C. curvata by Copeland, but that can hardly be right. A specimen identified as P. curvans is Riobamba, Ecuador, 3,200 m., Rimbach 332; the material that I have seen is fragmentary. According to Dr. A. Murray Evans, P. curvans belongs in Polypodium and not Grammitis. Polypodium melanopus Hook. & Grev. Bot. Misc. 3:384, pl. 111. 1833. Ctenopteris melanopus Copel. Phil. Journ. Sci. 84:404. 1955. Type: “Hanging vertically from the trunks of trees at Lurencucho (Surucucho], near Cuenca, at an elevation of about 9000 feet above the level of the sea, Prof. W. Jameson.”’ (not seen). From the description and figure, the species is dubious to me, as it was also to Copeland. Polypodium oligosorum Mett. ex Kuhn, Linnaca 36:132. 1869, non Klotzsch, 1847. Syntypes: Venezuela, Moritz 460, Fendler 208, Karsten 10. A specimen that has been referred here is Ecuador, Jameson 790, p.p.; it is too fragmentary for identification, cf. p. 103. Polypodium pilosissimum Mart. & Gal. var. tunguraguense Hieron. Hedw, 48:252. 1909 (as “tunguraguensis’’). “Differt a forma typica foliis majoribus, usque ad % m longis; petiolis in speciminibus c. usque ad 1% dm longis; laminis 30-43 cm longis, usque ad 3 cm latis; pinnis et segmentis usque ad 1% cm longis, 5-6 mm basi latis, quam in forma typica latioribus; soris majoribus, usque ad 1% mm diametientibus.—Rhizoma deest. ‘‘An melius species propria (P. tunguraguense Hieron.)? ‘“Aequatoria: habitat ad radices montis Tunguragua prope Bafios in valle Pastaza (n. 844a).”’ A fragment of the holotype (Stuebel 844a) and a drawing by Brause were received by Dr. Maxon from Berlin, but the material is not adequate to place this plant, which does not seem to be exactly matched among the specimens in the U.S. National Herbarium. It is certainly not a variety of P. pilosissimum Mart. & Gal., and may well represent a new species, as indicated tentatively by Hieronymus. ?Gymnogramma sinuata Moore ex Baker, Ann. Bot. 5:483. 1891, non Presl, 1836. Type: Andes of Quito, Ecuador, Jameson (holotype, K). Dr. Jarrett has kindly examined the type and reports that the 118 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM frond is membranous and coarsely lobed; the sori are superficial, linear and up to 6 mm. long, and lie on the main vein-branches; the veins are free, with lax, pinnately arranged branches; at the point where the main vein to each lobe is given off from the midrib there is a vein directed toward the sinus. The name G. sinuata Moore is illegiti- mate, being a later homonym, and so cannot affect the name of any of the species accepted in this treatment. The following names, overlooked by Copeland in his treatment of Ctenopteris, apparently are referable to Grammitis, but they cannot be placed definitely without an examination of the types, which are not available for consultation. Father Bosco’s types were originally at the Istituto Salesiano S. Michele, Foglizzo, Italy, but their present location is uncertain. Dr. R. E. G. Pichi-Sermollii is attempting to locate them. Inasmuch as the original descriptions are not every- where readily available they are reproduced below. The names were all based on collections of Father Crespi from Ecuador. The plants of this collection were so wildly identified that one cannot be sure that they are even in the right genus. For instance, the plant described as Polypodium induens var. subdentatum Bosco, which should be a Grammitis, is actually a synonym of Dryopteris pusilla (Mett.) Kuntze, as shown by an isotype in the U.S. National Herbarium. Father Bosco kindly sent to the National Herbarium a duplicate set of the Crespi collections, but unfortunately there were no duplicates available for a number of the reputed new species and varieties. Polypodium allioni Bosco, Nu. Giorn. Bot. Ital., n. ser. 45:150. 1938. “Species rhizomate legnoso, squamibus numerosis, lanceolatis, castaneis, caducis vestito; stipitibus remotis, subglabris, nigrescen- tibus, 4-6 cm. longis; frondibus coriaceis, oblongo-ovatis, in apicem acutum desinentibus, 25-40 cm. longis, 3-6 cm. latis, usque ad rachim crassam, nigram ac pilosam divisis in foliola ligulata, subacuta, integra, basi dilatata in rachim decurrentia; inferiora diminuta, paulatim orbicularia, superiora vero in apicem linearem, acutum, integrum desinentia; venis mediis nigricantibus patentibus, parvulis, venis secundariis immersis, haud visibilibus; soris 6-9 ex utraque inferiorum venarum parte singularum foliarum positis, crassis, fere contiguis. ‘“‘“Hab.—Epiphyta in Paramos de Potrerillos (3200 m.) et ad Plan de Sapote (2100 m.). “Nota.—S’avvicina al P. mollissimum Fée, da cui perd differisce specialmente per le fronde coriacee e molto pit sviluppate; e al P. alternifolium Hk. da cui differisce per le foglioline fra loro riunite per la base decorrente sulla rachide, stipiti e rachidi pid robusti, sori piuttosto pit vicini al nervo medio che al margine. MORTON—GENUS GRAMMITIS 119 “Dedico questa specie al Missionario Salesiano, Padre M. Allioni, botanico e insigne esploratore dell’Equatore Orientale.” Polypodium crespianum Bosco, Nu. Giorn. Bot. Ital., n. ser. 45:150. 1938. “Specie stipitibus ut rachibus nigricantibus, permultis pilibus setiformibus obtectis; frondibus 40-60 cm. longis, 5-8 cm. latis, oblongo-lanceolatis, basim versus et ad apicem gradatim diminutis; membranaceo-coriaceis, flaccidis, pendulis, in superiori pagina foliarum pilis tenuibus, subrubris et sericeis obtectis, in inferiori vero pilis albescentibus, lanosis, copiose obtectis; usque ad rachim in foliola ligulata, subacuta divisis, inter se basi dilatata et in rachim decurrente reunitis, gradatim ad apicem distantibus, in quo paulatim foliola acuta vel irregulariter circinnata fiunt et pilosissima; venis immersis, vena media parum visibili, caeteris invisibilibus; soris 10-12 in utraque parte inferioris omnium foliarum paginae dispositis, vena media et margine aequaliter distantibus, in densis pilis nidulantibus. ‘Hab.—Epiphyta in silvis regionis nomine Plan de Sapote (2100 m.). ‘“Nota.—E una specie le cui dimensioni e l’aspetto generale fanno pensare al P. sericeo-lanatum Hk. da cui pero differisce sia per la mag- gior quantita di peli sulla pagina inferiore delle foglioline, sia per V’aspetto dei sori, le venuzze affatto invisibili, causa i molti peli che ricoprono le foglioline e le rachidi pid grosse.” Polypodium moniliforme Lag. ex Swartz var. culebriliense Bosco, Nu. Giorn. Bot. Ital., n. ser. 45:149. 1938. “Differt a typo stipitibus longius distantibus inter se et brevissimis; lamina 8-15 cm. longa, 3-5 cm. lata, in apicem acutum, serrulatum desinente; segmentis triangularibus, subacutis; soris unicis in omni segmento, crassis, totam fere foliarum paginam inferiorem tegentibus. “Hab.—In humidis petris ad Paramos de Culebrilla (3500 m.) “Nota.—E una specie questa diffusa dal Messico al Cile e di cui esistono gia, oltre il tipo, altre due varieta: Peruvianum Desv. del Pert e del Venezuela e Rigescens Bory dell’Equatore, Cile, Brasile, Cuba, Isola Fernando Po. A queste aggiungo ora la varieta Cule- briliense v. n. la quale differisce dalle altre specialmente per le dimen- sioni e la forma delle foglioline. “La collezione che no tra mano ha esemplari, della forma tipica e delle altre varieta, esclusa la Peruvianum Desv. la quale continua ad essere ancora endemica del Peri e del Venezuela.” Polypodium subsessile Baker ex Hook. & Bak. var. longipinnatum Bosco, Nu. Giorn. Bot. Ital., n. ser. 45:151. 1938. ‘“Frondes glabrae; pinnae lineares, filiformes, mediae 3-5 cm. longae; sori inter se adpropinquantes immo confluentes et total inferiorem foliarum paginam obtegentes. 120 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM “Hab.—Epiphyta, prope viam quae a regione nomine Indanza (950 m.) ducit ad flumen Santiago (540 m.).” Polypodium tonellii Bosco, Nu. Giorn. Bot. Ital., n. ser., 45:148. 1938. “Rhizoma tenue, erectum; squamibus parvis et paleis fulvo- castaneis vestitum; stipites brevissimi (1-2 em. longi), fasciculati, paleis numerosis, erectis, tenuibus ut in rachibus et in margine seg- mentorum quibus est quasi corona vestiti; fronda 10-12 cm. longa, sublinearis, acuminata, basin versus gradatim decrescens, 3-5 mm. lata, acuminata, pennato-partita; segmenta triangularia, acuta, inter se separata senibus concavis, a suis basibus productis; venae immersae, haud visibiles, sori subrotundi, crassi, ad costam mediam adpressi, fere totam segmentorum paginam inferiorem obtengentes. “Hab.—In terris humidis ad Plan de Sapote (2100 m.) et Plan de Milagro (1500 m.). “Nota.—Per le dimensioni parrebbe il P. serrulatum Mett. dal quale pero si differenzia sia per la copiosita di peli, sia per i sori non affatto confluenti e uno su tutti i segmenti, e non raggruppati nella meta superiore delle foglie; oltre anche alla forma dei segmenti. Differisce pure bene dal P. Truncicula Klotzs. al cui gruppo appartiene. ‘“Dedico questa specie al Rev. Prof. A. Tonelli gid Insegnante ordinario di Scienze Naturali al Liceo Pareggiato “Don Bosco” in Valsalice (Torino), il conservé con cura la presente collezione e ne incitd lo studio (1).’’ 24 * (1) Defunto il 3-1-1938. Di Lui rimangono pubblicazion{ di etnologiae glottologia di alcune triba del Brasile e specialmente studi notevoli e pregiati sul valore scientifico e sull’autenticitaé della 8. Sindone che si conserva nel Duomo di Torino.” CONTR. NAT. HERB. VOL. 38 MORTON—PLATE 1 gal ANTS OF ECUADOR a pe SLOT 29 Aug 45 Epiphyte. Near the laguna. PROV, AZUAY--"ORIENTE” BORDER: Piramo del Castillo snd sar @ forested areas lores of the eastern conlillers on the trail between evilln de Cho smd Memes): 9000-4 1000 £1. elev Lievirvbuted by the Sew York Botanica! Garden. Holotype of Grammitis aphelolepis Morton CONTR. NAT. HERB. VOL 38 MORTON-—PLATE 2 Holotype of Grammitts eminens Morton CONTR. NAT. HERB. VOL. 38 MORTON—PLATE 3 Ks PLANTS OF BCUADOR collected on Cinchena Minion of Formgn Econom / In rather deep shade old fronda persisting to forr neath ourrent live fronds, servir moisture; fronds soft-hairy,; at So. Province of PICHINCHA of bos a heavy mt Guamani Pass, ¢. of Pifo, Cordillera Oriental JOREPM A eWAN wo 1456 10 MOVEMERH 1944 Holotype of Grammitis intricata Morton CONTR. NAT. HERB. VOL. 38 MORTON—PLATE 4 Holotype of Grammitis alsopterts Morton Index [Synonyms in italics. New species, new names, and new combinations in boldface. Page numbers of principal entries in ztalics.] Acrostichum graminoides, 86 serrulatum, 86, 90, 95, 96 Antrophyum, 85 Cochlidium, 87 graminoides, 86 Cryptosorus blumei, 90 dionaea, 90 Ctenopteris, 85-90, 100, 111, 118 aromatica, 110 assurgens, 116 attenuatissima, 112 crispata, 100, 101 cuencana, 111 curvata, 117 ecuadorensis, 111 elastica, 107 farinosa, 99 firma, 110 heteromorpha, 102, 104 lanigera, 105, 106 leucosticta, 112 longa, 105, 106 longiuscula, 112 major, 109 megaloura, 114 melanopus, 117 montliformis, 116 myrtophylla, 108 peruviana, 115 pichinchae, 111 psuedocapillaris, 108 pseudonutans, 114 pteropus, 107 rigens, 114 rigescens, 114, 115 rosarum, 116 semiadnata, 101 semthirsuta, 113 sentlis, 104 sericeo-lanata, 105 stella, 106 Ctenopteris— Continued subflabelliformis, 104 subsessilis, 107 subtilis, 104 taxifolia, 110 tenuicula, 109 truncorum, 85 tunguraguae, 109 vernicosa, 110 Dicranopteris, 85 Dryopteris pusilla, 118 Glyphotaenium crispatum, 100, 101 Grammitis, 85-88, 89, 100, 117, 118 alfarii, 91, 103 alsopteris, 94, 112 andina, 98 anfractuosa, 93, 109 aphelolepis, 90, 97 assurgens, 95, 116 attenuatissima, 93, 112 australis, 87 billardieri, 88 blepharolepis, 90, 98 caucana, 90, 96 crispata, 91, 100 cuencana, 93, 111 cultrata, 92, 106 daguensis, 98 delitescens, 97 dependens, 92, 104 eminens, 91, 99 farinosa, 91, 99 firma, 93, 110 flabelliformis, 94, 115 graminea, 86-88 graminotdes, 86 heteromorpha, 91, 92, 102, 104 intricata, 91, 101 jamesonii, 90, 96 jamesonioides, 92, 108 jungermannioides, 90, 94 lanceolata, 86 121 122 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM Grammitis—Continued Polypodium, 85—88, 117 lanigera, 92, 105 alfarii, 103 var. stella, 92, 106 allioni, 118 laxa, 92, 106 alternifolium, 105, 106, 118 lehmanniana, 92, 104 andinum, 98 leucosticta, 94, 112 anfractuosum, 109 linearis, 86-88 aromaticum, 110 major, 93, 109 assurgens, 116 marginella, 86-88 astrolepis, 86 moniliformis, 94, 95, 116 aitenuatum, 110 myriophylla, 92, 108 azuayense, 108 peruviana, 94, 115 blepharolepis, 98 phlegmaria, 103 brownii, 110 pichinchae, 93, 111 caucanum, 96 pichinchensis, 93, 111 crespianum, 119 pilosissima, 114 crispatum, 100 pseudoaustralis, 88, 89 cuencanum, 111 pseudocapillaris, 92, 108 cultratum, 106, 107 pseudonutans, 94, 114 var. minus, 107 repanda, 90 curvans, 117 rigens, 94, 114 curvatum, 117 rosarum, 95, 116 daguense, 99 semiadnata, 91, 101, 102 dependens, 104 semihirsuta, 94, 110, 113 delitescens, 97 senilis, 91, 103 duale, 96 sericeo-lanata, 104-106 ecostatum, 100, 101 serrulata, 86, 90, 96 ecuadorense, 111 sodiroi, 94, 114 elasticum, 107 sprucei, 90, 96 euchlorum, 107 subflabelliformis, 92, 104 extensum, 96 subsessilis, 92, 107 farinosum, 99 subtilis, 92, 104 jfirmum, 110 taxifolia, 93, 109 flabelliforme, 115 trichomanoides, 99 gontopteroides, 100 trifureata, 91, 100 gracilimum, 98 truncicola, 90, 98 gramineum, 87 tunguraguae, 93, 109 heteromorphum, 102 variabilis, 91, 92, 102 induens, 109 vernicosa, 93, 110 var. subdentatum, 118 wolfii, 94, 115 jamesonit, 96 yarumalensis, 90, 96 jamesonioides, 108 sect. Chilopleris, 88 jungermannioides, 90 sect. Cryptosorus, 90 lanigerum, 105, 106 sect. Grammitastrum, 88, 89 lasiosorum, 88 sect. Grammitis, 88, 89, 90 lazum, 105 ; subg. Grammitis, 89 lehmannianum, 104, 105 subg. Melanoloma, 88, 89 leucosticta, 112 . . leucosticton, 112 sect. Xiphopteris, 90 longisetosum, 108 Gymnogramma sinuata, 117, 118 longiusculum, 112 Lomaphlebia, 87 longum, 105 Loxogramme, 87 marginellum, 89 lanceolata, 86 melanopus, 117 MORTON—GENUS GRAMMITIS 123 Polypodium—Continued molissimum, 118 moniliforme, 115 var. culebriliense, 119 myosuroides, 87 myriophyllum, 108 obliquatum, 90 oligosorum, 103, 117 pastoense, 105 pectinatum, 85 peruvianum, 115, 119 phlegmaria, 103 pichinchae, 111 pichinchense, 111 pilosissimum, 114, 117 var. tunguraguense, 117 plumula, 85 pseudoaustrale, 88 pseudocapillare, 108 pseudonutans, 114 pteropus, 107 rigens, 114 rigescens, 115, 119 saxicola, 109 semiadnatum, 101 semihirsutum, 113 senile, 103 sericeo-lanatum, 105, 106, 119 serrulatum, 88, 96, 120 var. jamesonit, 96 var. majus, 96 var. strictissimum, 96 Sodiroi, 114 sprucet, 95 var. costaricense, 90 var. furcativenosum, 90 strictissimum, 96 f. minus, 96 Poly podium—Continaed subcrenatum, 116 subflabelliforme, 104 var. minus, 104 sublongipes, 105 subscabrum, 111 subsessile, 107 var. longipinnatum, 119 subtile, 104 taxifolium, 109 tenuiculum var. acrosorum, 109 var. brasiliense, 109 tenuisectum, 108 tonellii, 120 trichomanoides, 88 trichosorum, 100 trifurcatum, 100 var. brevipes, 100 truncicola, 98, 120 var. majus, 98 var, minus, 98 tunguraguae, 109, 117 undulatum, 99 vulgare, 85, 86 wolfiz, 115 yarumalense, 95 {subsect.] Grammitis, 87 [subsect.] Lomaphlebia, 87 Vittaria, 85 Xiphopteris, 87-89, 90 blepharolepis, 98 caucana, 96, 97 extensa, 96 jamesonii, 96 myosuroides, 87 serrulata, 86, 96 truncicola, 98 U.S. GOVERNMENT PRINTING OFFICE: 1967 BULLETIN OF THE UNITED STATES NATIONAL MUSEUM CONTRIBUTIONS FROM THE UNITED STATES NATIONAL HERBARIUM Vo.tumeE 38, Part 4 THE PERUVIAN SPECIES OF BESLERIA (GESNERIACEAE) By C. V. Morton SMITHSONIAN INSTITUTION PRESS e WASHINGTON, D.C. ¢ 1968 THE PERUVIAN SPECIES OF BESLERIA (GESNERIACEAE) C. V. Morton Some years ago I published a revision of Besleria,! a rather common genus of woody Gesneriaceae of the American tropics extending from southern Mexico south to Bolivia and Brazil. Like most Gesneriaceae, these are plants of virgin forests and rarely are collected in secondary erowth, and for this reason when forests are cut down the gesneriaceous flora is exterminated, often completely, because many of the species are extremely local in distribution. Many of the species are still known only from the original collections, and many will probably never be found again. In Peru the species are mostly confined to the wetter, eastern part, being especially numerous in the Departments of Ama- zonas, San Martin, Loreto, Hudnuco, Junin, and Cuzco. Some few additional species are found in Cajamarca, Ayacucho, and Puno. Since my revision was published some 28 years ago many additional collections have come in from Peru; especially noteworthy ones have been received from César Vargas, Ramén Ferreyra, H. A. Allard, J.J. Wurdack, and P. C. Hutchison. As is to be expected, these collec- tions contain many new records for previously undescribed species and also several very interesting novelties, including one species of a section of the genus new to Peru. I have grouped the species in the sections and subsections that I recognized in the revision, although I admit that these are undoubtedly somewhat artificial. The key to the Peruvian species that I have pro- vided is wholly artificial but hopefully more usable than the original one. The present paper does not account for all the Peruvian Beslerias, for several additional species are represented in the U.S. National Herbarium and doubtless others elsewhere. These specimens are in- adequate in various ways, and so I do not wish to describe them at the present time without additional material. I am listing some of these below so that collectors can be on the lookout for them: Rfo Azul, 30 km. from Tingo Marfa, Hudnuco, 750-850 m., Ferreyra 12736; Porvenir, Pichis Trail, Junin, 1,500-1,900 m., Killip & Smith 25960; 1 Contr. U.S. Nat. Herb. 26: 395-474. 1939. 125 126 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM Paso Gacsa, Exito to Puerto Durand, Hudnuco, 2,800 m., Vargas 5357; Divisoria, 59 km. from Tingo Marfa on highway to Pucallpa, San Martin, 1,600 m., Allard 21289; near Pajonal, Sandia, Puno, 1,000- 1,300 m., Metcalf (8rd Univ. Calif. Bot. Gard. Exped. 30621); Pilco- pata, Paucartambo, Cuzco, 580 m., Vargas 011605; Hacienda Schunke, La Merced, Macbride 5607; and Valley of Rio Marajion above Cascadas de Mayasf, near Campamento St. Montenegro (km. 280-284 of Marafion road), Bagua, Amazonas, 450-600 m., Wurdack 1843. Some of these collections represent species widely different from any previ- ously known, and so additional collections are much to be desired. Key to the Peruvian Species of Besleria Flowers solitary, fasciculate, or in subsessile umbels, a common peduncle lacking or much less than 1 em. long. Disk glabrous (unknown in B. horizontalis). Stems at apex, and costae and veins of the leaf blades beneath glabrous, stri- gose, strigillose, puberulous, or sericeous-tomentose, but not hirsute or spreading-villous. Corolla tube glabrous externally. Ovary and style glabrous (style slightly pubescent in B. aggregata). Calyx 3-5 mm. long, green. Corolla orange, 13-15 mm. long, lacking a hairy ring within. Stems strigillose toward apex; leaf blades strigillose on the veins beneath, elliptic; calyx lobes 3-4 mm. long, nearly glabrous, ciliolate. Cajamarca... . ... . . 1. B. solanoides Stems only microscopically ‘strigillose;, leaf blades microscopically strigillose, narrowly or broadly lanceolate; calyx lobes 5-6 mm. long, glabrous or _ Sparsely hirsute externally, ciliate. Hudnuco, 2,400m. ... i. eee eh)h)h)6 dB. gracilenta Calyx 9-16 mm. long, green or orange. Calyx green, 9-15 mm. long, the lobes ovate-lanceolate, nearly glabrous; corolla tube yellow, the limb white. Stems sparsely strigose; petioles elongate, up to 11.5 cm. long; leaf blades elliptic or ovate-elliptic, up to 16 em. long and 7.5 cm. broad, membranaceous, the veins 8 or 9 pairs; corolla 18 mm. long, lacking a hairy ring within. San Martin, 260m. ... . .. . . 15. B. membranacea Calyx orange, 13-16 mm. long in . flower, the lobes ovate-lanceolate or lanceolate; corolla orange. Stems and leaves completely glabrous; calyx lobes glabrous, ciliate; veins 8 or 9 pairs; corolla ca. 17 mm. long. Hudnuco, 650 m, 16. B. angusta Stems and leaves beneath obviously strigillose; calyx lobes densely sericeous-pilose externally; veins 5-8 pairs; corolla 20-27 mm. long, the lobes ciliate, the tube with a hairy ring within (not known in B. aggregata var. mexiae). Petioles up to 4 cm. long; leaf blades oblanceolate, up to 22 em. long and 9 cm. broad, long-decurrent at base. Junin, Loreto, and Cuzco, up to 700 m. alt. 17. B. aggregata var. aggregata Petioles 7-10 cm. long; leaf blades elliptic, up to 22 em. long and 11 cm. broad or more, cuneate at base but not long-decurrent. Loreto, 110m, .... .. . 17a. B. aggregata var. mexiae MORTON—PERUVIAN BESLERIA 127 Corolla tube puberulous, strigillose, pilose, or villous externally. Calyx 4-8 mm. long, yellow (green in B. densiflora), the lobes orbicular or ovate, strigillose or sparsely pilose externally. Corolla more than 30 mm. long, subhorizontal in calyx, spurred at base posteriorly. Leaf blades obviously serrate above middle. Cuzco,2,700m........+.++...-. SLB. horizontalis Corolla less than 20 mm. long, erect in calyx, not spurred at base. Corolla orange, the tube merely puberulous externally, with a hairy ring within; stems strigillose-tomentose at apex; leaf blades oblanceolate, up to 10 em. broad; calyx green, ca. 4 mm. long, the lobes orbicular, strigillose without; ovary glabrous; style pubescent. Loreto, 1,200m. ........ .3 B. densiflora Corolla lemon-yellow, the tube villous without, lacking a hairy ring within; stems sparsely strigose at apex; leaf blades oblong- lanceolate, up to 5 cm. broad; calyx lemon-yellow, 6-8 mm. long, the lobes ovate, acutish, sparsely pilose without; ovary and style pubescent. Loreto, 1,000-1,110m. ....... 18. B. citrina Calyx 13-16 mm. long, orange, the lobes ovate-lanceolate, acuminate, densely sericeous-pilose without. Stems densely appressed-pilose at apex; leaf blades strigillose on the veins beneath; corolla orange, 20-27 mm. long, the lobes ciliate, the tube with a hairy ring within (not known in B. aggregata var. mexiae), without only sparsely strigillose or pilosulous. Petioles up to 4 em. long; leaf blades oblanceolate, up to 22 cm. long and 9 em. broad, long-decurrent at base. Junin, Loreto, and Cuzco, up to700m.alt. ........ . 17. B. aggregata var. aggregata Petioles 7-10 cm. long; leaf blades elliptic, up to 22 cm, long and 11 cm. broad or more, cuneate at base but not long-decurrent. Loreto, 110m..........4... . 17a, B. aggregata var. mexiae Stems hirsute or spreading-villous at apex; costa and veins of the leaf blades hirsute or spreading-villous beneath. Leaf blades glabrous above; corolla tube glabrous without (or sparingly pilosulous in B. aggregata var. loretensis). Ovary glabrous. Calyx 15 mm. long; larger leaf blades up to 30 cm. long and 11 em. broad, long-decurrent at base. Style pubescent. Loreto and San Martin, 100-625m. ........ . . 17b. B. aggregata var. loretensis Calyx 5-7 mm. long; leaf blades smaller. Corolla orange, ca. 12 mm. long. Leaf blades up to 26 cm. long, 10-13 cm. broad, the petioles up to 4 cm. long; calyx 6.5-7 mm. long, the lobes rounded at apex, long-ciliate. San Martin, 100-200m. .......-.-+2. + 4& B. barbata Leaf blades up to 9 em. long and 3.5 cm, broad, the petioles up to 2 cm. long; calyx ca. 5 mm. long, the lobes oblong, obtuse. Style pubescent. Cajamarca. 2... ee ee ee ee 5. B. imberbis Leaf blades hirsute or pilose on the upper surface; corolla tube hirsute without (except in B. furva). Corolla white, 13 mm. long, the tube glabrous without. Leaf blades oblique- ly ovate, up to 24 em. long and 13 cm. broad, abruptly acuminate, unequally rounded at base; calyx 11 mm. long, the lobes ovate- lanceolate, acuminate, scarcely denticulate; ovary, style, and disk glabrous. San Martin . 2. 6 6 ee ee ee es 19. B. furva Corolla yellow-orange, or red, 13-20 mm. long, the tube hirsute without, at least near apex. 128 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM Calyx ca. 6 mm. long, less than half as long as the corolla (14-16 mm, long). Calyx lobes free, elliptic, narrowed to base, equal, entire, obtuse or acutish, mucronulate, hirsute, long-ciliate; corolla lacking a hairy ring within, the lobes not ciliate; ovary pilosulous at apex; style pilosulous; disk glabrous. San Martin, Hudnuco, 625-1,100 m . . .. .... . 6. B. placita Calyx 10-18 mm. long, at least two-thirds as long as the corolla (13-20 mm, long). Calyx 13-18 mm. long, the lobes united to middle, the free parts triangular-subulate, entire; ovary, style, and disk glabrous. Corolla tube lacking a hairy ring within, the lobes not ciliate. Loreto, 400-900 m. . . .. .... . 20. B. emendata Calyx 10-11 mm. long, the lobes free or nearly so; ovary pilose at least at apex; style pilosulous. Calyx lobes narrow, not more than 3 mm. broad, entire; leaf blades serrate. Corolla erect in calyx, not spurred at base, brick-red, the tube with a hairy ring within; leaf blades lanceolate or elliptic- lanceolate; filaments pilose at base. Loreto, 180-325 m. 21. B. immitis Corolla oblique or horizontal in calyx, slightly spurred at posterior base, yellow except for the 2 dark purplish posterior lobes of the limb, the tube lacking a hairy ring within; leaf blades elliptic-lanceolate; filaments glabrous. Ovary pilose; style pilosulous at base. Amazonas, 425-500 m. 32. B. wurdackii Calyx lobes ovate-lanceolate to lanceolate, 3-5 mm. broad near base, denticulate or serrate, hirsute without. Calyx lobes thick, lanceolate, ca. 16 mm. long, almost equaling the corolla. Loreto, 1,500-1,600 m.. . . . 22. B. ferreyrae Calyx lobes thin, ovate-lanceolate, ca. 11 mm. long, much shorter than the corolla. Corolla pale lemon-yellow, 19 mm. long, lacking a hairy ring within; filaments glabrous. Loreto and San Martin, 220-1,500 m.. . ... .. . 23. B. compta Flowers in peduncled inflorescences, the common peduncle (except sometimes in B. sprucei) 2 cm. long or more, the flowers umbellate, racemose, or subcorymbose. Stems and costae and veins of the leaf blades beneath glabrate, strigose, strigillose, puberulous, or tomentose (in B. corallina and B. rupestris) but not hirsute or spreading villous, the leaf blades nearly glabrous above at maturity (except in B. hutchisonii and B. rupestris). Calyx 2.5-7 mm. long. Ovary glabrous. Corolla tube glabrous externally; style glabrous. Common peduncle 2.5-5 em. long. Leaf blades elliptic, nearly entire; calyx lobes suborbicular, ciliate. Cololla orange-red, ca. 15 mm. long; calyx 4-5 mm. long; pedicels only 2 or 3 to an umbel, elongate, up to 2.5 cm. long; stems densely strigose toward apex; leaf blades up to 12.5 em. long and 5 cm. broad, strigillose on the veins beneath. Hudnuco, 2,300-2,700 m 2... . 7. Be lucida MORTON—PERUVIAN BESLERIA 129 Corolla whitish, 5-6 mm. long; calyx 2.5 mm. long; pedicels numerous, corymbose, short, ca. 5 mm. long; stems glabrate; leaf blades up to 28 cm. long and 12 em. broad, minutely puberulous on the veins beneath. Loreto... ...... . . .8.B. minutiflora Common peduncle 5.5-10 cm. long. Stems glabrous; petioles glabrous; leaf blades membranous; calyx lobes glabrous (ciliate in B. vari- abilis), entire, mucronate; corolla 20-26 mm. long, lacking a hairy ring within. Corolla horizontal in calyx, spurred at posterior base, the tube strongly inflated upwardly, the orifice of the limb dorsal and not terminal. Petioles ca. 10 mm. long; leaf blades obliquely cuneate at base, obviously serrate, glabrous on the veins beneath, the primary veins 12-15 pairs; calyx 7-7.5 mm. long, the lobes not ciliate; corolla lobes glabrous, not ciliate. Loreto, 250-350 m. 35. B. caligula Corolla erect in calyx, not spurred, the tube scarcely ventricose, the limb terminal. Petioles 4-11 cm. long; leaf blades obtuse at base, minutely denticulate, minutely strigillose on the veins beneath, the primary veins 8-10 pairs; calyx 3-4 mm. long, the lobes ciliate; corolla lobes sometimes ciliate. Junin, Ayacucho, San Martin, Loreto, Hudnuco, 220-700 m. Corolla greenish white; plants herbaceous. 9b. B. variabilis f. pallida Corolla orange or brick-red. Exterior corolla lobes not ciliate; plants fruticose. 9. B. variabilis f. variabilis Exterior corolla lobes barbate; plants arborescent. 9a. B. variabilis f. barbatula Corolla tube puberulous or pilosulous externally: style pubescent (except in B. vargasit). Corolla tube with a hairy ring within at insertion of stamens; common peduncle 3-5 cm. long. Stems sparsely pubescent; petioles 1-4 cm. long; leaf blades elliptic-lanceolate, up to 208 cm., subentire, puberulous on veins beneath, subtomentose on margins; pedicels up to 5 mm. long; calyx lobes oblong-lanceolate, submucronate, subglabrous; corolla orange; Puno, 900 m. . . 10. B. concolor Corolla tube glabrous within, stipitate-glandular within throat; common peduncle 5-16 cm. long. Calyx lobes ovate, obtuse, not venose; corolla ochroleucous; leaf blades elliptic-lanceolate, 2-5 cm. broad. Filaments stipitate-glandular upwardly; anthers stipitate-glandular; Junin, 2,100 m. 11. B. pallidifiora Calyx lobes suborbicular, rounded at apex; corolla yellow or red. Stems nearly glabrous; calyx 4-7 mm. long, the lobes not venose; leaf blades up to 14 cm. long and 7 cm. broad, cuneate at base, puberulous beneath on veins. Filaments and anthers stipitate- glandular. Junin, Arequipa. . . . . . . 12. B. tetrangularis Stems densely coarse-strigose; calyx 3-4 mm. long, prominently venose; leaf blades 15-19 em. long, 8-11 cm. broad, rounded at base, strigose beneath on veins. Style glabrous. Cuzco, 1,200 m. 13. B. vargasii 130 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM Calyx 6.5-15 mm. long. Flowers umbellate or subcorymbose. Stems densely ferruginous-strigillose at apex when young. Petioles slender, up to 10 cm. long; leaf blades elliptic, ca. 15X10 cm., rounded at base, sparsely strigillose on the veins beneath; common peduncle ca. 10 cm. long; calyx ca. 15 mm. long, the lobes lanceolate, united to middle, acuminate; corolla tube glabrous without. Amazonas, Loreto, Cuzco, Puno, and Hudnuco, 250-1,950 m. .. . .24. B. divaricata Stems lanate-tomentose or tomentose at apex. Corolla erect in calyx, not spurred at base; disk completely annular or semiannular, glabrous; leaf blades subentire or remotely and in- conspicuously denticulate. Inflorescences 8- to 15-flowered, on common peduncles 18-20 em. long; leaf blades nearly glabrous on the upper surface at maturity; calyx lobes glabrous within or essentially so. Cuzeo, 390 m. 25. B. moorei Inflorescences 2- or 3-flowered, on common peduncles 1.5-1.8 em. long; leaf blades obviously and persistently sericeous on the upper surfaces; calyx lobes puberulous within toward apex. Hudnuco, 2,800m. ... . .. . . 26. B. hutchisonii Corolla horizontal in calyx, spurred at ‘posterior base: disk reduced to a posterior gland or semiannular, pubescent; leaves crenate-serrate. Calyx lobes broad, incised-serrate, pilose on both sides, ca. 7 mm. long; corolla scarlet, 15-17 mm. long, the tube densely pubescent with- out; ovary glabrous; style glabrous; disk semiannular; common peduncle 8-17 cm. long, the flowers subeorymbose, the pedicels numerous, up to 15 mm. long, pilosulous; leaf blades tomentulose on the veins beneath, glabrous above, the primary veins 9-11 pairs. Loreto, 100-1,200m. ....... . . 33. B. corallina Calyx lobes unequal, lanceolate to ovate-lanceolate, sharply serrate, sparsely pilose without, glabrous within, 6.5-8 mm. long; corolla yellow with maroon spots on tube and lobes within, 22-32 mm. long, sparsely pilose without; ovary villous; style pubescent; disk reduced to a thick posterior gland; common peduncle 6-7.5 cm. long, the flowers subumbellate, the pedicels fewer, ca. 6 mm. long, glabrous; leaf blades puberulous on the veins beneath, puberulous on upper surface, the primary veins 9-15 pairs. Junin, Hudnuco, and Cuzco, 1,500-1 ,900 m. . 34. B. rupestris Stems at apex and costae and veins of the leaf blades beneath hirsute or tomentose-hirsute. Calyx lobes united to about the middle, ca. 15 mm. long, lanceolate, acumi- nate, not mucronate. Leaf blades up to 15 em. long and 6.5 cm. broad, nearly entire, cuneate at base, sparsely or densely hirsute above; common peduncle 4-6 cm. long; corolla orange-red, 20-22 mm. long, not much ventricose, the tube glabrous without, the throat sparingly stipitate- glandular within; ovary glabrous. Corolla lobes pilose without. Junin and Hudnuco, 700-2,500 m. 27. B. peruviana var. peruviana Corolla lobes entirely glabrous. Junin, 700 m. 27a. B. peruviana var. occulta Calyx lobes free or nearly so, not more than 11 mm. long, hirsute without, glabrous within; style pilosulous; ovary pilosulous at least at apex (except in B. trichiata). MORTON—PERUVIAN BESLERIA 131 Calyx 5-10 mm. long, the lobes linear-lanceolate, acuminate, entire, not mucronate. Leaf blades elliptic to oblanceolate, up to 25 cm. long and 10 em. broad, membranous, serrate, cuneate at base; common pedun- cles up to 2.6 cm. long, the pedicels up to 1.5 cm. long; corolla 18-20 mm. long, the tube densely pilosulous without, conspicuoulsy ven- tricose, the throat glandular within. Calyx lobes ca. 5 mm. long; corollas red; flowers numerous. Cuzco, 700-800 m. .. . .. . . .14, B. sprucei var. sprucei Calyx lobes ca. 10 mm. long; corollas yellow; flowers 2 or 3 on a peduncle. Amazonas, 320m. ....... . 14a. B. sprucei var. flavescens Calyx 10-11 mm. long, the lobes slightly or prominently mucronate. Cor- olla with a hairy ring within at insertion of stamens. Calyx lobes oblong-lanceolate, with an elongate mucro arising below the apex; corolla yellow, ca. 13 mm. long, scarcely ventricose, the lobes scarcely pilose without, not ciliate, the tube glandular within throat; common peduncle up to 4 cm. long; filaments pubescent upwardly; anthers more or less pubescent; primary leaf veins ca, 10 pairs. Hudnuco... . ... . 28. B. capitata Calyx lobes oblong, short- mucronate. “Corolla 15-16 mm. long, the lobes hirtous or ciliate; pedicels thick. Corolla brick-red, the tube densely hirto-pubescent without; common peduncle 9-15 cm. long, thick, the flowers racemose; primary leaf veins 13-15 pairs; leaf blades 17-26 cm. long, 8-15 cm. broad, rounded at apex, oblique and rounded at base, San Martin, Junin, 230-1,900 m .. . .... . 29. B. racemosa Corolla yellow, the tube glabrous without: common peduncle up to 3.5 em. long; filaments pilose at base; primary leaf veins 8-11 pairs; leaf blades up to 12 cm. long and 6.5 cm. broad, acute at apex, cuneate at base. Junin, Cuzco, 700-1,900 m. 30. B. trichiata Section Besleria Leaf blades mostly subentire; common peduncle elongate or obsolete, the pedicels mostly numerous, rarely solitary; calyx lobes free or variously connate, mostly entire, sometimes mucronate; corolla erect in the calyx, not spurred at the posterior base, the tube mostly ventri- cose above the middle, the throat contracted, the limb terminal, small, subregular; disk mostly completely annular, glabrous. Type: Besleria lutea L. Subsection Sessiles Morton, Contr. U.S. Nat. Herb. 26: 405. 1939. Common peduncle obsolete or not more than 4 mm. long; calyx lobes free or rarely united at base but not higher than middle, mostly obtuse or rounded, 3-7 mm. long, less than half as long as the corolla tube, entire, glabrous within. Tyre: Besleria solanoides H. B. K. 267-632—68——2 132 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM 1, Besleria solanoides H. B. K. Nov. Gen. & Sp. 2: 398. 1818. This species, common in Colombia north to Honduras and Jamaica, is still known in Peru only from Tambillo, Department of Cajamarca, Jelski 24 (B); this Berlin specimen was doubtless destroyed, and so there may not be any specimens of this number existing. 2. Besleria gracilenta Morton, Contr. U.S. Nat. Herb. 26: 421. 1939. The type came from Mirador, Hudnuco, 2,400 m. alt., Mexia 04136 (GH). A more recent collection that may represent this species is: Tingo Maria side of Carpish, Dept. Hudnuco, 2,550 m., October 1945, Sandeman 5245; it differs in having the calyx lobes sparsely hirsute without, rather than entirely glabrous; it is described by Sandeman as a subshrub with waxy, apricot-colored flowers growing in semishade beside a stream. 3. Besleria densiflora Fritsch, Notizbl. Bot. Gart. Berlin 11: 965. 1934. Known only from Cerro de Ponasa, Loreto, Ule 6667 (holotype B, photograph US); since the holotype is presumably destroyed, the isotype at Geneva (Field photograph 28346) is here designated as lectotype. 4. Besleria barbata (Poepp. & Endl.) Hanst. Linnaea 34: 327. 1865. Hypocyrta barbata Poepp. & Endl. Nov. Gen. &Sp.3: 3.1845. Type: Pampayaco, San Martin, Poeppig (isotype B, photograph US). Poeppig 1508 (W, Field photograph 32788) from Cuchero, Hudnuco, is presumably the same species. 5. Besleria imberbis Morton, Contr. U.S. Nat. Herb. 26: 430. 1939. Still known only from the type: Tambillo, Dept. Cajamarca, Raimondi 5334 (B); the type is presumably destroyed at Berlin; there is a photograph of it in US. 6. Besleria placita Morton, sp. nov. Caules dense hirsuti; folia opposita subaequalia, longe petiolata; laminae foliorum acuto-ellipticae, magnae, apice breviter acuminatae, basi cuneatae, remote denticulatae, utrinque hirsutae; pedicelli in axillis conferti, brevissimi; calyx parvus, lobis liberis, aequalibus, ovali-ellipticis obtusis vel acutiusculis, integris, vix mucronulatis, externe hirsutis, intus glabris, marginibus longe albo-ciliatis; corolla erecta, non calcarata, paullo ventricosa, tubo externe dense hirsuti- tomentoso, intus glabro, limbo terminali, parvo, lobis subaequalibus, erectis, eciliatis, intus glabris; filamenta glabra; ovarium apice paullo pilosulum; stylus vix pilosulus; discus brevis, annularis, glaber. Shrub 1-1.7 m. high; stems rather thick, 3-5.5 mm. in diameter at 30 cm. below apex, unbranched upwardly, densely and persistently MORTON—PERUVIAN BESLERIA 133 white-hirsute; leaves opposite, subequal, the petioles 2-6 cm. long, hirsute; leaf blades pointed-elliptic, the larger 126 cm. up to 209 cm., sharply short-acuminate at apex, cuneate equally at base, remotely and minutely denticulate above the middle, herbaceous, sparsely but persistently hirsute above, paler green beneath, hirsute on the midrib, veins, and surface, the primary veins 7-9 pairs; flowers numerous in the leaf axils, crowded, the pedicels very short, less than 5 mm. long, hirsute; calyx ca. 6 mm. long, less than half as long as the corolla, the lobes free, equal, erect, oval-elliptic, 2.5-3 mm. wide at middle, slightly narrowed toward the base, obtuse or acutish but not acuminate, slightly mucronulate at apex, membranaceous, not venose, slightly hirsute externally, glabrous within, conspicuously long-white-ciliate; corolla orange, erect in calyx, not spurred at base but slightly gibbous above the base, 14-16 mm. long, 4-5 mm. wide at base, slightly ventricose, becoming 6.5-7 mm. wide at the middle, slightly contracted to the throat, this 5 mm. wide, externally densely hirsute-tomentose, glabrous within and lacking a hairy ring at the insertion of the stamens, the limb terminal, small, ca. 6 mm. wide, the lobes erect, subequal, ca. 1.5 mm. long, 1.5-2 mm. wide at base, glabrous and eglandular within, not ciliate; filaments glabrous, con- torted; anthers connate, broader than long, 1-1.5 mm. long, 1.7-2 mm. wide, the cells confluent; ovary pilosulous at apex; style thick, ca. 6 mm. long, slightly pilosulous; disk completely annular, low, hardly 0.5 mm. high, glabrous. Type in the U.S. National Herbarium, no. 1,999,822, collected in deep ravine in jungle on the slope east of Tingo Maria, Department of San Martin, Peru, at 625-1,100 m. elevation, Oct. 30, 1949-Feb. 19, 1950, by H. A. Allard (no. 20851). PaRATYPES (all in the U.S. National Herbarium): San Marrin: Allard 21441,? 21593a,3 21601, 22243, 22422. Hvudnuco: Rio Supte, Mar. 9, 1947, Ferreyra 1901. This species belongs in the subsection Sessiles; it is closely related to B. boliviana Morton, which differs in having the calyx lobes acuminate, the corolla tube only slightly pilose, and the flowers solitary or few in the leaf axils instead of being crowded. 2 The corolla is indicated by Allard as “pinkish,” but this is probably an erroneous observation; a truly pinkish color is not found among plants of this alliance. 3 “On tree; flower pure white” on the label, but this is probably due to a faulty memory. Plants of this alliance are not usually epiphytes, and the other specimens of this species are clearly terrestrial. White flowers are characteristic of some few species of Besleria, but in this case the flower color does not appear to be different from that of the other specimens of this species. In most species of Besleria the corolla color is lost in drying. 134 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM Subsection Pedunculatae Morton, Contr. U.S. Nat. Herb. 26: 409. 1939. Leaf blades entire or denticulate; common peduncle mostly elongate, the pedicels umbellate or subcorymbose; calyx lobes entire, mostly free, often obtuse or rounded, 2.5-7.5 mm. long, less than half as long as the corolla tube, glabrous within. Type: Besleria variabilis Morton. 7. Besleria lucida Poepp. & Endl. Nov. Gen. & Sp. 3: 2. 1845. The type came from Cochero [Cuchero], Hudnuco, Peru, collected by Poeppig (not seen). Two additional specimens were cited in my revision, both from Hudnuco, at elevations from 2,300 to 2,700 meters. 8. Besleria minutiflora Fritsch, Repert. Sp. Nov. Fedde 18: 9. 1922, Type: Leticia, Loreto, Peru, Ule 6211 (B). The holotype was doubt- less destroyed in Berlin. A photograph of it is in US; it is still known only from the type collection. 9. Besleria variabilis Morton, Contr. U.S. Nat. Herb. 26: 435. 1939. Forma variabilis. The type came from Balsapuerto, Loreto, Klug 2912. Several variants, from San Martin, Ayacucho, and Junin were mentioned in the original description. One recent collection has been seen: Tingo Marfa, Tulumayo, Hudnuco, Ferreyra 1624. The species is known also from Colombia (Putumayo) and from between Palmar and Chapa, Bolivia, Cdérdenas 5993. The taxonomic significance of the two formae described is uncertain. 9a. Besleria variabilis f. barbatula Morton, op. cit. 436. Type: Between San Nicolés and Azupizi, Junin, 650-900 m., Killip & Smith 22666 (US). 9b. Besleria variabilis f. pallida Morton, loc. cit. Type: Rio Paucartambo Valley, Junin, 700 m., Killip & Smith 25293 (US). 10. Besleria concolor Fritsch, Notizbl. Bot. Gart. Berlin 11: 970. 1934. Known only from the type: Chunchusmayo, Sandia, Puno, Weber- bauer 1194 (B); the holotype was presumably destroyed at Berlin; there is a photograph of it in US. 11. Besleria pallidiflora Fritsch, Notizbl. Bot. Gart. Berlin 11: 968. 1934. Still known from the type only: Yanangu, Junin, Weberbauer 2117 (B). The holotype is probably destroyed; there is a photograph of it in US. 12. Besleria tetrangularis Ruiz ex Hanst. Linnaea 34: 331. 1865. Cited in my revision from Arequipa, Puno, and Junin, at elevations from 650 to 900 m. No additional collections have come in, but I have seen two additional photographs of isotypes (Vitoc, Ruiz & Pavon) from Madrid (Killip photograph 477, Field photograph 29614). The MORTON—PERUVIAN BESLERIA 135 sheet in Madrid represented by the last cited photograph is designated as lectotype, since the original holotype in Berlin has presumably been destroyed. 13. Besleria vargasii Morton, Journ. Washington Acad. Sci. 35: 126. 1945. Still known only from the type collection: Mandor, Marcapata, Prov. Quispicanchis, Dept. Cuzco, 1200 m., Vargas 3732 (US). 14. Besleria sprucei Britt. Bull. Torrey Club 27: 31. 1900. At the time of my revision, this species was known only from Bolivia (type: Mapiri, Bolivia, Rusby 2149, NY; isotype US) and Brazil (State of Amazonas, Ule 5927, type of B. uleana Fritsch). Several collections from Peru are now known: Near Atalaya on road to Pilcopata, Prov. Paucartambo, Cuzco, 750 m., Vargas 13969 (US); Consuelo to Asuncién, Prov. Paucartambo, Cuzco, 850 m., Vargas 7356 (US); road to Fortaleza, Prov. Quispicanchis, Cuzco, 800 m., Vargas 9717 (US). According to notes by Vargas the plants are 1 to 1.5 m. high and the corollas red-orange. The following variety appears recognizable: 14a. Besleria sprucei Britt. var. flavescens Morton, var. nov. Pedunculus communis brevis, 12-15 mm. longus, pauciflorus, floribus 2 vel 3; calycis lobi lineares, ca. 10 mm. longi; corolla flava (sec. Sagdstegui). Type in the U.S. National Herbarium, no. 2,469,215, collected at Montenegro, Provincia Bagua, Department of Amazonas, Peru, at 320 m. elevation, Oct. 15, 1965, by A. Sagdstegui A. (no. 5867). Subsection Confertae Morton, Contr. U.S. Nat. Herb. 26:411. 1939. Common peduncle obsolete, the pedicels aggregate in the leaf axils; calyx lobes mostly nearly free, mostly acute or acuminate, 6-17 mm. long, more than half as long as the corolla tube. Type: Besleria immitis Morton. 15. Besleria membrancea Morton, Contr. U.S. Nat. Herb. 26: 445. 1939. Known only from the Department of San Martin: Chazuta, Rio Huallaga, 260 m., Klug 3994 (holotype US; isotypes F, GH); Rio Mayo, near Tarapoto, Spruce s.n. (IX); T rocha Agua Blanca-Desquite, on road to Tarapoto, Prov. Lamas, 600-800 m., Ferreyra 7975 (US). 16. Besleria angusta Morton, sp. nov. Caules glaberrimi; folia opposita aequalia, breviter petiolata; laminae foliorum anguste lanceolatae 3.5-plo vel 4.5-plo longiores quam latiores, attenuatae, basi late cuneatae, subcoriaceae, integrae, fere concolores, utrinque glaberrimae; flores bini axillares in umbellis sessilibus, pedicellis brevibus crassis glabris; calycis lobi liberi imbricati 136 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM coriacei ovato-lanceolati, supra medium ligulati, obtusi, submucronati, utrinque glabri, margine ciliati; corolla in calyce erecta non calcarata, quam calyx vix longior, tubo non ventricoso externe glabro in fauce non contracto, limbo terminali subregulari intus piloso. Shrub 24 m. high, the stems terete, unbranched above the base, ca. 5 mm. in diameter at 20 cm. below the tip, entirely glabrous; leaves of a pair subequal, short-petiolate, the petioles 2.5-4.5 cm. long, thick, glabrous; leaf blades narrowly lanceolate, 3.5-4.5 times as long as broad, the larger 24 cm. long and 6 em. broad, long-acuminate at apex, broadly and subequally cuneate at base, entire, subcoriaceous, nearly concolorous, completely glabrous on both sides, the primary veins 8 or 9 pairs, obscure above, more or less immersed beneath or at least not elevated; flowers in 2-flowered sessile umbels in the leaf axils, the pedicels 10-15 mm. long, thick, ca. 1 mm. in diameter, glabrous, slightly enlarged at apex; calyx in flower ca. 13 mm. long (19 mm. long in fruit), the lobes coriaceous, free, imbricate, entire, ovate- lanceolate, ca. 7 mm. wide near base, narrowed and subligulate above the middle, costate above the middle, obtuse at apex and slightly mucronate, completely glabrous on both sides, conspicuously ciliate throughout; corolla orange, not much longer than the calyx, ca. 17 mm. long, erect in calyx, not spurred at base, fleshy in texture, the tube apparently not ventricose, ca. 4 mm. broad at middle, hardly contracted in throat, completely glabrous externally, the limb terminal, ca. 5.5 mm. wide, slightly pilose within the throat, the lobes subequal; ovary and style glabrous. Type in the U.S. National Herbarium, no. 2,322,768, collected between Santa Isabel and Cucharas, Department of Hudnuco, Peru, at 650 meters elevation, June 2, 1958, by Felix Woytkowski (no. 5023). Two isotypes in U.S. National Herbarium. Belongs in section Besleria, subsection Confertae, but it is not nearly allied to the other species. 17. Besleria aggregata (Mart.) Hanst. in Mart. Fl. Bras. 8 (1): 400. 1864. Hypocyrta aggregata Mart. Nov. Gen. & Sp. 3: 51. 1829. Type: Manacuru, on Rio Japura, Vaupés, Colombia, Martius. A specimen in Miinchen “in sylvis Japurensibus, R. N., Martius’ (Morton photograph 8813) is doubt- less an isotype. Besleria ignea Fritsch, Notizbl. Bot. Gart. Berlin 11: 966. 1934, Type: Leticia, Loreto, Ule 6182 * * * (holotype B, photograph US; isotype G, Field photograph 28347). Inasmuch as the holotype is presumably destroyed, the isotype at Geneva is here designated lectotype. In my revision of 1939, I adopted the name B. ignea and placed B. aggregata among the dubious species, remarking that the two might be identical. An examination of the isotype at. Miinchen shows that the two are indeed the same, and so the earlier name B. aggregata MORTON—PERUVIAN BESLERIA 137 must be taken up. The typical variety occurs in Loreto and Junin, as reported previously, as well as in Amazonas, Brazil, and may now be reported also from Cuzco: Near San Lorenzo, Prov. Quispicanchis, 700 m., Vargas 11759 (US). The following varieties may eventually be found to represent distinct species. 17a. Besleria aggregata var. mexiae (Morton) Morton, comb. nov. Besleria ignea var. mexiae Morton, Contr. U.S. Nat. Herb. 26: 444. 1939. Type: Rancho Indiana to Rio Napo, Loreto, 110 m., Mexia 6447a (holo- type US; isotype F). 17b. Besleria aggregata var. loretensis (Morton) Morton, comb. nov. Besleria ignea var. loretensis Morton, Contr. U.S. Nat. Herb. 26: 444. 1939. Type: Between Yurimaguas and Balsapuerto, lower Rio Huallaga Basin, Loreto, 135-150 m., Killip & Smith 28337 (holotype US). In addition to the specimens, all from Loreto, previously reported, the following new collections are at hand: Guallabamba, near Iquitos, Loreto, 106-110 m., Ferreyra 3379, 3383. Ridge east of Tingo Maria, Department San Martin, 625-1100 m., Allard 21576, 22299. 17c. Besleria aggregata var. semiannularis (Morton) Morton, comb. nov. Besleria ignea var. semiannularis Morton, Contr. U.S. Nat. Herb. 26: 444. 1939. Type: Florencio, Caquet4é, Colombia, Pérez Arbelaez 751B (holotype US). This variety is known also from Putumayo, Colombia (Cuatrecasas 10582) and is to be expected in Loreto, Peru. It differs in having the leaf blades softly pilose beneath and hirsute on the veins and the disk semiannular rather than completely annular. 18. Besleria citrina Fritsch, Notizbl. Bot. Gart. Berlin 11: 963. 1934. Known only from the type: Moyobamba, Loreto, Peru, Weberbauer 4618 (B); the holotype was presumably destroyed at Berlin; there is a photograph in US. 19. Besleria furva Morton, Contr. U.S. Nat. Herb. 26: 447. 1939. Still known only from the type collection: Tarapoto, Dept. San Martin, Spruce 4955 (K). 20. Besleria emendata Morton, sp. nov. Caules ad apicem versus dense hirsuti; folia opposita inaequalia, longe petiolata, ovata, breviter acriter acuminata, membranacea, denticulata, supra et infra hirsuta; flores axillares solitarii, pedicellis gracilibus elongatis; calyx magnus tubum corollae fere aequans, membranaceus viridis, lobis ca. ad medium connatis, lobis liberis triangulari-subulatis externe hirsutis, intus elabris; corolla aurantiaca demum rubra, in calyce erecta, tubulosa, tubo apice piloso, intus glabro, limbo brevi subbilabiato, lobis externe pilosis, intus glabris 138 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM eciliatis; filamenta glabra; ovarium glabrum; stylus glaber; discus annularis glaber. Subshrub up to 1.2 m. high; stems sparingly branched, slender, 2.5-4 mm. in diameter at 30 em. below tip, densely hirsute apically with white, septate hairs; leaves opposite, the opposed shghtly un- equal, the larger with petioles 2-6 cm. long, the smaller on petioles 1-2 cm. long, all petioles obviously white-hirsute; larger leaf blades ovate, 9-14 cm. long, 4-8 cm. wide,‘ sharply short-acuminate, broadly cuneate at the unequal base, membranaceous, minutely and remotely glandular-denticulate, white-hirsute above on the surfaces, paler green beneath and white-hirsute on veins and surfaces, the primary veins 6-8 pairs; smaller leaf blades similar but only about two-thirds as long; flowers solitary in the leaf axils, the pedicels 1.2-2.5 em. long, slender, white-hirsute; calyx green, membranaceous, venose, 13-18 mm. long, the lobes united to about the middle, the calyx tube 6-9 mm. long, the lobes entire, triangular-subulate, narrow and sharp at apex, 2.5-4 mm. wide at base, 6.5-9 mm. long, not mucronate, white-hirsute externally, glabrous within; corolla orange at first, becoming red later (fide Ferreyra), 17-20 mm. long, erect in calyx, not spurred at base but slightly gibbous above base, the tube 4-6 mm. wide at base, cylindric and hardly ventricose, 5-8 mm. wide above middle, slightly contracted in throat and 4-6 mm. wide, externally glabrous below, sparsely pilose at apex, glabrous within and without a hairy ring at the insertion of the stamens, the limb terminal, small, 6-10 mm. wide, slightly bilabiate, the lobes erect, suborbicular, the larger 3.5-4 (5) mm. long, the smaller 2-3.5 (4) mm. long, pilose externally, glabrous and eglandular within, not. ciliate ; filaments glabrous, contorted; anthers quadrately united, broader than long, ca. 1.5 mm. long, 2.5 mm. wide, the connective thickened, the cells confluent; staminodium ca. 3 mm. long, glabrous; ovary conic, glabrous; style thick, 9-10 mm. long, glabrous; stigma bilobed; disk completely and regularly annular, ca. 0.5 mm. high, glabrous; berry globose, enclosed in the persistent calyx. Type in the U.S. National Herbarium, no. 2,025,103, collected at Boquerén, ca. 92 km. from Tingo Marfa on the highway to Pucallpa, Department of Loreto,’ Peru, at 410 meters elevation, Dec. 16, 1949- Jan. 5, 1950, by H. A. Allard (no. 21713). Paratypss (all in U.S, National Herbarium except as indicated) : ‘ Only 52.5 cm. in depauperate (sun?) plants, e.g., Hutchison & Wright 60434. 5 The label gives the department as San Martin, but Boquerén is actually in the Department of Loreto, in the Provinee of Coronel Portillo, not far from the border with San Martfn. MORTON—PERUVIAN BESLERIA 139 Loreto: Boquer6n [type locality], Allard 21716, 22131, Swingle 101, Ferreyra, 1189, 8104, Hutchison & Wright 6043A (UC), 6053 (US). Between Divisoria and Boquerén, Ferreyra 1101.8 Rance: Known only from the vicinity of Boquerén, Department of Loreto, at elevations variously given as between 400 and 900 meters. This species may be the same as B. peruriana var. dissimilis Morton,’ but I regard it now as specifically different from B. peruviana Fritsch, which has the flowers borne in long-pedunculate umbels rather than solitary in the axils of the leaves. There are instances in Besleria where the common peduncle has been so reduced that the flowers appear to be borne in sessile umbels in the leaf axils, but in such cases at least a trace of the peduncle remains and the flowers are several. In B. emen- data there is no trace of a peduncle and the flowers are solitary. I consider these characters as of more than varietal importance now, and in fact they are the basis for a part of my sectional division of the genus. I continue to place B. peruriana in subsection Elongatae, but B. emendata should be in subsection Confertae. However, it does not seem to be wise to base this species on B. peruviana var. dissimilis because it is not absolutely certain that the two are identical. This variety was based on Tessmann 3170 from the lower Ucayali River, Department of Loreto, Peru, but unfortunately the type was in Berlin and was lost during the war. There is a photograph of it in the National Herbarium but it is not sufficiently good to make an identification certain. Unless a duplicate of Tessmmann 3170 can be located in some other herbarium, it is best to list B. peruviana var. dissimilis among the dubious plants of this genus. The only other species of section Confertae with calyx lobes connate to the middle is B. tuberculata Morton, of Colombia, which differs strongly in the tuberculate upper leaf surface, the villous ovary, the stipitate-glandular throat of the corolla, long-ciliate co- rolla lobes, and other characters. 21. Besleria immitis Morton, Contr. U.S. Nat. Herb. 26: 449, 1939. The type came from Umbria, Putumayo, Colombia (Klug 1912, holotype US; isotype S). Known also from the Department of Loreto, Peru, at low elevations, as mentioned in the original description. 22. Besleria ferreyrae Morton, sp. nov. Frutex, vix ramosus, caulibus crassis, dense lutescenti-villosis; folia longe petiolata, laminis ellipticis, acuminatis, basi late cuneatis, remote denticulatis, utrinque hirsutis; inflorescentiae fasciculatae, 6 At least one specimen of this number has been labeled incorrectly as “‘Be- tween Sinchono and Boquerén, Dept. San Martfn.” 7 Contr. U.S. Nat. Herb. 26: 452. 1939. 140 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM pedunculo communi nullo, pedicellis brevibus crassis hirsutis; calyx corollam fere aequalis, lobis lanceolatis acuminatis denticulatis, externe dense hirsutis, intus basi glabris apicem versus pilosis; corolla non calcarata, mediocris, supra medium valde ventricosa, externe dense tomentosa, intus glabra, limbo parvo terminali; filamenta glabra; ovarium pubescens; stylus basi pubescens; discus annularis integer glaber. Shrub 0.8-1.2 m. high, hardly branched above the base, the stems very thick, ca. 6 mm. in diameter ca. 20 cm. below the apex, densely shaggy-yellowish-hirsute; leaves of a pair equal, long-petiolate, the petiole 4-4.5 cm. long, densely hirsute; leaf blades elliptic, 104.5 to 16X6 cm., sharply acuminate, broadly cuneate and subequal at base, remotely denticulate, obviously hirsute on both sides, the primary veins 8 or 9 pairs, upwardly curved; flowers fasciculate in the leaf axils, a common peduncle absent, the pedicels 2 or 3, short and stout, 5-10 mm. long, densely hirsute; calyx green, almost as long as the corolla, the lobes nearly free, lanceolate, ca. 16 mm. long, slightly unequal, one ca. 3 mm. broad at middle, the others ca. 2 mm. broad, acuminate, remotely denticulate, outside densely yellow-tomentose, within glabrous below middle, slightly pilose toward apex; corolla yellow, erect in calyx, not saccate at base, 4 mm. broad at base, ventricose above middle and ca. 6 mm. broad, slightly contracted in throat and 5 mm. broad, densely white-tomentose, glabrous within and lacking a hairy ring, the limb terminal, small, the lobes erect, ca. 2 mm. long and broad; filaments glabrous; anthers included, connate, 1 mm. long, 2 mm. broad; ovary pubescent; style pubescent; disk annular, low, entire, glabrous. Type in the U.S. National Herbarium, no. 2,057,476, collected at Plantacién Margarita, near Divisoria, Province of Coronel Portillo, Department of Loreto, Peru, in tropical woods at 1,500-1,600 meters elevation, Aug. 14, 1946, by Ramén Ferreyra (no. 989). This species, named in honor of Sr. Ramén Ferreyra, of the Uni- versity of San Marcos, in recognition of his invaluable services to Peruvian botany, belongs in the section Besleria, subsection Confertae. It may be allied to B. compta Morton, which differs in its slenderer, less strongly hirsute stems, less obviously hirsute leaves, more slender pedicels, and shorter, thinner, broader, and less densely hirsute calyx lobes, which are much shorter than the corolla. 23. Besleria compta Morton, Contr. U.S. Nat. Herb. 26: 449. 1939. The type came from Balsapuerto, Loreto, 220 m., Klug 2926 (holo- type US; isotypes B, GH, K, S); also known from the Department of San Martin, as noted in the original description, at elevations up to 1,500 meters. MORTON—PERUVIAN BESLERIA 14] Subsection Elongatae Morton, Contr. U.S. Nat. Herb. 26: 412. 1939. Common peduncle elongate, the pedicels umbellate, racemose, or subcorymbose; calyx lobes free or connate to the middle, mostly en- tire, 10-17 mm. long, more than half as long as the corolla tube, mostly acuminate. Typs: Besleria elongata Urban. 24. Besleria divaricata Poepp. Nov. Gen. & Sp. 3: 2. 1845. In my revision this species was reported in Peru from the Depart- ments of Hudnuco and Loreto and from some places in Ecuador. It appears now that this is one of the more common and widespread species in Peru, the following additional collections being at hand: Northern slopes of Cerros Campanquiz, Pongo de Manseriche, Loreto, 550-750 m., Wurdack 2376. Rain forest along Rio Santiago, Dept. Amazonas, 230-300 m., Wurdack 2198. Trail from La Oroya to Santo Domingo, Prov. Sandia, Dept. Puno, 900-1,200 m., Hodge 6026. Three km. above Santo Domingo, Prov. Sandia, Dept. Puno, 1,950 m., Metcalf (U. Calif. Bot. Gard. Exped. 30649). Between San Pedro and Santa Isabel, Prov. Paucartambo, Dept. Cuzco, 1,400 m., Vargas 6769. Between Atalaya and Carbén, Prov. Paucartambo, Dept. Cuzco, 700 m., Vargas 13434. The plant is described by the collectors as a herb or bush 1 to 2.5 meters high, with waxy, orange-red corollas. 25. Besleria moorei Morton, sp. nov. Herba alta, caulibus apice dense tomentosis; folia longe petiolata, laminis elliptico-oblongis, magnis, abrupte acuminatis, basi late cuneatis, remote denticulatis, supra fere glabris maturitate, subtus praecipue in venis tomentosis; inflorescentiae subcorymbosae, longe pedunculatae, multiflorae, pedicellis brevibus; calycis lobi_ liberi, subaequales, late ovati, mucronati, integri, externe hirsuti, intus fere glabri; corolla aurantiaca, in calyce erecta, ecalcarata, tubo medio ventricoso, in fauce contracto, externe hirsuto, limbo parvo subregulari, tubo intus basi annulum pilosum gerente; filamenta basi pilosa; ovarium glabrum; stylus pilosus; discus annularis, glaber. Stout, branched herb up to 1 m. high, the upper stems ca. 5 mm. in diameter, densely brownish tomentose near apex; leaves of a pair apparently slightly unequal, the larger on petioles 9 cm. long and 2 mm. thick, these strigose or subtomentose when young; larger leaf blades elliptic-oblong, 27 cm. long and 11 cm. broad, the smaller 14 cm. long and 6 cm. broad, abruptly acuminate, the base broadly cuneate and subequal-sided, remotely denticulate, herbaceous, nearly glabrous above at maturity, pilosulous when young, subtomentulose beneath, especially on the veins, the primary veins ca. 15 pairs; inflorescences solitary, axillary, subcorymbose, 8- to 15-flowered, the common 142 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM peduncle very elongate, 18-20 cm. long, sparsely pilose, the pedicels thickish, 8-10 mm. long, § elabrate; calyx lobes orange (fide Moore) free, subequal, imbricate, broadly ovate, 9.5 mm. long, 4-5 mm. broad, entire, hirsute externally, essentially glabrous within with only occasional scattered hairs, not veiny externally, venose within, mucronate; corolla orange, erect in calyx, not spurred at base, ca. 28 mm. long, the basal orifice ca. 3 mm. broad, the tube ca. 5.5 mm. broad just above base, rather abruptly ventricose at the middle and here 8-8.5 mm. broad, contracted in the throat, here ca. 6 mm. broad, externally pilose, more densely so near the throat, the limb terminal, 7-8 mm. broad, the lobes erect, semicircular, subequal, ca. 3 mm, long and broad, cucullate with recurved margins, pilosulous externally, glabrous within, not ciliate, within glabrous except for a puose ring at the insertion of the fllaments; filaments inserted near base of corolla tube, connate with each other, flattened, slightly pilose near base, glabrous upwardly, coiled; anthers included, per- sistently connate, broader than long, ca. 15 mm. long, 2.6 mm. broad, the cells confluent; ovary ovoid, glabrous; style thick, pilose; disk completely annular, ca. 0.6 mm. high, unlobed, glabrous. Type in the U.S. National Herbarium, no. 2,370,542, collected on roadside bank 16.6 km. from Quincemil, on the road to San Lorenzo, Province of Quispicanchis, Department of Cuzco, Peru, at ca. 390 m. elevation, June 11, 1960, by H. E. Moore, Jr., Adolfo Salazar C., and Karl E. Smith (no. 8583). Isotype presumably in Bailey Hortorium (not seen). Besleria moorei belongs in section Besleria, subsection Elongatae, and in my key (1939) it will key out to B. ovalifolia Rusby (Mem. Torrey Bot. Club 4: 240. 1895), which is still known only from the original collection from Songo, Bolivia (Bang 862), but it cannot be very closely allied to that, which differs in having the calyx lobes essentially glabrous, long-acuminate, and united to above the middle. In its general characters and appearance it is much closer to B. racemosa Morton, of the Departments of San Martin and Junin, Peru, which differs in several ways, as follows: Pubescence densely hirsute-villous; petioles 1-3 em. long; common peduncles 9-15 em. long; calyx lobes serrulate: corollas ca. 15 mm. long; ovary slightly pilosulous at apex. 2...) ee . . . B. racemosa Pubescence densely scriceous-tomentose; petioles 6-9 em. long; common peduncles 18-20 em. long; calyx lobes entire; corollas ca. 18 mm. long; ovary entirely glabrous...) . ee B. moorei 26. Besleria hutchisonii Morton, sp. nov. Herba vel frutex parvus, vix ramosus, ramis crassis, sursum dense albido-tomentosis; folia breviter petiolata, petiolo crasso tomentoso, laminis mediocribus ellipticis, acuminatis, basi cuneatis, integris, supra MORTON—PERUVIAN BESLERIA 143 ubique dense sericeis, subtus praecipue in venis dense sericeis; 1n- florescentiae umbellatae, 2- vel 3-florae, pedunculo communi brevi, pedicellis antrorse hirsutulis; calyx magnus, lobis imbricatis liberis ovalibus rotundatis, non mucronatis, integris, extus dense sericels, intus apice excepta glabris; corolla aurantiaca, parva, utrinque elabra, medio ventricosa, limbo terminali magno, lobis ciliolatis; filamenta glabra; antherae inclusae connatae latiores quam longiores; ovarium glabrum; stylus hirsutulus; discus annularis integer glaber. Herb or small shrub up to 75 cm. high, the stems not branched above the middle, thick, ca. 5 mm. in diameter ca. 30 cm. below the tips, upwardly persistently densely whitish-tomentose; leaves of a pair equal, short-petiolate, the petioles 10-15 mm. long, thick, ca. 2 mm. in diameter, densely tomentose; leaf blades elliptic, 9-14 cm. long, 4-5.2 cm. broad, acuminate, cuneate at the base, entire, persist- ently and densely sericeous above, paler beneath and sericeous, espe- cially on the veins, the primary veins 9-11 pairs, curved, ascending, strongly elevated beneath; inflorescences axillary, few, 2- or 3-flowered, the common peduncle short, 15-18 mm. long, sericeous, the hairs ascending, the pedicels ca. 10 mm. long, antrorsely hirsutulous; calyx 7-9 mm. long, more than half longer than the corolla, the lobes im- bricate, free, oval, ca. 4 mm. broad, the sides curved, the apex rounded, not mucronate, entire, densely sericeous without, glabrous within except near the sparsely sericeous apex, 7-nerved, the nerves visible within, obscure without; corolla orange (fide Hutchison), erect in the calyx, ca. 15 mm. long, not spurred or gibbous at base, the tube ca. 4 mm. broad at base, upwardly a little ventricose and becoming ca. 6 mm. broad, the throat slightly contracted and 5.5 mm. broad, with- out glabrous, the limb terminal, ca. 8 mm. in diameter, the lobes erect, imbricate, broad, ca. 4 mm. long and 3.5 mm. broad, rounded, entire, glabrous without, slightly ciliate when young, the tube glabrous within; filaments inserted ca. 2 mm. above the base of the corolla tube, flattened, glabrous; anthers included, connate, broader than long, ca. 1.3 mm. long, ca. 2.2 mm. broad, elabrous, the locules con- fluent; ovary glabrous; style short, thick, hirsutulous; disk completely annular, slightly higher posteriorly, glabrous, entire. Type in the U.S. National Herbarium, no. 2,466,984, collected at Carpish, above Acomoyo, Province of Hudnuco, Department of Hudnuco, Peru, at 2,800 meters altitude, July 17, 1964, by Paul C. Hutchison, J. Kenneth Wright, and R. M. Straw (no. 5937). Isotypes in F, UC, USM (not seen). This species belongs in section Besleria subsection Elongatae of my revision. It will key out near B. laziflora Benth., but that species is obviously not nearly allied, being distinguished by lesser pubescence, larger, serrulate leaf blades glabrate (rather than densely sericeous) 144 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM above, by elongate common peduncles, more numerous flowers in an umbel, generally larger calyx lobes, these attenuate into a mucro (rather than rounded and not mucronate), and by Jarger corollas, with a hairy ring within at base and glandular within the throat. 27. Besleria peruviana Fritsch, Repert. Sp. Nov. Fedde 18: 7, 1922. Besleria peruviana var. peruviana. Known definitely only from the type collection, from La Merced, in Chanchamayo Valley, Prov. Tarma, Dept. Junin, at 700 to 800 m. alt., Weberbauer 1958 (B). Presumably the Berlin specimen was destroyed, but a photograph of it is in US. A second specimen possibly referable here is Sandeman 5170, from Chinchao Valley, Carpish Divide, Hufnuco, 2,250 m., October 1945, which is more densely hirsute, and with somewhat smaller calyces and corollas; it is a subshrub or herb, said to be fre- quent in the forest in deep shade. 27a. Besleria peruviana var. occulta Morton, Contr. U.S. Nat. Herb. 26: 452. 1939. Still known only from the type: La Merced, Junin, at 700 m., Killip & Smith 24086 (US). It was described by the collectors as a herb 3 to 5 feet high, with elongate branches and bright orange-red flowers. 27b. Besleria peruviana var. dissimilis Morton. For a comment on this plant, see the account of B. emendata above. 28. Besleria capitata Poepp. & Endl. Nov. Gen. & Sp. 3: 2. 1845. Known only from the type, Cuchero, Hudnuco, Poeppig 1622 (W, Field photograph 32789) and Cuchero, Poeppig 1623 (B, photo- graph US). 29. Besleria racemosa Morton, Contr. U.S. Nat. Herb. 26: 450. 1939. This species is still known only from the material originally cited, from San Martin and Junin, the specimens from Junin being dubiously exactly the same as the type from San Martin (Pongo de Cainarachi, Rio Cainarachi, San Martin, 230 m., Klug 2660; holotype US, iso- types B, NY, S). 30. Besleria trichiata Morton, Contr. U.S. Nat. Herb. 26: 451. 1939. Previously known only from the type collection, from Porvenir, Department of Junin, 1,500-1,900 m., Killip & Smith 25955 (US), but now represented by an additional collection: Between Fortaleza and San Lorenzo, Prov. Quispicanchis, Dept. Cuzco, 700 m., Oct. 22, 23, 1950, Vargas 9724 (US). According to Vargas’ note, the corolla is red-salmon colored. Section Neobesleria Morton, Contr. U.S. Nat. Herb. 26: 414. 1939. Herbs or shrubs, the stems glabrate, strigillose or tomentose (hirsute in B. wurdackii); leaf blades mostly serrate, membranaceous, acu- minate; common peduncle elongate or obsolete; calyx lobes free, often MORTON—PERUVIAN BESLERIA 145 toothed; corolla horizontal in the calyx, spurred at the posterior base; disk often reduced to a thick posterior gland, this often pubescent. Typs: Besleria corallina Fritsch. Subsection Axillares Morton, Contr. U.S. Nat. Herb. 26: 416. 1939. Herbs or soft woody shrubs; leaf blades serrate above the middle; common peduncle obsolete, the pedicels aggregate in the leaf axils; calyx lobes free, entire; corolla contracted in the throat, the limb terminal, small, regular. Typr: Besleria gibbosa (Poepp.) Hanst. 31. Besleria horizontalis Morton, sp. nov. Suffrutex, caulibus apice strigillosis, mox glabris; folia per paria aequalia, breviter petiolata, laminis ellipticis acriter acuminatis, basi aequaliter cuneatis, supra medium serratis, subcoriaceis, supra glabris, subtus pallidioribus glabris costa strigillosa excepta; flores solitarii axillares, pedicellis elongatis; calycis lobi aequales imbricati sub- orbiculares rotundati integri externe strigillosi; corolla flava, tubo purpureo-maculato, in calyce horizontali, basi postice calcarato, externe dense piloso, supra medium ampliato in fauce non contracto, limbo terminali, lobis subaequalibus, externe vix pilosis eciliatis. Woody subshrub, the stems unbranched upwardly, ca. 5 mm. in diameter ca. 30 cm. below tip, pale stramineous, with a few short, appressed hairs at the tip, below entirely glabrous; leaves of a pair equal, short-petiolate, the petioles up to 3 cm. long, thick, glabrous; leaf blades elliptic, 9-16 cm. long, 4-6 em. wide, shortly and sharply acuminate at apex, subequally cuneate at base, subcoriaceous, ob- viously serrate above the middle, dark green and glabrous above, paler beneath, some few appressed hairs on midrib, otherwise glabrous, the primary veins 5 or 6 pairs, elevated beneath; flowers solitary, axillary, the pedicels ca. 2.5 cm. long, slightly enlarged at apex; calyx ca. 7 mm. long, the lobes nearly free, inbricate, suborbicular, ca. 6 mm. long and 5.5 mm. wide, rounded at apex, not mucronate, entire, apparently yellowish, thinly silky-strigillose externally; corolla yellow, the tube with dark purple spots, horizontal in calyx, ca. 30 mm. long, slightly spurred at the posterior base (ca. 3 mm.), the tube ca. 5 mm. wide at middle, enlarged to 9 mm. wide in throat, densely pilose externally, the limb terminal, ca. 11 mm. wide, the lobes subequal, ca. 3 mm. long and broad, sparsely pilose externally, entire. Type in the U.S. National Herbarium, no. 1,877,283, collected in dense high forest, Alturas de Pintobamba, Province of Convencién, Department of Cuzco, Peru, at 2,700 meters elevation, March 4, 1943, by C. Vargas C. (no. 3252). 146 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM This, the second species of section Neobesleria, subsection Azillares, known from Peru, differs from B. wurdackii in many characters, notably in the glabrous (or only sparsely strigillose apically) stems, glabrate leaves, solitary flowers, suborbicular, imbricate calyx lobes, and especially in the large corolla, with the tube enlarged to the throat rather than contracted. Besleria gibbosa (Poepp.) Hanst., of Amazonian Brazil, differs in the crowded, short (only 8 mm. long) pedicels, and the short (only 4.5 mm. long), ovate-oblong, acute, nonimbricate calyx lobes, among other characters. 32. Besleria wurdackii Morton, sp. nov. Caules hirsuti; folia opposita aequalia, longe petiolata, lamina membranacea, late elliptica, acriter acuminata, basi obliqua, utrinque laxe hirsuta, conspicue serrata fere usque ad basim; flores axillares fasciculati, pedicellis gracilibus elongatis hirsutis; calycis lobi liberi subaequales ovato-lanceolati integri, apice attenuati et subulati, externe hirsuti eciliati; corolla flava 2 lobis posticis purpureis exceptis, in calyce subhorizontalis, basi postice leviter calcarata, ventricosa, externe hirsuta, intus glabra, in fauce contracta, limbo terminali sub- irregulari; filamenta glabra; ovarium pilosum; stylus basi pilosus; stigma bilobum, discus annularis leviter 5-lobatus altus glaber. Soft-woody shrub ca. 1 m. high, the stems unbranched above, ca. 4 mm. in diameter at 30 cm. below tip, densely yellowish hirsute toward apex; leaves opposite, subequal, the lower long-petiolate (4-5 cm.), the upper short-petiolate (1.5 em.), the petioles densely hirsute; leaf blades obliquely broadly elliptic, about twice as long as broad, 10-14 em. long, 5-7 em. wide, abruptly and sharply short-acuminate at apex, oblique at base, the lower base extending at nearly a right angle, the upper at an acute angle, conspicuously serrate nearly to the base, thin-membranaceous, persistently but finely hirsute on both sides, the primary veins 8-10 pairs; flowers several, fasciculate in the leaf axils, a common peduncle obsolete, the pedicels 2-2.5 cm. long, very slender, hirsute; calyx ca. 10 mm. long, the lobes free, ovate- lanceolate, ca. 3 mm. wide at the middle, subequal but the posterior reflexed, entire attenuate and subulate at apex, hirsute externally, glabrous within, not ciliate; corolla yellow except for the dark pur- plish two posterior lobes of the limb, oblique in the calyx or subhori- zontal, slightly spurred at the posterior base (1 mm.), 14-16 mm. long, strongly ventricose, becoming ca. 8.5 mm. wide at the middle, contracted to the throat, this ca. 5.5 mm. wide, the tube strongly pilose externally, glabrous within, the limb terminal, slightly bilabi- ate, the two posterior lobes smaller, dark purplish, the other yellow; filaments glabrous; anthers broader than long, ca. 1.5 mm. long and 2mm. wide, connate, the cells confluent; staminodium conspicuous, MORTON—PERUVIAN BESLERIA 147 ca. 5 mm. long; ovary strongly pilose; style pilose at base, ca. 7 mm. long; stigma bilobed; disk annular but irregularly 5-lobed, high, ca. 1.5 mm. long, glabrous. Type in the U.S. National Herbarium, no. 2,404,369, collected on the left bank of Rio Marafion, above Cascadas de Mayasi, opposite km. 278 of the Marafion road, Province of Bagua, Department of Amazonas, Peru, at 425-500 m. elevation, Sept. 12, 1962 by J. J. Wurdack (no. 1955). This species belongs in section Neobesleria, subsection Azillares. The only previously known species of this section with a hirsute type of pubescence is B. obliqua Morton, of Panama, which differs in the glabrous corolla tube, smaller and differently shaped calyx lobes, smaller, not serrate leaves, and in many other ways. Subsection Corallinae Morton, Contr. U.S. Nat. Herb. 26: 414. 1939. Leaf blades serrate or crenate; common peduncle elongate, longer than the pedicels; calyx lobes free, imbricate, serrate or denticulate, pubescent on both sides; corolla contracted in the throat, the limb terminal, small, regular; disk semiannular, pubescent. Tyre: Besleria corallina Fritsch. 33. Besleria corallina Fritsch, Repert. Sp. Nov. Fedde 18: 13. 1922. Known from the Department of Loreto and adjacent Putumayo, Colombia. The type came from Cerro de Ponasa, Ule 6671 (holotype B; photograph US). A collection that has not been previously re- ported is: Gamitanacocha, Rio Mazin, Loreto, 100-125 m., J. M. Schunke 198 (US). Subsection Maculatae Morton, Contr. U.S. Nat. Herb. 26: 415. 1939. Shrubs, the stems densely lanate-tomentose; leaf blades serrate; common peduncle elongate, 6-7.5 cm. long, the pedicles much shorter; calyx lobes free, ovate to ovate-lanceolate, acuminate, sharply serrate, glabrous within; corolla yellow with conspicuous small maroon spots on tube and within lobes, the throat not contracted, the limb terminal, slightly bilabiate; ovary villous; disk reduced to a thick, puberulous, posterior gland. Tyre: Besleria rupestris Morton. 34. Besleria rupestris Morton, Proc. Biol. Soc. Washington 48: 57. 1935. The type came from Ecuador (between Bafios and Cashurco, Prov. Tungurahua, 1,300-1,800 m., Hitchcock 21820, US). This is one of the more widely distributed species, for it is known from Colombia (Dept. of El Cauca and El Valle) as well as from Peru, in the Depart- ments of Junin (Killip & Smith 25759) and Hudnuco (Mezia 8237), as previously reported (Contr. U.S. Nat. Herb. 26: 462. 1939), and 148 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM from Cuzco: Between Tanamayo and Santa Isabel, Prov. Paucar- tambo, 1,600 m., Dec. 6, 1947, Vargas 6794. There may be some varia- tion in habit, for some collections indicate this species to be an erect herb ca. 80 cm. high, but one says that it is scandent and up to 1.5 m. high. The corollas are probably unique in the genus in having the lobes and tube within conspicuously but minutely spotted. Section Gasteranthus (Benth.) Fritsch, in Engl. & Prantl, Nat. Pflanzenfam. 4(3b): 159. 1895 Gasteranthus Benth. Pl. Hartweg. 233. 1846. Herbs or shrubs; leaf blades mostly toothed; common peduncle elongate; calyx lobes ovate to lanceolate or obovate, less than half as long as the corolla tube; corolla horizontal in calyx, spurred at pos- terior base, strongly ventricose, the tube inflated at apex and sur- passing the limb, the limb small, regular, dorsal, facing upwardly; disk thickened posteriorly or reduced to a posterior gland. Type: Besleria quitensis (Benth.) Hanst. In my revision I erroneously attributed the sectional name Gas- teranthus to Bentham and Hooker, but Bentham and Hooker did not formally recognize any sections in Besleria, as they did in some other genera of Gesneriaceae. 35. Besleria caligula Morton, sp. nov. Suffrutex, caulibus non ramosis, glaberrimis; folia inaequalia, breviter petiolata, lamina anguste elliptica, longe acuminata, basi oblique cuneata, serrata, utrinque glaberrima; inflorescentiae subum- bellatae vel subcorymbosae, longe pedunculatae, pedicellis crassis, glabris; calycis lobi inaequales, membranacei, integri, obtusi vel rotundati, glabri, apiculati, posticus deflexus latus; corolla horizon- talis, aurantiaca, calcarata, valde inflata, utrinque glabra, limbo dorsali non terminali parvo regulari; androecium et gynoecium glabra; discus postice valde incrassatus glaber. Subshrub ca. 50 cm. high; stems unbranched, ca. 4 mm. in diameter ca. 30 cm. below apex; leaves of a pair usually strongly unequal, one less than one-fifth as long as the other, all short-petiolate, the petioles ca. 10 mm. long, glabrous; larger leaf blades narrowly elliptic, 115 cm. up to 19X6 cm., long-acuminate at apex, obliquely cuneate at base, conspicuously serrulate, the teeth 18-24 on each side, thin- herbaceous, green above, paler beneath, entirely glabrous on both sides, the primary veins 12-15 pairs, ascending, anastomosing within the margin; inflorescences axillary, subumbellate or subcorymbose, the common peduncle 5-7 cm. long, thick, glabrous, the pedicels 4-6, short and thick, 8-10 mm. long, glabrous; calyx in flower ca. 7 mm. high and 15 mm. broad, irregular, the lobes free, thin-membranous, MORTON—PERUVIAN BESLERIA 149 entire, glabrous on both sides, 5-nerved, the midrib thickened up- wardly and excurrent in a short apiculus, the 2 anterior obovate, ca. 7 mm. long and 5.5 mm. broad, narrowed to base, broadest above middle, rounded at apex, the 2 lateral broadly elliptic, ca. 7 mm. long and 5 mm. broad, broadest at middle, slightly narrowed downwardly and upwardly, obtuse, the posterior deflexed in anthesis, ca. 7.5 mm. long, broader than long and ca. 8 mm. broad, carinate and somewhat folded around the corolla spur, prominently apiculate, the midrib thickened; corolla orange, ca. 25 mm. long, horizontal in calyx, spurred at posterior base, the spur ca. 2.5 mm. long, acutish, the basal orifice ca, 2.5 mm. broad, the tube ca. 4 mm. broad at little above the base, strongly enlarged upwardly and becoming ventricose-saccate at apex, here 15 mm. broad, glabrous without and within, the limb regular, dorsal and not terminal, facing upwardly, ca. 6 mm. broad, the lobes minute, semicircular, ca. 1.5 mm. long, 3 mm. broad, glabrous, not ciliate; filaments thick, ca. 12-15 mm. long, not coiled, glabrous; anthers included within the saccate portion of the tube, free at an- thesis, glabrous, broader than long, ca. 1.5 mm. long, 2 mm. broad, the cells completely confluent in anthesis; ovary conic, 3-4 mm. long, entirely glabrous; style included, thick, glabrous, ca. 8 mm. long; stizma bilobed; disk essentially reduced to a large, thick, fleshy, bilobed, glabrous, posterior gland ca. 1.2 mm. high and 1.3 mm. thick, the anterior portion reduced to a mere remnant; calyx lobes somewhat accrescent after anthesis; fruit unknown. Type in the U.S. National Herbarium, no. 2,404,445, collected in rain forest on the lower northwestern slopes of Cerros Campanquiz, near the Rio Marajion just above Pongo de Manseriche, Province of Alto Amazonas, Department of Loreto, Peru, at 250-350 meters elevation, Oct. 17, 1962, by J. J. Wurdack (no. 2277). Isotypes in Lima (not seen) and Kew. This is the first species of the section Gasteranthus to be known from Peru. The plants of this section have slipper-like flowers, with the inflated corolla tube terminal and the orifice of the corolla minute and facing upwardly on the upper side some distance back from the tip of the corolla. The peculiar form may be correlated with the mode of pollination, but this is unknown; the brightly colored flowers of yellow, orange, or red probably indicate an insect pollinator. The other species of this section are confined to Ecuador and Colombia, with one outlier in Guatemala. The most closely related species 1s Besleria calceolus Fritsch (Repert. Sp. Nov. Fedde 18: 12. 1922), which is still known only from the type collection from the Montafa de Canelos, Ecuador (Spruce 5069); since the holotype at Berlin is probably destroyed, I here designate the specimen of this collection in the Royal Botanic Gardens, Kew, as lectotype. The epithet ‘“‘cal- 150 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM ceolus” chosen by Fritsch means “‘little slipper” in Latin, and I have chosen another equivalent ‘‘caligula.” These two species may be distinguished by the following key: Calyx lobes unequal, the 2 anterior obovate, the 2 lateral broadly elliptic, the posterior broader than long, all 7-7.5 mm. long, entirely glabrous; petioles ca. 1 cm. long; stems and leaves entirely glabrous; primary veins 12-15 pairs; corolla glabrous externally; ovary glabrous. . . . . . . B. caligula Calyx lobes ovate-lanceolate, acuminate, subequal, the posterior somewhat broader, all 6 mm. long or less, slightly pilosulous externally; petioles 2-4 cm. long; stems at apex and leaf veins beneath minutely strigillose; primary veins 9 or 10 pairs; corolla tube sparsely pilosulous without; ovary with a few hairsat apex ............4.4.4.4. 4... . B. ealceolus (Synonyms in italics. New species, new names, and combinations in boldface. Page numbers of principal entries in italics.) Besleria, 125, 126, 133, 148 aggregata, 126, 136 var. aggregata, 126, 127 var. loretensis, 127, 137 var. mexiae, 126, 127, 137 var. semiannularis, 137 angusta, 126, 135 barbata, 127, 132 boliviana, 133 calceolus, 149, 150 caligula, 129, 148, 150 capitata, 131, 144 citrina, 127, 137 compta, 128, 140 concolor, 129, 134 corallina, 128, 130, 145, 147 densiflora, 127, 132 divaricata, 130, 141 elongata, 141 emendata, 128, 137, 139 ferreyrae, 128, 139 furva, 127, 137 gibbosa, 145, 146 gracilenta, 126, 132 horizontalis, 127, 144 hutchisonii, 128, 130, 142 ignea, 136 var. semiannularis, 137 imberbis, 127, 132 immitis, 128, 135, 139 laxiflora, 143 lucida, 128, 134 lutea, 131 membranacea, 126, 135 minutiflora, 129, 134 moorei, 130, 141, 142 obliqua, 147 ovalifolia, 142 pallidiflora, 129, 134 peruviana, 139, 144 var. dissimilis, 139, 144 var. occulta, 130, 144 var. peruviana, 130, 144 placita, 128, 132 quitensis, 148 racemosa, 131, 142, 144 rupestris, 128, 130, 147 solanoides, 126, 131, 132 sprucei, 136 var. flavescens, 131, 136 var. sprucei, 131 tetrangularis, 129, 134 trichiata, 130, 131, 144 tuberculata, 139 vargasii, 129, 135 variabilis, 134 f. barbatula, 129, 134 f. pallida, 129, 134 f. variabilis, 129, 134 wurdackii, 128, 144, 146 sect. Besleria, 131 sect. Gasteranthus, 148, 149 sect. Neobesleria, 144, 147 subsect. Axillares, 145, 146, 147 subsect. Confertae, 135, 136, 139, 140 subsect. Corallinae, 147 subsect. Elongatae, 139, 141, 142, 143 subsect. Maculatae, 147 subsect. Pedunculatae, 134 subsect. Sessiles, 131, 133 Gasteranthus, 148 Hypocyrta aggregata, 136 barbata, 132 U.S. GOVERNMENT PRINTING OFFICE: 1968 BULLETIN OF THE UNITED STATES NATIONAL MUSEUM CONTRIBUTIONS FROM THE UNITED StatTEs NATIONAL HERBARIUM VouumE 38, Part 5 THE GENERA, SUBGENERA, AND SECTIONS OF THE HYMENOPHYLLACEAE By C. V. Morton ERRATA Page 182, line 21: delete "T, cellulosum kl." Page 1%, line 16: read ",..includes only the three species T. arbuscula Desv., I. bicorne Hook., and T. cellulosum Kl...." Page 211: under "I. cellulosum," delete 182, add 1%. SMITHSONIAN INSTITUTION PRESS e WASHINGTON, D.C, ¢ 1968 THE GENERA, SUBGENERA, AND SECTIONS OF THE HYMENOPH YLLACEAE Cc. V. Morton Pteridologists have been attracted to the filmy-fern family Hy- menophyllaceae for many years because of the variety and beauty of its species and its peculiar morphological structure. Presl, van den Bosch, Mettenius, Prantl, and Giesenhagen were all interested in the group and each contributed many important observations. Most of these workers recognized two large traditional genera, T’richomanes L. and Hymenophyllum J. E. Smith, although Pres] and Prantl did segregate some additional genera, in Presl’s case sometimes by faulty observations. The oriental species were revised some 30 years ago by the late E. B. Copeland; no real monograph has ever been produced, although one is needed since these are among the commonest plants of the tropical forests at middle elevations in both the Old and the New Worlds. Christensen in the “Index Filicum” gave two classifications of Trichomanes that are contradictory in part..On page 634 of the Index (issued July 2, 1906) he named seven subgenera and arranged the species by means of symbols, following Prantl’s classification. In the introductory pages of the Index (pp. XIV, XV, after Sept. 15, 1906) he gave a different and better classification, which has, however, been less known. Here he recognized eight subgenera with many sections, almost all with a lectotype indicated, although this was wrongly chosen in some instances according to our present Interna- tional Code of Botanical Nomenclature. The late E. B. Copeland published in 1933 a revision of the Old World species of Trichomanes! and a similar treatment of the Old World Hymenophyllum in 1937,? both beautifully illustrated with line drawings. In Trichomanes he recognized a number of groups called Pyzidifera, Scandentia, Grandia, Apiifolia, Rigida, etc., but these cannot be considered as validly published names, since “group”’ is not a recognized taxonomic category. In Hymenophyllum, however, he 1 Phil. Journ. Sci. 51: 119-280, pls. 1-61. 2 Op. cit. 64: 1-188, pls. 1-89. 154 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM recognized a number of groups definitely called subgenera. In 1988, Copeland ® in his “‘Genera Hymenophyllacearum”’ raised most of the previously recognized groups and subgenera to generic rank, a total of 33 genera being recognized. A few years later, I published a review 4 critical of this work and questioned the necessity of recognizing so many genera, some of which were separated by characters hardly more than specific it would seem, such as “fronds harsh in texture’ opposed to ‘fronds soft in texture,” or “fronds once-pinnate” opposed to “fronds more divided,” or “margin naked” opposed to “margin hairy,” and so forth. Essentially the same treatment of the family was presented by Copeland in his ‘Genera Filicum” (1947), but during continued work in the last 25 years I have seen no reason for changing my previous opinion. Due to the prestige of Copeland, his system has been accepted by many workers without a truly critical examination of the basis of this system; however, the most eminent students, such as Christensen, Alston, Madame Tardieu-Blot, E.A.C.L.E. Schelpe, W. A. Sledge, R. M. Tryon, and R. E. Holttum have not accepted it. Copeland’s defense of his system is that the traditional character of distinction, a 2-valved, only partly immersed involucre in Hymenophyllum and a tubular, almost wholly immersed involucre in Trichomanes, is a vari- able and unreliable character. Nevertheless, he uses only this character of “involucre valvate” as opposed to “involucre tubular or obconic” to separate his groups of genera in the keys both in the ‘Genera Hymenophyllacearum” and the ‘Genera Filicum” without seeming to realize that he is contradicting himself. If a multitude of genera can be separated by this character then the two traditional ones can be also. As a matter of fact, about 98 percent of the species fall easily into one group or the other without difficulty. The presence of a few species that are somewhat equivocal in this character need not in- validate the genuine unity of the traditional genera Trichomanes and Hymenophyllum, which are not separated solely by this character, as Copeland would seem to imply. There are other differences, as one would expect between good genera, and a fern specialist with some experience with the family can almost always place sterile specimens in one genus or the other correctly and without difficulty. The differ- ences are a little subtle perhaps. Hymenophyllum usually has a more elastic frond, which often curls up on the sheet when dried; the fronds are usually a red-brown when dry, whereas those of Trichomanes are more usually persistently green. Hymenophyllum is always epiphytic or rupicolous, whereas Trichomanes is not infrequently terrestrial. The rhizones of Hymenophyllum are always widely creeping and the 3 Op. cit. 67: 1-110, pls. 1-11. ‘ Amer. Fern Journ. 32: 30, 31. 1942. MORTON—HYMENOPHYLLACEAE 155 fronds scattered, and they are always relatively slender and with pale rhizome hairs. Trichomanes often has thicker rhizomes, and in some groups these are contracted so that the fronds are clustered; they often have dark, blackish rhizome hairs. The stipes, too, tend to be thicker. The fronds of Hymenophyllum are usually more divided and simple blades do not occur, except in the aberrant H. marginatum. The cells of the fronds of Trichomanes often are larger and coarser, with thicker walls, these sometimes pitted. A number of species of Trichomanes are characterized also by the presence of false veins, which do not occur in Hymenophyllum. The segment margins are sometimes toothed in Hymenophyllum, never in Trichomanes. Stalked, stellate hairs occur in many Hymenophyllum (although there are many glabrous species also), and such hairs are perhaps not found in T7- chomanes, although a few species do have sessile, stellate hairs. The venation pattern in Hymenophyllum is always anadromous and the soriation “paratact” (see p. 158), as Prantl termed it, but in Trichom- anes many species are catadromous and “epitact,” a character that seems to me to be fundamental but which Copeland essentially ignored. As pointed out by Bower, Stokey, and others, important gametophytic differences also occur between the two genera, although relatively few species have been investigated in this regard. Trichom- anes and Hymenophyllum seem to differ in a great many more characters than merely the form of the involucre, important as that is. Therefore, it seems right that the traditional division into two major genera be maintained, and that the convenience of the general botanist and taxonomist will be better served by this treatment. It is clear that by the characters noted above the general botanist can recognize the traditional genera rather readily, but that only a special- ist could recognize and remember the 33 genera of Copeland. A word as to Copeland’s general philosophy may well point up our basic difference in views. Although he would have vigorously dis- claimed it, Copeland was an old-fashioned botanist in his general ideas, having been brought up under the influence of L. M. Underwood and N. L. Britton in the ‘New York” school of taxonomy back at the turn of the century. In general, he kept to the ideas of this school of thought that a plant was either a good species or it was nothing at all, a mere synonym. In his later years, at least, he never recognized any subspecies or varieties (cf. his treatments of the ferns of New Guinea, the “Fern Flora of the Philippines” and his “monographs” of Grammitis and Ctenopteris). Similarly, his larger groups were either full-fledged genera or else synonyms, with never or seldom any subgenera, sections, or subsections; instead all the species were massed together with no grouping. Similarly at the family level, Copeland split up the traditional Polypodiaceae into a number of 156 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM families, but he never recognized any subfamilies, tribes, or subtribes within them. There was no classification of the genera. He was thus essentially destroying the classification that had been carefully built up by Diels, Christensen, and previous workers. It was a flat- tening and leveling out process whereby only essentially three cate- gories, family, genus, and species, were recognized. With such a philos- ophy it is understandable that he would find smaller families and smaller genera more “convenient.” However, this is not the traditional conception that has come down to us from the botanical masters of the past, who recognized that there are such things as subfamilies and tribes, that subgenera and sections do exist, and that species are not uniform but are divisible into subspecies and varieties. It is true that one can go to an extreme in subdividing ad infinitum, as was the case with some of the workers of the Englerian school, but this is preferable to an absence of classi- fication, provided of course that the subdivisions are reasonably homogeneous and natural, as they are apt to be if subdivided finely enough. A middle course is certainly preferable. Just how much segre- gation into named groups is desirable depends largely on the size and variability of the group concerned. In the case of the Poly- podiaceae, no one has shown that it is necessary to recognize a large number of segregate families (or even a smaller number), when the recognition of these groups as subfamilies is equally convenient. The only justification would be on theoretical grounds that these “families” had a different origin, i.e., that the family Polypodiaceae sens. lat. is polyphyletic. This has been claimed but not proved and is logically unlikely considering the relative uniformity of the Poly- podiaceae sens. lat. in fundamental characters. Certain subfamilies like Adiantoideae show some similarities to the more primitive family Schizaeaceae, for instance, but in their sporangia, annulus, stomium, spores, embryos, gametophytes, and sex organs the Adiantoideae are similar to the Dryopteridoideae, Asplenioideae, and other sub- families of Polypodiaceae, and a great gap exists between all of these and the Schizaeaceae. If there are relationships with the Schi- zaeaceae it would be with the whole family Polypodiaceae and not just with the subfamily Adiantoideae. The situation is quite different in the Hymenophyllaceae, a very natural group and an obvious family it would seem; however, an extreme splitter like G. Kunkel instead of minimizing the differences between Zrichomanes and Hymenophyllum has overvalued them and proposed to separate them into two different families, the Hymeno- phyllaceae and Trichomanaeaceae,® without having brought out any 5 Repert. Sp. Nov. Fedde 70: 155. 1965. MORTON—HYMENOPHYLLACEAE 157 new characters of family importance, or indeed without an adequate discussion of the existing characters. Several workers have divided the family into subfamilies, one for each of the traditional genera, but this too seems unnecessary, for the fineness of subdivision should de- pend on the size of the groups concerned. In their inclusive sense, both Hymenophyllum and Trichomanes are large enough genera that a classification of the species is surely needed, and therefore a division into several subgenera, some of them with several sections, seems to be a logical and usable procedure. The system set forth below is not in any sense original. It has not been expressed in exactly this way before but is adapted from the work of my predecessors. I have tried to pick out the best ideas of Presl, van den Bosch, Prantl, Christensen, and Copeland, and to unite them into a system. It is based in part on general knowledge of the family, acquired over about 38 years of study, but not on a really intensive study of all the species, as it ideally should. A number of species are almost unknown to me, and these are some of the most critical ones so far as their placing in the system is concerned. For this reason I leave some groups in an uncertain position, pending further study, by me or others. No doubt their relationships can be determined definitely when proper material is available and is studied completely. Unfortunately, some of these critical species are extremely rare and hard to come by in good condition. The trickiest and in a sense one of the most difficult problems has been to typify properly the various names that have been proposed. Our present Code rightly requires that names be typified, or other- wise their application is indefinite and subject to change. Types were suggested for many names by Christensen in the preliminary pages of the “Index Filicum” (after Sept. 15, 1906) and by Copeland, but the choices have not always been in accord with the current rules set down in the Code. In particular, Christensen often designated as type a species that was not one of the original species of the genus, but which he thought, rightly or wrongly, to be a taxonomic synonym. Copeland’s chief idea in choosing a type was to take the first species cited, but this arbitrary system is not only unacceptable but is spe- cifically condemned in the Code, which indicates that all the species cited must be taken into account, and in particular that the designa- tion of a lectotype by a previous author must stand unless it can be shown that is does not agree with the original description or was selected by some sort of erroneous conception. However, this study of typification has shown that almost all the groups I accept already have available acceptable names. It might be thought that chromosome numbers would give some useful information as to the proper subgeneric and sectional distri- 158 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM bution, and it is likely that they will ultimately. However, only relatively few species have been counted and hundreds of additional counts of properly identified material will have to be taken before any definite statements can be made. At the present time the informa- tion is extremely indecisive and inconclusive. The Hymenophyllaceae have relatively low chromosome numbers, compared to the ferns in general, and so it would appear that counts would be made more easily, but this does not seem to be the case. It appears that several basic numbers are in the family (t=7, 9, 11, 13, or 17), but different basic numbers appear in closely allied species that obviously belong in the same section, the group known as “Meringium” having basic numbers of 7, 11, 13, and 21 reported. Indeed within the typical and small section Hymenophyllum itself there is no agreement: The type species of the genus H. tunbridgense is reported by Manton as n= 13; Brownlie reported that H. peltatum has n=11, and yet Manton reported n=18 in H. unilaterale, which is not only close to H. peltatum but is generally regarded and probably correctly as a complete synonym. It is likely that the species counted by Manton was really FH. wilsonii Hook., generally considered another synonym of H. peltatum, and may not differ other than in chromosome count. In all events they are closely allied and belong in the same section. So far as known, the species of Copeland’s groups “Mecodium” and “Sphaero- cionium” are more uniform, often having n=36, which is assumed to be from a base of z=9. The genus Trichomanes gives an indication of different base numbers from Hymenophyllum, although z=9 does occur too. Trevor Walker has given a most interesting and valuable summary ° of the results of his counts of 19 Jamaican species, which seems to show that base numbers of z=7 and z=8 are operative, although r=11 and 2=17 also occur. It appears from the data available that subg. Didymo- glossum is uniform with n=34 or 68 (presumably 2=17), subg. Achomanes often at least has n=32 or 64 (presumably z=8), and subg. Trichomanes (as I recognize it) at least in part as z==9 (although other numbers such as z=17 also occur). The study of chromosome numbers in the future will be fascinating and will certainly shed new light on relationships and appropriate taxonomic groupings. However, some of the reported counts should be regarded witb great discretion, for the numerous discrepancies may indicate errors in counting or errors in identification, or both. A word should be said about the key characters used as primary ones. The terms “paratactic” or “paratact” and “epitactic” or “epitact”’ were invented by Prantl’ to describe the positioning of the sori in * Trans. Roy. Soc. Edinburgh 66: 169-237, pls. 1-5. 1966. 7 Untersuch. Morph. Gefisskrypt. 1: 3-14. 1875. MORTON—HYMENOPHYLLACEAE 159 the Hymenophyllaceae, and this positioning is correlated with the venation type. In anadromous venation, the first vein of a primary pinna or the first veinlet of a secondary pinnule points toward the apex of the blade or the pinna, respectively. The sorus is borne on the tip of a proximal veinlet, so that a distal vein or veinlet is free to con- tinue growth and branch further; in other words there is at least the possibility of indeterminate growth. This is the paratactic type. Catadromous venation, in which the first vein faces downward, toward the base of the frond is accompanied by sori that terminate distal veins or veinlets. A proximal veinlet may or may not also be terminated by a sorus. The result is that such fronds with epitactic soriation are determinate, both in the growth of the fronds themselves and the individual pinnae, at least theoretically. After a pinna or a frond apex becomes fertile no more growth is possible since the distal veins that could have continued growth have been stopped by the production of a sorus. Of course, some fronds do continue to grow at the apex because they have not become fertile, and so this character is best shown by the lateral pinnae, and especially by the positioning of the sorus on a distal branch of a veinlet rather than a proximal. Although Copeland knew of this character from Prantl’s exposition, he minimized its importance and did not truly employ it in his classi- fication. It is probably a fundamental character, although sometimes it may be difficult to interpret, especially in plants where the fronds have become very dwarfed or even simple (as in some species of subg. Didymoglossum), when the character of catadromous venation can hardly be made out, for this is a character, like rhachis structure, that always is more determinable in compound fronds. Occasionally, especially in simple-bladed types like Hymenoglossum and Cardiomanes, all the veinlets may be fertile, a type of soriation called “pantotactic” by Prantl, an advanced type obviously correlated with simplicity. The lists of species belonging to the various sections given in this paper do not represent in many cases My own personal study. I do know many of the American species, but the Old World ones are largely placed following the list given by Copeland in his ‘‘Genera Hymenophyllacearum.” This list is in turn partly the result of Cope- land’s own studies and partly taken from Christensen’s ‘Index Filicum.” And so, although most species will be properly placed, some errors may need to be corrected by future studies. There are many more names than are mentioned here. Most of these are syno- nyms, the disposition of which can be found from the “Index Filicum”’ and from Copeland’s papers. Some others have not been studied and are of doubtful disposition at present. 160 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM Key to the Genera of Hymenophyllaceae Involucre bivalved throughout or at least to the middle, the immersed part, if any, cuplike or conic and not tubular; fronds usually brown or red when dried, glabrous or often hairy, generally elastic, the margins entire or toothed; rhizomes always epiphytic or rupicolous, long-creeping with scattered fronds, usually slender, with pale rhizoids; cells of blade often small, with smooth walls, but not always; false veins absent; venation anadromous; sori paratactic or pantotactic; receptacle typically included or only slightly exserted from involucre, typically sporangiiferous toward apex only. Frond simple, large, long-stipitate, ovate to lanceolate, subentire, the marginal cells in two rows and forming a dark border, the laminal veins parallel. Chile. 2... . ee. ee 2. 1 Hymenoglossum Fronds not simple, large, long-stipitate, or subentire. Fronds minute and mosslike, simply pinnate with entire, leaflike pinnae, the stipe bearing rhizoids and scarcely distinguishable from the rhizomes, the axis red-pilose. Antarctic South America . . . . IL Serpyllopsis Fronds otherwise, the stipe not bearing rhizoids. Ultimate segments (pinnae) large, dimidiate, with several dark black, thick, forked veins, the outer curving around in a submarginal position but leaving several rows of undifferentiated marginal parenchyma cells. New Caledonia... . 2... 0.0.0.0... . .. ID. Rosenstockia Ultimate segments without plural veins, always with a single vein, this not black (or in H. heimii the blade simply pinnatilobate). IV. Hymenophyllum Involucre tubular or conic throughout with a truncate apex, or with two lips less than half as long as the tube; fronds often green, mostly glabrous but some- times hairy, but then usually without stellate hairs, generally not especially elastic, the margins entire; rhizomes terrestrial or epiphytic, often thicker than in Hymenophyllum, widely creeping or contracted or even erect, with scattered or approximate fronds; cells of blade often large, and sometimes with thickened walls; false veins in the position of veins or marginal false veins sometimes present; venation anadromous or catadromous; sori paratactic or epitactic, or rarely pantotactic; receptacle typically exserted, typically sporangiiferous to the base but not to the apex. Fronds simple, reniform, large, long-stipitate, ca. 4 cells thick. New Zealand. Vv. Cardiomanes Fronds not simple, reniform and long-stipitate. . . . . . VI. Trichomanes I. Hymenoglossum Hymenoglossum Presl, Hymen. 35. 1843, Hymenophyllum sect. Hymenoglossum (Presl) Moore, Ind. Fil. exii. 1857. Type: Hymenophyllum cruentum Cav.= Hymenoglossum cruentum (Cav.) Presl. Although in general characters this genus is not particularly different from Hymenophyllum, it is so peculiar and isolated that its affinities really are not known and it can rank as a monotypic genus. The 2- rowed, differentiated band of black marginal cells are unlike any other marginal cells in Hymenophyllum, although they may be somewhat comparable to those of Hymenophyllum subg. Craspedophyllum, in which the marginal cells are dark and thickened , but differently placed MORTON-——HY MENOPHYLLACEAE 161 and in only one row. This species appears to be quite common in certain parts of the western slopes of the Andes in Chile, but appears not to have been able to cross the Andes into the comparable region of Argentina on the eastern slopes. II. Serpyllopsis Serpyllopsis v. d. Bosch, Versl. Meded. Konink. Akad. Wetens. Afd. Natuurk, 11: 318. 1861. Hymenophyllum subg. Cycloglossum Presl, Hymen. 32. 1843. Lectotype: There were six original species: H. caespitosum Gaud., H. cumingit Presl, H. semibivalve Hook. & Grev., H. decurrens (Jacq.) Swartz, H. kohautianum Presl, and H. schomburgkii Pres (ined.). Christensen 8 indicated the type as H. decurrens (Jacq.) Swartz, but inasmuch as this species has never been identified and presumably was as unknown to Presl as it has been to others, it is not a suitable choice of lectotype. Copeland ® indicated the type as H. caespitosum Gaud., probably on the basis that this was the first species listed by Presl, and this must be accepted as a choice of lectotype. It is in fact a desirable lectotype, for if another lectotype were to be chosen subg. Cycloglossum would have priority over and would have to replace subg. Mecodium, which is now a fairly well-known name due to Copeland’s use. Trichomanes subg. Serpyllopsis (v.d. Bosch) Christ, Farnkr. Erde 23. 1897. Hymenophyllum subg. Euhymenophyllum sect. Cycloglossum (Presl) C. Chr. Ind. Fil. XV. 1906. Typ: Trichomanes caespitosum Gaud.=Serpyllopsis caespitosa (Gaud.) C, Chr. This small, monotypic Antarctic genus is of uncertain relationship, showing characters of both Hymenophyllum and Trichomanes. In its general appearance it suggests a bryophyte rather than a member of the Hymenophyllaceae, although it is certainly of the family. III. Rosenstockia Rosenstockia Copel. Gen. Fil. 36. 1947. Type: Hymenophyllum rolandi-principis Rosenst.= Rosenstockia rolandi-prin- cipis (Rosenst.) Copel. This is one of the most peculiar and distinctive of all fern genera, and probably one of the rarest, known only from about three collections from New Caledonia. The structure of the dark, coarse veins has not been investigated anatomically, but it may very well be quite different from the normal vein structure of the Hymenophyllaceae. The alliance, by reason of the involucres, is surely with Hymenophyllum. The structurally somewhat similar group in Trichomanes with large seg- ments with plural veins (subg. Phlebiophyllum) is less distinctive and more typical of its genus in other characters. 8 Ind, Fil. XV. 1906. * Phil. Journ. Sci. 67: 14. 1938. 162 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM IV. Hymenophyllum Hymenophyllum is a large, common genus in the tropics of both hemispheres, and it is well known and usually recognized by the general botanist. It really is quite uniform and its segregation into small genera is not only inconvenient but unnecessary, since all the species are closely related. Several writers have remarked that Cope- land’s segregates from Hymenophyllum are less distinct than those from Trichomanes, as by Miss Tindale in her recent fine treatment of the species of New South Wales,!° where most of the segregates are kept within Hymenophyllum as subgenera but those from Trichomanes recognized as genera. This is possible and somewhat a matter of opinion. However, the easier recognizability of some of the segregates from Trichomanes (but not all, by any means) does not require that they be recognized as genera rather than subgenera. Trichomanes is the larger genus and has varied more, for instance by sometimes assuming a terrestrial habitat, which necessitates a change in the rhizomes and in frond disposition. In its general structure and espe- cially in its involucres Trichomanes is rather uniform and readily separated from Hymenophyllum, with the exception of a few aberrant species. There is no reason to suspect a polyphyletic origin and thus no theoretical or practical need to recognize a large number of segregate genera. Convenience in identification work, in anatomical and cytological study, and in herbarium arrangement is equally well served by recognizing a suitable number of subgenera and sections. Key to Hymenophyllum Margins of the segments toothed, the teeth not tipped by hairs or if rarely so, the hairs several cells long; involucre often somewhat conic at base. Subg. 1. Hymenophyllum Margins of the segments entire, or if rarely apparently slightly denticulate, the teeth tipped by one-celled or stellate hairs; receptacle typically included within the involucre, but sometimes exserted, the involucre rounded, cup- like, or conic at base. Blades variously hairy, at least on the margins or veins beneath. Subg. 2. Sphaerocionium Blades glabrous. Blades with a conspicuous row of black, differentiated marginal cells, simple or forked, one-veined. Australia, Tasmania, and New Zealand. Subg. 3. Craspedophyllum Blades without a black margin or differentiated cells, pinnatifid to decompound. Involucres immersed and urceolate and cuplike at base. Australia. Subg. 4. Hemicyatheon Involucres mostly free at base or if immersed then usually not rounded andcuplike. ..... Subg. 5. Mecodium 0 Contr. New South Wales Nat. Herb., Flora Ser. 201: 1-49, pls. 1-7. 1963. MORTON—HYMENOPHYLLACEAE 163 Subg. 1. Hymenophyllum Lamina with normal laminal surface, sometimes with accessory teeth. Sori directed in a plane more or less perpendicular to the plane of the lamina, arched at the base and somewhat pedunculate; receptacle clavate, almost covered with sporangia to the base, only slightly exserted. Sect. 1. Hymenophyllum Sori directed in the plane of the lamina, not arched at the base or pedunculate; receptacle cylindrical to globose, generally without sporangia at the base, often exserted. Teeth or wings present in planes perpendicular to the blade; receptacle globose to clavate. Peruand Ecuador. ..... . . Sect. 2. Buesia Teeth or wings absent in planes perpendicular to the blade; receptacle clavate to linear. Pinnae symmetrical, more or less equally developed on both sides; recep- tacle cylindric, thickened at base, exserted. . Sect. 3. Ptychophyllum Pinnae asymmetrical, with segments developed mostly or entirely on the acroscopic side; receptacle filiform, somewhat slender at the base, thickened at the middle, included. South America. Sect. 4. Eupectinum Lamina reduced mostly to discrete teeth spreading in various planes. New Guinea. Sect. 5. Myriodon Subg. 2. Sphaerocionium Blades essentially reduced to filiform processes bearing stellate marginal hairs. New Zealand ..... to. ew ew ee a ew ee 6 Sect. 7. Apteropteris Blades with normal laminar surfaces . ee a we ee . Sect. 6. Sphaerocionium Blades with hairs borne only on the margins and veins (sometimes only on the veins beneath) . Loe wee . .. . . . Subsect. Ciliata Blades with the hairs borne on ‘the leaf-surfaces as well as on the margins and veins. Veins with accessory wings not in the plane of the fronds. West Indies; Central and South America . . . . . . + « Subsect. Plumosa Veins lacking accessory wings not in the plane of the frond. Hairs, or at least some of them, stellate and stalked . . Subsect. Hirsuta Hairs all simple or basally paired. Southern South America. Subsect. Leptocionium Subg. 3. Craspedophyllum A single section. Australia, Tasmania, and New Zealand. Sect. 8. Craspedophyllum Subg. 4. Hemicyatheon A single section. Australia. . ...... .. ss. . Sect. 9. Hemicyatheon Subg. 5. Mecodium Blades one cell thick throughout (except the veins). . . . Sect. 10. Mecodium Veins without accessory wings not in the plane of the lamina. Subsect. Mecodium Veins with accessory wings not in the plane of the lamina. Java and New Guinea. Subsect. Amphipterum 164 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM Blades 2- or 3-celled thick at least in part. Australia and New Zealand. Subsect. Diplophyllum Subg. 1. Hymenophyllum Hymenophyllum sect. 1. Hymenophyllum Hymenophyllum J. i. Smith, Mem. Acad. Turin 5: 418. 1793. Lectotype: Hymenophyllum tunbridgense (L.) J. E. Smith, selected by Presl, Hymen. 31. 1843. Copeland in his revision of Hymenophyllum "| indicated that H. tunbridgense was the type because it was the sole original species, which is not true; there were nine original species, of which Trichomanes tunbrid- gense 1.., the first named, was the logical choice for a type, since it was the oldest and best known. Ptychomanes Hedwig, Fil. Gen. & Sp. in nota sub Trichomanes asplenioides. 1800. An illegitimate change of name of Hymenophyllum J. E. Smith on etymological grounds; it has therefore the same type species, 7'richomanes tunbridgense L. Hymenophyllum subg. Euhymenophyllum Pres, Hymen. 31. 1843. Hymenophyllum subg. Sphaerodium Presl, Hymen. 31. 1843. Lectotype: Hymenophyllum wilsonii Hook., selected by Copeland, Phil. Journ. Sci. 67: 14. 1938. In the “Index Filicum” (p. XVI) the type is erroneously indicated as HH. tunbridgense 1., which was not one of the original species of this sub- genus, and which in fact Presl was trying to distinguish from Euhymeno- phyllum, as typified by H. tunbridgense, but on what grounds is not en- tirely clear. Hymenophyllum subg. Leptocionium sect. Sphaerodium (Presl) C. Chr. Ind. Fil. XVI. 1906 (lectotype 7. tunbridgense incorrect, as indicated above). Hymenophyllum subg. Euhymenophyllum [sect.] Serrulata Presl, Hymen. 32. 1843. Lectotype: H. tunbridgense L. Hymenophyllum subg. Euhymenophyllum [sect.] Serrulata [subsect.] Fvoluta Pres], Hymen. 32. 1843. Lectotype: H. tunbridgense L. Hymenophyllum subg. Euhymenophyllum [sect.] Serrulata {subsect.] Dimidiata Presl, Hymen. 32. 1843. Lectotype: H. unilaterale Bory. The section Hymenophyllum contains according to Copeland the following presumed species: FH. alveolatum C. Chr., H. antarcticum Presl, H. asperulum Kunze, H. barbatum Baker, H. brachypus Sodiro, Hf. ceratophylloides Christ?, H. cupressiforme Labill., H. dentatum Cav., LH. falklandicum Baker, H. gracilescens Domin, HH. herzogi Rosenst., H. mooret Baker, H. nahuelhuapiense Diem & Lichtenstein, H. nanuwm Sodiro, H. peltatum (Poir.) Desv. (incl. presumably H. wilsonii Hook.), ET. perfissum Copel., H. pumilum C. Moore, H. revolutum Colenso, H. rugosum C. Chr. & Skottsb., H. secundum Hook. & Grev., H. simon- stanum Hook., H. spicatum Christ, H. torquescens v.d. Bosch, H. tun- bridgense (L.) J. E. Smith, and H. uncinatum Sim. Hymenophyllum sect. 2. Buesia (Morton) Morton, comb. nov. Hymenophyllum subg. Buesia Morton, Bot. Gaz. 93: 336. 1932. Type: Hymenophyllum mirificum Morton. Buesia (Morton) Copel. Phil. Journ, Sci. 67: 47. 1938. " Phil. Journ. Sci. 64: 77. 1937. MORTON—-HYMENOPHYLLACEAE 165 When I described this as a subgenus back in 1932 I really knew very little about the Hymenophyllaceae in general. The type species, H. mirificum, was so strange that it seemed to be subgenerically distin- guishable from Meringium (or Leptocionium, as it was called at that time). The ‘species is peculiar and still known only from the original material, but relatives are now known that are less distinctive. I am now uncertain that it even needs to be distinguished as a section, but it can perhaps separate a small group of species characterized by hav- ing some accessory wings not in the plane of the frond. The subglobose receptacle found in the type species is not uniformly present in others. The additional species referred here are: H. eristatum Hook. & Grev., H. jamesonii Hook., H. megistocarpum (Copel.) Morton, comb. nov. (Buesia megistocarpa Copel. Univ. Calif. Publ. Bot. 19: 295: 1941), H. sodirot C. Chr., and possibly one or two additional undescribed ones. Hymenophyllum sect. 3. Ptychophyllum (Presl) C. Chr. Ind. Fil. XVI. 1906. Didymoglossum subg. Chilodium Presl, Hymen. 23. 1843. Lectotype: T'ri- chomanes neesit Blume. The subgenus originally consisted of two unrelated species, 7’. neesit Blume and Didymoglossum longisetum Presl; the first is a Hymenophyllum and the second a T'richomanes of the section Pachy- chaetum. The first to typify the group was Christensen in the ‘Index Filicum” (p. XVI), who chose the Hymenophyllum element, citing as type H. denticulatum Swartz, which he considered the correct name, taxonomically, for Trichomanes neestt Blume. This fixes the lectotype as T'. neesit. Meringium Presl, Hymen. 24, pl. 8. f. B. 1843. Type: Meringium meyenianum Presl= Hymenophyllum meyenianum (Presl) Copel. There were two original species, but the second, M. blumeanum (Spreng.) Presl, was referred to the genus only with a query, and so M. meyenianum is unquestionably the type. Myrmecostylum Presl, Hymen. 27, pl. 10, f. A. 1843. Lectotype: Myrmecosiylum tortuosum (Hook. & Grev.) Presl= Hymenophyllum tortuosum Hook. & Grev. (chosen by ©. Chr. Ind. Fil. XVI. 1906). There were two other original species, M. dichotomum (Cav.) Presl and MM. clavatum (Swartz) Presl; the former was referred to the genus only with a query and the latter disagrees with the description, so Christensen’s choice was the only possible one. Piychophyllum Presl, Hymen. 28, pl. 11, f. E. 1843. Type: Ptychophyllum plicatum (Kaulf.) Presl= Hymenophyllum plicatum Kaulf., the only original species. In the “Index Filicum” (p. XVI. 1906) Christensen indicated the type as H. dichotomum Cav., but this is an obvious error, since it was not one of the original species. Tetralasma Philippi, Linnaea 30: 208. 1860. An invalidly published provi- sional name, proposed in case the character of a 4-parted indusium should prove of generic significance. Copeland indicated (Phil. Journ. Sci. 67: 45. 1938) that this indusial character is probably teratological. Based on Hymenophyllum quadrifidum Philippi, considered by Diem & Lichtenstein ¥ as a synonym of Hymenophyllum plicatum Kaulf. 122 Darwiniana 11: 675. 1959. 166 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM Hymenophyllum subg. Leptocionium sect. Chilodium (Presl) C. Chr. Ind. Fil. XVI. 1906. Hymenophyllum subg. Leptocionium sect. Myrmecostylum (Presl) C. Chr. Ind. Fil. XVI. 1906. Hymenophyllum sect. Acanthotheca Nakai, Bot. Mag. Tokyo 40: 242. 1926. Type: Didymoglossum acanthoides v. d. Bosch= Hymenophyllum acan- thoides (v. d. Bosch) Rosenst. Hymenophyllum subg. Meringium (Presl) Copel. Phil. Journ. Sci. 64: 14. 1937. INegitimate, because superfluous, the earlier subgeneric name Chilo- dium Presl being available for the same concept. Hymenophyllum sect. Meringium (Presl) Diem & Lichtenstein, Darwiniana 11: 666. 1959. Illegitimate, because superfluous, the earlier sectional names Chilodium, Ptychophyllum, Myrmecostylum, and Acanthotheca all being legitimate and available. As a generic name, Meringium Presl would be correct for this group, although Copeland might better have chosen Ptychophyllum or Myrmecostylum, both of the same date as Meringium, better described, and based on better known species. However, as a subgenus or as a section Meringiwm may not be used, since an earlier subgeneric name (Chilodium, which has not however been transferred to Hymeno- phyllum as a subgenus) is available, and several sectional names (Chilodium, Ptychophyllum, Myrmecostylum, and Acanthotheca) are earlier than Meringium and available. I choose Ptychophyllum, which is clearcut in its application. The competing sectional names, sect. Chilodium (Presl) C. Chr. and sect. Myrmecostylum (Presl) C. Chr. are of the same date (1906) as sect. Ptychophyllum (Presl) C. Chr., and no one has previously made a choice between them. Copeland included within Meringium four species that were said to be anomalous in having entire-margined segments. '* Hymenophyllum edentulum (v. d. Bosch) C. Chr. is not really “edentate” but has ob- vious although sparse teeth, and 77. holochilum (v. d. Bosch) C. Chr. has obviously toothed segements typical of Ptychophyllum. The other two species H. macroglossum v. d. Bosch and H. pachydermicum Cesati really do have entire segments; they were placed in Meringium on presumed affinities, but these were not demonstrated, and in fact absolutely nothing is known regarding their phylogeny or real relation- ships. They could have arisen as hybrids between two sections or they could be the result of parallel or convergent evolution. However, they probably can be accommodated in subg. Sphacrocionium subsect. Ciliata without undue strain. In this way it is possible to define Ptychophyllum more accurately, even if conceivably more artificially. Aside from the common Hymenophyllum fucoides (Swartz) Swartz, which may be an aggregate species as currently recognized, the American species of sect. Ptychophyllum are little known; they include 13 Phil. Journ. Sci. 67: 40. 1938. MORTON—HYMENOPHYLLACEAE 167 H. calodictyon v. d. Bosch, H. durandu Christ, H. ectocarpon Fée, and H. peruvianum Hook. & Grev. The following four species of southern South America have been carefully studied and accurately defined in a recent study by Diem and Lichtenstein: H. krauseanum Philippi, H. magellanicum (Klotzsch) Willd. ex Kunze, H. plicatum Kaulf., and H. tortuosum Hook. & Grev. The Old World species are much more numerous; they include: H. acanthoides (v. d. Bosch) Rosenst., H. affine Brack., H. archboldii (Copel.) Morton, comb. nov. (Meringium arehboldii Copel. Phil. Journ. Sci. 73: 462, fig. 3. 1941), H. bakeri Copel., H. bartlettii (Copel.) Morton, comb. nov. (Meringium bartlettit Copel. Phil. Journ. Sci. 73: 464. 1941), H. batwense Rosenst., H. bicolanum Copel., H. bivalve Copel., H. blandum Raciborski, H. bontocense Copel., H. brachy- glossum A. Braun, H. brevidens van Alderw. van Rosenb., H. bryo- philum C. Chr., H. campanulatum Christ, H. cardunculus C. Chr., H. cincinnatum Gepp, H. denticulatum Swartz, H. dimidiatum Mett., H. edentulum (v. d. Bosch) C. Chr., H. elbertu Rosenst., H. ellipticosorum van Alderw. van Rosenb., H. feejeense Brack., H. firmum van Alderw. van Rosenb., H. foersteri Rosenst., H. gorgoneum Copel., H. halliert Rosenst., H. hieronymii C. Chr., H. holochilum (v. d. Bosch) C. Chr., H. hoset Copel., H. howense Brownlie, H. johorense Holttum, H. kerianum Watts, H. klabatense Christ, H. latifolium (Copel.) Morton, comb. nov. (Meringiwm latifolium Copel. Phil. Journ. Sci. 73: 461. 1941), H. laxum (Copel.) Morton, comb. nov. (Meringium laxum Copel. Univ. Calif. Publ. Bot. 18: 217. 1942), H. lobbit Moore, H. macrosorum van Alderw. van Rosenb., H. melanosorum (Copel.) Morton, comb. nov. (Meringium melanosorum Copel. Phil. Journ, Sci. 73: 462. 1941), H. merrillii Christ, H. meyenianum (Presl) Copel., H. minimum A. Rich., H. multifidum (Forst.) Swartz, H. nutantefolvum van Alderw. van Rosenb., H. ovatum Copel., H. penangianum Mat- thew & Christ, H. pediculariifolium Cesati, H. piliferum C. Chr., H. perparvulum van Alderw. van Rosenb., H. poilanei Tardieu, H. pollenianum Rosenst., H. praetervisum Christ, H. pseudotunbridgense Watts, H. pulchrum Copel., H. pumilio Rosenst., H. ramosii Copel., H. reductum Copel., H. reinwardtii v. d. Bosch, H. ricewfolium Bory, H. rosenstockit Brause, H. rubellum Rosenst., H. rufifolvum van Alderw. van Rosenb., H. rufifrons van Alderw. van Rosenb., H. samoense Baker, H. subdimidiatum Rosenst., H. taiwanense (Tagawa) Morton, comb. nov. (Mecodium taiwanense Tagawa, Acta Phytotax. Geobot. 9: 141. 1940), HZ. tenellum Kuhn. H. thuidium Harrington, H. torri- cellianum van Alderw. van Rosenb., H. triangulare Baker, and H. vittatum Copel. 284-188—68——2 168 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM Hymenophyllum sect. 4. Eupectinum Diem & Lichtenstein Hymenophyllum sect. Eupectinum Diem & Lichtenstein, Darwiniana 11: 635. 1959. Type: Hymenophyllum pectinatum Cav. This small group of South American species, distinguished by having one-sided pinnae, contains only J. pectinatum, H. quetri- huense Diem & Lichtenstein, and H. umbratile Diem & Lichtenstein. Hymenophyllum sect. 5. Myriodon (Copel.) Morton, comb. nov. Hymenophyllum subg. Myriodon Copel. Phil. Journ. Sci. 64: 73. 1937. Type: Hymenophyllum odontophyllum Copel. Myriodon (Copel.) Copel. Phil. Journ. Sci. 67: 47. 1938. A small aberrant group of New Guinea, perhaps monotypic, con- taining H. brassii C. Chr., of which H. odontophyllum Copel. may be a variant. Subg. 2. Sphaerocionium Hymenophyllum subg. 2. Sphaerocionium (Presl) C. Chr. Sphaerocionium Presl, Hymen. 33, pl. 4, f. B, pl. 10, f. B, C. 1848. Lectotype: Sphaerocionium hirsutum (L.) Presl= H ymenophyllum hirsutum (L.) Swartz, chosen by Copeland, Phil. Journ. Sci. 64: 10. 1937. A large number of species were referred to the genus originally. Copeland chose the first named, a suitable choice for maintaining the application of the name in its current sense. Christensen had refused to indicate a type in the “Index Filicum.” Hymenophyllum subg. Euhymenophyllum [sect.] Pilosa Presl, Hymen. 32. 1843. Lectotype: Hymenophyllum valvatum Hook. & Grev. Presl used the sectional name “Pilosa” twice, once in Hymenophyllum and once in his genus Sphaerocionium, both for essentially the same kind of plant. The first species cited, H. valvatum, agrees with the diagnosis better than the second, H, blepharodes Presl. Sphaerocionium sect. Pilosa Presl, Hymen. 34. 1843. Lectotype: Sphaeroc- tonium schiedeanum Presl= Hymenophyllum crispum IH. B. K. (ex char.). Pres] distinguished his section by the simple rather than stellate marginal hairs. No lectotype has been designated previously. Hymenophyllum subg. Charitophyllum v. d. Bosch, in Junghuhn, Plant. Junghuhn. 562. 1856, nom. nud. Based on Hymenophyllum lineare Swartz. Hymenophyllum subg. Euhymenophyllum sect. Sphaerocionium (Presl) C. Chr. Ind. Fil. XV. 1906. Hymenophyllum subg. Sphaerocionium (Pres!) C. Chr. Ind. Fil. Suppl. 3: 5. 1934. This subgenus is divisible into two sections, Sphaerocionium and Apteropteris. Hymenophyllum sect. 6. Sphaerocionium The section Sphaerocionium is divisible into four subsections, as follows: Hymenophyllum sect. Spaerocionium subsect. Ciliata Prantl Hymenophylium subsect. Ciliata Prantl, Untersuch. Morph. Gefasskrypt. 1:55. 1875. Lectotype: Hymenophyllum lineare Swartz, the first species mentioned by Prantl and which agrees with the characterization. This MORTON—HYMENOPHYLLACEAE 169 group contains the species that are hairy on the margins or veins only and not on the surfaces. Hymenophyllum sect. Sphaerociontum subsect. Ciliata Morton, Contr. U.S. Nat. Herb. 29: 144. 1947. Invalidly proposed without Latin diagnosis. No type indicated, but assumed to be H. ciliatum (Swartz) Swartz. The subsection Ciliata includes the following American species: Hymenophyllum adiantoides v. d. Bosch, H. caparavense Brade, H. consangwineum Morton, H. erispum H. B. K., H. delicatulum Sehnem, H. dimorphum Christ, H. elegans Spreng., H. glaziovii Baker, H. hirsutum (L.) Swartz (incl. H. cilratum Swartz), H. hirtellum Swartz, H. lineare (Swartz) Swartz, H. mazonii Christ ex Morton, H. microcarpum Desv., H. molle Morton, H. prionema Kunze, H. pulchellum Schlecht. & Cham., H. roraimense Morton, H. ruizianum (Klotzsch) Kunze, H. sampaioanum Brade & Rosenst., H. semiglabrum Rosenst., H. silvaticum Morton, H. silveirae Christ, H. subrigidum Christ, H. tenerrimum v. d. Bosch, H. trapezoidale Liebm., H. trichophyllum H. B. K., H. urbanii Brause, and H. valvatum Hook. & Grev. Among the Old World species are: H. aeruginosum (Poir.) Carm., H. capillare Desv., H. hygrometricum (Poir.) Desv., H. wwohibense Tardieu, H. lanceolatum Hook. & Arn., H. macroglossum v. d. Bosch, H. marlothu Brause, H. obtuswum Hook & Arn., H. pachydermicum Cesati, H. pooli Baker, H. subobtusum Rosenst., and H. viguiert Tardieu. Hymenophyllum sect. Sphaerocionium subsect. Plumosa Prantl Dermatophlebium Presl, Epim. Bot. 17, 258. 1849 [1851], nom. nud. Hymenophyllum [subsect.] Plumosa Prantl, Untersuch. Morph. Gefasskrypt. 1:55. 1875. Lectotype: Hymenophyllum plumosum Kaulf. Hymenophyllum sect. Dermatophlebrum C. Chr. Ind. Fil. XV. 1906, nom, nud. In the “Index Filicum,” Christensen indicated the type of the unpub- lished section Dermatophlebium as H. sericeum (Swartz) Swartz, which does belong in this group, but was not the original species listed by Presl. Copeland ™ indicated several species as belonging to a group Dermatophlebium, but without a Latin diagnosis or any indication of the rank of the group intended. In my revision of the American species of Sphaerocionium I did not assign any special rank to this group of species (my nos. 45-52), H. sericeum (Swartz) Swartz, H. horizontale Morton, H. lobato-alatum Klotzsch, H. pyramidatum Desv., H. plumo- sum Kaulf., H. multialatum Morton, and H. tomentosum Kunze, of which a synonym is H. fusugasugense Karst. (taken up as a correct name in my revision), but they do deserve some nomenclatural distinction. Hymenophyllum sect. Sphaerocionium subsect. Hirsuta Prantl Sphaerocionium sect. Stellata Presl, Hymen. 34. 1843. Lectotype: Hymeno- phyllum interruptum Kunze. 14 Phil. Journ. Sci. 67: 30. 1938. 170 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM Hymenophyllum sect. Pilosa Prantl, Untersuch. Morph. Gefiisskrypt. 1: 55, 1875. Lectotype: Hymenophyllum raddianum C. Muell. Hymenophyllum [subsect.] Hirsuta Prantl, Untersuch. Morph. Gefisskrypt. 1: 55. 1875. Lectotype: Hymenophyllum raddianum C. Muell. Prantl listed also H. hirsutum (L.) Swartz, but since this species was misinterpreted and does not agree with the characterization, H. raddianum must be the lectotype. This species is a synonym of H. fragile (Hedw.) Morton. Hymenophyllum sect. Sphaerocionium subsect. Lanata Morton, Contr. U.S. Nat. Herb. 29: 144. 1947. Invalidly published without a Latin diagnosis. No type was indicated, but H. lanatum Fée assumed typical. The American species of subsect. Hirsuta are: Hymenophyllum amabile Morton, H. angustum v. d. Bosch, H. antillense (Jenm.) J enm., HI. dependens Morton, H. elegantulum v. d. Bosch, H. ferrugineum Colla, H. fragile (Hedw.) Morton, H. hemipteron Rosenst., £7. inter- ruptum Kunze, H. karstenianum Sturm, H. lanatum Fée, H. latifrons v. d. Bosch, H. lindenit Hook., H. plumieri Hook. & Grev., H. rufum Fée, H. sieberi (Presl) v. d. Bosch, H. simplex Morton, H. speciosum v. d. Bosch, and H. superbum Morton. The only Old World species are perhaps H. pilosissimum C. Chr., and H. splendidum van den Bosch. Hymenophyllum sect. Sphaerocionium subsect. Leptocionium (Pres!) Morton, comb, nov. Leptocionium Presl, Hymen. 26, pl. 11, f. D. 1843: Type: Leptocitonium dicranotrichum Presl. A second species was referred to the genus, LL. fucoides (Swartz) Presl, but since this was indicated only with a query it may not be considered the type. Hymenophyllum sect. Leptocionium (Pres}) Prantl, Untersuch. Morph. Gefasskrypt. 1: 54. 1875 (as to type, not concept). Hymenophyllum subg. Leptocionium (Presl) Christ, Farnkr. Erde 20. 1897. This subsection includes only the southern South American H. dicranotrichum. Christensen misunderstood it and misapplied the name Leptocionium as a subgenus to include all the species of sube. Hymeno- phyllum and “Meringium” in the “Index Filicum,” Supplement ITI. Hymenophyllum sect. 7. Apteropteris (Copel.) Morton, comb. nov. Hymenophyllum subg. Apteropteris Copel. Phil. Journ. Sci. 64: 176. 1937. Type: Hymenophyllum malingii (Hook.) Mett. Apteropteris (Copel.) Copel. Phil. Journ. Sci. 67: 34. 1938. The single species of this group, H. malingii, is so peculiar in its vegetative structure that it can perhaps be regarded as representing a monotypic section. The structure was well described by Holloway. Subg. 3. Craspedophyllum Hymenophyllum subg. Craspedophyllum Presl Hymenophyllum subg. Craspedophyllum Presl, Hymen. 33. 1843. Type: Hymenophyllum marginatum Hook. & Grev. 16 Trans. New Zeal. Inst. 54: 596, pls. 66, 76, 1923. MORTON—HYMENOPHYLLACEAE 171 Pachyloma v. d. Bosch, Vers]. Meded. Konink. Akad. Wetens. Afd. Natuurk. 11: 318. 1861, non DC., 1828. Type: Hymenophyllum marginatum Hook. & Grev. Hymenophyllum sect. Pachyloma (v. d. Bosch) Christ, Farnkr. Erde 15. 1897. Hymenophyllum sect. Craspedophyllum (Presl) C. Chr. Ind. Fil. XV. 1906. Craspedophyllum (Presl) Copel. Phil. Journ. Sci. 67: 27. 1938. Hymenophyllum sect. 8. Craspedophyllum Only two species are in this subgenus, H. marginatum Hook. & Grev., of Australia and Tasmania, and H. armstrongit Baker, of New Zealand. Subg. 4. Hemicyatheon Hymenophyllum subg. Hemicyatheon Domin Hymenophyllum subg. Hemicyatheon Domin, Bibl. Bot. 20, Heft. 85. 20. 1913. Lectotype: Hymenophyllum batleyanum Domin, chosen by Christensen, Ind. Fil. Suppl. 3: 5. 1934. Hemicyatheon (Domin) Copel. Phil. Journ. Sci. 67: 27. 1938. Hy menophyllum sect. 9. Hemicyatheon I do not feel confident that this entity deserves subgeneric rank. Copeland’s reasons for maintaining it are: ‘Unless the group be recognized as such, either its two members must be widely separated contrary to nature, or it must be included in Meringium where H. baileyanum would be very much out of place, or in Mecodium, where H. deplanchei would be altogether misplaced. I have already shown that the occurrence of such a group as this is explicable by assuming hybridization between members of the two great groups the characters of which are here combined.” Copeland’s argument will hold only if it is true that H. baileyanum and H. deplanchei, the first with entire margins and the second with toothed margins, are really more closely allied to each other than to any other species, but this has not actually been demonstrated. It could be that they do really belong to different subgenera and are similar by reason of convergent evolution rather than by reason of a common origin, which is by no means impossible on a priori grounds. No doubt H. baileyanum, the type of Hemicyatheon, is very close to Mecodium, and frankly the only reason that I don’t actually refer it to that subgenus is that as a subgenus Hemicyatheon has priority, and thus the great group of species known as Mecodium would all have to become Hemicyatheon, an argument that could be defended more on the grounds of practicality than theory. For the present I refer H. deplanchei to subg. Hymenophyllum. Subg. 5. Mecodium Hymenophyllum subg. Mecodium Copel. Mecodium Presl, Epim. Bot. 258. 1849 [1851], nom. nud. 172 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM Hymenophyllum subg. Euhymenophyllum sect. Integra Presl, Hymen,. 32. 1843. Lectotype: Trichomanes asplenioides Swartza=Hymenophyllum as- plenioides (Swartz) Swartz. Presl described essentially the same group of species twice, once under Hymenophyllum as sect. Integra, and once under his genus Sphaerocionium as sect. Glabra. Sphaerocionium sect. Glabra Presl, Hymen. 34. 1843. Lectotype: Hymeno- phyllum caudiculatum Mart. Hymenophyllum subg. Euphorophyllum v. d. Bosch in Junghuhn, Plant. Junghuhn, 563. 1856, nom. nud. Based on Hymenophyllum polyanthos (Swartz) Swartz. Hymenophyllum subg. Hapalophyllum v. d. Bosch, in Junghuhn, Plant. Junghuhn. 566. 1856, nom. nud. Based on Hymenophyllum javanicum Spreng. Hymenophyllum sect. Glabra Prantl, Untersuch. Morph. Gefisskrypt. 1: 54. 1875. Lectotype: Hymenophyllum polyanthos (Swartz) Swartz. Hymenophyllum sect. Globosa Prantl, Untersuch. Morph Gefisskrypt. 1: 55. 1875. Lectotype: Hymenophyllum junghuhnii v. d. Bosch. A group dis- tinguished by having the tip of the receptacle enlarged. Its significance is unknown, but eventually it might turn out to be of sectional importance. Hymenophyllum subg. Euhymenophyllum sect. Mecodium C. Chr. Ind. Fil. XV. 1906, nom. nud. Hymenophyllum subg. Mecodium Copel. Phil. Journ. Sci. 64: 93. 1937. Mecodium (Copel.) Copel. Phil. Journ. Sci. 67: 17. 1938. Hymenophyllum sect. Mecodium (Copel.) Diem & Lichtenstein, Darwiniana 11: 637. 1959. Lecrotype: Hymenophyllum polyanthos (Swartz) Swartz, selected by Copeland, Phil. Journ. Sci. 64: 10. 1937. Later, Copeland (Phil. Journ. Sci. 67: 14. 1938) stated the type to be H. sanguinolentum Forst., which would have been the most logical original choice, since it was the species that Presl mentioned in proposing Mecodium as a nomen nudum, but nevertheless, Copeland’s first choice may not be rejected. Hymenophyllum sect. 10. Mecodium (Copel.) Diem & Lichtenstein. Divisible into the three subsections: Hymenophyllum sect. Mecodium subsect. Mecodium Copeland (Phil. Journ. Sci. 67: 19. 1938) included within Mecodium three species (H. reinwardtti v. d. Bosch, H. thuidium Harrington, and H. samoense Baker) with toothed-margined segments. This destroys the unity of the group and is really unnecessary, since it was done on & priori assumptions regarding the relationships of these species, concerning which nothing is really known. As mentioned previously in discussing the inclusion by Copeland of two entire-margined species in Meringium, the origin of these few aberrant species ought not to be allowed to confuse the classification. It is possible that they represent ancient (or recent) hybrids or are the result of a kind of convergent evolution. Therefore I do not find it unnatural to in- MORTON—HYMENOPHYLLACEAE 173 clude the three species mentioned above in “Meringium.” In this way a reasonable and workable classification can be produced and these aberrant species placed where they fit best so far as our present knowledge goes, leaving the question of their ultimate relationships to be decided later by detailed anatomical, morphological, and cyto- logical evidence. Copeland had no such evidence, only intuition. The subsect. Mecodium is the largest group of the genus, at least so far as accepted names is concerned. It is also the most difficult group, for lacking characters of toothing and pubescence many of the species look quite alike, and many really are perhaps only variations of the common and worldwide H. polyanthos. They include: Hymenophyllum abruptum Hook., H. alfredii Ros., H. andinum v. d. B., H. angulosum Chr., H. apiculatum Mett., H. asplenioides Sw., H. atrovirens Col., H. avillare Sw., H. badium Hook, & Grev., H. balansae Fourn., H. bam- lerianum Rosenst., H. bismarckianum Christ, H. botryoides v. d. Bosch, H. breve Rosenst., H. brevistipes Liebm., H. carnosum Christ, H. caudiculatum Mart., H. constrictum Christ, H. contextum Rosenst., H. contiguum (D. A. Smith) Morton, comb. nov. (Mecodium con- tiguum D. A. Smith, N. Queensl. Nat. 14: 4, f. 1. 1946), H. contor- tum v. d. Bosch, H. copelandii Morton, nom. nov. (Mecodium arch- boldii Copel. Phil. Journ. Sci. 73:458. 1941, non ZH. archboldii (Copel.) Morton, a species of sect. Ptychophyllum), H. corrugatum Christ, H. costaricanum v. d. Bosch, H. crispato-alatum Hayata, H. crispatum Wall., H. cuneatum Kunze, H. daruinit Hook. f., H. dendritis Rosenst., Hi. diversilabium (Copel.) Morton, comb. nov. (Mecodium diver- silabium Copel. Occas. Pap. Bern. Bish. Mus. 14: 49, pl. 2. 1938), H. edanoi (Copel.) Morton, comb. nov. (Mecodium edanoi Copel. Phil. Journ. Sci. 81: 2. 1952), H. emarginatum Swartz, H. endivi- ifolium Desv., H. exsertum Wall., H. farallonense Hieron., H. fecundum vy. d. Bosch, H. fendlerianum Sturm, H. feraxz v. d. Bosch, H. fimbriatum J. Smith, H. flabellatum Labill., H. flerile Makino, H. flecuosum A. Cunn., H. fumarioides Willd., H. funckii v. d. Bosch, H. gardneri v. d. Bosch, H. gollmeri v. d. Bosch, H. hevmu Tardieu (of Madagascar, one of the most peculiar species, with a simple blade with pinnate veins and a deeply pinnatilobate margin), H. helicoideuwm Sod., H. humbertii C. Chr., H. imbricatum Blume, H. inaequale Desv., H. intricatum v. d. Bosch, H. involucratum Copel., H. gavanicum Spreng., H. junghuhnii v. d. Bosch, H. kuhnii C. Chr., H. laciniosum Christ, H. lehmannii Hieron., H. leratit Rosenst., H. levingit C. B. Clarke, H. longifolium van Alderw. van Rosenb., H. macrothecum Fée, H. mexiae (Copel.) Morton, comb. nov. (Mecodiwm mexiae Copel. Univ. Calif. Publ. Bot. 19: 294, pl. 48. 1941), H. micans Christ, H. macro- phyllum Mett., H. microsorum v. d. Bosch, H. mnioides Baker, H. montanum Kirk, H. myriocarpum Hook., H. nigrescens Liebm., H. 174 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM nitiduloides Copel., H. oligosorum Makino, H. ooides Muell. & Bak., H. opacum Copel., H. osmundoides v. d. Bosch, H. paniculiflorum Presl, H. parvulum C. Chr., H. paucicarpum Jenm., H. polyanthos (Swartz) Swartz, H. productoides J. W. Moore, H. productum Kunze, H. protrusum Hook., H. pulcherrimum Col., H. rarum R. Brown, H. recurvum Gaud., H. remotipinna Bonap., H. reniforme Hook., H. retusilobum Hayata, H. rimbachii Sod., H. riukiuense Christ, H. rufescens Kirk, H. salakense Racib., H. sanguinolentum Swartz, H. siliquosum Christ, H. skottsbergit C. Chr., H. streptophyllum Fourn., H. tablaziense Christ, H. tenerum v. d. Bosch, H. todjambuense Kjellb., HT, trewbti Racib., H. trianae Hieron., H. trichomanoides v. d. Bosch, H. undulatum Swartz, H. veronicoides C. Chr., H. walleri Maiden & Betche, H. whitet Goy, and H. wrightii v. d. Bosch. Hymenophyllum sect. Mecodium subscect. Amphipterum (Copel.) Morton, comb. nov. Amphipterum Presl, Epim. Bot. 258. 1849 [1851], nom. nud. Hymenophyllum sect. Amphipterum C. Chr. Ind. Fil. XV. 1906, nom. nud. Based on Amphipterum Presl. Hymenophyllum subg. Amphipterum Copel. Phil. Journ. Sci. 64: 68. 1937. Type: Trichomanes fuscum Blume= Hymenophyllum fuscum (Blume) v. d. Bosch. Amphipterum (Copel.) Copel. Phil. Journ. Sci. 67: 14. 1938. This small group perhaps contains only Hymenophyllum fuseum and fH. ledermannii Brause. The other two species referred here by Copeland, H. geluense Rosenst. and H. laminatum Copel., are probably referable to subg. Hymenophyllum on the basis of their toothed seg- ments. The sect. Amphipterum is hardly worth recognizing, except perhaps for consistency, since it differs only in the presence of accessory wings not in the plane of the frond. Since I assign names to such groups in the subgenera Hymenophyllum and Sphaerocionium, a sim- ilar group can be distinguished here although it is perhaps not really necessary. I do not undertake to place Amphipterum humatoides Copel., which is unknown to me. Hymenophyllum sect. Mecodium subsect. Diplophyllum (v. d. Bosch) Morton, comb. nov. Hymenophyllum subg. Platynophyllum v. d. Bosch, in Junghuhn, Plant. Junghuhn. 569. 1856, nom. nud. Based on Hymenophyllum dilatatum (Forst.) Swartz. Diplophyllum v. d. Bosch, Versl. Meded. Konink, Akad. Wetens. Afd. Natuurk. 11: 322. 1861, non Diplophyllum Lehm., 1814. Lectotype: T'ri- chomanes dilatatum Forst.= Hymenophyllum dilatatum (Forst.) Swartz, chosen by C. Chr. Ind. Fil. XV. 1906 (as “Diploopbyllum”’). Hymenophyllum sect. Diplophyllum (v. d. Bosch) C. Chr. Ind. Fil. XV. 1906. Hymenophyllum dilatatum and the related H. scabrum A. Richard, of New Zealand, are peculiar among the species of Hymenophyllum MORTON—IHYMENOPHYLLACEAE 175 in having the fronds three cells, rather than one cell, thick. This would seem to be a fundamental and important difference, but two other species, H. australe Willd. and H. demissum (Forst.) Swartz, are said to be sometimes partly two cells thick, and these would therefore perhaps go into the subsect. Diplophyllum technically. The matter needs to be investigated further. V. Cardiomanes Cardiomanes Presl, Hymen. 12. 1843. Trichomanes subg. Cardiomanes (Presl) Christ, Farnkr. Erde 33. 1897 (errone- ously attributed to Presl). Trichomanes sect. Eutrichomanes [subsect.] Cardiomanes (Presl) van Alderw. van Rosenb. Malay. Ferns 83. 1908. Type: Trichomanes reniforme Forst.—Cardiomanes reniforme (Forst.) Pres]. This peculiar plant, endemic in New Zealand, has always excited interest by its strange shape for a member of the filmy-fern family. In its indusial structure it is typical of Trichomanes, and the panto- tactic sori are, of course, correlated with the simple blade with forked veins. The only character that might really be distinctive is that the blades are four cells in thickness, and this too doubtless is correlated with the blade shape. A blade of this size would be almost too delicate and flaccid if it were only one cell thick, although this is not necessarily the case, for in the comparable Chilean Hymenoglossum an equally large blade is still only one cell thick except at the very margin. Still, for convenience and because its near affinities are unknown, Cardiomanes can be recognized as a monotypic genus, as has been the tendency by those who otherwise do not favor splitting up the large genus Trichomanes. VI. Trichomanes Trichomanes L. Sp. Pl. 1097. 1753. Typr: There has been some uncertainty as to the correct lecto- type for Trichomanes L. The 11 original species, arranged as was usual in the time of Linnaeus in the order of division of the frond, with the simplest first were: 7. membranaceum, T. crispum, T. polypodioides, T. hirsutum, T. pyxidiferum, T. tunbridgense, T. adi- antoides, T. scandens, T. chinense, T. canariense, and T. capillaceum. Obviously Linnaeus did not really understand Trichomanes because he included several species with somewhat similar involucres but which do not belong to the family Hymenophyllaceae. The first species removed were: adiantoides, an Asplenium; chinense, Adiantum chinense (.) Burm.; * and canariense, Davallia canariensis (L.) J. E. 18 Now Sphenomeris chinensis (L.) Maxon. 176 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM Smith. J. E. Smith also removed tunbridgense as the type of Hymen- ophyllum, and Swartz later transferred hirsutum to Hymenophyllum. The remaining species have all been left in Trichomanes, except by Copeland, and I consider them congeneric. Christensen in the “Index Filicum” (XV. 1906) did not indicate the type of his subg. Eutri- chomanes, but his concept is shown by the species he included, namely only T. pyzidiferum, scandens, and capillaceum. In this he was follow- ing the practice of previous specialists like Presl, van den Bosch, and Prantl. Copeland upset this concept completely in his “Genera Hymeno- phyllacearum” (1938) and “Genera Filicum” (1947) by coining the new name Vandenboschia and calling Trichomanes what had been termed subg. Achomanes or subg. Ptilophyllum, typified by T. crispum L. Because of his prestige Copeland has been followed in his typifica- tion by most subsequent workers without a critical evaluation of his reasoning. However, in this case as in many others, Copeland’s typifications are not to be trusted since they were made on a highly personal basis not in accord with the rules of the International Code of Botanical Nomenclature or with practices currently accepted in Copeland’s own time. Copeland’s reasoning was given in the “Genera Hymenophyllace- arum’’ (pp. 68, 69), as follows: Although the era of our botanic nomenclature began with the publication of Linnaeus’s Species Plantarum in 1753, genera are not defined in that work. For Linnaean generic definitions and, therefore, for the typification of his genera, we must go to earlier works of the same author. 1? In the case of Trichomanes it is defined in the Corollarium Genera Plantarum (and, teste Underwood, in Hortus Cliffortianus, also dated 1737): “Calyx turbinatus, solitarius, erectus, ex ipso margine follii. Stylus setaceus capsulam terminans.” No species is mentioned, but reference is made to Plumier !8 for an illustration. This is Trichomanes crispum. As Linnaeus’ generic concept finds its first expression here, and no other species is mentioned or referred to, this species must be accepted as the type species of the genus. This idea of Copeland’s that to typify Linnaean genera we must go back into “‘pre-Linnaean”’ (i.e., pre-1753) literature in order to interpret the first or original application of a generic name is entirely contrary to the Code and to current practice. It is the same idea that Otto Kuntze had back in 1891, when he adopted the 1737 edition of Linnaeus’ Genera as the starting date, and consequently made his name anathema to most of the botanists of the 1890’s and early 1900’s. In fact even today Kuntze’s application of names is suspect and rightly so. Linnaeus’ generic concepts were much changed and amplified “The latest rule sends us to the next subsequent edition of the ‘Genera,’ but, to interpret the definition found there, we must. still go back to the first edition, where the same definition is amplified by references which fix its type.” 18 “Wil, 86.” MORTON—-HYMENOPHYLLACEAE 177 between 1737 and 1753, and to try to go back and find out his original concept would lead to wholesale changes in names, such as Kuntze actually proposed. For Copeland to try to revive and justify this discredited notion is almost inconceivable. Dr. R. E. Holttum in “Ferns of Malaya,” 87. 1954, criticized Copeland’s choice of type and indicated that it should be reconsidered, but he did not actually make any change himself. The present rules regarding lectotypes and the ‘Guide for the Determination of Types” of the International Code for Botanical Nomenclature (1961 ed.) lead us in exactly the opposite direction, not backward to the maze of pre-Linnaean usages but forward to subsequent usage. The first subsequent author who has definitely or inferentially designated a suitable lectotype must be followed. The pertinent sections of the Guide are sections 4e and 4f: In cases when two or more elements were included in or cited with the original description, the reviewer should use his best judgment in the selection of a lecto- type, but if another author has already segregated one or two elements as other taxa, the residue or part of it should be designated as the lectotype if its essential characters correspond with the original description. If it can be shown that the element best fitting the protologue has been removed, it should be restored and treated as the lectotype. Whenever the original material of a taxon is hetero- geneous, the lectotype should be so selected as to preserve current usage unless another element agrees better with the protologue. The first choice of a lectotype must be followed by subsequent workers unless the original material is redis- covered, or unless it can be shown that the choice was based upon a misinterpreta- tion of the protologue. After the above-mentioned species were removed to other genera (Asplenium, Adiantum, Davallia, and Hymenophyllum) there re- mained in Trichomanes only T. membranaceum, T. crispum, T. poly- podioides, T. pyxidiferum, T. scandens, and Tf. capillaceum. These were all retained unchanged in the genus without any designation of a type until Presl’s ‘“Hymenophyllaceae” (1848). He removed T. membranaceum to a distinct genus, Lecanium, and placed T. crispum in a different subgenus, Trichomanes subg. Achomanes, thus removing it from consideration as a possible lectotype. Trichomanes polypodioides and T. capillacewm were not mentioned by Presl. In his Trichomanes subg. Eutrichomanes Presl retained only T. pyzidiferum and Tf. scandens, thus limiting the choice of lectotype to one of these. The choice was definitely made by John Smith '* by citing 7’. scandens L. as the type of J'richomanes. This is the eminently suitable choice: It agrees with the protologue as well as any other species, it was not misinterpreted by Presl or John Smith, and in fact is one of the well- known West Indian species of the genus, and it maintains the usage of all authors up to the time of Copeland asto the true subg. 19 Hist. Fil. 347. 1875. 178 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM Eutrichomanes, or subg. Trichomanes as it must now be called. The generic name Vandenboschia Copeland thus proves to be a superfluous, illegitimate name for the true Trichomanes L. to be abandoned by those who wish to continue to split up the Hymenophyllaceae into smaller genera. Key to the Subgenera of Trichomanes Venation anadromous; sori paratactic; 2° fronds mostly bipinnatifid to quad- ripinnatifid, rarely only pinnatifid, simple or merely lobed; segments not ciliate (except in 7. scandens) or with stellate marginal hairs; false veins absent (except in sect. Crepidomanes). Rhizomes wide-creeping and fronds distant; stipes often relatively slender; plants epiphytic .... » se. . .Subg. 1. Trichomanes Rhizomes erect or short-creeping, the fronds somewhat clustered; stipes often rather coarse; plants mostly terrestrial. . . . . Subg. 2. Pachychaetum Venation catadromous; sori epitactic; fronds not more than pinnate-pinnatifid, sometimes simple or merely lobed; segments often ciliate or with sessile stellate marginal hairs; false veins present or absent. False veins present between the true veins or in a submarginal position; fronds small, simple to Pipinnatifid but not truly pinnate; plants always epiphytic .... - . . . . . subg. 3. Didymoglossum False veins absent (except i in sect. N europhyllum with dark cross veins perpen- dicular to the true veins); fronds never simple, mostly deeply pinnatifid to once-pinnate; often terrestrial . . . . . . . . .Subg. 4. Achomanes Subg. 1. Trichomanes Stipes distinctly different from the rhizomes; rhachis not proliferous. Blades one cell thick throughout, at least mostly not pilose. Cells not perpendicular to the veins, the distal walls not thickened. Marginal cells not differentiated. Blades bipinnatifid to quadripinnatifid, the ultimate segments with a single, unforked vein. Blades ciliate, the hairs simple or forked at base. False veins absent. West Indies... ..... 2.2.2... .Sect. 1. Trichomanes Blades not ciliate. False veins absent . .......... . Sect. 2. Lacosteopsis False veins present... . . - . . . . Sect. 3. Crepidomanes Blades simply pinnate or pinnate-pinnatifid, the segments broad, with a continuous midrib and forked veinlets. Australia, Tasmania, and New Zealand ..... » ... . . Sect. 4. Phlebiophyllum Marginal cells (one or two rows) slightly differentiated from the others, more elongate. Philippines and Sumatra to Tahiti and New Zealand. Sect. 5. Crepidium Cells of the segments more or less perpendicular to the veins and elongate, the distal walls thickened and giving the appearance of false veins. Philippines . .. » . .. . . Sect. 6. Abrodictyum Blades two or three cells thick near ‘the principal veins. Blades delicate, often glaucous or whitish, sometimes pilose on the surfaces near the veins but 20 See the introduction (p. 158) for a definition of this term. MORTON—HYMENOPHYLLACEAE 179 not ciliate; rhizomes Tong: creeping; very delicate. Ceylon and Philippines to Tahiti. . ... .... . . Sect. 7. Pleuromanes Stipe bases bearing thizoids ‘and “hardly distinguishable from the rhizomes; rhachis often divided once or more times (“rhachis proliferous”). Old World. Sect. 8. Gonocormus Subg. 2. Pachychaetum Cell walls usually thick, pitted or wavy. Fronds bipinnate to tripinnate. Fronds one cell thick . . ... ... . . . Sect. 9. Pachychaetum Fronds ca. three cells thick. American tropics . . . Sect. 10. Davalliopsis Fronds once-pinnate. Old World ...... . . . Sect. 11. Cephalomanes Cell walls thin, straight, not pitted. Stipes and rhachises with stiff, bristle-like hairs. Sumatra to Tahiti and Hawaii. Sect. 12. Callistopteris Stipes, rhachises, and blades with only minute, submicroscopic, clavate hairs. Ryukyu Islands, Philippines, and Java to Polynesia . Sect. 13. Nesopteris Subg. 3. Didymoglossum Fronds simple to bipinnatifid, without marginal ‘“‘scales.” Fronds with dark, stellate hairs along the margins, lacking a differentiated submarginal row of cells ........ . .Sect, 14. Didymoglossum Fronds glabrous on the margin, a submarginal row of cells slightly different from the others sometimes present .... . . . Sect. 15. Microgonium Fronds simple or slightly lobate, with paired, marginal, scalelike projections. American tropics... . . ee + es ee ee es 6 Sect, 16, Lecanium Subg. 4. Achomanes Fronds not ciliate, the hairs of the costae and stipe simple, coarse. Fertile and sterile fronds not dimorphic. Blades simply pinnate, radicant at the apex; sterile pinnae toothed. Dark cross-walls (false veins) perpendicular to the true veins present. American tropics ... . ... - sect. 18. Neurophyllum Dark cross-walls (false veins) absent. “South America. Sect. 19. Odontomanes Blades simply pinnate to tripinnatisect, not radicant at the apex; sterile segments not toothed. False veins absent. Rhizomes long-creeping and fronds distant. American tropics. Sect. 20. Lacostea Rhizomes erect or short-creeping and fronds more or less adjacent. Amer- ican tropics .... . es . . . « Sect, 21. Trigonophyllum Fertile and sterile fronds dimorphic. Rhizomes widely creeping; fronds scattered, distichous. South America. Sect. 22. Homoeotes Rhizomes erect; fronds clustered ... . . 1... ees . . Sect. 23. Feea Fertile fronds dissected, the involucres solitary and free. American tropics... . . . . . Subsect. Feea Fertile fronds entire, ¢ ca. “3 ‘cells thick, ‘the involueres completely immersed. American tropics . ...... . .. . Subsect. Hymenostachis 180 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM Fronds ciliate. Cilia simple. Cilia with a dark bulbous base; blades pinnate-pinnatifid. American tropics... . ... . Sect. 24. Ragatelus Cilia without a markedly bulbous dark base; blades simply-pinnate. American tropics and West Africa .. . ... . . Sect. 17. Achomanes Veins without accessory wings not i in the plane of the lamina. Subsect. Crispa Veins with accessory wings not in the plane of the lamina. Subsect. Lamellata Cilia stellate or forked at base. Blades pinnatifid to pinnate-pinnatifid. Amer- ican tropics. . . ........... =... Sect. 25. Acarpacrium Subg. 1. Trichomanes Trichomanes subg. Trichomanes Pyxidaria Gleditsch, Syst. Plant. 291.1764. Lectotype: Trichomanes scandens L. Copeland (Gen. Hymen. 13) dismisses this name with the statement “This name has no status at all,” which is hardly comprehensible. There is no category of names ‘without status’; names are either valid or in- valid, legitimate or illegitimate. Pyridaria Gleditsch is validly published. It is illegitimate, since it was superfluous, being merely a change of name for Trichomanes L., because Gleditsch took up the name Trichomanes in the “pre-Linnaean”’ sense as used by Tournefort for Asplenium tricho- manes L. It must be remembered that Gleditsch was writing only about 10 years after the publication of Linnaeus’ “Species Plantarum” (1753) and “Genera Plantarum” (ed. 5, 1754) and did not know that botanical nomenclature was later to be accepted as beginning at that time. From his point of view it was quite proper to reestablish Tournefort’s genus Tri- chomanes and to rename Linnaeus’ Trichomanes as Pyzidaria. No species was indicated for Pyxidaria and since the diagnosis is essentially the same as the Linnaean diagnosis for Trichomanes,?! Pyzidaria is to be considered as based on the same type as Trichomanes and to be an absolute nomen- clatural synonym.” Trichomanes subg. Eutrichomanes Presl, Hymen. 16. 1843. Lectotype: Tri- chomanes scandens L. Trichomanes subg. Eutrichomanes [sect.] Pinnata Presl, Hymen. 16. 1843. Lectotype: T'richomanes scandens L. This section originally included many species, but since it included the lectotype species of Trichomanes it should be considered as a nomenclatural synonym of sect. T'richomanes, with the same lectotype. T'richomanes sect. Eutrichomanes (Prest) Moore, Ind. Fil. ex. 1857. Trichomanes [subg.] Minora Prantl, Untersuch. Morph. Gefiasskrypt. 1: 21 Gen. Plant. ed. 5. 1754. * Incidentally, it might be mentioned that Pteris Gleditsch (Syst. Plant. 289. 1764), omitted from the “Index Filicum,” is not the same genus as Pteris L. Gleditsch describes his Pteris as having a dorsal, centrally petiolate, peltate in- dusium, which indicates that he intended it to apply to the genus now called Polystichum (i.e., P. lonchitis). The Linnaean Pteris is renamed Cincinalis Gle- ditsch (Syst. Nat. 290. 1764); however, the species name Cincinalis aquilina Gleditsch does not occur here as it is indicated in the “Index Filicum.” MORTON—HYMENOPHYLLACEAE 181 51. 1875. Lectotype: Trichomanes intramarginale Hook. & Grev. Prantl divided his restricted Trichomanes into two groups, ‘“‘Minora” and ‘ Ma- zora,”’ each with a description. Since these both contained sections so named they must be considered to be subgeneric names, although Prantl may have thought of them as merely descriptive. Both of them refer to subg. Trichomanes as I delimit that subgenus, and so they do not cause any nomen- clatural complications. A lectotype has been arbitrarily selected, the type of the first section included by Prantl within his subg. Minora. Trichomanes [subg.] Maiora Prantl, Untersuch. Morph. Gefasskrypt. 1: 52. 1875. Lectotype: Trichomanes scandens L. See the remarks above under subg. Minora concerning this name. Since this subgenus included the type of Trichomanes, it follows that the lectotype must be the same, T. scan- dens L. Trichomanes subg. Holophlebium Christ, Farnkr. Erde 27. 1897. Lectotype: Trichomanes scandens L. Christ recognized four subgenera, the mono- typic Serpyllopsis and Cardiomanes, Hemiphlebium (with the sections Microgonium, Lecanium, and Euhemiphlebium) characterized by the presence of false veins, and Holophlebium, without false veins, which con- tained all the rest of the genus, including the section Hutrichomanes. Holophlebium was therefore another name for the typical subgenus, with the same type as subg. Trichomanes. As I construe it, the subg. Zrichomanes is divisible into eight sections. Trichomanes sect. 1. Trichomanes LectotyPe: Jrichomanes scandens L. See the discussion above concerning the typification of 7richomanes. This typical section may be monotypic, containing only the com- mon West Indian 7. scandens. This has always been grouped with and considered closely allied to 7. radicans Swartz, but it does differ somewhat in habit, in the ciliate fronds, and also in chromosome number as indicated by T. Walker. In T. scandens the number is n=64, presumably a polyploid on the base number x=17, whereas T. radicans (and related species so far as known) has n=36 or 72 (i.e., with a base r=9). Trichomanes sect. 2. Lacosteopsis Prantl. Trichomanes sect. Lacosteopsis Prantl, Untersuch. Morph. Gefasskrypt. 1: 53. 1875. Lectotype: Trichomanes luschnathianum Presl. There were two original species, 7'. radicans Swartz and T. luschnathianum Presl. In the “Index Fllicum,’’ Christensen selected as type T. rupestre (Raddi) v. d. Bosch, not one of the original species. He intended T. luschnathianum, which he considered a taxonomic synonym. Copeland (Gen. Hymen. 15 1938) wrongly picked the first cited species, which he indicates as “T’richomanes ‘radicans’ = T. rupestre(?).’’ This is a misinterpretation. The 7’. radicans cited by Prantl was the true radicans of Swartz, and not 7’. rupestre, as shown by the specimens cited. Copeland’s typification must be rejected be- cause of the prior typification of Christensen. The species 7. rupestre is rather peculiar and isolated, but the relationship is evidently with T. radi- cans Swartz and not with the group of 7. crispum L. where Copeland placed it. 182 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM Trichomanes sect. Eutrichomanes [subsect.] Lacostea [series] Trichomanopsis van Alderw. van Rosenb. Malay. Ferns 84. 1908. Lectotype: Trichomanes aphlebioides Christ. Although no definite rank was assigned by van Alderwerelt van Rosenburgh to his group Trichomanopsis, by working downward one can consider it to have been a series. Since no rank has been assigned, this rank will stand as indicated here, by Art. 35 of the Code (1961 ed.). Vandenboschia Copel. Phil. Journ. Sci. 67: 51. 19388. Type: Vandenboschia radicans (Swartz) Copel. =Trichomanes radicans Swartz. A superfluous illegitimate name since the type of T’richomanes is cited as a synonym. Trichomanes subg. Vandenboschia (Copel.) Allen, Fl. N. Zeal. 34. 1961.8 This section Lacosteopsis includes T. radicans Swartz and most of the species included by Copeland in his genus Vandenboschia, with the exception of T. scandens L. (which I refer to sect. Trichomanes) and T. pyxidiferum L. (which I refer to sect. Crepidomanes, although with some slight doubt). Among the additional species referable here, mostly following Copeland, are: T. abrotanifolium v. d. Bosch, T. amabile Nakai, T. angustatum Carm., T. aphlebioides Christ, T. auri- culatum Blume, T. axillare Sodiro, 7. beckert Krause, T. birmanicum Bedd., 7. boschianum Sturm, T. brachyblastos Mett., 7. bradei Christ, T. capillaceum L., T. cellulosum K1., T. clathratum Tag., T. cocos Chr., T. colensot Hook. f., T. collariatum v. d. B., T. eyrtotheca Hillebr., 7. cystoceiroides Christ, T. davallioides Gaud., T. debile v. d. Bosch., T. diaphanum H. B. K., T. draytonianum Brack., T. exsectum Kunze, T fallax Christ, T. fargesia Christ, 7. giganteum Bory, T. goetzii Hieron., LT. haughttt Morton, T. herzogit Rosenst., T. hymenophylloides v. d. Bosch, T. hypnoides Christ, T. ingae C. Chr., T. gohnstonense Bailey, T. gunceum Christ, T. latifrons v. d. Bosch, T. latisectum Christ, T. liuki- uense Yabe, T. longifrons Nakai, T. maximum Blume, T. melanotri- chum Schlecht., T. miyaker Yabe, T. naseanuwm Christ, T. nipponicum Nakai, 7. orientale C. Chr., T. parvum Copel., T. philippianum Sturm, T. schmidianum Zenker, T. serratifolium Rosenst., T. sinuatum Bonap., T. somae Nakai, JT. stenosiphon Christ, T. subclathratum (K. Iwatsuki) Morton, comb. nov. (Vandenboschia subclathrata K. Iwat- suki, Acta Phytotax. Geobot. 17: 70.1958), 7. teneruwm Spreng., T. tenuissimum v. d. Bosch, T. titibuense (Ito) Morton, comb. nov. (Vandenboschia titibuensis Ito, Journ. Jap. Bot. 24: 125, f. 1.1949), 7. ulei Christ, T. virgatulum v. d. Bosch, and T. wildu Bailey (?). Trichomanes sect. 3. Crepidomanes (Presl) Prantl. Trichomanes subg. Crepidomanes Presl, Epim. Bot. 17. 1849 [1851]. Type: Trichomanes intramarginale Hook. & Grev., the sole original species. In his “Genera Hymenophyllacearum”’ (p. 14), Copeland correctly had the type 3 Holttum, Ferns Malaya 105. 1954, has Trichomanes subg. Vandenboschia, but since the basionym was not indicated and since it was published after the date requiring the definite citation of basionyms for valid publication, Holttum’s subgeneric name must be considered invalid. MORTON—HYMENOPHYLLACEAE 183 as 1’. intramarginale, from Ceylon, but in the ‘‘Genera Filicum’”’ he arbi- trarily changed it to 7’. brevipes Presl, on the grounds that Presl’s descrip- tion was based largely on specimens of T’. brevipes; this is a mere supposi- tion on his part; Presl’s description may have been based in part on this Philippine species but it was also based in part on the cited true type T. intramarginale Hook. & Grev. Crepidomanes (Presl) Presl, Epim. Bot. 258. 1849 [1851]. A raising, in the Addenda et Corrigenda, of his subg. Crepidomanes to generic rank, with the same single species. Taschneria Presl, Epim. Bot. 258. 1849 [1851]. Type: This is a nomen nudum based on T'richomanes filicula Bory. No subsequent author took up the name Taschneria until Christensen’s Trichomanes subg. Eutrichomanes sect. Taschneria. Trichomanes sect. Crepidomanes (Presl) Prantl, Untersuch, Morph. Gefiss- krypt. 1: 51. 1875. Trichomanes sect. Didymoglossum sensu Prantl, Untersuch. Morph. Gefiiss- krypt. 1: 52. 1875, not Didymoglossum Desv. (1827). Trichomanes sect. Taschneria C. Chr. Ind. Fil. XV. 1906. Type: Based on the description of Trichomanes sect. Didymoglossum sensu Prantl (1875), not Didymoglossum Desv. (1827). The type was wrongly indicated by Christensen as 7’. bipunctatum Poir., because he considered the true type T. filicula Bory as a taxonomic synonym of 7’. bipunctatum Poir. The type is T’. filicula Bory, as Presl had it. Copeland, in his paper on Trichomanes (p. 174) recognized a Group Taschneria, but “group” is not a recognized taxonomic category. Later, Copeland considered Taschneria as a synonym of Crepidomanes, apparently correctly. The small section Crepidomanes contains probably only one New World species, 7. pyxidiferum L., somewhat doubtfully referable to the section, and the following Old World species: T. acutilobum Ching, T. acuto-obtusum Hayata, T. bilabiatum Nees & Blume, T. bilobatum van Alderw. van Rosenb., 7. bipunctatum Poir., T. boninense Koid- zumi, T. brevipes (Presl) Baker, 7. clarenceanum Ballard, T. christit Copel., 7. insigne Beddome, 1. intramarginale Hook. & Grev., T. kurzit Bedd., T. latealatum (v. d. Bosch) Christ, 7. latemarginale D. C. Eaton, 7. majorae Watts, T. makinoi C. Chr., T. megistostomum Copel., T. nanophyllum (Tagawa) Morton, comb. nov. (Crepr- domanes nanophyllum Tagawa, Acta Phytotax. Geobot. 9: 142. 1940), T. nymanii Christ, T. pervenulosum van Alderw. van Rosenb., 1. plicatum (v. d. Bosch) Beddome, T. pseudocapillatum van Alderw. van Rosenb., 7. pseudonymanii (Hosokawa) Morton, comb. nov. (Crepidomanes pseudonymanit Hosokawa, Trans. Nat. Hist. Soc. For- mosa 31: 44. 1941), 7. rothertit van Alderw. van Rosenb., 7. rupicola Racib., 7. tagawanum (K. Iwatsuki) Morton, comb. nov. (Crepi- domanes tagawanum K. Iwatsuki, Acta Phytotax. Geobot. 17: 161, f. 1, 2. 1958), T. tosae Christ, T. venulosum (Rosenst.) Copel., and T. walleri Watts (?). Trichomanes sect. 4. Phlebiophyllum (v. d. Bosch) Prantl. Phlebodium v. d. Bosch, Ned. Kruid. Arch. 4: 377. 1859, non J. Smith, 1841. 284-138—68——3 184 T and CONTRIBUTIONS FROM THE NATIONAL HERBARIUM This name is cited by Copeland * as a validly published name, but it appears at the place cited only in synonymy as Phlebodium brownii v. d. Bosch in sched., as a synonym of T'richomanes venosum R. Brown. Before he published on this plant in 1861 van den Bosch evidently realized that the name Phlebodium was preoccupied and chose to call it Phlebiophyllum, which also turned out to be a homonym, of Phlebophyllum Nees (1832). Phlebiophyllum v. d. Bosch, Versl. Meded. Konik. Akad. Wetens. Afd. Natuurk. 11: 321. 1861. Type: Trichomanes venosum R. Brown, the sole original species. Trichomanes sect. Phlebiophyllum (v. d. Bosch) Prantl, Untersuch. Morph Gefasskrypt. 1: 52. 1875 (erroneously attributed to v. d. Bosch). Polyphlebium Copel. Phil. Journ. Sci. 67: 55. 1938. New name for Phlebio- phyllum v. d. Bosch (1861), non Phlebophyllum Nees (1832). Trichomanes subg. Polyphlebium (Copel.) Allan, Flora New Zeal. 34. 1961. his group was assigned sectional rank by Prantl and Christensen formed Copeland’s “group” Phlebiophyllum,® which was later recognized as a genus Polyphlebium Copel. No doubt T. venosum is a peculiar, isolated species by reason of its broad pinnae with forked veinlets, but it does not seem necessary to recognize it as a genus or even subgenus. It can take its place as a section, as Prantl and Chris- tensen had it. The fact that as a genus Phlebiophyllum is a later homonym does not invalidate the use of this name at the sectional leve l, As a subgenus the proper name would be subg. Polyphlebium (Copel.) Allan. Trichomanes sect. 5. Crepidium (Presl) C. Chr. T Didymoglossum subg. Crepidium Presl, Hymen. 23. 1843. Type: The sole original species is Didymoglossum humile (Forst.) Presl= Trichomanes humile Forst. In his “Epimeliae,’”’ Presl suggested that Crepidium might be regarded as a genus but he did not actually adopt it. As a genus, the name would have been a later homonym of Crepidium Blume (1825), a genus of Orchidaceae. In his treatment of Trichomanes, Copeland treated this as one of his informal “groups,” but in his ‘‘Genera Hymenophyllacearum’’ he raised it to a genus, as Crepidopteris Copel., a later homonym of Crepi- dopteris Sternb. (1838) and was therefore renamed Crepidophyllum C. F. Reed. As a genus Reed’s name is correct, but I do not regard this group of species as sufficiently distinct to rank even as a subgenus. It can take its place as a section of subg. Trichomanes. Trichomanes sect. Crepidium (Presl) C. Chr. Ind. Fil. XV. 1906. Crepidopteris Copel. Phil. Journ. Sci. 67: 57. 1938, non Sternb. 1838. Based on Didymoglossum subg. Crepidium Presl. Crepidophyllum C. F. Reed, Amer. Fern Journ. 38: 88. 1948. Based on Didymoglossum subg. Crepidium Presl. Trichomanes subg. Crepidopteris (Copel.) Allen, Fl. N. Zeal. 35. 1961.26 he only species are probably: 7. apiculare Fourn., 7. endlicheri- anum Presl, T. gracillumum Copel., T. humile Forst. f., T. samoense 44 Phil. Journ. Sci. 67: 14. 1938. 25 Phil. Journ. Sci. 64: 131, 138. 1933. 6 Wolttum, Ferns Malaya 98. 1954, proposed subg. Crepidopteris, but invalidly without definite citation of basionym. MORTON—HYMENOPHYLLACEAE 185 C. Chr., T. vaupelii Brause, T. vieillardiit v. d. Bosch, and 7. werneri Rosenst. Trichomanes sect. 6. Abrodictyum (Presl) Moore. Abrodictyum Presl, Hymen. 20, pl. 7. 1843. Type: Abrodictyum cumingii Presl, the only original species=Trichomanes cumingii (Presl) C. Chr. Trichomanes sect. Abrodictyum (Presl) Moore, Ind. Fil. ex. 1857. Habrodictyon Pres] ex v. d. Bosch, Versl. Meded. Konink. Akad. Wetens. Afd. Natuurk. 11: 321. 1861. This is an orthographic correction that probably ought not to be adopted, since it involves the first syllable of the name, although it is more correct orthographically. The case is similar to the general adoption of the original spelling Eleocharis (Cyperaceae) in preference to the orthographically more correct Heleocharis, or of Aplopap- pus rather than Haplopappus, although the latter has obtained some currency in the United States through its adoption by H. M. Hall in his monograph of the genus. Trichomanes sect. Leptomanes Prantl, Untersuch. Morph. Gefasskrypt. 1: 52. 1875. Type: Five species were listed originally: T. tenerum Spreng., T. schiedeanum C. Muell., 1’. trichotideum Swartz, T’. exsectum Kunze, and T. smithii Hook. Under the last-named species Prantl correctly listed Abrodictyum cumingii Presl as a synonym. Since Presl’s genus Abrodictyum had been reduced to a section of Trichomanes by Moore and since there was no obstacle to its use as a sectional name, the name sect. Leptomanes Prantl was superfluous when published, since the prior and available name sect. Abrodictyum (Presl) Moore should have been adopted. It is therefore to be considered as merely an unacceptable change of name for Abrodictyum and therefore with the same type, Abrodictyum cumingit Presl. Christensen’s choice 2” of 7’. capillaceum L. as type therefore has to be rejected; it was chosen because Christensen considered T. trichoideum Swartz as a taxonomic synonym of T. capillaceum L. Similarly, Copeland’s choice % of 7’. tenerum on the basis that it was the first species listed by Prantl must be rejected also. Trichomanes sect. Eutrichomanes [subsect.] Abrodictyum (Presl) van Alderw. van Rosenb. Malay. Ferns 83. 1908. Trichomanes sect. 7. Pleuromanes Presl. Trichomanes sect. Pleuromanes Presl, Epim. Bot. 17. 1849 [1851]. Type: Trichomanes acutum Presl, the sole original species. Pleuromanes (Presl) Presl, Epim. Bot. 258. 1849 [1851]. A raising in the Addenda et Corrigenda of his sect. Pleuromanes to generic rank with the same species 7’. acutum and the addition of a second species, 7’. pallidum Blume. In the Index Filicum (XV. 1906), Christensen erroneously gave the type as 7. pallidum.?® Leucomanes Presl, Epim. Bot. 258. 1849 [1851]. A nomen nudum, based on Trichomanes album Blume; the name Lewcomanes has never been used by any author either as a genus or section, because it represents the same group as Presl’s Pleuromanes. 27 Ind. Fil. XV. 1906. *% Phil. Journ. Sci. 67: 15. 1908. ® Holttum (Ferns Malaya 97. 1954) proposed Trichomanes subg. Pleuromanes, but invalidly without specific citation of basionym. 186 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM Craspedoneuron v. d. Bosch, Versl. Meded. Konink. Akad. Wetens. Afd. Natuurk. 11: 322. 1861. Lectotype: Trichomanes album Blume. The name Craspedoneuron first appeared in van den Bosch’s treatment of the Hymenophyllaceae in Junghuhn’s ‘Plantae Junghuhnianae”’ (p. 550. 1856), where it appeared as an accepted name, apparently as a subgenus, but without a description. It is indicated as containing “7. album etc.,”’ which fixes the type as album. The name appeared later in van den Bosch’s Synopsis *! as a subgroup of Trichomanes but again without a description or stated rank. Copeland ®? also indicated the type as Trichomanes album Blume. Trichomanes sect. Craspedoneuron (v. d. Bosch) Prantl, Untersuch. Morph. Gefisskrypt. 1: 52. 1875. Although recognized by Copeland as a distinct genus, the group of Trichomanes acutum and T. pallidum does not appear to be very different from other groups of subg. Trichomanes. This section is distinguished by having the fronds more than one cell wide along the veins; although this character is well developed in J. pallidum, it is less evident in the type species 7. acutum and seems to be absent in T. latifrons, which Copeland associated with the other two in his 1933 paper on Trichomanes. I recognize Pleuromanes as a section only with hesitation, awaiting further anatomical study. It will be limited to T. acutum and T. pallidum, the latter including T. album Blume, tentatively, and TZ. retusum (Copel.) Morton, comb. nov. (Pleuromanes retusum Copel. Phil. Journ. Sci. 73: 466. 1941). The whitish or bluish color of the fronds may be distinctive. Trichomanes sect. 8. Gonocormus (v. d. Bosch) Christ. Gonocormus v. d. Bosch, Vers]. Meded. Konink. Akad. Wetens. Afd. Natuurk. 11: 321. 1861.3 Lectotype: T'richomanes proliferum Blume (selected by Christensen). Van den Bosch refers here species 1, 2, 9, and 10 of his “Synopsis,” namely 7. parvulum Poir., T. saxifragoides Presl, T’. proliferum Blume, and 7. minutum Blume, which W. A. Sledge (pers. comm.) con- siders a synonym of 7’. proliferum. Prantl (1875) accepted Gonocormus as a genus and listed four species in sect. Gonocormus, of which two are G. prolifer and G. minutus, thus limiting the choice of lectotype to one of these two species. In the Index Filicum (XIV. 1906) Christensen definitely selected 7’. proliferum as the type; therefore Copeland’s selection (Gen. Hymen. 14. 1938) of 7. minutum as type must be rejected. Copeland selected the first species listed by van den Bosch in his “‘Hymenophyllaceae Javanicae,” G. minutus, but the first-named species is not automatically the type of course. This does not change the application of the name, since T. minutum and T. proliferum are not different specifically. %0 Also published in Verhandl. K. Akad. Wetens. Amsterdam 9, Hym. Jav. 12. 1861, without description. 3! Ned. Kruid. Arch. 4: 361. 1859. 32 Phil. Journ. Sci. 67: 14. 1938. 8 Also published Verhandl. K. Akad. Wetens. Amsterdam 9, Hym. Jav. 7. 1861, without description. MORTON—HYMENOPHYLLACEAE 187 Gonocormus sect. Gonocormus Prantl, Untersuch. Morph. Gefasskrypt. 1:51. 1875. Trichomanes sect. Gonocormus (v. d. Bosch) Christ, Farnkr. Erde 27. 1897 (erroneously attributed to v. d. Bosch). Trichomanes subg. Gonocormus (v. d. Bosch) C. Chr, Ind. Fil. XIV, 634. 1906. Trichomanes sect. Eutrichomanes [subsect.] Gonocormus van Alderw. van Rosenb. Malay. Ferns 83. 1908. This somewhat peculiar group, best distinguished by its ‘‘prolif- erous’’ fronds, can rank best merely as a section of subg. Z’richomanes. The following are probably referable here: 7. alagense Christ, T. assimile Mett., T. bonapartet C. Chr., T. boninicola Nakai, T. brooksi Copel., 7. gracile v. d. Bosch, 7’. latilabiatum E. Brown, T. mannii Hook., 7. matthewii Christ, T. novoguineense Brause (?), 7’. ruwen- zoriense Taton, T. subtilissimum Brause, T. teysmannii v. d. Bosch, and T’. trinerve Baker. Subg. 2. Pachychaetum Trichomanes subg. 2. Pachychaetum Presl Trichomanes subg. Pachychaetum Presl, Hymen. 16. 1843. Lectotype: Six species were originally assigned to this subgenus: 7. luschnathianum Presl, T. rigidum Swartz, 7’. firmulum Presi, T. pyramidale Wall. T. speciosum Willd., and 1. brevisetum Spreng. Since these species do not really form a natural group, a type could be chosen only arbitrarily. In proposing Trichomanes subg. Eutrichomanes sect. Pachychaetum (Presl) C. Chr., Christensen *4 designated 7’. rigidum Swartz as the type, and this therefore must stand as the lectotype. Copeland’s choice of T’. luschna- thianum Presl as lectotype (Gen. Hymen. 14. 1938) merely on the basis that this was the first-cited species cannot be upheld. Trichomanes subg. Macroglena Presl, Abhandl. Boehm. Gesell. Wiss. V, 5: 333. 1848.35 Type: Five species were originally named: T. gemmatum J. Smith, 7. cellulosum Klotzsch, T. foeniculaceum Bory ex Willd., T. angustissimum Presl, T. meifolium Bory ex Willd. T'richomanes meifolium was selected as lectotype by Christensen #6 and there is no reason to make any change. The group was distinguished by Copeland as a genus distinct from Selenodesmium by the “bristle-like segements,” i-e., with the laminae only 1 or 2 cells wide on either side of the midveins, which is true of the 34 Ind. Fil. XV. 1906. 35 As noted elsewhere in this article, the groups in Presl’s work denoted by the symbol § are not sections, as this sign usually denotes, but are subgenera, which Presl calls them several times. He also calls them ‘‘paragraphs’’ on occasion but this is not a recognized botanical category. Presl used the term “‘sectio’’ (or rather misused it, according to our present concepts) for a suprageneric group, as for instance in the ‘“‘Epimeliae Botanicae,” p. 103, where the family Blechnaceae is divided into sectio Blechneae and sect. Blechnopsideae, each with several genera. Therefore, the combination Trichomanes subg. Macroglena (Presl) Allan, Flora New Zeal. 36. 1961, is superfluous, since it is the same as the original publication of Presl. 36 Ind. Fil. XV. 1906. 188 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM type species, T’. meifolium, and some others, such as 7’. asa-grayi; some species referred to Macroglena by Copeland, as for instance the New Zealand 1’, strictum Menzies and T'. caudatum Brack., definitely do not have setaceous segments and approach closely 7’. rigidum Swartz, the type species of Selenodesmium. Trichomanes sect. Selenodesmium Prantl, Untersuch. Morph. Gefasskrypt. 1; 53. 1875.37 Type: There were three original species, 7’. rigidum Swartz, I’. obscurum Blume, and 7’. elongatum Cunningham. By placing sect. Selenodesmium as a synonym of sect. Pachychaetum, Christensen * effectively chose 7’, rigidum Swartz as lectotype, also the lectotype of sect. Pachy- chaetum. Christensen was wrong in adopting sect. Pachychactum as the correct sectional name under the subgenerie name Eutrichomanes, inasmuch as Selenodesmium had priority as a sectional name, Pachychaetum having been originally proposed as a subgeneric name. However, this has no bear- ing on the effectiveness of the choice of a lectotype. When, as here, Pachy- chaetum is recognized as a subgenus distinet from Trichomanes [‘‘Hutri- chomanes’’] then the proper sectional name must be Pachychaetum, repeat- ing the subgeneric name according to the Code. Macroglena (Presl) Copel. Phil. Journ. Sci. 67: 82. 1938. Selenodesmium (Prantl) Copel. Phil. Journ. Sci. 67: 80. 1938. The subgenus Pachychactum is divisible into five sections, as follows, Trichomanes sect. 9. Pachychaetum. Among the species referable to sect. Pachychaetum are: T. angusti- marginatum Bonap., T. asa-grayt v. d. Bosch, T. batrachoglossum Copel., T. caudatum Brack., T. compactum van Alderw. van Rosenb., T. cupressoides Desv., T. dentatum v.d. Bosch, T. elongatum A. Cunn., T. extravagans Copel., T. ferrugineum Fourn., T. flavofuscum v. d. Bosch, L’. gemmatum J. Smith, T. laetum v. d. Bosch, T. longicollum v. d. Bosch, T. meifolium Bory ex Willd., T. obscurum Blume, 7. obtusum (Copel.) Morton, comb. nov. (Macroglena obtusa Copel. Phil. Journ. Sci. 84: 163. 1955), T. parviflorum Poir., T. schlechteri Brause, T. schultzet Brause, T. setaceum v. d. Bosch, J. siamense Christ, 7. strictum Menzies, T. stylosum Poir., T. tamarisciforme Jacq., T. tereti- cawlum Ching, T. trichophyllum Moore, T. truncatum (Copel.) Morton, comb. nov. (Macroglena truncata Copel. Occas. Pap. Bern. Bishop Mus. 14: 51, pl. 4. 1938), and 7. warburgit Christ (?). Trichomanes sect. 10. Davalliopsis (v. d. Bosch) Prantl. Davalliopsis v. d. Bosch, Versl. Meded. Konink. Akad. Wetens. Afd. Natuurk. 11: 323. 1861. Type: Copeland (Gen. Hymen. 82. 1938) commented that “Van den Bosch established this genus imperfectly, publishing no specific name under it; but his name fortunately was retained for subgeneric or section use by Prantl and Christensen.” He evidently believed that it was necessary to publish a specific name for a genus to be validly described, but there is no such provision in the nomenclature rules. It was not necessary for any 37 Holttum, Ferns Malaya 108, 1954, proposed Trichomanes subg. Selenodes- mium, but invalidly without definite citation of the basionym. 38 Ind. Fil. XV. 1906. MORTON—HYMENOPHYLLACEAE 189 species to be named prior to the present requirement that types be indicated after Jan. 1, 1958. Although he did not make any combinations under the name Davalliopsis, van den Bosch did name two species as belonging to his genus: species 86 and 87 of his “Synopsis,” 7. prieur?i Kunze and T. anceps Hook. (p.p.). In reducing this genus to a section of Trichomanes, Prantl listed only T. prieurti, thus effectively choosing this species as lectotype. Christensen (Ind. Fil. XV. 1906) wrongly stated the type to be T. elegans L. C. Richard, which is considered to be the same as 7. prieurit. Trichomanes sect. Davalliopsis (v. d. Bosch) Prantl, Untersuch. Morph. Gefas- skrypt. 53. 1875. Trichomanes sect. 11. Cephalomanes (Presl) Morton, comb. nov. Cephalomanes Presl, Hymen. 17, pl. V. 1843. The type and sole original species is Cephalomanes atrovirens Presl=T'richomanes atrovirens (Presl) Kunze. The type stated by Christensen in Ind. Fil. XIV. 1906, T'r:chomanes javanicum, is impossible, since it was not mentioned by Presl in the original description. In 1848 Presl enlarged the genus to include five species, of which 1’. javanicum was one;** it was further enlarged to nine species by van den Bosch.!® Prantl in 1875 made Cephalomanes a section of Lacostea v. d. Bosch (1861), which is nomenclaturally impossible, since Cephalomanes has priority. Cephalomanes may tentatively be regarded as a section, Trichomanes sect. Cephalomanes (Presl) Morton. The character by which Presl separated it, the globose-inflated tip of the receptacle, was either an illusion or an abnormality. Copeland in his key in “Genera Hymeno- phyllacearum” separates Cephalomanes from Trichomanes (as typified by T. crispum) merely by the former being ‘Oriental’ and the latter “American,” hardly a generic distinction, and Copeland’s discussion does not really bring out any characters. However, Cephalomanes may be distinguished by having the venation anadromous and the sori paratact, whereas T'richomanes subg. Achomanes (as typified by 7’. crispum) has the venation catadromous and the sori epitact. However, I have not examined all species with regard to these characters, which may not hold. Trichomanes [subg.] Schizophlebium v. d. Bosch in Junghuhn, Plant. Junghuhn. 551. 1856, nom. nud. Based on T'richomanes javanicum Blume. In his first paper on Hymenophyllaceae, in ‘‘Plantae Junghuhnianae”’ (1856), van den Bosch adopted several of Presl’s subgenera, such as Crepidomanes, and indicated them with the same sign as Presl used; therefore I consider that the new names proposed here by van den Bosch were also subgenera. Since no rank has been assigned by any other author, my decision will stand under the Code. All of van den Bosch’s own names were nomina, nuda, which he later ignored, and so the matter is of little importance in that respect. But if van den Bosch’s groups were to be considered sections, then the various Pres] names adopted would all be new combinations as sections and this would cause some changes in the sectional nomenclature. Therefore, it seems best to consider all of these groups as subgenera. Lacostea v. d. Bosch sect. Cephalomanes (Presl) Prantl, Untersuch. Morph. Gefasskrypt. 1: 50. 1875. Trichomanes sect. Eutrichomanes [subsect.] Lacostea [series] Cephalomanes (Presl) van Alderw. van Rosenb. Malay. Ferns 84. 1908. Trichomanes subg. Cephalomanes (Presl) C. Chr. Ind. XIV. 1906." 39 Abhandl. Bochm. Gesell. V, 5: 334. 1848. 40 Ned. Kruid. Arch. 4: 350-352. 1859. ‘t Also proposed by Bonap. Notes Pterid. 8: 26. 1919. 190 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM Probably the only species of sect. Cephalomanes, all Old World, are: T’. acranthum Ito, T. acrosorum Copel., T. atrovirens (Presl) Kunze, I’. boryanum Kunze, T. crassum Copel., T. densinervium Copel, T. foerstert. Rosenst., T. infundibulare van Alderw. van Rosenb., T. javanicum Blume, T. kingit Copel., T. ledermannii Brause, 7. mada- gascariense Moore, T. maluense Brause, T. preslii Morton, nom. nov. (Trichomanes asplenioides Presl, Hymen. 37. 1843, non Swartz, 1788), LT’. singaporianum (v. d. Bosch) van Alderw. van Rosenb., 7. suffrutex van Alderw. van Rosenb., and 7. swmatranum van Alderw. van Rosenb. Trichomanes sect. 12. Callistopteris (Copel.) Morton, comb. nov. Callistopteris Copel. Phil. Journ. Sci. 67: 64. 1938. Type: Trichomanes apiifolium Presl, chosen by Copeland. The only known species, all Old World, are: Trichomanes api- vfolium Presl, T. bawerianum Endlicher, T. baldwinit D. C. Eaton, TL. calyculatum (Copel.) Morton, comb. nov. (Callistopteris calyculata Copel. Occas. Pap. Bern. Bishop Mus. 14: 50, pl. 3. 1938), and T. societense J. W. Moore (T. polyanthum Hook., non T. polyanthos Swartz). Trichomanes sect. 13. Nesopteris (Copel.) Morton, comb. nov. Nesopteris Copel. Phil. Journ. Sci. 67: 65. 1938. Type: Trichomanes grande Copel., chosen by Copeland.” This and the preceding section are similar and perhaps should be united. However, for the present they should be left as separate sections, since they may not be truly allied. The only species, all Old World, are: 7. blepharistomum Copel., T. grande Copel., T. intermedium v. d. Bosch, T. pseudoblepharistomum Tagawa, 7. superbum Backhouse, and 7’. thysanostomum Makino. Subg. 3. Didymoglossum Trichomanes subg. 3. Didymoglossum (Desv.) C. Chr. Didymoglossum Desv. Mém. Soe. Linn. Paris 6: 330. 1827. Lectotype: Eight species were originally included: D. muscoides, pusillum, punctatum, dect- piens, alatum, lineare, filicula, and magellanicum (lineare and magella- nicum are referable to Hymenophyllum rather than Trichomanes). The first subsequent author to recognize Didymoglossum was Presl (1843), who divided the tribe Trichomanoideae into two groups, Trichomancae and Didymoglosseae, distinguished by having the limb of the involucre entire or 2-lipped, respectively. These groups are termed “‘sectio” by Presl, and so they must be considered invalid, since the category “sectio” was mis- used for a suprageneric rather than infrageneric group as required by the Code. Presl divided Didymoglossum into three subgenera—Eudidymo- glossum, Chilodium, and Crepidium; in Eudidymoglossum he included D. muscoides, punctatum, alatum, decipiens, and filicula, made D. pusillum the type of a new genus, Hemiphlebium, and removed lineare and magella- “” Holttum, Ferns Malaya 108, 1954, proposed Trichomanes subg. Nesopteris, but invalidly, without the definite citation of the basionym. MORTON—HYMENOPHYLLACEAE 191 nicum from the group Didymoglosseae entirely. In his “FEpimeliae Botanicae”’ (1851), Presl indicated that the species filicula should be removed from Didymoglossum as the genus Taschneria, thus effectively limiting the choice of lectotype to the four species remaining in the genus, D. muscozdes, punctatum, alatum, and decipiens. Unfortunately, van den Bosch, followed by Prantl, limited Didymoglossum to D. filicula (now properly known as Trichomanes bipunctatum Poir.) and its allies, thus seeming to provide a lectotype. However, such a typification would upset current nomenclature and it can be rejected on the basis that Presl had excluded filicula from the genus, even though he failed to provide his segregate genus Taschneria with a validating description; in 1851 the publication of the combination Taschneria filicula could have been taken as sufficient for a monotypic new genus. No further treatment of Didymoglossum was published until Christensen * proposed Trichomanes subg. Didymoglossum with the type T. hymenoides Hedw., which Desvaux had listed as a synonym of his D. muscoides. This effectively chooses 7’. muscoides as lectotype, which fortunately does maintain the current application of the name. Didymoglossum subg. Eudidymoglossum (Desv.) Presl, Hymen. 23. 1843. Didymoglossum subg. Eudidymoglossum [sect.] Flabellata Presl, Hymen. 23. 1843. Lectotype: Didymoglossum sphenoides (Kunze) Presl=T'. sphenoides Kunze. Didymoglossum subg. Eudidymoglossum [sect.] Pinnata Presl, Hymen. 23. 1843. Lectotype: Didymoglossum muscoides (Swartz) Desv.=T7'. muscozdes Swartz. Trichomanes sect. Didymoglossum (Desv.) Moore, Ind. Fil. CX. 1857. Trichomanes subg. Didymoglossum (Desv.) C. Chr. Ind. Fil. XIV. 1906. As I construe it, subg. Didymoglossum contains three sections, as follows. Trichomanes sect. 14. Didymoglossum The sect. Didymoglossum, the American species recently mono- eraphed by Wessels Boer, contains the following American species: Trichomanes angustifrons (Fée) W. Boer, T. curtit Rosenst., T. gourl- ianum Grev., T. hymenoides Hedw., T. krausit Hook. & Grev., T. lineolatum (v. d. Bosch) Hook., T. melanopus Baker, T. nummularium (v. d. Bosch) C. Chr., 7. ovale (Fourn.) W. Boer, 7. petersit A. Gray, T. pinnatinervium Jenm., T. punctatum Poir., T. pusillum Swartz, T. reptans Swartz, and T. rhipidophyllum Slosson. In addition, a very few Old World species are presumed to belong here, such as T. exiguum (Beddome) Baker, J. giesenhageniit C. Chr., T. liberiense Copel., and 7. wallit Thwaites. Trichomanes sect. 15. Microgonium (Presl) Christ. Microgonium Presl, Hymen. 19, pl. 6, f. A, B. 1843. Type: Two species were originally assigned to the genus, M. cuspidatum (Willd.) Pres] and M. berteroanum Presl. The first of these, Trichomanes cuspidatum Willd., was chosen lectotype by Christensen (Ind. Fil. XIV. 1906). Hemiphlebium sect. Microgonium (Presl) Prantl, Untersuch. Morph. Gefas- skrypt. 1: 46. 1875. 48 Ind. Fil. XIV. 1906. 192 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM Trichomanes sect. Microgonium (Presl) Christ, Farnkr. Erde 24. 1897 (erro- neously attributed to Presl) .44 The four American species of sect. Microgonium were treated by Wessels Boer, namely: Trichomanes ekmanii W. Boer, T. godmanii Hook., 7. hookeri Presl, and 7. kapplerianum Sturm. The Old World species are more numerous and include: Jrichomanes ballardianum Alston, 7. bimarginatum v. d. Bosch, 7. chamaedrys Taton, 7. craspe- doneurum Copel., T. cultratum Baker, T. cuspidatum Willd., 7. erosum Willd., 7. falsivenulosum (Nishida) Morton, comb. nov. (Micro- gonium falsivenulosum Nishida, Journ. Jap. Bot. 32: 156. 1957), 7. fulgens C. Chr., T. henzaianum Parish, T. mindorense Christ, 7. motleyt v. d. Bosch, T. omphalodes (Vieillard) C. Chr., 7. papuanum Brause, 7. sayert F. Muell. & Baker, and 7. sublimbatum K. Muell. Trichomanes sect. 16. Lecanium (Presl) Christ. Lecanium Presl, Hymen. 11, pl. 1. 1843, non Reinw. 1825. Type: The only original species Lecanium membranaceum (L.) Presl= Trichomanes mem- branaceum IL. Copeland stated that Lecanium Reinw. was not validly published because Reinwardt proposed both Lecanium and Lecanopteris as substitute names for his illegitimate genus Onychium Reinw. (1824) non Onychium Kaulf. (1820). However, according to the Code,‘® such alternative names are validly published if proposed prior to Jan. 1, 1953. The implication is that the genus Lecanopteris Reinw. was also validly published in 1825 by Reinwardt and not by Blume in 1828, as Copeland would have it. The species 7. membranaceum L. is certainly peculiar but I do not think it represents a distinct genus, and therefore Lecanium Pres] does not need to have a new generic name provided. It can con- veniently rank as a section of subg. Didymoglossum, differing especially only in the strange marginal “scales.” The fact that Lecanium is illegiti- mate as a genus does not mean that the name cannot properly be used at the sectional level. Hemiphlebium sect. Lecanium (Presl) Prantl, Untersuch. Morph. Gefas- skrypt. 1: 46. 1875. Trichomanes sect. Lecanium (Presl) Christ, Farnkr. Erde 25. 1897 (erro- neously attributed to Prantl). Subg. 4. Achomanes Trichomanes subg. Achomanes Pres] Trichomanes subg. Achomanes Presl, Hymen. 15. 1843. Lectotype: Presl adopted the name Achomanes from Necker (1790), whose “genus” Acho- manes is cited by Presl as a synonym of Trichomanes on p. 13. However, since Necker’s names are all invalid by the Code, Presl must be considered as the sole author of the subgenus, and Necker should not be cited as a parenthetical author, as has been done. The typification must be from Pres] and not from Necker. The three groups into which Presl divided “* Holttum, Ferns Malaya 91. 1954, proposed Trichomanes subg. Microgonium, but invalidly without the definite citation of the basionym. 45 Acta Bot. Neerl. 11: 277-330. 1962. 401961 ed., Art. 34. MORTON—HYMENOPHYLLACEAE 193 Trichomanes are preceded by the sign §, which usually denotes a section. However, on p. 31, where Presl is speaking of the subdivision of Hymeno- phyllum into four groups, he terms these groups subgenera. These groups are preceded by the same sign as the groups in Trichomanes, and there- fore must also be considered subgenera and not sections. The instance noted above in Hymenophyllum is not the only place where such groups are called subgenera by Presl. As noted elsewhere in this paper, Presl used the term “sectio” for a suprageneric rather than infrageneric group. The matter is of great importance when deciding on the priority of sub- generic and sectional names in the Hymenophyllaceae. Presl included 16 species in his subg. Achomanes, among them T. crispum L. When Presl removed 7. crispum from Eutrichomanes, he effectively eliminated it from consideration as a possible lectotype for Trichomanes itself, that is for subg. Trichomanes. As a section, the name Achomanes dates from van den Bosch (1859), where the group is preceded by the sectional sign and also referred to in a footnote as a section. By following the name Achomanes with “TT. crispum, ete.” van den Bosch inferentially chose T. crispum as lectotype, a choice confirmed later by Christensen in citing T. crispum as the type of Trichomanes subg. Achomanes." Trichomanes sect. Achomanes (Presl) v.d. Bosch, Ned. Kruid. Arch. 4: 358. 1859. Ptilophyllum sect. Achomanes (Presl) Prantl, Untersuch. Morph. Gefisskrypt. 1: 47. 1875. Trichomanes sect. Euachomanes C. Chr. Ind. Fil. XIV. 1906. To those who wish to continue to segregate many genera in the Hymenophyllaceae, the question of the proper generic name for this group will arise at once. It was called Trichomanes L. by Copeland, but as I have shown above it was due to an incorrect and unacceptable typification on J. crispum L. It was called Ptilophyllum by Prantl, but this name is illegitimate and unavailable. Neither is Achomanes Necker available as a generic name, since Necker’s names are speci- fically outlawed by the Code. The name to be adopted will depend on how finely workers wish to split up the group. If they accept the cir- cumscription of Copeland, the earliest available name is Ragatelus Presl (1843). However, if they regard the species with dimorphic fronds as coming within the group, then Feea Bory (1824) will be the oldest name for the whole group. Other generic names that I include within subg. Achomanes are Hymenostachis Bory (1824), Odontomanes Presl (1851), Newromanes Trev. (1859), and Lacostea yv.d. Bosch. Trichomanes subg. Achomanes sect. 17. Achomanes The section Achomanes may be divided into two subsections, as follows. Trichomanes sect. Achomanes subsect. Crispa (Prantl) Morton, comb. nov. Ptilophyllum sect. Achomanes [subsect.] Crispa Prantl, Untersuch. Morph. Gefisskrypt. 1: 47. 1875. LecroryPe: Trichomanes crispum L, 47 Ind. Fil. XIV. 1906. 194. CONTRIBUTIONS FROM THE NATIONAL HERBARIUM The species of sect. Achomanes, subsect. Crispa, are numerous but little known. This would be a fine group for monographic study for the plants are common and widely distributed and need to have reliable names. At the present time there are the following nominal species, but the actual number is probably much smaller: 7. accedens Presl, T. adscendens Kunze, T. anadromum Rosenst. (?), T. aureovestitum Proctor, T. badium Fourn., T. corcovadense v. d. Bosch, T. crassipilis Weatherby, 7. crispum L., T. eristatum Kaulf., T. dactylites Sodiro, T. daguense Weatherby, 7. delicatum v. d. Bosch, J. ertophorum v. d. Bosch, 7. fureatum v. d. Bosch, T. galeottii Fourn., T. gardnert v. d. Bosch, T. guianense Sturm, T. holopterum Kunze, T. imbricatum Sodiro, T. kalbreyert Baker, T. killipii Weatherby, T. lambertianum Hook., 7. lazwm Klotzsch, T. lindigit F ourn., T. lucens Swartz, L’. ludovicianum Rosenst., T. macilentum v. d. Bosch, T. micayense Hieron., T. opacum v. d. Bosch, 7. ornatulum v. d. Bosch, T. pellucens Kunze, T. procerum Fée, T. robustum Fourn., T. sellowianum Presl, and T. sublabiatum v. d. Bosch. The above are all tropical American. One species is in western tropical Africa, T. crispiforme Alston. Trichomanes sect. Achomanes subsect. Lamellata (Prantl) Morton, comb. nov. Ptilophyllum sect. Achomanes [subsect.] Lamellata Prantl, Untersuch. Morph. Gefaisskrypt. 1: 48. 1875. Type: Trichomanes martiusii Presl, the only original species (presumed to be a taxonomic synonym of 7', pilosum Raddi). Inasmuch as the presence of accessory wings on the veins not in the plane of the frond is recognized as of subsectional importance in Hymenophyllum subg. Sphaerocionium, it seems only logical to recog- nize a similar group here in Trichomanes, as Prantl did. The only known species are Trichomanes pilosum Raddi (probably an aggregate) and 7. anomalum Maxon & Morton. Trichomanes sect. 18. Neurophyllum (Presl) Moore. Neurophyllum Presl, Hymen. 18, pl. 4, f.C. 1843, non Torr. & Gray (1840). Type: There were three original species: Neurophyllum vittaria (DC.) Presl, N. pinnatum (Hedw.) Presl, and N. pennatum (Kaulf.) Presl. Since Neurophyllum Presl was an illegitimate later homonym, van den Bosch in 1859 replaced it with Neuromanes Trevisan, which is thus to be considered as a renaming rather than a new genus and to be based on the same type. The original Neuromanes Trevisan #3 was based on different material and was not intended as a new name for Neurophyllum Presl, but it was a mere nomen nudum and therefore not validly published and need not be con- sidered further. Christensen 4? chose 7. pinnatum Hedw. as the type of his sect. Neuromanes, and placed sect. Neurophyllum as a synonym, and this effectively chose 7. pinnatum as lectotype. Copeland’s choice of T. vittaria DC. as lectotype must therefore be rejected; he evidently chose this Atti Ist. Veneto II, 2: 163. 1851. 4 Ind. Fil. XIV. 1906. MORTON—HYMENOPHYLLACEAE 195 species merely because it was the first one listed by Presl. This small group of species can be recognized as a distinct section. The fact that Neuro- phyllum is illegitimate as a generic name does not mean that it may not correctly be used as a sectional name, and since it has priority as a sectional name over Neuromanes it must be adopted. Trichomanes sect. Neurophyllum (Presl) Moore, Ind. Fil. cx. 1857. Neuromanes Trev. ex v. d. Bosch, Ned. Kruid. Arch. 4: 347. 1859. Based on Neurophyllum Presl (1843), not Torr. & Gray (1840). Trichomanes sect. Newromanes (Trev. ex v. d. Bosch) C. Chr. Ind. Fil. XIV. 1906. (Superfluous and illegitimate). This small section contains only Trichomanes pinnatum Hedw., N. pennatum Kaulf., if that is distinguishable, and the simple-bladed 7. vittaria DC. It is distinguished by the presence of dark false veins, per- pendicular to the true veins, Trichomanes sect. 19. Odontomanes (Presl) C. Chr. Odontomanes Presl, Epim. Bot. 20. 1849 [1851]. Type: The only original species was Odontomanes hostmannianum (Klotzsch) Presl= Trichomanes hostmannianum (Klotzsch) Kunze. Neuromanes Trevisan, Atti Ist. Veneto IT, 2: 163. 1851, nom. nud. Trichomanes sect. Odontomanes (Presl) C. Chr. Ind. Fil. XIV. 1906. This very small group is perhaps too close to sect. Neurophyllum, differing chiefly in the absence of false veins. In addition to T. host- mannianum it contains perhaps only T. huberi Christ and T. laciniosum Alston. Trichomanes sect. 20. Lacostea (v. d. Bosch) Christ. Lacostea v. d. Bosch, Versl. Meded. Konink. Akad. Wetens. Afd. Natuurk. 11: 320. 1861. Type: Van den Bosch based his genus Lacostea on species 81 and 82 of his “Synopsis Hymenophyllacearum,” namely T. ankersii Parker ex Hook. & Grev. and T. brachypus Kunze. Prantl (1875) recog- nized this group as a genus and included both of these species. Christensen in his ‘Index Filicum”’ gave two treatments of the group. On p. 634 in the body of the text (this part published July 2, 1906) he followed Prantl more or less and recognized Lacostea as a subgenus. In the preliminary pages of the ‘Index Filicum,” published last (with a preface dated Sept. 15, 1906), he gave his own ideas of a suitable arrangement and here he recognized Lacostea as a section of his subg. Cephalomanes, citing T. pedicellatum Desv. as the type species, which was not one of the original species. However, this was in line with Christensen’s practice of citing recognized species as the types. In the body of the text he refers T. brachy- pus Kunze to the synonymy of T. pedicellatum, and therefore really in- tended the type to be 7. brachypus Kunze, now correctly called T’.. pedzcel- latum, a taxonomic synonym. This fixes the lectotype as T. brachypus Kunze. Copeland’s choice ® of T. ankersit Parker as type must therefore be rejected. Lacostea sect. Lacostea Prantl, Untersuch. Morph. Gefasskrypt. 1: 50. 1875. Trichomanes sect. Lacostea (v. d. Bosch) Christ, Farnkr. Erde 29. 1897 (erroneously attributed to v. d. Bosch). Trichomanes subg. Lacostea (v. d. Bosch) C. Chr. Ind. Fil. 634. July 2, 1906. 50 Phil. Journ. Sci. 67: 15. 1938. 196 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM Copeland did not recognize Lacostea either as a genus or section, but reduced it outright to his genus Trichomanes, my Trichomanes subg. Achomanes, but it seems to deserve recognition as a section. The species of Lacostea, all American, are probably only the following: Trichomanes ankersit Parker, T. pedicellatum Desv., T. tanaicum Baker, and 7. tuerckhetmii Christ. Trichomanes sect. 21. Trigonophyllum (Prantl) C. Chr. Ptilophyllum sect. Trigonophyllum Prantl, Untersuch. Morph. Gefasskrypt. 1: 48. 1875. Type: There were two original species, 7. bancroftii Hook. & Grev. and 7. bicorne Hook. In selecting a type, Christensen 5 chose 7’. arbuscula Desv., not one of the original species, but considered by Christen- sen to be an older taxonomic synonym of 7. bancroftei Hook. & Grev., which he reduced to synonymy. This fixes the lectotype as 7. bancroftii Hook. & Grev. Trichomanes sect. Trigonophyllum (Prantl) C. Chr. Ind. Fil. XV. 1906. This small group perhaps still includes only the two species T. arbuscula Desv. and T. bicorne Hook., which do not appear to be especially closely allied. Trichomanes sect. 22. Homoeotes (Presl) C. Chr. Homoeotes Pres], Abhandl. Boehm. Gesell. V, 5: 331. 1848. Type: Trichomanes heterophyllum Humb. & Bonpl. ex Willd., the sole original species. Ptilophyllum sect. Homoeotes (Presl) Prantl, Untersuch. Morph. Gefisskrypt. 1: 48. 1875. Trichomanes sect. Homoeotes (Presl) C. Chr. Ind. Fil. XIV. 1906. This monotypic section differs from Feea and Hymenostachis primarily in having the rhizomes elongate and the fronds distichous. Two other species that have been considered allied are probably not referable here, T. amazonicum Christ and TJ. spruceanum Hook. Their proper position remains to be determined, perhaps as a subsection of Feea. Trichomanes sect 23. Feea (Bory) Christ Feea Bory, Dict. Class. d’Hist. Nat. 6: 446. 1824. Lectotype: Feea poly- podina Bory=Trichomanes osmundoides DC., taxonomically. There were two original species, I’. polypedina and F. nana Bory. The first to choose a lectotype was apparently J. Smith,’ who chose Trichomanes spicatum Hedw.—not one of the original species but was considered, doubt less rightly, a taxonomic synonym of Feea polypodina Bory by Presl. This fixes the lectotype as F. polypodina. In the “Index Filicum” (XIV. 1906) Christensen indicated the type as Trichomanes botryoides Kaulf., an impossible choice and not one of the original species. Copeland ® erro- neously indicated F. nana Bory as the type. In the “Genera Filicum” (1947), Copeland corrected this to F. polypodina. Trichomanes subg. Feea (Bory) Hook. Sp. Fil. 1: 114. 1844. 51 Ind. Fil. XV. 1906. 52 Hist. Fil. 349. 1875. 53 Phil. Journ. Sci. 64: 13. 1937. MORTON—-HYMENOPHYLLACEAE 197 The section [cea is so close in its general morphology to the species of sect. Achomanes that is seems unnatural to segregate it as a distinct genus, as Copeland does solely on the basis of the dimorphism of the fronds. The section Homoeotes is perhaps a connecting link, for its leaves are only slightly dimorphic. The following two subsections may be recognized. Trichomanes sect. Feea subsect. Feea. Maschalosorus v. d. Bosch, Versl. Meded. Konink. Wetens. Afd. Natuurk. 11: 320. 1861. Type: 7. mougeotit v. d. Bosch. =7. osmundoides DC., taxonomically. Ptilophyllum sect. Feea (Bory) Prantl, Untersuch. Morph. Gefasskrypt. 1: 48, 1875. Prantl recognized Feea as a section and reduced Hymenostachis to a straight synonym; therefore, as sectional name Feea has priority. Trichomanes sect. Feea (Bory) Christ, Farnkr. Erde 29. 1897 (erroneously attributed to Bory). Trichomanes sect. Maschalosorus (v. d. Bosch) C. Chr. Ind. Fil. XIV. 1906. The only species of subsect. Feea, which has free, stalked involucres, are T’. osmundoides DC. and T. botryoides Kaulf. Trichomanes sect. Feea subsect. Hymenostachis (Bory) Morton, comb. nov. Hymenostachis Bory, Dict. Class. d’Hist. Nat. 6: 588. 1824. Type: T’richomanes elegans Rudge, the only original species =Trichomanes diversifrons (Bory) Mett.. The name T. elegans Rudge (1805) was a later homonym of 7. elegans L. C. Rich. (1792). The name ‘“Hymenestachis’”’ has been altered by almost all later authors to “Hymenostachys” on orthographical grounds, but probably unjustifiably. Bory doubtless knew a good deal of Greek and deliberately chose to use the transliteration ‘-stachis.” Trichomanes subg. Hymenostachis (Bory) Hook. Sp. Fil. 1: 114. 1844. The first to unite Feea and Hymenostachys as subgenera was apparently Chris- tensen,®4 who chose Feea and reduced Hymenostachis to a section, and therefore, for those who may wish to recognize F’eea as a subgenus rather than a section, the correct name is subg. Feea (Bory) Hook. In subsect. Feea the involucres are separated and somewhat stalked, and very little leaf-tissue is on the fertile fronds. ‘The two known species of subsect. Hymenostachis, T. diversifrons (Bory) Mett. and T. trollii Bergd., have the fertile frond entire, about three cells thick, and with the involucres deeply and completely immersed in the leaf-tissue. The group has not been recognized as distinct recently but can conveniently be separated. Trichomanes sect. 24, Ragatelus (Presl) C. Chr. Ragatelus Pres], Hymen. 14. 1843. Type: The type and sole original species is Ragatelus crinitus (Swartz) Presl=Trichomanes crinitum Swartz. Ptilophyllum v. d. Bosch, Versl. Meded. Konink. Akad. Wetens. Afd. Natuurk. 11: 321. 1861. Type: As originally proposed, this genus included a large number of species, nos. 17 to 24 and 26 to 43 of van den Bosch’s “Synopsis.” Inasmuch as species 48, Trichomanes crinitum Swartz, had already received the generic name Ragatelus Presl in 1843, the name 54 Ind. Fil. XIV. 1906. 198 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM Ptilophyllum was superfluous, since van den Bosch could have and should have used the prior name Ragatelus. Moreover, Ptilophyllum was later homonym of Ptilophyllum Reichenb. (1841) and thus doubly illegitimate. Nevertheless, Prantl adopted Ptilophyllum, reducing to synonymy such validly published earlier names as Feea, Hymenostachis, Ragatelus, and Neuromanes. In fact, Prantl even created a suprageneric group, the Ptilo- phylleae, but since this was called a “family” and yet was subordinated to a ‘tribe’ it is an invalid name, since the Code prohibits such changes in the relative order of categories, i.e., a family can be divided into tribes but a “‘tribe cannot be divided into families as Prantl tried to do. Since Ptilophyllum included the type of the previously published * Ragatelus Presl, it must be considered as an illegitimate change of name for the latter and to have the same type, namely J'richomanes crinitum Swartz. Trichomanes sect. Ptilophyllum (v. d. Bosch) Christ, Farnkr. Erde 27. 1897 (as to type, not as to concept). Trichomanes subg. Ptilophyllum (v. d. Bosch) C. Chr. Ind. Fil. 634. 1906 (as to type, not as to concept). Trichomanes sect. Ragatelus (Presl) C. Chr. Ind. Fil. XIV. 1906. The genus Ragatelus was erected on the impression that the recep- tacle was bifid, which was an erroneous observation or else a casual teratological form. This does not, of course, invalidate the publication of the name. The species 7. crinitum Swartz is rather distinctive and isolated and can tentatively be accepted as representing a monotypic section, as Christensen did. Copeland did not give it any recognition. Trichomanes sect. 25. Acarpacrium (Prantl) C. Chr. T'richomanes subg. Pseudachomanes Presl, Epim. Bot. 16. 1849 [1851]. Type: Five species were originally assigned to this subgenus by Presl: 7’. sinuosum L. C. Richard, 7. cognatum Presl, T. holopterum Kunze, 7. bancroftii Hook. & Grev., and 7. alatum Swartz. In the ‘Index Filicum,’’ Christensen chose 7’. polypodioides L. as the type, which was not one of the original species, but considered by Christensen as a taxonomic synonym and the correct name for 7’. sinuosum L. C. Richard; therefore, the latter must be considered as having been designated as lectotype. The matter is of some importance, for the species listed originally by Presl are not really closely allied. ; Ptilophyllum sect. Achomanes [subsect.] Sinuosa Prantl, Untersuch. Morph. Gefisskrypt. 1: 47. 1875. Lectotype: Trichomanes sinuosum L. C. Richard. Ptilophyllum sect. Acarpacrium Prantl, Untersuch. Morph. Gefisskrypt. 1: 48. 1875. Type: There were six species referred originally to the section. Christensen in the “Index Filicum’’ chose 7. alatum Swartz as the type, an impossible choice since this was not one of the original species, but this inferentially selects as lectotype 7'richomanes ptilodes v. d. Bosch, the first named of the original species, which Christensen considered a taxonomic synonym of 7’. alatum. Trichomanes sect. Acarpacrium (Prantl). C. Chr. Ind. Fil. XIV. 1906. Trichomanes sect. Pseudachomanes (Presl) C. Chr. Ind. Fil. XIV. 1906. This small group of species can be distinguished from sect. Acho- manes by the presence of stellate (or at least bifid) hairs on the margins of the segments and also on the veins to some extent. It does not MORTON—HYMENOPHYLLACEAE 199 appear necessary to distinguish between the merely pinnatifid species as sect. Pseudachomanes, such as T. polypodiides L. (syn. T. sinuosum L. C. Richard) and the more dissected species like T. alatum Swartz of sect. Acarpacrium, as Christensen did. The character of the rhizome hairs (termed “‘paleac” by Prantl) being straight in 7. sinuosum and its allies and peltate in T. crispum L. and its allies of the section Achomanes needs further study; apparently this has received no attention since the time of Prantl. The number of species is uncertain ; they include Trichomanes adscendens Kunze, T. alatum Swartz, T. attenuatum Hook., T. curranii Weath., T. fimbriatum Backh., T. pinna- tifidum v. d. Bosch, J. polypodioides L., T. ptilodes v. d. Bosch, and 7’. trigonum Desv. Unplaced Sectional Name Trichomanes sect. Flabellata Presl, Hymen. 16. 1843. Gonocormus sect. Microtrichomanes Mett. ex Prantl, Untersuch. Morph. Gefisskrypt. 1: 51. 1875. Lectotype: 7’. digitatum Swartz (chosen by Christensen). Trichomanes sect. Microtrichomanes (Mett. ex Prantl) C. Chr. Ind. Fil. XIV 1906. Microtrichomanes (Mett. ex Prantl) Copel. Phil Journ. Sci. 67: 35. 1938.5 LecroryPe: T'richomanes digitatum Swartz. This group ‘“Microtrichomanes” has been one of the stumbling blocks in the recent classification of the family, because in some ways it seems to bridge the gap between Trichomanes and Hymenophyllum and does not fit into either comfortably; but this fact ought not influence our classification unduly. If it were a homogeneous group as delimited by Copeland, the species obviously more closely related to each other than to any others and to have had a common ancestry, then one would have some hope of placing it definitely in a system, in Trichomanes, in Hymenophyllum, or in a distinct genus intermediate between them. But this is not the case. It is decidedly heterogeneous and would seem to be polyphyletic. It may have arisen anciently from hybridization between various species of Trichomanes and Hymeno- phyllum, perhaps belonging to quite different groups, in which case it 5 Mettenius proposed this previously (Abhandl. Math.-Phys. Class. Konig. Sachs. Gesell. Wiss. 7: 413. 1864) as “gruppe” Micro-trichomanes, but this can- not be considered a valid publication; since “‘gruppe”’ is not a recognized taxonomic category, it must be considered only an informal grouping. As members of the group Mettenius listed T. digitatum Swartz, T. palmatifidum C. Muell., and T. dichotomum Kunze; these same species were listed by Prantl when he recognized the section, along with two others, 7’. nitidulum v. d. Bosch and T. flabellatum v. d. Bosch. In recognizing this group as a section Christensen chose 7. digitatum as lectotype. % Holttum, Ferns Malaya 94. 1954, proposed Trichomanes subg. Microtricho- manes, but invalidly without the definite citation of the basionym. 284-138—68——_4 200 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM would have little or no unity as a “bybrid genus,” and would not really deserve separate generic status. Apparently most of the species can be accommodated in existing 7richomanes sections without undue problems, but they are too little known to me for me to make the attempt. The nine species referred here by Copeland are distributed from the East African Islands through Malaya, Malesia, and Polyne- sia to Australia, but they are everywhere rare. They should each be investigated morphologically from good fresh material and each one placed according to its merits, which is not something that is likely to be done in the near future. The somewhat doubtful species (generically) referred here are: T. aswijkit Racib., T. dichotomum Kunze, T. digitatum Swartz, T. flabellatum v. d. Bosch, 7. francit Christ, T. lyallit (Hook. f.) Hook., T. nitidulum v. d. Bosch, T. palmatifidum K. Muell. (Hymenophyllum borneense Hook.), T. parvulum Poir., T. piliferum van Alderw. van Rosenb., T. ridleyt Copel., T. taeniatum Copel., T. vitiense Baker, and Microtrichomanes zamboanganum Copel. Rejected Names Achomanes Necker, Elem. Bot. 3: 313. 1790. Although often cited as a synonym of Trichomanes L., this name must be rejected as invalid, as are all of Necker’s names according to Art. 20, Note 2, of the International Code of Botanical Nomenclature (1961 ed.). Adiantopsis v. d. Bosch. Christensen *” and Copeland ® cite this name as a synonym of Davalliopsis v. d. Bosch, but both without citing any place of publica- tion. I have been unable to locate such a name among van den Bosch’s works. If the name were published, it would be an illegitimate later homonym of Adiantopsis Fée (1852). Boschia Copel. Phil. Journ. Sci. 64: 11. 1937. A name not otherwise mentioned by Copeland, probably a pro- visional name that be intended to use (perhaps for Mecodium?) and later abandoned but overlooked here. Hemiphlebium Presl, Hymen. 25, pl. 9. 1843. Type (the only original species) : Hemiphlebium pusillum (Swartz) Presl= Trichomanes pusillum Swartz. Presl did have authentic type material of 7. pustllum Swartz in hand. However, evidently one or more fronds of a different species were intermixed, which in Jamaica could only have been T’. hookeri Presl, since Presl described and illustrated an intramarginal false vein, such as occurs in 7’. hookeri but not in 7. pusillum. That the drawing was partly made from the true 7’. pusillum 51 Ind. Fil. XV. 1906. 58 Gen. Hymen. 14. 1938. MORTON—HYMENOPHYLLACEAE 201 is shown by the marginal stellate hairs, which occur in pusillum and not in hookeri. Since Presl’s differentiation of this genus Hemiphlebium from Didymoglossum was primarily on the presence of this false vein and since such a false vein does not occur in the cited type, it seems that there is no alternative but to abandon Hemiphlebium as a nomen confusum. This seems to be the position taken by J. G. Wessels Boer in his recent ‘“The New World Species of Trichomanes sect. Didymoglossum and Microgonium.” ® Trichomanes sect. Hemiphlebium (Presl) Moore, Ind. Fil. cx. 1857. Trichomanes subg. Hemiphlebium (Presl) Christ, Farnkr. Erde 23, 1897. Trichomanes sect. Euhemiphlebium (Presl) Christ, Farnkr. Erde 23. 1897. Trichomanes sect. Eutrichomanes [subsect.] Hemiphlebium (Presl) van Alderw. van Rosenb. Malay. Ferns 83. 1908. The following names appear in Christ’s ‘Farnkratiter der Erde” (1897) apparently intended as sections, but they are all nomina nuda, without descriptions: Hymenophyllum sect. Ciliata, Crispata, Elastica, Exserta, Flabellata, Fuciformia, Pendula, Pilosa, Pinnatifida, and Polyantha. In those instances where Christ has indicated that these names are adopted from Presl, Prantl, or van den Bosch, I have con- sidered that these are validly proposed new combinations, where they differ from the original combination. Some names, none original, appear in Sadebeck’s treatment of the Hymenophyllaceae in Engler and Prantl’s ‘Die Natiirlichen Pflanzen- familien,” (1899) but without any indication of rank or any consistency in usage, some groups receiving names and others of coordinate rank without names. I have ignored these, since Sadebeck obviously did not intend to present a complete classification into subgenera, sections, and subsections but merely applied names to groups that had already received names by Presl or other previous authors, without designating or deciding on any rank. Copeland in his paper on Trichomanes (Phil. Journ. Sci. 51: 131. 1933) used several new names: Pyzidifera, Taschneria, Scandentia, Grandia, Apiifolia, and Rigida; these were not considered either as subgenera or sections but were merely called “groups,” which is not a recognized botanical category. I consider them as merely informal groupings and unpublished. Some of these same groups were indicated as sections by Alston in his key in the “Ferns and Fern-allies of West Tropical Africa” (1959) but without validating Latin descriptions or references to previously published descriptions, and these too are un- published. 6 Acta Bot. Neerl. 11: 277-330. 1962. Index (Synonyms in italics. New combinations in boldface. Page numbers of principal entries in italics.) Abrodictyum, 185 cumingit, 185 Abrodictyum, sect., 178, 185 Abrodictyum, subsect., 185 Acanthotheca, sect., 166 Acarpacrium, sect., 180, 198, 199 Achomanes, 192, 193, 200 Achomanes, sect., 180, 192, 193, 194, 198, 199 Achomanes, subg., 158, 178, 179 (key), 189, 192, 193, 196 Adiantopsis, 200 Adiantum, 177 chinense, 175 Amphipterum, 174 humatoides, 174 Amphipterum, sect., 174 Amphipterum, subg., 174 Amphipterum, subsect., 163, 174 Apttfolia, group, 153, 201 Aplopappus, 185 Apteropteris, 170 Apteropteris, sect., 163, 168, 170 Apteropteris, subg., 170 Asplenium, 175, 177 trichomanes, 180 Boschia, 200 Buesta, 164 megistocarpa, 165 Buesia, sect., 163, 164 Buesia, subg., 164 Callistopteris, 190 calyculata, 190 Callistopteris, sect., 179, 190 Cardiomanes, 159, 160, 176 reniforme, 175 Cardiomanes, subg., 175 Cardiomanes, subsect., 175 Cephalomanes, 189 atrovirens, 189 Cephalomanes, sect., 179, 189 Cephalomanes, ser., 189 Cephalomanes, subg., 189 Chilodium, sect., 166 Chilodium, subg., 165, 190 Ciliata, sect., 201 Ciliata, subsect., 163, 166, 168 Cincinalis, 180 aquilina, 180 Craspedoneuron, sect., 186 Craspedoneuron, subg., 186 Craspedophyllum, 171 Craspedophyllum, sect., 163, 170, 171 Craspedophyllum, subg., 160, 162, 170 Crepidium, 184 Crepidium, sect., 178, 184 Crepidium, subg., 184, 190 Crepidomanes, 183 nanophyllum, 183 pseudonymanit, 183 tagawanum, 183 Crepidomanes, sect., 178, 182, 183 Crepidomanes, subg., 182, 189 Crepidophyllum, 184 Crepidopteris, 184 Crepidopteris, subg., 184 Crispa, subsect., 180, 193, 194 Crispata, sect., 201 Ctenopteris, 155 Davallia, 177 canariensis, 175 Davalliopsis, 188, 189 Davalliopsis, sect., 188, 189 Dermatophlebium, 169 Dermatophlebium group, 169 Dermatophlebium, sect., 169 Didymoglossum, 183, 190, 191, 201 subg. Chilodium, 165, 190 subg. Crepidium, 184, 190 subg. Eudidymoglossum, 190, 191 sect. Flabellata, 191 sect. Pinnata, 191 acanthoides, 166 alatum, 190 decipiens, 190 jilicula, 190, 191 humile, 184 lineare, 190 203 204: Didymoglossum—Continued longisetum, 165 magellanicum, 190 muscoides, 190, 191 punctatum, 190 pusillum, 190 Didymoglossum, sect., 179, 183, 191, 201 Didymoglossum, subg., 158, 159, 178, 179 (key), 190, 191, 192 Dimidiata, subsect., 164 Diploophyllum, 174 Diplophyllum, 174 Diplophyllum, sect., 174 Diplophyllum, subsect., 164, 174, 175 Elastica, sect., 201 Eleocharis, 185 Euachomanes, sect., 193 Eudidymoglossum, subg., 190, 191 Euhemiphlebium, sect., 181, 201 Euhymenophyllum, subg., 164, 172 EKupectinum, sect., 163, 168 Eutrichomanes, sect., 181, 187, 189 Eutrichomanes, subg., 177, 178, 181, 183, 187, 188, 193 Evoluta, subsect., 164 Exserta, sect., 201 Feea, 193, 196, 198 nana, 196 polypodina, 196 Feea, sect., 179, 196, 197 Feea, subg., 196 Feea, subsect., 179, 197 Flabellata, sect., 191, 199, 201 Fuciformia, sect., 201 Glabra, sect., 172 Globosa, sect., 172 Gonocormus, 186 sect. Microtrichomanes, 199 minutus, 186 prolifer, 186 Gonocormus, sect., 179, 186, 187 Gonocormus, subg., 187 Gonocormus, subsect., 187 Grammitis, 155 Grandia, group, 153, 201 Habrodictyum, 185 Haplopappus, 185 Heleocharis, 185 Hemicyatheon, 171 Hemicyatheon, sect., 163, 171 Hemicyatheon, subg., 162, 171 182, 185, CONTRIBUTIONS FROM THE NATIONAL HERBARIUM Hemiphlebium, 190, 200, 201 sect. Lecanium, 192 sect. Microgonium, 191 pusilum, 200 Hemiphlebium, sect., 201 Hemiphlebium, subg., 181, 201 Hemiphlebium, subsect., 201 Hirsuta, subsect., 163, 169, 170 Holophlebium, subg., 181 Homoeotes, 196 Homoeotes, sect., 179, 196, 197 Hymenoglossum, 159, 160, 175 cruentum, 160 Hymenophyllum, 153-58, 160, 161, 162 (key), 164, 165, 170, 174, 176, 177, 190,193,199 sect. Acanthotheca, 166 sect. Amphipterum, 174 subg. Amphipterum, 174 subsect. Amphipterum, 163, 174 sect. Apteropteris, 163, 168, 170 subg. A pleropteris, 170 sect. Buesia, 163, 164 subg. Buesia, 164 subg. Charitophyllum, 168 sect. Chilodium, 166 sect. Ciltata, 201 subsect. Ciliata, 163, 166, 168, 169 sect. Craspedophyllum, 170, 171 subg. Craspedophyllum, 160, 162, 163, 170 sect. Crispata, 201 sect. Cycloglossum, 161 subg. Cycloglossum, 161 sect. Dermatophlebium, 169 subsect. Dimidiata, 164 sect. Diplophyllum, 174 subsect. Diplophyllum, 164, 174 sect. Elastica, 201 subg. Huhymenophyllum, 164, 168, 172 sect. Eupectinum, 163, 168 subg. Euphorophyllum, 172 subsect. Evoluta, 164 sect. Hxserta, 201 sect. Flabellata, 201 sect. Fuciformia, 201 sect. Glabra, 172 sect. Globosa, 172 subg. Hapalophyllum, 172 sect. Hemicyatheon, 171 | subg. Hemicyatheon, 162, 163, 171 | subsect. Hirsuta, 163, 169, 170 INDEX 205 Hymenophyllum—Continued sect. Hymenoglossum, 160 sect. Hymenophyllum, 163, 164 subg. Hymenophyllum, 162, 163 (key), 164, 171, 174 sect. Integra, 172 sect. Leptocionium, 170 subg. Leptocionium, 165, 170 subsect. Leptocionium, 163, 170 sect. Mecodium, 163, 172, 174 subg. Mecodium, 162, 163 (key), 166, 171, 172 subsect. Mecodium, 163, 172, 173 sect. Meringium, 166 subg. Meringium, 165, 166 sect. Myriodon, 163, 168 subg. Myriodon, 168 sect. Myrmecostylum, 166 sect. Pachyloma, 171 sect. Pendula, 201 sect. Pilosa, 168, 170, 201 sect. Pinnatifida, 201 subg. Platynophyllum, 174 subsect. Plumosa, 163, 169 sect. Polyantha, 201 sect. Ptychophyllum, 163, 166 sect. Serrulata, 164 sect. Sphaerocionium, 163, 168, 169, 170 subg. Sphaerocionium, 162, 163 (key), 166, 168, 174, 194 sect. Sphaerodium, 164 subg. Sphaerodium, 164 Hymenophyllum abruptum, 173 acanthoides, 166, 167 adiantoides, 169 aeruginosum, 169 affine, 167 alfredii, 173 alveolatum, 164 amabile, 170 andinum, 173 angulosum, 173 angustum, 170 antarcticum, 164 antillense, 170 apiculatum, 173 archboldii, 167, 173 armstrongii, 171 asperulum, 164 asplenioides, 172, 173 atrovirens, 173 australe, 175 Hymenophyllum —Continued axillare, 173 badium, 173 baileyanum, 171 bakeri, 167 balansae, 173 bamlerianum, 173 barbatum, 164 bartlettii, 167 batuense, 167 bicolanum, 167 bismarckianum, 173 bivalve, 167 blandum, 167 blepharodes, 168 bontocense, 167 borneense, 200 botryoides, 173 brachyglossum, 167 brachypus, 164 brassii, 168 breve, 173 brevidens, 167 brevistipes, 173 bryophilum, 167 calodictyon, 167 campanulatum, 167 caparavense, 169 capillare, 169 cardunculus, 167 carnosum, 173 caudiculatum, 172, 173 ceratophylloides, 164 ciliatum, 169 cincinnatum, 167 consanguineum, 169 constrictum, 173 contextum, 173 contiguum, 173 contortum, 173 copelandii, 173 corrugatum, 173 costaricanum, 173 crispato-alatum, 173 crispatum, 173 crispum, 168, 169 cristatum, 165 cruentum, 160 cuneatum, 173 cupressiforme, 164 darwinii, 173 delicatulum, 169 demissum, 175 206 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM Hymenophyllum—Continued dendritis, 173 dentatum, 164 denticulatum, 165, 167 dependens, 170 deplanchei, 171 dichotomum, 165 dicranotrichum, 170 dilatatum, 174 dimidiatum, 167 dimorphum, 169 diversilabium, 173 durandii, 167 ectocarpon, 167 edanoi, 173 edentulum, 166, 167 elbertii, 167 elegans, 169 elegantulum, 170 ellipticosorum, 167 emarginatum, 173 endiviifolium, 173 exsertum, 173 falklandicum, 164 farallonense, 173 fecundum, 173 feejeense, 167 fendlerianum, 173 ferax, 173 ferrugineum, 170 fimbriatum, 173 firmum, 167 flabellatum, 173 flexile, 173 flexuosum, 173 foersteri, 167 fragile, 170 fucoides, 166 fumarioides, 173 funckii, 173 fuscum, 174 fusugasugense, 169 gardneri, 173 geluense, 174 glaziovii, 169 gollmeri, 173 gorgoneum, 167 gracilescens, 164 hallieri, 167 heimii, 173 helicoideum, 173 hemipteron, 170 herzogii, 164 Hymenophyllum—Continued hieronymii, 167 hirsutum, 168-170 hirtellum, 169 holochilum, 166, 167 horizontale, 169 hosei, 167 howense, 167 humbertii, 173 hygrometricum, 169 imbricatum, 173 inaequale, 173 interruptum, 170 intricatum, 173 involucratum, 173 ivohibense, 169 jamesonii, 165 javanicum, 172, 173 johorense, 167 junghuhnii, 172, 173 karstenianum, 170 kerianum, 167 klabatense, 167 krauseanum, 167 kuhnii, 173 laciniosum, 173 laminatum, 174 lanatum, 170 lanceolatum, 169 latifolium, 167 latifrons, 170 laxum, 167 ledermannii, 174 lehmannii, 173 leratii, 173 levingii, 173 lindenii, 170 lineare, 168, 169 lobato-alatum, 169 lobbii, 167 longifolium, 173 macroglossum, 166, 169 macrosorum, 167 macrothecum, 173 magellanicum, 167 marginatum, 170, 171 marlothii, 169 maxonii, 169 megistocarpum, 165 melanosorum, 167 merrillii, 167 mexiae, 173 meyenianum, 165, 167 Hymenophyllum—Continued micans, 173 microcarpum, 169 microphyllum, 173 microsorum, 173 minimum, 167 mirificum, 164, 165 mnioides, 173 molle, 169 montanum, 173 moorei, 164 multialatum, 169 multifidum, 167 myriocarpum, 173 nahuelhuapiense, 164 nanum, 164 nigrescens, 173 nitiduloides, 174 nutantifolium, 167 obtusum, 169 odontophyllum, 168 oligosorum, 174 ooides, 174 opacum, 174 osmundoides, 174 ovatum, 167 pachydermicum, 166, 169 paniculiflorum, 174 parvulum, 174 paucicarpum, 174 pectinatum, 168 pediculariifolium, 167 peltatum, 158, 164 penangianum, 167 perfissum, 164 perparvulum, 167 peruvianum, 167 piliferum, 167 pilosissimum, 170 plicatum, 165, 167 plumieri, 170 plumosum, 169 poilanei, 167 pollenianum, 167 polyanthos, 172-174 poolii, 169 praetervisum, 167 prionema, 169 productoides, 174 productum, 174 protrusum, 174 pseudotunbridgense, 167 pulchellum, 169 INDEX 207 Hymenophyllum—Continued pulcherrimum, 174 pulchrum, 167 pumilio, 167 pumilum, 164 pyramidatum, 169 quadrifidum, 165 raddianum, 170 ramosii, 167 rarum, 174 recurvum, 174 reductum, 167 reinwardtii, 167, 172 remotipinna, 174 reniforme, 174 retusilobum, 174 revolutum, 164 ricciifolium, 167 rimbachii, 174 riukiuense, 174 rolandi-principis, 161 roraimense, 169 rosenstockii, 167 rubellum, 167 rufescens, 174 rufifolium, 167 rufifrons, 167 rufum, 170 rugosum, 164 ruizianum, 169 salakense, 174 samoense, 167, 172 sampaioanum, 169 sanguinolentum, 172, 174 scabrum, 174 secundum, 164 semiglabrum, 169 sericeum, 169 sieberi, 170 siliquosum, 174 silvaticum, 169 silveirae, 169 simonsianum, 164 simplex, 170 skottsbergii, 174 sodiroi, 165 speciosum, 170 spicatum, 164 splendidum, 170 streptophyllum, 174 subdimidiatum, 167 subobtusum, 169 subrigidum, 169 9208 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM Hymenophyllum—Continued superbum, 170 tablaziense, 174 taiwanense, 167 tenellum, 167 tenerrimum, 169 tenerum, 174 thuidium, 167, 172 todjambuense, 174 tomentosum, 169 torquescens, 164 torricellianum, 167 tortuosum, 167 trapezoidale, 169 treubii, 174 trianae, 174 triangulare, 167 trichomanoides, 174 trichophyllum, 169 tunbridgense, 158, 164 uncinatum, 164 undulatum, 174 unilaterale, 158, 164 valvatum, 168 veronicoides, 174 viguieri, 169 vittatum, 167 walleri, 174 whitei, 174 wilsonii, 158, 164 wrightii, 174 Hymenostachis, 193, 197, 198 Hymenostachis, sect., 196 Hymenostachis, subg., 197 Hymenostachis, subsect., 179, 197 Hymenostachys, 197 Integra, sect., 172 Lacostea, 189, 193, 195, 196 sect. Cephalomanes, 189 Lacostea, sect., 179, 195, 196 Lacostea, subg., 195 Lacostea, subsect., 182, 189 Lacosteopsis, sect., 178, 181 Lamellata, subsect., 180, 194 Lecanium, 177, 192 membranaceum, 192 Lecanium, sect., 179, 181, 192 Lecanopteris, 192 Leptocionium, 170 dicranotrichum, 170 fucoides, 170 Leptocionium, sect., 170 Leptocionium, subg., 165, 170 Leptocionium, subsect., 163, 170 Leptomanes, sect., 185 Leucomanes, 185 Macroglena, 188 obtusa, 188 truncata, 188 Macroglena, subg., 187 Mazora, subg., 180, 181 Maschalosorus, 197 Maschalosorus, sect., 197 Mecodium, 158, 171, 172, 200 archboldit, 173 contiguum, 173 diversilabium, 173 edanot, 173 mexiae, 173 taiwanense, 167 Mecodium, sect. 163, 172, 174 Mecodium, subg., 162, 163, 171, 172 Mecodium, subsect., 163, 172, 173 Meringium, 158, 165, 166, 171-173 archboldit, 167 bartletti?, 167 blumeanum, 165 latifolium, 167 laxum, 167 melanosorum, 167 meyenianum, 165 Meringium, sect., 166 Meringium, subg., 165, 166, 170 Microgonium, 191 berteroanum, 191 cuspidatum, 191 falsivenulosum, 192 Microgonium, sect., 179, 181, 191, 192, 201 Microgonium, subg., 192 Microtrichomanes, 199 zamboanganum, 200 Microtrichomanes, sect., 199 Microtrichomanes, subg., 199 Minora, subg., 181 Myriodon, 168 Myriodon, sect., 163, 168 Myriodon, subg., 168 Myrmecostylum, 165, 166 clavatum, 165 dichotomum, 165 tortuosum, 165 Myrmecostylum, sect., 166 Nesopteris, 190 Nesopteris, sect., 179, 190 Neuromanes, 193-195, 198 INDEX Neuromanes, sect., 194, 195 Neurophyllum, 194, 195 pennatum, 194 pinnatum, 194 vittaria, 194 Neurophyllum, sect., 179, 194, 195 Odontomanes, 193, 195 hostmannianum, 195 Odontomanes, sect., 179, 195 Onychium, 192 Pachychaetum, sect., 179, 187, 188 Pachychaetum, subg., 178, 179 (key), 187, 188 Pachyloma, 171 Pachyloma, sect., 171 Pendula, sect., 201 Phlebiophyllum, 184 Phlebiophyllum, sect., 178, 183 Phlebiophyllum, subg., 161 Phlebodium, 183 brownit, 184 Phlebophyllum, 184 Pilosa, sect., 168, 170, 201 Pinnata, sect., 181, 191 Pinnatifida, sect., 201 Platynophyllum, subg., 174 Pleuromanes, 185 retusum, 186 Pleuromanes, sect., 179, 185, 186 Pleuromanes, subg., 185 Plumosa, subsect., 163, 169 Polyantha, sect., 201 Polyphlebium, 184 Polyphlebium, subg., 184 Polystichum, 180 lonchitis, 180 Pseudachomanes, sect., 198, 199 Pseudachomanes, subg., 198 Pieris Gled., 180 Ptilophyllum, 193, 197, 198 sect. Acarpacrium, 198 sect. Achomanes, 193, 194 subsect. Crispa, 193 sect. Feea, 197 sect. Homoeotes, 196 subsect. Lamellata, 194 subsect. Sinuosa, 198 Ptilophyllum, sect., 198 Ptilophyllum, subg., 198 Ptychomanes, 164 Ptychophyllum, 165, 166 plicatum, 165 Ptychophyllum, sect., 163, 165, 166 209 Pyzidaria, 180 Pyzxidifera, group, 153, 201 Ragatelus, 193, 197, 198 Ragatelus, sect., 180, 197, 198 Rigida, group, 153, 201 Rosenstockia, 160, 161 rolandi-principis, 161 Scandentia, group, 153, 201 Schizophlebium, subg., 189 Selenodesmium, 187, 188 Selenodesmium, sect., 188 Selenodesmium, subg., 188 Serpyllopsis, 160, 161 caespitosa, 160 Serpyllopsis, subg., 161 Serrulata, sect., 164 Sinuosa, subsect., 198 Sphaerocionium, 158, 168 sect. Glabra, 172 sect. Pilosa, 168 sect. Stellata, 169 hirsutum, 168 schiedeanum, 168 Sphaerocionium, sect., 163, 168, 169, 170 Sphaerocionium, subg., 162, 166, 168, 174, 194 Sphaerodium, sect., 164 Sphaerodium, subg., 164 Sphenomeris chinensis, 175 Stellata, sect., 169 Taschneria, 183, 191 filicula, 191 Taschneria, group, 183, 201 Taschneria, sect., 183 Tetralasma, 165 Trichomanes, 153-158, 160-162, 174, 177, 178 (key), 180, 184, 188-190, 193, 194, 196, 199, 200, 201 sect. Abrodictyum, 178, 185 subsect. Abrodictyum, 185 sect. Acarpacrium, 180, 198, 199 sect. Achomanes, 180, 193, 194, 198, 199 subg. Achomanes, 176-178, 179 (key), 189, 192, 193, 196 group Apiifolia, 201 sect. Callistopteris, 179, 190 subg. Cardiomanes, 175 subsect. Cardiomanes, 175 sect. Cephalomanes, 179, 189, 190 ser. Cephalomanes, 189 subg. Cephalomanes, 189 sect. Craspedoneuron, 186 210 Trichomanes—Continued CONTRIBUTIONS FROM THE NATIONAL HERBARIUM subg. Craspedoneuron, 186 sect. Crepidium, 178, 184 sect. Crepidomanes, 178, 182, 183 subg. Crepidomanes, 182 subg. Crepidopteris, 184 subsect. Crispa, 180, 193, 194 sect. Davalliopsis, 188, 189 sect. Didymoglossum, 179, 183, 191, 201 subg. Didymoglossum, 178, 179 (key), 190, 191, 192 sect. Huachomanes, 193 sect. Huhemiphlebium, 181, 201 sect. EHutrichomanes, 175, 180, 181, 182, 185, 187, 189 subg. Eutrichomanes, 176-178, 180, 181, 183, 187, 188, 193 sect. Feea, 179, 196, 197 subg. Feea, 196 subsect. Feea, 179, 197 sect. Flabellata, 199 sect. Gonocormus, 179, 186, 187 subg. Gonocormus, 187 subsect. Gonocormus, 187 group Grandia, 201 sect. Hemtphlebium, 201 subg. Hemiphlebium, 181, 201 subsect. Hemiphlebium, 201 subg. Holophlebium, 181 sect. Homoeotes, 179, 196, 197 sect. Hymenostachis, 196 subg. Hymenostachis, 197 subsect. Hymenostachis, 179, 197 sect. Lacostea, 179, 182, 195 subg. Lacostea, 195 subsect. Lacostea, 182, 189 sect. Lacosteopsis, 178, 181, 182 subsect. Lamellata, 180, 194 sect. Lecanium, 179, 181, 192 sect. Leptomanes, 185 subg. Macroglena, 187 subg. Mazora, 180, 181 sect. Maschalosorus, 197 sect. Microgonium, 179, 181, 191, 192, 201 subg. Microgonium, 192 sect. Microtrichomanes, 199 subg. Microtrichomanes, 199 subg. Minora, 180, 181 sect. Nesopteris, 179, 190 subg. Nesopteris, 190 sect. Neuromanes, 194, 195 Trichomanes—Continued sect. Neurophyllum, 179, 194, 195 sect. Odontomanes, 179, 195 sect. Pachychaetum, 179, 187, 188 subg. Pachychaetum, 178, 179 (key), 187, 188 sect. Phlebiophyllum, 178, 1783, 184 subg. Phlebiophyllum, 161 sect. Pinnata, 181 sect. Pleuromanes, 179, 185, 186 subg. Pleuromanes, 185 subg. Polyphlebium, 184 sect. Pseudachomanes, 198, 199 subg. Pseudachomanes, 198 sect. Ptilophyllum, 198 subg. Piilophyllum, 176, 198 group Pyxidifera, 201 sect. Ragatelus, 180, 197, 198 group Rigida, 201 group Scandentia, 201 subg. Schizophlebium, 189 sect. Selenodesmium, 188 subg. Selenodesmium, 188 subg. Serpyllopsis, 161, 181 group Taschneria, 183, 201 sect. Taschneria, 183 sect. Trichomanes, 178, 181, 182 subg. Trichomanes, 178 (key), 180, 181, 184, 186, 187 ser. T'richomanopsis, 182 sect. Trigonophyllum, 179, 196 subg. Vandenboschia, 182 Trichomanes abrotanifolium, 182 accedens, 194 acranthum, 190 acrosorum, 190 acutilobum, 183 acuto-obtusum, 183 acutum, 185, 196 adiantoides, 175 adscendens, 194 alagense, 187 alatum, 198, 199 album, 185, 186 amabile, 182 amazonicum, 196 anadromum, 194 anceps, 189 angustatum, 182 angustifrons, 191 angustimarginatum, 188 angustissimum, 187 ankersii, 195, 196 Trichomanes—Continued anomalum, 194 aphlebioides, 182 apiculare, 184 apiifolium, 190 arbuscula, 196 asa-grayi, 188 asplenioides, 164, 172, 190 assimile, 187 aswijkii, 200 atrovirens, 189, 190 attenuatum, 199 aureovestitum, 194 auriculatum, 182 axillare, 182 badium, 194 baldwinii, 190 ballardianum, 192 bancroftii, 196, 198 batrachoglossum, 188 bauerianum, 190 beckeri, 182 bicorne, 196 bilabiatum, 183 bilobatum, 183 bimarginatum, 192 bipunctatum, 183, 191 birmanicum, 182 blepharistomum, 190 bonapartei, 187 boninense, 183 boninicola, 187 boryanum, 190 boschianum, 182 botryoides, 196, 197 brachyblastos, 182 brachypus, 195 bradei, 182 brevipes, 183 brevisetum, 187 brooksii, 187 caespitosum, 161 calyculatum, 190 canariense, 175 capillaceum, 175-177, 182, 185 caudatum, 188 cellulosum, 182, 187 chamaedrys, 192 chinense, 175 christii, 183 clarenceanum, 183 clathratum, 182 cocos, 182 INDEX 211 Trichomanes—Continued cognatum, 198 colensoi, 182 collariatum, 182 compactum, 188 corcovadense, 194 craspedoneurum, 192 crassipilis, 194 crassum, 190 crinitum, 197, 198 crispiforme, 194 crispum, 175-177, 181, 189, 198, 194, 199 cristatum, 194 cultratum, 192 cumingii, 185 cupressoides, 188 curranii, 199 curtii, 191 cuspidatum, 191, 192 cyrtotheca, 182 cystoceiroides, 182 dactylites, 194 daguense, 194 davallioides, 182 debile, 182 delicatum, 194 densinervium, 190 dentatum, 188 diaphanum, 182 dichotomum, 199, 200 digitatum, 199, 200 dilatatum, 174 diversifrons, 197 draytonianum, 182 ekmanii, 192 elegans Rich., 189, 197 elegans Rudge, 197 elongatum, 188 endlicherianum, 184 eriophorum, 194 erosum, 192 exiguum, 191 exsectum, 182, 185 extravagans, 188 fallax, 182 falsivenulosum, 192 fargesii, 182 ferrugineum, 188 filicula, 183 fimbriatum, 199 firmulum, 187 flabellatum, 199, 200 212 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM Trichomanes—Continued Trichomanes—Continued flavofuscum, 188 latilabiatum, 187 foeniculaceum, 187 latisectum, 182 foersteri, 190 laxum, 194 francii, 200 ledermannii, 190 fulgens, 192 liberiense, 191 furcatum, 194 lindigii, 194 fuscum, 174 lineolatum, 191 galeottii, 194 liukiuense, 182 gardneri, 194 longicollum, 188 gemmatum, 187, 188 longifrons, 182 giesenhagenii, 191 lucens, 194 giganteum, 182 ludovicianum, 194 godmanii, 192 luschnathianum, 181, 187 goetzii, 182 lyallii, 200 gourlianum, 191 macilentum, 194 gracile, 187 madagascariense, 190 gracillimum, 184 majorae, 183 grande, 190 makinoi, 183 guianense, 194 maluense, 190 haughtii, 182 mannii, 187 henzaianum, 192 martiusi, 194 herzogii, 182 matthewii, 187 hirsutum, 175, 176 maximum, 182 holopterum, 194, 198 megistostomum, 183 hookeri, 192, 200, 201 meifolium, 187, 188 hostmannianum, 195 melanopus, 191 huberi, 195 melanotrichum, 182 humile, 184 membranaceum, 175, 177, 192 hymenoides, 191 micayense, 194 hymenophylloides, 182 mindorense, 192 hypnoides, 182 minutum, 186 imbricatum, 194 miyakei, 182 infundibulare, 190 motleyi, 192 ingae, 182 mougeotii, 197 insigne, 183 muscoides, 191 intermedium, 190 nanophyllum, 183 intramarginale, 181, 182, 183 naseanum, 182 javanicum, 189, 190 neesit, 165 johnstonense, 182 nipponicum, 182 junceum, 182 nitidulum, 199, 200 kalbreyeri, 194 novoguineense, 187 kapplerianum, 192 nummularium, 191 killipii, 194 nymanii, 183 kingii, 190 obscurum, 188 krausii, 191 obtusum, 188 kurzii, 183 omphalodes, 192 laciniosum, 195 opacum, 194 laetum, 188 orientale, 182 lambertianum, 194 ornatulum, 194 latealatum, 183 osmundoides, 196, 197 latemarginale, 183 ovale, 191 latifrons, 182, 186 pallidum, 185, 186 INDEX 213 Trichomanes—Continued Trichomanes—Continued palmatifidum, 199, 200 sellowianum, 194 papuanum, 192 serratifolium, 182 parviflorum, 188 setaceum, 188 parvulum, 186, 200 siamense, 188 parvum, 182 singaporianum, 190 pedicellatum, 195, 196 sinuatum, 182 pellucens, 194 sinuosum, 198, 199 pennatum, 195 smithii, 185 pervenulosum, 183 societense, 190 petersii, 191 somae, 182 philippianum, 182 speciosum, 187 piliferum, 200 spicatum, 196 pilosum, 194 spruceanum, 196 pinnatifidum, 199 stenosiphon, 182 pinnatinervium, 191 strictum, 188 pinnatum, 194, 195 stylosum, 188 plicatum, 183 subclathratum, 182 polyanthos, 190 sublabiatum, 194 polyanthum, 190 sublimbatum, 192 poly podioides, 175, 177, 198, 199 subtilissimum, 187 preslii, 190 suffrutex, 190 prieurit, 189 sumatranum, 190 procerum, 194 superbum, 190 proliferum, 186 taeniatum, 200 pseudoblepharistomum, 190 tagawanum, 183 pseudocapillatum, 183 tamarisciforme, 188 pseudonymanii, 183 tanaicum, 196 ptilodes, 198, 199 tenerum, 182, 185 punctatum, 191 tenuissimum, 182 pusillum, 191, 200, 201 tereticaulum, 188 pyramidale, 187 teysmannii, 187 pyxidiferum, 175-177, 182, 183 thysanostomum, 190 radicans, 181, 182 titibuense, 182 reniforme, 175 tosae, 183 reptans, 191 trichoideum, 185 retusum, 186 trichophyllum, 188 rhipidophyllum, 191 trigonum, 199 ridleyi, 200 trinerve, 187 rigidum, 187, 188 trollii, 197 robustum, 194 truncatum, 188 rothertii, 183 tuerckheimii, 196 rupestre, 181 tunbridgense, 164, 175, 176 rupicola, 183 ulei, 182 ruwenzoriense, 187 vaupelii, 185 samoense, 184 venosum, 184 saxifragoides, 186 . venulosum, 183 sayeri, 192 _ . - scandens, 175-177, 180-182 vieillardii, 185 schiedeanum, 185 virgatulum, 182 schlechteri, 188 vitiense, 200 schmidianum, 182 vittaria, 194, 195 schultzei, 188 walleri, 183 214 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM Trichomanes—Continued wallii, 191 warburgii, 188 werneri, 185 wildii, 182 Trichomanopsis, ser., 182 Trigonophyllum, sect., 179, 196 Vandenboschia, 176, 178, 182 radicans, 182 subclathrata, 182 titibuensis, 182 Vandenboschia, subg., 182 U.S. GOVERNMENT PRINTING OFFICE: 1968 BULLETIN OF THE UNITED STATES NATIONAL MUSEUM CONTRIBUTIONS FROM THE UNITED States NATIONAL HERBARIUM VotumE 38, Part 6 STUDIES OF FERN TYPES, II By C. V. Morton SMITHSONIAN INSTITUTION PRESS Washington, D.C. 1973 OFFICIAL PUBLICATION DATE is handstamped in a limited number of initial eopies and is recorded in the Institution’s annual report, Smithsonian Year. SI PRESS NUMBER 4803. STUDIES OF FERN TYPES, II C. V. Morton This paper is the second and final part of C. V. Morton’s general work on fern types. This work was accomplished in large part through grants from the John Guggenheim Memorial Foundation and the Na- tional Science Foundation (grants G-4080 and GB-1243). The first part, “Studies of Fern Types, I” (Contr. U.S. Nat. Herb. 38 :29-83, 31 Aug. 1967), included an introduction that provides background for this paper as well; readers may wish to refer to it. At the time of his death, in July 1972, Mr. Morton had all but com- pleted the present paper. The work is entirely his, except for this in- troduction and some slight editorial changes that I made in assembling the manuscript for publication.—D. B. Lellinger. 1. ACROSTICHUM NEGLECTUM F. M. Bailey, Trans. Linn. Soc. New South Wales 5:32. 1881. =Bolbitis neglecta (F. M. Bailey) Morton, comb. nov. Leptochilus neglectus (F. M. Bailey) C. Chr. Ind. Fil. 14. 1905. Type: Trinity Bay Ranges, Queensland, Australia, May, 1877, F.M. Bailey (iso- type BM, Morton photograph 7357). This species is very different in the cutting of the sterile blades from other Australian species of Bolbitis. Apparently it is endemic in Queensland, where it is rare. 3. ACROSTICHUM OBDUCTUM Kaulf. ex Spreng. in L. Syst. Veg. ed. 16, 4:34. 1827. =Elaphoglossum lancifolium (Desv.) Morton, Contr. U.S. Nat. Herb. 38 :32. 1967. Type: A renaming of Acrostichum lancifolium Desv., and thus based on the same type as that. It is likely that Kaulfuss intended his A. obductum to be based on Mauritius, Sieber, Syn. Fil. no. 25, and it has been perhaps generally so considered, although never definitely lectotypified. But when Sprengel published the name he did not mention Sieber or indeed cite any specimens, merely citing Acrostichum lancifolium Desv. as a synonym. Since the name A. lancifolium was a legitimate and avail- able name, there was no need to propose a different name, A. obductum. There is nothing in the original description to suggest that A. lanci- 215 216 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM folium was cited as a synonym with an implied query, and so by Art. 63 of the Code the name A. obductum must be considered superfluous, and by Art. 7 it must be typified on the basis of the type of the name that ought to have been adopted, thus on the type of A. lancifolium. For the plant identified as Llaphoglossum obductum, see under Acrostichum tomentosum Bory. 3. ACROSTICHUM TOMENTOSUM Bory ex Willd. in L. Sp. Pl. ed. 4, 5:101. 1810. =Elaphoglossum tomentosum (Bory) Christ, Farnkr. 37. 1897. Acrostichum heterolepis Fée. Mém. Foug. 2:56, t. 15, f. 1. 1845. Based on Bourbon [Réunion], Bory and Sieber, Fl. Mix. 281 (probably also from Réunion). I do not choose to lectotypify it at present. Elaphoglossum heterolepis (Fée) Moore, Ind. Fil. 10. 1857. Elaphoglossum obductum sensu auctt. (eg., Tardieu, Notul. Syst. 16:185. 1960), non Acrostichum obductum Kaulf. ex Spreng. See under A. obductum, Tyre: Bourbon Island [=Réunion], Bory. In the Willdenow Herbarium in Berlin there is a single sheet (no. 19508) named A. tomentosum, this received from Fliigge. I do not believe that Fliigge ever collected in Réunion, but he is known to have exchanged plants widely ; his specimen of A. tomentosum may well have come from Bory. The specimen in the Willdenow Herbarium is poor, just a detached frond, but it agrees well enough with Willdenow’s description. In Madame Tardiew’s treatment of Zlaphoglossum in Madagascar and the Mascarene Islands (Notul. Syst. 15 425-443. 1959), the species Acrostichum tomentosum Bory ex Willd. is not placed or mentioned except (p. 429) under £. obductum (Kaulf.) Moore, where a synonym is “A. tomentosum Bory in herb. (non Willd. Spec.).” If A. tomento- sum Bory ex Willd is different from obductum, then it must be the same as one of the other species treated by Tardieu, since it is from Réunion, and it would very likely be the earliest and thus the correct name for one of them. But this does not seem likely, for there is no a priori reason to think that the Bory specimens in Paris labeled A. tomento- sum are different from the species described by Willdenow. If they are indeed different, then the proper name for the plant called Z. ob- ductum by Madame Tardieu will be /’. heterolepis (Fée) Moore. 4, ADIANTUM CUNEATUM var. ANGUSTIFOLIUM Mart. & Gal. Mém. Acad. Sci. Brux. 15 :70. 1842 [as “angustifolia”], =Adiantum andicola Liebm. Dansk. Vid. Selsk. Skrift. V, 1 :266. 1849. Syntypres: Mexico, Galeotti 6266 and 6359. Lectotype: Mount Orizaba, Vera- cruz, Mexico, 9,500 feet, Galeotti 6266 (BR, Morton photograph 5062). Although Martens and Galeotti cite only two numbers they indicate four localities. Fournier (Mex. Pl. 128. 1872) did not mention var. angustifolium by name, but he cited Gu/eott: 6266 under A. multiforme A. Braun var. 8 and 6539 under A. amabile Liebm. Adiantum multiforme A. Braun ex Fourn. (Mex. Pl]. 128. 1872) is essentially without a description as a species, although the two varieties are distinguished as var. a, pinnis STUDIES OF FERN TYPES, II—MORTON 217 latioribus, inferioribus aliquando suborbicularibus, and var. 8, pinnis angustioribus. Under var. a there are several synonyms, all “sensu” names except A. ewtensum Fée. At the places cited by Fournier (Fée, Gen. Fil. 119. 1852 and Mém. Foug. 9 :6. 1857) this name is also a nomen nudum based on Schaffner 40 and 41, but the species was validly de- scribed in Mém. Foug. 8:72. 1857, based on Schaffner 40 and 41 from Orizaba and Huatusco, Veracruz. I do not choose one of these as lecto- type at the moment, since I have not seen either specimen; however, A. multiforme A. Braun can be considered as validly published, a sub- stitute name for A. extenswm Fée. As such it is a superfluous and con- sequently illegitimate name which must be typified on the same types as A. extensum, namely Schaffner 40 and 41. Under var. Fournier cited A. mewicanum Presl, Tent. Pterid. 158. 1836. From the entry in the “Index Filicum” it would appear that this is a validly published name, but apparently it is only a nomen nudum. Presl gave no de- scription, only referring to A. cuneatum Schlecht. (non Langds. & Fisch.) , by which he seemingly intended A. cuneatum sensu Schlecht. & Cham. Linnaea 5:615. 1830, but in Schlechtendal and Chamisso there is no word of description that might validate the name. Fournier cited also Mettenius, but I do not find any use of the name by Mettenius. 5. ADIANTUM TENERUM var. DISSECTUM Mart. & Gal. Mém, Acad. Sci. Brux. 15:71. 1842, —Adiantum andicola Liebm. Dansk. Vid. Selsk. Skrift. V, 1:266 (seors. 114). 1849. Type: Oaxaca, Mexico, “terre froide,” 8,000 ft. alt., Galeotti 6361 (holotype BR, Morton photograph 5063). 6. ASPIDUM BRACHIATUM Zoll. & Moritzi. Natuur Geneesk. Arch. Neerl. Indié 1:399. 1844. =Tectaria brachiata (Zoll. & Moritzi) Morton, comb. nov. Aspidium zollingerianum Kunze. Bot. Zeit. 4:462, 1846. Syntypes: Java, Zollinger 655, 655A. The specimens of these numbers in Geneva (Morton photographs 3803 and 3804 respectively ) have the names in Kunze’s hand, and are undoubtedly the syntypes. No. 655 is here chosen lectotype, since it is the larger and more mature specimen. The labels of the Geneva speci- mens lack specific locality data, and consequently Kunze did not cite this. A duplicate of no. 655 in Leiden gives the locality as on calcareous rocks on stream banks near Tjikoya, Java, i.e., the same locality as A. brachiatum. Aspidium variolosum Wall. ex Hook. Sp. Fil. 4:51. 1862. Syntypes: Hooker cites seven syntypes, but since the name is adopted from Wallich, Num. List no. 379. 1829 (nom. nud.) and is attributed to Wallich, one of the collections under Wallich Cat. 379 must be the lectotype. Hooker, however, cited three localities for no. 379, namely Amherst, Tavoy, and Penang. Ching (Sinensia 2:21. 1931) indicated Wallich 379 as “type” (i.e., lecto- type) and illustrated it in his Plate III, presumably from a specimen in the Kew Herbarium. This plate ought to be matched up with the Kew specimens to determine the type locality. Tectaria variolosa (Wall. ex Hook.) C. Chr., Contr. U. 8. Nat. Herb. 26 :289. 1931. 218 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM Type: “In caleareo-argillosis ad ripas rivulorum prope Tjikoya prov. Batav., raro,’ Java, Zollinger. These data correspond to Zollinger 655 and 655A, the same numbers that A. zollingerianum Kunze was based on. Since the publication of Zollinger and Moritzi is rare, at least out- side of Dutch libraries, I give below the original description: Fronde heteromorpha membranacea glabriuscula margine ciliolata ternata, foliolis lateralibus bipartitis cordatis, laciniis acuminatis, medio 3-partito, laciniis lateralibus sessilibus, mediali elongato-acuminata pinnatifida ; frondibus fertilibus longe stipitatis frondibus sterilibus similibus sed omnibus partibus elongato- angustatis, soris majusculis subseriatis, stipite glabriusculo. Filis vix pedalis. Folioli lateralis pars superior inferiore longior subpinnatifida ; folioli medii laciniae laterales subfalcatae. Venulae subtus hirsutulae. Stipes frondis sterilis minus elatus; fertilis folioli breviter petiolati. Sori venularum ramulos superiores terminantes. It is too bad that the rather well-known name 7ectaria variolosa must be abandoned, but there seems no help for it. In publishing Aspidium variolosum for Wallich, Hooker completely overlooked the earlier species A. brachiatum and A, zollingerianum. Moore, who was adept in picking up overlooked names, listed A. bra- chiatum as a doubtful species, with A. zollingerianum as a synonym. Christensen in the “Index Filicum” also recognized brachiatum as a valid species. Ching in his “A Revision of the Genus Tectaria from China and Sikkime-Himalaya” (Sinensia 2:9-36. 1931) adopted 7. variolosa (Wall. ex Hook.) C. Chr. and reduced both A. brachiatum and A. zollingerianum to synonymy, even though these were earlier names. This is perhaps explained by the fact that he attributes (twice on page 21) the date “1826” to volume 4 of Hooker’s “Species Filicum,” whereas the correct date is really 1862. Holttum, in his “Ferns of Ma- laya” (p. 506. 1954), recognized 7. variolosa, and commented on its occurrence in Java, but also overlooked the earlier names brachiatum and zollingerianum. 7. ASPIDIUM IMMERSUM Blume, Enum. Pl. Jav. 2:156. 1828. =Thelypteris immersa (Blume) Morton, comb, nov. Lastrea immersa Moore, Ind. Fil. LXX XIX. 1857, Nephrodium immersum Hook. Sp. Fil. 4112, 1862. Dryopteris immersa Kuntze, Rev. Gen. Pl. 2 :813. 1891. Dryopteris besukiensis f. laviloba van Alderw. van Rosenb. Bull. Jard. Bot. Buit. II, 1:8. 1911. Type: Besoeki Idjen, Java, Koorders 15436 B (isotype or syntype L, Morton photograph 1064). Dryopteris besukiensis f. confertiloba van Alderw. van Rosenb. loc. cit. Type: Besoeki Idjen, Java, Koorders 19830 B (isotype or syntype L, Morton photo- graph 1064). Type: Gaenaeng Parang, Java, Blume (holotype L, Morton photograph 1164). This species has been reported also from Borneo, New Guinea, and the Philippine Islands, but its real range remains to be determined. My STUDIES OF FERN TYPES, II—MORTON 219 notes on the type are: Veins oblique, bearing yellow glands beneath and no long hairs; indusia flat, glabrous, persistent. 8. ASPIDIUM MEXICANUM var. SERRATUM Mett. Abhandl. Senckenb. Naturf. Gesell. 2:349. 1858. =Dryopteris patula var. serrata (Mett.) Morton, comb. nov. Aspidium chaerophylloides Moritz ex Mett. loc. cit. pro syn. Dryopteris patula var. chaerophylloides C. Chr. Dansk. Vid. Selsk. Skrift. VII, 10:71. 1913. Typr: Caracas, Venezuela, Moritz 435. The holotype is presumably in the Met- tenius Herbarium in Berlin. I have seen a specimen labeled Aspidium chaerophyl- loides Moritz but bearing the number Moritz 434 (Leiden, Morton photographs 1087, 1088). It seems likely that this is an isotype, and that it bears the wrong number “434” or that Mettenius erred in citing the number originally as “435.” Christensen in describing Dryopteris patula var. chaerophylloides credited the epithet to “Bak. Syn. 276,” but this is an error. Baker never used the generic name Dryopteris; he had this plant as “Nephrodium mexicanum B A. chaerophylloides Moritz” (Hook. & Bak. Syn. Fil. 976. 1868), an illegitimate combination, using a binomial for a variety. Moreover, var. serratum has a clear priority of ten years. The taxonomic validity of this eglandular variety needs further investiga- tion, as do the forms of this variable species generally. 9. ASPIDIUM ULIGINosUM Kunze, Linnaea 20:6. 1847. =—Thelypteris torresiana (Gaud.) Alston, Lilloa 30 :111. 1960. LEcToTyPe: Described from cultivated specimens in the botanical garden in Leipzig, which Kunze had received from the botanical garden in Bonn, Since Kunze’s herbarium in Leipzig was destroyed, there is no holotype extant. So far as I know, no lectotype has been designated. There is a specimen (Morton photo- graph 19739) in the Jardin Botanique National de Belgique, Brussels, from the botanical garden in Leipzig with the notation that the plant was originally from Java; this specimen is identified in Kunze’s own hand as “Aspidium uliginosum mihi,” and is thus fully authentic; I therefore designate it as lectotype. The above lectotype is quite the same as the species treated by Holt- tum (Blumea 17:27. 1969) as Macrothelypteris torresiana (Gaud.) Ching. 10. Asplenium adiantum-nigrum var. capense Schlecht. Adumbratio 31, 32, t. 17. 1825-32. Sometimes, as in Kuhn’s “Filices Africanae,” a “var. capense Schlecht.” is cited as though validly published, and the name appears also on some herbarium labels. Schlechtendal mentioned that As- plenium adiantum-nigrum occurs in two varieties, one with acute seg- ments in the Canary Islands, Italy, and Porto Rico, the other with obtuse segments in the Cape of Good Hope and Germany. He did not actually propose varietal names for these, however, nor did he indi- 220 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM cate which, if either, included the type of A. adiantum-nigrum L. The name “capense” occurring only on the plate and the explanation of the plate merely indicates material drawn from plants collected in the Cape region and is not a formal naming of a variety. 11. ASPLENIUM CAUDATUM var. SEcTUM Hillebr. Fl. Haw, 603. 1888. = Asplenium polyodon var. sectum (Hillebr.) Morton, comb. noy. Asplenium sectum (Hillebr.) Copel. Phil. Journ. Sci. 9 :439, 1914. Asplenium faleatum var. sectum (Hillebr.) Skottsb. Acta. Hort. Gotob. 15 :101. 1942. SyNtTypes : Haleakala, Maui, Hawaii, Hillebrand and Oahu, Hawaii, Hillebrand. The former collection from Haleakala, right-hand frond, is here designated lecto- type (B, Morton photograph 9666). This is a form with long-attenuate pinnae, these prominently lobed at the base only or nearly throughout. Skottsberg recognized three or four forms. 12. Asplenium decipiens Zippel ex Kunze, Bot. Zeit, 6:193. 1848, in adnot. TYPE: Java, Kollmann (not seen). I am calling attention to this species because the name is overlooked in the “Index Filicum” and its supplements. Although mentioned in a note only, it is accepted as new by Kunze and given sufficient descrip- tion to validate it, namely that the caudex is flexuous and glabrous and the blade irregularly incised-dentate. Kunze indicates that this might be a Diplazium, but this does not of course invalidate the name, which is definitely indicated as Asplenium decipiens Zippel. The matter is of some importance because this name antedates Asplenium decipiens Mett. (1859) and A. decipiens Kuhn (1879). Backer and Posthumus (Varenfl. Java 126. 1939) place A. decipiens Zippel as a synonym of Diplazium subserratum (Blume) Moore. 13. ASPLENIUM DENTICULATUM Blume, Enum. Pl. Jay. 186. 1828. —Asplenium aethiopicum (Burm. f.) Becherer, Candollea 6:23. 1935, forma. Type: Java, Blume (holotype L, Morton photograph 527). The type is mounted on a sheet with four other Blume types. All the labels, in Blume’s own hand, are grouped on the right-hand margin of the sheet, and there is no indication as to which label goes with which plant. The other types are Asplenium tripartitum Blume, A. furcatum var. fissulum Blume, A. furcatum var. fragrans Blume, and A. furcatum var. depauperatum Blume, I have compared these plants with Blume’s descriptions, and I was able to decide on the various types with some degree of certainty. The large central plant is A. furcatum var. fissulum, The plant at the lower left is A. tripartitum; the plant at the upper right is A. furcatum var. fragans, The small fragment at the bottom center is A, furcatum var. depaupera- tum, And the plant at the lower right is the type of A. denticulatum. In the “Index Filicum,” Asplenium denticulatum Blume is listed in italics as a dubious species. In Backer and Posthumus’ “Varenflora STUDIES OF FERN TYPES, II—MORTON 221 voor Java” (1939), it is placed as possibly the same as Asplenium stereophyllum Kunze (Bot. Zeit. 6: 175. 1848), in which case it would be the prior and correct name. I have seen the type of A. stereophyllum (Java, Zollinger 2249, holotype G, Morton photograph 3811; isotype FI-Webb!). The Zollinger collections from Java that Kunze described were evidently lent to Kunze from Geneva, for they have the annota- tions in Kunze’s own hand and the data as published by Kunze; they are thus indeed holotypes, which is fortunate since Kunze’s own her- barium was destroyed during World War II. This holotype shows that A. stereophyllum has been correctly interpreted in the naming of recent collections. The pinnae are strongly unequal-sided, the lower side being cut away to about the middle of the pinnae and so it has fewer segments than the upper side. Asplencum denticulatum Blume is by no means this species but is only a form of the widespread, pantro- pic A. aethiopicum, not unlike some of the American forms that are called A. praemorsum Swartz. In fact, all of the types on this sheet (A. tripartitum and the varieties of A. furcatum) are referable to A. aethiopicum in a broad sense, which is evidently just as variable in Java as it is elsewhere. 14. ASPLENIUM DISTANS Fée, Gen. Fil. 198. 1852, non D. Don, 1825. == Diplazium camptocarpon Fée, Mém., Foug. 8:84. 1857. Type: Oaxaca, Mexico, Galeotti 6529 (isotype P, Morton photograph 4165). Fée wrongly gave the number as 6579, which was corrected by Fournier to 6529 The holotype, with the name in Fée’s hand, has not been found, but the isotype seen agrees with the description and is surely authentic. In the “Index Filicum,” Asplenium distans Fée is considered a dubious species of Asplenium,; however, the isotype shows that it is not an Asplenium, but a large, bipinnate (almost subtripinnate at base) Diplazium. It is very likely the same as D. camptocarpon Fée, the type of which is from Cérdoba, Veracruz, Mexico, A. Nieto (Schaffner 69, not seen), a species that is similar at least to the more divided forms of D. franconis Liebm. 15. ASPLENIUM FALCATUM Var. SUBCAUDATUM Skottsb. Acta Hort. Gotob. 15 :100. 1942, —Asplenium polyodon var. subcaudatum (Skottsb.) Morton, comb. nov. Type: Nuuanu-Kalihi Ridge, Koolau Mountains, Oahu, Hawaii, Skottsberg 146 (not seen). This variety includes the plants referred to A. caudatum Forst. in Hillebrand’s “Flora of the Hawaiian Islands.” Skottsberg, after an exhaustive study of the variation in Asplenium falcatum Lam. and ite allies, excluded A. caudatum from the Hawaiian Islands, most of the variations being proposed as varieties of A. falcatwm. As I indicated 499-018—-73——2 222 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM in my earlier paper on fern types (Contr. U.S. Nat. Herb. 38 :39-41. 1967), Asplenium falcatum Lam. was a superfluous and illegitimate name. The earliest correct name for this species is A. polyodon Forst., and consequently the various Hawaiian varieties are here transferred to that species (see also A. knudsenii, A. nitidulum, and A. caudatum var. sectum). 16. ASPLENIUM GRACILE. D. Don, Prodr. Fl. Nepal. 8. 1825. =Athyrium setiferum ©. Chr. Ind. Fil. 146. 1905. Based on Asplenium tenellum Hope, non Roxb." Allantodia tenella Wallich, mss. Based on Nepal, Wallich. Asplenium tenellum Hope, Journ. Bombay Nat. Hist. Soc. 22 :529, t. 4. 1899. non Roxb. 1816, nee Fée, 1852. Lectotype: Cori Valley, above Bugdiar, Kumaon, India, 9,000 feet alt, R. W. MacLeon in 1893 (BM, Morton photograph 6980). Type: “Hab. in summis alpibus reg. Himalayae dict. Nepaliae,” Wallich (holo- type BM, Morton photograph 6986, four right-hand plants; the left-hand plant on this sheet is also Nepal, Wallich, but was probably added later after the time of Don). Asplenium gracile D. Don is considered in the “Index Filicum” as possibly the same as Athyrium nigripes (Blume) Moore (type from Mount Burangrang, Java, Blume, L, Morton photograph 665) ; there is some resemblance, but I do not believe that they are quite the same. On the other hand, it does seem clear that A. gracile is the same as Athyrium setiferum C. Chr., which is of course a later name. There is already an Athyrium gracile Fourn. (1872), however, and so Don’s epithet may not now be transferred to Athyrium. Hope did not desig- nate a type and cited many syntypes. He evidently chose the epithet “tenellum” from the cited synonym Allantodia tenella Wallich, which is a manuscript name only, not even appearing in Wallich’s “Numerical List.” It would thus seem that a Wallich specimen ought to be the lecto- type, but Hope specifically indicated that the Wallich specimen repre- sented a variant, thus making it ineligible as a lectotype. The specimen chosen above as lectotype seems to be a normal and typical specimen of Hope’s concept. I have seen several other syntypes also, not all of which seem identical; in particular, the Mackinnon specimen from Sowarna Nala, Dehra Dun District, is larger and more divided and may not belong here at all, at least so far as the left-hand plant on the sheet in the British Museum (Natural History) is concerned ; the speci- men on the right may be rather typical (Morton photograph 6982). These Athyriums of the filix-femina group in the Himalayas are ex- tremely perplexing and much confused in herbaria. There are probably too many names proposed for them. * Christensen wrongly cites Allantodia tenella Wallich as the basionym, but this was a manuscript name only and so cannot be a basionym. STUDIES OF FERN TYPES, II—MORTON 223 17. ASPLENIUM GRACILESCENS Mett. Ann, Sci. Nat. V, 2:287. 1864. =Diplazium gracilescens (Mett.) C. Chr. Ind. Fil. 233. 1905. SyntryPes: Ocafia, Colombia, Hngel 250 (isotype BM, Morton photograph 70734), Schlim 69 (isotype L, Morton photograph 931). Since I have not seen the original syntypes in Berlin, I do not choose a lectotype. In the “Index Filicum,” the authority for the name D. gracilescens is given as “Moore, Ind. Fil. 329. 1861,” but the name was a nomen nudum at this place. Since the species was not described until 1864, a new combination of the epithet under Diplazium could not be validly published three years previously in 1861. Therefore, Christensen must be cited as the author. 18, Asplenium gueinzianum Mett. ex Kuhn, Fil. Afr. 102. 1868. Asplenium caespitosum Wallich, Num. List. no. 217. 1829, nom. nud., non Blume, 1828. Asplenium laciniatum sensu Hook. Sp. Fil. 3:164. 1860, non D. Don. Type: Port Natal, South Africa, Gueinzius (holotype B, Morton photograph 9714). Hieronymus stated (Hedwigia 61 :34. 1919) that this South African species was not distinct from the Himalayan plant that he called A. laciniatum, and he was probably right, for the holotype of A. guein- sianum does look indistinguishable from Wallich 217, the basis of the unpublished A. caespitoswm Wallich (non Blume). 19. ASPLENIUM HAHNI Fourn. Mex. Pl. 109. 1872. =Diplazium camptocarpon Fée, Mém. Foug. 8 :24. 1857. Typr: Escamela, near Orizaba, Veracruz, Mexico, Aug. 2, 1865, Hahn (holo- type P, Morton photograph 4087). Diplazium hahnii (Fourn.) C. Chr. Ind. Fil. 233. 1905. The type is very similar to that of A. distans Fée (non D. Don); it seems to agree with the description of Diplaziwm camptocarpon Fée. 90, ASPLENIUM JAPONICUM var. CHATTAGRAMMICUM Clarke, Trans. Linn. Soc. IT, Bot. 1:499. 1880. =Diplazium chattagrammicum (Clarke) Ching, Lingnan Sci. Journ. 15 :277. 1936. Type: Clarke cited merely Kasalong, Chittagong, India. Ching indicated the type to be at Kew, but did not designate a specimen. There are four specimens of Clarke’s collecting from Kasalong, all much alike and agreeing with Clarke’s description, but none of them is named var. chattagrammicum, Three are named merely Asplenium japonicum, and one Asplenium japonicum var. elongata (an unpublished name). Since the latter specimen agrees with the original descrip- tion, it seems that Clarke first intended to call this plant var. elongata and changed it ultimately to var. chattagrammicum. Therefore, I take this sheet, the first collected, to be the lectotype of var. chattagrammicum: Kasalong, Chitta- gong, Jan. 10, 1869, Clarke 8254 (K, Morton photograph 18917). The other speci- mens that I believe to represent var. chattagrammicum also are from the same locality, but with later dates—Clarke 19074D, 19740A (this sheet bearing an annotation slip by Ching), and 19818. 294 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM Diplazium japonicum has recently been segregated as a distinct genus, Athyriopsis Ching (Acta Phytotax. Sinica 9:63. 1964), and as Lunathyrium Koidz. sect. Athyriopsis (Ching) Ohba (Sci. Rep. Yoko- suka City Mus. 11:52. 1965). It is said to differ from Diplaztum in having the lateral ridges of the rhachis above continuous and not open at the junction of the pinnae, and the prothallia with papillae on the margin and on the surface near the growing point. Rhachis characters are important, but especially for bipinnate or tripinnate plants. Simply pinnate-pinnatifid plants like D. japonicum do not show very different rhachis characters. Characters of the prothallia may indeed offer sub- stantiating characters in segregating groups, but they are not useful as key characters, and in any case only a relatively few of the species of Diplazium have been studied from this viewpoint. The generic status needs to be investigated further. 21, ASPLENIUM KNUDSENII Hillebr. Fl. Haw. 601. 1888, —Asplenium polyodon var. knudsenii (Hillebr.) Morton, comb. nov. Type: Waimea, Kauai, Hawaii, Knudsen. A specimen bearing this name and data is not in the Hillebrand Herbarium in Berlin; however, there are two specimens from Kauai in the Hillebrand numbered 125 and 144 and named A. knudsenii. These may well be and probably are the types. The left-hand frond numbered 125 is here designated lectotype (Morton photograph 9677). This variety represents the least cut form of A. polyodon found in the Hawaiian Islands. It is finely doubly serrate, but not at all lobed. 22. Asplenium laciniatum D. Don, Prodr. Fl. Nepal. 8. 1825. Asplenium varians Hook. & Grey. Icon. Fil. 2: t. 178. 1830. Type: Nepal, common name “Dawecow,” Wallich (holotype K or possibly E, not seen). Tyre: “In alpibus,” Nepal, Wallich (holotype BM, Morton photograph 6686). It is indeed unfortunate that the application of the name A. laciné- atum D. Don must be changed, but I see no alternative. The name has been applied, following Hooker, to a rather characteristic plant of Nepal and the Himalayas (and elsewhere) that is allied to A. plani- caule Wallich, but Hooker did not explain how he determined the identity of Don’s species. There is in the British Museum (Natural History) a sheet containing a number of plants, indicated as “1,” “.” and “3.” Those at the top of the sheet (and the one in the right bottom corner) were collected on the banks of the Rapty River by Wallich in 1820; these are not types. The three plants numbered “2” are on the left center and are from Nepal collected by Wallich in December, 1817, and the common name in Nepal is given as “Dawecow”’; these data are the same as those given by Hooker and Greville for their Asplenium varians except that they give the date as January, 1818; the specimens agree with Hooker and Greville’s figure. The three specimens at right STUDIES OF FERN TYPES, II—MORTON 225 center and the one center bottom are numbered “3,” and they quite agree with each other; these are labeled, I think in Don’s hand, “laciniatum” and “in Nepalia alphibus, D. Don Prod.”; these specimens are indeed clearly the holotype, for they not only have Don’s name and data but agree with Don’s description, so far as that goes. Don described his plants as only three to four inches long and the blades thin (“tenerae”), characters that clearly agree with the plants on this sheet labeled Jaciniatwm, for these are small plants of thin texture. The plant described by Hooker as Zaciniatum is much larger, the stipes alone being stated by Hooker to be four to six inches, and the blades a span to a foot long, i.e., about nine to twelve inches; the texture is thick, and the species itself belongs to a different section of Asplenium, viz. sect. Sphenopteris. On the other hand, the true A. laciniatum is identical with the plants described as A. varians Hook. & Grev., and they probably came from the same Wallich gathering in either December, 1817, or January, 1818. Unfortunately, Hieronymus (Hedwigia 61:33. 1919), in discussing the differences between A. laciniatum D. Don and A. planicaule Wallich, followed Hooker’s concept, apparently without any doubt about the matter, but he must not have looked carefully at Don’s description or tried to locate Don’s type in the British Museum (Natural History). The plant that has been wrongly called A. laciniatum is best called, at least tentatively, Asplenium gueinzianum Mett. (see entry No. 18). 23, Asplenium laserpitiifolium sensu D. Don, Prodr. Fl. Nepal. 9. 1825, non Lam., 1786. AUTHENTIC MATERIAL: Chitlong, Nepal, April 11, 1802, Buchanan-Hamilton (BM, Morton photograph 6678). This sheet has been marked by someone (Gepp?) as the type of “Asplenium laserpitiifolium (Hamilton) Don,” and the label does say “Asplenium laserpitiifolium B.,” the “B” evidently standing for “Buchanan.” When Don published the name in his “Prodromus,” however, he attributed the name to Lamarck, and so there is no “A. laserpitiifolium Buch.-Ham.” even though this name appears in the “Index Filicum” and this sheet is marked as the type. This authentic specimen was first identified as A. bulbéferwm Forst. f., to which it was attributed in the “Index Filicum,” but it can hardly be that, although it does resemble it from my photograph, since the true A. bulbiferum grows in New Zealand and perhaps in New Caledonia, but nowhere in Asia. Someone has crossed out bulbiferum and written bullatum Wallich, which possibly is correct, although I doubt it. The matter is not of importance, however, since this is a “sensu” name only, a misidentification, and it does not have to be placed 226 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM at all, although it should be identified correctly in a definitive treat- ment of the ferns of Nepal. 24, ASPLENIUM LONGIFOLIUM D. Don, Prodr. Fl. Nepal. 7. 1825, non Sehrad., 1824. =Diplazium longifolium Moore, Ind. Fil. 151. July, 1859. Asplenium lobulosum Wallich, Num. List. no. 210, 1829, nom. nud.; Mett. Abhandl. Senckenb. Naturf. Gessell. 3 :207. Sept. 1859. Type: Nepal, 1821, Wallich (isotype BR, Morton photograph 21170, left-hand plant). Diplazium lobulosum (Wallich ex Mett.) C, Chr. Ind. Fil. 234, 1905 (wrongly attributed to Presl). Typre: Based on Nepal, Wallich. A specimen that is doubtless an isotype is in Brussels (Morton photograph 21170, right-hand plant); this was received from the Lambert Herbarium (where Don was curator) in 1826. The name Diplazium lobuloswm has been generally adopted since the publication of the “Index Filicum,” since the first name applied to the species, Asplenium longifolkium D. Don, was an illegitimate later homonym. By the Code (Art. 72, Note), however, Moore’s name Diplazium longifolium must be considered as a new name for A. longifolium D. Don (1825), non Schrader (1824), rather than as a transfer. Since Moore’s name dates from July, 1859, it is earlier than the synonymous name Asplenium lobulosum Wallich ex Mettenius, which according to Stearn was published in September, 1859. The epithet longifolium thus has a priority of two months. 25. ASPLENIUM NITIDULUM Hillebr. Fl. Haw. 601. 1888. =Asplenium polydon var, nitidulum (Hillebr.) Morton, comb. nov. Asplenium falcatum var. nitidulum (Hillebr.) Skottsb. Acta Hort. Gotob. 15 :101. 1942, SYNTYPES: Kauapali, Maui, Hawaii, Hillebrand (B, Morton photograph 9676) ; hills near Hilo, Hawaii, Hillebrand. The first of these is the lectotype. This is a form that is just a little more cut than var. knudsenii, the pinne being short and broad, mostly with an evident superior basal lobe. 26. Asplenium normale D. Don, Prodr. Fl. Nepal. 7. 1825. Asplenium unilaterale Buch.-Ham. mss, ex D. Don, loc. cit. in syn, Asplenium multijugum Wallich, Num. List. no. 207. 1829, nom, nud. Asplenium multijugum Wallich ex Mett. Abhandl. Senckenb. Naturf. Gesell. 5 :179, 1859. Syntypes : India orientalis, Wallich 207 (dupl. US) and Ceylon, Gardner 25. The Wallich specimen is here designated lectotype. Type: Narainhetty, Nepal, Buchanan-Hamilton (holotype BM, Morton photo- graph 6675) The type is a good specimen, complete with rhizome, representing Asplenium normale in its usual sense. Asplenium multijugum Wallich is identical; Wallich ignored the name A. normale D. Don. STUDIES OF FERN TYPES, II—MORTON 227 27, ASPLENIUM OBTUSILOBUM Desy. Mag. Naturf. Freund. Berlin 5:323. 1811. —Asplenium affine Swartz. Journ. Bot. Schrad. 1800(2) : 56. 1802. Type: Bourbon Island [Réunion], collector unknown (holotype P, Herb. Des- yaux, Morton photograph 4161). In the “Index Filicum,” Asplenium obtusilobum is referred to the West Indian A. cuneatum Lam. (holotype P, Herb. Jessieu Cat. 1255, Morton photograph 3033, the locality and collector unknown). This disposition was accepted by Weatherby in his paper on the Desvaux fern types (Contr. Gray Herb. 114:21. 1936), probably on the basis that someone has written on the type sheet “Habitat in America calidi- ore,” although pointing out that it differs somewhat from typical A. cuneatum. Actually the type does not match any American material, but it does match specimens from Réunion and Mauritius of A. affine Swartz (type not stated in original publication, but given later by Swartz, Syn. Fil. 84, 279. 1806, as Mauritius, Groendal; holotype S-PA, Herb. Swartz, Morton photographs 5770, 5771). Therefore, Desvaux’ original ascription of the species to Bourbon was correct and the locality of tropical America on the type sheet is wrong. Asplenium affine is similar to A. cuneatum, but obviously it is different in having the ultimate segments trapezoidal and bluntly toothed or lobed, whereas those of A. cwneatum are obovate and sharply toothed. The Old World specimens, mostly from the South Pacific, that have been identified as A. cuneatum are probably referable to A. affine or other allied species. 298. ASPLENIUM PLANICAULE Wallich ex Mett. Abhandl. Senckenb. Naturf. Gesell. 3:201. 1859, non Lowe, 1858. =Asplenium yoshinagae var. planicaule Morton, nom. nov. Asplenium planicaule Wallich, Num. List. 8, no. 189. 1829, nom. nud. Asplenium truncatum D. Don ex K. B. Presl, Tent. Pterid. 107. 1836, nom. nud. A new name for A. planicaule Wallich, nom. nud. Authentic material : Nepal, Wallick (BM, Morton photograph 6670; B, Morton photograph 9737). Tarachia truncata K. B. Presl, Epim. Bot. 78, 1849 [1851], nom. nud. Asplenium indicum Sledge, Bull. Brit. Mus. (Nat. Hist.) Bot. 3:264. 1965. A new name for Asplenium planicaule Wallich ex Mett., non Lowe. Type: Based on A. planicaule Wallich ex Mett., non Lowe. The syntypes are Himalaya Mountains, Wallich no. 189 and Fielding. The name planicaule origi- nated with Wallich and so the Wallich specimen no. 189 in Berlin ex Herb. Met- tenius is here designated lectotype (Morton photograph 9739, a sheet with a detached frond at the left, and two fronds with a rhizome at the right, and a label saying “Typus”; there are other sheets of Wallich 189 at Berlin, all of them also with the label “Typus”’). Sledge noted that the well-known name Asplenium planicaule Wal- lich ex Mett. (1859) was a later homonym of A. planicaule Lowe (1858) and so proposed the new name A. indicum, remarking that a 228 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM variety grew in Japan. He overlooked the fact that the Japanese vari- ety had been originally proposed as a species, Asplenium yoshinagae Makino (1900). So long as these are considered as only varietally dif- ferent, which appears to be best, the name A. yoshinagae has priority as a species name for the combined species, and Sledge’s name A. indi- cum is unnecessary. The Japanese plant has been called A. planicaule var. yoshinagae (Makino) Tagawa (Acta Phytotax. Geobot. 14:95. 1951), but now this variety must become the species and the species the variety. Even though A. planicaule Wallich ex Mett. is not correct as a species name, the epithet planicaule is available as a varietal name and is therefore adopted here. Sledge cited as a synonym Asplenium falca- tum var. abbreviatum Kunze (Linnea 24 :260. 1851), and so it might appear that abbreviatum is the earliest varietal epithet, but an exami- nation of Kunze’s publication shows that Sledge misinterpreted Kunze’s intention, for Kunze clearly indicated that his names at this place were forms. Since epithets have priority only within their cate- gory, it is not necessary to raise Kunze’s forma abbreviatum to varietal status, and I do not do so because I am not absolutely convinced that the name really refers to this plant. Kunze’s original type specimens were destroyed in Leipzig and so his plant cannot really be definitely identified ; the f. abbreviatum was based on plants from the Niligiri Mountains (Schmid-Koch 49 and 182; duplicates may exist elsewhere). 29, ASPLENUM RIEDELIANUM Bongard ex Kuhn, Linnaea 86:102. 1869. =—Di- plazium riedelianum (Bongard ex Kuhn) Kuhn ex C. Chr. Ind. Fil. 238. 1905. TYPE: Brazil, Riedel (holotype B, Morton photograph 10449). The authority for Diplazium riedelianum is usually stated as Kuhn, Linnaea 36 :102. 1869, but at this place the name appears as a synonym only. I cannot find that anyone properly made a combination under Diplazium until Christensen in 1905. 30. Asplenium tenuifolium D. Don, Prodr. Fl. Nepal. 8. 1825. Asplenium concinnum Wallich, Num. List No, 216. 1829, nom. nud. Based on Nepal, Wallich, doubtless on a part of the same collection as the type of A, tenuifolium D. Don. Type: Nepal, Wallich. One would expect a holotype in the British Museum (Natural History), but I was unable to locate one there; a specimen labeled Asplenium tenuifolium D. Don is in Brussels, received from Lambert (the em- ployer of D. Don), which is surely an isotype at least and may even be a holo- type; it is here designated lectotype until such time as a true holotype may be found in the British Museum (Natural History) or elsewhere (Morton photograph 20949). Isolectotype: Nepal, Wallich 216 (US). Asplenium tenuifolium D. Don is a common species of the Himalayas that has always been correctly understood. So far as I know, there are no validly published synonyms. STUDIES OF FERN TYPES, II—MORTON 229 21. ASPLENIUM UNILOBUM Poir. in Lam. Encycl. Méth, Suppl. 2 :505. 1810. =Diplazium unilobum (Poir.) Hieron. Hedwigia 59 :332. 1917. Asplenium semihastatum Kunze ex Mett. Abhandl. Senckenb. Naturforsch. Gesell. 3:206. 1859. Type: Monte Libano, Oriente, Cuba, Linden 1921 (holo- type probably B; isotype FI!). Typs: Santo Domingo (West Indies), Nectoua, FI (holotype, Morton photo- graph 16116). The type and type locality of this species has been in doubt. There is no specimen in Paris labeled by Poiret that could be the type. Poiret did not cite a specimen or locality, merely “A merique meridionale,” a term applied vaguely by Poiret, Desvaux, and other early workers to almost any part of tropical America. There is a specimen in the Jus- sieu Herbarium (Cat. no. 1235) under the name Asplenium erosum L. that I formerly thought might be the type (Morton photograph 3017), but this now seems unlikely, for the specimen is from Peru, collected by Joseph de Jussieu, and the species that has been known as S. uni- lobum is known only from the Greater Antilles. In Florence I found a specimen labeled Asplenium unilobum Poir. nov. sp., from the Des- fontaines Herbarium. Poiret described some new species from the Desfontaines Herbarium (e.g., Polypodium elegans) and this is an- other one, very evidently the holotype. It is from Santo Domingo, collected by Nectoux. The specimen agrees exactly with Zinden 1921 in Florence, an isotype of Aspleniwm semihastatum Kunze, which agrees with the opinion expressed by Hieronymus. 32. ATHYRIUM FOSBERGII Copel. Occas. Papers Bern. P. Bishop Mus. 14:60, t. 13. 1938. =Diplazium fosbergii (Copel.) Morton, comb. nov. Type: Nitiperu, Rapa Island, July 18, 1934, Fosberg 11577 (isotype US). This rare species is apparently a Rapa Island endemic. Copeland considered Diplazium as a synonym of Athyrium, but pteridologists are now generally agreed that the two may be distinguished, although a few species are a little hard to place. One of these is our eastern United States narrow-leaved spleenwort, Athyrium pycnocarpon, which has been considered a Diplazium by some authors. Since A. pycnocarpon has n= 40 (as always in Athyrium), it isan Athyrium that is somewhat aberrant rather than a Diplazium, in which the chromosome number is n= 41, so far as known. 33. ATHYRIUM MATANGENSE Copel. Sarawak Mus. Journ. 2 :377. 1917. —Diplazium falcinellum C. Chr. Ind. Fil. Suppl. 3:73. 1934. Asplenium matangense Hose, Journ. Str. Br. Roy. As. Soc, 32:58. 1899, nom. prov. Syntrypes: Matang, Borneo, Hose 74 (presumably K), and 287 (K, Morton photograph 18934). In the “Index Filicum,” Christensen overlooked this species, which he apparently noted only in preparing Supplement 3. In the meantime, 499-018—73——3 230 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM he had applied in 1917 the name Diplazium matangense C. Chr. to an entirely different species, and therefore renamed the species of Hose D, faleinellum C. Chr. Hose listed this plant as merely Asplenium (Diplazium) sp. in his list of the ferns of Borneo in 1899. He gave a rather elborate description and ended up by saying that he named the species provisionally A. matangense. Since the author himself indi- cates that he considered his name provisional, we have no option but to consider it a nomen provisorium also, and consequently invalid. Ap- parently no one discussed the species again until Copeland, who ac- cepted it as a valid species. The epithet matangense therefore dates from Copeland. Under Athyrium, the name A. matangense Copel. is correct; under Diplazium the name must be D. falcinellum C. Chr. 34. ATHYRIUM MEGISTOPHYLLUM Copel. Phil. Journ. Sci. 56:475, t. 7. 1935. =Diplazium megistophyllum (Copel.) Morton, comb. nov. Tyre: Penibuken, Mount Kinabalu, Borneo, 5,000 ft. alt., Oct. 19, 1933, J. & M.S. Clemens 40806 (isotype K, Morton photograph 18948). This is one of the few Diplaziums that has a “spiny” stipe. In this case at least, the “spines” are the knobby persistent and hardened bases of the numerous deciduous stipe scales. This is one of the largest Di- plaziums, as indicated by the specific epithet, the middle pinnae being almost. a meter long. The caudex is unknown, but it may be conjectured that it is elongate and erect, thus making this a little “tree fern.” 35. ATHYRIUM SANCTI-JOHANNIS Copel. Occas. Papers Bern. P. Bishop Mus, 14:59, t, 10.1938. =Diplazium sancti-johannis (Copel.) Morton, comb. nov. Type: Tubuai Island, Austral Islands, Aug. 20, 1934, H. St. John 16440 (isotype US). This rare endemic of Tubuai is probably known only from the type collection. It is a relative of Diplazium membranaceum (Mett. ex. Kuhn) C. Chr., a more widespread Polynesian species. In the “Index Filicum,” the authority is given as “Mett. Linn. 36: 103. 1869 (syn.)”- however, a name published as a synonym is not published at all. Ap- parently no one else treated this species as a Diplazium until Christen- sen, and so Christensen must be considered the author of the combination D. membranaceum. 36. BLECHNUM SOULEYETIANUM Gaud. Voy. Bonite Atlas t. 184. f. 1-7. 1845-50. =Sadleria souleyetiana (Gaud.) Moore, Ind. Fil. XXVI. 1857 [as “souleytiana’’]. Type: Sandwich Islands [Hawaii], Gaudichaud. Tn the “Index Filicum,” 2B. soweyetianum [as “souleytianum”] is cited as “sine descr.” and consequently the name 8. souleytiana is indicated as first published by Hillebrand in 1888. Although there STUDIES OF FERN TYPES, II—MORTON 231 is no description, Gaudichaud’s ¢. 734 gives a full illustration with analyses and consequently provides a valid publication by Art. 42 of the Code. The exact date of publication of ¢. 34 is still uncertain. Johnston (Journ. Arn. Arb. 25:487. 1944) gives dates for most of the plates in Gaudichaud’s work, but leaves ¢. 134 only as “1845-50.” Further bibliographic work is indicated. 37. BLECHNUM sSQuUARROSUM Gaud. Voy. Bonite Atlas f. 2, f. 1, t. 6, f. 1-6, 1841. =Sadleria squarrosa (Gaud.) Moore, Ind. Fil, XXVI. 1857. Blechnum polystichoides Brack. in Wilkes, U.S. Expl. Exped. 16:134, 1854. Type: “Saw Mill,” Hawaii, Wilkes Exped. (holotype US). Sadleria polystichoides (Brack.) Heller, Minn. Bot. Stud. 1:788. 1897. Type: Sandwich Islands [Hawaii], Gaudichaud. As lectotype I choose the specimen in Firenze, with the name in the hand of Gaudichaud (Morton photo- graph 16428); an isotype is in Berlin (Morton photograph 10229). There is probably an isotype in Paris also. In the “Index Filicum” and in the “Revised List of Hawaiian Pteridophyta” (Bern. P. Bishop Mus. Bull. 25 :16. 1925), Christensen adopted the name Sadleria polystichoides, remarking in the “Index” that B. squarrosum Gaud. was “sine descr.” ; however, although there is no description by Gaudichaud, there is a plate with analyses and so B. squarrosum is validly published by Art. 42 of the Code. The ‘llustration is fine and readily identifiable as the same as the later B. polystichoides Brack., which is confirmed also by the isotypes that I have seen. I have seen the name given as “S. sqguarrosa (Gaud.) H. Mann” on some herbarium labels, which I do not understand. The combination was validly published by Moore in 1857, although wrongly attributed to Gaudichaud. 38. BRACHYSORUS WOODWARDIOIDES K. B. Presl, Epim. Bot. 70. 1849 [1851]. =Diplazium woodwardioides (K. B. Presl) Morton, comb. nov. Allantodia sylvatica Blume, Enum. Pl. Jav. 173. 1828, non Diplazium silva- ticum (Bory) Swartz. Type: Burangrang, Java, Blume (holotype L, Morton photograph 674). Athyrium basilare Fée, Gen. Fil. 197. 1852. Based on Brachysorus wood- wardioides K. B. Presl. Since the epithet woodwardioides was legitimate and available under Athyrium, there was no need to change it to “basilare,” which is thus an illegitimate, superfluous name, Asplenium woodwardioides (K. B. Presl) Baker in Hook. & Bak. Syn. Fil. 229. 1867, non Bernh. 1802. Athyrium woodwardioides (K. B. Presl) Christ, Bull. Herb. Boiss. 6 :154. 1898. Type: Luzon, Philippine Islands, Cuming 153 (holotype presumably in Prague, not seen; isotypes BM, Morton photograph 6999, and FI, Morton photograph 16203). Pres] distinguished his genus Brachysorus from Diplazium pri- marily by the short sori occupying the lower part of the veinlets and 232 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM not extending beyond the vein-fork, and by the sori being solitary rather than back to back. He did not mention the genus Allantodia R. Brown, although that is the obvious alliance. The sori are mostly in the lower part of the veins, but they do sometimes extend somewhat above the fork. I have not seen young sori, but it is likely that the indusia are somewhat inflated and at least partly covering the young sporangia, as in Aldlantodia australis R. Brown of Australia and A. umbrosa (Ait.) Kaulf. of Madeira. The habit of these plants is entirely that of Diplaziwm rather than Athyrium, and the cytological evidence is in favor of this, so far as it goes. Brownlie has found n= 123 in A. umbrosa and Sledge indicates that the related Diplazium muricatum (Mett.) van Alderw. van Rosenb. of Java, India, and Ceylon is n= 82, an indication that the base number of AJlantodia is w=41 as in Diplazium rather than v=40 as in Athyrium. Sledge has suggested that this group, Diplazium subg. Pscudallantodia (C. B. Clarke) Sledge (Bull. Brit. Mus. [Nat. Hist.] 2:312-315. 1962), may ultimately deserve generic status, in which case the name Allantodia is available, Pending a thorough study, however, it is better to place it as a Subgenus of Diplazium. 39. CHEILANTHES DEALBATA D. Don, Prodr. FI. Nepal, 16. 1825, non Pursh, 1814. =Cheilanthes farinosa (Forsk.) Kaulf, Enum. Fil. 212. 1824, sens. lat. Gymnia pectinata Buch.-Ham. ex D. Don, loc. cit. in synon. Hemionitis dealbata Wallich, mss. ex D. Don, loc. cit. Cheilanthes dealbata Wallich, Num. List. no. 71. 1829, nom. nud. There is no mention of Don and so this must be considered as a different name, although a nomen nudum. Aleuritopteris farinosa (Forsk.) Fée, Gen, Fil. 154, t. 12 B, f. 1. 1852. SYNTYPES: Narainhetty, Nepal, March 138, 1803, Buchanan-Hamilton (BM, Morton photograph 6694, left-hand plant) and Nepal, Waillich (BM, Morton photograph 6694, two right-hand plants). The Buchanan-Hamilton specimen is here designated lectotype. There is a duplicate of the Wallich collection in Brus- sels (Morton photograph 5098, right-hand specimen). Ching (Hong Kong Nat. 10:201. 1941) places C. dealbata D. Don as a synonym of typical Aleuritopteris farinosa, but it is doubtful that he saw any types. I must admit that the Himalayan plants do look like those from East Africa, which are presumed to be typical, but there may be some minute differences. 40. CoLYSIS HEMIONITIDEA K. B. Presl, Epim. Bot, 147, 1849 [1851]. =Polypodium hemionitideum (K. B. Presl) Mett. Abhandl. Senckenb. Naturf. Gesell. 1:112. 1856. Polypodium hemionitideum Wall. Num. List 10, no. 284, 1829, nom. nud. Selliguea hemionitidea K. B. Presl, Tent, Pterid. 216, t. 9, f. 17. 1836, nom. nud. The illustration is only an analysis, and not a plate with analyses, as required by the Code for valid publication. STUDIES OF FERN TYPES, II—MORTON 233 Drynaria hemionitidea J. Smith in Hook. Journ. Bot. 4:61. 1841, nom. nud. Type: Nepal, 1822, Wallich no. 284. There may be a specimen of this number in Presl’s herbarium in Prague, which will be the holotype. Duplicates are in the East India Company Herbarium at Kew, and elsewhere, including US. The authority for this species is usually cited, as in the “Index Filicum,” as “Wallich” or “Wallich ex Mett.,” but from the citations above it is clear that Presl must be considered as the author of the epithet, since he was the first to give a description. It cannot be “Wall. ex Presl” because Wallich had it under Polypodium, whereas Presl described it as a Colysis. It is still retained in Colysis by those recog- nizing this genus, such as Ching, even though the sori are not confluent and linear, as they ought to be in this genus. There is a Wallich specimen of this species in the British Museum (Natural History) (Morton photograph 7 616) that is labeled in Don’s own hand as “Polypodium membranaceum D. Don,” but it does not represent that species as usually considered, and it does not agree with Don’s description. Don’s description calls for a very large frond about four inches wide, very membranous, and with scattered, rotund sori, characters that agree with the common Himalayan species usually sdentified as P. membranaceum. Polypodium hemionitideum is a smaller plant of thicker texture, with the often somewhat elongate sori in a single median row. 41, CTENOPTERIS AMYLACEA Copel. Phil. Journ. Sci, 84:469. 1955. —Grammitis amylacea (Copel.) Morton, comb. nov. Type: Cargadira, Bolivia, L. Williams 1134 (US 700281). 42. CTENOPTERIS FRAGILLIMA Copel. Phil. Journ. Sci. 84:470. 1955. =Grammitis fragillima (Copel.) Morton, comb, nov. Type: Mount Roraima, Steyermark 58882. The holotype is in US, although not so stated by Copeland. 48. CTENOPTERIS HERRERAE Copel. Phil. Journ. Sci. 84 :-467, 1955. —Grammitis herrerae (Copel.) Morton, comb. nov. Type: Copeland cited two collections, Bues 1268 and 1269, but without in- dicating a type or providing locality data. In the United States National Herbarium, he indicated no. 1269 as the type, and it is here designated lectotype ; it is from Huadquifia, Department of Cuzco, Peru, collected November, 1920 (US 1515537). The other collection would not have been a suitable type because it is sterile. 44, CTENOPTERIS RHIZOPHORAE Copel. Phil. Journ. Sci. 84 425. 1955. =Grammitis rhizophorae (Copel.) Morton, comb, nov. Type: “Colombia, in mangrove swamp along Rio Dagua, alt. 0 to 5 m.” is all that Copeland cited, the rest having somehow been lost in the manuscript or printing. The holotype is in the United States National Harbarium (no. 1140066) ; the remaining data are: Buenaventura, Dept. El Valle, May 7, 9. 1922, H. P. Killip 5334. 234 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM 45. CTENOPTERIS SUBCRASSA Copel, Phil. Journ. Sci. 84:468. 1955. =Grammitis subcrassa (Copel.) Morton, comb. nov. Type: Copeland’s citation of the type is somewhat defective in that he omits the collector’s number, omits an indication of the herbarium where the type is deposited, and cites the collector as “L. A. Brade,” in which the “L” is not an initial but stands for the Latin word “Legit” (i.e,, collected by). The holotype is A. C. Brade 5838, collected at Campo Grande, Serra do Mar, Sido Paulo, Brazil, November, 1913 (US 1198707). 46. CYCLOPHORUS FOVEOLATUS Alston, Journ. Bot. 78:226. 1940. =Pyrrosia foveolata (Alston) Morton, comb. nov. Tyre: Boridi, Papua, Carr 13039 (holotype BM, not seen; isotype L, Morton photograph 826). This rather common New Guinea species is distinguished by the sori being sunken in pits. 47. DAVALLIA SCABRA D. Don, Prodr. Fl. Nepal. 9, 1825. =Microlepia marginata (Houtt.) C. Chr. Ind. Fil. 427. 1906. Davallia villosa Wallich, Num. List. no. 244. 1829, nom. nud., non D. villosa D. Don, 1825. Based on Nepal, Wallich in 1821, Microlepia scabra (D. Don) J. Smith, Journ. Bot. Hook, 1:427. 1842. Davallia villosa Wallich ex Hook. Sp. Fil. 1 1172, t. 48A. 1846, Since Hooker cited Davallia scabra D. Don as a synonym and since this name was legitimate, prior, and available, Hooker's D. villosa must be considered a superfluous renaming of D. scabra and based on the same type. Also, D, villosa Wallich ex Hook. is an illegitimate later homonym of the different species D. villosa D. Don, 1825. Microlepia villosa (Wallich ex Hook.) K. B. Presl, Epim. Bot. 95. 1849 [1851]. Type: Nepal, Wallich (holotype BM, Morton photograph 6901, a sheet with a full frond with a detached stipe and a single detached pinna in the upper right hand corner; the label in the upper right hand corner may be in Don’s hand). 48, DICRANOPTERIS PALMATA Underw. Bull. Torrey Bot. Club 34 :259. 1907. =Gleichenia palmata (Underw.) C. Chr. Ind. Fil. Suppl. 1:113. 1913. Mertensia palmata Schaffner ex Fée, Mém. Foug. 9:32. 1857, nom. nud.; Fourn. Mex. Pl. 1:137, 1872, nom. nud. Gleichenia palmata Moore, Ind. Fil. 380. 1862, nom. nud. Type: Orizaba, Veracruz, Mexico, Pringle 6129 (NY; isotypes B, US). Distrib- uted originally as G. pubescens H.B.K. The first valid publication of this species was by Underwood as Dicranopteris palmata, who designated the type as Pringle 6129. Since this is a different collection from the Schaffner specimen intended as the basis for the nomina nuda Mertensia palmata Schaffn. and Glet- chenia palmata Moore, the species must be considered as wholly Under- wood’s. The first acceptance of the species under the name Gleichenia after Underwood's description was by Christensen in the Corrigenda (1913) to the “Index Filicum,” where the species is accepted under the name Gleichenia palmata (Schaffner) Moore. These authorities are STUDIES OF FERN TYPES, II—MORTON 235 not in accordance with the Code, since they refer to nomina nuda and also to a species based on a different type. The proper authority is Gleichenia palmata (Underw.) C. Chr. Ind. Fil. Suppl. 1:118. 19138 [incorrectly attributed to “ (‘Schaffn.) Moore’’]. Other collections refer- able to this species in the Berlin herbarium are: Orizaba, Veracruz, Mexico, Afiller ; San Andres, Veracruz, Mexico, Schiede & Deppe 727; Jalapa, Veracruz, Mexico, Khrenberg 843; Mexico, Schaffner; and Mabess River, Jamaica, Harris 7597. All of these are filed as G. fur- cata, following identifications of Hieronymus, but they are not close to the true furcata of the Lesser Antilles. 49, DIPLAziUM CAUDATUM J. Smith in Hook. Journ. Bot. 3:408. 1841, nom. nud. ; C. Chr. Ind. Fil. Suppl. 3:72. 1934, nom. nud. =Diplazium melanopodium Fée, Mém, Foug. 8 :85. 1857. Athyrium melanopodium (Fée) Copel. Fern Fl. Phil. 401. 1960 [wrongly attributed to Fée]. 2Diplazium meyenianum K. B. Presl, Epim. Bot. 88. 1849 [1851]. Type: Manila, Philippine Islands, Meyen (Herb. Presl, Prague). VAsplenium aspidioides Goldm. Noy. Acta Acad. Leop. Caes. Nat. Cur. 19, Suppl. 1:461. 1843, non Schlecht., 1825. Type: Manila, Philippine Islands, Meyen (presumably B). Type: Since D. caudatum J. Smith is a nomen nudum, it does not strictly speak- ing have a type, but it was intended to be based on Luzon, Philippine Islands, Cuming 158. Christensen was in error in taking up the same Diplazium caudatum J. Smith, for this was a nomen nudum only. As a synonym he cited Athyrium meyenianum Copel. (Phil. Journ. Sci. 3C :295. 1908), but this was an error also, for A. meyentanum was not a new species of Copeland but merely a new combination based on Diplazium meyenia- num K. B. Presl, which Christensen specifically excluded from his concept of D. caudatum. Doubtless, Christensen intended Athyrium meyenianum sensu Copel. excl. synon., but there was no description by Copeland at the place cited that would serve to validate the name D. caudatum. The other synonym cited by Christensen, D. melanopodium Fée, was cited only with a query, and so this cannot validate the publi- cation of the name D. caudatum J. Smith either. Hieronymus (Hed- wigia 59 :337. 1917) believed that D. meyentanum K. B. Pres] was iden- tical with the Hawaiian D. arnottii Brack., of the Hawaiian Islands, but Holttum, in his recent paper on Pres] types, indicates that this is doubtful and that the species needs further study. 50. Diplazium crenulans Fée ex Baker in Hook. & Bak. Syn. Fil. 232. 1867, in obs. Diplazium legalloi Proctor, Rhodora 68 :466. 1966. Diplasium celtidifolium sensu auctt. as to plants from the Lesser Antilles. Diplazium callipteris sensu auctt. as to plants from the Lesser Antilles. Type: Guadeloupe, L’Herminier 67 (holotype K, Morton photograph 18581). 236 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM Diplazium erenulans is mentioned by Baker only in an observation, but there is enough description to validate the name. This species has usually been called D. callipteris, as by Fée himself, who kept that species distinct from D. celtidifolium Kunze, which he also attributed to the Antilles. Proctor, however, states that the type of D. callipteris came from Venezuela, rather than from Cuba as stated by Fée, and that it represents D. celtidifolium. He does not state that he has seen authentic material, but that is the inference. He is very likely right, because no species closely allied has ever been collected since in Cuba. There are two other specimens at Kew of D. crenulans, but they came from the T. Moore Herbarium, which was purchased in March, 1885, and so these were not studied by Baker presumably; they are also from Guadeloupe, collected by L’Herminier, and may be part of the same gathering (Morton photographs 18582, 18582a) although the pinnae are merely crenate-dentate in one (rather than lobulate) and subentire in the other ; this species appears to vary in these respects, perhaps according to the size of the plants or the position of the pinnae gathered. It would appear that the lowest and largest pinnae are lobulate, the middle crenate-dentate, and the upper subentire. Another specimen mounted on two sheets is also at Kew, collected in Guadeloupe by L’Herminier (Morton photographs 18583, 18584) ; it is identified by Fée as D). callipteris var. undulatum, an unpublished name, probably intended originally to have something to do with Liplazium undulosum Swartz, which was based on Plumier, Tract. Fil. ¢, 107. and which is listed in the “Index Filicum” as a dubious species. Plumier’s illustration is one of his more esoteric ones, but it is clear that it cannot possibly represent a Diplazium. The illustration and the ac- companying description show large sized pinnae and many transverse veinlets. Among the plants known from Martinique it can only repre- sent Lhelypteris [Meniscium] reticulatum (I.) Proctor. Fée (Mém. Foug. 11:40. 1866) indicated that D. wndulosum Swartz was probably a Meniscium. 51. DIPLAZIUM FALCATUM D Don, Prodr. Fl. Nepal. 13. 1825. =Coniogramme fraxinea (D. Don) Diels in Engl. & Prantl, Nat. Pflanzenfam. 1(4) :262. 1899. Hemionitis falcata Buch.-Ham. mss. ex. D. Don. loc. cit. Gyumnogramma falacta (D. Don) J. Smith, Journ. Bot. Hook. 4:51. 1842. Coniogramme falcata (D. Don) Salomon, Nomencl. 139. 1888. Type: Narainhetty, Nepal, March 1, 1803, Buchanan-Hamilton (holotype BM, Morton photograph 6691). In his monograph of Coniogramme, Hieronymus (Hedwigia 57 :325. 1916) considered D. falcatum as a dubious species, perhaps a synonym of C. fraxinea, which appears to be true. It seems odd that Don would STUDIES OF FERN TYPES, II—MORTON 237 describe the same species twice, but this is explained by the fact that his D. frawineum had the lower pinnae ternate or pinnate and his D. falcatwm had them simple. It seems that this species varies in this respect, but the matter deserves study in the field. If the form with simple pinnae should be considered different, the name C. falcata (D. Don) Salomon is available for it. 52, DIPLAZIUM PROLIFEROIDES Bory in Bélanger, Voy. Bot. 2:38, 1833. =Diplazium proliferum var. proliferoides (Bory) Morton, comb. nov. Athyrium accedens var. proliferoides (Bory) Tardieu, Bull. Mus. Hist. Nat. Paris IT, 29 :290. 1957. Type: Mauritus, Bélanger (holotype presumably P; isotype FI, Morton photo- graph 16154). This is a form of the widespread Diplazium proliferum with the pinnae rather deeply lobed; consequently, the venation is simpler, most of the veins being connivent to the sinuses and only occasionally anastomosing. Madame Tardieu, who has seen abundant material from Mauritius and Réunion, however, considers it not specifically distinguishable from typical D. proliferum. Whether the Javan D. accedens Blume is specifically different remains an open question. 53. DRYMOGLOSSUM SUBCORDATUM Fée, Mém. Foug. 3:29. 1852. =Lemmaphyllum microphyllum K. B. Presl, Epim. Bot. 263. 1849 [1851]. Syntyres: Coast of China, Gaudichaud (FI. with the name in Fée’s hand, Morton photograph 16218), and Amboina, Labillardiére (FI, with the name in Fée’s hand, Morton photograph 16219). Christensen (Dansk. Bot. Ark. 6(8) :46, 47, 86. 1929) discussed the identity of D. subcordatum. He had not seen the syntypes, but decided that Fée’s description agreed best with material from China, and so chose the Gaudichaud collection as lectotype. Judging from the locality “Amboina,” he believed that the second syntype represented an en- tirely different species, Drymoglossum fallax van Alderw. van Rosenb., and from my photograph of the Labillardiére specimen it appears that he is right, although there is only one fertile frond on the sheet and the photograph is not sufficiently detailed to show the arrange- ment of the sporangia, which is very peculiar in D. fallax, the sporan- gia being arranged in a single row, as Christensen expresses it “like cakes in a tinbox.” Christensen thought that Fée’s figure (Gen. Fil. t 9A, f. 1.1852) represents the Amboina plant, but it does not seem so to me; judging from the disposition of the sporangia, it seems that the fertile leaf at least is Lemmaphyllum microphyllum, and I think that the sterile leaves are also L. microphyllum. The stellate hair shown was probably not drawn from material of D. subcordatum at all, since it is indicated as representing the hairs of three different species. The sterile fronds of the Labillardiére specimen are definitely cuneate at 238 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM the base, and therefore Fée’s description of these as subcordate is not applicable, nor is the specific epithet subcordatum. But sometimes Chinese specimens of L. microphyllum from Hong Kong really are subcordate, another point indicating that the Gaudichaud plant is the correct lectotype. 54, DRYNARIA STENOPHYLLA Fée, Mém. Foug. 6:18, t. 8, f. 3. 1853, non J. Smith. 1841. =Polypodium stenophyllum Blume, Enum. Fil. Jav. 124. 1828, Drynaria stenophylla (Blume) J. Smith, Journ. Bot. 3:397. 1841. Based on Polypodium stenophyllum Blume. Crypsinus stenophyllus (Blume) Copel. Gen. Fil. 206. 1947. Tyrr: Luzon, Philippine Islands, Cuming 122 (isotype FI, Morton photograph 16028). Drynaria stenophylia Fée (1853) is usually mentioned (if at all) as though it were the same as D. stenophylla J. Smith (1841), but it is not. Smith’s name was a new combination based on Polypodium stenophyllum Blume, and the type is therefore the same as Blume’s type from Java. Fée attributed his D. stenophylla to J. Smith, but he described it as a new species based on Cuming 122 from Luzon, Philippine Islands, and made no mention of Blume. Therefore, Drynaria stenophylla Fée must be considered as nomenclaturally a different species since it has a different type; it is thus a legitimate name under Drynaria. The Philippine Island plant represented by Cuming 122 is considered by Copeland as taxonomically the same as the Javan P. stenophyllum Blume (Java, Blume, holotype L, Morton photograph 1969), but Copeland did not see the Blume type, which may or may not be the same. 55. DRYOPTERIS CHEILANTHINA ©. Chr. Kungl. Svensk. Vetensk. Akad. Handl. (Stockholm) III, 16(2) :34. March, 1937. =Ctenitis cheilanthina (C. Chr.) Morton, comb. nov. Type: Morne Cabaio, near Robergeau, Nouvelle Touraine, Selle, Haiti, Ekman 1689 (isotype US). Collected several times by Ekman but not otherwise. 56. DRYOPTERIS MERIDIONALIS (Poir.) C. Chr. var. SPELUNCAE C. Chr. Dansk. Vid. Selsk. Skrift. VIII, 6:47. 1920. =Ctenitis meridionalis var. speluncae (C. Chr.) Morton, comb. nov. Tyre: Six collections were cited, all from Bermuda, none of which was designated as the type. As lectotype I choose: Sink hole, Paynter’s Vale, Ber- muda, Aug. 3-Sep. 20, 1905, 8S. Brown & N. L. Britton 266 (US 524898). This is a good specimen, identified by Christensen. 57. DRYOPTERIS SANCTA var. PORTORICENSIS C, Chr. Smiths. Mise. Coll. 52, no. 1867 :380. 1909. =Thelypteris sancta var. portoricensis (C. Chr.) Morton, comb. nov. Aspidium sanctum var. portoricense Kuhn, Bot. Jahrb. Engler 24 :115. 1897, nom. nud. STUDIES OF FERN TYPES, II—MORTON 239 Thelypteris sancta var. portoricensis (Kuhn) Morton, Amer. Fern Journ. 53 :64, 1963, nom. nud. Type: There were three syutypes cited, but no lectotype has been designated. As lectotype I choose the second syntype: Utuado, Puerto Rico, Sintenis 5956 (US, with the name in Christensen’s hand). The third syntype, Underwood ¢& Griggs 60 is also in US, as is a duplicate of the first syntype, Sintenis 403, the latter, however, without the name in Christensen’s hand. It has been pointed out to me that when, in 1968, I attempted to publish a new combination for this variety under 7helypteris that I wrongly cited a nomen nudum as the basionym. Although the epithet portoricense originated with Kuhn the variety was really first de- scribed later by Christensen. 58, DRYOPTERIS STEGNOGRAMMA var. ASPLENIOIDES C. Chr. Acta Hort. Gotob, 1:56. 1924. =Thelypteris dasypoda Morton, nom. nov. Stegnogramma asplenioides (C. Chr.) Ching, Sinensia 7:94. 1936, non Thelypteris asplenioides (Swartz) Proctor. Synryres: Christensen cited Szechuwan, China, H. Smith 2116, and Khasia, India, collector not stated. Ching chose Khasia, Griffith as lectotype. Christensen indicated the author of his var. asplenioides as “(J.Sm.) C. Chr.,” as though it were a new combination of some validly pub- lished name, but I cannot find that Smith ever published any name that could be a basionym, and therefore the variety must be attributed to Christensen only. Similarly, Ching in transferring this variety to Stegnogramma attributed the name Stegnogramma asplenioides to “J. Sm. mss.,” but since the first describer was Christensen, the proper authority must be “(C. Chr.) Ching.” As to the distinctness of Stegnogramma as a genus, see my remarks in Amer. Fern Journ. 56:177-179. 1966. Since the epithet “asplenioides” has already been used in Thelypteris, a new name is necessary for this species. 59. Elaphoglossum alatum Gaud. in Vaillant, Voyage Autour du Monde Bonite, Bot. Atlas ¢. 135, 1845-50. In my previous paper on fern types (Contr. U. S. Nat. Herb. 38 :44. 1967) I stated that Llaphoglosswm alatum Gaud. was a synonym of E. gorgoneum (Kaulf.) Brack. When this was written I had not seen the paper “A Revision of the Hawaiian Species of Elaphoglossum,” by W. R. Anderson and M. R. Crosby (Brittonia 18 :380-397. 1966), in which it is indicated that a study of the type specimen has shown that Acrostichum gorgoneum Kaulf. does not represent #’. gorgoneum as usually named but is a synonym of Acrostichum aemulum Kaulf. [=Elaphoglossum aemulum (Kaulf.) Brack]. Anderson and Crosby were probably the first to unite these two species, and therefore their choice of the epithet aemulum must be followed; this is the best any- way for avoiding confusion, since the epithet gorgoneum has been so 240 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM widely used for the different species /’. alatum Gaud., which is the correct name for this species. Aconiopteris obtusa Fée (1845) is per- haps earlier, but the epithet obtusa may not now be transferred to Elaphoglossum because of the different £. obtuswm A. Peter (1929). 60. EpIDRYOPTERIS LYCOPODIOMUS Rogas, Bull. Acad. Géogr. Bot. [Le Mans] 28 :156. 1918. =Polypodium vacciniifolium Langsd. & Fisch. Icon. Fil. 8, t. 7. 1810. In a paper entitled “Addenda ad Floram regionis Chaco Australis,” N. Rojas Acosta published a new genus and species called peculiarly enough E'pidryopeteris lycopodiomus, based on material from Para- guay, presumably collected by himself. In 1958, I wrote asking about this plant to the Jardin Botdnico in Asuncién, where the Rojas her- barium is presumably preserved, but I never received a reply. Although Christensen in the “Index Filicum,” Supplement ITI, referred the genus to Polypodium with a query, the species has never been definitely placed. The original description is as follows: Scandens, filiformis, ramosa, setoso-paleacea vel albida, 20-60 em. longa, frondi- bus alternis, glabris, rotundo-ellipticis, supra nitidis, fertilibus linearibus, soris solitariis. Oct. Nov. Voisin du Polypodium lycopodioideum de Mexico. Croit dans les foréts humides et ombreuses sur les trones et les rameaux des arbes, surtout aux bords du Parana et du Rio Bermejo. There are not many epiphytic ferns in Paraguay, and the only one that agrees with the description in having an elongate, scandent, epi- phytic, whitish-scaly rhizome, rotund-elliptic and shining sterile blades, and linear fertile blades is Polypodium vaceiniifolium Langsd. & Fisch. The identification can be considered certain, even in the absence of type material. 61. GLEICHENIA BIFURCATA Blume, Enum. Pl. Jav. 250, 1828, —Gleichenia trun- cata (Willd.) Spreng. in L. Syst. Veg. ed. 16, 4:25. 1827. Type: Boerengrang, Java, Blume (holotype L, with the name in Blume’s hand, Morton photograph 1421). This species was overlooked in Holttums’ account of Gleichenia in the “Flora Malesiana” (II, 1(1) :4-27. 1959), but the type was an- notated by Holttum in 1957 as G. truncata. 62. GLEICHENIA LAEVIGATA Var. BRACTEATA Rosenst. Repert. Sp. Nov. Fedde 5 :370. 1908. =—Gleichenia truncata var. bracteata (Rosenst.) Holttum, Rein- wardtia 4 :271. 1957. TYPE: Rosenstock gave no description, but his varietal name is validated by his reference to Hooker and Baker’s characterization of G. bracteata Blume (Syn. Fil. 14. 1865) ; the Hooker and Baker description was based on Java, Blume in Herb. Hook., and the holotype is therefore at Kew. Holttum considered the species Gleichenia bracteata Blume ex Hook. & Bak. validly published, and consequently cited “Blume ex Hook. & STUDIES OF FERN TYPES, II—MORTON 241 Bak.” as the parenthetical author. However, even though Hooker and Baker did mention a few of the characters of Blume’s G. bracteata, they did not accept the species; since it was published as a synonym only it cannot be valid. Holttum indicated that the holotype of G. bracteata is in Leiden, but the description by Hooker and Baker was of a Blume plant in the Hooker Herbarium at Kew, and this must there- fore be the holotype. 63. GLEICHENIA LANIGERA D, Don. Prodr. Fl. Nepal. 17. 1825. =? Dicranopteris linearis (Burm. f.) Underw. Bull. Torr. Bot. Club 34 :250, 1907. Type: Sirinagur, India, Kamroop (not seen). There is in the British Museum (Natural History) a sheet annotated Gleichenia lanigera Don but this is from “in alpibus Nepaliae,” not Sirinagur. It was probably collected by Wallich and is a part of his no. 157, i.e., Gleichenia gigantea Wallich. Not only is the locality different but the plant is also different, apparently, for this sheet is one of the pinnate type like G. glauca (Thunb.) Hook., whereas Don described his species as variously dichotomous. Don’s species is considered to be Dicranopteris linearis (Burm. f.) Underw., according to J. Smith (quoted by Hooker, Sp. Fil. 1:5. 1844, under Gleichenia dichotoma). 64, GLEICHENIA LINEARIS var. BIDENTATA van Alderw. van. Rosenb. Bull. Jard. Bot. Buitenzorg III, 5:204. 1922. =Dicranopteris linearis (Burm. f.) Underw. var. linearis. Type: Several syntypes were cited; of these I have seen Biinnemeijer 7881, from Mount Djantan, Poeloe Karimon Island, Riouw Archipelago (L, Morton photograph 1425). This variety was overlooked in Holttum’s revision of Gleichenia in the “Flora Malesiana.” From the photograph of the syntype cited above it appears that the variety can be safely referred to var. linearis. 65. GLEICHENIA LINEARIS Var. CRASSIFRONS van Alderw. van Rosenb. Bull. Jard. Bot. Buitenzorg III, 5:204. 1922. =Dicranopteris linearis (Burm. f.) (Blume) Holttum, Reinwardtia 4 :277, 1957. Type: Foramadiahi, Ternate Island, Mar. 11, 1921, Beguin 1501 (holotype L, Morton photograph 1426). This variety was overlooked in Holttum’s revision of Gleichenia in the “Flora Malesiana.” The occurrence of var. rigida in Ternate is not mentioned either, but it was to be expected since it occurs in Tidore and the Celebes. Another variety that was overlooked is Gleichenia linearis var. irregularis van Alderw. van Rosenb. (Malay. Ferns Suppl. 84, 1917) ; no collections were cited for this variety and so the type (ora lectotype) can be determined only by consultation with van Alder- werelt’s material so named in Bogor and Leiden; the variety appar- 242 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM ently can be referred to Dicranopteris linearis var. alternans (Mett.) Holttum going on description and probability. 66. Gleichenia longissima var. nivea Blume, Enum, P1. Jav. 251. 1828. Type: Java, Blume (holotype L, with the name in Blume’s hand, Morton pho- tograph 1441). This is apparently a form in which the segments are even more strongly glaucous beneath than usual. The varietal name was over- looked by Holttum in the “Flora Malesiana.” 67. GRAMMITIS SCOLOPENDRIOIDES Gaud. in Freycinet, Voy. Monde Uranie 310. Aug. 1828. =Loxogramme scolopendrioides (Gaud.) Morton, comb. nov. Grammitis scolopendrina Bory in Duperrey, Voy. Monde Coquille, Crypt. 267, t, 80, 7. 1, Nov. 1829. Type: New Ireland, d@’Urville (holotype P, Morton photograph 21514). The locality was originally stated as “New Zealand,” but this was surely a slip, for the species is not known from New Zealand, and the holotype clearly came from New Ireland, in the New Hebrides. Lorogramme scolopendrina (Bory) K. B. Presl, Tent. Pterid. 215. 1836. Type: Rawak [=Lawak] Island, near New Guinea, Gaudichaud (holotype P, with the name in Gaudichaud’s hand, Morton photograph 21512). In the “Index Filicum,” Christensen indicated that Grammitis scolo- pendrioides Gaud. was probably the same as G. scolopendrinum Bory, but apparently no one has ever studied the type and confirmed this. These two species were published at about the same time, but Stafleu’s “Taxonomic Literature” shows that Gaudichaud’s name clearly has priority by over a year. The species is close to Lorogramme involuta (D. Don) K. B. Presl but is kept distinct by Copeland in his “Gram- mitidaceae of New Guinea” (Phil. Journ. Sci. 81 :117, 1952). 68. GYMNOGRAMMA EGGERSII Christ, Bull. Soc. Roy. Bot. Belgique 32(2) :92. 1894 [June 27, 1895]. =Asplenium monteverdense Hook. Second Cent. ¢. 41. 1860. Anogramma eggersti Christ in C. Chr. Ind. Fil. 58. 1905. Asplenium mortonti Duek, Adansonia 11:718. 1971. Based on Anogramma eggersii Christ, non Asplenium eggersii Sodiro. Types: Jagiiey, Oriente, Cuba, alt. 500 m., March, 1889, Eggers 4882a (holo- type P, Morton photograph 3593). Christ failed to see the indusia and so referred this plant to @ymno- gramma,; however, the indusia are quite obvious, and the specimen dif- fers in no way from the type of Aspleniwm monteverdense (Wright 1029), which came from Monteverde, Oriente, Cuba, not very far from Jagiiey, which was misread by Christ as “Jaqueij.” In its rather elon- gate ultimate segments A. monteverdense Hook. is only slightly differ- ent from Asplenium myriophyllum (Swartz) Presl, to which it has been referred as a synonym; however, it may tentatively be main- tained as a species, pending cytological study. STUDIES OF FERN TYPES, II—MORTON 243 69, HEMIONITIS DEALBATA Willd. Hort. Berol. 1:#. 40. 1805. =Pityrogramma tartarea (Cay.) Maxon, Contr. U.S. Nat. Herb. 17 :173. 1913. In his paper on Pityrogramme (Contr. Gray Herb. 189 :65. 1962), Tryon stated that Wemionitis dealbata Willd. is a superfluous and illegitimate name, a renaming of Acrostichum tartareum Cav., but this is an error, due to Tryon’s having cited the wrong place of publication. Hemionitis dealbata was not first published in Willdenow’s edition of the “Species Plantarum” in 1810, but was published in 1805, with a detailed description. In this 1805 publication, Acrostichum tartareum Cav. is not mentioned. Acrostichum ebenum L. is cited as a synonym, but only with a query, showing that the species was based on a speci- men. The locality is cited as “in America australi ad fluvium nigrum.” The description and plate of Willdenow were based on material culti- vated in the botanical garden in Berlin. No cultivated material col- lected by Willdenow is known. In the Willdenow Herbarium in Berlin there are three sheets determined as H. delabata under the number 19568, two of them collected by Humboldt and one by Swartz. I trans- late Willdenow’s locality “ad fluvium nigrum” as referring to the Rio Negro in Brazil. Swartz never collected in South America, and so the cultivated material could not have been obtained from his collection. According to the account of H. dealbata in Humboldt, Bonpland, and Kunth’s “Nova Genera,” Humboldt collected this species near Guana- juato, Mexico, and so the Humboldt collection cannot have been the original source of the material either. Therefore, there is no herbarium specimen available as a holotype or lectotype, and the species must be typified on the original description and plate, which are fortunately rather good and complete. These show that Tryon was right in con- sidering this species a synonym of Pityrogramma tartarea (Cav.) Maxon. The plant described and depicted was, however, a very small one of this species, probably due to its being unhappy under cultivation. 70. Hymenophyllum capillare Desvaux, Mém. Soc. Linn. Paris 6 :333. 1827. In my first paper on fern types (Contr. U.S. Nat. Herb. 88 :76. 1967), I placed H. capillare Desv. as a synonym of Hymenophyllum aerugi- nosum (Poir.) Carm., going on the locality Tristan d’Acunha, since that is the only species of this alliance on that island. Dr. EK. A.C. L. E. Schelpe kindly drew my attention to a note by Christensen in his paper “The Pteridophytes of Tristan da Cunha” (1940), in which it is indi- cated that the holotype in Paris has the locality corrected from Tristan d’Acunha to Bourbon. I have photographed this holotype (Morton photograph 4548). It is identical with H. pendulum Bory (1883) and 244 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM is the correct oldest name for that species of the Mascarene Islands. It is thus to be excluded from the flora of Tristan d’Acunha. 71. Hymenophyllum fusugasugense Karst. ex Sturm, Bot. Zeit. 1859 :297. 1859. A specimen in the Hamburg herbarium collected at Fusugasuga, Bogota, Colombia, 2,700 m. alt. by H. Karsten is probably the type. It agrees with the illustration by Karsten (FI. Col. 2:4. 155, f. 1-8. 1862-69) and with my delimitation (Contr. U.S. Nat. Herb. 29 :186. 1947), but as I have shown (Amer. Fern Journ. 48:175. 1958), this species should be called H. tomentosum Kunze var. fusugasugense (IXarsten) Morton. 72. Hymenophyllum interruptum Kunze, Linnaea 9 :107. 1834. Type: Pampayacu, Peru, Poeppig 1104. A specimen in the Hamburg herbarium merely indicated as collected in Peru by Poeppig is probably an isotype. It is a detached and some- what depauperate frond that matches closely a frond on the collection Schunke 183, from Chanchamayo Valley, Junin, Peru, as cited by me (Contr. U.S. Nat. Herb. 29:178. 1947). It will not key out very well, however, because the rhachis is, exceptionally, alate to the base, and even the stipe is a little alate at the apex. Still, there is no doubt as to the correctness of the usual concept of this species. 73. Hymenophyllum molle Morton. Contr. U.S. Nat. Herb. 29 :149. 1947. A specimen in Hamburg collected at Agapata, Peru, by Lechler is typical of this local Peruvian species. It bears the name Hymenophyl- lum myriocarpum Hook. and is doubtless Lechler 2250a, listed by Met- tenius as “ZH. myriocarpum Hooker ?” (Filices Lechleriane 1:26. 1856). It is, however, by no means H. myriocarpum, which is a glabrous species belonging to a different subgenus. Someone, probably C. Miil- ler, has written “77. myriocarpoides C. Mill. n. sp.,” but this name is unpublished. 74, HYMENOPHYLLUM NITENS Wercklé ex Christ, Bull. Herb. Boiss. II, 4:946. 1904, non R. Brown, 1810. =Hymenophyllum crispum H. B. K. Nov. Gen. Sp. 1:26. 1816. Hymenophyllum micans Christ, Bull. Herb. Boiss. II, 5:260. 1905. Based on H. nitens Wercklé, non R. Brown. Type: Costa Rica, without specific locality, Wercklé 249 and 252. Syntypes presumably P; two apparent isosyntypes BR, Morton photographs 19866, 19867. The two apparently authentic collections that I have seen in Brus- sels agree with the brief original description. They are clearly typical H. crispum as it grows in Costa Rica. In my revision of the section STUDIES OF FERN TYPES, II—MORTON 245 Sphaerocionium I omitted this species, believing at the time that it belonged in section Mecodium. 75. Hymenophyllum tenellum D. Don, Prodr. Fl. Nepal. 12. 1825. Hymenophyllum exsertum Wallich, Num. List. no. 170. 1829, nom. nud. Based on Nepal, Wallich in 1821. Hymenophyllum exsertum Wallich ex Hook. Sp. Fil. 1:109, t. 384. 1844. Syntypes: H. exsertum Wallich Cat. no. 171, Nepal, Wallich, and H. densum Wallich Cat. no. 170, Nepal, Wallich. Hooker here has the names reversed from those in Wallich’s “Numerical List,” where H. exsertum is no. 170 and H. densum no, 171. This is surely a slip of the pen, and the species should be considered based on the plant called ersertum by Wallich, namely no. 170; the specimen of this number in Kew is here designated lectotype. An isolectotype is in US. Another reason for not choosing no, 171 is that this number perhaps does not represent the species described. At least the specimen in US of this number, which bears the name Hymeno- phyllum densum Wallich, is not a Hymenophyllum, but is Trichomanes bipunctatum Poir.; our specimen is so annotated by Copeland. I do not know what the Kew material of this number may be. Trichomanes bipunc- tatum does resemble this species of Hymenophyllum a little, and so it could be that both Wallich and Hooker were mistaken in referring their plants to Hymenophyllum ; the Trichomanes can be distinguished at once by the thicker rhizome, which is densely covered by short, black trichomes. Mecodium exsertum (Wallich ex Hook.) Copel. Phil. Journ. Sci. 67 :23. 1938. LEcTOTYPE: Nepal, Buchanan-Hamilton (BM, Morton photograph 6592). The other syntype is Nepal, Wallich, which I have not seen. Presumably it is a part of the same collection that was the basis of Wallich List no. 170, i.e., H. exsertum Wallich. In the “Index Filicum,” Hymenophyllum tenellum D. Don is listed as a dubious species. Copeland in his revision of the Old World Hymenophylium listed it as a synonym of H. polyanthos (Swartz) Swartz; he had not seen the type but merely says: “Wallich 172 must be discussed, because, as it is recognized as H. polyanthos by Hooker, Sp. Fil. 1:102, I see no reason to doubt that it is HZ. tenellum Don, which Hooker listed, page 112, as a ‘dubious species.’ ” This is a non sequitur, for the fact that Hooker listed H. tenellum as a dubious species is no reason for associating it with the plant called HZ. poly- anthos by Hooker. Apparently it did not occur to Copeland that there are other species in Nepal to which Don’s name might apply, in par- ticular HZ. exsertum. This species is somewhat similar to 1. polyanthos, but the division of the blade is a little different, and a distinguishing character is the presence of hairs on the rhachis; the rhachis of H. polyanthos in all its many forms is entirely glabrous. 76. LASTREA ACUMINATA Houlston, Gard. Mag. Bot. Hort. Floricult. Nat. Sci. 1851 :317. 1851. =Lastreopsis acuminata (Houlston) Morton, comb. nov. Aspidium shepherdii Kunze ex Mett. Fil. Hort. Lips. 94. 1854. Mettenius was the first to give a description of Aspidium shepherdii Kunze, a nomen 499-018—73——3 246 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM nudum when first published (Linnaea 23 :230. 1850). Kunze’s plants were cultivated at Kew in 1822, Berlin in 1842, and Leipzig in 1843. The syn- types were destroyed in Leipzig during the war, but there are probably some isosyntypes in Berlin that will provide a suitable lectotype, presum- ably one in the Mettenius Herbarium. Dr. Tindale indicates that the holotype is in the Christensen Herbarium in the British Museum (Natural History), but this is impossible. As indicated above there was no holotype, only three syntypes, and none of these could be in the Christensen Her- barium (unless Christensen stole one from Leipzig before the war, which is not likely). Lastrea atrovirens J. Smith, Cat. Cult. Ferns 59, 1857. Type: Cultivated at Kew (holotype J. Smith Herbarium, BM 2 sheets, Morton photographs 6463, 6464). Dryopteris shepherdii (Kunze) C. Chr. Vict. Nat. 60 :155. 1944. Lastreopsis shepherdii (Kunze) Tindale, Viet. Nat. 73 7182, 1957. Type: A plant cultivated by Loddiges in 1842 (holotype Herb. Houlston, BM, 2 sheets, Morton photographs 6771, 6772). In her fine “A Monograph of the Genus Lastreopsis Ching” (Contr. New South Wales Nat. Herb. 3 :249-339. 1965), Dr. Tindale has over- looked Lastrea acuminata Houlston, perhaps because in Supplement 3 of the “Index Filicum” (p. 117. 1934) the locality is given as “Nepal.” Houlston’s plant, however, was cultivated from some unknown source and he only guessed that it might be from Nepal. His type, now in the British Museum (Natural History), is identical with the type of L. atrovirens J. Smith, and their plants probably came from the same source. It is indeed quite likely that the original cultivated plants of Aspidium shepherdii Kunze also came from the same source at Kew, where it was said by Kunze to be in cultivation as early as 1822. Houlston described his Z. acuminata as a new species, with no refer- ence or suggestion that it might be the same as Aspidium acuminatum Willd. The confusion with Willdenow’s A. acuwminatum was due to Lowe, who later illustrated Houlston’s species (Ferns Brit. & For. 6:2. 12. 1857) but attributed the name to Willdenow, for what reason cannot be imagined, because Aspidium acuminatum Willd. is an utter! y different plant, considered to be a synonym of NVephrolepis biserrata. Inasmuch as the original description was published in a rather rare journal, rare at least outside of England, I give below a transcript of it: L. acuminata J. H.—An evergreen warm greenhouse Fern, the native country of which is doubtful; prob, from Nepal. Frond glabrous, triangularly elongate, bipinnate, a foot or more long, dull green : pinnae triangularly elongate-acuminate ; pinnules oblong-ovate, lower ones entire, and slightly pinnatifid, upper ones decurrent at the base, and rather rigidly toothed. Sori medial. Stipes sealy at the base, terminal, adherent to a somewhat tufted rhizome. 77, LASTREA SPINESCENS Houlston, Gard. Mag. Bot. Hort. Floricult. Nat. Sci. 1851 :318, 1851. =Lastreopsis decomposita (R. Brown) Tindale, Vict. Nat. 73 :183. 1957. STUDIES OF FERN TYPES, II—MORTON 247 Type: Cultivated by S. Rucker, Wandsworth, England, 1850 (holotype BM, ex Herb. Houlston, Morton photograph 6769). This species is omitted in Dr. Tindale’s monograph of Lastreopsis, but she wrote me some years ago that ZL. spinescens was probably a synonym of L. decomposita. An examination of the type shows that this is indeed the case. Houlston believed that his species was from the “Fast Indies” but it was surely from Australia. Since the original journal is rare, I give below a transcript of the original description : L. spinescens J. H.—An ornamental, evergreen stove Fern, from the East Indies. Fronds pubescent, deltoid, one and a half to two feet long, dull green, tripinnate ; pinnules linear-lanceolate, pinnatifid, decurrent at the base, with rather ovate slightly dentate segments, terminating in a long spinous mucro. Sori submarginal. Stipes scaly. Fronds lateral, adherent to a scaly creeping rhizome. This species is also rare in cultivation, although introduced ten years ago, among some orchids, by 8S. Rucker, Esq., of Wandsworth. 78. LEPTOSTEGIA LUCIDA D. Don, Prodr. Fl. Nepal. 14. 1825. =Onychium japoni- cum var. lucidum (D. Don) Christ, Bull. Soc. Bot. France 52, Mém. 1:60. 1905. Cheilanthes lucida Wallich, Num. List. no. 69. 1829, nom. nud. Type: Narainhetty, Nepal, Feb. 22, 1803, Buchanan-Hamilton (holotype BM, Morton photograph 6697, right-hand plant; the left-hand plant is Kumaun [Kamaoun], Duthie 3675). There has been some doubt about the identity of this species, the type and only species of the genus Leptostegia D. Don. For a long time it was considered a synonym of Onychium japonicum. In 1934, Christen- sen and Ching, working entirely independently, came to opposed con- clusions. Christensen in the third supplement of the “Index Filicum” took up the name Q. lucidum (D. Don) Spreng as a valid species, with Onychium contiguum (Wall.) Hope as a synonym. Ching on the other hand, recognized O. contiguum as a valid species and placed L. lucida as a variety of O. japonicum. Apparently neither Christensen nor Ching ever saw the type of Leptostegiu lucida, but Ching was right in his placement. Don’s type is not at all like the type of Cheilanthes contiqua Wall., judging by two sheets of Wallich (2 in the U.S. Na- tional Herbarium. They represent a species with the blade very finely cut, with the ultimate segments linear, and the sori short. The type of L. lucida is like Wallich 69, which has broader segments and large sori, these mostly 5 mm. long or more. Ching is very likely right in consider- ing this as merely a variety of O. japonicum, although further study may show that it is specifically distinct from the Japanese plants. The matter is complicated by numerous Chinese specimens, some of which resemble typical O. japonicum and some 0. lucidum. 248 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM 79, LOMARIA BIFORMIS Baker, Journ. Linn. Soc. London 15:415, Oct. 23, 1876. =Blechnum biforme (Baker) Christ, Farnkr, Erde 180. 1897. Blechnum microbasis (Baker) C. Chr. var. biforme (Baker) Tardieu in Humbert, Fl. Madag. Fam. 5, 2:14. 1960. Type: Antananarivo, Madagascar, April, 1876, Pool s.n. (holotype K, 2 sheets, Morton photographs 11422, 11423). Lomaria biformis Baker has priority by four years over L. micro- basis Baker (1880), and so Madame Tardieu has the names backward, since biformis should be the species and microbasis the variety. It seems somewhat doubtful if these species represent more than forms of each other, however, and so the following combination seems pref- erable: Blechnum biforme (Baker) Christ f. microbasis (Baker) Mor- ton, comb. nov. (Lomaria microbasis Baker, Journ. Bot. Brit. & For. 18 :328. 1880. Type: between Tamatave and Antananarivo, Madagas- car, May, 1880, Z. Kitching (holotype K, Morton photograph 11382; isotype B, Morton photograph 10193). 80. LoMARIA DEcoMPsITA D. Don, Prodr. Fl. Nepal. 14. 1825. =Onychium sili- culosum (Desy.) C. Chr. Ind. Fil. 468. 1906. Pteris siliculosa Desv. Naturforsch. Freund. Berlin Mag. 5 :324. 1811. Type said by Desvaux to be from South America, but Weatherby confirms (Contr. Gray Herb, 124:18. 1989) that it represents the Asiatic species known since the time of the “Index Filicum” as Onychium siliculosum, which had been generally called Onychium auratum Kault. Pteris chrysocarpa Hook. & Grev. Icon. Fil. 1: t. 107. 1828. Syntypes: Nepal, Buchanan-Hamilton, and Katmandu, Nepal, Wallich (presumably K but possibly E). These two collections are doubtless duplicates of the same two collections that served as the syntypes of Lomaria decomposita D. Don. Hooker and Greville noted the similarity of their species to that of Don, but remarked that so careful an observer as Don would not have failed to describe the yellow coloration (but Don was not all that careful and, furthermore, his descriptions were very brief). Syntypes: Narainhetty, Nepal, Buchanan-Hamilton (BM, Morton photograph 6696) ; Nepal, Wallich (not seen). The Buchanan-Hamilton specimen is here designated lectotype. In the “Index Filicum,” Zomaria decomposita D. Don is left as a dubious species, perhaps a synonym of Onychium japonicum (Thunb. ) Kunze. It was correctly placed by Ching (Lingnan Sci. Journ. 13 :495. 1934) as a synonym of 0. siliculosum; however, Ching did not desig- nate a lectotype. 81. LoMARIA MARGINATA Schrad. Goett. Gel. Anz, 1824 :871. 1824. =Lomariopsis marginata (Schrad.) Kuhn in von Decken, Reise Ost-Afr. Bot. 3(3) :22. 1879. Acrostichum erythrodes Kunze, Flora 22(1): Beibl. 46. 1839. Type: Ilheos, Bahia, Brazil, Martius 366 (holotype presumably M; isotypes BR, Morton photographs 5077, 5078, and L, Morton photographs 2259, 2260). STUDIES OF FERN TYPES, II—MORTON 249 Lomariopsis erythrodes (Kunze) Fée. Mém. Foug. 2 :67. 1845. Stenochlaena marginata (Schrad.) C. Chr, Ind. Fil, 624. 1906. TypE: Brazil, Prince Neuwied (holotype BR, Morton photograph 19888). In the “Index Filicum,” Christensen correctly placed Lomaria mar- ginata Schrad. under Stenochlaena, and indicated Acrostichum japur- ense Mart. and A. erythrodes Kunze as synonyms, but for some un- explained reason in the first supplement he indicated that LZ. marginata was a Blechnum, and in the third supplement he equated it with Blechnum regnellianum Kunze with a query. It must be that a speci- men in Copenhagen or Stockholm named Z. marginata is really a Blechnum. But the holotype in Brussels is a Stenochlaena, or a Lo- mariopsis as Holttum now delimits these genera. The type itself is annotated by Kunze as equaling his own A. erythrodes, and Kuhn in his transfer of Z. marginata to Lomariopsis also indicates that he has seen the type and that it is the same as A. erythrodes, which I can now confirm. In his brief synopsis of American Lomariopsis (Kew Bull. 1939 :618. 1939), Holttum adopts the name L. erythrodes, pre- sumably not having inquired further into the identity of Z. marginata Schrad.2 Lomariopsis marginata is confined to southern Brazil. The closely allied Z. japurensis (Mart.) J. Smith of northern Brazil and the Guianas differs in having more numerous and closer veins in the sterile pinnae. 82, LOMARIA PYROPHILA Blume, Enum. Pl. Jay. 202. 1828. =Blechnum pyrophilum (Blume) Morton, comb. nov. Type: Summit of the volcano, Ternate, August, 1821, Reinwardt 1712 (holotype L, Morton photograph 731). This species was considered as possibly the same as Lomaria vestita Blume in the “Index Filicum,” and also in Backer and Posthumus, but the type shows that it is different in its fewer pinnae, these rounded at the apex and not long acuminate, in its thicker texture, and in its dark rhachis, this only sparingly scaly. The type of L. vestita is also in Leiden (Java, Blume, Morton photograph 732) ; it shows a much larger plant of thin texture, the pinnae being elongate and long- acuminate, and the rhachis pale and densely scaly. Lomaria vestita Blume var. 8 (Blume. Enum. 203) from Mount Tjerimai, Java, Blume (Morton photograph 733) is less scaly and has the veins not raised beneath, but it is doubtless a form of the same species. The proper status of Z. vestita remains to be determined; it is close to Blechnum procerum (Forst. f.) Swartz, of which it may be a geo- 2 Although I have not seen the types, it seems likely that Lomariopsis speciosa Holttum (loc. cit. Type: Bahia, Luschnath) is the same as L. elongata Fée (Mém. Foug. 2:67. 1845, also described from Bahia, Luschnath), which Holttum does not compare it with. At least, from the description I cannot distinguish them. 250 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM graphical race. It is rather unfortunate that the name Blechnum pyrophilum is so much like B. pyrophyllum Blume, but I think that they must be considered different names, because they have different derivations and meanings and because they differ in spelling by two letters. The epithet “pyrophila” was evidently chosen because of the plant having been found growing by the crater of the volcano on Ternate Island. 83. LOMARIA XIPHOPHYLLA Baker, Journ. Bot. Brit. & For. 22 :142. 1884. =Blech- num biforme var. xiphophyllum (Baker) Morton, comb. nov. Blechnum viphophyllum (Baker) C. Chr. Ind. Fil. 161. 1905. Blechnum simillimum var. viphophyllum (Baker) Tardieu in Humbert, FL Madag. Fam. 5, 2:11. 1960. SynTypes: Madagascar, Humblot 257 and 442 (K, Morton photographs 11391, 11892, 11393). Following Madame Tardieu, it seems best to consider this plant as only varietally distinct from B. béforme, which Madame Tardieu calls B. simillimum. 84. LYCOPODIUM CRASSINERVIUM Desv. Ann. Soc. Linn. Paris 6 :190. 1827. =Selagi- nella crassinervia (Desv.) Spring. in Mart. Fl. Bras. 1(2) :120. 1840. Lycopodium brasiliense Raddi, Pl. Nov. Bras. Nov. Gen. 82, t. 1, f. 1. 1825. Lycopodium pallidum Beyrich ex Gaud. in Freye. Voy. Bot. 1:285. 1827. Not definitely accepted by the publishing author, Gaudichaud, and therefore not validly published. Selaginella muscosa Spring in Mart. Fl. Bras. 1(2) :120. 1840. Type: Macahé, Sebastianopolis, Brazil. Luschnath (presumably BR or M, not seen). Selaginella brasiliensis (Raddi) A. Braun, Ann. Sci. Nat. Paris V, 3:290. 1863, non Spring, 1838. Tyre: “Habitat in Brasilia.” In Alston’s “The Brazilian species of Selaginella” (Repert. Sp. Nov. Fedde 40:313. 1936) the earliest available epithet crassinervia was rejected, apparently on the ground that the type specimen is infected with a fungus. This is true and was even noted in the original descrip- tion by Desvaux, but this does not mean that the name is based on a “monstrosity” and so must be rejected. Many plants are affected more or less by fungi, but if this does not cause so much distortion that the plant is unidentifiable (as in a “witches’ broom”), the names are legitimate. Therefore the name S. crassinervia should be restored for the common south Brazilian species and the name S. muscosa consid- ered asynonym. For additional synonymy, see Alston. 85. Meniscium puncta lunulatum L. C. Rich. Act. Soc. Hist. Nat. [Paris] 1:114. 1792. In the “Index Filicum” this is indicated as a validly published but dubious species, but there is no such thing. The words “puncta lunula- STUDIES OF FERN TYPES, II—MORTON 201 tum” are merely the first words of Richard’s generic description of Meniscium. 86. MENISCIUM SCREBERI L. C. Rich., Act. Soe. Hist. Nat. [Paris] 1:114. 1792. =Thelypteris reticulata (L.) Proctor, Bull. Inst. Jamaica, Sci. Ser. 5:63. 1953. In the “Index Filicum” this is indicated as a nomen nudum, but it is not. The genus Meniscium Schreber dates from 1791 and Richard must have known of it even though he does not cite a reference to it, for he is unlikely to have invented the generic name independently. In fact, it is quite evident that he did know of Schreber’s work, for he names the species after him, misspelling the specific name as “screberi.” Schreber himself did not name a species, and Richard was providing a name. He cites for Jf. screberi “Plumier Fil. ¢. 110,” and since Plumier had both a plate and a full description the species screberi is validly published by reference to this “pre-Linnaean” description, which refers to the plant described later as Polypodium reticulatum L. 87. MICROLEPIA INCISA Fée, Gen. Fil. 328. 1852. —Dennstaedtia obtusifolia ( Willd.) Moore, Ind. Fil. 306, 1861. Type: Gmadeloupe, L’Herminier (Herb Mougeot). In the “Index Filicum” Mierolepia incisa Fée is regarded as a dubi- ous and unplaced species. The present location of the Mougeot fern her- barium has not been ascertained, although Dr. Lellinger has followed several leads. There is in Paris a L’Herminier specimen from Guade- loupe (Morton photograph 21327) that has a label in Fée’s own hand with the name Microlepia incisa Fée. It came from the Houllet Herbar- ium. I do not know anything about B. Houllet (1815-1890), whose herbarium is now partly in Paris and partly in Brussels, but it seems possible that the specimen of this species from the Mougeot Herbarium (or the whole Mougeot Herbarium?) came to Houllet, and that this specimen is actually the holotype of df. incisa. In any case it can be considered the lectotype until another specimen is discovered. It represents the same species as that later described (in 1866) by Fée as Dicksonia incisa Fée, Fée having apparently chosen the epithet “én- cisa” independently for both the Microlepia and the Dicksonia. The latter is Dennstacdtia incisa (Fée) Kuhn, which Tryon considers to be synonymous with D. obtusifolia; it is perhaps not quite identical and it may be that the material from the Lesser Antilles deserves some taxonomic recognition. 88. MIcROSORIUM LONGISSIMUM Fée, Gen. Fil. 268, t. 20B, f. 2. 1852. =Polypodium myriocarpum Mett. Abhandl. Senckenb. Naturforsch. Gesell. 2 :105. 1856. Phymatodes myriocarpa Presl, Tent. Pterid. 198, t. 8, f. 72. 1836, nom. nud. Drynaria longissima J. Smith. Tourn. Bot. 3:397. 1841, nom. nud. 252 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM TYPE: Luzon, Cuming 66 (isotype FI Morton photograph 16016). Under the genus Microsorium, the name MM. longissimum Fée is correct, but the epithet Zongiss¢mum is not available under Polypodium because of P. longissimum Blume (1828). The basis of Presl’s Phyma- todes myriocarpa is unknown, but very likely it was a Haenke collec- tion from the Philippine Islands; there is no description, and the figure given is an analysis only rather than the figure with analyses which would be required for valid publication. Polypodium myrio- carpum Mett. was validly published with a description; it was based partly on Cuming 66 and partly on a Mérat collection from Cochin China. In the “Fern Flora of the Philippines,” Copeland placed Drynaria rubida J. Smith as a synonym of Microsorium longissimum Fée, but this is wrong, a confusion with the quite different Polypodium longissimum Blume, of which DP. rubida has been considered a synonym. The true identity of D. rubida remains to be established ; it was originally a nomen nudum but was validly published as Polypodium rubidum Kunze, later, although Kunze included in his species material from Java collected by Zollinger as well as the original Cuming material of D. rubida J. Smith. A second sheet of Cuming 66 at Firenze (Morton photograph 16015) shows a form with deeply lobed blades; it is just a monstrosity or possibly is a hybrid with one of the regularly pinnatifid species. 89. NEPHRODIUM SPARSUM D. Don, Prodr. Fl. Nepal. 6. 1825. =Dryopteris sparsa (D. Don) Kuntze, Rev. Gen, P1. 2:813. 1891. Polypodium sparsum Buch.-Ham. mss. ex D, Don, loe. cit. Aspidium sparsum (D. Don) Spreng. in L. Syst. Veg. ed 16, 4:106, 1827. Aspidium densum Wallich, Num. List no. 390. 1829, nom. nud. Based on Nepal, Wallich in 1821 (presumably E. Ind. Co. Herb., K). Lastrea sparsa (D. Don) Moore, Ind. Fil. 104. 1858. Aspidium densum Wallich ex Mett. Abhandl. Senekenb. Naturf. Gesell. 2 :349. 1858, pro parte, at least as to Wallich no. 390. Mettenius cited as synonyms Aspidium catophoron Kunze (1848) and Aspidium weigleanum Kunze (1851), both of which were legitimate and available names. Aspidium densum Wallich was therefore an unnecessary name, super- fluous and illegitimate by Art. 63 of the Code. Aspidium densum Wallich ex Mett. is therefore to be typified as a renaming of the oldest available name that ought to have been adopted, namely Aspidium catophoron Kunze and based on the same type as that. Nephrodium densum D. Don was cited not as a straight synonym but as a variant or perhaps possible synonym, according to Mettenius’ usual way of disposing of names that were doubtful or unknown to him. Type: Suembu, Nepal, May 15, 1802, Buchanan-Hamilton (holotype BM, Morton photograph 6448). This species has generally been understood correctly, as by Ching in his treatment of Sikkim—Himalaya PDryopteris (Bull. Fan. Mem. STUDIES OF FERN TYPES, II—MORTON 253 Inst. Biol. Bot. 8:470. 1938), at least as to plants from Nepal and the Himalayas. The species is variable and possibly an aggregate in the wide range stated by Ching—India, Japan, Formosa, China, Indo-China, Philippine Islands, Malesia, and Polynesia. 90. NEPHRODIUM sUBFUSCUM Baker in Hook. & Bak. Syn. 267, 1867. =Thelyp- teris leprieurii (Hook.) Tryon, Rhodora 69 :6. 1967. Dryopteris subfusca (Baker) Kuntze, Rev. Gen. Pl. 2 :813. 1891. Type: Cayenne [French Guiana], Leprieur 23 bis (holotype K!). In his monograph of Dryopteris (p. 167), Christensen remarked on this species and indicated that the type could not be located at Kew; it is there, however, with the name on a label at the side of the sheet where it could be easily overlooked. The name Vephrodium leprieuri Hook. var. a is also on the sheet in Hooker’s hand, and it is indeed possi- ble that this same sheet is also the type of Hooker’s species, although I think not. Hooker says that his plant was collected in marshy woods of central French Guiana by Leprieur, but there is no information of this sort on this sheet. Moreover, Hooker comments especially on the false vein recurrent from the sinus [typical of this section Szezro- pteris |, but this vein is not evident on this specimen. Baker’s V. subfuscum may indeed be a different species from 7’. leprieurii, for it differs not only in the absence of the recurrent false vein, but in having a rhachis that is only slightly pubescent (rather than densely short villous as in /epricurii), and fewer pairs of veinlets (about 8 pairs in subfuscum and about 12 pairs in leprieurii, according to Baker, Syn. Fil. 266). Still, 7’. Zeprieurii appears to be a variable species, at least as treated by Christensen, and Hooker’s var. 8 from Peru appears even more different than subfuscum. 91, NEPHROLEPIS TRICHOMANOIDES J. Smith ex Presl, Epim. Bot. 44. 1849 [1851]. =Arthropteris palisotii var. trichomanoides (J. Smith ex Presl) Morton, comb. nov. Synrypes: Luzon, Cuming 101 and Java, Zollinger 2812. Since the name was attributed to J. Smith, the Cuming specimen is the suitable lectotype, for it was named by Smith. This specimen is presumably in the Presl Herbarium in Prague, but it is not mentioned in Holttum’s recent paper on the Presl types that he found. There are two isolectotypes in Paris (Morton photographs 3613, 3614). The proper citation for this species is omitted in the “Index Filicum” and its supplements; at the reference cited there, “J. Smith in Hook. Journ. Bot. 3:418. 1841,” the name is a nomen nudum. In Copeland’s “Fern Flora of the Philippines,” V. trichomanoides is placed as a synonym of Arthropteris obliterata (R. Brown) J. Smith, but the basis of that (Nephrodium obliteratum R. Brown) is actually a Nephrolepis, the identity of which is being studied by Dr. Jarrett. The 499-018—73——_4 254 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM Philippine specimens in general have the pinnae smaller and narrower (often only 4-7 mm. wide) than in the typical African specimens of A. palisotit, and they are often obviously crenate rather than entire or subentire. They can conveniently be recognized as a variety. The holo- type of Aspidium palisotii Desv., from Oware, Nigeria, Africa, Palisot de Beauvois, is in the Desvaux Herbarium in Paris (Morton photo- graph 8620) ; it consists of a piece of rhizome with one frond attached and a broken part of another frond; it is likely that this was removed from the holotype sheet of Aspidium ramosum Palisot de Beauvois (Fl. d’Oware 2:54, ¢. 91, f. 1.1818), which is a better specimen with six fronds attached to a rhizome, one of these broken in a way that suggests that the broken fragmentary frond on the Desvaux type was removed from this sheet (Morton photograph 3619). There is a good isotype of A. palisotii in the Jussieu Herbarium (Cat. 1106, Morton photograph 2955). 92, PHEGOPTERIS ATROVIRIDIS van Alderw. van Rosenb. Bull. Jard. Bot. Buitenzorg II, 16:26. 1914. =Diplazium atroviride (van Alderw. van Rosenb.) Morton, comb. nov. Dryopteris atroviridis (van Alderw. van Rosenb.) C. Chr. Ind. Fil. Suppl. 2:13. 1917. At this place, Christensen attributes this combination to v. A. v. R., but van Alderwerelt published it only as a synonym and did not ac- cept it; therefore as an accepted combination it must be attributed to Christensen himself. Tyre: Gunong Tanditat, Sumatra, in a damp ravine at 6,000 feet elevation, Jan. 21, 1913, C. G. Mathew 678 (holotype presumably BOG; isotype K, Morton photographs 18912, 18913, 18914). From the isotype at Kew, this is clearly not a Phegopteris or Dry- opteris but a Diplazium, one of the compound-leaved types. The mid- dle pinnae closely resemble those of D. polypodioides Blume, but the lower are quite unlike, being tripinnate, with the ultimate segments elongate and strongly lobed. It is a distinct species thus far known only from the type collection. 93. PITYROGRAMMA TARTAREA var. FALLAX Domin. Véstn. Kral. Ceské Spol. Nauk, Tf. Mat. 1941(15) :8. 1942. =Pityrogramma x distans (Link) Domin, Rozpr. Geské Akad. II, 38(4) :49. 1929. TYPE: “Mexico: San Marcos, M. H. Jones 21. VI. 1892, no. 518a (H. Dom. ex Nat. H. Wash.).” T have received on loan from Pomona College, through the kindness of Dr. Lyman Benson, the specimens of Jones 513 and 518a, both of which bear the same data. Jones 513 is small and juvenile but rather typical P. tartarea (Cav.) Maxon. Joncs 513a, the isotype of var. fallax, is a large, mature specimen that agrees with Domin’s descrip- tion and comments. This specimen is designated lectotype. There is in STUDIES OF FERN TYPES, II—MORTON 255 the U.S. National Herbarium a specimen with the same data and with the number 513; this does not agree with Jones 513 in Pomona College but rather with 518a, and it is thus an isotype of var. fallax, and an “isolectotype.” Domin’s var. fallax is not typical P. tartarea, for the specimen differs in aspect, in the more oblique pinnae and segments, the more acuminate pinnae, and the softer texture. It appears to represent a hybrid between P. calomelanos and P. tartarea, which is known as Pityrogramma x distans (Link) Domin. The original Gymnogramma distans Link was not considered by Link as a hybrid; it was based on cultivated material in the botanical garden in Berlin of unknown origin. From the illustration of an isotype of G. distans given by Domin in his paper on hybrid Pityrogramma, it appears that he is right in considering this taxon as representing the hybrid P. calomelanos X tartarea. This photograph agrees quite well with var. fallax, which probably repre- sents a naturally occurring hybrid. The two species P. calomelanos and P. tartarea are both common and frequently grow in the same areas, and so natural crosses are to be expected. 94, PoLYPODIUM ARGYRATUM Bory ex Willd. in L. Sp. Pl. ed. 4, 5:175. 1810. =Grammitis argyrata (Bory) Morton, comb. nov. Ctenopteris argyrata (Bory) Tardieu, Notul. Syst. 15 :445. 1959. Type: Bourbon Island [Réunion], Bory (holotype B, Herb. Willd. no, 19660, microfiche photograph US). The holotype is a poor specimen, a single detached frond lacking a rhizome. It agrees with a topotype collected in Bourbon by Commerson (P, Herb. Jussieu cat. 1091, right-hand plant, Morton photograph 92946). In their white-ceraceous blades and spongy texture, this species and two others of the Mascarene Islands and Madagascar show a clear and unexpected relationship to several tropical American species such as Grammitis curvata (Swartz) Ching, and the following allied species: G. amylacea (Copel.) Morton, @. fragillima (Copel.) Morton, G. herrerae (Copel.) Morton, and G@. suberassa (Copel.) Morton. 95. PoLYPODIUM aTHYRIOIDES Hook. Sp. Fil. 4:224, t. 277 B. 1862. =Grammitis athyrioides (Hook.) Morton, comb. nov. Ctenopteris athyrioides (Hook.) Copel. Phil. Journ. Sci. 84 :406. 1955. Type: Pangoa, Peru, Mathews 1103 (K). This distinctive species has been known only from the type and two other collections: Rio Lachac, Valle de Lares, Cuzco, Peru, Biies 1822 (US), and Cerros Calla Calla, 26 km. above Leimebamba, on the road to Balsas, Province of Chachapoyas, Department of Amazonas, Peru, 3360 m., Hutchison &:} Wright 6988 (UC). Superficially, this species rather resembles Polypodium chochetangense Rosenst., but this is an 256 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM instance of a parallel evolution apparently, for the latter belongs in Polypodium subg. Polypodium in the group of P. pectinatum L. The rhizome of Grammitis athyrioides has been unknown until the recent collection of Hutchison and Wright. It is short-creeping and about 2 mm. in diameter. The rhizome scales are abundant. They are castaneous, 8-+ mm. long, 0.5-0.6 mm. broad near the base, narrow- lanceate, acuminate, clathrate, the cells in about 8 rows near the base, the cells toward the apex of the scales with thickened walls and obscure lumina, and the scale margins with several conspicuous, brown, stiff, one-celled setae 0.3-0.4 mm. long. 96. POLYPODIUM BINERVE Hook. Sp. Fil. 4:175, ¢t. 273 B. 1862. —Blechnum binerve (Hook.) C. Chr. Cat. Pl. Madag. Pterid. 44. 1932. Lomaria simillima Baker, Journ. Bot. Brit. & For. 22 :141, 1884, Type: Madagascar, Humblot 307 (KK). Lomaria stenophylla Baker, Journ. Bot. Brit. & For, 22:142. 1884. Type: Madagascar, Humblot 305 (K, Morton photograph 11302). Blechnum simillimum (Baker) Diels in Engl. & Prantl, Nat. Pfilanzenfam. 1(4) :248. 1899. Blechnum humblotii C. Chr. Ind. Fil. 155. 1905. Based on Lomaria stenophylla Baker, 1884, non Klotzsch, 1847. Blechnum simillimum f. binerve (Hook.) Tardieu in Humbert, Fl. Madag. Fam. 5, 2:11. 1960. Typr: Madagascar, Lyall (holotype K, Morton Photograph 11303). Madame Tardieu reduced Blechnum binerve to a form of B. similli- mum, but that may not be, for the basionym Polypodium binerve Hook. dates from 1862, whereas the basionym of the other, Lomaria simillima Baker, dates from 1884. Doubtless she was influenced not by the dates of publication of the species but because P. binerve Hook. was based on juvenile plants. The fact that the type of a species is juvenile, however, does not mean that the species loses its priority. It is rather unfortunate perhaps that the type of the species should be an atypical, juvenile plant, but that cannot be helped. Since the juvenile and mature forms of the same species cannot be recognized as different forms, it follows that Lomavia simillima is a synonym of B. binerve. Lomaria stenophylla Baker was based also on a juvenile plant of this species. 97. POLYPODIUM CONJUNCTISORUM Baker, Ann. Bot. 8:129. 1894. —Grammitis conjunctisora (Baker) Morton, comb. nov. ?Polypodium roemerianum Rosenst. Nova Guinea 8:725. 1912. Lectotype: Hellwig Mountains, Dutch New Guinea, 1600-2300 m., November 1909, von Roemer 1227, The other syntype in Leiden, von Roemer 741 (Morton photograph 1969) is a poor specimen bearing the same data as the lecto- type; this sheet is annotated by Rosenstock and it is thus truly the syntype, the published number “744” being a typographical error. The STUDIES OF FERN TYPES, II—MORTON 257 true 744 is a quite different plant, cited by Rosenstock (op. cit. p. 727) as Polypodium rupestre Blume var. leucolepis Rosenst. Xiphopteris conjunctisora (Baker) Copel. Phil. Journ, Sci. 81:94. 1952. Type: Mount Suckling, New Guinea, Macgregor (holotype K, not seen). 98. PoLypopruM FLOccULosuM D. Don, Prodr. Fl. Nepal. 1. 1825. —Pyrrosia flocculosa (D. Don) Ching, Bull. Chin. Bot. Soc. 1:66. 1935. Polypodium acrostichoides Buch.-Ham. ex D. Don, Prodr. Fl. Nepal. 1. 1825, in synon., non Forst. f., 1786. Niphobolus flocculosus (D. Don) Spreng. in L. Syst. Veg. ed. 16, 4:45, 1827. Cyclophorus flocculosus (D. Don) C. Chr. Ind. Fil. 199. 1905. Type: Narainhetty, Nepal, Novy. 18, 1803, Buchanan-Hamilton (holotype BM, Morton photograph 6738). This species has been generally understood. It was well described by Giesenhagen (Die Farngattung Niphobolus 125. 1901). 99. PoLYPODIUM HIRSUTISSIMUM Var. SERICEUM Mart. & Gal. Mém. Acad. Brux. 15:42. 1842. =Polypedium rosei Maxon, Contr. U.S. Nat. Herb. 17 :594. 1916. Tyrer: Near Morelia, Michoacéin, Mexico, Galeotti 6432 (holotype BR, Morton photograph 5024, left-hand plant and right-hand label). In Maxon’s revision of the group of Polypodium squamatum, this variety was overlooked. There is only one sheet at Brussels that could be the holotype. It contains two specimens—a large plant at the right that is Polypodium pyrrholepis (Fée) Maxon and a smaller plant that represents P. rosei Maxon. This smaller plant agrees with the brief diagnosis of Martens and Galeotti. The sheet bears two labels, which are mounted in such a way that they seem to refer to the wrong plants. The label for Galeotti 6482, from Morelia, is the one at the right, whereas the one at the left is Galeotti 6276 from Mirador, Veracruz, referring to the large plant of P. pyrrholepis. The labels must go this way because P. rosei does occur near Morelia, where the type of var. sericeum came from, and P. pyrrholepis does occur at Mirador, in Veracruz, but not near Morelia in western Mexico, Another sheet of P. pyrrholepis in Brussels from Mirador bears the number Galeotti 6432, evidently an error for Galeotti 6276. 100. PoLyPopIUM INCANUM var. FIMBRIATUM Mart. & Gal., Mém. Acad. Brux. 15:36. 1842. —Polypodium thyssanolepis A. Braun ex Klotzsch, Linnaea 20 :392. 1847. Syntypes: Barranca, Guadalajara, Jalisco, Mexico, Galeotti 6438 and 6443. Weatherby stated (Contr. Gray Herb. 124 :34, 1939) that he could not find either of these two numbers in the Galeotti Herbarium in Brussels, and he therefore indicated Galeotti 6443 in Paris as the suitable type (i.e., lectotype). But there is a collection of no. 6438 in Brussels from the Martens Herbarium. This agrees with the original description and is here designated lectotype, on the reasoning that a lectotype ought to be chosen from among the original syntypes rather than 258 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM from an isosyntype. It appears that no. 6438 is a mixture, for Weatherby states that the specimens under this number in Kew and Paris are different, being P, polypodioides var. aciculare Weatherby. 101, PoLypopiuM IRvineii Kuhn, Fil. Afr. 147. 1868. =Polypodium glaucophyllum var. irvingii (Kuhn) Ballard, Kew Bull. Mise. Inf, 1937 :348. 1987. Polypodium glaucophyllum var. B Hook. Sp. Fil. 5:18. 1864. Based on Irving 41 from Abbeokuta, Nigeria, Africa. Type: Abbeokuta, southern Nigeria, Irving 41 (K, holotype, Morton photograph 13895. ) In the “Index Filicum,” P. irvingit Kuhn is indicated as a “nomen.” Kuhn, however, cited P. glaucophyllum var. 8 Hook. as the basis of the species, and since Hooker did give a description of his var. 8, the name P. ervingzi Kuhn must be considered as validly published by a reference to this description and not as a nomen nudum. In proposing var. irvingii, Ballard indicated this as a new variety with himself as author; but since he cites P. trvingti Kuhn as a synonym, this must be considered as a transfer rather than a new variety, and the authority should be “(Kuhn) Ballard.” 102. PoLYPODIUM LIGUSTIFOLIUM Poir. in Lam. Encycl. Méth. 5:553. 1804. =Rumohra adiantiformis (Forst.) Ching, Sinensia 5:70. 1934. TYPE: Buenos Aires, Argentina, Commerson (holotype P, Herb. Lam., Morton photograph 17237). In the “Index Filicum,” Polypodium ligustifolium Poir. was re- ferred to the synonymy of Polystichum denticulatum (Swartz) J. Smith, which is now generally known as Arachniodes denticulata (Swartz) Ching (Acta Bot. Sinica 10:260. 1962). Christensen in his monograph of American Dryopteris (Dansk. Vid. Selsk. Skrift. VIII, 6 :126, 1920) listed P. ligustifolium among the “Species Inquirendae,” with the remark that it was probably Polystichum adiantiforme sens. lat. The type in the Lamarck Herbarium shows that this is right, for it is identical with specimens of Rumohra adiantiformis collected near Buenos Aires by others. This variable species is widespread through- out the world; possibly someone may be able to distinguish segregates sometime. In that case, P. ligustifolium will not be the earliest name for the Argentine plant, for it is antedated by V'ectaria calaguala Cav. (Descr. 252. 1802), described two years previously also on material from Buenos Aires. In the Jussieu Herbarium (Cat. no. 1197B, Mor- ton photograph 2996) is another specimen labeled P. ligustifolium, but not by Poiret. It is a sterile specimen from Montevideo, Uruguay, also collected by Commerson, and represents the same species as the Buenos Aires collection; it is, however, not a type. There are in Firenze two specimens determined as P. ligustifolium that appear to be authen- tic; they are labeled as having been collected by Commerson in the STUDIES OF FERN TYPES, II—MORTON 259 Ile de France, ie., Mauritius, but doubtless they were also really collected either near Buenos Aires or Montevideo (Morton photo- graphs 16476-78). 103. PoLYPoDIUM MEDICINALE Rojas, Bull. Acad. Géogr. Bot. [Le Mans] 28 :156. 1918. =Polypodium phyllitidis L. In the “Index Filicum,” Supplement III, the species is left as dubious. The original description is as follows: Radice fibrosa, gracili, tereti; fronde simplici, subcoriacea, angustato, elongato, integro, glabro, acuminato, basi attenuato, subtus reticulatovenoso, bipedali et ultra; soris sparsis. Foréts et sur les vieux troncs, prés de Cangayam Quitilipi. Intermédiaire entre P. taeniosum H. B. et P. crassifolium de la Jamaique et du Venezuela. Reméde populaire aux missions du Corrientes du temps de Bonpland. An inquiry to the Jardin Botanico in Asuncién, where the type is doubtless conserved, was without response. The species, however, 1s identifiable from the description by the process of elimination, for Polypodium phyllitidis L. is the only epiphytic species in Paraguay with scattered sori and elongate, simple, entire, glabrous blades two feet long and more. N. Rojas Acosta gave the common name as “Cala- cuala,” a name commonly applied to P. phyllétidis and other species of Polypodium with simple blades, as well as apparently to various glabrate species of Elaphoglossum also, the blades of which resemble Polypodium casually. 104. Polypodium normale D. Don, Prodr. Fl. Nepal. 1. 1825. Phymatodes normalis (D. Don) K. B. Presl, Tent. Pterid. 196. 1836. Polypodium longifrons Wallich, Num. List. no. 274. 1829, nom, nud. Polypodium longifrons Wallich ex Hook. & Grev. Icon. Fil. 1:t. 65. 1829. Syntypes: “India orientalis,” Wallich; Nepal, Buchanan-Hamilton. Since the specific epithet was adopted from Wallich, I designate the Wallich specimen (K) as lectotype. Drynaria longifrons (Wallich ex Hook. & Grev.) J. Smith, Journ. Bot. Hook. 3:397, 1841. Drynaria normalis (D. Don) J. Smith, Journ. Bot. Hook. 4:61. 1841. Pleopeltis normalis (D. Don) Moore, Ind. Fil. 347. 1862. Pleopeltis longifrons (Wallich ex Hook. & Grev.) Moore, Ind. Fil. 346. 1862. Colysis normalis (D. Don) J. Smith, Hist. Fil. 100. 1875. Microsorium normale (D. Don) Ching, Bull. Fan. Mem. Inst. Biol. Bot. 4 :299. 1933. Neolepisorus normalis (D. Don) Ching, Bull. Fan. Mem. Inst. Biol. Bot. 10:13. 1940. Neocheiropteris normalis (D. Don) Tagawa, Journ. Jap. Bot. 27:217. 1952. Type: Nepal: Buchanan-Hamilton (holotype BM, a single frond mounted at left on same sheet as the holotype of Polypodium scolopendrium PBuch.-Ham. ex D. Don, Morton photograph 7612). 260 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM From the fact that this species has been referred to so many different segregate genera—Phymatodes, Pleopeltis, Colysis, Microsorium, Neolepisorus, and Neocheiropteris—one might assume that this is very strange species of uncertain relationships, but actually it is rather nondescript without any very distinctive characters. I would suppose that the latest combination proposed, under Neocheiropteris, is prob- ably right if one accepts all the microgenera suggested. Two species may be passing as P. normale; the type has the sori irregularly scat- tered in two or three rows, but there are some plants referred to P. normale in which the sori are strictly in a single row. 105. Polypodium oxylobum Wall. ex Kunze, Linnaea 24 :255. 1851. Polypodium oxylobum Wallich, Num. List. no, 294, 1829, nom. nud. Phymatodes oxyloba K. B. Presl, Tent. Pterid. 196. 1886, nom. nud. Although no description had been published of Polypodium oxylobum Wallich or Phymatodes oxyloba K. B. Presl, Ching adopted the name as validly pub- lished by Presl in 1836. Pleopeitis oryloba (Wallich ex Kunze) Beddome, Ferns So. Ind, ¢. 175. 1863-65. Pleuridium orylobum (Wallich ex Kunze) J. Smith, Ferns Brit. & For. 96. 1866. Polypodium hastatum Thunb. var. orylobum (Wallich ex Kunze) Clarke, Trans. Linn. Soc. II, Bot. 1 :563. 1880. Phymatodes oxyloba (Wallich ex Kunze) Ching, Contr. Bot. Inst. Nat. Acad. Peiping 2 :67, 1933. Wrongly attributed to Presl. Crypsinus oxylobus (Wallich ex Kunze) Sledge, Bull. Brit. Mus. (Nat. Hist.) Bot. 2 :145. 1960. Syntypes: Nepal, Wallich 294; Emodo, near Mossuri, Hiigel: Nilgiris, Schmid- Koch 5 and 151; Nilgiris, Kurr. The valid publication of P. orylobuwm has gen- erally been attributed to Mettenius, but Sledge pointed out that the species was published previously by Kunze. Sledge remarked that the species was based on collections by Wallich and Hiigel and that Kunze also reported it from the Nilgiris, but this is a little misleading. Kunze adopted the name from Wallich, but he did not indicate a type and he referred the Schmid-Koch and Kurr specimens from the Nilgiris to the species without any question, and so they are as much a part of the protologue as the Wallich and Hiigel specimens, Sledge re- moved these Nilgiris specimens to his new species Crypsinus montanus, thus leav- ing the Wallich and Hiigel specimens to typify P. orylobum. Since no lectotype has been formally designated, I choose Nepal, Wallich no. 294 in the East India Company Herbarium, Kew. Kunze’s own original specimen was destroyed in Leipzig. Isolectotype, US. The proper disposition of this species is a little doubtful to me. Sledge transferred it to Crypsinus without any question. But it seems to me that if Crypsinus can be maintained as a group it can only be on the character of the notched margins of the blades, as it is keyed out in Copeland’s key to Polypodiaceae in his “Genera Filicum” and as is required by Copeland’s description of Crypsinus. Copeland did include two species with entire margins in his Crypsinus, noting them as aber- STUDIES OF FERN TYPES, II-—MORTON 261 rant, but their inclusion was not justified by any arguments. Copeland did not mention P. oxylobum, and he presumably included it in Aftero- sorium, which in his opinion included Phymatodes. The supposed re- lationship between P. ovylobum and P. hastatum Thunb., a true Crypsinus, is not obvious, even though Clarke had the former as a variety of the latter. There is perhaps sufficient likeness in a general way to Polypodium scolopendrium Burm. f. to indicate that that might indeed be the relationship. 106. PoLYPODIUM PARVULUM Bory ex Willd. in L. Sp. Pl. ed. 4, 5 :182, 1810. =Grammitis parvula (Bory) Morton, comb. nov. Ctenopteris parvula (Bory) J. Smith, Hist. Fil. 185. 1875. Typr: Bourbon [Réunion], Bory (B, not seen; isotype FI, Morton photograph 16017). This species belongs in Grammitis sect. Cryptosorus (Fée) Fourn. (cf. Morton, Contr. U.S. Nat. Herb. 38:90. 1967) but is less deeply pin- natifid than most species of the section. It is probably confined to the Mascareno Islands, although it has been reported from Africa. It is given the wrong citation of “Ctenopteris parvula (Willd.) Tardieu, Notul. Syst. 15:445. 1959” in the fourth supplement of “Index Fili- cum”; this combination was validly published by J. Smith in 1875. 107. PotypopIuM PLATYNoTUS Kunze ex Zollinger, Syst. Verz. Ind. Arch. 1842- 48 :37. 1854. =Prosaptia alata (Blume) Christ, Ann. Jard. Buitenzorg II, 5 1127. 1905. In the “Index Filicum” this is listed as though it were a validly pub- lished but dubious species, and in Backer and Posthumus’ “Varenflora voor Java” (p. 210. 1939) it is given as a synonym of Polypodiwm obliguatum Blume. This was, however, originally a nomen nudum only, based on Zollinger 3012, from Mount Salak, Java. A specimen in Brus- sels (Morton photograph 20966) is authentic, for it has the right data and the name is in Kunze’s own hand. This specimen shows that the plant is not at all a synonym of P. obliguatum Blume, but rather is identical with Prosaptia alata (Blume) Christ. 108. PoLYPoDIUM PROPINQUUM Wallich ex Mett. Abhandl. Senckenb. Naturf. Gesell. 2:120. 1856. =Drynaria propinqua (Wallich ex Mett.) J. Smith, Cat. Cult. Ferns 13. 1857. Polypodium propinquum Wallich, Num. List. no, 293. 1829, nom. nud. Phymatodes propinqua KX. B. Presl, Tent. Pterid. 198. 1836, nom. nud. Drynaria propinqua J. Smith, Journ. Bot. Hook. 4:61. 1841, nom. nud. Type: Nepal, Wallich List no. 293 (presumably holotype B; isotype US) In the “Index Filicum,” the combination Drynaria propinqua is attributed to J. Smith in 1841, but obviously it is impossible to have a valid combination published in 1841 prior to the valid publication of 262 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM the basionym in 1856. J. Smith republished the combination in 1857 shortly after Mettenius had provided a description. He did not refer to Mettenius, but the basis of his new combination is obvious. 109, PoLyPopIUM PRUINATUM Baker in Hook. & Bak. Syn. Fil. ed. 2, 508. 1874, non Swartz, 1802. =Grammitis pruinosa (Maxon) Morton, comb. nov. Polypodium pruinosum Maxon, Proc. Biol. Soc. Wash. 52:117. 1939. New name for P. pruinatum Baker, non Swartz. Ctenopteris pruinosa (Maxon) Copel. Phil. Journ. Sci, 84:470. 1955. Type: Chontales, Nicaragua, 1867-68, J'ate 44 (holotype K, Morton photograph 15430). There is a fragment of the type collection in US, kindly sent by the Director of Kew to Dr. Maxon. This apparently distinct species is still known only from the type collection. It is a small plant; the fronds are only 1-2 inches long, rather broad, and broadly round-lobed, each lobe having two sori. The blades have minute, white hairs on the upper surface, and the margins are obviously ciliate. 110. POLYPODIUM RADICANS Poir. in Lamarck, Encycl. Méth. 5:530, 1804, non Burm. f., 1768. =Thelypteris reptans (Gmel.) Morton in Steyerm. Fiel- diana Bot. 28 :12. 1951. Type: From “Caroline merid.” in Herb. Lamarck, Paris (Morton photograph 2691). The locality “Caroline merid.,” ie, South Carolina, is an error, and the specimen undoubtedly came from the West Indies. The small stellate hairs present on the surfaces of the pinnae, both above and beneath, show that this species is a synonym of Thelypteris reptans, as it 1s currently recognized, although that may be an aggre- gate, since it is so highly variable. The present plant belongs to the form with elongate leaves radicant at the apex but with the upper pinnae (segments) all fully adnate. The more typical form has the upper pinnae all sessile or even short-petiolulate. Poiret cites Polypodium rhizophyllum Swartz as a synonym, but the species P. radicans Poiret may not be considered as based on that species, for in his comments Poiret says that he merely suspects that his species is the same as that of Swartz. The Poiret name is illegitimate, being a later homonym of P. radicans Burm. f., i.e., Nephrolepis radicans (Burm. f.) Kuhn. 111. PoLypopIuM sAFForpDII Maxon, Amer. Fern Journ. 2:19, fig. 1912. =Gram- mitis saffordii (Maxon) Morton, comb. nov. Polypodium minimum Brack. U.S. Expl. Exped. 16 5, t. 1, f. 3. 1854, non Aublet, 1775. Tyre: Mountains behind Honolulu, Oahu, Hawaiian Islands, Wilkes Exped. (holotype US). This is a species of the sect. Xiphopteris endemic to the Hawaiian Islands, where, apparently, it is not uncommon. STUDIES OF FERN TYPES, II—MORTON 263 112. PoLypPopIUM SCANDENS Léveillé & Rojas, Bull. Acad. Géogr. Bot. [Le Mans] 28 :156. 1918. =Polypodium lycopodioides L. Sp. Pl. 1082. 1753. This species has been omitted from the supplements of the “Index Filicum,” but I think it must be considered validly published. The entire publication is as follows: “Le Dr. Plukenet l’avait nommé Dry- opteris scandens jamaicensis. It serait plutot a rattacher au genre Polypodium: P. scandens Lévl. et Roj.” This is thus a new species based on a reference to a pre-Linnaean publication of Plukenet. The Plukenet reference is obviously to the Almagestum Botanicum 156, 1696, ¢. 290, f. 3. 1694, where the plant appears as Filia sarmentosa bifrons s. Bryopteris [sic! = Dryopteris| scandens jamaicensis inter filicem et Lycopodium media.” Plukenet also quotes as a synonym Phyllitis scandens cauliculis squamosis Plumier t. 42, a reference to Plumier, Descr. Plant. Amér. 29, ¢. 42, 1693. Plukenet has no descrip- tion, but the reference to Plumier’s long description [reprinted and en- larged in Plumier, Tract. Fil. 104., ¢. 179. 1705] serves to validate the name P. scandens Lévl. & Rojas, which is a later homonym of P. scan- dens Forst. f. (1786). The Plukenet figure and the Plumier plate both represent the species Polypodium lycopodioides L., and in fact both references were cited by Linnaeus in publishing this species in 1753. Since Linnaeus had ad- ditional material and cited additional references, however, P. scandens must be considered as a segregate rather than a simple renaming of the Linnaean species. The rather quaint and naive remark of Plukenet that the plant is intermediate between a fern and Lycopodium is due to the numerous spreading scales on the elongate rhizome; when the fronds have fallen, the rhizome thus resembles slightly the leaves and stem of a Lycopodium. 113. PoLypopIuM SCOLOPENDRIUM Buch.-Ham. ex D. Don. Prodr. Fl. Nepal. 1. 1825, non Burn, f., 1768 [as “scolopendria”], =Polypodium leiopteris Kunze, Linnaea 23 :279, 319. 1850. Syntypes : Cultivated from the Hast Indies (Hort. Van Houtte, 1848, and Hort. Lips., 1849, not seen). Lepisorus excavatus (Bory) Ching var. scolopendrium (Buch.-Ham.) Ching, Bull. Fan. Mem. Inst. Biol. Bot. 4:69, 1933. Pleopeltis scolopendrium (Buch.-Ham.) Alston & Bonner, Candollea 15 :207. 1956. Lepisorus scolopendrium (Buch.-Ham.) Tagawa ex Hara, Fl. East. Himal. 494, 1966. Tyre: Nepal, Buchanan-Hamilton (holotype BM, Morton photograph 7612, right-hand plant; the left-hand plant is the holotype of Polypodium normale D. Don). Polypodium scolopendrium was for a long time considered to be only a variety of the African Polypodium excavatum Bory, but recent authors Alston and Bonner, and Tagawa have considered it specifically 264 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM different. The epithet “scolopendrium” has been used, but to me it seems too close to the earlier P. scolopendria Burm. f., a conclusion with which Dr. Stafleu and Dr. Rickett agree. In my opinion P. scolo- pendria Burm. f., although validly published, is incorrect orthographi- cally. It must have been proposed because of the similarity of the simple-bladed fronds to the European Hart’s-tongue, Scolopendriwm vulgare, but if so then the specific epithet ought to have been “scolo- pendrium,” the old generic name unchanged. “Scolopendria” would be the same name in the plural; but specific epithets must be in the nominative singular, and so in my opinion the Burmann species ought to be corrected to Polypodium scolopendrium Burm. f., in which case Buchanan-Hamilton’s later P. scolopendrium would be identical with it. But whether corrected or not the earlier P. scolopendria must make the later P. scolopendrium illegitimate. I have considered P. leiopteris Kunze the same on the authority of Ching; I have not seen any authen- tic material myself. The syntypes were doubtless destroyed in Leipzig. 114, Polypodium trichodes Houlston & Moore, Gard. Mag. Bot. Hort. Flor. Nat. Sci. 3:18. 1851. Since the original publication of this species is in a rather uncommon periodical, I reproduce the description below: An ornamental evergreen stove species from the East Indies. Fronds three to five feet high, hairy, terminal, adherent to a thick, creeping rhizome, bi-tripin- nate, fragile, pale green, with linear-lanceolate pinnae, the pinnules pinnatifid, with somewhat ovate, bluntly lobed segments. Sori round, medial. Stipes and rachis covered with a fine powder; stipes scaly, especially near the rhizome. This fern is in cultivation under the name Lastrea paludosa. The name is attributed to Reinwardt, but there is no real reason to do so. Polypodium trichodes Reinw. ex J. Smith, Journ. Bot. Hook. 3 :394, 1841, was a nomen nudum, and there is reason to think that perhaps J. Smith interpreted Reinwardt’s herbarium name wrongly. Apparently no one has ever really examined Reinwardt’s original specimen to which he applied the name ¢richodes; it is presumably in Leiden, but it was not the basis for Smith’s use of the name or Houl- ston and Moore’s use either. The latter authors do not give the origin of their material, other than that it was cultivated, but Lowe (Ferns: Brit. & Exot. 2:.2. 1872), who also describes the species and attributes it to Reinwardt, says that it appeared spontancously at Kew in 1848. Holttum (Blumea 17:28. 1969) states that the epithet ¢richodes was always a nomen nudum until used by Rosenstock as Dryopteris tri- chodes in 1917, but as shown above the epithet was validly published in 1851. The identity of Houlston and Moore’s species, however, is not quite certain. I could not find any possible type at Kew, but there may STUDIES OF FERN TYPES, II—MORTON 265 be one in Moore’s own herbarium, also at Kew, which I did not con- sult, or there may be one at the British Museum (Natural History), which also has many plants from Moore’s herbarium. Very likely, how- ever, the species is a synonym of 7'helypteris torresiana (Gaud.) Alston. Dryopteris trichodes Rosenst. was based on entirely different material. According to Holttum, the material was mixed, partly 7’. torresiana and partly a related species that he calls Macrothelypteris polypodioides (Hook.) Holttum. Holttum did not select a lectotype. Since Rosenstock stresses the soft hairs and the nonasperous rhachises, his name should apply to the torrcsiana element. Therefore, I designate Java, Zollinger 354 (1) as lectotype, which fixes D. trichodes as a taxonomic synonym of 7’. torresiana. 115. Potypoprum TRIFIDUM D. Don, Prodr. Fl. Nepal. 3. 1825, non Hoffm., 1790. =Polypodium oxylobum Wall. ex Kunze, Linnaea 24 :255, 1851. Type: Nepal, Wallich (holotype BM, with the name in the hand of Don, Morton photograph 7618). The type is the two small plants, trifid and bifid, in the lower left-hand corner; the other large, pinnatifid plants on the type sheet are also Nepal, Wallich, and are doubtless duplicates of the original specimens of Polypo- dium oxylobum Wallich. 116. PoLypopruM UMBILICATUM Poir. in Lam. Encycl. Méth. 5 :528. 1804. =?Dry- opteris filix-mas (L.) Schott, Gen. Fil. ad t. 9. 1834, var. Type: Ile de France [i.e., Mauritius], 1774, Commerson 131 (holotype Herb. Lamarck, P. Morton photographs 2680, 2681). This species is mentioned here merely to point out the need for further study. In the “Index Filicum” it is referred to Dryopteris filiz-mas (L.) Schott, but this species is not reported from Mauritius by Baker in his “Flora of Mauritius and the Seychelles” nor by Madame Tardieu in her list of the ferns of the Mascarene Islands. The type specimen shows that P. wmbilicatum is indeed a close ally of the European D. filix-mas. It is perhaps the same as D. paleacea (Swartz) C. Chr. var. madagascariensis C. Chr. The proper name for the latter is doubtful, since D. paleacea (Swartz) C. Chr. (1911) is an illegiti- mate later homonym of D. paleacea (D. Don) Hand.-Mazzet. (1908). 117. Pteris amoena Blume, Enum. Pl. Jav. 210. 1828. Type: Java, Blume (holotype L, with name in Blume’s hand, Morton photo- graph 2158). In the “Index Filicum” and in Backer and Posthumus’ “Varenflora voor Java” (p. 179. 1939), Péeris amoena Blume is listed as a synonym of P. biaurita L., following Agardh (Rec. Gen. Pterid. 27. 1889), but Agardh placed it there only with a query, for he had not seen any material. The holotype in Leiden shows that P. amoena is not the same as P. biaurita, for the veins are free and not joined to form 266 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM costal areoles as they are in 2. biaurita. In its broad pinnae with cuneate bases and rather broad costal wings (the pinnae not being cut to the base), venation, and other characters it agrees with P. longipinnula Wall. ex Agardh var. b of Holttum’s “Ferns of Malaya” (p. 405. 1954), which differs from typical dongipinnula in having the lowest pinnae basally forked. The typical P. longipinnula, with un- forked basal pinnae, may be known as Péeris amoena var. longipinnula (Wall. ex Agardh) Morton, comb. nov. (Pteris longipinnula Wall. ex Agardh, Rec. Gen. Pterid. 19. 1889. Type: Penang, Wadllich Cat. 108. Agardh saw two specimens of Wadlich 108, one in the Hooker Herbarium and one in the East India Company Herbarium, both now at Kew. I designate the one in the Hooker Herbarium as the lectotype, the sheet bearing the indication “Wallich 1829,” which is the date of receipt of the specimen and not the date of collection, which was 1822). 118. Pteris aspericaulis Wallich ex Agardh, Ree. Sp. Gen. Pter, 22. 1839. Pteris aspericaulis Wallich, Cat. 107. 1828, nom. nud. Pteris quadriaurita var. aspericaulis (Wallich ex Agardh) Bedd. Handb. Ferns Br, Ind. 111. 1883. LecroryPe: The specimen of Wallich 107 in the general herbarium at Kew (Morton photograph 14805) is here designated lectotype. Agardh saw this sheet and also the sheet of Wallich 107 in the herbarium in the Linnean Society in London, which is now in the East India Company Herbarium at Kew. In the “Index Filicum” no description of P. aspericaulis is indicated until that of Hieronymus (Hedwigia 55:348. 1914), but Agardh accepted the species and gave a description in 1839. 119. Pteris deltea Agardh, Rec. Sp. Gen. Pter. 33. 1839. The holotype is in the Hooker Herbarium at Kew, collected in Tahiti (“Otaheite”) by Menzies (Morton photograph 14810). Agardh er- roneously stated that it was collected by “Mathews.” 120. PTERIS ENDLICHERIANA Agardh, Rec. Pterid. 66. 1839. =Pteris comans Forst. Prodr. Fl. Ins. Austr. 79. 1786. Tyre: Agardh cited merely “Hab. in sylvaticis umbrosis Insule Norfolk (Hb. Hookeri!).”” The only specimens in the Hooker Herbarium at Kew that could be the types are two collections from shady woods [“sylvaticis umbrosis’], Norfolk Island, A. Cunningham 39 and 60 (Morton photographs 14922, 14923). These are surely both syntypes, no. 39 being a blade apex (Agardh deseribes the upper pinne from this sheet) and no. 60 being four lower pinne (and Agardh describes these). Dr. Tindale has marked no, 39 as the “holotype,” but it is clear that there is no holotype but two syntypes. I choose no. 60 as lectotype, because it is more identifiable, consisting of fully grown lower pinne, and especially because it has the name endlicheriana in the hand of Agardi; no. 39 does not have the name endlicheriana on it, only Pteris comans. Cunningham did not think his two collec- tions the same species, for he wrote on no. 60 “Habit of no, 39 but distinct” and he may possibly be right. Agardh did not compare his new species with comans but with berteroana. STUDIES OF FERN TYPES, II—MORTON 267 121, Preris 1npica Poir. (var. A) in Lam. Encycl. Méth. 5:712. 1804. =Pteris vittata L. Type: Molucca Islands, Labillardiére (holotype FI-Webb, Morton photograph 16404). In the “Index Filicum,” P. indica Poir. is considered dubious, a syno- nym of either P. longifolia L. or P. moluccana Blume. It was based on two specimens collected by Labillardiére, and the other from the Moluccas; the latter is indicated and described as “variety A,” without a varietal epithet. Both specimens must be considered as syntypes of P. indica, since they were included within the circumscription, but the one indicated as a variety cannot be the lectotype. The Java specimen that must be the lectotype of P. indica was cited by Poiret as being in the Desfontaines Herbarium, which is now incorporated in the Webb Herbarium in Florence, but I did not find it when I was there in 1967. It is doubtless there, however, but some doubt still remains whether it is properly P. vittata (P. longifolia in part of the “Index Filicum”) or P. moluccana. One indication that it may be P. moluccana is pro- vided by a specimen of P. moluccana collected on Didang Island, Indo- nesia, by Gaudichaud, and identified by him as P. indica Poir. (cf. Morton photograph 16408). 122, PrerIS NEVILLEI Baker, Ann. Bot. 5:219. 1891. —Pteris pseudolonchitis Bory ex Willd. in L. Sp. Pl. ed. 4, 5:389. 1810. Type: Bourbon Island [=Réunion], Bory (isotype L, Morton photograph 2219). Tyre: Bourbon Island [=Réunion], @. Neville (holotype K, Morton photograph 14920). This rather characteristic bipinnate-pinnatifid species of the section Litobrochia is confined to the Mascarene Islands. Pteris nevillet was recognized as a valid species in the “Index Filicum.” It was overlooked in Madame Tardiew’s list of the ferns of the Mascarene Islands (Notul. Syst. 16 :162-164. 1960). 123. Preris PECTINATA D. Don, Prodr. Fl. Nepal. 15. 1825, non Cav., 1802. =Pteris biaurita L. Sp. Pl. 1076. 1753, sens. lat. Hototyre: Nepal, Wallich (BM, Morton photograph 6710). A specimen (US) of Wallich 106 from Nepal, originally identified as P. nemoralis Willd. by Wallich, agrees with Don’s holotype and is in all probability an isotype. However, Wallich 106 at Kew is free-veined and is P. quadriaurita Retz., sens, lat.). Wallich ignored P. pectinata Don in his List. There has been confusion about the identity of Pteris pectinata D. Don. Agardh placed it as a synonym of his Pteris aspericaulis Wal- lich,? not adopting Don’s name because of its being a later homonym of ® Pteris aspericaulis Wallich ex Agardh, Rec. Pterid. 22. 1839, is a validly pub- lished name that has never yet appeared in the “Index Filicum” or its supple- ments with its correct earliest citation. In the original “Index Filicum,” the only (Continued ) 268 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM P. pectinata Cav.* Agardh stated that he had seen an authentic speci- men of P. pectinata in the herbarium of the Linnean Society of Lon- don, now the East India Company Herbarium at Kew. This specimen, however, is apparently not authentic, for P. aspericaulis, as described by Agardh, has free veins. The specimen in the British Museum (Natural History) with the name P. pectinata in Don’s own hand is by no means the same thing; Don’s description does not mention the vena- tion, but this holotype shows that the venation is of the Campteria type, 1.e., with a single transverse costal areole between the adjacent lateral midribs of the segments. Thus, P. pectinata D. Don isa synonym of P. biaurita L. in a broad sense, as it applies to specimens from Nepal and the Himalayas generally. This species is characterized also by hav- ing the lowest pair of pinne with a pectinate basal segment, by having raised veinlets beneath, and by entire segments. The holotype sheet bears also the identification P. wallichiana Agardh, but this is an error; the true P. wallichiana also has the venation of Campteria, but the blades are very large and pedately divided, the basal pinne being twice-parted. The veins are not raised beneath and the segments are toothed at the apex. The type of P. wallichiana is P. wumbrosa Wall. List. no. 109. 1829, nom. nud., from Kumaon, India, Wal/ich in Herb. Hook. (K) and East India Company Herbarium (K). Wallich 109, however, is evidently a mixture, for the specimen distributed by the British Museum (Natural History) to the United States National Herbarium does not represent this species, but is P. aspericaulis, it is from Kumaon, collected by R. Blinkworth, and thus evidently mis- numbered; it ought to be no. 106-6. A specimen in Leiden (Morton photograph 2171) distributed as P. wnbrosa Wall. no. 109 is also free- veined; however, it is from Penang and represents the species called P. asperula J. Smith in Holttum’s “Ferns of Malaya”; this sheet (Continued) reference is to Wallich “List,” no. 107, 1829, where it is a nomen nudum; in Sup- plement IT, there is a reference to the later description of Hieronymus in 1914. Pteris aspericaulis cannot be considered as a renaming of P. pectinata D. Don, for as indicated Agardh misunderstood Don’s species; the type must be a speci- men that agrees with Agardh’s description, namely Nepal, Wallich in Herb. Hooker (K) and Herb. Linn. Soc., i.e., East India Company Herbarium (K) ; the specimen in the Hooker Herbarium is here designated lectotype. * Pteris pectinata Cav. is not a dubious species from the Masearene Islands, as indicated in the “Index Filicum,” but represents a species that occurs only in the Andes of South America, namely Belchnum lorense (H.B.K.) Hieron., the origi- nal locality being thus an error. (Cavanilles, however, did not indicate “Masca- rene Islands” but “Marianna Islands,” a very different place indeed.) (cf. C. Christensen, Ark. fér Bot. 9(11) :43. 1910, and Dansk. Bot. Ark. 9(8) :22. 1937). Yortunately, the epithet pectinata cannot now be transferred to Blechnum be- cause of the existence of B. pectinatum K. B. Pres] (1825). STUDIES OF FERN TYPES, II—MORTON 269 doubtless represents Wallich 106-2, which was identified by Wallich as P. nemoralis Willd., which is doubtless the alliance. 124, Pteris subquinata Wallich ex Agardh, Rec. Sp. Pter. 21. 1889. Pteris subquinata Wallich, Cat. 104. 1828, nom, nud. Pteris quadriaurita var. subquinata (Wallich ex Agardh) Bedd. Handb. Ferns Br. Ind. Suppl. 23, 1892. Agardh accepted and described this species, but in the “Index Filicum” no description is indicated until that of Hope in 1901. Agardh saw specimens of Wallich 104 in the Hooker Herbarium and in the herbarium of the East India Company; the better of these two speci- mens should be designated lectotype. 125. PreRIs TRIPARTITA Swartz var. MILNEANA Hook. Sp. Fil. 2 :226, t. 188B. 1858. =Pteris milneana (Hooker) Baker in Hook. & Bak. Syn. Fil. 170. 1867. Syntypes: “The same collectors [i.e., MacGillivray and Milne] and by Dr. Harvey. Society Islands, Nightingale.” There are at Kew collections named var. y [i.e., milneana] by Hooker from Maala, Fiji, M ine (Morton photograph 14910) and Fiji, Harvey (Morton photographs 14911, 14912), from woods above Makerie Harbour, San Cristoval Island, Solomon Islands, Milne 511 (Morton photograph 14909), and Society Islands, Nightingale (Morton photograph 14913). On page 227 Hooker indicated his variety y as from the Feejee [Fiji] Islands, and this would eliminate the San Cristoval Island specimen from being the lectotype (although it was doubtless included in the concept by Hooker). Also on page 297 Hooker commented on the Nightingale specimen as being peculiar and some- what different, and this also eliminates it as lectotype. One might think that the Milne specimen from Fiji would be the logical choice of lectotype, since the variety was named for Milne. Hooker’s drawing of the variety (t. 138B), how- ever, shows the lower pinnae forked, and this Milne specimen has no lower pinnae. The illustration must have been drawn from the Harvey specimen that does have forked lower pinnae, and this specimen is here designated lectotype. In the above discussion I have termed the lower pinnae “forked” for convenience, but they are really not forked ; the lowest inferior seg- ment is elongate and again pinnatifid, in the manner of P. biaurita and P. quadriaurita. The venation is, however, not like these species but is of the fully areolate Litobrochia type. 126, Preris vittosa Baker, Ann. Bot. 5:218. 1891, non Swartz, 1802, nec Fée 1852. —Pteris decurrens K. B. Presl, Del. Prag. 2:1883. 1822. Type: Brazil, Pohl (not seen). Syntypes: Hort. Linden, 1859 and 1861. Lectotype: Hort. Linden, 1859 (K, Morton photograph 14919). Pteris villosa Baker (in the “Index Filicum” erroneously as “Lin- den”) has never been placed, being listed as dubious in the “Index Filicum.” It was said to have been introduced from Assam, India, but no similar species is known from India. In its cutting and pubes- cence the type matches Pteris decurrens K. B. Presl, a rather common 499-018—73——5 270 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM Brazilian species distinguished among other things by having the costae sparsely (or sometimes densely) villosulous beneath. 127. SAGENIA HETEROCARPA Bedd. Ferns Brit. Ind. t. 47. 1866. =Tectaria hetero- carpa (Bedd.) Morton, comb. nov. Nephrodium heterosorum Baker in Hook. & Bak. Syn. Fil. ed. 2. 504. 1874. Based on Sagenia heterocarpa Bedd., non Nephrodium heterocarpum (Blume) Moore, 1858. Aspidium heterocarpum (Bedd.) Bedd. Ferns Brit. Ind. Suppl. 16. 1876, non A, heterocarpon Blume, 1828. Asidium heterosorum (Baker) Bedd. Handb. Ferns Brit. Ind. Suppl. 46. 1892, Tectaria heterosora (Baker) Ching, Sinensia 2:29. 1931. Type: Khasya Hills, India, Thomson (Holotype K). In transferring this species to Nephrodium, Baker was quite right to rename it V. heterosorum, because the epithet heterocarpum was not available under Vephrodium. When Ching transferred the species to Tectaria in 1931, however, he ought to have used the earlier epithet heterocarpa, which was available under the generic name J ectaria, and which must now be restored. 128. SALVINIA LAEVIGATA Humb. & Bonpl. ex Willd. in L. Sp. Pl. ed. 4, 5 :537. 1810. =Limnobium laevigatum (Humb. & Bonpl.) Morton, comb. nov. Trianca bogotensis Karst. Linnaea 24 :424, 1856. Type: Santa Fé de Bogota, Colombia, Humboldt & Bonpland (holotype not seen ; isotype P-Humb. Herb., Morton photograph 3351). It has been known, for instance in the “Index Filicum” and Reed’s “Index Marsileata et Salviniata,” that Salvinia laevigata Humb. & Bonpl. was not a Salvinia but an aquatic floweri ng plant of the family Hydrocharitaceae. The fact that this plant was wrongly ascribed to the ferns does not make the epithet laevigata illegitimate and un- available; since it has priority, it must replace bogotensis as the correct name. It is possible that Z. laevigatum is the same as 1. ydromystria stolonifera G. F. W. Meyer (Prim. Fl. Esseq. 152. 1818), which is Limnobium stoloniferum (G. F. W. Meyer) Griseb (F]. Brit. W. Ind. 506. 1862), but the epithet Zaevigatwm is older than stoloniferum also. 129, SYNAPHLEBIUM PULCHRUM Brack, U.S. Expl. Exped. 16 :228. 1854. =Lind- saea pulchra (Brack.) Mett. Ann. Sci. Nat. IV, 15.65. 1861. SYNTYPES: Tutuila, Samoa, and Sandalwood Bay, Fiji, Wilkes Expedition (US). In the “Index Filicum,” Christensen had an entry “Lindsaya pulchra var.” of Mettenius in 1861 and gave the author of L. pulchra as a species as “Carr. in Seem. FI. Vit. 337. 1873.” It is obviously impossible nomenclaturally to have a variety published 12 years prior to a specific combination. In point of fact, the entry of Mettenius’ is “27. Z [indsaea] STUDIES OF FERN TYPES, II—MORTON 271 pulchra var.? (Synaphlebium Brach. [sic!] expl. 223). ‘Balade’ (1553).” This is in Mettenius’ paper on the ferns collected by Vieillard in New Caledonia, and the locality “Balade” and the number “1553” refer to Vieillard’s collecting locality and collection number. The “var?” refers to Mettenius thinking that this Vieillard collection was a doubtful variety of Z. pulchra, it does not mean that Mettenius con- sidered L. pulchra itself as a possible variety of something else, for it is given a species number and accepted as a valid species the same as the preceding “26. L. lanceolata Billard.” and “27. L. vieillardii n. sp.” Therefore, Mettenius is the author of the specific combination and not Carruthers. 130. Trichomanes collariatum van den Bosch. Nederl. Kruid. Arch. 4:368. 1859. Trichomanes martinezit Rovirosa, Pterid. Sur Mex. 106, ¢. 7A, figs. 1-3. 1909. Type: Rocky banks of the Rio Lacanja, Chiapas, Mexico, Leandro Mar- tinez 1103 (not seen). Type: Tabasco, Mexico, 300 ft., Linden. Van den Bosch cited no number or spe- cific locality in Tabasco. I have seen a specimen from Teapa, Tabasco, Linden 1508 (K, Morton photograph 19047) ; it is very likely an isotype, for it agrees with the description, and also with the description of T. martinezii Rovirosa, which seems surely a synonym. This species has been little known, most specimens probably having been identified as 7’. radicans Swartz, from which it differs in having a definitely two-lipped flaring involucre instead of a truncate one. The fronds are usually short-stipitate, whereas those of 7’. radicans usually are long-stipitate, but there is some variation. Trichomanes collaria- tum occurs from Tabasco and Chiapas in Mexico south to Colombia, and is not at all rare in Central America. 131. Trichomanes crenatum van den Bosch, Ned. Kruid. Arch. 5(8) :205. 1863. Trichomanes crispiforme Alston in Exell, Cat. Vasc. Pl, Sao Tomé 57. 1944, nom, superfl. Based on 7. crenatum van den Bosch 1863, non Gilibert, Pxerc. Phyt. 2 :556. 1792. Lreororype: Niger, West Africa, Barter 1918 (K, Morton photograph 19067). The other syntype is Niger, West Africa, Barter 1917 (K). Alston renamed 7’. crenatum van den Bosch believing that it was an illegitimate later homonym of 7’. crenatum Gilibert. McVaugh, how- ever, has convincingly shown that names published in Gilibert’s “Exer- citia Phytologica” are not validly published by Art. 23 of the Code, because Gilibert did not consistently employ the Linnaean system of binary nomenclature. On the contrary, as McVaugh shows, Gilibert changed some Linnaean binomials into polynomials, such as renaming Anthericum lithago L. as “Anthericum non ramosum” and Lythrum salicaria L. as “Lythrum salicis folio.” Gilibert also described species with names like “Agaricus totus luteus,” “Samolus beccabungae facie,” 272 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM “Jasione scabiosae capitulo,” and so forth. This being so, the name 7’. erenatum van den Bosch is correct and the name 7’. crispiforme Alston is superfluous. 132. TRICHOMANES DEBILE van den Bosch, Nederl. Kruid. Arch. 5(2) :154. 1861. =Trichomanes diaphanum H. B. K. Nov. Gen. Sp. 1:25. 1816. Type: Carabobo, Venezuela, Funck & Schlim 596 (not seen). In the “Index Filicum” this is recognized as a distinct species. I have not seen the type, but there is a sheet at Kew identified by van den Bosch; it contains two collections, one from Ecuador, Jameson, and one from Venezuela, Fendler 459. These plants are similar and both can be matched with rather small, less divided, exiguous forms of 7’. dia- phanum H. B. K., as it is currently recognized. 133. Trichomanes idoneum Morton, nom. nov. Trichomanes gemmatum J. Smith, Journ. Bot. Hook, 3:417. 1841, nom. nud. Based on Mount Ophir, Malacca, Cuming 400. Trichomanes bifidum Pres], Hymen. 43. 1843, non Vent. ex Willd., 1810. Type: Mount Ophir, Malacca, Cuming 400 (holotype presumably PR). Trichomanes gemmatum sensu Copel. Phil. Journ. Sei. 51 :269, 1933, non Baker, 1867. The name 7’. gemmatum J. Smith was never validated by anyone until, for some reason not readily understandable, Baker took it up (in Hooker & Baker, Syn. Fil. 87. 1867 ), citing as synonyms 7’. cel- lulosum Sturm, 7’. filiforme Sturm, 7’. longisetum Brack., 7’. asae-grayi v.d. Bosch, and 7. foeniculaceum Hook. (p.p.). Of these, 7’. Longise- tum is merely a misidentification by Brackenridge (a “sensu” name) and 7’. foentculaceum sensu Hooker was only “pro parte,” but the other three species cited as synonyms were all validly published and legiti- mate, and there was no reason to displace them by accepting a nomen nudum in their place. Thus 7’. gemmatum Baker is a nomen abortivum, a superfluous name. Baker’s description was based probably in part on Cuming 400, the plant assigned the name gemmatum by Smith, but it was based also on varied plants from Venezuela, Brazil, Polynesia, Java, and the Philippines, and so it cannot be said that all these synonyms were cited by Baker in error. Copeland ignored all this and adopted the name gemmatum for a local species native to Malaya and Borneo, saying merely “As Baker takes up the name of J. Smith, the type is the only specimen cited by Smith, Cuming 400, from Malacca.” In fact, Baker did not cite Smith’s publication of 1841, nor did he men- tion Cuming 400. Since 7. gemmatum Baker is a superfluous name, it must be lectotypified on the basis of the earliest name that ought to have been adopted, 7. celluloswm, which Baker cites as “Sturm, Hk. 2nd Cent. Ferns, ¢. 63.” But a check of this reference shows that the STUDIES OF FERN TYPES, II—MORTON | 273 plant described and figured by Hooker is 7’. celduloswm Klotzsch, Lin- naea 18:531. 1844, a species which had been adopted by Sturm in Martius’ “Flora Brasiliensis.” The type of 7’. celluloswm Klotzsch is Schomburgk 1186, from the Kanuku Mountains, British Guiana, and this must therefore be the lectotype of 7’. gemmatum Baker. Holttum in his “Ferns of Malaya” adopted Copeland’s concept of “Z’. gem- matum,” apparently without looking into the nomenclature carefully. One further important point to be mentioned is that Presl described Cuming 400 as a new species, 7’. bifidum Presl, a name which has been overlooked in the “Index Filicum,” presumably because it was de- scribed only in a footnote; Presl’s name is a later homonym of 7. bi- fidum Ventenat ex Willd. (1810). Presl’s description, though brief, is fully adequate to validate his name. 134. TRICHOMANES LONGIFOLIUM Desy. Mag. Naturf. Freund. Berlin 5 :328. 1811. =Trichomanes pinnatum Hedw. Fil. Gen. Sp. t. 4, f. 1. 1799. Type: “In America calidiore” (holotype P ex Herb. Desvaux, Morton photo- graph 22053). In the “Index Filicum,” 7. longifolium Desv. is indicated as a synonym of 7’. crispum L., possibly on the basis of Desvaux’s com- parison of his species with 7’. crispum. An examination of the types, however, shows that the species is clearly a synonym of the common and widespread 7. pinnatum Hedw. 135. TRICHOMANES sTRIATUM D. Don, Prodr. Fl. Nepal. 11. 1825. =?Trichomanes giganteum Bory ex Willd. in L. Sp. Pl. ed. 4, 5:514. 1810. Type: Bourbon [Réunion], Bory (Herb. Willdenow 20216, B, not seen). Type: Nepal, Buchanan-Hamilton (holotype BM, Morton photograph 6578). In the “Index Filicum,” T'richomanes striatum is considered dubious but is referred with a query to 7’. bipunctatum Poir., which is a wild guess, for the type as well as the original description shows a plant widely different from 7’. bipunctatum. The type sheet has an annotation indicating that the plant on the right-hand side is the type of 7’. striatum (which it is) and that the plant at the left is probably the type of Hymenophyllum ramosis- simum D. Don. This it clearly is not, for it does not bear the name in Don’s hand, does not have the locality data, and quite disagrees with the original description; this plant is sterile, whereas Don described the sori and indusia of 17. ramosissimum. This left-hand plant is in fact the sterile part of the holotype of 7’. striatum. Don indicated that he saw both sterile and fertile fronds, for he commented: “fructiferae sunt opaciores et tenuius sectae,” which is true. The holotype shows that 7’. striatum belongs to the group of 7’. radicans Swartz. From the nonalate stipe and rhachis and the somewhat greater division of 274 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM the blade, I take it to be the plant called 7. gigantewm Bory by Cope- land (Phil. Journ. Sci. 51:220. 1988) in his revision of Old World Trichomanes. 'This may well be correct, but neither Copeland nor I have seen the type of that species or any material from the type local- ity. The holotype agrees with a specimen (US) from Mongpo, Sik- kim, 5,500 ft. alt., Oct. 7, 1884, Clarke 36382A, originally distributed as 1’, radicans Swartz, and identified by Copeland as “7. giganteum Bory ?.” 136. TRICHOMANES TENELLUM Hedw. Fil. Gen. Sp. ad ¢. [6], text & f. 1, la, 10d. 1799. =Trichomanes capillaceum L. Sp. Pl. 1099. 1753. Trichomanes trichoideum Swartz, Journ. Bot. Schrad. 1800 (2) :98. 1802; Syn. Fil. 144. 1806. Illegitimate renaming of 7. tenellum Hedwig. Trichomanes trichodes Swartz, Fl. Ind. Oce, 3:174. 1806. Renaming of T. trichoideum. Trichomanes trichoideum Swartz was a renaming of 7’. tenellum Hedw., seemingly due to a misunderstanding. In the unnumbered Hedwig plate the habit illustration of 7. tenellwm is so placed that it might seem to belong with the name 7. pusilum rather than with 7’. tenellum. Since this does not by any means represent 7’. pusillum Swartz (1788), Swartz renamed this plant 7’. trichoideum in 1802, cit- ing 7’. pusillum Hedw. as a synonym. In Hedwig’s plate, however, YT’. pusillum is represented by figs. 5, dg, and 5h, an entirely different plant of the sect. Didymoglossum. Swartz evidently realized his error later and corrected it in his “Synopsis Filicum” by citing 7’. tenellum Hedw. as a synonym of his own 7’. trichoidewm and omitting the “7. pusillum” Hedwig. Since 7’. tenellum was validly published and had priority, however, Swartz ought to have adopted that name and re- duced his own 7’. trichoidewm to synonymy. The matter is not of im- portance, since both names are presumably synonyms of the older LT. capillaceum IL. (Synonyms in italics. New species, new names, and combinations in boldface. Index Page numbers of principal entries in italics.) Aconiopteris obtusa, 240 Acrostichum aemulum, 239 ebeneum, 243 erythrodes, 248, 249 gorgoneum, 239 heterolepis, 216 japurense, 249 lancifolium, 215, 216 neglectum, 219 obductum, 215, 216 tartareum, 243 tomentosum, 216 Adiantum amabile, 216 andicola, 216, 217 cuneatum, 217 var. angustifolium, 216 extensum, 217 lexicanum, 217 multiforme, 216, 217 tenerum var. dissectum, 217 Aleuritopteris farinosa, 232 Allantodia, 232 australis, 232 sylvatica, 231 tenelia, 222 umbrosa, 232 Anogramma eggersii, 242 Arachniodes denticulata, 258 Arthropteris obliterata, 253 palisotii, 254 trichomanoides, 253 Aspidium acuminatum, 246 brachiatum, 217, 218 catophoron, 252 chaerophylloides, 219 densum, 252 heterocarpon, 270 heterocarpum, 270 heterosorum, 270 immersum, 218 Aspidium—Continued mexicanum var. serratum, 219 palisotii, 254 ramosum, 254 sanctum var. portoricense, 238 shepherdii, 245, 246 sparsum, 252 uliginosum, 219 variolosum, 217, 218 weigleanum, 252 zollingerianum, 217, 218 Asplenium, 221, 230 adiantum-nigrum, 219, 220 var. capense, 219 aethiopicum, 220, 221 affine, 227 aspidioides, 235 bulbiferum, 225 caespitosum, 223 caudatum, 221 var. sectum, 220, 222 concinnum, 228 cuneatum, 227 decipiens, 220 denticulatum, 220, 221 distans, 221, 223 erosum, 229 faleatum, 221 f. abbreviatum, 228 var. abbreviatum, 228 var. nitidulum, 226 var. sectum, 220 var, subcaudatum, 221 furcatum var. depauperatum, 220 var. jissulum, 220 var. fragrans, 220 gracile, 222 gracilescens, 223 gueinzianum, 223, 225 hahnii, 223 276 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM Asplenium—Continued indicum, 227, 228 japonicum, 223 var. chattagrammicum, 223 var. elongata, 223 knudsenii, 222 laciniatum, 223-225 laserpitiifolium, 225 lobulosum, 226 longifolium, 226 matangense, 229, 230 monteverdense, 242 mortonti, 242 multijugum, 226 myriophyllum, 242 nitidulum, 222, 224, 226 normale, 226 obtusilobum, 227 planicaule, 224, 225, 227, 228 var. yoshinagae, 228 polyodon, 222, 224 var. knudsenii, 224, 226 var. nitidulum, 226 var. sectum, 220 var. subcaudatum, 221 praemorsum, 221 riedelianum, 228 sectum, 220 semihastatum, 229 stereophyllum, 220 tenellum, 222 tenuifolium, 228 tripartitum, 220, 221 truncatum, 227 unilaterale, 226 unilobum, 229 varians, 224, 225 woodwardioides, 231 yoshinagae, 228 var. planicaule, 227, 228 sect. Sphenopteris, 225 Athyriopsis, 224 Athyrium, 222, 229, 231, 232 accedens var. proliferoides, 237 basilare, 231 fosbergii, 229 gracile, 222 matangense, 229, 230 megistophyllum, 230 melanopodium, 235 meyenianum, 285 Athyrium—Continued nigripes, 222 pycnocarpon, 229 sancti-johannis, 230 setiferum, 222 woodwardioides, 231 Blechnum, 249 biforme, 248, 250 f. microbasis, 248 var. xiphophyllum, 250 binerve, 256 humblotti, 256 loxense, 268 microbasis var. biforme, 248 pectinatum, 268 polystichoides, 231 procerum, 249 pyrophilum, 249, 250 regnellianum, 249 simillimum, 256 f. binerve, 256 var. viphophyllum, 250 souleyetianum, 230 squarrosum, 231 riphophyllum, 250 Bolbitis, 215 neglecta, 215 Brachysorus, 231 woodwardioides, 231 Campteria, 268 Cheilanthes contigua, 247 dealbata, 282 farinosa, 232 lucida, 247 Colysis, 233, 260 hemionitidea, 232 normalis, 259 Coniogramme, 236 falcata, 237 fraxinea, 236 Crypsinus, 260, 261 orylobus, 260 stenophyllus, 238 Ctenitis cheilanthina, 238 meridionalis var. speluncae, 238 Ctenopteris amylacea, 233 argyrata, 255 Ctenopteris—Continued athyrioides, 255 fragillima, 233 herrerae, 283 parvula, 261 pruinosa, 262 subcrassa, 234 Cyclophorus flocculosus, 257 foveolatus, 234 Davallia scabra, 284 villosa, 234 Dennstaedtia obtusifolia, 251 Dicksonia, 251 incisa, 251 Dicranopteris linearis, 241 var. alternans, 242 var. bidentata, 241 var. crassifrons, 241 var. irregularis, 241 var. linearis, 241 var. rigida, 241 palmata, 234 Diplazium, 220, 221, 223, 224, 298-232, 236, 254 accedens, 237 arnottii, 235 atroviride, 254 callipteris, 235, 236 var, undulatum, 236 ecamptocarpon, 221, 223 caudatum, 235 celtidifolium, 235, 236 chattagrammicum, 223 crenulans, 235, 236 falcatum, 236, 237 falcinellum, 229, 230 fosbergii, 229 franconis, 221 fraxineum, 237 gracilescens, 223 hahnii, 223 japonicum, 224 legalloi, 235 lobulosum, 226 longifolium, 226 matangense, 230 megistophyllum, 230 melanopodium, 235 INDEX 277 Diplazium—Continued membranaceum, 230 meyenianum, 235 muricatum, 232 polypodioides, 254 proliferoides, 237 proliferum, 237 var. proliferoides, 237 riedelianum, 228 sancti-johannis, 230 silvaticum, 231 subserratum, 220 undulosum, 236 unilobum, 229 woodwardioides, 231 subg. Pseudoallantodia, 232 Drymoglossum fallax, 237 subcordatum, 2387 Drynaria, 238 hemionitidea, 233 longifrons, 259 longissima, 251 normalis, 259 propinqua, 261 rubida, 252 stenophylla, 238 Dryopteris, 252-254, 258 atroviridis, 254 besukiensis f. confertiloba, 218 f. laziloba, 218 cheilanthina, 238 filix-mas, 265 immersa, 218 meridionalis var. speluncae, 238 paleacea, 265 var. madagascariensis, 265 patula var. chaerophylloides, 219 var. serrata, 219 sancta var. portoricensis, 238 shepherdii, 246 sparsa, 252 stegnogramma var, asplenioides, 239 subfusca, 253 trichodes, 264, 265 Elaphoglossum, 216, 240, 259 alatum, 239, 240 278 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM Elaphoglossum—Continued gorgoneum, 239 heterolepis, 216 lancifolium, 215 obtusum, 240 tomentosum, 216 Epidryopteris lycopodiomus, 240 Gleichenia, 234, 240, 241 bifurcata, 240 dichotoma, 241 fureata, 285 gigantea, 241 glauca, 241 laevigata var. bracteata, 240 lanigera, 241 longissima var. nivea, 242 palmata, 284, 235 truncata, 240 var. bracteata, 240, 241 Gramnitis amylacea, 233, 255 argyrata, 255 athyrioides, 255, 256 conjunctisora, 256 curvata, 255 fragillima, 233, 255 herrerae, 238, 255 parvula, 261 pruinosa, 262 saffordii, 262 scolopendrina, 242 scolopendrioides, 242 subcrassa, 234, 255 sect. Cryptosorus, 261 sect. Xiphopteris, 262 Gymnia pectinata, 232 Gymnogramma, 242 distans, 255 eggersii, 242 falcata, 236 Hemionitis dealbata, 232, 243 falcata, 236 Hydromystria stolonifera, 270 Hymenophyllum, 245 aeruginosum, 243 capillare, 243 crispum, 244 densum, 245 exsertum, 245 Hymenophyllum—Continued fusugasugense, 244 interruptum, 244 micans, 244. molle, 244 myriocarpoides, 244 myriocarpum, 244 nitens, 244 pendulum, 243 polyanthos, 245 ramosissimum, 273 tenellum, 245 tomentosum var, fusugasugense, 244 sect. Mecodium, 245 sect. Sphaerocionium, 245 Lastrea acuminata, 245, 246 atrovirens, 246 immersa, 218 paludosa, 264 sparsa, 252 spinescens, 246, 247 Lastreopsis, 247 acuminata, 245 decomposita, 246, 247 Shepherdii, 246 Lemmaphyllum microphyllum, 237, 238 Lepisorus exrcavatus var. scolopendrium, 263 scolopendrium, 263 Leptochilus neglectus, 215 Leptostegia, 247 lucida, 247 Limnobium laevigatum, 270 stoloniferum, 270 Lindsaea lanceolata, 271 pulechra, 270, 271 vieillardii, 271 Lomaria biformis, 248 decomposita, 248 marginata, 248, 249 microbasis, 248 pyrophila, 249 simillima, 256 stenophylla, 256 > Lomaria—Continued vestita, 249 ciphophylla, 250 Lomariopsis, 249 erythrodes, 249 marginata, 248, 249 Loxogramme involuta, 242 scolopendrina, 242 Lunathyrium sect. Athyriopsis, 224 Lycopodium, 263 brasiliense, 250 crassinervium, 250 pallidum, 250 Macrothelypteris torresiana, 219, 265 Mecodium exsertum, 245 Meniscium, 236, 251 puncta lunulatum, 250 sereberi, 251 Microlepia, 251 incisa, 251 marginata, 234 scabra, 234 villosa, 234 Microsorium, 252, 260, 261 longissimum, 251, 252 normale, 259 Neocheiropteris, 260 normatlis, 259 Neolepisorus, 260 normalis, 259 Nephrodium, 270 densum, 252 heterocarpum, 270 heterosorum, 270 immersum, 218 leprieurii, 253 mexicanum var. chaerophylloides, 219 obliteratum, 253 sparsum, 252 subfuscum, 2538 Nephrolepis, 253 biserrata, 246 radicans, 262 trichomanoides, 253 Niphobolus flocculosus 257 Onychium auratum, 248 INDEX Onychium—Continued contiguum, 247 japonicum, 247, 248 var. lucidum, 247 lucidum, 247 siliculosum, 248 Phegopteris, 254 atroviridis, 254 Phymatodes, 260, 261 hastata, 261 myriocarpa, 251, 252 normalis, 259 oryloba, 260, 261 propinqua, 261 Pityrogramma, 243, 255 calomelanos, 255 calomelanos X tartarea, 255 x distans, 254, 255 tartarea, 243, 254, 255 var. fallax, 254, 255 Pleopeltis, 260 longifrons, 259 normalis, 259 oxytoba, 260 scolopendrium, 263 Pleuridium oxrylobum, 260 Polypodium, 233, 240, 252, 259 acrostichoides, 257 argyratum, 255 athyrioides, 255 binerve, 256 chochetangense, 255 conjunctisorum, 256 crassifolium, 259 elegans, 229 excavatum, 263 flocculosum, 257 glaucophyllum var. irvingii, 258 var. B, 258 hastatum var. oxylobum, 260 hemionitideum, 232, 233 hirsutissimum var. sericeum, 257 incanum var. fimbriatum, 257 irvingti, 258 leiopteris, 263, 264 ligustifolium, 258 longifrons, 259 longissimum, 252 279 280 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM Polypodium—Continued lycopodioides, 240, 263 medicinale, 259 membranaceum, 233 minimum, 262 myriocarpum, 251, 252 normale, 259, 260 obliquatum, 261 oxylobum, 260, 261, 265 parvulum, 261 pectinatum, 256 phyllitidis, 259 platynotus, 261 polypodioides var. aciculare, 258 propinguum, 261 pruinatum, 262 pruinosum, 262 pyrrholepis, 257 radicans, 262 reticulatum, 251 rhizophyllum, 262 roemerianum, 256 rosei, 257 rubidum, 252 rupestre var. leucolepis, 257 saffordii, 262 scandens, 263 scolopendria, 263, 264 scolopendrium, 259, 261, 263, 264 sparsum, 252 squamatum, 257 stenophyllum, 238 taeniosum, 259 thyssanolepis, 257 trichodes, 264 trifidum, 265 umbilicatum, 265 vacciniifolium, 240 subg. Polypodium, 256 Polystichum adiantiforme, 258 denticulatum, 258 Prosaptia alata, 261 Pteris amoena, 265 var. longipinnula, 266 aspericaulis, 266-268 asperula, 268 berteroana, 266 Pteris—Continued biaurita, 265-267, 268, 269 chrysocarpa, 248 comans, 265 decurrens, 269 deltea, 266 endlicheriana, 266 indica, 267 var. A, 267 longifolia, 267 longipinnula, 266 milneana, 269 moluceana, 267 nemoralis, 267, 269 nevillei, 267 pectinata, 267, 268 pseudolonchitis, 267 quadriaurita, 267, 269 var. aspericaulis, 266 var. subquinata, 269 siliculosa, 248 subquinata, 269 tripartita var. milneana, 269 umbrosa, 268 villosa, 269 vittata, 267 wallichiana, 268 sect. Litobrochia, 267 Pyrrosia flocculosa, 257 foveolata, 234 Rumohra adiantiformis, 258 Sadleria polystichoides, 231 souleyetiana, 230 squarrosa, 231 Sagenia heterocarpa, 270 Salvinia, 270 laevigata, 270 Scolopendrium vulgare, 264 Selaginella brasiliensis, 250 crassinervia, 250 muscosa, 250 Selliguea hemionitidea, 232 Stegnogramma, 239 asplenioides, 239 Stenochlaena, 249 marginata, 249 Synaphlebium pulchrum, 270 Tarachia truncata, 227 Tectaria, 270 brachiata, 217 calaguala, 258 heterocarpa, 270 heterosora, 270 variolosa, 217, 218 Thelypteris, 239 asplenioides, 239 dasypoda, 239 immersa, 218 leprieurii, 253 reptans, 262 reticulatum, 236, 251 sancta var. portoricensis, 238, 239 torresiana, 219, 265 sect. Steiropteris, 253 Trianea bogotensis, 270 Trichomanes, 274 asae-grayi, 272 bifidum, 272, 273 bipunctatum, 245, 273 capillaceum, 274 cellulosum, 272, 273 INDEX 281 Trichomanes—Continued collariatum, 271 crenatum, 271, 272 crispiforme, 271, 272 crispum, 273 debile, 272 diaphanum, 272 filiforme, 272 foeniculaceum, 272 gemmatum, 272, 273 giganteum, 273, 274 idoneum, 272 longifolium, 273 longisetum, 272 martinezii, 271 pinnatum, 273 pusillum, 274 radicans, 271, 273, 274 striatum, 278 tenellum, 274 trichodes, 274 trichoideum, 274 sect. Didymoglossum, 274 Xiphopteris conjunctisora, 257 BULLETIN OF THE UNITED STATES NATIONAL MUSEUM CONTRIBUTIONS FROM THE UNITED STaTES NATIONAL HERBARIUM VoLuME 38, Part 7 {end of volume] WILLIAM ROXBURGH’S FERN TYPES By C. V. Morton SMITHSONIAN INSTITUTION PRESS Washington, D.C. 1974 OFFICIAL PUBLICATION DATE is handstamped in a limited number of initial copies and is recorded in the Institution’s annual report, Smithsonian Year. SI PRESS NUMBER 5061. For sale by the Superintendent of Documents, U.S. Government Printing Office Washington, D.C. 20402 - Price $1.70 (paper cover) Stock Number 4701-00117 WILLIAM ROXBURGH’S FERN TYPES C. V. Morton This paper originally was titled “William Roxburgh, His Life, Collections, and Fern Types,” but at the time of his death, in 1972, Mr. Morton left as complete only that portion of the paper dealing with Roxburgh’s fern types. Although he had gathered some materials and references on Roxburgh’s life and collections, time did not permit him to write about them. At those few points in the manuscript where Mr. Morton queried what he had written or where he intended to double-check information, I have done so and thus have removed any ambigu- ity. Fortunately, almost none of these cases proved doubtful. I would like to acknowledge the help of Drs. F. M. Jarrett and R. E. Holttum, who helped Mr. Morton by letter and during his visits to Kew and who also submitted comments on portions of the manuscript.—D. B. Lellinger. 1. ACROSTICHUM ALATUM Roxburgh, Calcutta Journ. Nat. Hist. 4:480. 1844. =Bolbitis appendiculata (Willd.) K. Iwatsuki, Acta Phytotax. Geobot. 18:48, 1959. Type: A specimen in the Brussels Herbarium with the name in the hand of Roxburgh (Morton photographs 5055 and 19999, right-hand plant). Native of the Malay Islands according to Roxburgh, which means either Penang Island or the Molucca Islands, where all of Roxburgh’s specimens came from. In the “Index Filicum,” A. alatum Roxb. is referred to Stenochlaena sorbifolia var. 8, i.e., Lomariopsis cochinchinensis Fée, with which the type of A. alatum has nothing to do; also Roxburgh’s description indicates that his species is not L. cochinchinensis. The sheet cited above is the only one found in any herbarium bearing the name A. alatum in the hand of Roxburgh, and the plant at the right agrees with Roxburgh’s description of the fertile fronds. This has been identified in Baker’s hand as Lomaria alpina Spreng., but it is not that. Roxburgh also describes the sterile fronds, but these have ap- parently been lost. The other plant on the type sheet is a sterile 283 284 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM plant marked “Acrosticum [sic] native place and species un- determined” not in Roxburgh’s hand. This plant appears to have been added at a later date. It does not agree with Rox- burgh’s description of the sterile leaves of his A. alatwm and is to be eliminated from consideration. Since this plant is poor, sterile, and without locality or collector, it is a bit hard to place, but it may be Blechnum aggregatum (Colenso) Tindale from New Zealand. 2. ACROSTICHUM EMARGINATUM Buchanan-Hamilton ex Roxburgh, Calcutta Journ. Nat. Hist. 4:480, t. 27, f. 2. 1844 =Acrostichum aureum L. Sp. Pl. 1069. 1753. LECTOTYPE: A specimen in the Delessert Herbarium, Geneva, with the tag “India Orient. Dr. Roxburgh” and the name Acrostichum emarginatum R. in the hand of Roxburgh; the sheet has a tag “Typus” (Morton photo- graph 16774). The locality is “Delta of the Ganges,’ India, according to Roxburgh. Roxburgh (Hort. Beng. 75. 1814) stated that it was collected in 1796 and grown in the Calcutta Botanical Garden. A Roxburgh drawing in Kew (no. 1743) represents the same plant (Morton photograph 15852), and a specimen is reported to be in Brussels. This species was presumably considered distinct, as indicated by the epithet chosen, on the basis of the pinnae being emarginate at the apex, i.e., somewhat cut in at the apex and with the midrib extending out as a short mucro in contrast with the generally rounded apices of the pinnae. This is a common variant, however, occurring in both the Old and New World, and is perhaps teratological; at least it does not appear to have taxonomic importance. 3. ACROSTICHUM RADIATUM Koenig ex Roxburgh, Calcutta Journ. Nat. Hist. 4:479, 1844. = Actiniopteris radiata (Swartz) Link, Fil. Sp. 80. 1841. LECTOTYPE, chosen by Pichi-Sermolli (Webbia 17:11. 1962): India, Koenig (S-PA). From a comparison with Koenig specimens in other herbaria that seem to be part of the type collection Pichi-Sermolli has determined that the material probably came from Tranquebar, Coromandel, India. Roxburgh did not consider this a new species, but rather an unpublished combination based on Asplenium radiatum Swartz (1802). Swartz had cited Acrostichwm radiatum Koenig as the basis for his new species, but he considered it an Asplenium rather than an Acrostichum. It was Roxburgh’s intention to revert to Koenig’s generic opinion. When Roxburgh was writing, probably prior to 1810, the name A. radiatum had not been published, but it was published subsequently as Acrostichum radiatum Koenig ex Poir. in Lam. Encycl. Méth. Suppl. 1:128. ROXBURGH'’S FERN TYPES—-MORTON 285 1810. Roxburgh cited the same illustration as Swartz, viz. Vahl, Symb. Bot. 1:t. 25. 1790. The typification of Asplenium radiatum Swartz offers some problems, since Swartz did not cite India as a locality, but rather Bourbon Island and Arabia. Since Vahl’s illustration cited by Swartz did not represent the true A. australe L.f. as was intended, it must have been based on the other synonym cited by Vahl, namely Acrostichum dichotomum Forssk. There would thus be some reason to typify Asplenium radiatum on the basis of a Forsskél specimen from Arabia. According to Pichi- Sermolli, originally a specimen of this species was in the Forsskal Herbarium in Copenhagen, but it has now been lost and only an impression remains. Under the circumstances it seems that Pichi-Sermolli was right in assuming that Swartz acciden- tally omitted mention of the locality India, and therefore that a Koenig specimen is the proper lectotype. It is not quite to be expected that plants from India and Arabia would be conspe- cific, but in this case it appears that they are. Only one specimen from Roxburgh has been seen, this in Ox- ford (Morton photograph 20200); it bears the date May 29, 1808. According to Roxburgh, it was collected in the mountains of Coromandel. There is a drawing by Roxburgh of this at Kew (no. 695, Morton photograph 15853); although stylized, the drawing is clearly identifiable as A. radiata. 4. ACROSTICHUM SEETACOONENSE Roxburgh, Calcutta Journ. Nat. Hist. 4:480. 1844. =Bolbitis appendiculata subsp. vivipara (Ham. ex Hook.) Hen- nipm. Blumea 18:146. 1970. Polybotrya vivipara Hook. Exot. Fl. 2:t. 107. 1824. Acrostichum viviparum Buch.-Ham, ex Spreng. Syst. Nat., ed. 16, 4:36. 1827, non L. f., 1781. Polybotrya nodiflora Bory in Bélanger, Voy. Bot. 2:17. 1833. Egenolfia nodiflora Fée, Gen. Fil. 48. 1852. TYPE: A specimen in the Brussels Herbarium with the name in Roxburgh’s hand (Morton photograph 20016); the plant was collected according to Roxburgh in Chittagong, near the Burning Wells at Seetacoond, then in East Bengal and now in East Pakistan. Since this is the only specimen of this species found in Kew, British Museum, Geneva, or Brussels, it may be considered to be unique and therefore the holotype. In his treatment of Egenolfia, Ching (Bull. Fan Mem. Inst. Biol. 2:304. 1981) confused the nomenclature of this species by citing the epithet vivipara as dating from ‘“‘Acrostichum vivipa- rum Hooker, Exot. Flora 2:t. 107. 1827,” thus telescoping the citations for Polybotrya vivipara Hook. and Acrostichum vivipa- rum Buch.-Ham. ex Spreng. Since Acrostichum viviparum was a later homonym, Ching rejected the epithet vivipara, 286 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM but Polybotrya vivipara Hook. was quite valid. Dr. E. Hennip- man is currently revising the genera Egenolfia and Bolbitis in the Asiatic and Malaysian regions and may have a different disposition for this species eventually. 5. ACROSTICHUM SEMIPINNATUM Roxburgh, Calcutta Journ. Nat. Hist. 4:480. 1844. =Tectaria semipinnata (Roxburgh) Morton, comb. nov. Gymnogramme maingayt Baker in Hook. & Bak. Syn. Fil. ed. 2, 517. 1874. Type: Malacca, Malaya, Maingay in 1809 (presumably K). Aspidium maingayi (Baker) Holttum, Gard. Bull. Str. Settl. 5:207. 1931. Tectaria maingayi (Baker) C. Chr. Ind. Fil. Suppl. 3:182. 1934. LECTOTYPE: A specimen in the British Museum with the name in the hand of Roxburgh and the number 2367 [or 23372] (Morton photograph 6800). A native of the Malay Islands, according to Roxburgh, which in this case probably means Penang Island, since plants matching the lectotype closely have been found on Penang but do not occur in the Moluccas, the other locality referred to by Roxburgh as the “Malay Islands.” A specimen matching the lectotype is in the Brussels Herbarium without any name or indication of collector; I feel sure that it is an isotype. In the “Index Filicum,” Acrostichum semipinnatum Roxb. is referred to Leptochilus latifolius (Meyen) C. Chr. with a query, doubtless on the basis of the original description only. The authentic specimen chosen as lectotype shows that this was an error, for the plant is clearly the Malayan species called Tectaria maingayi in Holttum’s “Ferns of Malaya.” A specimen in the U.S. National Herbarium from Penang, Curtis 577, matches the lectotype closely. Roxburgh referred his plant to Acrostichum and Baker his plant to Gymnogramma because of the sori not being round, as in Roxburgh’s and Baker’s ‘Polyodium,” but extended along the veins irregularly and thus appearing acro- stichoid or “gymnogrammoid.” They are exindusiate, as might be expected. Strangely enough, Roxburgh described another species, this one called Polypodium semipinnatum, which repre- sents the same species of Jectaria. He called attention to its resemblance to his Acrostichum semipinnatum. Holttum (Gard. Bull. Str. Settl. 5:207-209. 1931) cites the extensive synonymy of this species, which includes Polypodium heterosorum Baker (1874), Phegopteris subdecurrens Luerss. (1882), and Campylogramme trollit Goebel (1931). He also gives an interesting discussion of the ecology and variation in this species. 6. ADIANTUM CAUDATUM sensu Roxb. Calcutta Journ. Nat. Hist. 4:512, t. 34. 1844. —Adiantum indicum Ghatak, Bull. Bot. Surv. India 5:71, 74. 1963, at least pro parte. I have seen a herbarium specimen in Brussels (“Ind. or.” ROXBURGH'S FERN TYPES—MORTON 287 Roxburgh, Morton photograph 19885) identified by Roxburgh as A. caudatum; it agrees with Roxburgh’s rather good descrip- tion and with another specimen from Roxburgh, no. 364, June 8, 1808, at Oxford (Morton photograph 20198); these both repre- sent the species recently segregated from A. caudatum L. sens. lat. as A. indicum Ghatak. The published Roxburgh drawing, op. cit., t. 34, and the drawing labeled A. caudatum preserved at Kew (Roxburgh drawing 1756, Morton photograph 15854) presumably also represent A. indicum, so far as one can judge from rather crude drawings. Roxburgh gave the range as “Hin- doostan as well as Ceylon,” and doubtless included the true A. caudatum L. from Ceylon in his concept. In the “Hortus Bengalensis,” Roxburgh indicated that the plants cultivated in the Calcutta Botanical Garden had been collected by T. Colebrooke, Esq. Nayar, in his account of Adiantum in India (Bull. Nat. Bot. Gard. Lucknow no. 52. 1961), described an “A. caudatum var. assamicum” (p. 7) and a “var. flabellatum” (p. 8), but these are not validly published, because they lack Latin diagnoses and also lack a citation of a type; in fact no specimens are cited of the new varieties. Judging from the description, it appears that “var. assamicum” is probably A. indicum. Although Ghatak does not mention “var. assamicum,” he must have known of it. “Variety flabellatum” seems to represent A. incisum Forssk. 7. ADIANTUM MICROPHYLLUM Roxburgh, Calcutta Journ, Nat. Hist. 4:513. 1844, non Swartz, 1788. =Adiantum venustum D. Don, Prodr. Fl. Nepal. 17. 1825. TyPE: In the published version of Roxburgh’s A. microphyllum, A. venus- tum D. Don is cited as synonym. Roxburgh could not have added this synonym, since he was dead by the time Don’s “Prodromus” was published, and so it must have been added by the editor Griffith; even so, the published work must be taken at face value, and so A. microphyllum was a superfluous change of name for A. venustum; this being so, it must have the same type, namely, Nepal, Wallich. The only herbarium specimen found bearing the name A. microphyllum Roxb. is in the East India Company Herbarium at Kew (no. 81-4, Morton photograph 14709, the left-hand plant on the sheet); this specimen, authentic for Roxburgh’s concept, agrees with A. venustum D. Don. 8. ADIANTUM PROLIFERUM Roxb. Calcutta Journ. Nat. Hist. 4:512. 1844. —Adiantum incisum Forssk. Fl. Aegypt.-Arab. 187. 1775. Adiantum flagelliferum Wall. Num. List 61, no. 76-5. 1830, nom. nud. This is based on a Roxburgh specimen without locality (Morton photograph 14705) and is not the same as A. flagelliferum Wall. Num. 288 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM List. 4, no. 76. 1829, nom. nud., which is based on a plant collected by Wallich at Sylhet, Assam. LECTOTYPE: “Ind. or.,’ Roxburgh Herb. 2423 (BR, Morton photograph 19888). According to the original description this species came from the Molucca Islands, where it was doubtless collected by Christopher Smith either in Amboina or Honimoa. Adiantum incisum Forssk. has not usually been distinguished from A. caudatum L., but Pichi-Sermolli (Webbia 12:669-678. 1957) showed that it may be distinguished by characters of pubescence and cutting. The exact range of these species needs to be established. The three treatments of the group of Adiantum caudatum, all published within six years by Pichi-Sermolli (Webbia 12:669-678. 1957), Nayar (Bull. Nat. Bot. Gard. Lucknow No. 52. 1961), and Ghatak (Bull. Bot. Surv. India 5:71-77. 1963), came to different conclusions, and none of these papers referred to any of the others. I have not studied the original specimen of A. flagelliferwm Wall., but it may, from the Assam locality, refer to A. indicum Ghatak, if that is distinct from the true A. caudatum; according to Ghatak, A. caudatum has the one-celled hairs of the lower leaf veins hamate [hooked or uncinate] at the apex, whereas these hairs in A. indicum have straight apices; this is very likely true but a bit hard to see in practice. 9. ADIANTUM TENERUM Roxburgh, Calcutta Journ. Nat. Hist. 4:513. 1844, non Swartz, 1788, non L. von Buch, Abhandl. Akad. Wiss. Berlin 1816-17:360. 1819. —Adiantum capillus-veneris L. Sp. Pl. 1096. 1753. LecToTyPe: A Roxburgh specimen in the Brussels Herbarium with the name in the hand of Roxburgh (Morton photograph 19871). According to Roxburgh, the species was collected at the northern boundary of Oude by A. Gott. An isotype is in the East India Company Herbarium, no. 73-9 (Morton photograph 15729). In the “Index Filicum,” A. tenerwm Roxburgh is cited as though it were the same as A. tenerum L. von Buch., but Roxburgh’s species was published independently with no men- tion of von Buch. The von Buch species apparently was a mis- identification of A. capillus-veneris L. Roxburgh certainly did not intend his species as that of Swartz. 10. ASPLENIUM BIPINNATUM Roxburgh, Calcutta Journ, Nat. Hist. 4: 499. 1844. =Diplazium esculentum (Retz.) Swartz, Journ. Bot. Schrad. 1801(2): 312. 1803-4. LECTOTYPE: A Roxburgh collection in the Brussels Herbarium mounted on two sheets, both named bipinnatum in Roxburgh’s hand (Morton photo- graphs 19607, 19608). There are two isotypes in Geneva, labeled “Amboina, Dr. Roxburgh” (Morton photographs 20626, 20627). According to Roxburgh ROXBURGH’'S FERN TYPES—MORTON 289 it is a native of Amboina and introduced into the Botanic Garden in Cal- cutta in 1798, according to Voigt (Hort. Suburb. Calcutt. 735. 1845). Another specimen in Brussels from the Roxburgh Herbarium is named Asplenium hemionitis by Roxburgh, which is probably the name first assigned by Rox- burgh, who later realized that there already was an Asplenium hemionitis L., which is such an entirely different plant with a simple rather than bipinnate blade that Roxburgh could never have confused it with his own collection. A Roxburgh drawing at Kew (no. 2000) also represents this species (Morton photograph 15855). Asplenium bipinnatum Roxburgh was referred wrongly to Asplenium blumei Bergsm. by Mettenius, presumably going only on the description, and correctly to Callipteris ambigua (Swartz) Moore [=Diplazium esculentum] by Moore. 11. ASPLENIUM CICUTARIUM Roxburgh, Calcutta Journ. Nat. Hist. 4: 500. 1844, non Swartz, 1788. =? Asplenium tenuifolium D. Don, Prodr. Fl. Nepalensis 8. 1825, fide Mett. Abhandl. Senckenb. Naturf, Gesell. 3:172. 1859. TyPE: No herbarium specimen named A. cicutarium by Roxburgh has been found. Roxburgh’s plant came from “mountains north of Rohilcund,” i.e., the present Rohilkhand, Division of Agra, Northern United Provinces, India. Roxburgh evidently did not consider this species as new, for he lists it as “Asplenium cicutarium Linn.,”’ but there is no such Linnaean species. There is an Asplenium cicutarium Swartz, but there is really no reason to associate Roxburgh’s species with that American species. Roxburgh does not mention Swartz in this fern work and probably did not know Swartz’ work. Nor does Roxburgh’s description match the description of A. cicutarium Swartz. Under the circumstances, it is best to con- sider A. cicutarium Roxburgh as a new species, as did Mettenius in the paper cited above, Moore in his “Index Filicum,” and Christensen in his “Index Filicum.” The original description is as follows: “Shoots creeping, scaly. Stipes alternate, polished; fronds alternately subtripinnate, as broad as long (6-8 inches high); ultimate divisions subovate, with the anterior margins crenately-dentate, firm, and smooth on both sides.” As may be seen, this is one of Roxburgh’s poorer descriptions, and his plant is not definitely identifiable from it. Mettenius’ guess seems a possibility, for the segments in this are firm and smooth, they are toothed, although whether one could call them “crenate-dentate” is dubious; the fronds are subtripinnate, or usually fully tripinnate, and sometimes nearly as broad as long; the size too is within the size range of A. tenuifolium, which is 290 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM a common species in the area concerned. Moore guessed that Roxburgh’s plant was Asplenium praemorsum Swartz, but this does not seem possible unless one assumes that Roxburgh in- tended “Pinnae as broad as long,” rather than fronds as broad as long as stated in the description. Moore’s identification was retained in Christensen’s “Index Filicum.” Roxburgh’s descrip- tion is so vague that it could possibly apply to a wholly different plant, some species of Diplazium or Athyrium, such as the Athyrium spectabile (Wallich) K. B. Presl of the ‘Index Filicum,” which should perhaps be considered as a Diplazium rather than an Athyrium. It will be necessary to find a specimen from the Roxburgh Herbarium that agrees with the description in order to place this species definitely. 12. ASPLENIUM CORIACEUM Roxburgh, Calcutta Journ. Nat. Hist. 4:497. 1844, non Desv. 1827. =Asplenium macrophyllum Swartz, Journ. Bot. Schrad. 1800(2) :52, 1802. Type: Mauritius, Groendal (presumably S-PA). LECTOTYPE: A specimen in the Brussels Herbarium named A. coriaceum in Roxburgh’s hand, with the number 2407 (Morton photograph 19989). An isotype is in Geneva, marked “India Orient. Malay Islands, Dr. Roxburgh” (Morton photographs 6539 and 16807). Another presumable duplicate is in the East India Company Herbarium, no. 191-3 (Morton photograph 15726, left hand; the right-hand plant is Dindang, Buchanan-Hamilton in 1822) ; the last-named specimen differs from the others in having six rather than three pairs of pinnae, but it is probably a part of the same collection, since this species varies considerably in the number of pinnae, six being about the maximum number and three the minimum. By “Malay Islands” Roxburgh usually intended Penang Island, which may be presumed to be the type locality for A. coriaceum, since the type specimens agree with material collected in Penang. When Roxburgh named his plant A. coriacewm the name was available, but before its publication in 1844 it was a homonym twice over, the same epithet having been used by Desvaux in 1827 and by Bory in 1833 for different plants. It seems to have been a popular name, for both Fée and Baker used the same epithet again in 1852 and 1867, respectively, for still different species. In the original publication is cited ‘“‘Asplenium finlaysonianum Wall. 68, No. 191 (quod nomen delendum),” but this was added not by Roxburgh but by the editor, Griffith, who intended (as shown by the cited page 63) the citation of Wallich List 191-3 (the specimen of A. coriacewm Roxburgh as indicated by Wallich) and not Wallich no. 191 on page 8, which is based on a different collection. Actually Wallich was wrong in placing Roxburcgh’s A. coriaceum under his own A. finlaysonianum, for the plants represent different species. In A. finlaysonianum the ROXBURGH’S FERN TYPES—MORTON 291 veins are partly anastomosing, whereas in A. coriaceuwm Roxburgh they are all free. 13. ASPLENIUM CRENATUM Roxburgh, Calcutta Journ. Nat. Hist. 4:498. 1844, non Desv. 1827. =Diplazium repandum Blume, Enum. Pl. Jav. 2:191. 1828. Type: Java, Blume (holotype L, with the name in the hand of Blume, Morton photograph 979). Diplazium crenatum (Roxburgh) Moore, Ind. Fil. 121, 329. 1859, non Poir. 1811. LECTOTYPE: A specimen in Geneva labeled “India Orient. Malay Islands, Dr. Roxburgh,” and “Typus” (Morton photographs 6538, 16914). There is a duplicate in Brussels from the Roxburgh Herbarium, no. 2412 (no. 2413 in the original description) (Morton photograph 19631) but with the wrong name tag “Asplenitum serrulatum’; as noted under A. serrulatum, the name tags have become switched between Roxburgh’s A. crenatum and his A. serrulatum. For this reason, I choose the Geneva specimen as lecto- type, the specimen and the name that correspond with the described species. A Roxburgh drawing in the British Museum (Morton photograph 15761), labeled “Amboyna,” fixes the type locality, which is stated by Roxburgh as merely “Malay Islands.” It is likely that the specimen was collected by Christopher Smith, who provided Roxburgh with all his collections from Amboina. The group of Diplazium proliferum (Lam.) DuPetit Thouars, recognized as the genus Callipteris Bory by Copeland, is in need of monographic study. It appears that the true D. proliferum from the Mascarene Islands is different from the Malaysian plants in its deeper and slightly different type of cutting of the pinnae. The Malaysian plants are usually all called D. accedens Blume, of which D. repandum Blume is considered a synonym. It may be so. Blume distinguished his D. accedens and D. repandum in the same way that Roxburgh characterized his A. serrulatum and A. crenatum, by the pinnae being remotely serrulate in the one and rounded-crenate in the other. I keep them separate tentatively, for in typical specimens they can be distinguished. However, there are many somewhat intermediate specimens that need further study. In Mettenius’ monograph of Asplenium, both A. crenatum Roxburgh and A. serrulatum Roxburgh are referred to A. porrectum Wallich, which is clearly wrong. 14. ASPLENIUM CULTRIFOLIUM sensu Roxburgh, Calcutta Journ. Nat. Hist. 4:498. 1844, non L., 1753. =Asplenium polyodon Forst. Fl. Ins. Austr. Prodr. 80. 1786. AUTHENTIC SPECIMEN: A specimen in the Brussels Herbarium with the name “Asplenium cultratum?” in the hand of Roxburgh and the number 2408 (Morton photograph 19982). There is a matching specimen in the British Museum collected in Amboina by Christopher Smith from the Rox- 292 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM burgh Herbarium, and this is surely authentic. A Roxburgh drawing of a plant from Amboina, unnamed, is also in the British Museum (Morton photograph 15755), and this agrees well enough with the herbarium speci- mens mentioned. Asplenium cultrifolium L. (Sp. Pl. 1081. 1753) has remained an unknown species. Linnaeus cited his “Hortus Cliffortianus” (p. 474, Asplenium no. 5) and a Plumier reference to a plant from Martinique. There is no specimen in the Linnean Herbarium in London, but if there is one in Clifford’s Herbarium (BM), it ought to be the lectotype rather than a plant known to Linnaeus only from a literature reference. It would be possible to identify the herbarium specimen definitely with material col- lected in the wild, but the Plumier plate cited is so stylized and poorly drawn that it is scarcely identifiable except by guesswork. For this reason the name Asplenium cultrifolium L. has remained out of use in recent years. In the “Index Filicum,” it is con- sidered a Diplazium. Concerning the synonymy of A. polyodon Forst. see my paper on fern types (Contr. U. S. Nat. Herb. 38: 40, 41. 1967). The query in “Asplenium cultratum?” means—as it does in the similar case of “Polypodium ciliatum Roxburgh?”’—not that Roxburgh was doubtful about the identity of his Asplenium cultratum, but that he was doubtful about describing it as a new species under this name. In this case he abandoned the name Asplenium cultratum and instead referred the specimen in his final manuscript to Asplenium cultrifolium L. 15. ASPLENIUM HEMIONITOIDES Roxburgh, Calcutta Journ. Nat. Hist. 4: 498. 1844. =Diplazium tomentosum Blume, Enum. Pl. Jav. 2:192. 1828. Type: Java, Blume (holotype L, a sheet with the name in Blume’s hand and a hand-written description by Blume, Morton photograph 1012). Diplazium burchardii Rosenst. Repert. Sp. Nov. Fedde 4:293. 1907. Type: Lalah Indragiri, between Tjinaco and Pakan Herun, Sumatra, 1906, Burchard (isotype L, Rosenstock Fil. Sumatr. Exs. no. 22, Morton photograph 1010). A reduction to D. tomentosum Blume was made by Rosenstock on the annotation labelh on this sheet in Leiden. LECTOTYPE: A sheet in the Brussels Herbarium with the name in the hand of Roxburgh and the number 2409 (Morton photograph 19635). The locality was cited by Roxburgh as “Malay Islands.” A duplicate specimen in Geneva (Morton photograph 16919) fixes the type locality as Amboina, where it may be presumed to have been collected by Christopher Smith. Asplenium hemionitoides was reduced properly to the synonymy of D. tomentoswm Blume by Moore in his “Index Filicum,” without comment. ROXBURGH'S FERN TYPES—MORTON 293 The lectotype is an unusually small specimen, less deeply cut, but it can be matched by other small specimens of D. tomentosum, which is rather distinctive by its hairy rhachis and the strongly deflexed lowest pinnae. The species appears to be very rare everywhere it occurs. Perhaps no one since Smith has found it again on Amboina. Roxburgh’s name should perhaps have been spelled more properly “hemionitidoides.” Roxburgh’s choice of this name is explained by his comment: ‘In the double lines and involucres, it approaches to Smith’s character of Hemionitis,” which shows clearly that Roxburgh did not know the character of Diplazium, which is to have the involucres double, i.e, back to back. Hemionitis has, of course, no “involucres.” 16. ASPLENIUM LINGUIFORME Roxburgh, Calcutta Journ. Nat. Hist. 4:497. 1844. =? Syngramma alismifolia J. Smith, London Journ. Bot. 4:166. 1845. Lectotype: Singapore, Lobb 11 (K, Morton photograph 20639). TYPE: No Roxburgh material has been found, but it may be in Brussels filed as a Syngramma, The original description is as follows: “Stipes long, and polished; fronds tongue-shaped, entire, smooth. Fructification in numerous, approximated lines, over the whole disk, and ex- tending almost to the margin. (Involucre not visible in the dry specimen). Nat. of the Moluccas.” As may be seen, this descrip- tion is not quite adequate for a definitive determination. The long, shining stipe would exclude any simple-bladed species of Asplenium known in the Moluccas. In the “Index Filicum,” A. linguiforme is referred with a query to Polypodium feet (Bory) Mett., probably on the basis of the description only, but in my opinion the description does not suggest this species very much. It does suggest greatly Syngramma alismifolia in the elongate, shining stipes, tongue-shaped blades, and exindusiate sori extending in close lines almost to the margin of the fronds. This species is not known from the Moluccas, but is expected there since it occurs as far east as the Solomon Islands. The name Syngramma alismifolia is usually cited with the author (K. B. Presl) J. Smith, London Journ. Bot. 4:166. 1845, but this is incorrect. Syngramma alismifolia J. Smith was in- tended as a new species and Diplazium alismifolium K. B. Presl was excluded as a synonym, Smith believing Presl’s species to be indusiate and different generically. It seems from Holttum’s recent examination of Presl’s type in Prague that Diplaziwm alismifolium is really exindusiate and is the same as Syngramma alismifolia J. Smith. Since Presl’s Diplazium 294 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM alismifolium cannot now be transferred to Syngramma because that would create a later homonym, Smith’s name Syngramma alismifolia J. Smith is the oldest and correct name for the species, unless Asplenium linguiforme Roxburgh proves to be the same, in which case Roxburgh’s name will have a year’s priority. Incidentally, it may be mentioned that Selliguea feet Bory, the basionym of Polypodium feei (Bory) Mett., has a wrong citation in the “Index Filicum,” which gives it as “Bory, Dict. Class. 6:587. 1824; 17:18, t. 41,’ which would seem to date it from 1824. The genus does date from 1824, but at that time the species name Selliguea feei was not proposed, nor was it in the subsequent discussion in Dict. Class. 15:344. 1829. The name appears only in the volumes devoted to the plates, and so it should be cited Selliguea feet Bory, Dict. Class. 17:18, ft. 41. 1831; it is published here by a plate with analyses as well as by the reference to the previous descriptions of 1824 and 1829. Bory, in the 1829 paper, complains that Hooker had published the same species as a Ceterach (referring to Ceterach pedunculata Hook. & Grev. Icon. Fil. 1: t. 5. 1827) without mentioning his publication of 1824, but it turns out that Hooker and Greville’s species is different from Selliguea feet. 17. ASPLENIUM MIXTUM Roxburgh, Caleutta Journ. Nat. Hist. 4:499,. 1844, =Diplazium mixtum (Roxburgh) Morton, comb. nov. Asplenium prescottianum Wallich, Num. List. no. 235. 1829, nom. nud. Based on Singapore, 1822, Prescott (presumably in E. Ind. Co. Herb., K). Asplenium acuminatum Wallich, Num. List. no. 205. 1829, nom. nud. Based on Penang, Porter (Presumably in E. Ind. Co. Herb., K). A duplicate is in the E. Ind. Co. Herb. under no. 205B (Morton photo- graph 19588a, right-hand plant). Diplazium prescottianum (Wallich) Moore, Ind. Fil. 156, 234. 1859, nom. nud. Asplenium acuminatum Wallich ex Mett. Abhandl. Senckenb. Naturf. Gesell. 3:225. 1859, non Hook. & Arn., 1832. Type: Penang, Porter (Wallich no. 205) (isotype BR, Morton photograph 19625). Asplenium prescottianum Wallich ex Hook. Sp. Fil. 3:251. 1860. Based on Wallich List no. 235, i.e., Singapore, Prescott in 1822. Hooker also proposed a var. 6 from Singapore, Lobb and a var. y from Penang, Norris, but since these are considered varieties they cannot be con- sidered as syntypes of the species, and therefore the holotype is Wallich no. 235, presumably in the Hooker Herbarium, K. Diplazium silvaticum var. prescottianum (Wallich ex Hook.) Curtis, Journ. Str. Br. Roy. Asiat. Soc. 25:159. 1894. The species name is wrongly spelled “sylvaticum” and wrongly referred to Presl rather than (Bory) Swartz. ROXBURGH’S FERN TYPES—MORTON 295 Diplazium prescottianum (Wallich ex Hook.) C. Chr. Ind. Fil. 237. 1905, wrongly attributed to Moore. Moore did propose D. prescottianum (Wallich) Moore, but since the basionym Asplenium prescottianum Wallich was at the time a nomen nudum, Moore’s name is equally a nomen nudum and invalid; nevertheless, it has been accepted, e.g., by Holttum in his paper on Diplazium in Malaya (Gard. Bull. Str. Settl. 11:94, 1940). Athyrium prescottianum (Wallich ex Hook.) Holttum, Ferns of Malaya 557. 1954. LECTOTYPE:A specimen in the Brussels Herbarium with the name in the hand of Roxburgh (Morton photograph 19629). A second sheet in Brussels with the name mixtum? in a different hand agrees with the lectotype and is surely an isotype. Another isotype is in the East India Company Herbarium under the number 205B (Morton photograph 19583a, left-hand plant) ; it agrees with the specimens in Brussels. According to Roxburgh, this species was collected in Amboina, presumably by Christopher Smith. In the “Index Filicum,” A. mixtuwm was referred to Diplazium silvaticum (Bory) Swartz, following Moore’s lead. Christensen here recognized D. silvaticwm in a very broad sense as including plants from Asia, Polynesia, Australia, and Mauritius, and pos- sibly tropical America. It is likely that the true D. silvaticum is confined to the Mascarene Islands; it differs from those con- fused with it in cutting and, I believe, in toothed scales. Rox- burgh’s type material seems to agree quite well with collections from Penang Island that were described as Asplenium acumina- tum Wallich ex Mett. (an invalid later homonym that was over- looked in the “Index Filicum”’) and which are apparently currently referred by Holttum to Diplazium prescottianum. They represent the form of this without auricles on the pinnae, de- scribed from Penang Island as Asplenium prescottianum var. y by Hooker. Whether these plants are different from the true D. prescottianum from Singapore remains to be determined, which would be done best by field study in Malaya. This might not be easy, however, since the species appears to be excessively rare. So far as I am aware, D. prescottianum has not been re- ported from Amboina or anywhere outside Malaya and Singa- pore, and consequently it is possible that D. mixtum is different; Roxburgh’s plant is different in being smaller and slightly less cut. On the other hand, it is possible that Roxburgh was in error in stating that his plant came from Amboina, and that it really came from Penang, where many of his collections came from. This may be presumed if a search in herbaria and in the field for plants from Amboina of this type should be unsuc- cessful. In any case, however, the epithet mixtum would be older 296 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM than others of the group (except silvaticum), and so the com- bination D. mixtum would be a correct name. 18. ASPLENIUM MONANTHEMOIDES Roxburgh, Calcutta Journ. Nat. Hist. 4: 497. 1844. =? Asplenium normale D. Don, Prodr. Fl. Nepal. 7, 1825. TYPE: No specimen of the species from the Roxburgh collections has been found, but one is said to be in Brussels. The original description is as follows: “Stipes polished, round, with a groove; fronds (6-12 inches high) alternately-pinnate; leaflets delicate, smooth, trapeziform, very obtuse, anterior and exterior margins dentate-serrate; posterior entire. Fructifications in short lines, on both sides of the nerve; where the involucre expands they become round (as in Polypodiwm). Nat. of Chit- tagong.” In the “Index Filicum,” A. monanthemoides is identified without doubt as A. normale, and this appears to be right. The description agrees well enough except for the statement that the sori are round when the indusia are expanded; the sori in A. normale are too long for that. 19. ASPLENIUM MULTIFLORUM Roxburgh, Calcutta Journ. Nat. Hist. 4:499. 1844, =Diplazium multiflorum (Roxburgh) Moore, Ind. Fil. 147, 333. 1859. TYPE: A specimen in the Brussels Herbarium labeled no, 7, A. multiflorum, in the hand of Roxburgh (Morton photograph 19622). Since no other speci- men has been located in any herbarium named by Roxburgh and which agrees with the original description, this may be presumed to be a holotype. According to Roxburgh, it was collected in the “Malay Islands.” Since the type agrees with other material from Malaya, it may be presumed to have been collected on Penang Island where most of Roxburgh’s collections cited as Malay Islands came from. Very likely they were collected by William Roxburgh, Jr., and given to his father. Diplazium multiflorwm (Roxburgh) Moore is listed in the “In- dex Filicum” as a dubious species, and it has never been placed before. It is close to D. malaccense K. B. Pres] as represented by the type material. Some recently collected material by Mrs. B. E. G. Molesworth-Allen and identified as D. malaccense ap- pears as though it may represent something a little different, or this species may be quite variable. There are in Copenhagen two specimens from the F. von Mueller Herbarium identified as Asplenium multiflorum Rox- burgh, but not in Roxburgh’s hand, that do not agree with the original description, since the frond is completely bipinnate and disagrees in other respects. Some years ago I identified these specimens as D. dilatatum Blume, which may or may not be right. ROXBURGH’S FERN TYPES—-MORTON 297 90. ASPLENIUM NIDUS sensu Roxburgh, Calcutta Journ. Nat. Hist. 4:496. 1844. =Asplenium nidus L. Sp. Pl. 1079. 1753. AUTHENTIC MATERIAL: A Roxburgh specimen in the Brussels Herbarium (Morton photograph 19990), which very likely came from Amboina, collected by C. Smith in 1798 (cf. Hort. Bengal. 75. 1814). The species from Amboina are mostly represented by herbarium specimens in Roxburgh’s herbarium now in Brussels, but many of the species from Chittagong and other parts of India are not. Roxburgh evidently understood this species correctly from his description and from the specimen noted above. He recorded the species from “Chittagong, Malay Islands, etc.” 91. ASPLENIUM RETICULATUM Roxburg, Calcutta Journ. Nat. Hist. 4:497. 1844. —Loxogramme avenia (Blume) K. B. Presl, Tent. Pterid. 215. 1836. Grammitis avenia Blume, Enum. Pl. Jav. 2:117. 1828. Type: Java, Blume (holotype L, a sheet bearing a single frond, with the name in the hand of Blume, Morton photograph 1834). Grammitis coriacea sensu Blume, Enum. Pl. Jav. 2:117. 1828, non Kaulf. Lorogramme blumeana K. B. Presl, Tent. Pterid. 215. 1836. Based on the description of Grammitis coriacea sensu Blume, non Kaulf. Type: Java, Blume (holotype L, a sheet with a single frond determined as Grammitis coriacea in the hand of Blume (Morton photograph 1833). LECTOTYPE: Roxburgh, East India Company Herbarium 10-2 (K, Morton photograph 14638, two right-hand plants; the left-hand plant is Grammitis macrophylla Wall. Num. List 10. 1829, nom. nud.). According to Roxburgh, his species occurs in the Malay Islands, which in this case probably means Penang Island. The citation inserted after the name Asplenium reticulatum, namely “Grammitis macrophylla Wall. Cat. 61, No. 10,” was inserted by the editor, Griffith, and does not mean that A. reticulatum was based on G. macrophylla Wall. but merely that Roxburgh’s species is represented in the Wallich Herbarium under no. 10, actually 10-2. There is reported to be a specimen of A. reticulatum Roxburgh in Brussels, which, if found, should be the lectotype. In the “Index Filicum,” A. reticulatum Roxburgh is referred to Polypodium scolopendrinum (Bory) C. Chr., i.e., Loxogramme scolopendrina (Bory) K. B. Pres]. From the characters given by Holttum in his “Ferns of Malaya,” Roxburgh’s species ap- pears to be L. avenia, rather than L. scolopendrina, because it has the midrib raised above rather than beneath. 29. ASPLENIUM SERRULATUM Roxburgh, Calcutta Journ. Nat. Hist. 4:498. 1844, non Cav., 1801. =Diplazium accedens Blume, Enum. Pl. Jav. 2: 192. 1828. Type: Java, Blume (presumably L, but not found by me in 1954). 298 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM Diplazium serrulatum (Roxb.) Moore, Ind. Fil. 167, 337. 1860, non Desv., 1827. LECTOTYPE: A specimen in the Geneva Herbarium from “India Orient. Amboyna, Dr, Roxburgh” and marked “Typus” (Morton photographs 6537, 16918). An isotype is in the Brussels Herbarium with the wrong tag reading Asplenium crenatum and the number 2413 (Morton photograph 19602). Roxburgh cited the locality as “Amboyna and other Malay Islands.” Roxburgh described his Asplenium serrulatum as having the margins “remotely serrulate’ and his A. crenatum with the margins “elegantly crenate.” The specimens in the Brussels Her- barium have the name labels reversed, for the one with the serrulate pinnae in labeled A. crenatum and the one with crenate pinnae is labeled A. serrulatum. It seems clear that the name tags have been accidently reversed, either by Roxburgh or some later botanist. This is clearly shown by the specimens in Geneva where the proper plants and names do correspond. For this reason I am designating the specimens in Geneva as lectotypes, rather than those in Brussels, as I do in most cases where the Brussels specimen has a name in Roxburgh’s hand. It is likely that A. serrulatum Roxburgh is typical D. accedens Blume, for Blume described his species as having serrate pinnae. 23. ASPLENIUM TRAPEZIFORME Roxburgh, Calcutta Journ. Nat. Hist. 497. 1844. = Asplenium unilaterale Lam. Encycl. Méth. 2:305. 1786. Type: Ile de France [i.e., Mauritius], Commerson (holotype P. Herb. Lam., Morton photograph 2771). Asplenium trapeziforme Roxburgh ex Wallich, Num. List. no. 2213, nom. nud. This entry referred to a plant in “Herb. Roxb.” and is the speci- men now in Brussels. LECTOTYPE: A specimen in the Brussels Herbarium determined as A. trapeziforme in the hand of Roxburgh, with the number 2408 and the locality “Malay Islands.” The specimen had been named first “Asplenium monan- themum,” and the epithet “monanthemum” crossed out (Morton photograph 19994). This lectotype is probably from Penang Island, a locality sometimes intended by Roxburgh when he wrote “Malay Islands”; the specimen agrees with plants collected in Penang. This species never has been identified properly. In Moore’s “Index Filicum,” A. trapeziforme Roxburgh is recognized as a distinct species, with the range “Malay Islands; India (Bombay, Mahabeleshuar); Bourbon.” In Christensen’s “Index Filicum,” the species is wrongly credited to Wallich (who merely published Roxburgh’s name as it was indicated on the specimen he re- ceived, his number 2313) and referred to “A. lunulatum var.?,” evidently going solely on Roxburgh’s description. The African A. lunulatum Swartz is indeed not very close. In “Supplement III,” Christensen referred A. trapeziforme to A. inaequilaterale ROXBURGH’S FERN TYPES—MORTON 299 Willd., citing the description and illustration in Beddome’s “Ferns of Southern India,” p. 45, t. 184. Beddome identified the plant that he illustrated as A. trapeziforme Roxburgh only with a query; the plant described and drawn does apparently represent A. inaequilaterale Willd. but is quite different from the true A. trapeziforme Roxburgh. Alston saw and photographed in Brussels the type of Roxburgh’s species and made a note ac- companying his photograph that A. trapeziforme is “a good species near A. camptorhachis Kunze.” I cannot agree with that; I have seen authentic material of A. camptorhachis from southern India, and it is quite different. It is possible that A. trapeziforme is indeed different from typical A. wnilaterale Lam., which seems to have a more slender rhizome with scattered fronds, but Roxburgh’s species is not different from material from Asia that is commonly referred to A. wnilaterale (as in Holttum’s “Ferns of Malaya”), as for instance the specimen Penang, Curtis 567 (US), in which the rhizome is thicker, the roots more numerous and thicker, and the fronds more numerous and congested (with- out being at all fasciculate). If this plant is really a different species, it is unlikely that Roxburgh’s name is the oldest, for there are several reputed synonyms of A. uwnilaterale that are older, including several originally from Java, described by Blume. Sledge mentions A. trapeziforme in his discussion of A. inaequi- laterale Willd. and excludes it from that species, remarking that “it also has an erect rhizome but photographs in the British Museum Herbarium of the type specimen, including enlargements of the pinnae, clearly show the veins running to the extremities of the marginal teeth. The fronds are also pinnate to the end, not terminating in a distal, lobed, pinnae-like extremity.” These remarks apply to A. unilaterale Lam. sens. lat., except for the statement that the rhizome is erect, which is a wrong observation, due to having studied only a photograph. 24. ASPLENIUM TRIPINNATUM Roxburgh, Calcutta Journ, Nat. Hist. 4:500. 1844. Type: A specimen in the Brussels Herbarium with the name in the hand of Roxburgh and the number 314 (Morton photograph 19988). Since no other specimen of this species has been found from the Roxburgh Herbarium, this may be presumed to be a holotype. Roxburgh indicated that his species was from the Moluccas, which means that he received it from Christopher Smith. There is a specimen in the British Museum collected in Amboina by Christopher Smith (Morton photograph 19511) that may well be authentic material; it is a more divided plant than the type but may be within the range of variation. If this is authentic, it would fix the type locality as Amboina. 300 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM In Moore’s “Index Filicum,” Asplenium tripinnatum is referred to A. laserpitiifolium Lam., and it does represent that species as it has usually been conceived in a very broad sense. Tardieu- Blot and Ching (“Revision des Espéces confondue avec l’Asplenium laserpitiifolium Lam. avec Description d’Espéces Nouvelles Asiati- ques de ce Groupe,” Notul. Syst. 5:184-154. 1936), however, showed that the true A. laserpitiifolium Lam. is apparently con- fined to the Seychelles Islands. Their illustration of the type shows that the species is not the same as A. tripinnatum. Tardieu-Blot and Ching described several species of this group from Indo- China—A. confusum, A. sublaserpitiifolium, A. pseudolaserpitit- folium, and A. neolaserpitiifolium—but none of these appears identical with A. tripinnatum. The common species of this alliance in Java, and perhaps also in Sumatra, New Guinea, and the Ad- miralty Islands, is A. robustum Blume (Enum. Pl. Jav. 2:189. 1828. Type: Salak, Java, Blume, holotype L, with the name in the hand of Blume, Morton photograph 543). This is a rather leafy plant, with the segments mostly obovate. Asplenium tripinnatum is more skeleton-like, with more distinct segments, these appearing oblanceolate or almost linear, due partly to the margins being somewhat inrolled. I am not sure that it is truly different from A. robustum, but it can stand as a distinct species tentatively. 25, ASPLENIUM VARIUM Roxburgh, Calcutta Journ. Nat. Hist. 4:498. 1844. =Diplazium crenatoserratum (Blume) Moore, Ind. Fil. 325. 1859. Diplazium roxburghii Moore, Ind. Fil. 176, 337. 1859. Asplenium crenatoserratum Blume, Enum. Pl. Jav. 177. 1828. Type: Bantam Mountains, Java, Blume. Not located in Leiden in 1954. There is a specimen labeled A. crenatoserratum in Blume’s hand (Morton photograph 960) which may be an isotype, but it lacks the locality “Bantam” and is sterile, whereas Blume clearly describes the sori; moreover, it does not agree very well with the description. Strangely enough, there is another specimen named A. crenatoserratum which was probably collected by Blume, but it is Coniogramme fraxinea (Morton photograph 771). It does not seem possible that Blume could have confused this widely different plant with an Aspleniwm (or Diplazium) ; it seems more likely that the labels have become mixed. A search should be made in Leiden under Coniogramme for a specimen of Diplazium agreeing with the description of A. crenatoserratum but with a label reading Gymnogramma javanica Blume or Gymnogramma serrulata Blume, the names under which Coniogramme fraxinea was originally described from Javan material. Diplazium phanerotis Kunze, Bot. Zeit. 4:443. 1846. Type: Java, Zol- linger 1491 (holotype G, Morton photograph 3830). Diplazium roxburghii Moore, Ind. Fil. 176, 337. 1859. Based on Asplenium varium Roxburgh, non Diplazium varium Gaud., 1827. Diplazium varium Gaud. is also from the Moluccas and is still a dubious species. In ROXBURGH’'S FERN TYPES—MORTON 301 Christensen’s “Index Filicum,” D. roxburghii Moore is left in limbo, so to speak, neither accepted nor placed in italics as a dubious species. Asplenium porrectum Mett. Abhandl. Senckenb. Naturf. Gesell. 3:220. 1859. Based on Asplenium porrectum Wallich Cat. 204, Asplenium porrectum Wallich Cat. 224, Asplenium multisorum Wallich Cat., p. 63, Asplenium polyodon Wallich Cat., p. 63, Asplenium cataractarum Moritzi, and Diplazium phanerotis. The Wallich names were nomina nuda; Asplenium cataractarum must be considered as sensu Moritzi, non Blume, 1828; Diplazium phanerotis Kunze was a validly and previously published name. The epithet phanerotis was available under the genus Asplenium, and Mettenius should have adopted it and made the new combination “Asplenium phanerotis (Kunze) Mett.” instead of taking up the nomen nudum A. porrectum Wallich. Therefore, Asplenium porrectum Mett. must be considered a superfluous name by Art. 63 of the Code and be typified by the type of the name that ought to have been adopted, namely, Diplaziwm phanerotis. (Type: Java, Zollinger 1491). Mettenius saw Zollinger 1491, but he wrongly cites it as from Malacca rather than Java. There are some other errors in Mettenius’ treatment of this species. Wallich used the name Asplenium porrectum twice, first under no. 204 for a plant from Penang and Singapore (wrongly cited by Mettenius as from Nepal) and again under no. 224 for a plant from Mauritius which is utterly different. Wallich realized this later and rejected both of his porrectum names, on p. 63 of his List, renaming no. 204 as Asplenium multisora- tum Wallich (wrongly cited by Mettenius as A. “multisorum”) and no. 224 as Asplenium polyodon Wallich. The latter, A. polyodon Wallich [not A. polyodon Forst., 1786], is still a nomen nudum, considered a synonym of A. protensum Schrad. Type: A specimen in the Brussels Herbarium with the name A. varium in the hand of Roxburgh and the number 2409 (Morton photograph 19600). According to Roxburgh it was collected in Amboina, which means probably by Christopher Smith. Since duplicates of this collection have not been seen in other herbaria, it is likely that this specimen is unique and a holotype. 26. ASPLENIUM WOODWARDIOIDES Roxburgh, Calcutta Journ. Nat. Hist. 4:500. 1844, non Bernh., 1803. =? Diplazium maximum (D. Don) C. Chr. Ind. Fil. 235. 1905 (as to basionym, excl. synonymy). Type: Not determined. The type came from Chittagong, East Bengal (now East Pakistan), where it was collected by Buchanan-Hamilton. No specimen from the Roxburgh Herbarium determined as A. woodwardioides has been located nor any collected at Chittagong by Buchanan-Hamilton. Don described two species from Nepal, Asplenium latifolium Don and A. maximum Don, both from collections made by Wal- lich. In the “Index Filicum,” these were united under the name Diplazium maximum, but it is clear from Don’s descriptions that there were two different species—A. latifolium D. Don (non Bory, 1803) having the pinnules merely toothed or slightly lobed (i.e., the same species or closely related to D. dilatatum Blume) and A. maximum D. Don with the pinnules deeply 302 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM pinnatifid (clearly related to D. polypodioides Blume). The holotypes of Don’s species have not been located; they are seem- ingly not in the British Museum where they might be expected. Although Don’s new species were based on his own collections, Wallich ignored Don’s names and provided new names in publishing his own “A Numerical List.” It will be necessary to examine closely the specimens of Wallich’s collection in the East India Company Herbarium to decide which of Wallich’s names apply to A. maximum D. Don; several are to be considered, especially A. diversifolium Wallich, A. frondosum Wallich, and A. polymorphum Wallich, and perhaps others. It seems clear, how- ever, that Don’s name applies to the plant of Nepal and northern India generally called D. polypodioides Blume (syn. D. asperum Blume). The true D. polypodioides of Java, Malaya, and southern India and Ceylon has a spiny stipe and rachis. The Himalayan plant is closely allied but, as pointed out by Sledge (Bull. Brit. Mus. Nat. Hist. Bot. 2:308. 1962), differs apparently constantly in having smooth rather than prickly stipes and rhachises. This Himalayan plant I take to be the true D. maximum. If it should prove to be the same as D. polypodioides, the name D. maximum has priority. Roxburgh’s description of his A. woodwardioides is “Stipes smooth; fronds (2 feet high), subovate, smooth, alternately- bipinnate; leaflets broad-ensiform; those of the inferior pinnae pinnatifid; of the upper more or less serrate. Fructifications in oblong spots along the nerve, but forming a sharp angle with it. Involucre separating towards the nerve. Found at Chittagong by Dr. Buchanan.” As may be seen, Roxburgh stressed that both the stipe and the fronds were “smooth,” not prickly or scaly, which agrees with D. maximum, and the identification with D. maximum, at least sens. lat., is confirmed by the de- scription of the pinnules of the lower pinnae as “pinnatifid,” which would exclude D. dilatatum Blume. Roxburgh’s name is a later homonym, and therefore it can never be adopted to re- place any other name that might be found to be identical with it. A plant that appears to be a variety of D. maximum, rather than D. polypodioides, because of the nonspiny stipe and rhachis follows: 27. DIPLAZIUM MAXIMUM (D. Don) C. Chr. var. vestitum (C. B. Clarke) Morton, comb. nov. . Asplenium polypodioides var. vestitum C. B. Clarke, Trans. Linn. Soc. II, Bot. 1:501. 1880. Type: “Central Himalaya,” with no specimens ROXBURGH'S FERN TYPES—MORTON 8038 cited. Lectotype: Darjeeling, 6500 ft alt, June 19, 1884, Clarke 35382D (K, Morton photograph 20640). This variety, which may prove to be a distinct species, has the rhachis and rhachillas rather densely scaly and pubescent. I have seen two recent collections from Lebong Forest, Dar- jeeling: Mehra 7 (US) and Bir 121 (US). The other syntypes of var. vestitum are: Darjeeling, 6500 ft alt, Sept. 2, 1875, Clarke 27329 (K) and Rungbee, Darjeeling, 5500 ft alt, Aug. 17, 1869, Clarke 8646 (2 sheets, K, Morton photograph 20641). 28. BLECHNUM ANGUSTIFOLIUM Roxburgh, Calcutta Journ. Nat. Hist. 4:501. 1844, non Willd., 1810. =Taenitis blechnoides (Willd.) Swartz f. angustifolia (Roxburgh) Morton, comb. nov. Type: A specimen of the Roxburgh Herbarium from the Martius Hebarium in Brussels annotated by me as B. angustifolium Roxb. (Morton photograph 19830). According to the unpublished list of Roxburgh species represented in the Martius Herbarium, Blechnum angustifolium Roxburgh is present, and this collection is the only possible one, even though the name B. angustifolium is not written on the sheet; the specimen agrees with the original description, and since the plant is distinctive there cannot be any doubt. No other speci- mens of B. angustifolium have been found in Kew, the British Museum, or Geneva, and so this specimen may be presumed to be unique and therefore the holotype. The locality was stated by Roxburgh to be the Molucca Islands, where it was probably collected by Christopher Smith. Taenitis blechnoides normally has pinnate blades. The juvenile blades are said to be simple, as is expected, but they are seldom collected. Roxburgh’s specimen of B. angustifolium is, however, not juvenile, because it has mature sori; it represents a form with simple mature blades, evidently rare, because, although I have seen many typical specimens of 7. blechnoides, this Rox- burgh collection is the only one with simple mature blades. It is to be doubted, however, that this condition would be constant. 99. BLECHNUM DECURRENS Roxburgh, Calcutta Journ. Nat. Hist. 4:502. 1844. —Blechnum finlaysonianum Wallich ex Hook. & Grev. Icon. Fil. 2: ¢. 225, 1831. Type: Penang?, “Herb. Finlayson.” Apparently this is not in the general herbarium at Kew, although there is an old collection without collector or locality that might be an isotype. There is an isotype—a fragment only—in the Greville Herbarium now in Edinburgh (Morton photograph 11560), and this can serve as lectotype until a better specimen is discovered. Wallich gave no locality, but Hooker and Greville guessed Penang. Blechnum finlaysonianum Wallich, Num. List 65, no. 2172, 1830, nom. nud. Asplenium penangianum Wallich, Num. List 8, no. 196. 1830, nom. nud. Based on Penang, Wallich in 1822 (K, Morton photograph 11558). Type: A specimen in the herbarium in Brussels with the name in the 304 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM hand of Roxburgh (Morton photograph 20002). This is the only collection found at Kew, Brussels, or Geneva, and so may be considered unique and the holotype. It came from Prince of Wales Island, i.e, Penang Island, Malaya, and was collected by Dr. Hunter, according to Roxburgh. Blechnum decurrens has not been properly placed. In the “Index Filicum,” it was referred doubtfully to Blechnum orien- tale L. It was properly determined as B. finlaysonianum by Bommer on an annotation slip with the type. 30. BLECHNUM GLABRUM Roxburgh, Calcutta Journ. Nat. Hist. 4:502. 1844. =Taenitis blechnoides (Willd.) Swartz, Syn. Fil. 24, 220. 1806. Based on Pteris blechnoides Willd. Phytogr. 13, t. 9, f. 3. 1794. Type: India, Klein (B, Herb. Willd.). LECTOTYPE: Two sheets in Brussels with the name in the hand of Roxburgh (Morton photographs 19828 and 19829). An isotype is in the East India Company Herbarium ex Herb. Roxburgh, no. 141-2 (Morton photograph 20658). The only locality cited by Roxburgh is Prince of Wales Island, i.e., Penang Island, Malaya, and so these specimens may be presumed to be from there. Another specimen is in the East India Company Herbarium Kew, no. 141-2 (Morton photographs 15722 and 19592b); this also has the name in the hand of Roxburgh and is doubtless a part of the same collection as the lectotype. The original de- scription cites “Taenitis blechnoides. Sw. Wall. Cat. 62, No. 141,” which refers to this specimen sent to Wallich by Roxburgh. The intent is Taenitis blechnoides sensu Wallich. In point of fact, Wallich did identify his no. 141 correctly as T. blechnoides. Dr. Holttum has recently published “A Re-definition of the Fern-genus Taenitis Willd.” (Blumea 16:87-95. 1968) in which Taenitis is enlarged to contain 15 species, including the types of the genera Holttumiella and Platytaenia and many species formerly referred to Syngramma and Schizoloma. 31. BLECHNUM MOLUCCANUM Roxburgh, Caleutta Journ. Nat. Hist. 4:502. 1844, non Desv., 1811. =Blechnum orientale L. Sp. Pl. 1077. 1753 (by typographical error as “B. occidentale,’ the epithets “occidentale” and “orientale” were accidentally reversed). LECTOTYPE: A collection in Brussels with the name Blechnum moluceanum written in the hand of Roxburgh and with the date May 28 (Morton photograph 19997). Roxburgh cited “Prince of Wales Islands, Moluccas, ete.” thus indicating that he had more than one specimen in hand. It cannot be determined if the specimen chosen as lectotype came from the Prince of Wales Island, ie., Penang Island, or the Moluccas. There is another sheet in Brussels (Morton photograph 19998) which is doubtless a part of the same collection as the lectotype. A sheet in Geneva (Morton photograph 16837) is also a part of the Roxburgh Herbarium and bears the locality ROXBURGH’S FERN TYPES—MORTON 305 “Moluccas”; however, I do not choose it as lectotype because the name is not in Roxburgh’s hand. The original description cites as a synonym “Blechnum orien- tale, Linn. Wall. Cat. 61, No. 57,” but this does not mean that Roxburgh was renaming the Linnaean species. This citation was added by Griffith and was intended to mean that Roxburgh’s species B. moluccanum was B. orientale sensu Wallich’s Cata- logue by no. 57 (actually no. 57-6, which is the Roxburgh collec- tion sent to Wallich (Morton photograph 15728 taken at Kew). 32. CYATHEA PINNATA Roxburgh, Calcutta Journ. Nat. Hist. 4:517. 1844. =Cyathea moluccana R. Brown ex Desv. Mém. Soc. Linn. Paris 6:322. 1827. Type: Molucca Islands, C. Smith (holotype BM). LECTOTYPE: A Roxburgh specimen in the Brussels Herbarium with the name in Roxburgh’s hand and the number 2429 (Morton photograph 19853). There is also a duplicate in Brussels (Morton photograph 19854). Additional isotypes are in Geneva, with a ticket reading “Prince of Wales Island, Dr. Roxburgh” (Morton photograph 16876), and in the East India Company Herbarium, no. 179-2 (K, Morton photograph 15724). The locality stated by Roxburgh is Prince of Wales Island, i.e., Penang Island, Malaya, where it was collected by W. Roxburgh, Jr. Holttum has this placed correctly as a synonym of C. moluc- cana; cf. Flora Malesiana II, 1(2):148. 1963, for a full synonymy. 33. CYATHEA TRIPINNATIFIDA Roxburgh, Calcutta Journ. Nat. Hist. 4:518. 1844, LECTOTYPE: Holttum (FI. Males. II, 1(2):156. 1963) cites the type as “Herb, Wallich n. 7076, Moluccas (CAL?; dupl. at K, BM) [Wallich 7076 was not collected by Wallich, but by Roxburgh],” but this can hardly be considered as a definitive choice of a lectotype, for Holttum seems to have thought that perhaps there was a holotype in Calcutta, but there are no Roxburgh specimens in Calcutta. Griffith, who published Roxburgh’s fern species, commented that not a scrap of Roxburgh’s material was left in Caleutta. The holotypes, where these can be determined, are in Roxburgh’s personal herbarium now in Brussels. I designate two Roxburgh sheets in the Brussels Herbarium as lectotype (Morton photographs 4881, 4882, 19856, 19857), which are evidently parts of the same frond; one of the sheets bears the name in Roxburgh’s hand. A duplicate, doubtless a part of the same frond, is in Geneva marked “Ind. Orient. Dr. Roxburgh” (Morton photographs 6536, 16893). Another duplicate is in the East India Company Herbarium, no. 7076 (Morton photograph 15724); in the general herbarium at Kew is a specimen labeled “Amboyna, Mr. Webb,” which means that this is probably collected on Amboina by C. Smith and received from Webb. It is doubtless an isotype. Another isotype is in the British Museum, as cited by Holttum. According to Holttum’s treatement in the “Flora Malesiana,” this species is endemic in Amboina. In the original description 306 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM Griffith added a reference to “C. excelsa, Sw. Wall. Cat. 63, No. 181 (sub nomine C. bipinnatifidae),”’ but this was only with a query. It appears that a specimen in Geneva (Morton photograph 6536) is part of a specimen collected in Amboina by Christopher Smith (no. 3279). 34. DAVALLIA ANGUSTIFOLIA Roxburgh, Caleutta Journ. Nat. Hist. 4:513. 1844, =Humata angustata (Wall. ex Hook. & Grev.) J. Smith, Journ. Bot. Hook. 3:416. 1841. Type: Singapore, Wallich in 1822 (isotype in E. Ind. Co. Herb. no. 242, Morton photograph 15731, lower left-hand specimen). Davallia angustata Wallich ex Hook. & Grev. Icon. Fil. 2: t. 231. 1831. LECTOTYPE: A Roxburgh specimen in the Brussels Herbarium with the name in the hand of Roxburgh, from Prince of Wales Island, i.e., Penang Island; it bears a field label in the hand of W. Roxburgh, Jr., reading: “Grows on the trunks of trees which are well shaded. It runs along this trunk for 6-8-12 feet and perhaps more. It is sent in the list you will receive.” An isotype is in the East India Company Herbarium, no. 242-2 (Morton photograph 15731), two top plants; the lower right-hand plant is from Penang, collected by Wallich in 1822, which was not mentioned by Wallich or by Hooker and Greville, but was probably included in Wallich’s original concept of his Davallia angustata. The citation added to Roxburgh’s original description by the editor, Griffith, “D. angustata Wall. Cat. 63, No. 242” does not mean that Roxburgh’s species was intended as a renaming of D. angustata Wallich, but merely that Roxburgh’s species was included by Wallich under his no. 242, actually no. 242-2, as shown by the reference to page 63. Roxburgh’s Davallia angustifolia is quite the same as D. angustata Wallich, which is a good species and not a synonym of Humata heterophylla as the “Index Filicum” has it. It differs in the leaves not being dimorphic and in the fertile ones not being lobed. 35. DAVALLIA CORDIFOLIA Roxburgh, Calcutta Journ. Nat. Hist. 4:514. 1844. =Humata trifoliata Cav. Deser. 273. 1802. Type: Marianna Islands, Née, according to Cavanilles; Christensen (Dansk Bot. Ark. 9(3) :26. 1937) suggested that Nee’s plant may have come from the Philippine Islands, since the species had not been collected again in the Mariannas, and since it could be matched by Philippine specimens; however, un- known to Christensen, Hosokawa had reported it again the previous year (Trans. Nat. Hist. Soc. Formosa 26:121. 1936) from Alamagan, one of the smaller Mariannas Islands, under the synonymous name Humata lepida (K. B. Presl) Moore, as pointed out by Wagner (Oce. Pap. Bern. P. Bishop Mus. 19:85. 1948). Thus the original locality was correct, but the species must be very rare in the Mariannas. ROXBURGH’S FERN TYPES—MORTON 307 Davallia serrata Willd. in L. Sp. Pl. ed. 4, 5:467. 1810. An illegitimate change of epithet in transferring Humata trifoliata to Davallia. Humata serrata (Willd.) Desv. Mém. Soc. Linn. Paris 6:323. 1827, An illegitimate name, since the earliest available epithet trifoliata was not adopted. Type: No herbarium specimen of D. cordifolia has been located. Roxburgh’s type came from “mountains north of Rohileund,” i.e., the present Rohilkhand, in the Northern United Provinces, Division of Agra. This species has had to be placed from description only. In the “Index Filicum,” it is referred to Humata repens (L. f.) Diels, which was an aggregate as treated. Humata repens was based on Adiantum repens L. f. Suppl. 446. 1781. The type came from the Ile de France, i.e., Mauritius, and was collected by Sonnerat and transmitted by Thouin. The holotype is in the herbarium of Linnaeus fil. in the J. E. Smith Herbarium (no. 1635.21) in the Linnean Society, London (Morton photograph 20298). It is a good specimen, typical of the species as it grows in Mauritius, where the plants are small, long-stalked, not very divided, and not dimorphic. The plants from the Philippine Islands, Malaya, and India that have been called H. repens agree with H. trifoliata Cav., according to the study by Christensen (Dansk Bot. Ark. 8(3):26. 1937). Additional synonymy for H. trifoliata is given by Copeland in his Fern Flora of the Philippines (1:177. 1958). Copeland also maintains H. repens as a native of the Philippines, although with doubt. Apparently true H. trifoliata occurs in Sumatra and Borneo. The larger and more divided plants of Java, New Caledonia, the Admiralties, and perhaps Fiji and Samoa (Humata serrata Brack., non Desv.) seem to be best called Humata alpina (Blume) Moore, Ind. Fil. XCII. 1857, as recognized in the “Index Filicum, Suppl. 3,” which is based on Davallia alpina Blume (Enum. Pl. Jav. 2:231. 1828). As lectotype of D. alpina, I designate a specimen in Leiden from Mount Gedé, Java, collected by Blume and with the name “alpina” in Blume’s hand (Morton photo- graph 1523). This probably is the actual holotype. A second sheet lacking the locality is in Leiden also, which probably rep- resents an isotype (Morton photograph 1514). 36, DAVALLIA LONGIFOLIA Roxburgh, Calcutta Journ. Nat. Hist. 4:514. 1844, —Grammitis alata (Blume) Morton, comb. nov. Davallia alata Blume, Enum. Pl. Jay. 2:230. 1828. “In fissuris rupium Javae,” fide Blume, Lectotype: Raab, Java, Zippel (1. Morton photo- graph 845). Several other syntypes are at Leiden, collected by Blume. Prosaptia alata (Blume) Christ, Ann. Jard, Bot. Buitenzorg 20:127. 1905. Ctenopteris alata (Blume) Holtt. Fl. Malaya, Ferns 2:232, 1954. 308 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM TYPE: No specimen from Roxburgh has been located. The type was collected on Prince of Wales Island, i.e., Penang Island, by W. Roxburgh, Jr. In the “Index Filicum,” D. longifolia Roxb. was referred with- out doubt to Davallia alata Blume, and from the description it appears that this is surely right. Prosaptia does not appear to differ from the species of Gram- mitis sect. Cryptosorus (cf. Morton, Contr. U.S. Nat. Herb. 38:90. 1967) that have sunken sori, except in having the sori submarginal. It could perhaps rank ag a distinct section. 37, DAVALLIA MOLUCCANA Roxburgh, Calcutta Journ. Nat. Hist. 4:516. 1844, non Blume. =Tapeinidium moluccanum (Blume) C. Chr. Gard. Bull. Str. Settl. 4:399. 1929. Based on Davallia moluccana Blume, Enum. Pl. Jav. 237. 1828. Davallia amboynensis Hook. Sp. Fil. 1:178, t. 56C. 1846. A renaming of Davallia moluccana Roxburgh, non Blume. Tapeinidium amboynensis (Hook.) C. Chr. Ind. Fil. 631. 1906. LECTOTYPE: A specimen in the Brussels Herbarium with the name in the hand of Roxburgh (Morton photograph 19825). There is an isotype in Brussels, with the name not in Roxburgh’s hand (Morton photograph 19824) , one in the British Museum marked “Amboyna, Chr. Smith” (Morton photo- graph 15589), and one in Geneva marked “Moluccas, Dr. Roxburgh” (Mor- ton photographs 6533, 16747). Roxburgh indicated his species as from the Molucea Islands; these collections show that the species was collected in Amboina by Christopher Smith. Roxburgh assigned the name Davallia moluccana to this spe- cies long before Blume chose the same name for one of his species. Hooker assumed (without having seen Blume’s type) that the Roxburgh species was different and renamed it Davallia amboynensis, since Roxburgh’s name although proposed in manu- script earlier than Blume’s was not published until later, in 1844; from Hooker’s extensive footnote, it appears that he had just received the Roxburgh publication while he was writing the account of Davallia; he cites it from a reprint rather than from the original Calcutta Journal. In addition to Roxburgh, Hooker cited collections from Amboina in “Herb. Banks” (i.e., British Museum) and in his own herbarium collected by “A. Smith” received from Webb. The specimen in the British Museum re- ferred to is clearly C. Smith, ie, Christopher Smith; Hooker’s “A. Smith” is either a typographical error or an error on the part of Webb; I believe that Christopher Smith was the only Smith who collected in Amboina in the early years. In his recent revision of Tapeinidium (Blumea 15:545-556. 1967), Kramer indicated as lectotype of D. amboynensis this Kew specimen collected by “A. Smith.” But as indicated above, D. ROXBURGH'S FERN TYPES—MORTON 309 amboynensis, was a renaming of D. moluccana Roxburgh, non Blume, and so it must have the same type as Roxburgh’s species. The Kew collection mentioned, although surely as isotype of Roxburgh’s species, was doubtless not seen by Roxburgh and so it is not a suitable lectotype. Kramer states that the type of Davallia moluccana Blume is a Saccoloma rather than a Tapeinidium, following Mettenius. But as indicated by Christensen (Gard. Bull. Str. Settl. 4:399. 1929), Mettenius obtained his concept from a specimen in Leiden that had been identified as Davallia moluecana Blume by Blume only with a query. The specimen I take to be the holotype, or at least the most suitable lectotype, is one in Leiden with the name Davallia “moluccensis nobis” in Blume’s own hand; it is from Amboina, collected by Reinwardt (Morton photograph 2281); Blume changed “moluccensis” to moluccana in his publica- tion. This type shows that Blume’s species is indeed the same as Roxburgh’s Davallia moluccana, both authors having hit upon the same specific epithet independently. 38. DAVALLIA MULTIFLORA Roxburgh, Calcutta Journ. Nat. Hist. 4:515, t. raat, left hand. 1844. —Nephrolepis multiflora (Roxburgh) Jarrett ex Morton, comb. nov. LECTOTYPE: A specimen in the Brussels Herbarium originally named Davallia with a species name now mostly cut off, this subsequently changed to Polypodium davallioides, and identified later as Nephrolepis exaltata by Baker and N. hirsutula by Bommer (Morton photograph 19638). According to Roxburgh the species is native to “the interior parts of Bengal, Nepaul, ete.” Judging from my photograph, the same species is represented in the herbarium of the East India Company at Kew, where it is filed as no. 1031, although it should be 1031-6, i.e., Polypodium davallioides Roxburgh (Morton photograph 19582). Griffith indicated that Davallia multiflora Roxburgh occurred under the name Polypodium davallioides also. In the “Index Filicum,” Davaillia multiflora Roxburgh is said to “=Humata gaimardiana vel Nephrolepis sp. (hirsutula?).” This confusion was occasioned by the editor Griffith, who added to Roxburgh’s manuscript a reference to “D. parallela, Wall. Cat. 63, No. 251,” which is plainly an error. Griffith should have added this Wallich reference under Davallia pectinata Smith of Roxburgh’s manuscript, since D. pectinata sensu Roxburgh is clearly the same as D. parallela Wallich, i.e., a species of Humata. Roxburgh’s description and published illustration, as well as the unpublished painting at Kew, show that Roxburgh’s plant was a Nephrolepis and not a Humata. Roxburgh’s description is fairly good, as is the published drawing (t. XXXI, left hand). The specimen selected as lecto- 310 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM type is believed authentic because it was originally determined by Roxburgh as a Davallia, and D. multiflora is the only species of Nephrolepis described as a Davallia by Roxburgh; the orig- inal species name has been mostly cut off leaving only the top of some of the letters visible, but they are consistent with the name multiflora. More importantly, the specimen agrees well with the original description, as for instance the crowded pin- nae, which are indeed almost imbricate in this specimen, in the shape of the blades and pinnae, including the auricles, and in the indusia being submarginal, reniform and opening out- wardly. Especially the apex of the blade and the basal, somewhat reduced pinnae are exactly like the published illustration. It is therefore likely that this lectotype is truly authentic and _ per- haps an actual holotype. Indian specimens of this species have been generally identi- fied as either Nephrolepis eraltata (L.) Schott or N. hirsutula (Forst.) K. B. Presl, but Dr. Jarrett regards them as distinct (both lack short hairs on the upper side of the pinna midribs). In the shape of the indusia and their submarginal position, the species appears to be closer to N. evaltata than to N. hirsutula, with which, however, it can be easily confused. Some of the material in cultivation as N. hirsutula is probably N. multiflora. New World specimens identified as N. multiflora are actually N. exaltata. 39. DAVALLIA PECTINATA sensu Roxburgh, Calcutta Journ. Nat. Hist. 4:514. 1844. —Humata pectinata (J. E. Smith) Desv. Mém. Soc. Linn. Paris 6:328. 1827. AUTHENTIC MATERIAL: East Ind. Co. Herb., no. 251-2 (K, Morton photo- graph 15732, upper plant). This collection is named Davallia multiflora Roxburgh and is so cited in Wallich’s “Numerical List,” p. 63 under no. 251-2, but this was an error, the name having been transposed somehow to the wrong plant. This plant, no. 251-2, is Roxburgh’s D. pectinata, which was not considered a new species but was credited to Smith, and not his D. multiflora. According to Roxburgh it was “found on the northern boundary of Oude by A. Gott,” which refers to the area later known as the United Provinces of Agra and Oudh, India. The true Davallia multiflora Roxburgh is a Nephrolepis, as shown by the description and drawing. Roxburgh identified his plant correctly, it appears, from his description and from this specimen. IT am not sure that Humata pectinata has ever again been found in Oudh, but there is no reason that it could not occur there. It grows in Burma, as indicated by Beddome (under the synonymous name Humata parallela). The type of Davallia pectinata J. E. Smith is in the J. E. ROXBURGH'S FERN TYPES—MORTON 311 Smith Herbarium, no. 1635-19, in the Linnean Society, London (Morton photograph 20297, excluding the small plant at upper left, which is Malacca, June, 1796, Christopher Smith). The specimen is labeled “Ind. or. Soc. Unit. Frat. 1786” and according to Smith was collected by D. Hurloch, presumably a missionary. Smith also cited “eandem forte in Otheite legit Nelson. H. Banks.” The word “forte” (=perhaps) indicates that this second specimen is not a type but was referred to the species with some doubt. This specimen is also in the J. E. Smith Herbarium (Morton photograph 20299, left-hand plant) marked as ‘“Ota- heite, Nelson ex hb. Banks.” It is marked as “D. pectinata var.” Smith described the lowest segments as “auriculatis semipin- natisve”; the ‘“‘auriculatis” is true of the type, and the “‘semi- pinnatisve” came from this Nelson specimen from Tahiti, which may or may not be separable taxonomically. It is not definite where the type of D. nectinata may have come from, since in the eighteenth century the term “India orientalis” did not mean eastern India but what we might call the ‘East Indies,” a sort of general term including Malay, Singapore, and Malaysian Is- lands, as well as India; since it agrees with plants from Singa- pore, it may well have come from there. Hooker (Sp. Fil. 1:153. 1845) misunderstood Smith’s D. pecti- nata, which he misapplied to the plant from Tahiti represented by the Nelson collection, and redescribed the true D. pectinata as D. parallela Wallich, the latter based in part on Wallich List no. 251 from Singapore, collected by Wallich in 1822 (Kast Ind. Co. Herb. no. 251, Morton photograph 15732, lower plant). However, since Hooker cited Nephrodium gaimardianum Gaud. as a straight synonym of D. parallela, the latter becomes a superfluous name, since the epithet “gaimardianum” was availa- ble in Davallia and should have been used. In fact, Pres] had tentatively proposed D. gaimardiana in his “Tentamen” (1886), only to delete it in the errata at the end (p. 290) in favor of his Nephrolepis gaimardiana. 40. DAVALLIA PILOSA Roxburgh, Calcutta Journ. Nat. Hist. 4:515, t. 2 (right hand). 1844, —Microlepia speluncae (L.) Moore, Ind. Fil. XCIII. 1857, var. speluncae. LECOTYPES: A specimen in the Brussels Herbarium named by Roxburgh Polypodium ciliatum? and reidentified by Baker as Davallia speluncae (Mor- ton photograph 19818). No specimen named Davallia pilosa Roxburgh has been found. The reason for connecting the specimen chosen as lectotype is that Roxburgh indicated that his illustration was under the 312 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM name of Polypodium ciliatum, and this sheet does bear the name Polypodium ciliatum. The query present doubtless indicated Roxburgh’s doubt about the species being truly a Polypodium. When he decided to describe it as a Davallia, he changed the specific epithet from ciliatwm to pilosa. Roxburgh’s published drawing and the copy at Kew are both poor, and one could not possibly identify the species from them. Roxburgh’s description, however, is rather detailed; the lecto- type chosen agrees with the description, and it may be con- sidered truly authentic and very likely the actual holotype. According to Roxburgh, it came from the Delta of the Ganges, from where it was introduced into the Botanic Garden in Cal- cutta by Buchanan-Hamilton. It is most likely that no herbarium specimen was made of the plant in the wild and that the speci- men at hand came from the plant cultivated in the Botanic Garden. In Sledge’s treatment D. pilosa is placed as a doubtful synonym of M. speluncae var. pubescens (Hook.) Sledge. The lectotype, however, shows that D. pilosa is typical M. speluncae as it grows in the type locality, Ceylon. 41. DAVALLIA SERRATA Roxburgh, Calcutta Journ. Nat. Hist. 4:514. 1844, non Willd., 1810. =Tapeinidium pinnatum (Cav.) C. Chr. Ind. Fil. 213. 1905. Type: Philippine Islands, Née. TYPE: A specimen in the Brussels Herbarium with the name in the hand of Roxburgh (Morton photograph 19916). Since this is the only specimen seen in any herbarium, it is likely unique and a holotype. It came from the Prince of Wales Island (i.e., Penang Island) according to Roxburgh, where it was collected by W. Roxburgh, Jr. In the “Index Filicum,” the citation “Hk. sp. 1:174. 1846 = biserrata Bl.),” following the entry for Davallia serrata Rox- burgh is an error. The “Davallia serrata” of Hooker at the place cited is not the same species as D. serrata Roxburgh, but an in- advertent error by Hooker for Davallia biserrata Blume, also a species of Tapeinidium, a somewhat dubious one but different from T. pinnatum (Cav.) C. Chr. A true synonym is Davallia flagellifera Wallich ex Hook. & Grev. Icon. Fil. 2: t. 183. 1830 [or 18312], the type of which also came from Penang Island (Wallich in 1822, Num. List no. 248, isotypes BR, Morton photo- graphs 19826, 19827); this species has been omitted from the “Index Filicum,” except for the citation of the invalid nomen nudum “D. flagellifera Wall. List n. 243. 1828.” Hooker and Greville gave a complete description and illustration. ROXBURGH’'S FERN TYPES—-MORTON 313 42. DAVALLIA TRAPEZIFORMIS Roxburgh, Calcutta Journ. Nat. Hist. 4:516. 1844. =Microlepia pilosiuscula (J. E. Smith) Morton, comb. nov. Davallia pilosiuscula J. E. Smith in Rees, Cycloped. 11: Davallia no. 10. 1808. Syntypes: Honimoa, July, 1797, Christopher Smith, and Am- boina, 1796, Christopher Smith. These two specimens are in the J. E. Smith Herbarium in the Linnean Society, London, nos. 1635-17 and 1635-18 (Morton photographs 20295 and 20296 respectively). I desig- nate the specimen from Amboina as lectotype. Microlepia trapeziformis Kuhn, Chaeopt. 27. 1882. Microlepia speluncae var. pubescens (Hook.) Sledge, Kew Bull. 1956: 525, at least in part. LECTOTYPE: A specimen in the Brussels Herbarium with the name Poly- podium saccatum and the number 2404 in the hand of Roxburgh (Morton photograph 19819). The species came from the Moluccas according to Roxburgh. The name “Polypodium saccatum” that Roxburgh originally applied to this species was based on the sorus, which Roxburgh described as “‘involucre . . . forming a pouch.” When he reconsid- ered and decided to place the plant in the genus Davallia rather than Polypodium, he changed the specific epithet to trapeziformis (referring to his description of the leaflets as ‘“‘subtrapeziform’’), doubtless for the reason that all Davallias have the “involucre”’ somewhat pouchlike or saccate. The lectotype is matched by a second specimen in Brussels also identified as Polypodium saccatum, although not in Roxburgh’s hand (Morton photograph 19820). These specimens were un- doubtedly collected for Roxburgh by Christopher Smith—like Roxburgh’s other Molucca collections—and they match the lecto- type above for Davallia pilosiuscula. Thus it appears that Rox- burgh’s species came from Amboina. J. E. Smith had a second collection of his D. pilosiuscula, that from Honimoa, which is slightly larger and more divided. Roxburgh had this too, and he annotated it as “not well ascertained, probably a luxuriant speci- men of saccatum as it has the same involucre.” This specimen was doubtless included in Roxburgh’s description, for he wrote that the fronds were “alternately bipinnate and tripinnatifid,” the bipinnate applying to the plant from Amboina and the tri- pinnatifid to the more divided plant from Honimoa. A duplicate of this larger plant also is in Brussels, which by comparison with the plants in the J. E. Smith Herbarium can be presumed to be from Honimoa collected by Christopher Smith. Other specimens of this species were evidently collected by Smith in some quantity. One of these is in the East India Com- pany Herbarium, no. 262-4 at Kew (Morton photographs 15734, 19583, at left), which agrees with the lectotype chosen; 314 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM undoubtedly it came from Amboina, collected by C. Smith. A similar specimen in Geneva, determined as Polypodium saccatum Roxb. (Morton photograph 16641), has only the locality ‘Ind. or.” Another sheet of the same thing in Geneva, marked “Am- boina, Christ. Smith,” lacks the name Polypodium saccatum but is determined as Davallia polypodioides (Morton photograph 6531). Another specimen in Geneva from ‘Moluccas, Dr. Rox- burgh” has been named Polypodium dubium wrongly by some- one other than Roxburgh (Morton photograph 6543). Another specimen collected by C. Smith in the Moluccas is from Gilolo Island (i.e., the present Halmahera Island) and was probably not seen by either J. E. Smith or Roxburgh, and so is not authentic. The group of species centering around Microlepia speluncae is exceedingly difficult to study from herbarium specimens. Alston in his paper on the ferns of the J. E. Smith Herbarium was the first to identify Davallia pilosiuscula J. E. Smith; he stated that the species was “probably conspecific with the type of M. speluncae from Ceylon in Hermann’s herbarium” (Phil. Journ. Sci. 50:177. 1933). Material from Ceylon identified by Sledge as agreeing with the Hermann type is, however, rather different. The hairs on the rhachillas of the pinnae are elongate, and many- septate, with conspicuous cross walls; the hairs on the segments are few and coarse. The plants from Amboina and Honimoa, and probably also at least most of those from Java and Sumatra, are finely and densely pilosulous, the hairs of both the rhachillas and segments being short, horizontally spreading, and only one or two cells long. I feel that these plants can hardly be conspecific. I venture to propose a new combination because the name of Smith is long prior to any other name that might apply to these Malay- sian plants, and so it will likely stand when these plants are monographically studied and better understood. Sledge based his M. speluncae var. pubescens on Davallia polypodioides var. pub- escens Hook. Sp. Fil. 1:182. 1846, but did not typify this name. Hooker cited three different numbers of Wallich, and the locali- ties of Singapore, Penang, Martaban, Assam, Mergui, Java, and Macalisberg in South Africa. Considering this geographically wide range, Hooker’s var. pubescens is probably a mixture. The plant from Java likely included plants that I refer to M. pilo- siuscula; however, I have not seen the collection cited—Zollinger 513. It is clear from the above discussion that the name Microlepia trapeziformis (Roxburgh) Kuhn has been misused for an entirely different species, an identification of D. trapeziformis Roxburgh ROXBURGH’S FERN TYPES—MORTON 315 as being synonymous with D. rhomboidea Wallich going back to Moore’s “Index Filicum.” Kuhn picked up the name, without discussing it or its type, and he was followed by Christensen in the ‘Index Filicum.” The name has been generally accepted since, as by Holttum in his “Ferns of Malaya” and by Sledge (Kew Bull. 1956:526), without verification. The species, misidentified as M. trapeziformis, is, as shown by Sledge, quite different from M. speluncae in its pubescence; the hairs on the rhachillas of the pinnae are coarse, stiff, and antrorsely appressed; the segments are broader and more rounded, and the whole plant coarser in appearance. This species is widespread in Asia (in Ceylon, India, Thailand, Pahang, Tonkin, and Yunnan) and occurs also in Java and Sumatra. I doubt that it occurs in the Molucca Islands, how- ever, where the type of D. trapeziformis came from. The proper name of this species appears to be Microlepia rhomboidea (Wal- lich ex Kunze) Prantl, Arb. Bot. Gart. Breslau 1:31. 1892, based on Davallia rhomboidea Wallich ex Kunze, Bot. Zeit. 8:158. 1850. The original Davallia rhomboidea Wallich, Num. List no. 257. 1829, was a nomen nudum. These collections should be restudied, both at Kew, the Linn- ean Society, and in Geneva, to make sure that I have them right. It may be that the two plants in the Smith Herbarium do repre- sent different species, the one from Amboina being M. trapezi- formis and the one from Honimoa being M. dubia. 43. DICKSONIA MOLUCCANA Roxburgh, Calcutta Journ. Nat. Hist. 4:517, 1844, non Blume, 1828. =Cystodium sorbifolium (J. E. Smith) J. Smith in Hook & Bauer, Gen. Fil. t. 96. 1841. Dicksonia sorbifolia J. E. Smith in Rees, Cycl. 11: unpaged. 1808. TYPE: Honimoa, Ceram, Indonesia, Christopher Smith, July, 1797 (Herb. Smith no. 1636.4, LINN, Morton photograph 20301). LECTOTYPE: Honimoa, Christopher Smith 326 in 1797, BR (Morton photo- graph 19642). Roxburgh’s Dicksonia moluccana was founded on a part of the same collection on which D. sorbifolia was based. Two collections are in the British Museum. One is a single pinna from “Ins. Molucc.” and one has parts of five pinnae match- ing the preceding marked as “Amboyna,” both collected by C. Smith; these may be and probably are part of the same collection and thus isotypes, the “Amboyna” being an error for Honimoa (Morton photographs 6875 and 6633). A good illustration of this interesting monotypic genus is given by Holttum (FI. Males. II, 1(2):163. 1963). An account of the gametophye was given by Lenette R. Atkinson (Amer. Fern Journ. 55:32-85. 1965) and of the anatomy by Sen and Mittra (Amer. Fern Journ. 56: 97-101. 1966). 316 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM Another specimen in Kew (Morton photograph 15652) is noted as “Moluccas, Wallich”; this agrees wholly with the other speci- mens seen and is surely a part of the same collection of Christo- pher Smith given to Wallich by either Roxburgh or J. E. Smith, for Wallich did not collect in the Moluccas. The reference added by Griffith was to Wallich’s List no. 2178, which was a typographical error for no. 2174, a specimen of D. moluccana Roxburgh doubtless entered from the lectotype in- dicated above. There is no specimen now under no. 2174 in the East India Company Herbarium. 44, EQUISETUM DEBILE Roxburgh ex DC. in Vaucher, Mem. Soc. Phys. Hist. Nat. Genéve 1:387. 1822 or 1821); Calcutta Journ. Nat. Hist. 4:468, t. 26 (middle plant). 1844, =Equisetum ramosissimum Desf. subsp. debile (Roxburgh) Hauke, Amer. Fern Journ. 52:38. 1962. LEcTOTYPE: A plant in the J. E. Smith Herbarium, Linnean Society, London, no. 1648-1, from Calcutta, India, collected by Roxburgh (Morton photograph 20335). A similar plant, but lacking strobiles, is in the British Museum (Natural History) (Morton photograph 7775). Because of the close similarity, a drawing by Roxburgh at Kew (no. 1921, Morton photo- graph 15887) might have been made from the plant in the Smith Herbarium, with some artistic rearrangement. In his treatment of FE. debile in his “A Taxonomic Monograph of the Genus Equisetum Subgenus Hippochaete” (Nova Hedwigia Beih. 8:1-123. 1963), Hauke cites the type as being in the De- Candolle Herbarium in Geneva, collected at Serampore, near Calcutta, by Griffith (no. 919), but this is obviously impossible, since Griffith was only about 11 years old at the time FE. debile was described in 1821 or 1822; he did not begin collecting in India until 1835. Alston searched in Geneva for a type unsuc- cessfully, but probably one is there, perhaps a specimen that is unlabeled or without the indication of Roxburgh as the col- lector. If a specimen is located eventually it will become the holo- type, and the lectotype will be abandoned. According to the orig- inal description, Vaucher did see a specimen, but only part of a branch; the description was taken from a manuscript by De- Candolle. 45, HEMIONITIS CORDATA Roxburgh ex Hook. & Grev. Icon. Fil. 1: ft. 64. 1828. —Hemionitis arifolia (Burm. f.) Moore, Ind. Fil. 114. 1859. Asplenium arifolium Burm. f. Fl. Ind. 231. 1768. Type: India, Burmann. Morton photo 3863, which is labeled TYPUS, is not Hemionitis but has been determined by Alston as Acrostichum aureum L., of which it is a juvenile specimen with a simple blade. If this really is the type, then the correct name for H. arifolia will be Hemionitis cordata Roxburgh ex Hook. & Grev. ROXBURGH'’S FERN TYPES—-MORTON 317 Hemionitis cordifolia Roxburgh. See entry 46. Hemionitis hastata R. Brown ex Wallich, Num. List. 65, no. 2170. 1830, nom. nud. Based on Mysore, India, Buchanan-Hamilton (BM, Morton photograph 7501, upper right-hand plants). Acrostichum trinerve Buchanan-Hamilton in sched. (BM, Mysore, India, Buchanan-Hamilton, Morton photograph 7501). TYPE: Hooker and Greville cited three collections: one from “Ind. Orient,” collected by Roxburgh; one from Madras by Shuter (K, probably the plant on the right mounted on the sheet with Wight 51, Morton photograph 20642) ; and one from low places near Calcutta “ad locis depressis veget. sub pluviis,” collected by Wallich in 1820 (K, Morton photograph 20643). The Roxburgh specimen should be the lectotype, if one can be located at Kew or Edinburgh, considering that the species was attributed to Roxburgh. 46. HEMIONITIS CORDIFOLIA Roxburgh, Calcutta Journ. Nat. Hist. 4:500. 1844. =Hemionitis arifolia (Burm. f.) Moore, Ind. Fil. 114. 1859. Hemionitis cordata Roxburgh ex Hook. & Grev. Icon. Fil. 1: t. 64. 1828. LECTOTYPE: Herb. East India Company, no. 44 (K, Morton photograph 14670). This sheet contains three collections according to the label: (1) Bengal, 1807 [Roxburgh]; (2) Rangoon, 1827 [Wallich]; (3) “var. frond. fert. lobatis.” The latter, which is not localized, is represented by a blade with the stipe cut off; it is not exactly “lobate” but rather irregularly sinuate. The middle plant on this sheet is the Roxburgh collection, since it matches a Roxburgh collection in the British Museum. The two plants at the sides are the Wallich collection. The isolectotype in the British Museum is marked “prope H. B. Calcuttae,” i.e., near the Calcutta Botanical Garden (Morton photograph 15781). Copies of authentic Roxburgh drawings are at Kew (no, 1750) and the British Museum (Morton photograph 15770). According to Voigt (Hort. Suburb. Calcutt. 734. 1845), the Roxburgh specimen came from Serampore, Bengal. In addition to the specimens cited above I have seen a Rox- burgh specimen at Oxford determined as “Acrostichum hasta- tum,’ which has also been identified as ‘“Cyclophorus” in some later hand (Morton photograph 20203). There is also a specimen in Brussels with the name Hemionitis cordata Roxburgh, from Madras, India, received from Hooker; this specimen cannot be a type, since it is from the wrong locality, but it may have been collected by Roxburgh while he was stationed in Madras prior to going to Calcutta. 47. HEMIONITIS RETICULATA Roxburgh, Calcutta Journ. Nat. Hist. 4:501. 1844, non Forst., 1768. =Antrophyum callifolium Blume, Enum. PI. Jav. 111. 1828. TYPE: Java, Blume 95 (holotype L, Morton photograph 349; this sheet is the holotype because it is the only specimen labeled in Blume’s hand and the only one collected by Blume). Lectotype: A sheet in the Brussels Herbarium collected in Honimoa, Ceram, Indonesia, Roxburgh Herb. 1174 (Morton photograph 19908). This sheet could be considered a holotype perhaps, since it is the only one seen with locality and with the name in the hand of Roxburgh. Roxburgh gave 318 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM the locality as “Moluccas,” which includes the island of Honimoa. As other Roxburgh specimens from the Moluccas, this one was undoubtedly collected by Christopher Smith. Another sheet in Brussels, also from the Roxburgh Herbarium, has two detached fronds: the one at the left is identical with the lectotype and is doubtless part of the same collection; the smaller one at the right is different—longer stipitate blades and with a midrib that is not prominent and not black. I exclude this frond from the type collection. Very likely from my photograph the same plant is represented by the Roxburgh collection in the East India Company Herbarium no. 40-2 (Morton photograph 14668, two bottom plants), and this too is excluded as type material. Hemionitis reticulata Roxburgh is omitted from the “Index Filicum,” presumably because Christensen assumed that Rox- burgh intended H. reticulata Forst. Roxburgh, however, does not mention Forster, and he puts an “R.” after the species name, thus indicating himself as the author of the species, as he did with the others he described as new. He selected the epithet “reticulata” independently from Forster; it is a natural choice, considering the conspicuously reticulate venation of the blades. Roxburgh’s species came from the Moluccas and Forster’s was definitely indicated by Forster as from the Society Islands. The citation added to H. reticulata Roxburgh, “Antrophyum reticula- tum Kaulf. Wall. Cat. 61, No. 40,” was added by the editor Griffith, since Kaulfuss published his A. reticulatum long after Roxburgh’s death, as did Wallich his “Catalogue”; Griffith intended merely to note that a Roxburgh collection was cited in Wallich’s ‘“Cata- logue”’ under number 40; this collection, actually no. 40-2, is the one mentioned above as prebably to be excluded from the type material. The species of Antrophyum are by no means clear. Someone, possibly Bommer, has identified the Roxburgh lectotype as A. semi- costatum Blume, but this is clearly wrong, for that species be- longs to the group of species having clavate paraphyses, whereas H. reticulata Roxburgh has delicate, hairlike paraphyses. The nearest species appears to be A. callifolium Blume, although the midrib may be more prominent and darker than is usual in that species. This character may have been stressed too much in de- limiting the species. The Roxburgh species is perhaps equally close to the true A. reticulatum (Forst.) Kaulf. from Tahiti. The latter may be based on a mixture. Christensen, in his “Ferns of Samoa,” indicated that the Forster specimen he saw in the British Museum had two elements, one corresponding to A. grevillei Balfour (with elongate, narrow fronds with elongate, parallel sori) and one he took to be typical A. reticulatwm. I have seen a different specimen in the British Museum (Morton photograph ROXBURGH'S FERN TYPES—MORTON 319 7610) which also has two elements, one identified as A. lessonti Bory (with short, broad blades) and another I take to be typical A. reticulatum. 48. ISOETES CAPSULARIS Roxburgh, Calcutta Journ. Nat. Hist. 4:470. 1844. =Vallisneria spiralis L. The original description is as follows: “Capsules peduncled, 1-celled. Leaves linear flat. “This plant grows in deep standing sweet water, with Valisneria spiralis, ete. “Stoles creeping, jointed, tufts of filiform roots descend from each joint, and from 4 to 12 leaves ascend; they are like those of Valisneria spiralis, very delicate, 2-3 feet long, a quarter of an inch broad, and slightly serrated near the apex. From the alae of the leaves arise several diaphanous, cordate capsules standing on short peduncles; they consist of 1-cell, formed of two valves, opening from the apex, the seeds are numerous connected to a conical receptaculum in the centre. “T have not seen the male flowers.” At first sight this is a truly amazing description, for this is a description of Vallisneria spiralis L., as shown by the description of the stolons as creeping and jointed (Isoétes has a fleshly corm, not creeping, jointed stems), of the leaves as two to three feet long and serrate near apex (the leaves of Isoétes are, at least usually, shorter than this and not serrate), and of the capsules as being pedunculate and bivalvate (Isoétes has of course “cap- sules,” i.e., sporangia, borne sunken within the bases of the leaves, and not pedunculate or valvate). It seems likely that Griffith, the editor and publisher of Rox- burgh’s posthumous work, accidentally inserted a description of Vallisneria for the intended description of an Isoétes. This is borne out by two facts. Roxburgh’s published drawing, op. cit. t. 26, left-hand plant, is labeled Jsoétes capsularis, and it is truly and obviously an Jsoétes. Secondly, Griffith later published a description and discussion of Isoétes capsularis, which is cited as being the J. capsularis of Roxburgh (Notul. ad Pl. Asiat. 2: 572-580. 1849; Icon. Pl. Asiat: ¢. 116-118. 1849), and Griffith’s description and drawings are obviously truly Isoétes, very likely I. coromandelina L. f. Nevertheless, since the original description of J. capsularis Roxburgh applies altogether to Vallisneria spiralis, except the illustration, it seems that one is forced to consider it a synonym of V. spiralis, even though this may not have been Roxburgh’s intent. 320 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM 49, ISOKTES COROMANDELINA sensu Roxburgh, Calcutta Journ. Nat. Hist. 4:470, 1884, prob. not L. =? Isoétes indica Pant & Srivastava, Proe. Nat. Inst. Sci. India 28, B:246. 1962. Type: Ram Nai village, Rews, Madhya Pradesh, India, Pant Pt. 2A (K, not seen). AUTHENTIC MATERIAL: A specimen from “Ind. or.” collected by Roxburgh in the J. E. Smith Herbarium, no. 1650.3, Linnean Society London (Morton photograph 20341). The plant came from the Circar Mountains, according to Roxburgh, i.e., in northern Madras Province, India. The above specimen indicated as authentic does not bear the name J. coromandelina in Roxburgh’s hand, but it agrees with Roxburgh’s description. Roxburgh considered his plant the same as the Linnaean species and not as new, and he may have been right. The sporangia, however, are enormous (for [soétes) in these plants in the Smith Herbarium, about 2 cm. long, in which they seem to agree better with the recently described /. indica than with I. coromandelina L., which has the sporangia only about 12 mm. long. The identity with J. indica, however, would have to be proved by a comparison between the holotype at Kew and the specimen in the Smith Herbarium, something I have not done. Jsoétes indica has been known only from the type locality in Madhya Pradesh, in central India. According to the paper “The Genus Isoétes in India,” by Pant and Srivastava, it differs from J. coromandelina as follows: Triradiate ridges of megaspores normally simple; ends of megaspore tu- bercles generally rounded; sterile cells absent; microspores smooth or rugose to papillate 2... .. 0... ccc cee ees I. coromandelina Triradiate ridges in megaspores often branched; ends of megaspore tubercles generally tapering; sterile cells present in outer megasporangia; micro- spores tuberculate 0.2... . cc ccc eee eee tenet ene I. indica 50. LINDSAEA BIPINNATA Roxburgh, Calcutta Journ. Nat. Hist. 4:511. 1844. —Lindsaea parasitica (Roxburgh) Hieron. Hedwigia 62:14. 1920. LECTOTYPE: A Roxburgh specimen in the Brussels Herbarium with the number 2242 and the names Lindsaea and Vittaria in Roxburgh’s hand (Morton photographs 5152, 19914). I believe that this specimen is not only a suitable lectotype, but that it is actually the holotype. According to Rox- burgh, the species was collected on Prince of Wales Island (i.e., Penang Island) by W. Roxburgh [Jr.]. There is a duplicate of this leectotype in Geneva, indicated as from the Prince of Wales Island, collected by Dr. Roxburgh (Morton photograph 6567). Roxburgh usually wrote his specific names on the specimens in his own collection, but he was lax about doing this with his spe- cies of Vittaria and Lindsaea. There are four Lindsaea specimens in Roxburgh’s collections in Brussels from Roxburgh’s own per- sonal herbarium, but only one of these (V. /unulata) has the name in Roxburgh’s hand. The other three specimens, however, can be ROXBURGH'S FERN TYPES—MORTON 321 matched up clearly with the described species. These four speci- mens represent the four species described from the Prince of Wales Island (Penang Island) for the collections of W. Roxburgh, Jr. Two of these (V. parasitica and V. interrupta) have labels in the hand of W. Roxburgh, Jr., giving something about the habitat. The remaining specimen, the one chosen above as lectotype of L. bipinnata, has no such label, but the plant corresponds with the description of L. bipinnata. Roxburgh remarked: “It is ex- ceeding like Vittaria parasitica, and only a little larger, and more robust.” This is the fact; the specimen is so much like V. parasitica that it does represent the same species, only the plant is a little larger and more robust than the type of V. para- sitica. It probably represents the terrestrial state of this species described as Lindsaea scandens var. terrestris Holttum, but it appears doubtful that this terrestrial form ought to be given any taxonomic recognition. Holttum indicates that it is larger than the epiphytic plants, but this might be expected since terrestrial plants would have better soil and nutrients than epiphytic plants. Roxburgh’s reason for describing this same species in two dif- ferent genera, Vittaria and Lindsaea, is that he misunderstood Vittaria. One has to remember that he was working in Calcutta only 10 or 15 years after the first publication of Vittaria, that he did not have Swartz’ “Synopsis Filicum,” of 1806 available, and that he had no authentic material of anything available for comparison. Roxburgh defined Vittaria as follows: ‘‘Fructifica- tions in an uninterrupted marginal line. Involucre double, unin- terrupted; one from the surface separating outwards; the other from the margin of the frond turned in, separating inwards.” Following is J. E. Smiths’ original description of Vittaria in 1793: “Fructif. in linea marginali continua. Involucrum duplex, continuum; alterum superficiarium, exterius dehiscens, aliud e margine ipsius frondis, inflexo, interius dehiscens.” This descrip- tion by Smith is highly misleading because it is completely wrong, and if we had only this description Vittaria would have to be listed as a dubious genus. The only way to identify Vittaria is by Smith’s citation of Pteris lineata L., for we know that this plant belongs to the genus Vittaria. But Pteris lineata does not have marginal sori and it does not have any “involucrum” (i.e., indusium), let alone an inner and an outer indusium, such as is present in such genera as Pteridium and Paesia. Roxburgh did have a true Vittaria, which he referred to V. lineata incorrectly, but his other Vittarias are all Lindsaeas. Where the leaf margin appeared to be slightly recurved, Roxburgh considered this as an 322 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM outer indusium and consequently described the “involucres” as “double” and the species as Vittarias. However, in the case of his Lindsaea bipinnata he wrote: “I am not certain if ever the thin edge of the leaf is turned in (over the inner) involucre. I am rather inclined to think not; at least I have not been able to discover that it is,’ and therefore he described this species as a Lindsaea, remarking how much it resembled his V. parasitica. There is a good reason why Roxburgh was not able to find an outer indusium, because this ‘‘difference” is illusory. In no Lind- saea is there ever anything corresponding to an outer indusium. In Lindsaea the margins of the fertile pinnules are sometimes just a little thinner and perhaps slightly recurved, but they never approach anything that might be called an outer indusium. 51, LINDSAEA ODORATA Roxburgh, Calcutta Journ. Nat. Hist. 4:511. 1844. Lindsaea cultrata sensu auctt. (e.g., Hook. & Grev. Icon. Fil. 2:t. 144. 1829 or 1830), non Adiantum cultratum Willd., 1794. LECTOTYPE: No herbarium specimen of this species has been seen. The drawing by Roxburgh at Kew, no. 2578 (Morton photograph 15860) was selected as lectotype by Kramer (Blume 15:567. 1967 [1968]). Roxburgh’s plant came from the Garrow Hills, i.e., Garo Hills, Assam, India. The illustration undoubtedly represents the species usually called L. cultrata incorrectly, as in the “Index Filicum.” 52. LYCOPODIUM ARISTATUM Roxburgh, Calcutta Journ. Nat. Hist. 4:473. 1844, non Humb. & Bonpl. ex Willd., 1810. = ? Selaginella tamariscina (Palisot) Spring, Bull. Acad. Brux. 10:136. 1843. Lycopodium aristatum Roxburgh, Hort. Bengal. 75. 1814, nom. nud. Cited is “Dill. Muse. t. 66, f. 7,” which might seem to validate the publication of Roxburgh’s name by this reference to Dillenius’ pre- Linnaean book, but in this case it does not, for there is no “t. 66, f. 7” in Dillenius’ work. This is a slip on Roxburgh’s part. It is not certain which Dillenius figure Roxburgh really intended. In the manuscript as published in 1844, Roxburgh states: “It resembles most Dillenius’s figure t. 66, f. 7,” thus repeating the same erroneous citation, and showing also that Roxburgh never intended his L. aristatum to be based on a Dillenius figure. 2*Stachygynandrum tamariscinum Palisot, Prodr. Aeth. 106. 1805. Type: East Indies (Isotype B, Willd. Herb. 19372-2, ex Palisot de Beauvois). TYPE: No herbarium specimens have been located. According to Roxburgh: “From China this pretty little species was brought to the Botanical Garden in Caleutta in 1812, where it grows freely in a rich, shaded, moist soil.” According to Roxburgh’s “Hortus Bengalensis” (p. 75. 1814) it was col- lected by W. Kerr in China in 1812. The original description is: “Patent (1-3 inches long), dichot- omous. Leaves as well as the superficial scales alternate, bifarious, ovate-falcate; membrane margined, apex ending in a bristle or arista.” In Reed’s “Index Selaginellarum” (p. 25. 1968), Lyco- r ROXBURGH’S FERN TYPES—MORTON 323 podium aristatum Roxburgh is equated to Selaginella uncinata (Desv.) Spring as though this were established. In conversation with Reed, it was stated that this disposition was taken from Alston’s manuscript “Index.” In this “Index,” in card form in the British Museum, under Selaginella aristata (Roxburgh) Scott, Alston has merely written casually “cfr. S. uneinata,” which is far from making a definite reduction to synonymy. It is apparent that Alston had not seen a specimen and that this was merely a guess from the description. I would not agree, for Selaginella uncinata is a rather large trailing species with blades often 12 inches long or more; also the leaves are not se conspicuously aristate that Roxburgh would have commented on it and even named his species “aristatum.’’ Roxburgh says his species is “little,” with the fronds only one to three inches long. This small size coupled with the conspicuously aristate leaves sug- gests at once S. tamariscina. I do not know where in China W. Kerr might have been, but from the early date (1812) it is likely that he was not too far from the coast, and indeed he might have been in Hong Kong, where S. tamariscina is common; this species does occur at other places along the China coast and lowlands of the interior. Since Roxburgh’s species is an illegitimate later homonym, its identification is not very important. 53, LYCOPODIUM CERNUUM sensu Roxburgh, Calcutta Journ. Nat. Hist. 4:472. 1844. —Lycopodium cernuum L. Sp. Pl. 1103. 1753, sens. lat. AUTHENTIC MATERIAL: Two specimens in the Brussels Herbarium with the name in the hand of Roxburgh (Morton photographs 19965, 19966). A specimen in the East India Company Herbarium, no, 139-6 [or 130-6?] with the name in Roxburgh’s hand (Morton photograph 19591 bis). Roxburgh understood L. cernuum in its usual sense. This spe- cies is really rather uniform, considering its extensive range in both hemispheres, and it is hard to see how Nessel could have recognized some 43 varieties. The segregation of any subspecies or varieties will be difficult or impossible. 54, LYCOPODIUM FILIFORME Roxburgh, Calcutta Journ. Nat. Hist. 4:473. 1844, non Swartz, 1806. =Lycopodium salvinioides (Herter) Alston in Lecomte, Fl. Gén. L’Indo-Chine 7(2) :5538. 1951. Urostachys salvinioides Herter, Bot. Arch., Konigsberg 3:18. Jan. 15, 1923.1 Type: Herter lists many syntypes; to my knowledge, no lecto- type has ever been chosen. ‘Republished in Phil. Journ. Sci. 22(1):67. Jan. 24, 1923, according to Herter (Ind. Lycopod. 102. 1944). 324 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM LECTOTYPE: A specimen in the Brussels Herbarium named in the hand of Roxburgh, the parts indicated by red letters as “c” and “d”; the fragment indicated as “e” has been considered different (Morton photographs 5218, 19976). The locality was indicated as Sumatra by Roxburgh. This and L. rotundifolium are the only species described by Roxburgh from Sumatra; there is no indication who may have supplied the material; so far as I am aware, Roxburgh never collected in Sumatra. In some herbaria, as in the British Museum, Lycopodium fili- forme Roxburgh is recognized under its own name as a distinct species. That can hardly be, however, since the name is an illegiti- mate later homonym. The type appears to represent a form of L. phlegmaria L. with rather small, spreading leaves and very elon- gate, slender spikes. The various varieties described by Blume are not clearly differentiated; it may be that Roxburgh’s species would agree better with var. pellucidum Blume or var. graciles- cens Blume than it would with var. laxwm. I have not seen the types of any of these varieties. 55. LYCOPODIUM FURCATUM Roxburgh, Calcutta Journ. Nat. Hist. 4:475. 1844. =Selaginella atroviridis (Wallich ex Hook. & Grev.) Spring, Flora 21:183. 1838. Lycopodium atroviride Wallich ex Hook. & Grev. Icon. Fil. 1: t. 39. 1827. Syntypes: Prince of Wales Island [=Penang Island, Malaya], Wallich and Herb. Hort. Soc. London. The Wallich specimen at Kew I designate as lectotype (Morton photograph 20656). The Wallich plants are the two upper and the center plants. The lower plant and perhaps the lower left-hand plant are from Tenasserim, collected by Griffith, and perhaps are a different species. An isosyntype is in the East India Company Herbarium, no. 120 (K, Morton photograph 19588e). The collection of the specimen that Hooker and Greville first saw in the herbarium of the Royal Horticultural Society was without indication of collector; the collector could have been Wallich, in which ease the two syntypes could have been part of the same gathering, or it is possibly a Roxburgh collection and an isotype of L. furcatum Roxburgh. Lycopodium intermedium Blume, Enum. Pl. Jav. 2:269, 1828. Type: Java, Blume (not seen). Selaginella intermedia (Blume) Spring, Bull. Acad. Brux. 10:144, 1843. LECTOTYPE: Roxburgh, East India Company Herbarium. no. 120-2 right- hand plant with name in hand of Roxburgh (K, Morton photographs 15719, 19589). The left-hand plant on this sheet is Wall. Cat. 120-4, Herb. Finlayson (not a type). An isotype is in the Brussels Herbarium labeled Lycopodium hymenophyllum in the hand of Roxburgh. Evidently Roxburgh first assigned the name L. hymenophyllum and later changed it to L. furcatum. Contrary to my usual designation of specimens in the Brussels Herbarium as lecto- types, I choose rather the one in the East India Company Herbarium which has the name furcatum in Roxburgh’s hand. Alston in his treatment of the Indian species of Selaginella adopted the name S. intermedia (Blume) Spring for this species, ROXBURGH’S FERN TYPES—MORTON 325 but he had the dates wrong. He indicated L. intermedium Blume as being published in 1830 and L. atroviride Wallich ex Hook. & Grev. in 1831, and if this were true then Blume’s name would have priority. Actually L. atroviride was published in 1827 and L. intermedium in 1828. It is clear that L. atroviride was pub- lished prior to Blume’s treatment, for Blume adopted the name atroviride for one of his species and cited the Hooker and Greville reference. This error has been perpetuated in Reed’s “Index Sela- ginellarum,” where also S. atroviridis is erroneously listed as “S. atrovirens.” 56. LYCOPODIUM IMBRICATUM Roxburgh, Calcutta Journ. Nat. Hist. 4:475. 1844, non Forssk., 1775. =Selaginella bryopteris (L.) Baker, Journ. Bot. Brit. & For. 22:376. 1884. Lycopodium bryopteris L. Sp. Pl. 1103. 1753. Concerning the typification, see Alston, Journ. Bot. Brit. & For. 69:252. 1931. TYPE: A Roxburgh specimen in the Brussels Herbarium with the name in the hand of Roxburgh (Morton photograph 19955). There is according to Alston a Roxburgh specimen of L. imbricatum in the DeCandolle Herbarium in Geneva. These specimens have been identified as Selaginella bryopteris by Alston (Proc. Nat. Inst. Sci. India 11:221. 1945), doubtless correctly. Roxburgh received his specimens from H. Colebrooke, who collected them in Behar, now Bihar, India. 57. LYCOPODIUM LAEVIGATUM sensu Roxburgh, Calcutta Journ. Nat. Hist. 4: 474, 1844, non Lam. 1791. =Selaginella plana (Desv.) Hieron. in Engl. & Prantl, Nat. Pflanzenfam. 1 (4) :7038. 1901. AUTHENTIC MATERIAL: A specimen in the Brussels Herbarium with the name in the hand of Roxburgh (Morton photograph 19956). A fragment of apparently the same plant is in the East India Company Herbarium, Kew, mounted on the same sheet as L. mimosoides, no. 128-2 (Morton photograph 19591a); this small fragment, named L. laevigatum by Roxburgh, was proba- bly overlooked by Alston in studying this sheet. Roxburgh indicated his ma- terial as from the “Malay Islands.” Roxburgh did not consider his species new, but cited Lamarck as the author. He, however, misidentified his plant, for L. laevi- gatum Lam. is a quite different plant from Madagascar, which is now properly Selaginella pectinata Spring [=Selaginella laev- igata (Lam.) Baker, 1867, non S. laevigata (Willd.) Spring, 1840] ; concerning the synonymy, see Alston im C. Chr., Dansk. Bot. Ark. 7:196. 1932. 58. LYCOPODIUM MIMOSOIDES Roxburgh, Calcutta Journ. Nat. Hist. 4:473. 1844. —Selaginella wallichii (Hook. & Grev.) Spring in Mart. Fl. Bras. 1(2) :124. 1840. Lycopodium elegans Wallich, Num. List. no. 128. 1829, non Desv. 1827. Based on a collection from Penang and Singapore, Wallich. Lycopodium wallichti Hook. & Grev. in Hook. Bot. Mise. 2:384. 1831. Type: Penang and Singapore, Wallich List. no. 128. 326 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM LECTOTYPE: A specimen in the Brussels Herbarium with the name in the hand of Roxburgh and the number 13 (Morton photograph 19958). Accord- ing to Roxburgh, the species was collected in the Moluccas. An isotype is in the East India Company Herbarium, no. 128-2, with the name in Rox- burgh’s hand (K, Morton photographs 15721, 19591). Another isotype is in the British Museum (Morton photograph 195380). This is a characteristic and easily recognized species. Rox- burgh remarked that the branches “are exactly like the pinnae of a fine-leaved Mimosa,” which is true and explains his choice of specific epithet. Alston (Gard. Bull. Str. Settl. 8:49. 1984) writes: “Roxburgh’s specimens of L. mimosoides Roxb. in Hb. Brit. Mus. are labelled ‘Ind. Orient.’; they were probably col- lected in Penang,” and a note by Alston on the sheet in the Brit- ish Museum referred to states that the “Moluccas” of Roxburgh included all the East Indies. This remark is not borne out by my studies. Roxburgh’s “Moluccas” really are the Molucca Islands, from where he received collections made by his assistant Christo- pher Smith; they were mostly from Amboina, but a few were from other islands—Honimoa, Ternate, and Gilolo. Selaginella wallichii has not since been collected in the Moluccas, but it could be there, since it does grow in Sumatra, not so far away. It is possible that in this instance Roxburgh did inadvertently write “Moluccas” rather than “Malay Islands,” the term he usu- ally used for Penang Island. 59. LYCOPODIUM PECTINATUM sensu Roxburgh, Calcutta Journ. Nat. Hist. 4:474, 1844, non Lam., 1791. =Selaginella willdenovii (Desv.) Baker, Gard. Chron. 1867:783, 950. 1867. AUTHENTIC MATERIAL: A specimen in the Brussels Herbarium named in the hand of Roxburgh (Morton photograph 19959, left-hand specimen). Ac- cording to Roxburgh a native of the Malay Islands. This sheet has a fragment of another species of Selaginella on it at bottom right, which is mounted here by mistake, for it is evidently a part of the authentic material of L. laevigatum sensu Roxburgh, i.e., Selaginella plana (Desv.) Hieron. An- other authentic specimen of L. pectinatum sensu Roxburgh is in the East India Company Herbarium, no. 126-7 (K, Morton photograph 19590c), with the name in the hand of Roxburgh and annotated as S. willdenovii by Alston; the other specimens on this sheet are different; they are from Cortallum and represent S. pouzolziana, fide Alston. Roxburgh did not intend this as a new species, but credited it to Lamarck. Lycopodium nectinatum Lam. was a mixture, based on plants from Mauritius and literature references to both East Indian plants and American plants. Desvaux (in Poiret in Lam. Eneycl. Méth. Suppl. 3:540. 1818 [1814]) essentially selected a lectotype by restricting the Lamarck name to the specimen from ROXBURGH’S FERN TYPES—MORTON 327 Mauritius and indicated that it was probably the same as Lyco- podium obtusum Swartz, i.e., Selaginella concinna (Swartz) Spring, the type of which is also from Mauritius. This disposi- tion has been generally accepted, but I do not know whether it has been actually verified. Swartz did not indicate any type for his plant, and so that will have to be determined by the consul- tation of Swedish herberia, first naturally Stockholm. However, Stachygynandrum obtusum Palisot (Prodr. Aethiog. 113. 1805), the type of which is from Mauritius, Bory, is generally considered to be the same, and since Palisot’s name clearly has priority over Swartz’, the proper name is Selaginella obtusa (Palisot) Spring, Bull. Acad. Brux. 10:228. 1843. 60. LYCOPODIUM PENDULUM Roxburgh, Calcutta Journ. Nat. Hist. 4:472. 1844, =Lycopodium carinatum Desv. in Lam. Encycl. Méth. Suppl. 3:555. 1813. LECTOTYPE: A specimen in the Brussels Herbarium with the name in the hand of Roxburgh (Morton photographs 5207, 19961). The locality is given by Roxburgh as Amboina, and therefore the collector was Christopher Smith. Since no other specimen named L. pendulum by Roxburgh has been found, this may be presumed a holotype. In Herter’s “Index Lycopodiorum,” L. pendulum is referred to L. carinatum Desv., and this appears to be correct. The type agrees with Robinson 1973 from Amboina (US). 61. LYCOPODIUM PHLEGMARIA sensu Roxburgh, Calcutta Journ. Nat. Hist. 4:471. 1844. =Lycopodium phlegmaria L. Sp. Pl. 1101. 1753, sens. lat. AUTHENTIC MATERIAL: Represented by a drawing at Kew, no. 1008 (Morton photograph 15885). No herbarium specimen has been seen. Roxburgh’s ma- terial came from “the Sunderbunds, on old trunks of trees, in flower during the rains.” The Sunderbunds are the present Sundarbans or Sundribuns, the swampy coastal region of the Delta of the Ganges in Bengal. Lycopodium phlegmaria as currently recognized in Asia is ob- viously a mixture of several species. The Roxburgh drawing is poor, but seems to correspond with a collection from Khasia, 2,000 feet alt., of Hooker and Thomson (dupl. US). It is a plant with somewhat sparse, narrowish, spreading leaves and short, slender, mostly simple spikes. According to Prain (Rec. Bot. Surv. India 2:231-370. 1903), L. phlegmaria is the only species of Lycopodium known from the Sundribuns. 62. LYCOPODIUM PLUMOSUM sensu Roxburgh, Calcutta Journ, Nat. Hist. 4:474. 1844, non L., 1753. =Selaginella roxburghii (Hook. & Grev.) Spring, Bull. Acad. Brux. 10:228. 1843. 328 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM Lycopodium plumosum Roxburgh ex Wallich, Num. List. no. 120-73. 1830, nom. nud. Lycopodium roxburghii Hook. & Grev. in Hook. Bot. Misc. 2:390. 1831. Type: Rorburgh ex. Herb. Wallich. Although it is not annotated as L. rorburghi by Hooker or Greville, the specimen labeled L. plumosum in the hand of Roxburgh in the East India Company Herbarium, originally sheet 120-93 but now filed under no. 122 (Morton photograph 15720, two small fragments at the lower right-hand bottom, indicated by Alston as “syn-type?” may be selected as lectotype until an un- doubted holotype is found, which might be in the Greville Herbarium in Edinburgh. In Reed’s “Index Selaginellarum” (p. 195) the type is indicated as “Dr. Wallich in Herb. Roxburgh, BM,” but I have not found any such specimen in the British Museum, although there may be one there; in any case, Reed has the names reversed, for it should be “Dr. Roxburgh in Herb. Wallich,” Roxburgh being the collector. AUTHENTIC MATERIAL: A specimen in the Brussels Herbarium with the name L. plumosum in the hand of Roxburgh (Morton photograph 19957). The specimen in the East India Company Herbarium mentioned above as the lectotype of L. roxburghii Hook. & Grev. is also authentic for Rox- burgh’s L. plumosum. In Reed’s “Index Selaginellarum” and by Alston, L. plumosum is listed as though it were a new species of Roxburgh, but Rox- burgh did not really consider it new. He did not place an “R” after the name, as he did with his truly new species, and he cited a reference to “Dill. musc. t. 66, f. 10,” which is one of the references cited under L. plumosum L. It seems therefore that Roxburgh considered his plant to represent L. plumosum L. This agrees with Crantz’ lectotypification of 1766, in which the Lin- naean L. plumosum is restricted to Dillenius’ t. 66, f. 10; accord- ing to this, L. plumosum L. is Selaginella plumosa (L.) K. B. Presl, a West Indian species, syn. S. stolonifera (Swartz) Spring. It is evident that Roxburgh misidentified his plants as L. plaumo- sum L., but he still did not describe a new species. Roxburgh stated that the species was “Native in various parts of India. In Bengal it is found on rotten wood.” I cannot reconcile this state- ment with the specimens mentioned above labeled L. plumosum by Roxburgh, for these plants represent Selaginella roxburghii, a rather characteristic species found in Malaya but not in India. In this instance Roxburgh likely used “India” to cover Malaya as well as India. From his comment about the species occurring in Bengal, it is clear that he included more than one species under his L. plumosum. 63. LYCOPODIUM ROTUNDIFOLIUM Roxburgh, Calcutta Journ. Nat. Hist. 4:473. 1844. =Lycopodium nummulariifolium Blume, Enum. Pl. Jav. 2:263. 1828. Type: Java, Blume (L, not seen). ROXBURGH’S FERN TYPES—MORTON 329 Type: Sumatra, Roxburgh Herbarium (E-Herb. Greville). According to Hooker and Greville (Icon. Fil. t. 212. 1831), Roxburgh’s collection is repre- sented in the East India Company Herbarium under no. 2183, but it is not there now; it may have been misplaced. This collection does not appear to be in the general herbarium at Kew, nor at the British Museum, Brussels, or Geneva. There is no doubt about the identification of L. rotundifolium Roxburgh, because Hooker and Greville’s plate was drawn from an authentic Roxburgh specimen, and it is identical with the characteristic species L. nummulariifolium Blume. The collector who supplied Roxburgh with the type material from Sumatra is unknown. 64, MARATTIA PINNATA Roxburgh, Calcutta Journ. Nat. Hist. 4:519, 1844. =Angiopteris pinnata (Roxburgh) Morton, comb. nov. Angiopteris ruttenii van Alderw. van Rosenb., Bull. Jard. Bot. Buiten- zorg II, 28:6. 1918. Type: Boren Kaloa, Ceram, 150 m., Sept. 21, 1917, Rutten 177 (holotype prob. BOG; isotype L, Morton photograph 439). LECTOTYPE: A specimen in Brussels with the name in the hand of Roxburgh (Morton photograph 19951). Isotypes in Geneva (Morton photograph 16993), British Museum (Morton photograph 6529), and Kew (East Ind. Co. Herb. 187-7, consisting of two pinnules only, Morton photograph 15725). This species has never been identified. In the “Index Filicum,” it is referred with a query to A. crassipes Wallich ex K. B. Presl, but this is only on the basis of Wallich’s having placed the single pinnule that Roxburgh evidently sent him under his A. crassipes. (The pinnules in this species are deciduous from the rhachilla; Roxburgh probably sent more material, but the pin- nules fell off and Wallich retained only two.) The species ap- pears to be different from the Indian and Nepalese A. crassipes, but it has priority, the Wallich species not having been published except as a nomen nudum until 1845, a year after Roxburgh’s species. Bommer identified the lectotype as A. dregeana deVriese (1853), which is according to the “Index Filicum’” a synonym of A. javanica K. B. Pres] (1845). It does not appear that either of these can be the same as Roxburgh’s species, since they belong in the subg. Angiopteris, whereas A. pinnata belongs in the subg. Pseudangiopteris K. B. Presl. Angiopteris is the fern genus most in need of monographic study. More than a hundred species have been proposed, but why is something of mystery. They all look much alike and do not differ in obvious characters like most good species of ferns. There may be only a very few good species in the genus. I ven- ture to transfer Roxburgh’s species, however, because it is the 330 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM earliest species belonging to the subg. Pseudangiopteris, all the others having been proposed later by Presl, deVriese, and others. Pseudangiopteris may not really be a good subgenus, although the monographer DeVriese thought it was, but at least the character on which it is based—the absence of false veins—seems to be a good specific character. Roxburgh gave the locality as “Molucca Islands.” The collections available to him from the Moluccas were collected by C. Smith in Amboina, Ternate, or Honimoa (Ceram). The lectotype and isotype specimens that I have examined agree with the isotype of A. ruttenii, from Ceram, and so A. pinnata likely came from Honimoa. 65. MARSILEA QUADRIFOLIA sensu Roxburgh, Calcutta Journ. Nat. Hist. 4:469. 1844. =Marsilea quadrifolia L. Sp. Pl. 1099. 1753. Type: Not deter- mined; India was included within the original geographic range assigned by Linnaeus. AUTHENTIC MATERIAL: A Roxburgh drawing in Kew (no. 1300) and a specimen in the herbarium at Edinburgh. Another specimen, determined as M. dentata Roxb. and dated 1809, should be at the British Museum (Natural History). Roxburgh illustrated the sporocarp, and it corresponds with M. quadrifolia as delimited by K. M. Gupta in his book “Marsilea” (Council Sci. Industr. Res., New Delhi, Bot. Monog. 2:96. 1962). Roxburgh indicated that he had seen this plant in various parts of India, and that he had searched for M. coromandelica Burm. f. without success. 66. OPHIOGLOSSUM CORDIFOLIUM Roxb. Calcutta Journ. Nat. Hist. 4:475. 1844. = Ophioglossum petiolatum Hook. Exotic Fl. 1:56. 1823. Ophioglossum cordifolium Roxb. ex Wall. Num. List., no. 47. 1828, nom. nud. LECTOTYPE: A sheet in the herbarium in Brussels with the name in the hand of Roxburgh and the number 1199 (Morton photograph 19952). There is a comment by Roxburgh on the label: “Yet I now doubt if it differs specifically from O. vulgatum.” Roxburgh’s description agrees with this herbarium specimen. The comment that it is native to ‘Bengal, Moluccas, ete.” indi- cates that Roxburgh had more than one specimen in hand, and that the sheet chosen as lectotype is thus not a holotype. The sheet of no. 47 in the herbarium of the East India Company at Kew (Morton photograph 14672) has several fronds on it. The three at the top left represent Wallich no. 47-1 from Viemora, Bengal, collected in 1808 (undoubtedly by Roxburgh) and are undoubtedly isosyntypes; the single frond at upper right ROXBURGH’S FERN TYPES—MORTON 331 appears to be a part of the same collection. The one larger plant, the second from the right in the top row, has a label “Inter Mangiferas, H. B. C. Sept. 1825”; “H. B. C.” refers to the Hort. Bot. Calcutta, and the collector was doubtless Wallich; this is not type material. The lower row of four plants are Wallich no. 47-2, collected in Courtallum by Heyne; they formed a part of Wallich’s concept of O. cordifolium but not that of Roxburgh, although they do represent the same species. A du- plicate specimen of Wallich 47-1 from Viemora, Bengal, Roxburgh is in US. In the “Index Filicum,” O. cordifolium is wrongly referred to O. pedunculosum Desv. There are drawings of O. cordifolium by Roxburgh in the British Museum (Morton photograph 15774) and at Kew. 67. OPHIOGLOSSUM FILIFORME Roxburgh, Hortus Bengalensis 75. 1814; Calcutta Journ. Nat. Hist. 4:476, t. 26, f. 3. 1844. —Lygodium micro- phyllum (Cav.) R. Brown, Prodr. Fl. Nov. Holl. 162. 1810. Ugena microphylla Cav. Icon. Descr. Pl. 6:76, t. 595, 1801. Syntypes: Marianna Islands and Philippine Islands [near Nabua (Prov. Cama- rines Sur, Luzon), and Samboangan], Née (MA, not seen). Type: In the Hortus Bengalensis there is no description, but Rheede, Hort. Malab. 12: t. 34, is cited, which is sufficient to validate the publication of the name O. filiforme. Rheede’s poor illustration t. 34 is identified by Willdenow (in L. Sp. Pl. ed. 4, 5:78. 1810) as a sterile plant of the species that he calls Hydroglossum scandens, of which Ugena microphylla Cav. is cited as a synonym; Rheede’s t. 34 is similarly identified with L. scandens by Prantl (Untersuch. Morph. Gefasskrypt. 2:82. 1881). A specimen from the Roxburgh Herbarium now in Brussels is labeled O. filiforme in Roxburgh’s hand (Mor- ton photograph 4764, 19945) and surely represents Roxburgh’s concept and the species that he intended to describe. According to Roxburgh (Hort. Beng. 75. 1814) it was collected in Bengal by Dr. F. Buchanan and was cultivated in the Calcutta Botanical Garden. This specimen does represent Lygodium microphyllum, and so it appears that Roxburgh did interpret Rheede’s plate correctly. This specimen also agrees with Roxburgh’s published plate ¢. 26, f. 3 and with the authentic Roxburgh drawings labeled O. filiforme in the British Museum (Morton photograph 15572) and at Kew (no. 1741), which are indicated as having been collected at Chittagong by Buch- anan-Hamilton; in Roxburgh’s time Chittagong (now in East Pakistan) was included within Bengal. Thus there are no confused elements in this species. This species is the one that has usually, at least since the time of Swartz’ “Synopsis Filicum” (1806), been identified as Lygodium scandens (L.) Swartz, and it is too bad that its name must be changed. Alston and Holttum (Reinwardtia 5:12-14. 1959), however, showed that the original Ophioglossum scandens L. was based on several elements, none of which represent L. 332 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM scandens as interpreted by Swartz and succeeding authors. There were four literature citations, the one from the “Hortus Cliffortianus” representing L. volubile Swartz, the one from the “Flora Zeylanica” L. flexuosum L. (fertile), and the one from Breyne (copied by Morison) L. venustum Swartz, and finally the one from Rheede representing L. flexcuoswm. They concluded that since L. venustum and L. volubile already had received names, and generally accepted names, that O. scandens L. ought to be typified by the “Flora Zeylanica” reference and the cor- responding specimen in the Hermann Herbarium (no. 474). Strangely enough, they make no mention of the Linnean Her- barium, in which there is a specimen (no. 1243.3 labeled scandens in Linnaeus’ hand and also numbered “5,” the number of O. scandens in the “Species Plantarum” (1753). It is generally agreed that when possible Linnaean species ought to be typified by specimens in the Linnean Herbarium rather than on literature citations, and in this case there is no obstacle. Fortunately for stability, this specimen in the Linnean Herbarium, which I designate as lectotype, does not represent L. venustum or L. volubile, but is clearly L. flexuosum, and thus O. scandens L. and L. scandens (L.) Swartz remain as synonyms of L. flexuosum (L.) Swartz, as indicated by Alston and Holttum. 68. OPHIOGLOSSUM FLEXUOSUM sensu Roxburgh, Calcutta Journ. Nat. Hist. 4:477, t. 26 p.p. (lower sterile leaflet). 1844. —Lygodium circinnatum (Burm. f.) Swartz, Syn. Fil. 153. 1806. The only herbarium specimen that I have seen determined by Roxburgh as O. flexuosum is one in Brussels, from Amboina, no. 320, collected in 1796 (Morton photograph 19941) ; although not stated on the label the collector was surely Christopher Smith. There is a duplicate at the British Museum (Morton photograph 20856). This is clearly L. circinnatum with twice-forked sterile pinnae. That this really represents Roxburgh’s concept at least in part is shown by his description of the sterile pinnae as generally palmate. Roxburgh did, however, include other material also in his concept, for he cites the Bengal name as “Bhootraj” and cites “Valli-panna. Rheed. Mal. 12, t. 32. bad,” the latter the same reference as cited by Linnaeus under his Ophioglossum flexruosum and cited also by Pres] as his L. salicifolium. 69. OPHIOGLOSSUM FURCATUM Roxburgh, Calcutta Journ. Nat. Hist. 4:478. 1844. =?Lygodium circinnatum (Burm. f.) Swartz, Syn. Fil. 153. 1806. No specimens of this species have been located. It was described from Pullo Pinang (i.e., Penang Island, Malaya). In the ‘Index ROXBURGH'S FERN TYPES—MORTON 333 Filicum” it is referred to L. circinnatum, which is probably cor- rect. The original description reads: ‘“Scandent. Fronds dichoto- mous, ultimate divisions linear, very long, finely pinnatifid, with numerous minute spikelets.” 70. OPHIOGLOSSUM SCANDENS Roxburgh, Calcutta Journ. Nat. Hist. 4:477, t. 26 p.p. 1844. =Lygodium flexuosum (L.). Swartz, p.p. and L. salici- folium K. B. Pres], p.p. Roxburgh intended his O. scandens to be that of Linnaeus, and so it is, as now lectotypified by Alston and Holttum, i.e., a syno- nym of L. fleruosum (L.) Swartz. The published ¢t. 26 (lower fertile pinnules) also represents L. flexuosum, as do three herbarium specimens that I have seen: Sin loc., Roxburgh (G, Morton photograph 6573), E. Ind. Co. Herb. 175-10 (K, Morton photograph 15723), and Amboina, Christopher Smith in 1797 (Linn. Soc., Smith Herb. 1647-7, Morton photograph 20333). It appears, however, that Roxburgh’s concept was confused, because the specimen in Brussels (Morton photograph 19944) labeled O. scandens in Roxburgh’s hand is L. salicifolium K. B. Presl. It appears that the drawing at BM (Morton photograph 15771) and at Kew (no. 1200) is a composite, the lower sterile pinnae being shown as forked somewhat as in L. circinnatum (Burm. f.) Swartz and the fertile pinnae pinnate as in L. flexuosum. In the paper mentioned under O. filiforme by Alston and Holttum, these authors lectotypify Ophioglossum flexuosum L. on the basis of “Fl. zeyl. 375,” i.e, a Hermann specimen from Ceylon, mistakenly stating that this is the only reference cited by Linnaeus, who in point of fact did cite another reference, namely Rheede, Hort. Malab. 12:6, t. 32. Since Linnaeus gives the locality as “India,” and not Ceylon, it might be supposed that the Rheede reference could be chosen as lectotype; still he did cite the “Fl. zeyl.” reference and so the “India” habitat is incom- plete. There is a further complication not mentioned by Alston and Holttum, namely that there is a specimen in the Linnean Herbarium named by Linnaeus as flerwosum and also indicated as being species 6 of the ‘Species Plantarum.” The Linnaean specimen represents the Jamaican species later described as Lygodium volubile Swartz. Since Linnaeus did not cite Jamaica or any American locality for his O. flerwosum, but did cite India, the Linnaean specimen must in this case be rejected as a type. It would be most unfortunate to transfer the application of the epithet flexuosum from a common Old World species to a common New World one. 334 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM Alston and Holttum (Reinwardtia 5:15. 1959) write that Singapore, Cuming 365 in Vienna is the type of L. salicifolium K. B. Presl, but this is an error, for Pres] cited material from Penang, Singapore, Silhet, Irawaddi, Ava, Rajmahl, Monghur, Gualpara, Tavoy, Martaban, Nepal (all distributed by Wallich), Singapore (Cuming 365), and Java (Drege 20). The Cuming collection is thus just one of many syntypes, all of which are not conspecific; however, Cuming 365 is a suitable lectotype, although there is no reason to designate a specimen in Vienna; on the contrary, there is a specimen in the Pres] Herbarium in Prague that should be, and is here, designated as lectotype (cf. Holttum, Novit. Bot. Inst. Bot. Univ. Carol. Prag. 1968:37. 1969). 71. OSMUNDA LANCEOLATA Roxburgh, Calcutta Journ. Nat. Hist. 4:478, t. 27, 1844, non Gmelin, 1768. =Leptochilus decurrens Blume, Enum. Pl. Jav. 206. 1828. Lectotype: “Ad pedem montis Burangrang inter lapida ad rivulos,” Java, Reinwardt (L, Morton photograph 1582). This specimen is probably the actual holotype, for it has the name in Blume’s hand; it shows one frond with the blade long-decurrent and another with the blade less decurrent, but this variation is common in this species. LECTOTYPE: The published drawing, t. 27, left-hand plant. No specimen named Osmunda lanceolata has been located in any herbarium. The drawing is good; it was a new species when Roxburgh described it (although he picked out a later homonym for his name), but the same plant was later described by Blume as Leptochilus decurrens. The published drawing has two sterile leaves and one fertile. Strangely enough, the other two drawings seen are different but represent the same species. The one at Kew (no. 694, Morton photograph 15865) has four sterile leaves and two fertile, and the one at the British Museum (Morton photograph 15753) has six sterile leaves and two fertile. If a specimen can be located, it will replace the drawing as lectotype. Roxburgh did not indicate a locality, but, according to Merrill, it is from Penang. The genus Leptochilus has been studied by Sledge (Ann. Mag. Nat. Hist. XII, 9:865-877. 1956), and additional synonymy for L. decurrens is given in his paper. 72. OSMUNDA ZEYLANICA sensu Roxburgh, Calcutta Journ. Nat. Hist. 4:478. 1844. =Helminthostachys zeylanica (L.) Hook. iz Hook. & Bauer, Gen. Fil. t. 47. 1844. No specimen collected by Roxburgh has been seen, but none was searched for, since Roxburgh’s use of the name O. zeylanica was surely correct. There is a drawing by Roxburgh (no. 1742) at Kew (Morton photograph 15866). According to Roxburgh ROXBURGH’S FERN TYPES—MORTON 335 (Hortus Bengalensis 75. 1814), his specimen came from Amboina where it was collected by Christopher Smith. 73. POLYPODIUM ACUMINATUM Roxburgh, Calcutta Journ. Nat. Hist. 4:490. 1844, non Houtt., 1786, non D. Don, 1825. =Thelypteris arida (D. Don) Morton, Amer. Fern Journ. 49:113. 1959. Aspidium aridum D. Don, Prodr. Fl. Nepal. 4. 1825. Type: Nepal, Wallich (BR? photograph BM). LectotyPeE: A Roxburgh specimen in the Brussels Herbarium with the name in the hand of Roxburgh (Morton photograph 19649). An isotype is in Geneva, indicated as “India orient. Dr. Roxburgh” (Morton photograph 16931). Roxburgh did not indicate a locality for this species. This Polypodium acuminatum Roxburgh was based on Rox- burgh’s original manuscript and was published independently and without mention of P. acwminatum Roxburgh ex D. Don (1825). The latter was based on a Nepal collection of Wallich, whereas Roxburgh’s own species was based on one of his own collections. Therefore, it seems better to regard these two species as nomenclaturally different and differently typified even though they do seem to be taxonomically equivalent. They are both later homonyms. The lectotype is unusually large for T. arida and unusually glabrous. It seems to agree with material from Penang Island, and it may have been collected there by W. Roxburgh, Jr. There is another Roxburgh specimen in Brussels with a label by Rox- burgh that reads: “agrees with P. acuminatum, but smaller” (Morton photograph 19650); Roxburgh’s comment is true, for it does represent the same species, but it is a smaller, hairier, and more typical specimen of TJ. arida. Another Roxburgh collection in Brussels is labeled by Roxburgh and _ reads: “Polypodium undetermined; differs from acuminatum in being pubescent and having a larger and more lasting involucre”’; this specimen too I identify as T. arida. 74. POLYPODIUM ACUTUM Roxburgh, Calcutta Journ. Nat. Hist. 4:492. 1844, non Burm. f., 1768. =Thelypteris subalpina (v.A.v.R.) Reed, Phytologia 17:317. 1968. Dryopteris subalpina v.A.v.R. Bull. Bot. Jard. Buit. III, 5:200. 1922. Type: North Foramadiahi, Ternate, 1200 m alt, Béguin 1496 (BO; isotype L, seen by Holttum). LEcTOoTYPE: A Roxburgh specimen in the Brussels Herbarium with the name in Roxburgh’s hand and the number 2390 (Morton photograph 19662). There is in Brussels also a sheet containing two pinnae only, doubtless a part of the same collection. According to Roxburgh, the species occurs in Amboina. An isotype is in Geneva marked as “Amboyna, Dr. Roxburgh,” which has been identified (not by Roxburgh) as “Polypodium cultratum 336 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM Rox.” There is no such species, and this is certainly an error for P. acutum Roxburgh (Morton photographs 6544, 16632). A specimen in the J. E. Smith Herbarium (no. 1622-75) in the Linnean Society, collected in Honimoa [Ceram] by Christopher Smith, April, 1797 (Morton photograph 20354), agrees with the lectotype and is probably an isotype; Roxburgh probably thought that all the Smith collections came from Amboina, but this one was likely among those that came from Honimoa. The lectotype chosen was identified by Baker as Nephrodium extensum, but the specimen is by no means that species or nearly allied. The lowest pair of veinlets unite into an excurrent veinlet; the second pair reach the margin well above the sinus, which is necessarily almost true since the pinnae are deeply pinnatifid with only a narrow costal wing. 75. POLYPODIUM AEMULUM sensu Roxburgh, Calcutta Journ. Nat. Hist. 4:496. 1844, non Aiton, 1789. —Hypodematium crenatum (Forssk.) Kuhn, von Deck. Reise 3(3) :37. 1879, Polypodium crenatum Forsk. FI. Aeg.-Arab. 185. 1775. Type: Forsskal (C not seen, photograph BM). AUTHENTIC MATERIAL: A Roxburgh specimen in the British Museum, marked “Polypodium aemulum?” in Roxburgh’s hand. According to Roxburgh, his material came from “mountains north of Rohileund,” i.e., the present Rohilkhand, Province of Agra, Northern United Provinces. The description by Roxburgh is: “Fronds quadri-pinnatifid, smooth, and delicate; ultimate divisions narrow-trapeziform- oblong, sides pinnatifid, or gashed and denticulate at tip. Fruc- tifications solitary on the segments. Involucre reniform, peltate and ciliate.” Roxburgh surely did not know Polypodium aemulum Aiton at first hand, but only through a description. His plant might seem to resemble that by a brief diagnosis, but it was different. Polypodium aemulum is a species of Dryopteris from the Azores, and nothing like it is known in India. The identity of Roxburgh’s plant required a little detective work. The descrip- tion of the indusia as “ciliate’ ruled out Dryopteris, Polystichum, and Arachniodes. The quadripinnatifid, delicate blades and hairy indusia suggested Hypodematium, and I did find a Roxburgh specimen under H. crenatum in the British Museum labeled by Roxburgh as “P. aemulum ?.” The query shows that Roxburgh had some doubt about the identity of his plant with that of Aiton. The indusium in Hypodematium is not truly peltate, as Roxburgh described it, although it might appear so to casual inspection; it is pubescent all over, rather than merely ciliate. 76. POLYPODIUM ARBORESCENS Roxburgh, Caleutta Journ. Nat. Hist. 4:495. 1844, =Cyathea batjanensis (Christ) Copel. Phil. J. Sci. 40:45. 1909. ROXBURGH’S FERN TYPES—MORTON 337 Alsophila batjanensis Christ in Warb. Monsunia 1:90. 1900. Type: Mt. Sibella, Batjan, Moluccas, 5000 ft alt, Warburg 17828 (P not seen; isotype B not seen). Polypodium arboreum Roxburgh ex Wallich, Num. List no. 2226. 1830, nom. nud. Alsophila latebrosa var. batjanensis Christ in Warb. Monsunia 1:89. 1900, nom. nud. Lectotype: A Roxburgh collection in the Brussels Herbarium with the name in Roxburgh’s hand and the number 2406 (Morton photograph 19859). An isotype (probably the apical part of the same frond) is in Geneva, labeled as from “Honimoa, Dr. Roxburgh” (Morton photograph 6545). Roxburgh gave the locality as Honimoa, i.e., Ceram, where it was doubtless collected by Christopher Smith. In the original description the editor, Griffith, added a refer- ence to Wallich, Num. List no. 2226 with a query, a number undoubtedly referring to the present species, whose name was wrongly transcribed by Wallich as “arboreum” rather than arborescens. The specimen noted by Wallich was surely the lectotype specimen noted above from “Herb. Roxb.,” which was at the time in the East India Company Herbarium under the custodianship of Wallich. 77. POLYPODIUM ATTENUATUM Roxburgh, Calcutta Journ. Nat. Hist. 4:482. 1844, non Humb. & Bonpl. ex Willd., 1810. —Polypodium lanceolatum L. Sp. Pl. 1082. 1753. Pleopeltis macrocarpa (Bory ex Willd.) Kaulf. Berlin Jahrb. Pharm. 21:41. 1820. See Pichi-Sermolli (Webbia 20:349-354. 1965) for addi- tional comments on the nomenclature of P. lanceolatum. Lectotype: A Roxburgh specimen in the Brussels Herbarium bearing the name P. attenuatum and a label “Typus” (Morton photograph 5007). A duplicate is in the British Museum (Morton photograph 7611). According to Roxburgh the species was collected in Pullo Pinang, i.e., Penang Island, Malaya. From my photographs and notes on the type specimens, it seems that the plants are truly P. lanceolatum L., but that species has been recorded in Asia only from India and not from Penang Island or elsewhere in Malaya. Perhaps Roxburgh’s locality was wrongly stated, and the specimens came from southern India. 78. POLYPODIUM CONFERTUM Roxburgh, Calcutta Journ. Nat. Hist. 4:493. 1844. =Arcypteris irregularis (K. B. Presl) Holttum, Reinwardtia 1:198. 1951. Polypodium irregulare K. B. Presl, Rel. Haenk. 1:25. 1825. Type: “Mexico,” Haenke (the locality Mexico is presumably wrong and the type probably came from the Philippine Islands; it is very likely in Prague). Aspidium difforme Blume, Enum. Pl. Jav. 1:160. 1828. Type: “In regione Buitenzorg,” Java, Blume (holotype L, with the name in Blume’s hand, Morton photograph 2318). 338 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM Polypodium eximium Kunze, Bot. Zeit. 4:424, 1846. Lectotype: Two sheets in Geneva from Java, Zollinger 514A, the first sheet (the blade apex) with the name “Polypodium eximium Kaze. n. sp.” in Kunze’s hand, the second sheet with part of the stipe and blade base (Morton photographs 3851, 3852). The other specimen cited by Kunze, Java, Zollinger 514, is also represented by two sheets in Geneva (Morton photographs 3853, 3854), but these do not have the name in Kunze’s hand. The two above collections may indeed be the original specimens used by Kunze in drawing up his description, and there may never have been any specimens in Kunze’s own herbarium, now destroyed. Tectaria irregularis (K. B. Presl) Copel. Phil. Journ. Sci. 2C:416. 1907. Polypodium macrodon Reinwardt, mss. Based on a specimen in Leiden marked “H. B. [i.e., Hort. Buitenzorg] Polypodium macrodon R.” (Morton photograph 2317). The epithet was picked up and used by Pres] as Dictyopteris macrodonta and by Fée under the same name, but for a species that proves to be different, a Philippine species typified by Cuming 9[=Areypteris macrodonta (Fée) Holttum, Rein- wardtia 1:194. 1951]. The Reinwardt specimen with the pinnae (except the basal) very little lobed is matched by a specimen collected near Buitenzorg in 1909 (Palmer & Bryant 93, US). Lectotype: A Roxburgh specimen in the Brussels Herbarium with the name in Roxburgh’s hand (Morton photographs 5188, 19800); this is the basal part of a frond, with the pinnae almost fully again pinnate. A second specimen in Brussels is probably the apical part of the same frond (Morton photographs 5187, 19801). According to Roxburgh, this species came from Chittagong, in East Bengal, now East Pakistan. In the “Index Filicum,” this species was left as dubious; it has never been placed, except as Baker has correctly annotated the second sheet mentioned above as Polypodiwm difforme Blume. I follow Holttum with some hesitation in recognizing Arcypteris as a genus distinct from Tectaria. The characters need to be investigated further. Sinus teeth, the presence of which is stressed by Holttum, seem to be more often absent than present, at least on old, mature fronds, and also on fronds with the pinnae only very slightly lobed. Not enough is known about the pubescence of Tectaria in general to know if the characters mentioned are distinctive. Holttum’s statement that the sori in Arcypteris are borne at the vein junctions is not borne out by his own figure 2, in which more sori are not at vein junctions than are. The character of sori irregularly scattered and ex- indusiate also occurs in various New World species. 79, POLYPODIUM CONFLUENS Roxburgh, Caleutta Journ. Nat. Hist. 4:494. 1844, non R. Brown, 1810. =Ctenitis rhodolepis (Clarke) Ching, Bull. Fan Mem. Inst. Biol. Bot. 8:300. 1938. Nephrodium rhodolepis Clarke, Trans, Linn. Soc. II, Bot. 1:526, t. 72. 1880. Syntypes: “Sikkim, Assam, Khasia, alt. 5000-7000 feet; Chit- tagong, 150 feet, Clarke (very small form).” It is evident that Clarke ROXBURGH’S FERN TYPES—MORTON 339 had several specimens in hand, which must be examined before a lectotype can be designated. Dryopteris rhodolepis (Clarke) C. Chr. Ind. Fil. 288. 1905. Type: A Roxburgh specimen in the Brussels Herbarium with the name in the hand of Roxburgh and a label reading “Typus” (Morton photograph 19722). According to Roxburgh, it came from Chittagong, in East Bengal, now East Pakistan. Since this is the only specimen found in any herbarium, it may be presumed to be a holotype. In the “Index Filicum,” P. confluens Roxburgh, a homonym several times over, is left as a dubious species. The type has been identified as Nephrodium intermedium Baker by Baker, which it doubtless is, but Baker’s name is a later homonym. The type is quite the smallest specimen of this species that I have seen, and it evidently corresponds to the other specimen from Chittagong mentioned by Clarke in the original description of N. rhodolepis as a “very small form”; it is not only smaller but less divided, being bipinnate only at the very base. The large, thin, beautifully clathrate scales are characteristic of this species. 80. POLYPODIUM CORIACEUM Roxburgh, Calcutta Journ. Nat. Hist. 4:481, t. 28 (left). 1844, non Swartz, 1788. =Pyrrosia angustata (Swartz) Ching, Bull. Chin. Bot. Soc. 1:49. 1935. Polypodium angustatum Swartz, Syn. Fil. 27, 224. 1806. Type: Tranque- bar, Rottler (prob. S-PA, not seen). 1 Niphobolus sphaerocephalus Hook. & Grev. Icon. Fil. 1:¢. 94. 1828. Type: Wallich (K-Hb. Hook., Morton photograph 20652), which was thought by Hooker and Greville to be perhaps from Nepal, but indi- cated properly by Wallich under his Num. List. no. 272. 1829, as from Singapore. Hooker and Greville described the sterile leaves as completely glabrous beneath, but this is not true for P. angustata. The sterile blades of the type do appear to be glabrous beneath, but they are old and the stellate hairs which were once present have fallen off, except for a few. Normally this species is not glabrescent; however, it is not clearly conspecific with P. angustata or with Wallich 272 in the East India Company Herbarium (Morton photo- graph 20657). ? Polypodium sphaerocephalum (Hook. & Grev.) Wallich, Num. List no. 272. 1829. Based on Niphobolus sphaerocephalus Hook. & Grev. LECTOTYPE: Roxburgh in East India Company Herbarium, no. 272-3 (K, Morton photograph 15737, left-hand specimen). The center and right-hand specimens on this sheet are Wall. Num, List. no. 272-2, Herb. Finlayson, not a type. According to Roxburgh, his species came from “Malay Islands and Pullo Pinang, whence introduced into this Garden [Calcutta] by Mr. W. Roxburgh, Jun.” This lectotype likely came from Penang Island, where the species is common, for it matches material from there. A probable duplicate of the lectotype is in US, marked as Penang, Dr. Wallich no, 272. Wallich’s own no. 272, the type material of Niphobolus sphaerocephalus Hook. & Grev., came from Singapore [Hooker and Greville guessed Nepal wrongly in the original publication], and Wallich’s only material of this 340 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM species from Penang would presumably have been received from Roxburgh, and so this specimen is doubtless collected by Roxburgh; it agrees with the lectotype cited above. A Roxburgh drawing at Kew (no. 1744, Morton photograph 15867) is labeled Polypodium coriaceum and is doubtless intended for this species; it shows the rhizome as too thick and is otherwise rather poor. 81. POLYPODIUM CUSPIDATUM Roxburgh, Calcutta Journ. Nat. Hist. 4:491. 1844, non D. Don, 1825. =Thelypteris repanda (Fée) Morton, comb. nov. Phegopteris repanda Fée, Gen. Fil. 251. 1852. Type: Penang, Gaudichaud (Hb. Webb-FI, not seen). Polypodium urophyllum Wallich, Num. List 299. 1829, nom. nud. Based on Penang, Wallich in 1822, Assam, Wallich in 1827, and a var. angusta (nom. nud.), Penang, Wallich in 1822. Goniopteris dalhousiana Fée, Mém. Soc. Hist, Nat. Strasb. 5(1) :92. 1857. Type: Penang, Lady Dalhousie (isotype K, not seen). Phegopteris urophylla Mett. Abhandl. Senckenb. Naturf. Gesell. 2:310. 1858. Syntypes: Nepal, Wallich; Borneo, Meissner; plus an unnamed variety, Borneo, Meissner. The locality “Nepal” is surely an error (although perhaps not Mettenius’ error if the specimen that he had was erroneously labeled); Mettenius’ description (in the sporangia being setose, the pinnae caudate, etc.) shows that he had a specimen of Wallich 299’from Penang, for his description agrees with that, and there are no Wallich specimens of this species from Nepal, where the species is not known to occur. I designate the Wallich specimen of this species (presumably under no. 299) in Berlin as the lectotype, if one is there; isotypes are Wallich 299 in Brussels (Morton photo- graphs 19744) and in the East India Company Herbarium, no. 299 (K, Morton photograph 15740). The Meissner specimen from Borneo is presumably 7. urophylla var. nitida (Holttum) K. Iwatsuki, Acta Phytotax. Geobot. 22:94. 1966.7 Polypodium urophyllum (Mett.) Hook. Sp. Fil. 5:9. 1864. Hooker wrongly attributed the species to Wallich, but Wallich’s name was a nomen nudum. Since the species was validly published only by Mettenius in 1858, long after Wallich’s death, Wallich cannot reasonably be said to have transferred it in 1864. Hooker cited Phegopteris uro- phylla Mett., but also some other synonyms erroneously, such as Polypodium asperum K. B. Presl. Polypodium pinwillii Baker, Ann. Bot. 5:460. 1891. Type: Malacca, Pinwill (K, not seen). Abacopteris urophylla (Mett.) Ching, Bull. Fan Mem. Inst. Biol. Bot. 8:251. 1938 (Ching cited the parenthetical author wrongly as “Wall.”). Thelypteris urophylla (Mett.) K. Iwatsuki, Southeast Asian Studies 3(3) :81. 1965 (parenthetical author wrongly cited as “Wall. ex Hook.”’). * Thelypteris urophylla var. nitida (Holttum) K. Iwatsuki was first pro- posed in Acta Phytotax. Geobot. 21:171. August, 1965, but the specific com- bination 7. urophylla was not made until December, 1965; it is not possible to make a legitimate varietal combination in advance of the valid publica- tion of its species combination. ROXBURGH’S FERN TYPES—MORTON 841 Pronephrium repandum (Fée) Holttum, Blumea 20:109. 1972. LECTOTYPE: Two Roxburgh specimens in the Brussels Herbarium, evidently a part of the same collection, the sheet with the apex with the name in the hand of Roxburgh and the date July 4, 1802 (Morton photographs 19742, 19743). According to Roxburgh, the species was collected in Pinang, i.e., Penang Island, Malaya, presumably by W. Roxburgh, Jr. Isotypes are in Geneva, indicated as from Prince of Wales Island, i.e., Penang Island (Morton photograph 16943) and in the East India Company Herbarium, no, 299-4 (K, Morton photograph 15739). The reference given in the original description to Polypodiwm urophyllum Wall. Cat. 64, no. 299, was added by Griffith and was intended to indicate that P. cuspidatum Roxburgh was represented in the Wallich Herbarium by no. 299, more cor- rectly 299-4 (as indicated by the page reference ‘64”), and not that Roxburgh’s species was based on a Wallich collection; this would be impossible, since the Wallich collection of this species was made in 1822, many years after Roxburgh’s death. As indicated in the synonymy, most authors have wrongly attributed the epithet to “Wallich ex Hook.,” but the first valid publication was by Mettenius, who must be cited as the author; Wallich’s name will disappear in this connection, since his orig- inal wrophyllum was a nomen nudum under Polypodium, and when it was first published by Mettenius it appeared under a different generic name, Phegopteris. 82. POLYPODIUM DICHOTOMUM sensu Roxburgh, Calcutta Journ. Nat. Hist. 4:493. 1844, non Houtt., 1783, non Thunb., 1784. =Dicranopteris line- aris var. montana Holttum, Reinwardtia 4:276. 1957. AUTHENTIC MATERIAL: A Roxburgh specimen in the Brussels Herbarium with the name in Roxburgh’s hand (Morton photograph 19939). There is a Roxburgh drawing in the British Museum that represents the same plant rather well; it is indicated as being from Amboina. According to Roxburgh, his plant came from the Molucca Islands. Holttum records var. montana from the Moluccas, but only from Ternate, where Roxburgh’s plant may have been col- lected by Christopher Smith; however, it could very well grow also in Amboina. The specimen cited above agrees well with a plant from the Admiralty Islands identified as var. montana by Holttum. Roxburgh did not consider his plant as new but referred it to P. dichotomum Thunb. Acrostichum furcatum L. is added as a synonym, but whether by Roxburgh or by Griffith is uncertain ; my guess is that Griffith added it and inserted it in the wrong place, intending to put it as a synonym of P. furcatum Roxburgh, assuming that Roxburgh’s species was the same as the Linnaean one from having the same specific epithet; however, he was 342 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM wrong about that, for the Linnaean species is a quite different plant of the West Indies. 83. POLYPODIUM DUBIUM Roxburgh, Calcutta Journ. Nat. Hist. 4:496. 1844. =Microlepia dubia (Roxburgh) Morton, comb. nov. Davallia polypodioides sensu D. Don. Prodr. FI. Nepal. 10. 1825, as to description not type. In Sledge’s treatment of Microlepia, D. polypodioi- des is listed and discussed as though it were a new species; however, Don clearly was making a new combination, based on his cited synonym Dicksonia polypodioides Swartz, and remarked that the reason for considering it a Davallia rather than a Dicksonia was because of the one-valved involucre. However, Don evidently misidentified D. polypodioides Swartz, for a specimen that is authentic for Don’s concept is in the British Museum (Nepal, Wallich, Morton photograph 6896), and this does not represent Dicksonia polypodioides Swartz, but rather the species later described as Polypodium dubium Roxburgh and Microlepia firma Mett. Davallia roxburghii Wallich, Num. List. 2218. 1830, nom. nud. Based on a Roxburgh specimen received as “Pol. dub. vel Cyathea.” Wallich evidently mistook the “Pol. dub.” to mean merely a dubious species of Polypodium, but as shown by his use of the name and his descrip- tion Roxburgh intended it as a specific epithet, his Polypodium dubium (Morton photograph 15743). Microlepia firma Mett. ex Kuhn, Linnaea 36:146. 1869. Syntypes: Mish- mee, Griffith (K, Morton photograph 20646) and Bhoton, Griffith (K, Morton photograph 20645). I designate the latter as lectotype. There is a Griffith collection from Mishmee in the British Museum which is presumably an isosyntype. LECTOTYPE: East India Company Herbarium no. 2218 (K, Morton photo- graph 15743), collected by Roxburgh. Roxburgh indicated his plant came from the Molueca Islands, where it was surely collected by Christopher Smith, There is a specimen at Kew in the Hooker Herbarium from Amboina, Webb, which matches the lectotype Wallich 2218 exactly. Webb was never in Amboina, and so the specimen was merely sent by Webb to Hooker; similar specimens from Amboina sent by Webb were collected by Christopher Smith. This indicates that the type of Polypodium dubium came from Amboina. There is in Geneva a specimen from Amboina collected by Christopher Smith that is unidentified except for the name “Trichomanes” written in, not by Roxburgh, which seems to match the lectotype chosen above and which is undoubtedly another isotype (Morton photograph 6551). In his most useful paper on Microlepia (Kew Bull. 1956:523-— 531), Sledge clearly delimited three species that were much confused in herbaria—M. speluncae, M. trapeziformis, and M. firma. He cited Davallia roxburghii Wall. as a synonym of M. firma, not adopting this name because it is a nomen nudum as published by Wallich and not realizing that Polypodium dubium Roxburgh was based on the same material. A variety of this species follows: ROXBURGH’S FERN TYPES—MORTON 343 84. Microlepia dubia (Roxburgh) Morton var. subglabra (Sledge) Morton, comb. nov. Microlepia firma var. subglabra Sledge, Kew Bull. 1956:527. Type: Ceylon, Thwaites C. P. 3272 (K, Morton photograph 20644, not annotated by Sledge). Microlepia firma var. hirta (Clarke) Sledge (Kew Bull. 1956: 527. 1956) was published as though Davallia polypodioides var. hirta was a new variety. It was not, however, but was based on the Hooker and Baker reference cited, on Davallia hirta Kaulf., which is from the Hawaiian Islands. Although it was not Sledge’s intention, his var. hirta applies to the plants of the Hawaiian Islands, which are different from those of Sikkim and Bhotan. 85. POLYPODIUM ELATUM Roxburgh, Calcutta Journ. Nat. Hist. 4:495. 1844. =Pleocnemia leuzeana (Gaud.) K. B. Presl, Tent. Pterid. 184, ¢. 7, f. 12. 1836 (as “leuceana’’). Polypodium leuzeanum Gaud. in Freyc. Voy. Bot. 361, t. 6. 1827. Type: Pisang, Molucca Islands, Gaudichaud. Aspidium leuzeanum (Gaud.) Kunze, Bot. Zeit. 4:474. 1846. Tectaria leuzeana (Gaud.) Copel. Phil. Journ. Sci. 2C:417. 1907. LectotyPE: A Roxburgh collection from Amboina in the Brussels Her- barium with the name in the hand of Roxburgh (Morton photographs 5189, 19806). A second sheet in Brussels is very likely a lower pinna from the same collection; it is marked as Polypodium elatum herb. Roxb. Amboyna, 1796. Roxb., but not in Roxburgh’s hand (Morton photographs 5190, 19805). A duplicate, very likely the apical part of the same frond, is in Geneva marked as “Amboyna, Dr. Roxburgh” (Morton photograph 16754). Another duplicate, surely a part of the same collection, is in the British Museum, collected in Amboina by Christopher Smith (Morton photograph 19529). In the “Index Filicum,” Polypodium elatum is left as a dubi- ous species. The lectotype was identified by Baker as Nephro- dium giganteum, a somewhat confused name. In recognizing Pleocnemia as a genus distinct from Tectaria, I am following Holttum (Reinwardtia 1:171-189. 1951), although with some reservations until the delimitation of Tectaria is better under- stood. 86. PoLYPODIUM EXCAVATUM Roxburgh, Hortus Bengalensis 75. 1814; Cal- cutta Journ. Nat. Hist. 4:485, ¢. 30 right (wrongly labeled “P. quercifolium”) 1844, non Bory ex Willd., 1810. —Polypodium scolo- pendria Burm. f. Fl. Ind. 232. 1768. Polypodium phymatodes L. Mant. 306. 1771. Type: This presents a unique problem. Some of Roxburgh’s species were listed in his “Hortus Bengalensis” (1814) in advance of their intended pub- lication in his “Flora Indica.” There was probably no intention of publishing new species here, but some of them have to be considered valid by reason of the citation references to Rheede’s “Hortus Indicus Malabaricus,” Rum- phius’ “Herbarium Amboinense,” or other pre-Linnaean publications. The 344 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM only fern published by reference to Rumphius is this one, for which Rumphius vol. 6, t. 35, f. 1 is cited. If this is taken at face value, Polypodium ex- cavatum Roxburgh would become a synonym of Stenosemia aurita (Swartz) K. B. Presl, which this plate of Rumphius’ represents. However, when Roxburgh’s full manuscript was published, this reference was corrected to t. 35, f. 2, which represents Polypodium scolopendria Burm. f. That this was Roxburgh’s real intent is shown by his description and published and un- published drawings, as well as by two authentic specimens in the Brussels Herbarium named excavatum in Roxburgh’s hand, for the description, the drawings, and the specimens all represent P. scolopendria, and are nothing like Stenosemia aurita. Therefore, common sense requires that we regard the original citation of “f. 1” as a typographical error for “f. 2.” This means that the type of Roxburgh’s species must be considered to be Rumphius’ description and illustration of his Polypodium indicum minus Herb. Amboin. 6:80, t. 35, f. 2. According to E. D. Merrill’s “An Interpretation of Rumphius’ Herbarium Amboinense,” this description and plate clearly apply to Poly- podium phymatodes L., now correctly known as P. scolopendria. Roxburgh’s description indicates that he had two forms in hand, one with a nearly simple blade and one with deeply pin- natifid blades. The two authentic Roxburgh specimens in Brus- sels named by Roxburgh illustrate these forms. One (Morton photograph 19926) has a simple blade just slightly lobed; the other (Morton photograph 19927) has a deeply pinnatifid blade typical of mature specimens of P. scolopendria. This kind of variation is well known in this species. According to Roxburgh’s “Hortus Bengalensis,” his material came from the Moluccas, collected by Christopher Smith; it was cultivated in the Cal- cutta Botanical Garden. In the “Index Filicum,” P. excavatum Roxburgh is referred to P. longissium Blume without question, but the basis of this opinion, which is an obvious error, is not known. Copies of Roxburgh’s original drawing are at Kew (no. 1747, Morton photograph 15868) and the British Museum (Morton photograph 15779). These drawings agree closely with the pub- lished drawing t. 30, right, which is erroneously labeled Poly- podium quercifolium, doubtless by the editor, Griffith. The true P. quercifoliwm was not illustrated by Roxburgh, so far as I know. 87. POLYPODIUM FELINUM Roxburgh, Calcutta Journ. Nat. Hist. 4:496. 1844. = Cyathea felinum (Roxburgh) Morton, comb. nov. Alsophila concinna Baker in Hook. & Bak. Syn. Fil., ed. 2, 439. 1874. Type: Louisiade Archipelago, MacGillivray. Alsophila polyphlebia Baker, Journ. Linn. Soc. Bot. 15:104. 1876. Type: Aru Island, Molucca Islands, Mosley. Alsophila sangirensis Christ in Warburg, Monsunia 1:90. 1900. Type: Sangihe Island, Molucca Islands, Warburg 16605, ROXBURGH’'S FERN TYPES—MORTON 345 Cyathea sangirensis (Christ) Copel. Phil. Journ. Sci. 40:37. 1909. LECTOTYPE: Amboina, Roxburgh (G, Morton photographs 6546, 16633). A duplicate closely agreeing and evidently a part of the same collection is in Brussels, with the name in Roxburgh’s hand (Morton photograph 19858). Specimens that also agree, doubtless isotypes, are in the J. E. Smith Herbarium, Nos. 1625-74 and 1625-75 (LINN, Morton photographs 20226, 20227); they were collected in Amboina by Christopher Smith in 1797. Roxburgh cited the locality as Moluccas and Pullo Pinang; I have not located any specimens from Penang, which would be different, since this species is not known in Penang. Cyathea sangirensis of Holttum’s treatment in the “Flora Malesiana” is a rather characteristic species because of its large size and strongly toothed segments. Nevertheless, it has been repeatedly renamed, for in addition to the synonyms quoted above, Holttum lists seven additional specific synonyms. Ap- parently, this is one of the common species of the Moluccas. I have seen authentic specimens or photographs of many of these synonyms, and it appears that Holttum is correct in plac- ing them all under a single species; the Roxburgh lectotype is quite in agreement with the other specimens. Roxburgh de- scribed his plant as being “scandent,” which would be unusual in this species, but not impossible. At least a few species of Cyathea are said to be scandent, e.g., C. biformis (Rosenst.) Copel., which Holttum describes as “Stem 1-114 cm. thick, climbing (clinging to supporting tree by its roots,” and C. scandens (Brause) Domin is said to have a similar habit. However, Rox- burgh may have been wrong about this; he did not know the plant in the field, since he never collected in Amboina, and probably not in Penang. 88. POLYPODIUM FERRUGINEUM Roxburgh, Calcutta Journ. Nat. Hist. 4:487. 1844, non Mart. & Gal., 1842. =?Nephrolepis hirsutula (Forst.) K. B. Presl, Tent. Pterid. 79. 1836. Type from the South Sea Islands, Forster (holotype or isotype BM, Morton photograph 6876); since this type specimen agrees altogether with plants from Tahiti, and since many of Forster’s collections are known to have come from Tahiti, this may be presumed the type locality. LECTOTYPE: A Roxburgh specimen in the Brussels Herbarium with the name in Roxburgh’s hand. According to Roxburgh his species came from Amboina. There is an isotype from the Roxburgh Herbarium in Geneva, collected in Amboina by Christopher Smith (Morton photograph 6571). An- other specimen that may be an isotype, although not named P. ferrugineum, is in the J. E. Smith Herbarium, no. 1622-104 in the Linnean Society (Morton photograph 20357). I do not have the lectotype in hand or a photograph of it, but I did see it at Kew while it was on loan to Dr. Jarrett. 346 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM According to my notes made at the time, the pinnae are strongly rufescent-pilose, with very shallow lobes, the upper base rounded and not auriculate, the lower base very broad and rounded also; the midrib is scaly above. It has been identified as Nephro- lepis rufescens (Schrad.) Wawra, which is considered in the “Index Filicum” as a synonym of N. hirsutula. Although most, but not all, specimens of N. hirsutula have the pinnae auriculate at the upper base, at least one collection from Amboina (C. B. Robinson 1957, US), identified as N. hirsutula by Merrill, agrees with the Roxburgh type. 89. POLYPODIUM FLAGELLIFERUM Roxburgh, Calcutta Journ. Nat. Hist. 4:487, t. XXXI (right hand). 1844. =Nephrolepis biserrata (Swartz) Schott, Gen. Fil. ad t. 3. 1834. LECTOTYPE: A specimen in the Brussels Herbarium with the name in the hand of Roxburgh (Morton photograph 19930a). According to Roxburgh the species is a native of Bengal, India. This type specimen was determined as N. acuta (Schkuhr) K. B. Presl by Baker, which is considered a synonym of N. biserrata. A second specimen may be in the Brussels Herbarium, also with the name in Roxburgh’s hand, for my notes indicate that one was on loan to Dr. Jarrett at Kew in 1969, It is the same as the one selected as lectotype. A painting at Kew (no. 1748) (Morton photograph 15869) agrees well enough, especially in the nearly medial sori, a characteristic of N. biserrata. In the “Index Filicum,” P. flagelliferum is referred to N. exaltata, but this is clearly wrong from the position of the sori. The indusia are wrongly drawn in the painting as reniform: they are actually suborbicular with a narrow, closed sinus, as they should be in N. biserrata. 90. POLYPODIUM FURCATUM Roxburgh, Calcutta Journ. Nat. Hist. 4:493. 1844, non Swartz, 1802, non Desv., 1827. —Gleichenia truncata (Willd.) Spreng. in L. Syst. Nat. ed. 16, 4:25. 1827. Mertensia truncata Willd. Kongl. Vet. Akad. Nya Handl. 25:169, t. V, f. A. 1804. Type: B-Hb. Willd. no. 19470, received from Jussieu; an isotype ought to be in the Jussieu Herbarium in Paris. Mertensia laevigata Willd. in L. Sp. Pl. ed. 4, 5:75. 1810. Gleichenia laevigata (Willd.) Hook. Sp. Fil. 1:10, 1844. Sticherus truncatus (Willd.) Nakai, Bull. Nat. Sci. Mus. Tokyo 29:20. 1950. LECTOTYPE: A Roxburgh specimen in the Brussels Herbarium with the name in Roxburgh’s hand on a ticket overlying a stem (Morton photograph 4743). An isotype, also with a label in Roxburgh’s hand, is also in Brussels and is essentially identical (Morton photograph 4744). According to Rox- burgh, the species was collected in Pullo Pinang, i.e., Penang Island, Malaya, where it was doubtless found by W. Roxburgh, Jr., or W. Hunter, ROXBURGH’S FERN TYPES—-MORTON 347 In the “Index Filicum,” P. furcatum Roxburgh is referred with a query to Gleichenia glauca (Thunb.) Hook., which is by no means the same and which does not even belong in the same subgenus. In Holttum’s treatment of Gleichenia in the “Flora Malesiana,” P. furcatuwm is omitted as a synonym. 91. POLYPODIUM GLABRUM Roxburgh, Calcutta Journ. Nat. Hist. 4:482. 1844, non Burm. f., 1768. =Polypodium polycarpon Cav. Descr. 1:246. 1802. Type: Née, MA (not seen). Cavanilles ascribed the type to San Antonio, Ecuador, but this species does not grow in South America; many of Née’s localities were mixed, and his plant doubtless came from either the Marianna Islands or the Philippine Islands, where Née did collect and where this species grows. Acrostichum punctatum L. Sp. Pl. ed. 2, 1524. 1763. Polypodium punctatum (L.) Swartz, Journ. Bot. Schrad. 1800(1) :21. 1802, non Thunb., 1784. Polypodium polycarpon Swartz, Journ. Bot. Schrad. 1800(1) :21. 1802, non Cav., 1802. In the “Index Filicum,” this is cited erroneously as “Cav.: Sw.,” but Swartz did not mention Cavanilles in publishing this species, he chose the name independently. No specimens were cited, and so the species must be typified from Swartz’ “Species Fillecum,” to which the account in Schrader’s Journal was a preliminary. The “Synopsis,” p. 227, cites Mauritius, Groendal and Java, Thunberg. A reference is added here to Polypodium polycarpon Cav. and no “Amer. merid.,” but only with a query, showing definitely that Swartz had not based his species on that of Cavanilles. The Swartz name was cited as 1801 and was thought to be earlier, but Stafleu has shown that it was actually published in 1802, and so we can assume for convenience that the Cavanilles name is earlier, for it also was formerly cited as of 1801. The Swartz species must be lectotypified by either the Groendal or Thunberg specimens, and very likely the Groendal will be chosen, for it is probably in the Swartz Herbarium in Stockholm, and the Thunberg specimen is probably only in the Thunberg Her- barium. It has been generally assumed that the species is taxonomically the same as P. polycarpon Cav., but since they are from different geo- graphic areas, they may not be identical. Polypodium irioides Poir. in Lam. Enc. Méth. 5:513. 1804. Type: Ie de France [Mauritius], probably collected by Commerson (P-Herb. Lam., Morton photograph 2696). Polypodium lingulatum Swartz, Syn. Fil. 30. 1806. A new name for P. punctatum (L.) Swartz, non Thunb. Polypodium glabrum Roxburgh ex Wallich, Num. List. no. 281. 1829, nom. nud. Based on Wallich in 1826 from Amherst and Martaban, no. 281-1; Herb. Roxburgh, no. 281-2; and Herb. Wright, no. 281-3. The specimen from the Roxburgh Herbarium cited by Wallich is not in the East India Company Herbarium at present, and the reference was doubt- less to the specimen then in the Roxburgh Herbarium in the India House, which is the same sheet that is in Brussels and cited as lectotype. 348 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM LECTOTYPE: A Roxburgh specimen in the Brussels Herbarium with the name in the hand of Roxburgh (Morton photograph 19930). According to Roxburgh it was “found near Calcutta on the trunks of large old trees, etc. where there is much shade and humidity.” The “ete.” locality mentioned refers at least in part to a collection from Silhet, by M. R. Smith (ex Herb. Roxburgh, BM, Morton photograph 20887). Roxburgh drawings of P. glabrum are at Kew, no. 1006 (Morton photograph 15870) and the British Museum (Morton photograph 15780); they are stylized, but probably do represent P. polycarpon. In most herbaria and books this species has continued to be called Polypodium punctatum (L.) Swartz, even though Swartz recognized that this name was an illegitimate later homonym and renamed it P. lingulatum. An older name, however, appears to be P. polycarpon Cav. (but see the note above on P. polycarpon Swartz, which might have priority and which seemingly refers to the same species). Under Microsorium, the epithet punctatwm is correct. 92. POLYPODIUM IMPUBER Roxburgh, Calcutta Journ. Nat. Hist. 4:494. 1844, = Cyathea alternans (Hook.) K. B. Presl, Abhandl. K. Bohm. Gesell. Wiss. V, 5:347. 1848. Polypodium alternans Wallich, Num. List. no. 329. 1829, nom. nud. Hemitelia alternans Hook. Icon. Pl. 7:t. 622. 1844. Syntypes: Penang, Wallich, Dalhousie. In the “Flora Malesiana,” Holttum indicates that the “type” is Wallich 329, but this was only one of two original syn- types; Holttum’s designation of Wallich List. no. 329 as type must be considered as a lectotype, the natural choice but still a lectotype rather than a type. LECTOTYPE: Amboina, Rorburgh (G, Morton photographs 6542, 16636), the apical part of a frond. The basal part of probably the same frond is in Brussels, with the name P. impuber in Roxburgh’s hand (Morton photograph 19851). Although I usually choose the Brussels specimens as lectotypes, since they have the names in Roxburgh’s own hand, I do not in this instance, since the Brussels specimen, consisting of basal pinnae only, is sterile, and the Geneva specimen is fertile. Roxburgh gave the locality as Amboina. Cyathea alternans is a rather characteristic species by reason of its small, entire, glabrous, rounded pinnules, most of which, except the basal, are adnate. Holttum (Fl. Males. II, 1(2):145. 1965) suggests that it has the appearance of being a hybrid between C. moluccana R. Brown and either C. squamulata (Blume) Copel. or C. ridleyi (Baker) Copel., but this is not quite com- prehensible since he considers squamiulata and ridleyi to be synonyms. One of these suggested parents does grow in Am- boina, C. moluccana, but C. squamulata is not reported from there, although it may well occur. Incidentally, Holttum cites the parenthetical authority for C. alternans as ‘Wall. ex Hook.,” ROXBURGH’S FERN TYPES—MORTON 349 but this may not be, for Wallich used the epithet alternans under Polypodiwm (as a nomen nudum), but the first valid pub- lication was under the generic name Hemitelia by Hooker, and therefore Wallich’s name disappears as an authority. The syn- onym P. impuber was overlooked by Holttum, although it was cited as a synonym of C. alternans in the “Index Filicum,” Suppl. 1:124. 19138. 93. POLYPODIUM INVOLUCRATUM Roxburgh, Calcutta Journ. Nat. Hist. 4:491. 1844. —Tectaria crenata Cav. Descr. 250. 1802. Type: Marianna Islands, Née (presumably MA). Aspidium pachyphyllum Kunze, Bot. Zeit. 6:259. 1848. Lectotype: Java, Zollinger 580z (the fertile specimen, L, Morton photograph 2296). Since Kunze’s herbarium is destroyed, it is necessary to designate lectotypes for his species; the present specimen is suitable, since it is marked “Aspidium pachyphyllum Kze n. sp.” in Kunze’s own hand; this may indeed be a holotype and there may never have been a col- lection in Kunze’s own herbarium. There is a second sheet in Leiden of the same number also with the name in Kunze’s hand, and it is undoubtedly a sterile part of the lectotype chosen above (Morton photograph 2294). LECTOTYPE: A specimen of the Roxburgh herbarium in the Brussels Herbarium with the name in Roxburgh’s hand; this specimen has a ticket in the hand of Christopher Smith reading “Amboyna, no, 333, 96,” - the “96” referring to the date 1796 (Morton photograph 5109, 19788). A duplicate of this collection is in Geneva with the data “Amboina, Dr. Rox- burgh” (Morton photograph 16746). There is a specimen in the British Museum of this species collected in Ternate by Christopher Smith (Morton photograph 19528), and this may be a part of the type collection and the correct locality, since Roxburgh may have wrongly assumed that all of the Smith collections came from Amboina, as indeed most of them did, but per- haps this one did not. In the “Index Filicum,” P. involucratum is listed as a dubious species of Dryopteris. The brief original description is entirely consistent with the type cited above and not with any Dryopteris known from the Moluccea Islands. 94, POLYPODIUM LONGIFOLIUM Roxburgh, Caleutta Journ. Nat. Hist. 4:492. 1844, non Cav., 1802, non Presl, 1822. =Thelypteris heterocarpa (Blume) Morton, Amer. Fern Journ. 49:1138. 1959. Aspidium heterocarpon Blume, Enum. Pl. Jav. 2:155, 1828. Syntypes: Two Blume collections in Leiden; the one of these from Java with the name in Blume’s hand I designate as lectotype (Morton photo- graph 2556), since it is a fine specimen with two complete plants with rootstocks. The other specimen from Boerengrang, Java, also with the name in Blume’s hand (Morton photograph 1152) is a larger plant with a much longer stipe and fewer reduced basal pinnae; it lacks a rootstock. 350 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM Dryopteris heterocarpa (Blume) Kuntze, Rev. Gen. Pl. 2:813. 1891. Cyclosorus heterocarpus (Blume) Ching, Bull. Fan Mem. Inst. Biol. Bot. 8:180. 1938. LECTOTYPE: Two Roxburgh collections in the Brussels Herbarium, both evidently a part of the same frond, one with the name in the hand of Roxburgh (Morton photographs 19683, 19684). The locality is given as Amboina by Roxburgh. An isotype in Geneva bears the data “Amboyna, Dr. Roxburgh” and a label “Typus” (Morton photographs 6541, 16660). I do not know that T. heterocarpa has been previously re- ported from Amboina, but there is no reason why it should not be there, for the species ranges as far east as New Guinea and the Bismarck Archipelago. It is rather easily recognizable among its numerous relatives by the large, glabrous, very persistent, vaulted indusia. 95, POLYPODIUM LUCIDUM Roxburgh, Calcutta Journ. Nat. Hist. 4:486. 1844. =Polypodium cuspidatum D. Don, Prodr. Fl. Nepal 2. Feb. 1, 1825. Neotype: Nepal, Wallich in 1818 (BM, Morton photograph 20885). This might be the holotype, but does not bear the name in Don’s hand. Polypodium leiorhizum Wallich, Num. List. no. 303. 1829, nom. nud. Based on three collections: Nepal, Wallich in 1820, Kumaon, R. Blinkworth, and Sylhet Mountains, DeSilva. Phymatodes leiorhiza K. B. Presl, Tent. Pterid. 196. 1836, nom. nud. Drynaria leiorhiza J. Smith, Journ. Bot. Hook. 4:61. 1841, nom. nud. Polypodium leiorhizum Wallich ex Mett. Fil. Hort. Bot. Lips. 39, t. 25, f. 7, May-June, 1856. Not formally described, but with sufficient de- scription of sorus and venation plus the figure to validate the name. Fully described by Mettenius, Abhandl. Senckenb. Naturf. Gesell. 2:103. Oct., 1856. Also published in Hook. Fil. Exot. t. 24. 1857. Type: Nepal, Wallich (presumably B). Phymatodes cuspidata (D. Don) J. Smith, Cat. Cult. Ferns 10. 1857. Pleopeltis leiorhizon (Wallich ex Mett.) Moore, Ind. Fil. 346. 1862. Phymatodes lucida (Roxburgh) Ching, Contr. Inst. Bot. Nat. Acad. Peiping 2:61. 1933. The type is wrongly stated to be Nepal, Wallich. Microsorium lucidum (Roxburgh) Copel. Gen. Fil. 196. 1947. Microsorium cuspidatum (D. Don) Tagawa, Fl. Eastern Himalaya 495. 1966. TYPE: No herbarium specimens from Roxburgh of this species have been located. Therefore, a Roxburgh drawing at Kew, no. 1922 (Morton photo- graph 15871), is designated lectotype temporarily until a specimen can be located. The plant, according to Roxburgh, is a “Native of Nepaul, from thence introduced into the Botanic Garden [at Calcutta] by Dr. Buchanan in 1802.” Roxburgh’s description, which is rather full, was probably drawn up from living plants and no herbarium specimens may have been made. I have given the synonymy in full, since this species is cur- rently being filed under three specific epithets—lucidum, leio- ROXBURGH’S FERN TYPES—-MORTON 351 rhizum, and cuspidatum. Polypodium cuspidatum D. Don is clearly the earliest of these, but it was rejected in the ‘Index Filicum” in the belief that Polypodium cuspidatum K. B. Presl, Rel. Haenk. 1:20, t. 1, f. 3. 1825, had priority. However, from Stafleu’s recent “Taxonomic Literature” it appears that Don’s work was published Feb. 1, 1825, and Presl’s between June and November 1825. This is another instance showing the impor- tance of careful dating. 96. POLYPODIUM MUCRONATUM Roxburgh, Calcutta Journ. Nat. Hist. 4:490. 1844, non Swartz, 1806. =Polystichum sp. TYPE: No authentic specimens seen. The species was described from plants collected in Sylhet (East Pakistan) by M. R. Smith in 1811 (Hort. Bengal. 75. 1814) and introduced in the Calcutta Botanical Garden in 1811 according to Voigt (Hort. Suburb. Calcut. 734. 1845). The description is of a plant with simply pinnate fronds, with ensiform, acutely serrate, acuminate pinnae; the texture is hard and glossy, and the serrations very acute, i.e, mucronate as in the specific epithet chosen. This description indicates that the species is a Polystichum. It is possible that no herbarium spec- cimens were ever made. By knowing which species of Polystichum grow in Sylhet, one could guess which species is meant. The matter is not of importance since Roxburgh’s name is an il- legitimate later homonym. 97. POLYPODIUM MULTIFLORUM Roxburgh, Calcutta Journ. Nat. Hist. 4:493. 1844, non Roth, 1797. =Arcypteris irregularis (K. B. Presl) Holttum, Reinwardtia 1:193. 1951. LeEcToTYPE: A Roxburgh specimen in the Brussels Herbarium with the name in Roxburgh’s hand and the identification Aspidium difforme Blume by Baker (Morton photographs 5184, 19804). Two other specimens, surely isotypes, are also in Brussels and likely represent different parts of the same frond, a base and an apex (Morton photographs 5185, 5186, 19802). The species was collected in Amboina, according to Roxburgh, where it was doubtless obtained by Christopher Smith. Two isotypes are in Geneva, one labeled “Ind. orient., Dr. Roxburgh” (Morton photograph 6565) and one “Amboina, Dr. Roxburgh” (Morton photograph 16755). For synonymy and comments on the genus Arcypteris see under Polypodium confertum. Roxburgh’s descriptions of his P. confertum and P. multiflorum are almost identical; he may have been influenced by the geography—one came from India and one from the Molucca Islands—but this species has a broad range and occurs in both places in apparently indistinguishable forms. In the “Index Filicum,” P. multiflorwm Roxburgh is left as a dubious species. 352 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM 98. POLYPODIUM NUDATUM Roxburgh, Calcutta Journ. Nat. Hist. 4:491. 1844. =Thelypteris nudata (Roxburgh) Morton, comb. nov. Polypodium multilineatum Wall. ex. Hook. Sp. Fil. 5:11. 1863. Type: Wallich (K not seen). Nephrodium moulmeinense Bedd. Ferns Brit. Ind. Suppl. 18. 1876. Type: Sylhet, Wallich 147 (K not seen). Thelypteris multilineata (Wall. ex Hook.) Morton, Amer. Fern Journ. 49:113. 1959. Pronephrium nudatum (Roxburgh) Holttum, Blumea 29:111. 1972. TYPE: Two Roxburgh specimens in the Brussels Herbarium, evidently a part of the same frond, the apical part with the name in Roxburgh’s hand and the number 2395 (Morton photographs 19688, 19689). Since this is the only specimen found in any herbarium, it is presumably a holotype. Accord- ing to Roxburgh, it is from Amboina, but this is surely an error. There is in the British Museum a specimen of the Roxburgh Herbarium—not named nudatum, however—that is multilineata and which agrees with the Roxburgh type of nudatwm. It was collected in Sylhet by M. R. Smith (Morton photo- graph 20854). Presumably, Roxburgh had his material (doubtless unmounted at the time) in a cover with the name “Smith” on it, and since he had so many specimens from Amboina collected by Christopher Smith, he errone- ously assumed that this one also was from Christopher Smith, rather than M. R. Smith. Therefore, the species is from Sylhet, and not from Amboina. 99. POLYPODIUM PARASITICUM sensu Roxburgh, Calcutta Journ. Nat. Hist. 4:507. 1844. =Thelypteris dentata (Forsk.) E. St. John, Amer. Fern Journ. 26:44. 1936, sensu lato. Polypodium dentatum Forst. Fl. Aegypt.-Arab. 185. 1775. AUTHENTIC MATERIAL: Amboina, Christopher Smith 310, collected in 1796, with the name in Roxburgh’s hand (BR, Morton photograph 19783). An ap- parent duplicate of this and probably the basal part of the same collection is also in Brussels,’ identified by Baker as Nephrodium molle var. glabratum (Morton photograph 19702). Roxburgh stated: “Nat. of various parts of India. The above definition is taken from Molucea plants.” Roxburgh did not attribute this species to Linnaeus, and Linnaeus was not mentioned. Griffith added “Aspidiam parasi- ticum, Hb. Madras, Wall. Cat. 67, No. 22392,” but this is not equivalent to citing Polypodium parasiticum L. as a synonym, for it is only a reference to a particular sheet mentioned in the Wallich “Numerical List” on page 67, and this only with a query, which may not have anything to do with the Linnaean species and is surely not a type. Nevertheless, it seems reasonable that Roxburgh really did consider his plant the same as that of Linnaeus. The specific name could hardly have been chosen * It was the custom in the Roxburgh Herbarium not to transfer data to the second sheets, but to pin the sheets together. When they are not fastened, it is not usually difficult to match them, as quite often the second sheet is clearly the base or apex of the frond on the first sheet. ROXBURGH’S FERN TYPES—MORTON 353 independently by Roxburgh, for it is peculiar in that it is so inappropriate; “parasitic” in the old days usually meant epi- phytic, but this species is always terrestrial. It is evident that Roxburgh had several collections in hand that he called P. parasiticum, which probably came from dif- ferent places since they are not identical with each other. A collection named parasiticum by Roxburgh in Brussels (Morton photograph 19701) is probably correctly named, but it is a young frond with only immature indusia and sporangia; a prob- able duplicate is also in Brussels (Morton photograph 19700), determined as Nephrodium molle and with the indication “Ind. or. Hb. Roxb.” These two specimens might be T. contigua (Rosenst.) Reed, according to Holttum. They do have smaller and narrower pinnae. In that case, they are doubtless from Penang. Another Roxburgh collection in Brussels has a label in Roxburgh’s hand reading “Seems to differ too little from P. parasiticum to form a distinct species”; Roxburgh was prob- ably right, for the specimen seems to be only a rather large, glabrate form; there is an apparent duplicate of this, without data, also in Brussels (Morton photograph 19674). These Holt- tum has identified as Cyclosorus vestigiatus (Copel.) Copel., with sessile, spherical glands. Another collection, perhaps slightly different, is represented by two sheets in Brussels, both identified as Nephrodium molle, with the indication “Ind. or., Roxburgh” (Morton photographs 19698, 19699); a similar speci- men is in the British Museum, indicated as collected in the Moluccas by Christopher Smith (Morton photograph 20971). Some specimens that Roxburgh identified as Polypodiwm sopho- roides appear to be rather 7. dentata with only the basal veinlets fertile; these are in Geneva, from the Molucca Islands (Morton photographs 6561, 16661); a duplicate is in the British Museum, from Gilolo Island, Christopher Smith, November, 1801 (Morton photograph 20870). See my comment under Polypodium sopho- roides sensu Roxburgh. 100. PoLYPODIUM PERTUSUM Roxburgh ex Hook. Exot. Fl]. 2:t. 162. 1825; Rox- burgh, Calcutta Journ. Nat. Hist. 4:483, t. 29 (left). 1844. =Pyrrosia lanceolata (L.) Farwell, Amer. Midl. Nat. 12:245. 1931. Acrostichum lanceolatum L. Sp. Pl. 1067. 1753. Lectotype: Chosen in- ferentially by Trimen (Journ. Linn. Soc. Bot. 24:152. 1886) as Ceylon, Hermann 380 (BM). Trimen considered this species erroneously to be the same as Polypodium adnascens Swartz, a somewhat similar species which, however, does not occur in Ceylon. According to a note by Alston quoted by Ching (Bull. Chin. Bot. Soc. 1:46, 1935) the lecto- type of A. lanceolatum in the Hermann Herbarium matches Polypodium 354 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM spissum Bory ex Willd. better than Indian specimens that Ching referred to P. adnascens. Niphobolus pertusus (Roxb. ex Hook.) Spreng. Syst. Veg. ed. 16., 4:44. 1827. Cyclophorus lanceolatus (L.) Alston, Journ. Bot. Brit. & For. 69:102. 1931 (excl. syn). TYPE: Hooker described P. pertusum Roxburgh from a Roxburgh specimen from the Delta of the Ganges transmitted to him by Wallich and from a plant cultivated in the Liverpool Botantical Garden from material received from the Royal Horticultural Society. The Roxburgh specimen that Hooker had in hand when he adopted Roxburgh’s name was not found at the British Museum (Natural History) or in the Hooker Herbarium at Kew. A duplicate is in Brussels with the name in the hand of Roxburgh (Morton photograph 19936). Another duplicate is in the East India Company Herbarium, no. 267-5 (K, upper right, the other specimens on the sheet are Herb. Wight; Morton photograph 15735). An authentic Roxburgh draw- ing is at Kew, no. 1745 (Morton photograph 15872). The other specimen cited by Hooker, the one cultivated in Liverpool, was said to have come from China, and according to Ching represents Pyrrosia adnascens; I have not seen it and probably neither did Ching. Ching (Bull. Chin. Bot. Soc. 1:71. 1935) indicates that P. lanceolata is a common fern in “the whole of tropical Asia,” and also cites Polynesia, but not the Mascarene Islands. He equates it with Polypodium spissuwm Bory ex Willd. and states that the type of that species came from “Insula Borbourn (or Malabar),” but this represents a misunderstanding. The type came from Bourbon, which is the present island of Réunion in the Mascarene Islands and not from Malabar in India. Whether P. spissum is really the same as the true P. lanceolata from Ceylon can be determined only from an examination of the type in the Willdenow Herbarium in Berlin, where it is mounted on two sheets under the number 19627. 101. POLYPODIUM PHYLLITIDIS Roxburgh, Calcutta Journ. Nat. Hist. 4:483. 1844, non L., 1753. =Polypodium phyllomanes Christ, Bull. Acad. Géogr. Bot. Le Mans 1902:210, figs. 1902. Neocheiropteris phyllomanes (Christ) Ching, Bull. Fan Mem. Inst. Biol. 4:110. 1933. Neolepisorus phyllomanes (Christ) Ching, Bull. Fan Mem. Inst. Biol. 10:14. 1940 (reference not seen). LECTOTYPE: A Roxburgh specimen in the Brussels Herbarium with a label reading “Polypodium an phyllitidis, Feb. 2, Journ. p. 185” not in Roxburgh’s hand, and a later annotation label “Nephrodium (Sagenia) singaporianum Baker,” the right-hand specimen (Morton photograph 19810). According to Roxburgh his species was collected near Chittagong [East Pakistan] by Buchanan-Hamilton (Hort. Bengal. 75. 1814). In the “Index Filicum,” P. phyllitidis Roxburgh (non L.) is referred to Aspidium singaporianum Wallich ex Hook. & Grev. ROXBURGH'S FERN TYPES—MORTON 355 with a query. This is based on the entry added to the original description by the editor, Griffith, reading: “Aspidiwm Singa- porianum Wall. Cat. P. 64, No. 3874?” Griffith may have been partly right, for the left-hand specimen on the sheet designated above as lectotype is actually Tectaria singaporiana (Wallich ex Hook. & Grev.) Copel. However, that this specimen was not Roxburgh’s type is shown by the fact that T. singaporiana— a very distinct species of Tectaria by reason of its simple blade resembling that of a Polypodiwm—is known only from the Malay Peninsula (Thailand, Malaya, and Singapore) and from not near Chittagong where Roxburgh’s plant came from. The fact that Roxburgh has a real Polypodium and not a Tectaria is also indicated by his statement “involucres obscure,” which would indicate that he saw no indusia; the indusia in Tectaria singa- poriana are conspicuous and persistent. The right-hand plant on the sheet mentioned above is a Polypodium and it agrees with Roxburgh’s description, so far as that goes, which is not very far. It appears to me from my photograph that this specimen can be referred to P. phyllomanes Christ as a form with a cuneate base, but it deserves further study. The venation of the Tectaria and the Polypodium are subtly different. 102. POLYPODIUM PHYMATODES sensu Roxburgh, Calcutta Journ. Nat. Hist. 4:484, 1844, non L., 1771. =Polypodium alternifolium Willd. 7x L. Sp. Pl. ed. 4, 5:168. 1810. AUTHENTIC MATERIAL: A Roxburgh specimen in Brussels identified as Polypodium phymatodes L. by Baker but without a name by Roxburgh (Morton photograph 19929). Roxburgh stated that his P. phymatodes was native in various parts of India. There is no proof that this specimen actually represents Roxburgh’s concept of P. phymatodes, but the specimen agrees with Roxburgh’s brief description and is likely the plant intended; it is one of the Indian plants generally identified as P. phymatodes. In the “Index Filicum,” Polypodium alternifolium Willd. is listed as a dubious species, said probably to equal P. longissimum Blume or P. nigrescens Blume. Since P. alternifolium has priority, it would displace one of these well-known names, but an exami- nation of photographs of the type in the Willdenow Herbarium kindly supplied by Dr. D. E. Meyer shows that this species is neither of these species but is a close ally or possible synonym of P. scolopendria Burm. f. (P. phymatodes L.). The type is mounted on three sheets under the number 19637; it came from “India orientalis,” and was collected by Klein. It differs from P. scolopendria in having narrower segments separated by broader sinuses. I have recently stated that P. alternifoliwm is the same as the recently described species Phymatodes baner- 356 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM jiana Pal & Pal, but I am not quite sure; that has the sori generally in two rows on each side of the midribs and P. alterni- folium seems to have but one row; further investigation is in- dicated. Pal and Pal point out a number of minute differences in their Phymatodes banerjiana that cannot be seen in the photo- graph of Polypodiwm alternifolium. 103. POLYPODIUM PILOSUM Roxburgh, Calcutta Journ. Nat. Hist. 4:492. 1844, non Schkuhr, 1806. =Tectaria fuscipes (Wallich ex Bedd.) C. Chr. Contr. U. S. Nat. Herb. 26:290. 19381. Aspidium fuscipes Wallich, Num. List. no. 361. 1829, nom. nud. Based on a plant from Chappendang Mountain, Tenasserim, India, Wallich in 1827. Aspidium fuscipes Wallich ex Bedd. Ferns Brit. Ind, Suppl. 15, t. 366. 1876. Syntypes: “Himalayas, Birma... The plant figured was collected by Mr. Clarke at Cachar, but I have it from various parts of North India, the Birma plant is of softer texture more pubescent and greener in color when dried, and the anastomosis not so copious.” Although Beddome attributed the name to Wallich, it is not clear that he actually drew his description from the Wallich specimen that was the basis of Wallich’s original unpublished name; on the other hand, Beddome’s illustration was specifically based on a Clarke collection from Cachar, and this I designate as lectotype; it is presumably in either Kew or the British Museum. Ctenitopsis fuscipes (Wallich ex Bedd.) C. Chr. ex Tardieu & C. Chr. Notul. Syst. 7:87. Oct. 1938. The same combination was later pub- lished by Ching (Bull. Fan Mem. Inst. Biol. Bot. 8:813. Nov. 1938). LECTOTYPE: A Roxburgh specimen in the Brussels Herbarium with the name in Roxburgh’s hand (Morton photograph 5196, 19798). There is an isotype agreeing in all ways in Geneva, marked as “Chittagong, Dr. Rox- burgh,” but not in Roxburgh’s hand (Morton photograph 16749). According to Roxburgh, his plant came from “Chittagong, near the Burning Wells,” now in East Pakistan. Roxburgh’s plant is a small one for this species and shows only a few of the dark scales on the stipe that are characteristic of this species. However, there is no doubt as to its identity. This is one of the species poorly described by Roxburgh, and in the absence of the authentic material cited above, it would be hard to identify. 104. POLYPODIUM PROLIFERUM (Retz.) Roxburgh ex Wallich, Num. List. 312. 1829, non Kaulf., 1824. =Thelypteris prolifera (Retz.) Reed, Phytologia 17:306. 1968. Hemionitis prolifera Retz. Obs. Bot. 6:38. 1791. Dryopteris prolifera (Retz.) C. Chr. Ind. Fil. 286. 1905. Ampelopteris prolifera (Retz.) Copel. Gen. Fil. 144. 1947. ROXBURGH'S FERN TYPES—MORTON 357 This is an instance of a name validly published in Wallich’s “Numerical List,” since a basionym is cited, a reference to Hemi- onitis prolifera Retz. The name was attributed to Roxburgh, but this was very likely a distortion of Roxburgh’s intent, for when Roxburgh published the name Polypodium proliferum, he attrib- uted the species only to himself and did not mention Retzius. The epithet “proliferum” is apt and evidently occurred independently to Retzius and Roxburgh. 105. PoLYPODIUM PROLIFERUM Roxburgh, Calcutta Journ. Nat. Hist. 4:489, t. 32, left. 1844, non Kaulf., 1824, non Roxburgh ex Wallich, 1829. =Thelypteris prolifera (Retz.) Reed, Phytologia 17:306. 1968. Type: No herbarium specimens of this species have been located; there- fore, the published plate is designated as lectotype temporarily until a specimen is found. A copy of the original plate is in Kew (no. 1007, Morton photograph 15873) identified as “Polypodium radicans,” probably the name first chosen by Roxburgh and probably changed to proliferum because of the earlier Polypodium radicans Burmann f., which Roxburgh knew and treated on page 488 of the published paper. According to Roxburgh, his species is a “native of Bengal, and the more interior parts of India. Grows among brushwood, long grass, etc. in moist shady places about Calcutta; fructifies during the latter part of the rainy season.” According to Voigt (Hort. Suburb. Caleut. 734. 1845), the species was native in Serampore (Bengal), the Khasia Mountains, Oude, and Nepal. It is evident that Roxburgh knew this species well in the field, for he gives a rather detailed description. He indicates that the fronds may reach up to 12 feet long, which is longer than I have seen reported elsewhere, but which may indeed be possible. As mentioned under P. proliferum Roxburgh ex Wallich, Rox- burgh intended his P. proliferum as a new species and not as a transfer of Hemionitis prolifera, as indicated by Wallich. The two species are taxonomically the same but have different types. 106. POLYPODIUM QUERCIFOLIUM sensu Roxburgh, Calcutta Journ. Nat. Hist. 4:484 (excl. t. 30, left). 1844. =Drynaria quercifolia (L.) J. Smith, Journ. Bot. Hook. 3:398. 1841. AUTHENTIC MATERIAL: A Roxburgh specimen in the Brussels Herbarium with the name in the hand of Roxburgh (Morton photograph 19911). Rox- burgh drawings at Kew (no. 1100 and Morton photograph 15874) and BM (Morton photograph 15767). According to Roxburgh, the species grows in various parts of India. Roxburgh did not mention Polypodium quercifolium L., but he did cite a Rheede illustration that was one of the original citations of Linnaeus. And Roxburgh did not put the “R” after the species name, which he generally did when he was describing a new species. Therefore, it seems that Roxburgh really intended 358 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM the Linnaean species and was not considering his plant as new. The published illustration labeled P. quercifolium has been wrongly labeled, presumably accidentally by Griffith; it repre- sents P. excavatum Roxburgh and not P. quercifolium. 107. POLYPODIUM RADICANS sensu Roxburgh, Calcutta Journ. Nat. Hist. 4:488. 1844, =Nephrolepis radicans (Burm. f.) Kuhn, Ann. Lugd. Bat. 4:285, 1869, AUTHENTIC MATERIAL: “Ind. or.,” Roxburgh (BR, Morton photograph 19917a); Malay Islands, Rorburgh (G, Morton photograph 6547). There is also a second specimen in Brussels with the name in Roxburgh’s hand and the number 2390 (not photographed). Roxburgh wrote “P. radicans. Burm. Fl. In. 233, t. 66, f. 3. is not unlike the sterile frond, and was most likely intended for the same plant,’ which shows that he had some doubt about the identity of his collection with Burmann’s species. His speci- mens, however, are clearly the same as Polypodium radicans Burm. f., which is a characteristic species easily recognized by its scrambling habit and thick, elongate runners. Roxburgh gave his locality as “Malay Islands, etc.” It may be that his plant came from Penang Island. 108. POLYPODIUM RUPESTRE Roxburgh, Calcutta Journ. Nat. Hist. 4:488. 1844, non R. Brown, 1810, non Blume, 1828. =Polypodium taeniatum Swartz, Journ. Bot. Schrad. 1800 (2) :26. 1802. Type: Java, Thunberg. Crypsinus taeniatus (Swartz) Copel. Gen. Fil. 206. 1947. Type: A Roxburgh specimen in the Brussels Herbarium with the name in the hand of Roxburgh and the number 22 (Morton photograph 19931) is designated as lectotype. A second Roxburgh specimen in Brussels without the name in the hand of Roxburgh (Morton photograph 19932) has fewer and somewhat larger pinnae but is probably an isotype. According to Rox- burgh the species was collected in Pinang [i.e., Penang Island, Malaya] on mossy shaded rocks of granite; it was doubtless collected by W. Roxburgh, Jr. In the “Index Filicum,” P. rupestre Roxburgh was stated to be “Nephrolepis sp.” but this was a very bad guess on the part of Christensen. The authentic specimen above designated as lectotype is clearly a Polypodium [Crypsinus]; it was identified as Polypodium palmatum Blume; although it is near that species, it is referable rather to P. taeniatum Swartz, which differs in having the blade fully pinnate rather than deeply pinnatifid, according to Holttum (Ferns of Malaya 194-196. 1954). 109, PoLYPODIUM scABRUM Roxburgh, Calcutta Journ, Nat. Hist. 4:491. 1844, non Pres], 1822. =Thelypteris ferox (Blume) Tagawa & Iwat- ROXBURGH’S FERN TYPES—MORTON 359 suki, Acta Phytotax. Geobot. 23:54. Aug. 1968. [Same combination made by Reed, Phytologia 17:276. Oct. 1968.] Aspidium ferox Blume, Enum. Pl. Jav. 2:153. 1828. Dryopteris ferox (Blume) Kuntze, Rev. Gen. Pl. 2:812. 1891. Cyclosorus ferox (Blume) Ching, Bull. Fan Mem. Inst. Biol. Bot. 8:167. 1938. LECTOTYPE: Two Roxburgh specimens in the Brussels Herbarium, evidently parts of the same frond, with a label in the hand of Christopher Smith reading “On the high mountains of Amboina, Nov. 1796, no. 319”; the second sheet does not have an original label (Morton photographs 19666, 19667). Isotypes are in Geneva, labeled “Ind. Orient. Dr. Roxburgh” (Morton photograph 16933) and the J. E. Smith Herbarium nos. 1625-68, 1625-69, and 1625-70, Linnean Society (Morton photographs 20220, 20221, 20222) ; these are indicated as having been collected in Amboina by Christopher Smith. In the “Index Filicum,” P. scabrum Roxburgh is correctly indicated as equivalent to Dryopteris ferox, one of the better known species of the sect. Cyclosorus, distinguished by its sub- arborescent habit, large size, glabrous pinnae, and especially by the abundant, dark, hairlike scales present on the stipes and rhachises. 110. PoLYPopIUM SCARIOSUM Roxburgh, Calcutta Journ. Nat. Hist. 4:494. 1844. —Polystichum scariosum (Roxburgh) Morton, comb. nov. Polystichum prolifieans van Alderw. van Rosenb. Bull. Jard. Bot. Buitenzorg III, 2:170. 1920. Type: Deli, Bandar-baroe, Sumatra, 1200 m., June 16, 1918, Lérzing 5743 (isotype L, Morton photograph 2148). LECTOTYPE: Two Roxburgh specimens in the Brussels Herbarium, the sheet with the rhizome and stipe base with the name in Roxburgh’s hand, the other consisting of the balance of a complete frond (Morton photographs 19780, 19779). There is an isotype in Geneva, marked “Amboyna, Dr. Roxburgh” (Morton photographs 6563, 16642). Roxburgh gave the locality as Amboina. Polypodium scariosum Roxburgh is clearly the same as Poly- stichum prolificans, which is fairly distinct among the numerous bipinnate Polystichums in the rounded, not at all mucronate or aristate, lobes on the distal margin of the pinnules, the fairly large pinnules, and the presence of a large bud on the lower side of the rhachis a short distance from the apex. This species must be rare. I have seen it only from Sumatra, Malaya, and from Amboina, but it may well grow elsewhere, since Poly- stichums have been little studied. However, this could be a case of a wrong locality, and the specimens could have come from Penang. This could be assumed if study of all Amboina col- lections does not reveal the species there. 360 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM 111. POLYPODIUM SEMIPINNATUM Roxburgh, Calcutta Journ. Nat. Hist. 4:486. 1844. =Tectaria semipinnata (Roxburgh) Morton (see no. 5. Acrostichum semipinnatum), TYPE: Two Roxburgh specimens in the Brussels Herbarium, the sterile one with the name in Roxburgh’s hand, the fertile one not (Morton photo- graph 5111). Since no other specimens have been found in any herbarium collected by Roxburgh and named P. semipinnatum, these are doubtless the holotype. They were probably collected by W. Roxburgh, Jr. A native of the Malay Islands according to Roxburgh, surely from Penang Island, since the specimen agrees with material from there and the species does not grow in the Moluccas, the other area that Roxburgh perhaps sometimes referred to as the “Malay Islands.” Roxburgh described the same species twice and with the same specific epithet, once as Acrostichum semipinnatum and once as Polypodium semipinnatum. This was intentional and not a lapse, for under P. semipinnatum Roxburgh remarked “is very like my Acrostichum semipinnatum.” It is indeed so like that species that it cannot be distinguished even specifically, let alone generically. 112. POLYPODIUM SEMISAGITTATUM Roxburgh, Calcutta Journ. Nat. Hist. 4:491. 1844. =Thelypteris semisagittata (Roxburgh) Morton, comb. nov. LEcTOTYPE; A Roxburgh specimen in the Brussels Herbarium with the name in the hand of Roxburgh (Morton photograph 19645). A very similar specimen and surely a duplicate of the lectotype is in Geneva, marked “India Orient. Dr. Roxburgh” (Morton photographs 6562, 16932). According to Roxburgh, his species is a native of the Delta of the Ganges, Chittagong, ete. The lectotype likely came from Chittagong, where it may have been collected by Buchanan-Hamilton, since it agrees with specimens from there. The lectotype was identified by Baker as Nephrodium arbuscula Desy., but it can hardly be the true Thelypteris arbuscula (Willd.) K. Iwatsuki (Acta Phytotax. Geobot. 2:170. 1965), the type of which is from Mauritius, and it may occur also in Ceylon and southern India. This species is distinct from, but apparently allied to, T. papilio (Hope) Iwatsuki, which is much larger, does not have semisagittate lower pinnae, and has the pinnae more deeply lobed than in T. semisagittata. In the “Index Filicum,” Polypodium semisagittatum Roxburgh was referred without a query to Dryopteris arida (D. Don) Kuntze, but this species does not agree with Roxburgh’s descrip- tion or specimens. Roxburgh’s specimen resembles in general but is not identical with a specimen from Chittagong, March, 1880, Gamble 7827 (US), which was originally identified as Nephrodium truncatum Presl, reidentified as Cyclosorus subpu- bescens (Blume) Ching by Ching, and recently as Cyclosorus ROXBURGH’S FERN TYPES—MORTON 361 latipinna (Hook.) Tardieu by Iwatsuki. This species can hardly be the same as Cyclosorus subpubescens, as delimited by Holttum, or C. sumatranus. It does not agree with C. latipinna, which was described from Hong Kong as Nephrodiuwm molle var. lati- pinna Benth. (Flora Hong Kong 455. 1861), which does not have numerous reduced, butterfly-like lower pinnae. Apparently, no lectotvpe has ever been proposed for var. latipinna. There were four original syntypes: Hong Kong, Champion, Harland, Hance; Little Hong Kong, Wilford; of these, I choose Hong Kong, Hance 135 as lectotype (K, Morton photograph 20647), for it is the best developed of the syntypes. None of the syntypes was annotated by Bentham, nor by Hooker. Rather curiously, all the authors who have discussed this species, such as Tardieu, Holttum, and Ching, have cited Hooker as the author of the basionym, but the epithet latipinna originated with Bentham, and Hooker merely raised Bentham’s variety to specific rank. 113. POLYPODIUM SOPHOROIDES sensu Roxburgh, Calcutta Journ. Nat. Hist. 4:489,. 1844, non Thunb., 1794. =Thelypteris sumatrana (v. A. v. R.) Reed and T. dentata (Forsk.) E. St. John, sensu lato, both pro parte. AUTHENTIC MATERIAL: Two Roxburgh specimens with the name sophoroides in Roxburgh’s hand, The frond on the right of the sheet labeled “Ind. or. Roxb.” (BR, Morton photograph 19696) is Thelypteris sumatrana. The left-hand plant is the basal part of the plant with apex only on the other sheet (BR, Morton photograph 19697). This plant is Thelypteris dentata, sl. It is hairier and has persistent indusia, and the superior segments of the basal pinnae are toothed. The specimen of T. sumatrana is glabrate, the indusia are smaller, and the basal segments are not toothed. According to Roxburgh, his material came from the “Moluccas, ete.,” which is true of the specimen of T. dentata, s.l. The specimen of T. sumatrana is probably from Penang, for it matches a specimen in the British Museum (Natural History) marked as from “Princes Island, Dr. Banks” (Morton photograph 20869). This means that the specimen was sent to Dr. Banks, who never was in Penang, by Roxburgh, for only Roxburgh called Penang “Prince of Wales Island’; other earlier collectors called it “Pullo Pinang.” That Roxburgh’s concept of P. sophoroides was confused is shown by his description of the sori as “sometimes in a single line on each side of the nerve; sometimes they form a nearly continued line near the margins of the incisures.’”’ The former condition is shown by the specimens, cited above as authentic, that represent T. swmatrana, and the same condition obtains in the true Polypodium sophoroides Thunb. By the second kind of sori Roxburgh intended to describe a condition where the sori are present only near the sinuses between the segments, i.e., that only the lowest pair, or the lowest two pairs perhaps, of veinlets are fertile; this condition is shown by specimens in Geneva 362 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM (Morton photographs 6561, 16661), one of them with the name sophoroides in Roxburgh’s hand; these specimens are matched by one in the British Museum from Gilolo Island, Moluccas, col- lected by Christopher Smith in November, 1801; these specimens with only the lowest veinlets fertile do not represent 7. swma- trana, but appear to be forms of Thelypteris dentata, s.1. Roxburgh’s identification of his Penang material as Polypo- dium sophoroides is not far wide of the mark, for his specimens do resemble that species, which, however, has a more northern range in Japan and China, and which differs in having most of the pinnae with a somewhat elongate superior basal segment and in having the basal pinnae not reduced, among other characters, especially of pubescence. Polypodium sophoroides Thunb. is a taxonomic synonym of Thelypteris acuminata (Houttuyn) Mor- ton. Strangely, in the fourth supplement of the “Index Filicum,” the authority for this species is given as Thelypteris acuminata (Panz. in Christm. & Panz.) Morton, with the basionym Poly- podium acuminatum Panz. in Christm. & Panz. Pflanzensyst. 13(1):204, t. 99, f. 2. 1786, which is entirely wrong for both author, book, and date; the author is properly Houttuyn, and the citation of the basionym is P. acuwminatum Houtt. Nat. Hist. II, 14:181, ¢. 99, f. 2. 1783, as given by Merrill (Journ. Arn. Arb. 19:313. 1938) and as I gave it (Amer. Fern Journ. 49:139. 1958 [1959]). 114. PoLYPODIUM SQUARROSUM Roxburgh, Caleutta Journ. Nat. Hist. 4:494. 1844, =Polystichum squarrosum (D. Don) Fée, Gen, Fil. 287. 1852. Aspidium squarrosum D. Don, Prody. Fl. Nepal. 4. 1825. Type: Narain- hetty, Nepal, Mar. 7, 1808, Buchanan-Hamilton (BM, Morton photo- graph 20886). LECTOTYPE: A Roxburgh specimen in the Brussels Herbarium with the name in Roxburgh’s hand (Morton photograph 19777). There is an isotype in Geneva, with the label “Mountains of Sirinagur, Capt. Hardwicke “(Mor- ton photographs 6560, 16703). According to Roxburgh, the species was “found by Captain Hardwicke on the tops of the mountains between Hurd- war and Sirinayur,” now Hardwar, United Provinces, and Srinagar, Kash- mir. In the “Index Filicum,” P. squarroswm Roxburgh is listed under Polystichum lobatum var. 4 as though it were a transfer of Aspidium squarrosum to Polypodium, but it obviously is not, for Don’s species was published more than ten years after Rox- burgh’s death. Roxburgh’s species was described as new, but taxonomically it appears the same as Don’s later species with the same epithet, an apt epithet referring to the squarrose and con- spicuous stipe scales. I have not seen the type of Don’s species, ROXBURGH’S FERN TYPES—MORTON 363 however, and cannot be sure that it is identical. The whole group of Polystichum aculeatum in India is in an inextricable tangle at present, and it will take monographic study to straighten it out. The name P. squarrosum, however, will likely remain as a correct name, since it is older than competing epithets other than P. aculeatum (L.) Schott, P. setiferum (Forssk.) Moore ex Woy- nar, and P. lobatum (Hudson) K. B. Presl; any one or all of these three may be correct for Indian species, although, all are founded on European types. 115. POLYPODIUM TENERUN Roxburgh, Calcutta Journ. Nat. Hist. 4:490. 1844. —?Thelypteris sericea (Scott. ex Bedd.) Reed, Phytologia 17:313. 1968. ?Lastrea sericea J. Scott ex Beddome, Ferns Brit. India ¢. 308. 1867. Type: Chittagong, J. Scott (BM, Morton photograph 20859, marked as original for t. 308.) ?Lastrea calcarata var. sericea (J. Scott ex Bedd.) Bedd., Handb. Ferns Brit. India 237. 1883. ?Dryopteris pseudocalcarata C. Chr. Ind. Fil. Suppl. 3:95. 1934. Based on Lastrea sericea J. Scott ex Bedd., non Dryopteris sericea C. Chr. 1913. TYPE: No specimens from the Roxburgh Herbarium have been located. The species was collected in Silhet, East Pakistan, by M. R. Smith in 1811, (cf. Hort. Bengal. 75. 1814), and was cultivated in the Calcutta Botanic Garden. The original description is: “Fronds alternately-pinnate; pin- nae linear-lanceolate, gash-serrate, acuminate. Fructification in a few spots on each side of the veins; involucre reniform. A na- tive of Silhet. In the Botanic Garden at Calcutta, it grows to the height of from 12 to 18 inches, is of a delicate soft texture, and somewhat villous.” This description seems surely to represent a species of Thelypteris sect. Lastrea, from the small size, simply pinnate blades, delicate texture, villous condition, and reniform indusia. The only species I have found that grows in the region of Silhet, which belongs in the region known as Chittagong formerly, is the one described as Lastrea sericea J. Scott ex Bedd. This species was growing in the Calcutta Botanic Garden as late as 1869, for C. B. Clarke collected it there on Nov. 15, 1869 (Clarke 10375, BM, Morton photograph 20860); Clarke reported that the cultivated plant had come originally from Chittagong. My guess is that the plant Scott collected and marked as from Chittagong, the type of Lastrea sericea, was actually collected in the Calcutta Botanic Garden, from the same plant or plants that served Clarke, and that these were the same plants or de- scendants of Roxburgh’s original Polypodium tenerum. Many of 364 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM Roxburgh’s plants did survive in the garden throughout the nineteenth century (and perhaps still do), but many lost their labels, if they ever had any. Thus Roxburgh’s P. tenerum, an appropriate name for a plant of this delicate texture, and Scott’s Lastrea sericea, also an appropriate name for a plant that is obviously pubescent, may have been based on the same material. I feel that this is true, and yet in the absence of any Roxburgh specimen I should hate to propose a new combination, even though Roxburgh’s name has priority. Moreover, this plant belongs to a troublesome group of species that are perhaps not quite properly delimited by Ching in his paper on Sikkim- Himalaya Thelpteris. In particular, the species called by Ching Thelpteris cana (J. Smith) Ching is impossibly confused. The basionym cited is a nomen nudum, based on material from northern India, and all the references cited by Ching refer to plants from northern India, although Ching states that his spe- cies occurs only in southern India. He cites as “type” a Wight col- lection not mentioned in any of the cited synonyms. It seems that Thelypteris cana Ching must be treated as a new species, and since it was published after 1935 yet has no Latin diagnosis, it is invalid and can be disregarded. Still, it would be interesting to know what plant was intended. 116. POLYPODIUM TOMENTOSUM Roxburgh, Calcutta Journ. Nat. Hist. 4:483, t. 29 (right). 1844,.non DuPetit Thouars, 1804, non Bory, 1833. =Pyrrosia flocculosa (D. Don) Ching, Bull. Chin. Bot. Soc. 1:66. 1935. Polypodium flocculosum D. Don, Prodr, Fl. Nepal. 1. 1825. Type: Narain- hetty, Nepal, Buchanan-Hamilton (not seen). Cyclophorus floceulosus (D. Don) C. Chr. Ind. Fil. 199. 1905. Polypodium detergibile Hook. Sp. Fil. 5:49. 1864. Type: Based on various collections from Bhotan (Griffith), Nepal (Wallich as P. vestitum in Herb. Hook.), Simla (Madden, Edgeworth), Kumaon, and Sylhet and Assam (Strachey & Winterbottom, Hooker f. & Thomson). Of these I choose Kopkut, Kumaon, Strachey & Winterbottom 402 as lectotype (K, Morton photograph 20653). Hooker’s comment “J.Sm.? vix Don?” is hardly comprehensible, since there are no species named detergibile by J. Smith or Don. LECTOTYPE: A Roxburgh specimen in the East India Company Herbarium, no. 269-8 (Morton photograph 15736, upper left-hand plant). According to Roxburgh, this species was collected in “Hindoostan, on trunks of trees’; by Hindoostan Roxburgh was probably referring to West Punjab (now in Pakistan), East Punjab, or Rajasthan in northwestern India as distinguished from Bengal in eastern India. An authentic drawing by Roxburgh agreeing with the published illustration is at Kew, no. 1746 (Morton photograph 15875). On the sheet with the lectotype are two plants at the right hand representing Polypodium mysurense Heyne ex Wallich, Num. List. no, 269. 1829, nom. nud. [=Pyrrosia mollis (Kunze) Ching]. The plant at the lower left is Herb. Wight, also P. mollis. ROXBURGH'S FERN TYPES—MORTON 365 117. POLYPODIUM TRIDENTATUM Roxburgh, Calcutta Journ. Nat. Hist. 4:495. 1844. =Thelypteris torresiana (Guud.) Alston, Lilloa 30:111. 1960. Polystichum torresianum Gaud. in Freyc. Voy. Bot. 333. 1824. Type: Marianna Islands, Gaudichaud (Holttum, Blumea 17:27. 1969, states the type is in Paris, but it has not been located there; the holotype is in Geneva, according to Dr. F. R. Fosberg). Macrothelypteris torresiana (Gaud.) Ching, Acta Phytotax. Sinica 8:310. 1963. LEcTOTYPE: A specimen (Morton photograph 19726) from the Roxburgh Herbarium in Brussels with the locality Banda, the number 327, and the date May, 1797; although not so indicated, the collector was surely Christopher Smith, who was collecting in the Molucea Islands in 1797. Two duplicates in Brussels are evidently a part of the same collection (Morton photograph 19727). An isotype is in Geneva, indicated as “Banda, Dr. Roxburgh” with a label reading “Typus” (Morton photographs 6559, 16648). A sheet in the J. E, Smith Herbarium (no. 1625-145) in the Linnean Society (Morton photograph 20246) is another probable isotype, for it agrees with the lecto- type above completely, however, it is indicated as having been collected by Christopher Smith in Amboina in 1797; the locality “Amboina” is likely an error for “Banda,” although it is quite possible that Smith did collect this species twice, once in Banda and once in Amboina. Holttum (Blumea 17:25-32. 1969) has recently recognized Thelypteris torresiana, T. setigera (Blume) Ching, and seven other Asiatic and Malesian species as a genus Macrothelypteris, distinct from Thelypteris, following the lead of Ching, but he did not ex- plain why it is necessary to recognize this or other groups as genera rather than as subgenera or sections. The only charac- ters used to separate the genera Macrothelypteris and Pseudo- phegopteris are the presence generally of hair-pointed scales on the axes and the presence of multicellular, hyaline hairs in the former; Pseudophegopteris has scales (but not hair-pointed scales), and hairs (but these are unicellular). Such vegetative dif- ferences may be important and indicate relationships, but there is no necessary reason that they should be considered generic characters, rather than subgeneric. The characters mentioned have very recently been discussed by Pichi-Sermolli (Webbia 24:713-717. 1970), who has also indicated the type of Macro- thelypteris as Nephrodium oligophlebium Baker correctly. Holt- tum proposed to change the designated type to another species, Polystichum torresianum Gaud., but this is not permissible. 118. POLYPODIUM UNITUM sensu Roxburgh, Calcutta Journ. Nat. Hist. 4:488. 1844, non L., 1759. =Thelypteris totta (Thunb.) Schelpe var. hirsuta (Mett.) Morton, Contr. U. 8S. Nat. Herb. 38:73. 1967 (with synonymy). AUTHENTIC MATERIAL: A Roxburgh specimen in the Brussels Herbarium with the name in the hand of Roxburgh (Morton photograph 19682). A Roxburgh drawing at Kew (no. 1749) (Morton photograph 15876) labeled 366 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM Polypodium unitum is apparently not the same thing and must remain unidentified. Roxburgh did not consider this a new species but referred it to Burmann, presumably Burm. f. Fl. Ind. 232. 1768, although not so stated, but Burmann merely quoted Linnaeus, who is the au- thor of Polypodium wunitum. Roxburgh cited “Burm. zeyl. t. 44, f. 1,” which is one of the original citations by Linneaus for this species. Roxburgh was not the only one to misapply the name unitum to Thelypteris totta, for this was the common opinion up until this century, and the matter is perhaps still not definitely settled. 119. PTERIS AMPLEXICAULIS Roxburgh, Calcutta Journ. Nat. Hist. 4:505. 1844. =Pteris vittata L. Sp. Pl. 1074. 1753. Type: China, Osbeck. LECTOTYPE: A specimen in the Brussels Herbarium with the name in the hand of Roxburgh (Morton photograph 19893). Roxburgh cited as locality: “Nat. of Bengal, in shady, moist places. Also among ruins of brick buildings. Is sometimes parasitical,” which indicates that he had several specimens in hand; this is borne out by the specimens seen, all of which represent P. vittata, but they are not identical and were probably collected at different times and places. These syntypes are: “Ind. or.,” Roxburgh in Herb. Mart. (BR, Morton photograph 19895); Bengal, Roxburgh (BM, Morton photograph 7563) ; without locality, Roxburgh (East Ind. Co. Herb. 112-4, K, with name in the hand of Roxburgh, Morton photograph 19587b); Botanical Garden [Calcutta], this sheet probably collected by Wallich, Dec. 11, 1814 (East Ind. Co. Herb. 112-4, K, Morton photograph 15718); drawing by Roxburgh (K, no. 1753, labeled P. vittata rather than P. amplexicaulis, Morton photograph 15883; same drawing BM, Morton photograph 15777). As mentioned under Pteris vittata sensu Roxburgh, Roxburgh misidentified another species as P. vittata L. and redescribed the true P. vittata as a new species P. amplexicaulis. 120. PTERIS ANGUSTIFOLIA sensu Roxburgh, Calcutta Journ. Nat. Hist. 4:503, t. 33 (left). 1844, non Swartz, 1788. =Vittaria sp. No herbarium specimens have been seen, only the published drawing and two original drawings representing the same plant at Kew (no. 1751, Morton photograph 15878) and the British Museum (Morton photograph 15775). The Kew drawing was first named “Pteris parasitica,’ which was changed in a different hand to “Pteris angustifolia Roxb.” The name “P. parasitica” was never published, and the name ‘“‘Pteris angustifolia” appears in the printed work attributed to Swartz rather than Roxburgh. (Roxburgh’s own species are, at least usually, indicated with an “R.”) Therefore, P. angustifolia must be considered as a mis- identification of P. angustifolia Swartz and not as a new species. ROXBURGH’S FERN TYPES—MORTON 367 The true P. angustifolia Swartz might appear from a diagnosis to be the same, but that is a quite different species, now known as Ananthacorus angustifolius (Swartz) Underw. & Maxon. Rox- burgh’s plant was from “the Delta of the Ganges, where it is found growing on the trunks of trees, intermixed with mosses, etc. parasitic plants, of various kinds.” It was collected in 1796, according to Roxburgh (Hort. Beng. 75. 1814), and cultivated in the Calcutta Botanical Garden. From the description and drawing I am unable to identify Roxburgh’s plant specifically. If one knew which species of Vit- taria grew in the Delta of the Ganges, one could probably identify the species correctly, but I do not have this information. The plant is shown as having a rather broad blade, and so it may be Vittaria amboinensis Fée as treated by Ching (Sinensia 1:189. 1931) or a small form of V. scolopendrina (Bory) Thwaites. 121. PTERIS BICOLOR Roxburgh, Calcutta Journ, Nat. Hist. 4:507. 1844. =Cheilanthes farinosa (Forssk.) Kaulf. Enum, Fil. 212. 1824. LectoTyPpE: A Roxburgh specimen in the Brussels Herbarium with the name in the hand of Roxburgh and the number 2416 (Morton photograph 20004). The loeality is given by Roxburgh as “mountains north of Rohileund,” i.e., the present Rohilkhand, in the Division of Agra, Northern United Provinces. Another specimen, an isotype, is in the East India Company Herbarium, no. 71-5 (K, Morton photograph 14698, upper and lower left- hand plants and the small plant upper center). To the original description Griffith added the reference “Cheil- anthes dealbata Wall. Cat. 61, no. 71,” which does not refer to the main entry under no. 71 (which is C. dealbata Wall.) but to the entry on page 61, referring to no. 71-5, the Roxburgh col- lection; the meaning is not that Pteris bicolor is a renaming of Cheilanthes dealbata Wall. but merely a particular entry in the Wallich List, namely no. 71-5. In any case, Cheilanthes dealbata Wallich is a nomen nudum. It cannot be assumed that it is the same as C. dealbata D. Don, which is not mentioned; and the name C. dealbata is attributed to himself by Wallich. As a matter of fact, Cheilanthes dealbata D. Don (1825) is an illegitimate later homonym of C. dealbata Pursh, a quite different plant of North America, and so was in need of renaming. If Pterts bicolor Roxburgh is the same as Cheilanthes dealbata D. Don, and if this species is distinct from C. farinosa, then the epithet bicolor will be correct. Pteris bicolor Roxburgh was correctly understood in the “In- dex Filicum.” There is a drawing in the British Museum of Cheilanthes farinosa that might be by Roxburgh (Morton photo- graph 15758). A revision of the Asiatic species of this alliance 368 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM was published by Ching (Hong Kong Natur. 10:194—-204. 1941), who, however, overlooked the name Pteris bicolor Roxburgh. 122. PTERIS DAUCIFOLIA Roxburgh, Calcutta Journ. Nat. Hist. 4:508. 1844. =Onychium siliculosum (Desv.) C. Chr. Ind. Fil. 468. 1906. Type: “America australis,’’ Herb. Desvaux, P. (fide Ching); the locality is surely an error, since Onychiwm does not occur in South America, and the species represented by the type is exclusively Asiatic; Ching (Ling- nan Sci. Journ. 13:493-501. 1934) suggested that the type came from the Philippine Islands, since a specimen collected by Gaudichaud in Manila matches it exactly. TYPE: Manipur, Assam, India, Roxburgh (holotype BR, Morton photograph 20008). This is one of the few Roxburgh collections in Brussels bearing a definite locality; the original locality was stated as “eastern parts of Bengal,” which included Assam in Roxburgh’s time. Since this is the only collection seen in any herbarium, it may be presumed to be a holotype. In the “Index Filicum,” Pteris daucifolia is referred with a query to Cheilanthes tenuifolia Swartz, on what grounds I do not know. The description, although brief, clearly indicates Ony- chium rather than Cheilanthes in the character of the fertile segments being linear with the margins completely occupied by the fructifications. 123. PTERIS DIMIDIATA Roxburgh, Calcutta Journ. Nat. Hist. 4:507. 1844, non Willd., 1810. —Pteris semipinnata L. Sp. Pl. 1076. 1753. LECTOTYPE: East India Company Herbarium, no. 97-3 (left-hand plant) (K, Morton photograph 14725). A duplicate is in the British Museum col- lected in Silhet by Smith, Herb. Roxburgh (Morton photograph 20853). Roxburgh stated that his plant came from Chittagong, where it was collected by M. R. Smith in 1811 (Hort. Bengal. 75. 1814). Silhet is in the region called Chittagong by Roxburgh and others in the early nineteenth century. Pteris dimidiata Roxburgh is omitted in the “Index Filicum,” probably because it was assumed the same as Pteris dimidiata Willd. (Sp. Pl. ed. 4, 5:381. 1810). However, Roxburgh did not cite Willdenow but himself as author, and indeed he does not mention Willdenow’s fern volume of the “Species Plantarum” in any place, which he surely would have if he had seen it. There- fore, Roxburgh must have written his description and assigned the name dimidiata independently, which is not unlikely con- sidering that the epithet dimidiata is a natural one for a plant with pinnae of this particular shape. Both Roxburgh’s and Will- denow’s species seem to be surely Pteris semipinnata L., and it is strange that they did not realize this. After Roxburgh’s name, Griffith added “Pteris semipinnata Linn. Wall. Cat. 62, No. 97,” but this is not a reference to P. semipinnata L. of 1753, but only that Roxburgh’s plant is P. ROXBURGH’S FERN TYPES—-MORTON 369 semipinnata sensu Wallich as to a plant listed on page 62 of his “List” under the number 97 (which should have been no. 97-3). The other specimen mounted on no. 97-3 on the right-hand side was collected at Sylhet by Di Silva, a collection not listed by Wallich; it is a topotype of Roxburgh’s species. 124, PTERIS GRACILIS Roxburgh, Calcutta Journ. Nat. Hist. 4:508. 1844, non Michx., 1803, =Pteris ensiformis Burm. f. Fl. Ind. 230. 1768. LECTOTYPE: A Roxburgh specimen in the British Museum, collected in Silhet by Smith (Morton photograph 20852). Roxburgh stated that his plant came from Chittagong, the name of his time for the region in which Silhet is found. The collector was surely M. R. Smith, who collected other plants for Roxburgh in Silhet. In the “Index Filicum,” Pteris gracilis Roxburgh is referred without a query to Cheilanthes tenuifolia Swartz, but this is an obvious error, for the original description does not agree with that species. The description does agree in all particulars with Pteris ensiformis. Although the name P. gracilis does not occur on the specimen selected as lectotype, there can be no doubt that it does represent Roxburgh’s species and is probably indeed the holotype, since no other Roxburgh specimen of this species has been found in other herbaria. A Roxburgh specimen of Cheilanthes tenuifolia Swartz in Brus- sels (Morton photograph 20005) is not named by Roxburgh and does not agree with Roxburgh’s description of his P. graci- lis. 125. PTERIS GRAMINIFOLIA Roxburgh, Caleutta Journ. Nat. Hist. 4:502, t. 33 (middle). 1844. = ?Vittaria elongata Swartz, Syn. Fil. 109, 302. 1806. Type: “India orientalis,” Rottler (holotype S-PA, Morton photograph 6105). TYPE: Roxburgh t. 33 (middle plant). Since no herbarium specimens of this species have been located, the published illustration will have to stand as the type, pro tem. According to Roxburgh (Hort. Beng. 75. 1814), it was collected in Silhet [Sylhet] by M. R. Smith in 1811. The entire original description is: “Parasitic. Fronds linear, very long (2-8 feet) entire pendulous. Nat. of the close dark forests of Silhet, where it is found suspended on trees, resembling long tufts of long, narrow-leaved grass.” In the “Index Filicum,” P. graminifolia is referred to Vittaria elongata Swartz, and this may be correct. If the discovery of a herbarium specimen should prove this wrong, the epithet “graminifolia,” although legitimate, could never be transferred to Vittaria, because there already exists a valid and different species Vittaria graminiifolia Kaulf. (1824). 370 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM 126. PTERIS LINEARIS Roxburgh, Calcutta Journ. Nat. Hist. 4:505. 1844, non Poir., 1804. =Pteris vittata L. Sp. Pl. 1074. 1753, sens. lat. Tyee: Amboina, 1796, Roxburgh 310 (BR, Morton photograph 19892). The type was surely collected by Christopher Smith, who was in Amboina in 1796. The detached stipe on this type sheet is large and coarse and surely does not go with the type frond. It may go with another sheet in the Brussels Herbari- um, also labeled Pteris linearis in the hand of Roxburgh, which bears the number 2415 (Morton photograph 19894). This second specimen is not marked as from Amboina, and it may conceivably represent a different species, since the blade is a good deal larger and the stipe thicker than that of P. vittata. Since it is not the type, the placing of it is not of importance, A specimen in the East India Company Herbarium, no. 111-7 (K, Morton photograph 15717), is determined as “P. vittata & linearis Hb. Roxb.”; the small fragment at the upper right seems to agree with the lectotype of P. linearis, and the larger plant with the specimen in Brussels indicated above as larger and possibly different. In the “Index Filicum,” P. linearis Roxburgh is listed as a dubi- ous species, with the notation ‘an Wall.?,” referring to the pre- ceding entry Pteris linearis Wall. List, no. 105. 1828. However, Roxburgh’s species is by no means the same as that, which is considered to represent Pteris tripartita Swartz. Pteris linearis Roxburgh is a form of the common and variable P. vittata L., which may be an aggregate; it has a number of somewhat doubt- ful synonyms. 127. Preris LOBATA Roxburgh, Calcutta Journ. Nat. Hist. 4:504. 1844, non Goldm. 1843. =Polypodium scolopendria Burm. f. Fl. Ind. 232. 1768. Type: A specimen in the Brussels Herbarium with the name in the hand of Roxburgh (Morton photograph 19928). According to Roxburgh the type came from the Moluccas, where it was doubtless collected by Christopher Smith in either Amboina, Ternate, or Honimoa. Since this is the only speci- men seen with the name in Roxburgh’s hand, this may be considered unique and a holotype. In the “Index Filicum,” Pteris lobata Roxburgh is listed as a dubious species, and with good reason, for the original descrip- tion is merely ‘“Petiole smooth, nearly as long as the thin, pol- ished, 2 or 3 lobed fronds,” quite the shortest and most inade- quate of Roxburgh’s descriptions. One could never guess the identity of the species if a type had not been located. Roxburgh’s type was sterile, but even so his reference to Pteris rather than Polypodium must be considered a temporary aberration. 128. PTERIS LUNULATA sensu Roxburgh, Calcutta Journ. Nat. Hist. 4:506. 1844, —Adiantum lunulatum Burm. f. Fl. Ind. 235. 1768. Lectotype: A Burmann specimen in Geneva with the notations “Capillaris malab. non ramosa folio rotundodentato, Petiv. Tab. 59, f. 10. H. Mal. tom. 12 tab. 40” (Morton photograph 16782). ROXBURGH'S FERN TYPES—MORTON 371 MATERIAL EXAMINED: A specimen in Brussels with the name Pteris lunulata Retz.? in the hand of Roxburgh (Morton photograph 19875); a specimen in the East India Company Herbarium, Kew, no. 77-8 (Morton photograph 15730, left-hand plant). Roxburgh did not localize his material, but stated that the species was common in most parts of India. Roxburgh attributed his plant to ‘“‘Retz. Obs. 2, No. 99, t. 4,” and wrote “Common in most parts of India, sometimes the mar- gin is broken, when it resembles an Adiantum, and is very likely A. luniulatum, but at all times sufficiently distinct.” This shows a little confusion, for P. lwnulata Retz. is based on Adiantum lunula- tum Burm. f., and is thus taxonomically identical. Roxburgh was attempting to distinguish between plants with the indusial flaps elongate and unbroken, calling these a Pteris, and plants with several separate indusia, calling these an Adiantum. Both forms occur in typical Adiantum lunulatum, as shown by Mehra and Verma (Journ. Indian Bot. Soc. 42A:110--121. 1963). They have not been given distinctive names. I am adopting the name Adiantum lunulatum Burm. f. rather than A. philippense L. in conformity with the arguments ad- duced by Verma (Nova Hedwigia 8:463-468, 1961). Adiantum philippense L. was based solely on a drawing by Petiver which is so poor as to be unidentifiable; the drawing cannot really be matched by any specimen, and so A. philippense L. ought to be regarded as a dubious and permanently unidentifiable species, for its identity as conspecific with A. lunulatwm can be guessed at but never proved. On the other hand, A. lunulatim has a good type specimen extant, and its identity is definitely established. Mehra and Verma in the publication cited are able to identify the type with a collection that has been cytologically analyzed, and find that it is a sterile triploid, which is apparently the com- mon form in much of India. Other forms do occur—apogamous and sexual diploids and sexual tetraploids—which may be dis- tinguished morphologically somewhat, but no names have been applied to these. 129. PTERIS MULTIFIDA Roxburgh, Calcutta Journ. Nat. Hist. 4:507. 1844, non Poir., 1804. =Doryopteris ludens (Wallich ex Hook.) J. Smith, Hist. Fil. 289. 1875. Basionym: Pteris ludens Wallich ex Hook. Sp. Fil. 2:210. 1858. LECTOTYPE: A specimen in the British Museum collected by Roxburgh, without further data (Morton photograph 15786). The specimen does not have the name P. multifida written on it (except by me in 1967), but it does agree clearly with Roxburgh’s description and is surely authentic. Since it is the only Roxburgh herbarium specimen seen of the species, it is desig- nated the lectotype. If another specimen should be found with the name 372 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM written on it, that specimen can replace the present one as a lectotype. Roxburgh’s plant came from Chittagong, East Bengal, now in East Pakistan. It is fortunate that Roxburgh chose an epithet “multifida”’ that had already been used in Pteris for another species, or otherwise the rather well-known name Doryopteris ludens would have to be replaced. Pteris ludens Wallich appeared in Wallich’s “Numerical List’ as no. 88 in 1829, but only as a nomen nudum, and the name was not validated until the publication by Hooker in 1858, 12 years after the publication of Roxburgh’s name. Tryon (Contr. Gray Herb. 1438:60. 1942) stated that the type of Pte- ris ludens is “India, Wallich 88, Kew, not seen, photo G, seen,” but this is not correct. When published, Pteris ludens Wallich ex Hook. contained no type but four syntypes. Tryon’s designation of Wallich 88 as the type might be considered as a selection of a lectotype, except for the fact that Wallich List no. 88 consists of two different collections: one Scendnea, Irawaddy, Wallich in 1826, and the other Caves of the Mountain Nidan, Assam River, Wallich in 1827. Tryon cannot really be properly credited with having selected a lectotype, since that can hardly consist of two specimens from different localities. I designate the collection from Irawaddy in Kew as lectotype; this is presumably the one intended by Tryon, since it is the one represented by a photograph by Una F. Weatherby, in the Gray Herbarium. 130. PTERIS PECTINATA Roxburgh, Calcutta Journ. Nat. Hist. 4:507. 1844, non Cav. 1802, non Desv. 1811, non Don, 1825. =Pteris longipinnula Wallich ex Agardh, Rec. Gen. Pterid. 19. 1839. Types: Mountains of Penang, Wallich List no. 108 (K-Hb. Hook., Herb. Linn Soc. London— now K-E. Ind. Co. Herb. no. 108—Morton photographs 19586, 19586a). The two sheets in the East India Company Herbarium are designated lectotype. TYPE: A specimen in the Brussels Herbarium with the name in the hand of Roxburgh (Morton photograph 19896). Since this is the only specimen seen of this species from the Roxburgh Herbarium, it must be considered a holotype. In the “Index Filicum,” P. pectinata Roxburgh is listed as a dubious species. The type has been identified by Baker as P. longipinnula Wallich, which is correct. Roxburgh indicates the type as being from the Molucca Islands, which may very well be right; however, I have seen no other specimens from there. It may be that “Moluccas” was a slip and that the plant actually came from Penang, where P. longipinnula came from also. Holt- tum in his “Ferns of Malaya” describes (but does not name) three forms of the species as it grows in Malaya aside from the typical form. Roxburgh’s P. pectinata (a homonym three times ROXBURGH’S FERN TYPES—-MORTON 373 over) agrees with the typical form. The other forms, which have basally forked lower pinnae, may represent hybrids, as Holttum remarks, possibly with P. asperula J. Smith ex Hieron. A varia- tion of this kind (in mature plants) is hardly to be expected within a normal species. The above identification of P. pectinata with P. longipinnula is a bit doubtful. The latter has three or four pairs of subop- posite lateral pinnae. Only the basal pair of pinnae are present in the Roxburgh type; Roxburgh noted that the pinnae are few and opposite, and so it is likely that the species is similar to P. longipinnula in this respect. There are plants from Java and Sumatra, probably not specifically different from P. longipinnula, that seemingly have only a single pair of pinnae, with an elon- gate, subconform terminal apex (“pinna’’). Such a one is Pteris megaphylla Mett. ex Miquel (Ann. Mus. Lugd. Bat. 4:96. 1868). I have seen a syntype—Sumatra, Korthals (L, Morton photo- graph 2178). This species was placed as a synonym of P. biaurita L. in the “Index Filicum,’” but it is not that, because the veins are free and not united into a costal arc. An extremely similar, if not identical, plant is Pteris salakensis van Alderw. van Ros- enb. (Bull. Jard. Bot. Buitenzorg IJ, 7:26. 1912), the type of which was a cultivated plant in the botanical garden at Bogor, said to have been brought originally from Mount Salak, Java. I have seen an isotype (or the holotype?) in Leiden (Morton photograph 2160). The latter species was said in the third supple- ment of the “Index Filicum” to equal P. reducta Baker, but from Baker’s description that hardly seems possible. Much more distinct is Pteris longipinnula var. hirtula C. Chr. (Contr. U. S. Nat. Herb. 26:312. 1931), and it seems to me that this must represent a distinct species—Pteris hirtula (C. Chr.) Morton, comb. nov. In P. longipinnula the stipe and rachis are a shining olive green and are absolutely glabrous. Pteris hirtula has the stipe and rhachis stramineous, and both are obviously and strongly pubescent. The latter character is most unusual in Pteris. This species is still known only from the material seen by Christensen—the type from the valley of Meh Len, Keng Tung Territory, Burma, Rock 2133 and 2091 (both US). The type has three pairs of lateral pinnae, whereas the paratype has only a single pair, thus paralleling the difference between typical P. longipinnula and P. megaphylla, mentioned above. 131. PTERIS PEDATIFIDA Roxburgh, Calcutta Journ. Nat. Hist. 4:508, 1844. =—Pteris tripartita Swartz, Journ. Bot. Schrad. 1800(2):67. 1802. Type (from Swartz, Syn. Fil. 293. 1806): Java, Thunberg. 374 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM Pteris intermedia Blume, Enum. Pl. Jav. 2:211, 1828. Type: Celebes, Reinwardt (holotype L, Morton photograph 2218). LECTOTYPE: A sheet in the Brussels Herbarium with the name in the hand of Roxburgh and the number 2419 (Morton photograph 19904). Roxburgh gave the localities as Molucca Islands and Malay Islands; this lectotype presumably came from the Moluccas where it was presumably collected by Christopher Smith, since it agrees with a Smith collection from Honimoa, April, 1797, now in the herbarium of J. E. Smith, no. 1634-3 (LINN, Morton photograph 20285), which is an isosyntype. Two other Roxburgh specimens in Brussels (Morton photographs 19905, 19906) are not localized and do not bear the name in Roxburg’s hand, but they are surely syntypes also. They are conspecific but just slightly differently dried, and so I judge that these are not duplicates of the lectotype chosen above but may represent the second locality cited by Roxburgh, namely “Malay Islands,” which in this case surely means Penang Island, Malaya. The species does grow in the Moluccas and in Penang. In the “Index Filicum,” Pteris pedatifida Roxburgh is listed as a dubious species. Another plant, Pteris attenuata Swartz (Journ. Bot. Schrad. 1800(2):66. 1802), was also collected in Java by Thunberg. In the “Index Filicum,” this is given as a synonym of P. tripartita, but it may be different, judging by a fragment in the Swartz Herbarium in Stockholm (Morton photograph 6259) which shows a plant with the pinnules pinnatisect almost or quite to the costa and the segments elongate and almost linear. This fragment was probably removed by Swartz from the holotype in the Thunberg Herbarium, but there is some confusion, for it is labeled as being from “Cap B. Spei,” i.e., the Cape of Good Hope, South Africa. It remains to be determined if the original Thunberg collection was from Java or from South Africa. If it is really from South Africa it may represent the later described Pteris buchananii Sim; if it is from Java it may be distinct from P. tripartita, sens. str., and probably the same as P. tripartita var. dissoluta van Alderw. van Rosenb. (Bull. Jard. Bot. Buitenzorg II, 23: 20. 1916), the type of which came from Benkoelen Lebong Tandai, Sumatra, Brooks 2238 [cited as 223/S] (isotype L, Morton photo- graph 2220). 132. PTERIS PILOSELLOIDES sensu Roxburgh, Calcutta Journ. Nat. Hist. 4:503. 1844. =Drymoglossum piloselloides (L.) K. B. Presl, Tent. Pterid. 227, t 10, f. 5,6. 1836. Basionym: Pteris piloselloides L. Sp. Pl. ed. 2, 1530. 1763. Roxburgh referred his P. piloselloides to Linnaeus and did not describe it as a new species. His description and drawing at Kew (no. 2576, Morton photograph 15880) show that his plant was indeed the Linnaean species. Pteris piloselloides L. came from ROXBURGH’S FERN TYPES—MORTON 375 “India orientalis,” but has not been typified. Christensen did not mention a type in his rather detailed account of Drymoglossum (Dansk Bot. Ark. 6(3):83-91. 1929). Since Linnaeus did not cite any literature references, it is clear that his species is based on a specimen. This is not in the Linnean Herbarium in London, and so should be sought elsewhere, first in Stockholm, Lund, and Uppsala. Roxburgh’s drawing is not a bad illustration of this species. No Roxburgh herbarium specimens have been found. His plant came from Chittagong, now in East Pakistan, where it was collected by John Roxburgh in 1810 (Hort. Beng. 75. 1814). 133. PTERIS QUADRIAURITA sensu Roxburgh, Calcutta Journ. Nat, Hist. 4:507. 1844, =Pteris quadriaurita Retz. Obs. Bot. 6:38. 1791, sens. lat. Roxburgh attributed the name P. quadriaurita to Retzius, rightly, and drew his description from a plant native in the Moluccas. The only specimen seen that is authentic for Roxburgh’s concept is in the J. E. Smith Herbarium, no 1681-13 (LINN, Morton photograph 20275, left-hand plant). This specimen was collected on Honimoa, Ceram, by Christopher Smith in 1797. The right-hand plant on this sheet may have been a part of Rox- burgh’s concept also; it is from Amboina, Christopher Smith in 1797; it is not the same as the Honimoa plant, since the lower pinnae lack the basal fork that is characteristic of plants of P. quadriaurita and its allies. The proper names of the Malaysian species of this group can hardly be determined without a detailed monographic study. 134. PTrERIS SCANDENS Roxburgh, Hort. Bengal. 75. 1814; Calcutta Journ. Nat. Hist. 4:505. 1844. —Stenochlaena palustris (Burm. f.) Underw. Bull. Torr. Bot. Club 33:38. 1906 (wrongly attributed to Beddome). Polypodium palustre Burm. f. Fl. Ind, 234. 1768. Lectotype: Ceylon, Burmann (selected by Underwood, Bull. Torr. Bot. Club 33:38, 1906). Type: Considered published in the “Hortus Bengalensis” by the citation of Rheede, Hort. Ind. Malab. 12:t. 35, which is thus the type. No Rheede specimens are presumed to exist, Roxburgh’s concept is illustrated by two specimens from the Roxburgh Herbarium (BR, Morton photographs 19841, 19842); there is also an authentic drawing at Kew (no. 1752, Morton photograph 15881). These specimens and the drawing seemingly represent the same species as the Rheede illustration. According to Roxburgh they came from India. There has been some confusion about this name, for in the “Index Filicum” it is cited under Stenochlaena palustris as though P. scandens Roxburgh were a transfer of Onoclea scandens Swartz (1806), but this is clearly wrong. Roxburgh did not cite this name of Swartz and probably did not even know of its 376 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM existence, for nowhere in his work does he refer to Swartz’ “Syn- opsis Filicum” of 1806, the place where Onoclea scandens was published. The epithet “scandens” would be a natural one for this fern and was doubtless arrived at independently. It might be thought that Pteris scandens Roxburgh is a super- fluous name for Polypodium palustre Burm. f., because the illus- tration of Rheede, its type, was also cited by Burmann f. under his Polypodium palustre. This would not be right, however, be- cause the Rheede illustration was only one of several elements in- cluded in Polypodium palustre Burm. f., and Roxburgh’s Pteris scandens is technically a segregate, a name applied to one of the elements included in Polypodium palustre and not a renaming of that species. Stenochlaena palustris has been passing with the author “(Burm. f.) Beddome, Ferns. Brit. Ind. Suppl. 26. 1876,” as in the “Index Filicum,” in Underwood’s paper cited above, and in Holttum’s paper on Stenochlaena (Gard. Bull. Str. Settl. 5:254. 1932), but I do not think this is correct. Beddome in the place cited listed the plant as “Stenochlaena palustre L. Ferns Southern India, tab. 201 (scandens).” There is no species “Stenochlaena palustre L.,” obviously, since Stenochlaena was not published un- til 1841, many years after the death of Linnaeus, nor did Lin- naeus publish this species under any genus or epithet. In the cited reference to the “Ferns of Southern India, tab. 201” one finds Stenochlaena scandens (J. Smith) and as a synonym Lomaria scandens Willd., but again no reference to Burmann or the epi- thet palustre. Therefore, it should not be assumed that Beddome really meant Stenochlaena palustris (Burm. f.) Beddome, when he wrote Stenochlaena palustre L. Diels (Nat. Pflanz. 1(4): 251. 1899) gave the authority as “Stenochlaena palustris (L.) Mett.,” but this cannot be accepted either, for Mettenius never published such a combination and again the parenthetical au- thority “L.” is incorrect. So far as I can determine, the first author to accept the name Stenochlaena palustris and cite Poly- podium palustre Burm. f. as a synonym was Underwood, whose paper appeared a few months before Christensen’s use of the same combination in the “Index Filicum.” The incorrect associa- tion of this species with Linneaus is attributable to Swartz (Syn. Fil. 112. 1806), who cited “Polypodium palustre Linn. Fl. Zeyl. p. 200” as a synonym of his Onoclea scandens. Linnaeus did describe this species in his ‘Flora Zeylanica,” page 200, as species no. 425 “Filix, fronde pinnata, pinnis lanceolato-ensiformibus integris striatis setaceo-serratis,’ but he did not use the bino- mial Polypodium palustre (he was not using binomials in this ROXBURGH’S FERN TYPES—MORTON 377 Flora) and in any case the Flora is of 1747, before the starting date for botanical nomenclature. For some reason, Linnaeus ig- nored or overlooked this species in his “Species Plantarum” and later works after 1753. 135, PTERIS SUCCULENTA Roxburgh, Calcutta Journ. Nat. Hist. 4:508. 1844. =Ceratopteris thalictroides (L.) Brongn. Bull. Soc. Philom, 1821:186. 1821. LECTOTYPE: A specimen from “Ind. or.’ collected by Roxburgh, in the J. E. Smith Herbarium, no. 1622-39 (LINN, Morton photograph 20348). A duplicate is in the same herbarium, no. 1622-38 (Morton photograph 20347). There is a presumable duplicate in the East India Company Herbarium, Kew, corresponding to Wallich, List no. 83-4. It is not indicated as collected by Roxburgh, but it agrees with the lectotype chosen above. There is no sheet indicated as 83-4, but this plant is where the Roxburgh collection might be expected, on the same sheet as 83-5 and 83-6 (Morton photograph 20664). Roxburgh did not cite a definite locality, merely “Nat. of various parts of India, in wet places.” It was from Bengal, according to Voigt (Hort. Suburb. Calcut. 736. 1845). Roxburgh had a drawing of his P. succulenta, and copies of this are at Kew, no. 1754 (Morton photograph 15882) and the British Museum, marked as from Jungholy, Bengal, collected by J. Law (Morton photograph 15756). These drawings agree with the herbarium specimen chosen as lectotype. In the original description is cited “Acrostichum thalictroides et siliquosum. Roxb. Ceratopteris thalictroides. Brongn. Wall. Cat. 61, No. 81.” These entires were obviously added by the editor Griffith, since they refer to entries in Wallich’s “Numerical List” published many years after Roxburgh’s death. The first part refers to the entry in Wallich’s Catalogue of ‘“Acrostichum thalictroides et siliquosum Herb. Roxb.” no. 83-4 (not “81” as stated by Griffith), and the “Ceratopteris thalictroides Brongn.” to the main entry in Wallich under no. 83. There is now no Roxburgh specimen in the East India Company Herbarium under 83-4, unless it is misplaced. Pteris succulenta Roxburgh cannot be considered a superfluous name, since the synonyms “Acrostichum thalictroides and siliquosum Herb. Roxb.’ mean only that the plant occurs under these names on a specimen in the Roxburgh Herbarium, i.e., sensu Roxburgh, with no indication that they are the original Acrostichum thalictroides L. and A. siliquosum L. The matter is not of any importance unless it develops that the specimen chosen as lectotype of P. succulenta could be shown to be different from the original Ceratopteris thalictroides (L.) Brongn., which is not likely. 136, PTERIS TRIPINNATIFIDA Roxburgh, Calcutta Journ. Nat. Hist. 4:508. 1844. =Histiopteris incisa (Thunb.) J. Smith, Hist. Fil. 295. 1875, sens. lat. 378 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM LECTOTYPE: A specimen from Honimoa, Ceram, no. 332, Roxburgh Her- barium (BR, Morton photograph 19907). This sheet does not bear the name P. tripinnatifida, but it does agree with Roxburgh’s description, and it is the only sheet of Roxburgh’s found that does agree. It may be presumed that the name tag has been lost and that this is truly the holotype; at least it is a suitable lectotype. Roxburgh gave the locality as merely “Moluccas,” which can now be stated more definitely as Honimoa. The collector was surely Christopher Smith, who sent his Honimoa collections to Roxburgh. Histiopteris incisa is a species of almost worldwide range as currently treated, but it may be an aggregate; however, lines on which it might be segregated are not clear. 137. PTERIS VITTATA sensu Roxburgh, Calcutta Journ. Nat. Hist. 4:504. 1844, non L. =Pteris moluccana Blume, Enum. Pl. Jav. 208. 1828. Type: Banda Island, Molucca Islands, Reinwardt (holotype L, 2 sheets, evidently part of the same plant, Morton photographs 2202, 2203). AUTHENTIC MATERIAL: A sheet in the Brussels Herbarium with the Rox- burgh number 1682 and the name Pteris spinulosa in Roxburgh’s hand, the “spinulosa” erossed out by Roxburgh and vittata substituted (Morton pho- tograph 19899). There is a drawing at Kew (no. 1105) by Roxburgh labeled Pteris vittata L., which is intended to represent the same species, as it probably does. Roxburgh did not actually mention Linnaeus in his treatment of Pteris vittata, but it may be presumed that he surely intended this, since he mentions “Osb. It. t. 4,” and the type of P. vittata L. was China, Osbeck. He gave a long careful description of P. vittata and a comparison with his own Pteris amplexicauilis. His observations are correct, but unfortunately he had the names wrong, for his P. amplexicaulis is the true P. vittata, and the plant that he identified as P. vittata was at the time an unde- scribed species that Blume later called P. moluccana. Roxburgh’s P. vittata was stated to be a “Native of the Delta of the Ganges, etc. Fructifies in the Botanic Garden most part of the year.” The plant mentioned above as authentic must have been a part of the “ete.” of Roxburgh’s localities, and it must have been cultivated in the Botanic Garden in Calcutta. It was doubtless brought or sent from the Moluccas by Christopher Smith and grown in the garden, for this species, Pteris moluccana, does not grow in the delta of the Ganges or elsewhere in India. Apparently, Roxburgh confused some plants that he had seen in the wild with the cul- tivated plant in the Botanic Garden. His description agrees with Pteris moluccana and not with any Indian species. 138. SALVINIA CUCULLATA Roxburgh ex Bory in Bélanger, Voy. Bot. 2:6. 1833. Salvinia cucullata Roxburgh ex Wallich, Num. List. no. 399. 1829, nom. nud. ROXBURGH'S FERN TYPES—MORTON 379 LECTOTYPE: A specimen in the Geneva Herbarium marked “India Orient. Dr. Roxburgh,” another label “Marsilea cyathoides, Mayo, 1276,” and a tag reading “Typus” (Morton photograph 16731). Roxburgh indicates that all three of his species of Salvinia were found “floating on lakes, or pools of sweet water, throughout Bengal.” Bory gave a short but adequate description of this species prior to the publication of the name by Roxburgh (Calcutta Journ. Nat. Hist. 4:470. 1844). In addition to material by Roxburgh, Bory cited also Calcutta, Wallich, and Hooglie, Bengal, Bélanger. There are thus three syntypes. Since the name is credited to Roxburgh, it is natural to choose the Roxburgh specimen as lectotype. I have seen the Wallich collection from Calcutta in the British Museum (Morton photograph 7721), and it is the same as Rox- burgh’s plant. I have not looked for a specimen of Bélanger’s collection, but it is presumably the same, since this is a rather characteristic species that has always been correctly understood. Griffith (Calcutta Journ. Nat. Hist. 5:255. t. XX, f. 21. 1845) gave a more detailed description of this species. 139. SALVINIA IMBRICATA Roxburgh, Calcutta Journ. Nat. Hist. 4:470. 1844. = Azolla imbricata (Roxburgh) Nakai, Bot. Mag. Tokyo 39:185. 1925. Azolla pinnata var. imbricata (Roxburgh) Bonaparte, Notes Pterid. 7:130. 1918. LECTOTYPE: A specimen in the Brussels Herbarium with the name in the hand of Roxburgh. Roxburgh indicated that his species was common through- out Bengal, and so he may have had more than one specimen in hand. Isotypes or syntypes are in Geneva (2 sheets, one Morton photograph 17029, the other with the name “Marsilea imbricata”) and in the British Museum (Morton photograph 7718). Griffith provided a very detailed description and drawings of this species in his paper “On Azolla and Salvinia” (Calcutta Journ. Nat. Hist. 5:257, t. XV-XVI. 1845) under the name Azolla pinnata. They represent actually A. imbricata. This species has sometimes been considered the same as Azolla pinnata var. africana (Desv.) Baker (Fern Allies 138. 1887), which is based on A. africana Desv. (Mém. Soc. Linn., Paris 6: 178. 1827). The proper disposition remains to be determined. 140. SALVINIA VERTICILLATA Roxburgh, Calcutta Journ. Nat. Hist. 4:469. 1844. =Salvinia natans (L.) All. Fl. Pedem. 2:289. 1785. LECTOTYPE: A specimen in the Geneva Herbarium labeled “Ind. Orient. Dr. Roxburgh” and with another label “Marsilea bengalensis, 1270” (Morton photographs 6555, 16732). Since this is the only specimen seen of a Roxburgh collection, it may well be a holotype. According to Roxburgh it occurs in lakes and pools throughout Bengal, India. 380 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM Griffith described this species in detail (Calcutta Journ. Nat. Hist. 5:254, t. XVIU-XX. 1845). If the plant of India should prove to be different from that of Europe, the name S. verticillata Roxburgh will be available. 141. SCOLOPENDRIUM LANCEOLATUM Roxburgh, Calcutta Journ. Nat. Hist. 4:501. 1844. =Polypodium pedunculatum (Hook. & Grev.) Mett. ex Salom. Nomencl. Gefasskrypt. 312. 1883. Ceterach pedunculata Hook. & Grev. Icon, Fil. 1:f. 5. 1827. Type: Sylhet, sent by Wallich and collected by “D. Smith.”* The holotype at Kew shows that it was sent by Smith, but the label does not read “D. Smith,” as Hooker and Greville did not use this form consistently. M. R. Smith, who collected ferns in Sylhet in 1812, was intended, the same Smith who sent specimens to Roxburgh. Grammitis hamiltoniana Wall. Num, List. no, 9. 1829, nom. nud, Based on Sylhet, Di Silva (E. Ind. Co. Herb.-K, Morton photograph 14636). Ceterach indivisa Hook. & Grev. ex Wall. Num. List no. 9-2, nom. nud. Based on Nibari, Nov. 26, 1808, Buchanan-Hamilton. Selliguea hookeri Presl, Tent. Pterid. 216, 1836, an illegitimate renaming of Ceterach pedunculata Hook. & Grev., and so with the same type as that. Selliguea hamiltoni K. B. Presl, op. cit., nom. illeg. Gymnogramma hamiltoniana Hook. Sp. Fil. 5:160. 1864. An illegitimate renaming of Ceterach pedunculata Hook & Grev., and so with the same type as that (although intended to be based on Grammitis hamiltoniana Wall.). TYPE: Chittagong, Roxburgh (G, Morton photographs 6554, 17027). This Roxburgh specimen from Chittagong is probably a part of the same collection as the type of Ceterach pedunculata, for the locality Sylhet was referred to by Roxburgh as Chittagong, a sort of general name for this district, and many of Roxburgh’s specimens from Chittagong were sent him by M. R. Smith; how- ever, this cannot be proved. In any case S. lanceolatum Roxburgh is legitimate, based on a different collection and not on the holo- type of Ceterach pedunculata Hook. & Grev., although the two are doubtless taxonomic synonyms. This species, known as Colysis pedunculata (Hook. & Grev.) Ching by those who segregate Polypodium, appears to be some- what variable. The Hooker and Greville type has the sterile blades broad and abruptly narrowed at the base, with the stipe slightly alate; the fertile fronds are small and exceed the sterile. The type of Scolopendrium lanceolatum has narrower blades that are gradually long-decurrent at the base, with the stipe alate; the fertile blades are shorter than the sterile. However, these dif- ferences are such as might occur within the same colony. On the ‘Dr. Jarrett has pointed out that the “D.” stands for “Dominus” (Master). ROXBURGH’S FERN TYPES—MORTON 381 other hand, there may be some differences in the venation, that of Ceterach pedunculata is a little simpler in spite of the broader blades; in S. /anceolatum there are about four secondary areoles between the main lateral veins, each of these with one or two free included veinlets. The type of S. lanceolatum is closely matched by the following collection: middle elevation of Khao Chong, Pukat, Thailand, 600-1100 m., Jan. 27, 1966, Tagawa, Iwatsuki & Fukuoka 6805 (US), and there are several quite similar specimens at Kew. 142, TRICHOMANES CAMPANULATUM Roxburgh, Calcutta Journ. Nat. Hist. 4:518. 1844. Hymenophyllum campanulatum Wallich, Num. List. 66, no. 2199. 1830, nom, nud, Didymoglossum plicatum van den Bosch, Ned. Kruid. Arch. 5(38) :139. 1863. Syntypes: Malacca, Malaya, Griffith (K, Morton photograph 19028); Ceylon, Thwaites C. P. 2985 (K, not seen); and Sumatra, Teysman (not seen). The first is designated lectotype. Trichomanes plicatum Bedd. Ferns Brit. Ind. t. 285, 1868. LECTOTYPE: A specimen in the Brussels Herbarium with the name in the hand of Roxburgh (Morton photograph 19846). Roxburgh’s descrip- tion states that it was collected in Chittagong [East Bengal, now East Pakistan] by Buchanan-Hamilton. Isotypes are in the British Museum (Morton photograph 6576), in Geneva (Morton photograph 6550), and in Kew (East Ind. Co. Herb. 2199, Morton photograph 15746). All these represent the same species and seem clearly a part of the same collection. In the “Index Filicum,” Trichomanes campanulatum is listed as a dubious species. It was overlooked by Copeland in his mono- graph and by Holttum in his “Ferns of Malaya.” It is a rather distinctive species, clearly the same as T. plicatum (van den Bosch) Beddome, as treated recently by Sledge in his “The Hym- enophyllaceae of Ceylon” (Journ. Linn. Soc. Bot. 60:289-308. 1968). The species has passed generally as T. latealatum (van den Bosch) Christ, but the type of that (East Indies, Griffith, K, Morton photograph 19027) shows that T. latealatum does not have the densely pubescent stipe wing that is characteristic of T. campanulatum. 143. TRICHOMANES CARUIFOLIUM Roxburgh, Calcutta Journ. Nat. Hist. 4:519. 1844. —Trichomanes obscurum Blume, Enum. PI. Jav. 2:227, 1828. Lectotype: “Ad terram Buitenzorg,” Java, Zippel (L, Morton photo- graph 2457). Blume indicated for his material: “Crescit in sylvis montanis humidioribus Javae et Moluccarum.” The only specimen from Java with the name in Blume’s hand is the one selected as lectotype. The syntypes from the Molucca Islands are represented by two specimens collected by Reinwardt: one is indicated as “Sylvae elatiores montis Tidore,” and the other has had a “Type” tag affixed and is a possible lectotype; but it is better to choose the one from 382 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM Java. The holotype of T. obscurum var. adnatum Blume (loc. cit) is Java, “crescit ad ripas fluvirum in sylvis Javae occidentalis,” Blume, according to the original description (L, Morton photograph 2430). The holotype of 7. obscurum var. obtusiusculum Blume (loc. cit) is Java, “crescit in Javae montis Salak,” Blume, according to the origi- nal description (L, Morton photograph 2429). These varieties of Blume do not appear to differ significantly from the typical variety. TYPE: A specimen in Brussels with the name T. caruifolium in the hand of Roxburgh and the number 2430 (Morton photograph 19848, right-hand plant). Since this is the only specimen seen in the herbaria at Kew, British Museum, Geneva, and Brussels, it may be considered a holotype. According to Roxburgh, it came from Prince of Wales Island, i.e., Penang Island, Malaya. In the “Index Filicum,” 7. caruifolium is listed as a dubious species, and the name is ignored by Copeland in his monograph of Trichomanes and by Holttum in his “Ferns of Malaya.” The type represents a small but typical plant of T. obscurum Blume, which is common in Malaya. 144, TRICHOMANES LACINIATUM Roxburgh, Calcutta Journ. Nat. Hist. 4:518. 1844. Trichomanes asplenioides K. B. Pres], Hymen. 37. 1848, non Swartz, 1788. Type: Philippine Islands, Cuming 184 (isotype US). Cephalomanes asplenioides K. B. Presl, Abhandl. Boehm. Gesell. Wiss. V, 5:334, 1848. To be considered a new name for T. asplenioides K. B. Presl, non Swartz. Cephalomanes asplenioides K. B. Presl, Abhandl. Boehm. Gesell. Wiss. Type: Philippine Islands, Cuming 169 p.p. (holotype PRC, photograph by Holttum US; isotype L, Morton photograph 2420; other isotypes reported by Holttum in K and BM). The specimen at Leiden gives the locality specifically as South Camarines. Trichomanes preslii Morton, Contr. U. S. Nat. Herb. 38:190. 1968. Based on Trichomanes asplenioides K. B. Presl, non Swartz, 1788. LECTOTYPE: A specimen in Brussels mounted on the same sheet as the type of Trichomanes caruifolium Roxb. (Morton photograph 19848). This sheet contains two species of Trichomanes, one of which agrees with the description of T. caruifolium and the other with the description of T. laciniatum. Since no specimen has been found labeled T. laciniatum, this is likely the holotype and the only material existing. According to Roxburgh, it was collected in the Molucca Islands, and doubtless was received by Roxburgh from Christopher Smith, who probably collected it either in Amboina, Ternate, or Honimoa. Amboina is likely, since most of Smith’s collections came from there, and since there is a specimen from Amboina (Robinson 1964, US) that agrees with the lectotype in every way. The lectotype has the involucres apical as described by Rox- burgh, and it therefore agrees with Trichomanes asplenioides K. B. Pres] as treated by Copeland in his monograph (Phil. Journ. Sci. 51:249, 1933). Although this species is mostly con- ROXBURGH’S FERN TYPES—-MORTON 383 fined to the Philippine Islands, Copeland did refer Robinson 1964 to it. Trichomanes javanicum Blume is supposed to differ in having the involucres borne along the distal margin rather than apically, but whether this is a true difference remains to be determined. Since 7. asplenioides K. B. Pres] (1843) is an illegitimate later homonym, the earliest name for the species is T. laciniatum Roxb., if it proves different from T. javanicum. The new name Trichomanes preslii Morton, proposed for T. asplenioides K. B. Presl, not Swartz, now proves to be a taxo- nomic synonym of 7. laciniatum. Another name that I did not consider when proposing T. preslii is Cephalomanes oblongifolium K. B. Presl. Pres! stated that his C. oblongifolium differed from his C. asplenioides in fronds, pinnae, and sori, but he did not state the differences; from the isotype specimens examined, and the photograph of the holo- type of C. oblongifolium, it does not appear that two species can be distinguished, and therefore C. oblongifoliwm also becomes a taxonomic synonym of T. laciniatum Roxb. The type of C. oblongifolium is Cuming 169 in part, the other part being the type of Cephalomanes atrovirens K. B. Presl, which is also close to C. asplenioides, but which is kept distinct by Copeland. 145. TRICHOMANES LUCIDUM Roxburgh, Calcutta Journ, Nat. Hist. 4:519. 1844. —Davallia denticulata (Burm. f.) Mett. ex Kuhn, Fil. von Deck. Reise 27. 1867. Adiantum denticulatum Burm. f. Fl. Ind. 236. 1768. Type: Java, Bur- mann (holotype G, Morton photograph 16909). Lectotype: A Roxburgh specimen in the Brussels Herbarium with the name in the hand of Roxburgh and the number 2431. According to Roxburgh, his plant came from the Prince of Wales Island, ie., Penang Island, Malaya, where it was collected by W. Hunter. There is an isotype in the Geneva Herbarium from “India Orient. Dr. Roxburgh” and a label “Typus” (Morton photographs 6558, 16908). Another isotype is in the East India Company Herbarium, no. 253-3 (Morton photograph 15733, the small plant in the upper right corner). The published account gives as a synonym “Davallia elegans Willd. Wall. Cat. 64, no. 253,” a citation added by the editor Griffith, which was intended to mean D. elegans sensu Wallich as to page 64, no. 253 (actually no, 253-38, which is the Roxburgh collection), and not that D. elegans Willd. is a synonym; actually there is no species D. elegans Willd., only D. elegans sensu Willd. (Sp. Pl. ed. 4, 5:471. 1810), for the account of Willdenow is based on three cited synonyms and the epithet elegans derives from Swartz. 384 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM The type of Adiantum denticulatum Burm. f. in Geneva has the involucres rather short and broad and the lateral teeth rather blunt; the Roxburgh plant has the teeth sharper and the false veins more obvious, but it is doubtless conspecific with the Javan D. denticulata. The lectotype of Roxburgh’s species has been identified as D. elegans var. bidentata Hook. There is in Brussels a second sheet of 7. lucidum that has much broader segments that are only slightly toothed: it is doubtless a part of the same collection, for this species is subdimorphic, the sterile blades having broader segments. 146. TRICHOMANES MALAYANUM Roxburgh, Calcutta Journ. Nat. Hist. 4:519. 1844. =Sphenomeris chinensis (L.) Maxon, Contr. U. S. Nat. Herb. 17:159. 1918. Type: China, Osbeck (S). Concerning the nomenclature of this species, see F, R. Fosberg (Taxon 18:596. 1969). The species has sometimes been wrongly called Sphenomeris chusana (L.) Copel. LecToTyPe: Malay Islands, Roxburgh (G, Morton photograph 6552). This specimen is determined as “Trichomanes sinense Roxb.,” but not by Rox- burgh, and this may be assumed to be a reidentification, an error for T. chinense L., which this specimen does represent. Roxburgh may have intended to withdraw his T. malayanum in favor of T. chinense, but Griffith published the species anyway as T. malayanum. This lectotype agrees with Roxburgh’s description, and so it is a suitable lectotype. There is a specimen in the British Museum collected in Ternate by Christopher Smith that is labeled T. malayanum Roxburgh, but not by Roxburgh. It also represents S. chinensis, but is a small plant, hardly more than 25 cm. high, whereas Roxburgh described his species as four to five feet high. The lectotype in Geneva is a large plant that agrees better with this description. 147. VITTARIA DIVERGENS Roxburgh, Calcutta Journ. Nat. Hist. 4:510. 1844. =Lindsaea divergens Hook. & Grev. Icon. Fil. 2: ¢. 226. 1831. Type: “India orientali, Herb. Roxb.” ex Wallich (presumably K, not seen). There ought to be an isotype in the East India Company Herbarium under no. 2191, but it is not there now; perhaps it has been misplaced. The specific locality was not known to Hooker and Greville; it is “Prince of Wales Island,” i.e., Penang Island, where it was collected by W. Roxburgh, Jr. Vittaria divergens Roxburgh ex Wallich, Num. List. no. 2191. 18380, nom. nud. Tsoloma divergens (Hook. & Grev.) J. Smith, Journ. Bot. Hook. 3:414. 1841. Schizoloma divergens (Hook. & Grev.) Kuhn, Chaetopt. 346. 1882. TYPE: Considered the same as the type of Lindsaea divergens Hook. & Grev., which is based on specimens of the same species sent to Wallich. No specimen in the Roxburgh collection in Brussels was located. ROXBURGH’S FERN TYPES—MORTON 885 According to Holttum (Ferns of Malaya, ed. 2. 337. 1966), this is a peculiar species found only in Malaya and Borneo. Its characters seem to be well shown in the Hooker and Greville plate. 148. VITTARIA INTERRUPTA Roxburgh, Calcutta Journ. Nat. Hist. 4:511. 1844. =Lindsaea interrupta (Roxburgh) Morton, comb. nov. Lindsaea interrupta Wallich, Num. List no. 2195. 130, nom. nud. Lindsaea cambodgensis Christ, Notul. Syst. 1:58. 1909. Type: Cambodia, Bouillod 61 (P, 2 sheets, photograph Kramer, US). LECTOTYPE: A specimen in the Brussels Herbarium first labeled “Pteris” by Roxburgh, this crossed out and Vittaria substituted (Morton photograph 19912). That this is authentic material and probably the actual holotype is likely. Of the eight species of “Vittaria” and Lindsaea described by Rox- burgh, this is the only Roxburgh collection that agrees with the description of V. interrupta in having the sori “interrupted” by the breaks in the margin, described by Roxburgh as “gash-dentate.” The other Roxburgh species of “Vittaria” have the margins entire and the sori continuous. That this is the type is further indicated by the handwriting of W. Roxburgh, Jr., on the label reading: “Grows on the ground in shady cool places. The roots are sent.” This was quoted directly by Roxburgh in the original description of V. interrupta as: “Found by Mr. W. Roxburgh, growing on the ground, in shady cool places on Prince of Wales’ Island.” 149. VITTARIA LINEATA sensu Roxburgh, Calcutta Journ. Nat. Hist. 4:509. 1844, non Swartz. =Vittaria ensiformis Swartz, Gesell. Naturf. Freunde Berlin Neu. Schr. 2:1384, t. 7, f. 1. 1799, at least as treated by Holttum (Ferns of Malaya 618. 1954). AUTHENTIC MATERIAL: A Roxburgh drawing, no. 1755, at Kew (Morton photograph 15884). No herbarium specimens have been noted, but not all herbaria were searched for this species. Roxburgh gave the locality as “Prince of Wales Island, from whence introduced into the Botanic Garden [Calcutta] by Mr. W. Roxburgh, Jun.” From the description, drawing, and locality, Roxburgh’s plant could only be Vittaria ensiformis Swartz or V. elongata Swartz, as treated by Holttum in the “Ferns of Malaya.” In the shape of the fronds and small size it agrees best with V. ensiformis. Whether this is the true V. ensiformis, the type of which is from the Mascarene Islands, remains to be determined. The American species Vittaria lineata has usually been attrib- uted to J. E. Smith in the original publication of the genus Vittaria, in 1793. But Smith merely indicated that Pteris lin- eata L. belonged in his genus and was in fact the only species in his genus, but he did not formally make the combination V. lineata (L.) J. E. Smith, as cited in the “Index Filicum” and elsewhere. I have not searched all the literature between 1793 and 1806, but it appears that the proper authority is V. lineata (L.) Swartz, Syn. Fil. 109. 1806. 386 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM 150. VITTARIA LUNULATA Roxburgh, Calcutta Journ. Nat. Hist. 4:510. 1844, =Lindsaea parasitica (Roxburgh) Hieron. Hedwigia 62:14. 1920 (simply pinnate form). Lindsaea scandens Hook. Sp. Fil. 1:205. 1846. One of the syntypes is from Penang Island, Dalhousie. This represents the same simply pinnate form as V. lunulata. Type: A specimen in the Brussels Herbarium with the name lunulata in Roxburgh’s hand, the number 2121, and a label “Typus” (Morton photo- graphs 5151, 19918). According to Roxburgh it was from Prince of Wales Island, i.e., Penang Island, where it was probably collected by W. Roxburgh, Jr. The above identification follows the identification of K. Kramer, who wrote on the label: “There is no proof that this is a type.” However, the evidence is clear that this is actually the holotype, for it is from Roxburgh’s personal herbarium, first in the custody of the Linnean Society, London, purchased in 1863 by Martius, and later purchased by Brussels; the specimen agrees with the original description and is the only Roxburgh specimen found that does, and the name V. luwnulata is in Roxburgh’s own hand. 151. VITTARIA PARASITICA Roxburgh, Calcutta Journ. Nat. Hist. 4:510. 1844. =Lindsaea parasitica (Roxburgh) Hieron. Hedwigia 62:14. 1920. Lindsaea parasitica Wallich, Num. List. no. 2196. 1830, nom. nud. Based on Vittaria parasitica Roxburgh, in 1830, also a nom. nud. LEcTOTYPE: A Roxburgh specimen in the Brussels Herbarium with a label reading in Roxburgh’s hand “between Lindsaea & Pteris,” which was crossed out and “Vittaria” written in, and in the hand of Roxburgh’s Jr.: “This is growing on part of the stem of a small tree I cut down. It grows for... 4-6 feet in cool shady places” (Morton photographs 5153, 19915). In the above quotation one word is illegible, but that is not important. This speci- men does not bear the name V. parasitica, but it is surely authentic. There is a duplicate of this lectotype in Geneva marked “Prince of Wales Island, Dr. Roxburgh” (Morton photograph 6568). Of Roxburgh’s species of “Vittaria,” this is the only one that agrees with the original description and the only one that is indicated in the description as being “parasitic’’—the term used in the early days for “epiphytic’—an unusual character in Lind- saea, most of the species of which are strictly terrestrial. That this character struck Roxburgh too is indicated by his choosing the specific epithet “parasitica.” Roxburgh obtained his informa- tion on the habitat from the label of Roxburg, Jr., stating that he had collected the specimen 4—6 feet up on a small tree. From my general survey of Roxburgh’s species, it seems that most of the species described definitely or probably from “Prince of Wales Island” (Penang Island) were collected by W. Roxburgh, Jr. There is no specimen in the East India Company Herbarium un- ROXBURGH’S FERN TYPES—MORTON 387 der no. 2196, which is listed by Wallich as “Lindsaea parasitica Herb. Roxb. (sub Vittaria),” and it is likely that Wallich did not have a specimen, but saw this Brussels specimen, which was at the time in the Roxburgh Herbarium in the Linnean Society, London. Lindsaea parasitica Wallich (Num. List. no. 2196. 1830) was a nomen nudum and based on Vittaria parasitica Roxburgh, which was also in 1830 a nomen nudum. Therefore Wallich’s name cannot be correct. The first worker to take up and accept the name Lindsaea parasitica after the publication of Vittaria parasitica Roxburgh in 1844 was apparently Hieronymus in the publication cited above; however, Hieronymus wrongly gave Wal- lich as the authority for the combination. He did not mention Roxburgh, but his combination is indirectly connected with Roxburgh’s Vittaria parasitica through Wallich’s citation under his no. 2196. The valid combination L. parasitica (Roxburgh) Hieron. has been overlooked in the “Index Filicum.” The ques- tion of the proper authority was discussed by Kramer (Blumea 15:570. 1967 [1968]), who came to the same conclusion. 152. VITTARIA RESECTA Roxburgh, Calcutta Journ. Nat. Hist. 4:510. 1844. =Lindsaea javensis Blume, Enum. Pl. Jav. 219. 1829. TYPE: No specimens have been found named V. resecta by Roxburgh nor any herbarium specimens of Roxburgh’s collections agreeing with the original description. Roxburgh’s plant came from Chittagong, East Bengal, now East Pakistan. Unless Roxburgh herbarium specimens can be found agree- ing with the original description, this species can be identified only from the description. Kramer (Gard. Bull. Singapore 26(1): 47. 1972) was unable to place the name definitely. The original description calls for a small plant only 12.5-25 cm. long, simply pinnate, with the pinnae subtrapeziform, obtuse, and the sorus in a continuous line on the anterior margin and around the apex. Index (Synonyms in italics. New species, new names, and new combinations in boldface. Page numbers of principal entries in italics.) Abacopteris urophylla, 340 Alsophila Acrostichum, 284, 286 batjanensis, 337 alatum, 283, 284 concinna, 344 aureum, 284, 316 latebrosa australe, 385 var. batjanensis, 337 dichotomum, 285 polyphlebia, 344 emarginatum, 284 sangirensis, 844 furcatum, 341 Ampelopteris prolifera, 356 hastatum, 317 Ananthacorus angustifolius, 367 lanceolatum, 353 Angiopteris, 329 punctatum, 347 crassipes, 329 radiatum, 284 dregeana, 329 seetacoonense, 285 javanica, 329 semipinnatum, 286, 360 pinnata, 329, 330 siliquosum, 377 ruttenii, 329, 330 thalictroides, 377 subg. Angiopteris, 329 trinerve, 317 subg. Pseudangiopteris, 329, 330 viviparum, 285 Antrophyum, 318 Actiniopteris radiata, 284, 285 callifolium, 317, 318 Adiantum, 287, 371 grevillei, 318 capillus-veneris, 288 lessonii, 319 caudatum, 286—288 reticulatum, 318, 319 var. assamicum, 287 semicostatum, 318 var. flabellatum, 287 Arachniodes, 336 cultratum, 322 Arcypteris, 338, 351 denticulatum, 383, 384 irregularis, 337, 351 flagelliferum, 287, 288 macrodonta, 338 incisum, 287, 288 Aspidium indicum, 286—288 aridum, 885 lunulatum, 370, 371 difforme, 337, 351 microphyllum, 287 ferox, 359 philippense, 371 fuscipes, 356 proliferum, 287 heterocarpon, 349 repens, 307 leuzeanum, 343 tenerum, 288 maingayi, 286 venustum, 287 pachyphyllum, 349 389 Aspidium—Continued parasiticum, 852 singaporianum, 354, 355 squarrosum, 362 Asplenium, 284, 291, 293, 300, 301 acuminatum, 294, 295 artfolium, 316 bipinnatum, 288, 289 blumei, 289 camptorhachis, 299 cataractarum, 301 cicutarium, 289 confusum, 300 coriaceum, 290, 291 crenatoserratum, 300 erenatum, 291, 298 cultratum, 291, 292 cultrifolium, 291, 292 diversifolium, 302 finlaysonianum, 290 frondosum, 802 hemionitis, 289 hemionitoides, 292 inaequilaterale, 298, 299 laserpitiifolium, 300 latifolium, 801 linguiforme, 293, 294 lunulatum, 298 macrophyllum, 290 maximum, 301, 302 mixtum, 294, 295 monanthemoides, 296 monanthemum, 298 multiflorum, 296 multisoratum, 801 multisorum, 801 neolaserpitiifolium, 300 nidus, 297 normale, 296 penangianum, 308 phanerotis, 801 polymorphum, 802 polyodon, 291, 292, 301 polypodioides var. vestitum, 302 porrectum, 292, 301 praemorsum, 290 prescottianum, 294, 295 protensum, 3801 pseudolaserpitiifolium, 300 radiatum, 284, 285 reticulatum, 297 robustum, 300 Asplenium—Continued serrulatum, 291, 297, 298 sublaserpitiifolium, 300 tenuifolium, 289 trapeziforme, 298, 299 tripinnatum, 299, 300 unilaterale, 298, 299 varium, 300, 301 woodwardioides, 301, 802 Athyrium, 290 prescottianum, 295 spectabile, 290 Azolla africana, 379 imbricata, 379 pinnata, 379 var. africana, 879 var. imbricata, 379 Blechnum aggregatum, 284 angustifolium, 303 decurrens, 308, 304 finlaysonianum, 303, 304 glabrum, 304 moluccanum, 804, 805 orientale, 304, 305 Bolbitis appendiculata, 283 subsp. vivipara, 285 Callipteris, 291 ambigua, 289 Campylogramme trollii, 286 Cephalomanes asplenioides, 382, 383 atrovirens, 388 oblongifolium, 383 Ceratopteris thalictroides, 377 Ceterach, 294 indivisa, 380 pedunculata, 294, 380, 381 Cheilanthes, 368 dealbata, 867 farinosa, 367 tenuifolia, 368, 369 Colysis pedunculata, 380 Coniogramme, 300 fraxinea, 300 Crypsinus, 858 taentatus, 358 Ctenitis rhodolepis, 338 Ctenitopsis fuscipes, 356 Ctenopteris alata, 307 390 Cyathea, 342 alternans, 348, 349 batjanensis, 336 biformis, 345 bipinnatifida, 306 excelsa, 306 felinum, 344 moluceana, 305, 348 pinnata, 305 ridleyi, 348 sangirensis, 345 squamulata, 348 tripinnatifida, 305 Cyclophorus, 817 lanceolatus, 354 ferox, 359 flocculosus, 364 heterocarpus, 350 latipinna, 361 subpubescens, 360, 361 sumatranus, 861 vestigiatus, 353 Cystodium sorbifolium, 315 Davallia, 307-313 alata, 307, 308 alpina, 807 amboynensis, 308, 309 angustata, 306 augustifolia, 306 biserrata, 3812 cordifolia, 306, 307 denticulata, 383, 384 elegans, 388 var. bidentata, 384 flagellifera, 312 gaimardiana, 311 hirta, 343 longifolia, 307, 308 moluecana, 308, 309 multiflora, 309, 310 parallela, 309, 311 pectinata, 309, 310, 3811 pilosa, 311, 312 pilosiuscula, 318, 314 polypodioides, 314, 342 var. hirta, 3438 var. pubescens, 314 rhomboidea, 315 roxburghii, 342 serrata, 307, 312 speluncae, 311 trapeziformis, 313-815 Dicksonia, 842 moluecana, 315, 316 polypodioides, 342 sorbifolia, 315 Dicranopteris linearis var. montana, 341 Dictyopteris macrodonta, 338 Didymoglossum plicatum, 381 Diplazium, 290, 292, 293, 300 accedens, 291, 297, 298 alismifolium, 293, 294 asperum, 302 burchardii, 292 crenatoserratum, 3800 crenatum, 291 dilatatum, 296, 301, 302 esculentum, 288, 289 malaccense, 296 maximum, 3801, 302 var. vestitum, 302, 303 mixtum, 294-296 multiflorum, 296 phanerotis, 300, 301 polypodioides, 302 prescottianum, 294, 295 proliferum, 291 repandum, 291 roxburghii, 300, 301 serrulatum, 298 silvaticum, 295 var. prescottianum, 294 tomentosum, 292, 2938 varium, 300 Doryopteris ludens, 371, 372 Drymoglossum, 3875 piloselloides, 374 Drynaria quercifolia, 357 Dryopteris, 336, 349 arida, 360 ferox, 359 heterocarpa, 350 prolifera, 356 pseudocalcarata, 363 rhodolepis, 339 sericea, 368 subalpina, 335 sect. Cyclosorus, 359 Egenolfia, 285, 286 nodiflora, 285 Equisetum debile, 316 391 Equisetum—Continued ramosissimum subsp. debile, 316 Gleichenia, 347 glauca, 347 laevigata, 346 truncata, 346 Goniopteris dalhousiana, 340 Grammitis alata, 307 avenia, 297 coriacea, 297 hamiltoniana, 880 macrophylla, 297 sect. Cryptosorus, 308 Gymnogramma, 286 hamiltoniana, 380 javanica, 300 maingayi, 286 serrulata, 300 Helminthostachys zeylanica, 334 Hemionitis, 293, 316 arifolia, 316, 317 cordata, 316 cordifolia, 317 hastata, 817 prolifera, 356, 857 reticulata, 317, 318 Hemitelia, 349 alternais, 348 Histiopteris incisa, 377, 378 Holttumiella, 804 Humata, 309 alpina, 307 angustata, 306 gaimardiana, 309 heterophylla, 306 lepida, 306 parallela, 310 pectinata, 310 repens, 807 serrata, 307 trifoliata, 306, 307 Hydroglossum scandens, 311 Hymenophyllum campanulatum, 381 Hypodematium, 336 crenatum, 336 Isoétes, 319, 320 capsularis, 319 coromandelina, 319, $20 indica, 320 Tsoloma divergens, 384 Lastrea calearata var. sericea, 363 sericea, 363, 364 Leptochilus, 334 decurrens, 334 latifolius, 286 Lindsaea, 320-822, 385, 386 bipinnata, 320-322 cambodgensis, 385 cultrata, 322 divergens, 384 interrupta, 385 javensis, 387 odorata, 322 parasitica, 820, 386, 387 scandens, 386 var. terrestris, 321 Lomaria alpina, 283 scandens, 376 Lomariopsis cochinchinensis, 283 Loxogramme avenia, 297 blumeana, 297 scolopendrina, 297 Lycopodium, 327 aristatum, 322, 323 atroviride, 324, 325 bryopteris, 325 carinatum, 327 cernuum, 323 elegans, 325 fiiforme, 323, 3824 furcatum, 324 hymenophyllum, 324 imbricatum, 325 intermedium, 324 laevigatum, 325, 326 mimosoides, 325, 326 nummulariifolium, 328, 329 obtusum, 327 pectinatum, 326 pendulum, 327 phlegmaria, 324, 327 var. gracilescens, 324 var. laxum, 324 var. pellucidum, 324 plumosum, 327, 828 rotundifolium, 324, 328, 329 roxburghii, 328 392 Lycopodium—Continued Nephrodium—Continued salvinioides, 323 rhodolepis, 338, 339 wallichii, 325 singaporianum, 354 Lygodium truncatum, 860 circinnatum, 332, 333 Nephrolepis, 309, 310, 358 flexuosum, 332, 333 acuta, 346 microphyllum, 331 biserrata, 346 salicifolium, 332-334 exaltata, 309, 310, 346 scandens, 331, 332 gaimardiana, 311 venustum, 332 hirsutula, 309, 310, 345, 346 volubile, 332, 333 multiflora, 309, 310 Macrothelypteris, 365 radicans, 358 torresiana, 365 rufescens, 346 Marattia pinnata, 329 Niphobolus Marsilea pertusus, 354 bengalensis, 379 sphaerocephalus, 339 coromandelica, 330 Onoclea scandens, 375, 376 cyathoides, 379 Onychium, 368 dentata, 330 siliculosum, 368 imbricata, 379 Ophioglossum quadrifolia, 330 cordifolium, 330, 331 Mertensia filiforme, 331, 333 laevigata, 346 flexuosum, 332, 333 truncata, 346 furcatum, 332 Microlepia, 342 pedunculosum, 331 dubia, 343 petiolatum, 330 var. subglabra, 343 scandens, 331-333 firma, 342 vulgatum, 330 var. hirta, 343 Osmunda var. subglabra, 343 lanceolata, 334 pilosiuscula, 313, 314 zeylanica, 334 rhomboidea, 315 Paesia, 321 speluncae, 312, 314, 315, 342 Phegopteris, 341 var. pubescens, 312-314 repanda, 340 var. speluncae, 311 subdecurrens, 286 trapeziformis, 3138-315, 342 urophylla, 340 Microsorium, 348 Phymatodes cuspidatum, 850 banerjiana, 355, 3856 lucidum, 350 cuspidata, 350 Neocheiropteris phyllomanes, 354 leiorhiza, 350 Neolepisorus phyllomanes, 354 lucida, 350 Nephrodium Platytaenia, 304 arbuscula, 360 Pleocnemia, 3438 extensum, 3386 leuzeana, 3438 gaimardianum, 311 Pleopeltis giganteum, 343 leiorhizon, 350 intermedium, 339 macrocarpa, 337 molle, 353 Polybotrya var. glabratum, 352 nodiflora, 285 var. latipinna, 361 vivipara, 285 moulmeinense, 352 Polypodium, 286, 312, 318, 341, 342, oligophlebium, 365 349, 354, 355, 358, 362, 370, 380 393 Poly podium—Continued acuminatum, 335, 362 acutum, 335, 336 adnascens, 858, 354 aemulum, 336 alternans, 348 alternifolium, 355, 356 angustatum, 339 arborescens, 336 arboreum, 337 asperum, 840 attenuatum, 337 ciliatum, 292, 311, 312 confertum, 337, 351 confluens, 338, 339 coriaceum, 339, 340 crenatum, 336 cultratum, 835 cuspidatum, 340, 341, 850, 351 davallioides, 809 dentatum, 352 detergibile, 364 dichotomum, 341 difforme, 338 dubium, 814, 342 elatum, 343 excavatum, 343, 344, 358 eximium, 338 feei, 298, 294 felinum, 344 ferrugineum, 345 flagelliferum, 346 flocculosum, 364 furcatum, 341, 346, 347 glabrum, 347, 348 heterosorum, 286 tmpuber, 348, 349 involucratum, 349 irioides, 347 irregulare, 337 lanceolatum, 337 leiorhizum, 350 leuzeanum, 343 lingulatum, 347 longifolium, 349 longissimum, 344, 355 lucidum, 350 macrodon, 338 mucronatum, 351 multiflorum, 351 multilineatum, 352 mysurense, 364 Polypodium—Continued nigrescens, 355 nudatum, 352 palmatum, 358 palustre, 375, 376 parasiticum, 352, 353 pedunculatum, 380 pertusum, 358, 354 phyllitidis, 354 phyllomanes, 354, 355 pilosum, 356 pinwillii, 340 polycarpon, 347, 348 proliferum, 356, 357 punctatum, 347, 348 radicans, 857, 358 rupestre, 358 saccatum, 3138, 314 seabrum, 358, 859 scariosum, 359 scolopendrinum, 297 semipinnatum, 286, 360 semisagittatum, 360 sophoroides, 358, 361, 362 sphaerocephalum, 339 spissum, 854 squarrosum, 362 taeniatum, 358 tenerum, 363, 864 tomentosum, 364 tridentatum, 365 unitum, 365, 366 urophyllum, 340, 341 vestitum, 364 Polystichum, 3386, 351 aculeatum, 368 lobatum, 362, 363 prolificans, 359 scariosum, 339 setiferum, 363 squarrosum, 362, 363 torresianum, 365 Pronephrium nudatum, 852 repandum, 841 Prosaptia, 308 alata, 307 Pseudophegopteris, 365 Pteridium, 321 394 phymatodes, 348, 344, 355 quercifolium, 344, 357, 358 scolopendria, 848, 344, 855, 370 Pteris, 885, 386, 870-378 amplexicaulis, 366, 378 angustifolia, 366, 367 asperula, 373 attenuata, 374 biaurita, 373 bicolor, 367, 368 blechnoides, 304 buchananii, 374 daucifolia, 368 dimidiata, 368 ensiformis, 369 gracilis, 369 graminifolia, 369 hirtula, 373 intermedia, 374 linearis, 370 lineata, 321, 385 lobata, 370 longipinnula, 372, 373 var. hirtula, 373 ludens, 871, 372 lunulata, 370, 3871 megaphylla, 373 moluccana, 378 multifida, 871 parasitica, 366 pectinata, 372, 3738 pedatifida, 373, 3874 piloselloides, 374 quadriaurita, 375 reducta, 373 salakensis, 373 scandens, 375, 876 semipinnata, 368, 369 spinulosa, 378 succulenta, 377 tripartita, 370, 373 var. dissoluta, 374 tripinnatifida, 877, 378 vittata, 366, 370, 378 Pyrrosia adnascens, 354 angustata, 339 flocculosa, 364 lanceolata, 353, 354 mollis, 364 Saccoloma, 309 Salvinia, 379 cucullata, 378 imbricata, 379 Salvinia—Continued natans, 379 verticillata, 379, 380 Schizoloma, 804 divergens, 384 Selaginella, 324, 326 atrovirens, 325 atroviridis, 324, 325 bryopteris, 325 concinna, 327 intermedia, 324, 825 laevigata, 325 obtusa, 327 pectinata, 325 plana, 325, 326 plumosa, 328 pouzolziana, 326 roxburghii, 327, 328 stolonifera, 328 tamariscina, 322, 323 uncinata, 323 wallichii, 325, 326 willdenovii, 326 Selliguea feei, 294 hamiltoni, 380 hookeri, 380 Sphenomeris chinensis, 384 chusana, 384 Stachygynandrum obtusum, 827 tamariscinum, 322 Stenosemia aurita, 344 Stenochlaena, 376 palustris, 375, 376 sorbifolia, 283 Sticherus truncatus, 346 Syngramma, 293, 294, 304 alismifolia, 293, 294 Taenitis, 304 blechnoides, 3038, 304 f. angustifolia, 303 Tapeinidium, 308, 309, 312 amboynensis, 308 moluccanum, 308 pinnatum, 3812 Tectaria, 286, 338, 348, 355 crenata, 349 fuscipes, 356 395 Scolopendrium lanceolatum, 380, 881 Tectaria—Continued Trichomanes—Continued irregularis, 388 asplenioides, 382, 883 leuzeana, 348 campanulatum, 381 maingay?, 286 caruifolium, 381, 882 semipinnata, 286, 360 chinense, 884 singaporiana, 355 javanicum, 383 Thelypteris, 364, 365 laciniatum, 382, 383 acuminata, 862 latealatum, 381 arbuscula, 360 lucidum, 383, 384 arida, 335 malayanum, 384 cana, 364 obscurum, 381, 382 contigua, 3538 var. adnatum, 382 dentata, 352, 3538, 361, 362 var. obtusiusculum, 382 ferox, 358 plicatum, 381 heterocarpa, 349, 350 preslii, 382, 883 multilineata, 352 sinense, 384 nudata, 352 Ugena microphylla, 331 papilio, 360 Urostachys salvinioides, 323 prolifera, 356, 357 Vallisneria, 319 repanda, 340 spiralis, 319 semisagittata, 360 Vittaria, 320, 321, 366, 367, 369, sericea, 363 385-887 setigera, 365 amboinensis, 367 subalpina, 335 divergens, 384 sumatrana, 361, 362 elongata, 369, 385 torresiana, 365 ensiformis, 385 totta, 366 interrupta, 821, 385 var. hirsuta, 365 lineata, 321, 385 wrophylla, 340 lunulata, 320, 386 var. nitida, 340 parasitica, 321, 322, 386, 387 sect. Lastrea, 363 resecta, 387 Trichomanes, 382 scolopendrina, 367 ve U.S. GOVERNMENT PRINTING OFFICE: 1974 O—531-696 396