————— JOURNAL #2657 OF THE iy ARNOLD ARBORETUM HARVARD UNIVERSITY B. G. SCHUBERT EDITOR L. I. NEVLING, JR. C. E. WOOD, JR. LAZELLA SCHWARTEN CIRCULATION VOLUME 47 PUBLISHED BY THE ARNOLD ARBORETUM OF HARVARD UNIVERSITY CAMBRIDGE, MASSACHUSETTS 1966 MissOURI BOTANICAL GaRDEN LIBRARY DATES OF ISSUE No. 1 (pp. 1-88) issued January 14, 1966. No. 2 (pp. 89-170) issued April 15, 1966. No. 3 (pp. 171-272) issued July 15, 1966. No. 4 (pp. 273-356) issued October 14, 1966. TABLE OF CONTENTS ARISTEYERA, A New Genus oF GEonomor Paums. Harold E. Moore, Jr. MorPHoLoGy AND ANATOMY OF THE INFLORESCENCE AXIS AND FLowers or A New Pam, ARISTEYERA SPICATA. Natalie W. Uhl NorEs ON THE VEGETATIVE ANATOMY OF ARISTEYERA SPICATA (PauMAE). P. B. Tomlinson Tue GENERA OF CAPRIFOLIACEAE IN THE SOUTHEASTERN UNITED Srates. I. K. Ferguson THe GENERA OF STERCULIACEAE IN THE SOUTHEASTERN UNITED States. George K. Brizicky IDENTIFICATION OF THE SPECIES AND HyBrips IN THE LONICERA TATARICA COMPLEX. P. S. Green On THE IDENTITY oF DrosERA BREVIFOLIA. Carroll E. Wood, Jr. Nores on THE NOMENCLATURE OF Cornus. I. K. Ferguson. ........ THE CoRNACEAE IN THE SOUTHEASTERN Unitep States. I. K. Ferguson Tur Nyssaceab IN THE SOUTHEASTERN UnitTep States. Richard . Eyde Tue GENERA OF ARALIACEAE IN THE SOUTHEASTERN UNITED States. Shirley A. Graham Nores on Some PLAnts or Purrto Rico. Richard A. Howard .... Stupres 1x Droscorea, I: A ContiecTion From British Hon- DURAS. Bernice G. Schubert ... ADDITAMENTA AD FLoRAM JorDANICAE. Baki Kasapligil ................ A REVISION oF THE MALESIAN Species OF ZANTHOXYLUM (RutTa- CEAE). Thomas G. Hartley A Karyonocicat Survey oF Lonicera, I. Lily Riidenberg and Peter S. Green ANATOMY OF THE PaLM RHAPIS EXCELSA, II. Ruizome. P. B. Tomlinson and Martin H. Zimmermann REDISCOVERY OF DAPHNOPSIS HELLERIANA. Lorin I. Nevling, Jr., and Roy Woodbury SCIAPHILA DOLICHOSTYLA (TRIURIDACEAE). Peter S. Green and Otto T. Solbrig POLYSCIAS VERTICILLATA (ARALIACEAE), A NEW SPECIES FROM THE OLOMON IsLaNnps. Benjamin C. Stone CoMPaRATIVE ANATOMY OF THE LEAF-BEARING CacTACEAE, XVI. THE DEVELOPMENT OF WATER-SOLUBLE CRYSTALS IN DEHY- DRATED LEAVES OF Pereskiopsis. I. W. Bailey ....................... THE SIGNIFICANCE OF THE REDUCTION OF VESSELS IN THE CACTA- iO sich ol uae ined ask oe THE GOODENIACEAE IN THE SOUTHEASTERN UNITED STATEs. PP OOOU Ta EN 6 ra ear adie ANATOMY OF THE PatmM RuApPIs EXCELSA, III. Juventite PuHase. P. B. Tomlinson and Martin H. Zimmermann ........... PHENOLOGICAL Nores on Latin AMERICAN PINUS AND ABIES. William B. Critchfield ... Tue Drrecror’s Report .. INDEX TO VOLUME 47 ‘ | VOLUME 47 NuMBER 1 : JOURNAL . OF THE RNOLD ARBORETUM HARVARD UNIVERSITY B. G. SCHUBERT EDITOR L, I. NEVLING, JR. Cc. E. WOOD, JR. LAZELLA SCHWARTEN THE JOURNAL OF THE | ARNOLD ARBORETUM Published quarterly by the Arnold Arboretum of Harvard University. Subscription price $10.00 per year. Volumes LXX, reprinted, and back issues of volumes XXI-XLV are \ the Kraus Reprint Corporation, 16 East 46TH Srreet, New York, NY. 10017. ms and remittances should be addressed to Mrs. Lazenna ARTEN, ARNOLD ARBORETUM, 22 Divinity AvENuE, CamBriDGE, Massa- 02138. . SPICATA. N atalie W. JOURNAL OF THE ARNOLD ARBORETUM Vor. 47 JANUARY 1966 NuMBER 1 ARISTEYERA, A NEW GENUS OF GEONOMOID PALMS ! Harorp E. Moore, Jr. THE GEONOMEAE CONSTITUTE A DISTINCTIVE TRIBE of monoecious are- coid palms (Palmae-Arecoideae) restricted to the Western Hemisphere. The most recent treatments of Burret (1930) and Burret and Potztal (1956) recognize seven genera — Asterogyne, Calyptrogyne, Geonoma, Kalbreyera, Pholidostachys, Taenianthera, and Welfia— to which Calyp- tronoma may be added as a genus distinct from Calyptrogyne as noted by Bailey (1938). Material recently received from Venezuela provides the basis for recognition of still another genus described herein as Aristeyera. The generic name is coined from the surnames of Lic. Leandro Ariste- guieta and Dr. Julian A. Steyermark of the Instituto Botanico at Caracas, who have most kindly loaned or given specimens and provided preserved materials of flowers and fruits for anatomical study. This fine material has provided the writer and Doctors Uhl and Tomlinson, whose articles accompany this one, an opportunity to present a standard morphological description in concert with anatomical studies. The Geonomeae are marked by some 12 characteristics common to all the component genera. The generic distinctions, however, are found chiefly ‘in the androecium, in the number of locules and ovules in the pistil at anthesis, in the fibers of the mesocarp of the fruit, in the arrangement of floral pits on the flowering axis or axes, and to some extent in the structure ‘of the inflorescence. The synoptic key that follows is intended to suggest the combinations of characteristics that separate genera and the relative levels of their specialization. Geonoma is the largest, mos uted, and, so far as the unilocular, uniovulate pistil 1s_ , U most advanced genus in the tribe. Burret (1930) recognized 172 species in the genus plus a few not well known, and some 26 species have been described since. If Taenianthera is to be included within the genus, as ?This study has been supported in part by National Science Foundation Grant -1354, GB , 2 Unt, N. W. Morphology and anatomy of the inflorescence axis and flowers of a xp new palm, Aristeyera spicata. Jour. Arnold Arb. 47: 9-22. 1966. : fact 1 — Notes on the vegetative anatomy of Aristeyera spicata (Palmae). Jour. Arnold Arb. 47: 23-29. 3 pls. 1966 2 JOURNAL OF THE ARNOLD ARBORETUM [vou. 47 seems reasonable, then the number of recognized species slightly exceeds 2 e remaining genera are small or monotypic and are characterized, so far as known, by a trilocular, triovulate mistil in combination with various modifications of the androecium and inflorescence Although the tribe is being studied morphologically and anatomically as part of a program to produce a broadly based revision of the Palmae at the generic level, it seems desirable to describe Arvisteyera separately and to put on record now some observations on the complex anatomy of the inflorescence axis as well as certain features of floral morphology and anatomy, and vegetative anatomy. The genus is of especial interest be- cause of the structure of the anthers which parallels that of anthers in Asterogyne, Geonoma, and Kalbreyera, though in other respects the staminate and pistillate flowers are basically similar to those of Welfia with which it seems most clearly allied. Key To GENERA OF GEONOMEAE 1. Staminate flowers with numerous stamens (21-42); eslaege flowers with numerous staminodes (15-18); pistil trilocular, triovulate at anthesis; meso- carp with a layer of closely placed slender fibers senate the endocarp; bracts subtending the ~enatie ni agree basally; floral pits arranged in ‘vertical series on the inflorescence a 2. Anthers erect in bud, sagittate, eve thecae united laterally with the con- nective ; inflorescence of several massive unbranched axes from a short dorso-ventrally compressed decurved peduncle and very short rachis, subtended by massive broad woody bracts; seed with the raphe sparingly branched basally and apically; large palms with stout stems and long pinnate ua leaf with 2-layered colorless hypodermis abaxially and Oe FoF eai eh rea oie atees lode ei eel eeias Wel bifid connective; inflorescence spicate, erect, lon subtended nm co 9 = | fet) =a Oo jon 2) ° = oO ca 77) = cl a a n o Lie | mt pet) oO he — ww n cr 5 oO = 77) | = n pce cS anes 0) a oO j=) io) a oO ca 7) = _ = a mn ae 9 3 ' mesocarp; floral pits arranged in vertical series; inflorescence and inser- tion of bracts various. 4. Upper bract of the inflorescence inserted at the base of the peduncle or somewhat above the base, never at the base of the flowering axis or axes. 5. Inflorescence once- or twice-branched; bracts subtending the floral pits with margins not overlapping adjacent bracts; staminodia of 1966 | MOORE, ARISTEYERA 3 pistillate flowers united in a solid receptacle below the spreading lobes; fruit with wrinkled exocarp when dry but the fibers of the mesocarp not clearly outlined. ................ Calyptronoma. Inflorescence spicate; bracts subtending the floral pits with margins overlapping the bases of adjacent bracts; staminodia united in a ne tube below the spreading lobes; ‘fruit with fibers of the carp prominently outlined when dry. siete Pholidostachys. on enclosing the spicate flowering axis in bud, caducous at anthesis, leaving a ring-like scar at apex of the long peduncle, the lower bract inserted at the base of the peduncle. .............. Calyptrogyne. (a F (unknown in Kalbreyera); bracts subtending the spicate to paniculately branched inflorescence inserted near or at the base, never at the base of the flowering axis; floral pits variously arrange 6. Pistil trilocular, triovulate at anthesis with elongate terminal styles loosely united Loa ad floral pits arranged in more or less clearly defined vertical ro Stamens 6; pistillode yaar as long as the stamens; Bginicesgrng tube with 6 spreading lobes; floral “in in 7 vertical rows; in- florescence long-pedunculate with 2- WisGlettne Rsiisicnn CE ges ce ee OE a ae ee erogyne. Stamens 3; pistillode short, trifid; staminodial tube gerlints floral pits in 4 vertical rows and nearly decussate; inflorescence short-pedunculate, terminating in a diffuse panicle of flowering Ce a POO ee ae IU WS AED Seer rs | AL ae iy 1 Kalbreyera. Pistil unilocular, uniovulate at anthesis with lateral or basal styles loosely united; floral pits spirally arranged or verticillate; inflorescence spicate to diffusely paniculate. ...... Geonoma (incl. Taenianthera). ~s ~ e Aristeyera H. E. Moore, gen. nov. [Palmae-Arecoideae-Geonomeae |. Genus Geonomearum Welfiae affinis sed folia praeter apicem bifidum indivisa pinnate nervata, inflorescentia spicata bracteis tenuis, florum masculorum staminibus 21—24 antheris in gemma inflexis thecis disjunctis in connectivo bifido, semine rapha indivisa vel furcata instructo. Solitary, slender, unarmed, monoecious, protandrous undergrowth palms. Leaves entire except for the bifid apex or rarely ruptured but scarcely divided laterally, pinnately nerved; sheath short, at first closed but soon rupturing opposite the petiole, this rounded below, shallowly grooved above, with sharp entire margins. Inflorescences interfoliar, erect, solitary at the nodes, subtended by 2 closely sheathing basally inserted bracts, the lower ancipitous, opening apically, the upper longer, rounded, acute and at first enclosing the spike in bud; peduncle elongate, terminating in a thick undivided spicate flowering axis, the flowers borne in triads of 2 lateral staminate and a central pistillate (or paired or solitary staminate through abortion of the pistillate toward the apex of the axis) sunken in prominent pits, these covered with a prominent bract in bud and arranged in 8 verti- 4 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 |Z We Ne ‘ \ i we Vl j ff ry | > 4 epal ON Be ee pistillode, x re i, anther, X 8; j, theca and connective, x 8; pistillate flower with bracteole in bud, x 4; 1, pistillate flower in vertical section, x 4; m, sepals, X 4; n, corolla, X 4; 0, corolla expanded and eae » stam inode, X 8; 4 pistil, ef, pistil i in vertical section, X 4; s, ovary in cross-section, x 8: , fruit, X 2; u, base of fruit and stigmatic eave M25 1966 | MOORE, ARISTEYERA 5 cal rows, bracteoles subtending the flowers 3, anatomically one per flower but in the pit placed so that superficially one appears to subtend one of the staminate flowers and two surround the base of the pistillate flower. Staminate flowers about 1/3 exserted at anthesis, the 3 sepals distinct above a short attachment to the receptacle, glumaceous, imbricate at least basally in bud; petals 3, connate, sometimes unequally so, for 1/3-3/4 their length in a soft tube, the free, slightly asymmetric lobes deltoid, glumaceous, valvate above briefly imbricate bases; stamens 21-24, ex- serted at anthesis, the filaments united with the base of the pistillode and with the corolla-tube about 1/4 their length in a fleshy solid receptacle, then expanded and connate in a tube about equally long surrounding a fleshy, shallowly and acutely, often irregularly 3-lobed pistillode, the free portions of the filaments only slightly shorter than the connate portions, erect, the anthers inflexed in bud, the thecae separate, longitudinally dehiscent, terminal on arms of a bifid connective, the arms of the connec- tive and thecae becoming erect or nearly so at anthesis. Pistillate sepals 3, glumaceous, distinct, imbricate in bud; petals 3, connate for about 2/3 their length in a soft fleshy tube, the deltoid free glumaceous lobes valvate in bud, probably spreading at anthesis; staminodes 15~18, connate basally and adnate to the corolla-tube about 1/2 their length, then free, angled- linear, often with shallow adaxial grooves and sometimes thickened api- cally, the apices dark and connective-like; pistil trilocular, triovulate, the ovary with 3 rounded lobes; styles terminal, elongate, loosely united; stigmas 3, prominent and probably recurved at anthesis; ovules hemi- tropous, axile, pendulous, attached near the top of the locule, only one normally maturing. Fruit ellipsoid-ovoid in outline, slightly compressed laterally (in relation to the flower but appearing dorso-ventrally com- pressed in relation to the axis) and slightly keeled apically when dry, the residual stigmas and abortive carpels basal; exocarp smooth; mesocarp dry with longitudinal fibers closely placed around the endocarp, this thin, crustaceous, shining, not adherent to seed; seed ellipsoid to obovoid and slightly compressed laterally; hilum small, excentrically basal; raphe en- circling the seed from hilum to excentrically basal embryo, somewhat impressed and unbranched to furcate or with a few parallel branches; endosperm homogeneous. TYPE species: Aristeyera spicata. Aristeyera spicata H. E. Moore, sp. nov. Caules solitarii ad 8 m. alti. Folia elongato-cuneata petiolo 35-36 cm. longo rachidi 68-70 cm. longa lamina utrinque 27-nervata. Inflorescentia 70-86 cm. longa spica dense fulvo-tomentosa 25-40 cm, longa. Flores masculi 7 mm. longi feminei in gemma 6 mm. longi. Fructus atroruber 12-14 mm. longus 6~7 mm. latus 6 mm. crassus Trunk short, 2-8 m. high, to 3.5 cm. in diameter or perhaps more, Vv, w, seed in adaxial and abaxial views, 2; x, seed in vertical section, 2. All from preserved material: L. Aristeguieta, Dec. 28, 1962. 6 JOURNAL OF THE ARNOLD ARBORETUM [VOL. 47 Ministeria de Agricultura y Orla nensamio NACIONAL DE VENEZUELA FLORA DE VENEZUELA Oricitngan sperala KE RT ATR? eee AX ve pee - Me ALTEIN A fiutian A, Mieeermerh FIGURE 2, Aristeyera spicata. Leaf from holotype, * ca. %. rather closely ringed with oblique leaf-scars. Leaves elongate-cuneate in outline, erect, subcoriaceous, deep green above, silvery green below, un- divided except for the caudately bifid apex, or the blade sometimes ir- regularly split but the segments scarcely separated, the rachis terminating in a short filament; sheaths at maturity open to or nearly to the base, frayed into a few stiffish fibers marginally, densely dull red-brown, lepidote-tomentose, ca. 15 cm. long and narrowed to a petiole ca. 36 cm. 1966 | MOORE, ARISTEYERA FicuRE 3. Aristeyera spicata. Part of stem with base of inflorescence and Portions of inflorescences in fruit, at anthesis, and in bud, from holotype, x ca, 34. 8 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 long, this deeply concave above, convex below, dark-brown-lepidote at least laterally below or becoming merely dark-puncticulate in age; rachis 6.8-7 dm. long, concave at base above but becoming triangular and elevated toward apex, rounded and brown-lepidote or -puncticulate below; blade ca. 30 cm. wide at apex of rachis, the caudate tips 19.5-33 cm. long on the inner margin, nerves ca. 27 per side, not elevated above, yellowish and prominent below when dry, brown scurfy toward base and with scattered brown mostly medifixed membranous scales to 6 mm. or more long on basal half, intermediate secondary nerves 2, tertiary nerves 7 or more, often minutely lepidote, the surface with numerous translucent dots. Inflorescence 70-86 cm. long; lower bract ca. 24 cm. long, upper ca. 39 cm. long, both more or less densely brown-lepidote-tomentose as the sheath; peduncle ca. 46 cm. long, densely fulvous-tawny lepidote-tomen- tose; spike 25-40 cm. long, densely fulvous-tawny lepidote-tomentose in bud and darker tomentose in fruit, usually narrowed to a sterile spinose tip ca. 2.5 cm. long. Staminate flowers with bracteole ca. 5 mm. high; sepals adnate basally to the receptacle for ca. 0.5 mm., nearly or quite as long as the petals, 6.5-7 mm. long, keeled, entire, and acute to emarginate at apex; petals 7 mm. long, united to receptacle for about 1.5 mm. then connate to 2/3 their length, the lobes 2.5—3 mm. long; stamen-filaments adnate to receptacle ca. 2 mm., then expanded in a tube ca. 2 mm. long, free filaments to 3 mm. long, connective to 1.5 mm. long, thecae to 1.5 mm. long; pistillode ca. 1 mm. high. Pistillate buds ca. 6 mm. high; sepals strongly keeled and about 6 mm. long, equalling the petals. Fruit dark red, 12-14 mm. long, 6-7 mm. wide, 6 mm. thick; seed 9-10 mm. long, 5 mm. wide, 4-5 mm. thick, 3 mm. across base. VERNACULAR NAMES: palma, palmito. Uses: fruit said to be edible, leaves used for durable thatch supposed to last ten years. DISTRIBUTION: on steep slopes of mixed evergreen forest. Venezuela. Estapo MirANpDA: Parque Nacional de Guatopo; moist rich mixed evergreen forest on steep slopes bordering Rio Santa Cruz, between Santa Teresa and Altagracia de Orituco, 14.5 kms. from Los Alpes, 12 kms. from Rancheria Mi Querencia, alt. 520 m., 23 Nov. 1961, J. A. Steyermark 90010 (BH, holotype; VEN, isotype); carretera Sta. Teresa-Altagracia de Orituco, June 1953, L. Aristeguieta 1772 (VEN); selva de Guatopo, June 1958, L. Aristeguieta 3185 (VEN); without definite locality, 20 Dec. 1961, G. Agostini S. 00010 (vEN). LITERATURE CITED BarLtey, L. H. The oo catia sac Problem — The Manac Palms. Gent. Herb. 4: 152-1 Burret, M. egg peeilinnt ee Jahrb. 63: 123-270. 1930. URRET, M., & E. Portztat. Systematische Ubersicht iiber die Palmen. 7 Arecoideae. Willdenowia 1: 350-385. 1956. L. H. Bartey Hortorrum, CORNELL UNIVERSITY IrHaca, NEw YorkK am a Sy Me 1966] UHL, MORPHOLOGY OF ARISTEYERA 9 MORPHOLOGY AND ANATOMY OF THE INFLORESCENCE AXIS AND FLOWERS OF A NEW PALM, ARISTEYERA SPICATA ! NATALIE W. UHL A DESCRIPTION OF THE INTERNAL STRUCTURE of the unusual inflores- cence and flowers of Aristeyera spicata is presented here to accompany the generic description (Moore, 1966) and an account of the vegetative anat- omy (Tomlinson, 1966). In the palms, branching of the major inflorescence axes is monopodial (H. E. Moore, Jr., personal communication), but in the ultimate flowering clusters, several different patterns of branching occur. Flowers have been described as solitary (Sadba/), in pairs (Calamus), in triads (Aristeyera), in glomerules (Livistona chinensis), in acervulae (Mascarena), and in cincinni (Borassus). More than half the known genera of palms bear flowers arranged in triads. Work is presently under way on the structure and relationships of these different flowering units. The anatomy of the Geonomoid palms (Palmae-Arecoideae-Geonomeae ) has not been studied previously. In this tribe, 3-flowered clusters (triads) are completely or partially sunken in pits in the inflorescence axes. This paper presents an anatomical analysis of the flowers and of the branching pattern of the triad as found in Avistevera spicata. MATERIAL AND METHODS The material studied, which was kindly collected and fixed in FAA by Lic. L. Aristeguieta on December 28, 1962, included inflorescences in bud, at anthesis, and in fruit. Material was softened by treatment for two weeks in one-third strength commercial (52 per cent) hydrofluoric acid. Serial sections of flowers and parts of the inflorescence, prepared by the usual paraffin method, and cut at 10 and 15 microns, were stained with safranin and fast green. Although tannins made complete clearing difficult, some flowers and portions of axes were cleared satisfactorily by treatment overnight in 5 per cent sodium hydroxide in an oven at 59° C., followed by two to four hours in one-third strength commercial sodium hypo- chlorite, and approximately 12 hours in a saturated solution of chloral hydrate. INFLORESCENCE AXIS The spicate inflorescence axis of Aristeyera (Fic. 1) consists of eight vertical rows of pits, each pit containing a cluster of three flowers (a * From work supported by National Science Foundation grant GB-1354. 10 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 triad), and closed completely in bud by an obovate bract. In Ficures 3 and 4, the letters A-H indicate successively higher pits in the phyl- lotactic succession (usually corresponding in mid-axis to an irregular 3/8 phyllotaxy) of a young inflorescence axis with buds. Below each pit, the Fics. 1 and 2. Aristeyera spicata. Fic. 1. Part of a spike with staminate flowers “ anthesis; es pit-closing bract, * 2. Fic. 2. Cleared preparation erchine three pits: A, B, and C. Refer to pits A, B, and C in Fie. 1 for orien- tation vascular strands to each triad (at the level A in Fics. 3 and 4) are ar- ranged in a single group. Slightly higher (represented by the vascular plexus at B), these bundles are divided into two groups. In the pit at C, the section passes through a pit cavity containing flower bases and bracteoles. A pit including flowers and bracteoles at a slightly higher level is cut at D. The four pits, E, F, G, and H are sectioned above the flowers, at successively higher levels. Enclosed in each pit are two lateral staminate flowers (Fics. 3 and 4, fs and ss), an inner central pistillate flower (pf), and three bracteoles (Fics. 3 and 4, ob, mb, ib). Flower clusters may be designated as left- handed or right-handed depending on whether the first staminate flower (Fics. 3, 4, and 9, fs), borne outside all bracteoles, is on the left or right side of the triad. Left-handed and right-handed triads are mirror images of each other and generally alternate in spiral succession. The sequence of the floral axes in a pit may be deduced from the posi- tion of the three bracteoles associated with the flowers. It appears that the initial axis (Fic. 10, broken line) of the 3-flowered group arises in the axil of the bract that covers the pit (pcb, Fics. 1, 3, 4, and 9). This axis bears the outer (Fic. 10, ob) bracteole of the triad and terminates with the first staminate flower (Fic. 10, fs). Arising in the axil of the outer bracteole (Fic. 10, ob) is a secondary axis that bears the bicarinate middle bracteole (Fic. 10, mb) and ends with the second staminate flower (Fic. 10, ss). A third-order branch arises in the axil of the bicarinate bracteole (Fic. 10, mb), bears the uppermost (inner) bracteole (Fic. 10, ib), and 1966 | UHL, MORPHOLOGY OF ARISTEYERA 11 ends with the pistillate flower. No vascular bundles are present in any of the three bracteoles, but each has an adaxial row of fibrous bundles having no connection with any other veins, The branching pattern of the triad here corresponds to that of a monochasium, as defined by Rickett (1955, p. 444). Thus, on the basis of general form and the pattern of bracteoles and flowers in each pit, the inflorescence of Aristeyera is interpreted as an indeterminate main axis bearing eight rows of determinate 3-flowered units (Fic. 10), each unit enclosed in a pit sunken in the main axis and covered by a subtending bract. Histological features. The inflorescence axis is densely covered with profusely branched, dendroid trichomes containing tannin and originating from columnar epidermal cells. The hypodermis is a single layer of short, isodiametric cells. In the periphery of the inflorescence axis, just inside the hypodermis, many small (5 to 10 cells in transection) bundles of fibers form an anastomosing system. The large bundles of the axis are arranged in a low spiral and oriented with the xylem toward the center of the axis. They gradually decrease in size from the largest, which occur in the second row inside the pits, to a few small vascular or entirely fibrous strands in the center. The xylem of the large vascular bundles consists of two large vessels and several smaller elements. Xylem and phloem meet in a straight line (type 3A, Cheadle and Uhl, 1948). The large vessels, often immature even in old axes, have long oblique, compound perforation plates with many (40 to 50) scalariform bars. Smaller xylem elements are helically thickened. A fibrous sheath, generally two cells in width, but sometimes wider (3-4 cells) around the phloem, surrounds each bundle. Two adjacent phloem areas in the same bundle occur below a bundle branch. Pits. Each cluster of three flowers and the associated bracteoles are deeply sunken in a pit. Around the inner edge of each pit is a single row of bundles (Fic. 3a, sr). Bundles in the centripetal part of this row are oriented as are the central bundles of the axis, with the xylem toward the center, but between adjacent pits (Fic. 3a, pi), the bundles appear partially inverted with reference to the center of the axis. This orientation of axis- bundles adjacent to the pits indicates that pits represent depressions or invaginations in the surface of the axis (Fic. 3a, arrows). Where vascular bundles between adjacent pit cavities (Fic. 3a) are crowded together, back to back, fusion sometimes occurs between the xylem (Fic. 3a, fb) of bundles extending around the edge of one pit and the corresponding row around the adjacent pit. The xylem of these fusion bundles is large, sometimes two-lobed, and between two opposite phloem groups. Distally, where the pit cavities are smaller in size and the vascular strands are less crowded, the bundles are usually separate. Each pit is covered by a bract with a fleshy, tapering proximal half which is completely embedded in the axis (Fic. 4, stippled areas, A, B, JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 Ics. 3, 3a. Aristeyera spicata. Fic. 3. Transection of a young spike. Phyl- lotactic a parade indicated by letters: A, B, C, D, E, F , D, E, F, G, and H. Centripetal to A and B vascular complexes below pits. Pits a bases and beasiegle . s : niente pcb, pit- -closing bract; fs, first stam inate flo ower; pf, pistillate flower; ss, se staminate flower; ob, outer bracteole; mb, middle bracteole; ib, inner peiceoale’ 1966 | UHL, MORPHOLOGY OF ARISTEYERA 13 C, D). Distally, where the bract is free, the edge of the axis (Fic. 3, G) is involute and the upper edge of the bract is revolute. When the inflores- cence is in bud, the inrolled edges of the axis “lock” the bract into position G. 4. Aristeyera spicata. Three-dimensional diagram of transection of young rhe illustrating structures seen in F1G. 3. Vascular bundles seen in FIG. 3a not represented. [Trac ed from material represented in FIG. 3 and a section 60 microns lower. } over the pit. Later, when the upper half of the bract becomes disengaged, the flowers are exserted through a mouth-like opening (Fic. 1). In fruit the bract is pushed farther away making a larger opening. These observa- pl, approximate position of polarized light pictures (Fics. 5-8), X 12. Fic. 3a. Diagram of lower part of transection (Fic. 3) to show orientation of bundles eee pits. Details: sr, single row of bundles around contour of a pit; pi, partially inverted bundle; fb, fusion bundle; phloem, qe xylem, white. 14 JOURNAL OF THE ARNOLD ARBORETUM [ VoL. 47 tions also support the conclusion that the pit represents an invagination of the surface of the axis. Subtending bract. The limits of the obovate bract covering each pit are easily recognized histologically. Except for an adaxial row of parallel vascular bundles (Fic. 2, B) the bract is composed entirely of parenchyma cells that are somewhat larger than those of the axis proper (Fic. 3, D). In cleared preparations, it is possible to remove the epidermal and outer fibrous systems of the axis and the thick outer parenchymatic part of the bract and thereby reveal the longitudinal shape of the bract and the vascular skeletons of bract and axis. A group of three pits (Fic. 2, A, B, and C, corresponding to pits A, B, and C in Fic. 1) were prepared in this manner. The bract bundles, shown in transverse section, abaxial to the pits at C and D in Fic. 3, are shown in longitudinal view below B in FicurE 2. These traces originate from bundles of the main axis, usually as branches of the row of strands (Fic. 3a, sr) that, at a lower level, curve around the pit cavity beneath. Near the tapered base of the bract, some of these veins may enter the bract directly, or they may divide to form an inner branch to a floral axis and an outer branch to the bract. The shape and extent of the lower part of the bract can be further verified by com- paring the longitudinal view (Fic. 2, B) with different levels of the bract as seen in transverse section (Fic. 4, stippled areas). In Ficure 3, the letters A and B are in the approximate centers of small areas of large parenchyma cells representing transverse sections of the lower, tapered bases of the bracts. Slightly higher, the bracts are larger (Fic. 4, C and D, stippled areas). Vascular patterns below pits. The configuration of bundles below the three flowers in a pit is illustrated by photomicrographs taken in polarized light (Fics. 5-8) in which the early metaxylem is anisotropic and appears as bright spots. The relative size and position of the areas in these pictures is indicated by the broken lines in Ficure 3, pl. The fibrous sheaths of the bundles in this young axis were not mature and did not exhibit birefringence. e first evidence of a pit appears as a bundle-free central area with small peripheral bundles (Fic. 5), approximately two-thirds to three- fourths the vertical distance above the top of the pit below. These small bundles originate as centripetal branches of the single row of strands around the pit. Slightly above the level in Ficure 5, those small bundles which at a higher level supply the pit triad are arranged in an oblong group (Fic. 6, bf). This group of bundles divides into a larger (Fic. 7, pfss) and a smaller cluster of strands (Fic. 7, fs). The veins of these two groups become oriented so that the xylem of each bundle faces the center of each group. The smaller cluster supplies the first staminate flower (Fics. 7-10, fs). At a higher level (Fic. 8), the second and larger group of bundles divides again to form two groups, the inner (pf) supplying the pistillate flower and the outer the second staminate flower (ss). 1966] UHL, MORPHOLOGY OF ARISTEYERA 15 Thus anatomy confirms an arrangement of the flowers in the triad which was suggested by the position of the bracteoles (Fics. 9 and 10). The group of irregularly arranged bundles (Fic. 6, bf) represents the base of the branch subtended by the pit-closing bract and is shown as a dotted line in Ficure 10. This group of bundles divides (Fic. 7 and Fic. 10, level ob), forming the vascular cylinder of a submerged branch on which the first staminate flower is terminal and a second branch which is represented by the vascular bundles in Ficure 7, pfss. The second branch terminates in the second staminate flower (ss) and bears a lateral branch (the cluster of bundles, Fic. 8, pf) terminating in the pistillate flower. Detailed study of the vascular bundles shows evidence of additional branching. The vascular supply below the first staminate flower always divides to form two, usually unequal, groups of bundles (Fics. 8 and 11, fs), each oriented with the xylem toward its center. In a mature axis, two vascular cylinders of approximately 9 bundles each can be seen surround- ing the ends of the arrows (Fic. 11, fs). The same may also occur below the second staminate flower (Fic. 11, ss). The supply to the pistillate flower is often separated into two groups without forming two completely reoriented circles of vascular bundles. Each of the two groups of bundles at this level resembles the vascular supply that eventually enters the receptacle of the individual flower, but above this level, some (2-4) of the bundles fuse, others (1—4) end blindly and the two vascular cylinders become reoriented into one. The new cylinder is composed of ten to twelve vascular bundles just below the insertion of the perianth parts of each flower. This abortive branching of the vascular supply below a flower is evidence that, ancestrally, additional flowers may have been present in each triad. In Avisteyera this unit seem to have originated from a more ramified type of inflorescence. FLOWERS In the following sections certain structural aspects common to both staminate and pistillate flowers are first discussed briefly and then details peculiar to each are described. Vascular bundles in the flowers are very small, the xylem containing no large vessels and consisting of a few helical elements. The number of traces to perianth parts is variable. Two other features are worthy of note: (1) in the petal tube of the staminate flowers and in the sepals of pistillate flowers, some of the traces appear to differen- tiate over a long period of time and as a result the number of traces in- creases with the age of the flower; (2) the traces developing later do not become connected to the vascular strands in the floral receptacles. Un- connected lateral traces have also been found in perianth parts of Borassus (unpublished) and in Chamaerops (Morrow, 1965). There is evidence in monocotyledons that the procambium of lateral leaf traces develops basipetally (Sharman, 1942). In certain palms this is also true for some lateral bundles of sepals and petals. 16 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 i metaxylem ; onigg Center of the main axis is to the right, pt eh to left, all * 25. 5. pproximately 720 microns above the pit below. Fic. 6. 170 microns alone Fic. 5; bf, bundles which, at a higher level supply the three flowers in the 1966] UHL, MORPHOLOGY OF ARISTEYERA 17 Staminate flowers. In staminate flowers the stalk has a wide paren- chymatic cortex and a small central vascular cylinder of a varying number of bundles (5, 6, 7, and 11, observed). Traces of all floral organs arise in a spiral pattern as branches of these few receptacular bundles (Fic. 18). The stalk elongates until at anthesis the upper one-third of the flower with its exserted stamens is pushed out between the pit-closing bract and the axis (Fic. 1). Anthesis of both staminate flowers occurs before the pistillate flower in the same pit is receptive; as the pistillate flower matures, the staminate flowers wither in the pit or are shed by rupture of the stalks The sepals are attached spirally, their few vascular bundles arising as branches from receptacular bundles opposite their bases (Fic. 18, $1, S2, S3). In each flower, the sepals are imbricate and progressively smaller. Corresponding to this sequence, the lowest sepal receives three, four, or five vascular traces and the second and third sepals, one, two, or three traces. In the adaxial row in each sepal most of the bundles are fibrous strands which alternate with the few vascular bundles (above), and extend from the sepal bases nearly to the tips, but these do not extend into the stalk nor have any connection with bundles of the floral receptacle. The petals of staminate flowers are connate for about one-third to three- fourths their length; above this they are imbricate for a short distance, then becoming valvate. Traces to the corolla arise from receptacular bundles in three successively higher groups (Fic. 18, pl, p2, p3; Fie. 17, pe), indicating a spiral arrangement of the petals. Some of the vascular bundles differentiate basipetally and are not connected to the floral recep- tacle. In the upper three-fourths of the petals many unconnected fibrous strands occur in the adaxial row of bundles. The morphology of the 21 to 24 stamens is unusual (Moore, 1966); the anatomy, however, is straightforward. Bundles of the receptacle branch above the origin of the petal traces in a flat and irregular spiral pattern to form a single trace to each stamen. In FicuRE 17, traces to 23 stamens (st) surround the remaining three bundles of the floral receptacle (center). The pistillode resembles the gynoecium of the pistillate flowers, con- sisting of a central axis with three rudimentary carpels. Bundles (3, 4, or pit. Fic. 7. 120 microns above Fic. 6; fs, supply to the first staminate flower; pfss, supply to the second staminate flower and the pisti illate base r. Fre: 8. 320 microns above Fic. 7; fs, supply to the first staminate flower, divided into flower and bracts occur at tulle this level but are extended to facilitate cates bracteoles; pcb, pit-closing bract; ob, outer bracteole; _mb, middle bicarinate bracte ae ib, inner bracteole; fs, branch bearing first staminate anch be s two vascular cylinders below the first staminate flower; ss, two vascular cylin- ders below the ae staminate flower; pf, group of bundles ik supply the Pistillate flower, < 2 18 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 & ee urs 4 es Ws - Mt #. = cs. 12-17. Aristeyera spicata, Fic. 12. Transection of part of the ovary showing the locule of one carpel with ovule; rec, floral receptacle above separa- tion of the traces to all thee carpels; cs, lateral traces to uppermost carpel, x 61. Fic. 13. yhagoage meal - ape of f the three carpels at the level of attach- ment of the ovule in the le carpel, X 61. Observe the difference in the length of the slits beside th pan Fic. 14. Transection of part of the pistillate flower including the ovary and a portion of the corolla- staminode tube; pe, aga trace; stm, ~ a staminode, x 40. Fic. 15. Transection of p f de thr rough ¢ one carpel it si the pen ndulous ovule, 44. Feo. 17. Trans —_ of a staminate flower; pe, arrows indicate one group of petal traces; st, stam trace, X 44. 1966] UHL, MORPHOLOGY OF ARISTEYERA 19 7 observed) remaining above the stamen traces (center bundles, Fic. 17) supply the pistillode. In one flower, the seven traces to the pistillode formed three groups and the xylem of each group coalesced, resembling the pattern found in the supply to the carpels of pistillate flower. Usually a single trace, sometimes formed by fusion of two or more receptacular strands, supplies each vestigial carpel. Pistillate flowers. In bud, the pistillate flower, situated below and adaxial to the male flowers, is attached by a short, thick base which ex- tends adaxially between the stalks of the staminate flowers to the inner central region of the pit. The vascular supply of the axis bearing the pistillate flower usually consists of a cluster of ten to 12 bundles. Several (2-4) of these fuse or end blindly and the remaining extend into the flower (Fic. 19). Vascular traces to the sepals are somewhat fewer than in the staminate flowers; two or three strands to the first sepal, two to the second, and one or none to the third were observed (Fic. 18, $1, $2). As in the staminate flowers, each sepal is successively smaller and the few vascular traces alternate with unconnected fibrous strands in an adaxial row of bundles. The petals are connate for about two-thirds their length, then briefly imbricate, and the free tips valvate. The corolla tube is supplied by three very eal traces (Fic. 19, P; Fic. 14, pe). These are the smallest bundles in the flowers and consist of one or two xylem elements and one or two sieve elements. In the upper valvate region of the petals are many unconnected fibrous bundles. The 15 or 16 staminodes (Fics. 14 and 15, stm) are adnate to the connate petals for about one-fourth the length of the corolla tube. Each receives one small bundle, that arises as a branch of a receptacular strand (Fic. 19) and ends between the vestigial anther sacs. Occasionally a trace to one staminode branches to provide the bundle to an adjacent one. The floral axis extends approximately to the top of the locule of the highest carpel (Fic. 14, rec). This extension of the floral receptacle above the origin of all carpel traces is evident in the group of small bundles remaining in the center (Fic. 12, rec) and, at a higher level, as a paren- chymatic tip (Fic. 14, rec). The gynoecium (Moore, 1966, Fig. 1, q) is composed of three carpels that are spirally arranged and closed basally, where they are adnate to the central axis. Near the upper end of the central axis (Fic. 14), the three carpels become free from each other and from the axis. Above the attachments of the ovule, the ventral sutures of each carpel are open; epidermal cells can be seen between individual ventral carpel sutures (Fic. 14). In the long stylar region (Moore, 1966, Fig. 1, q), the carpels are open and the degree of connation varies. At some levels glandular epidermal cells extend part way between adjacent carpels (Fic. 15). bove the origin of the traces to the staminodes, the floral receptacle consists of a ring of eight to 13 bundles (Fic. 19). Just below the ovary, these bundles separate into three groups and the xylem of the bundles in v 18 Fics. 18 and 19. Aristeyera spicata. Fic. 18. Diagram of the origin of floral traces in a young staminate flower. Details: S1, ge to lowest sepal; S2, traces to second sepal; S3, traces to highest sepal. P1, and P3, groups of successive petal traces. Asterisks, stamen traces; black cision traces we ‘tie pistillode. 1966] UHL, MORPHOLOGY OF ARISTEYERA 21 each group is united, forming a fusion bundle with one large central xylem area and three to five outer patches of phloem. The central part of each of these compound bundles extends radially across the syncarpous base of the ovary and becomes a dorsal carpel bundle which curves upward around the locule and divides tangentially at the base of the long style. One of the branches crosses the top of the locule for a very short distance, the other extends through the style to the stigmatic lobe. At the base of each carpel, some 20 lateral bundles diverge from the two opposite sides of the fusion bundle and form a complete ring of bundles which extend around each locule. Two of these lateral traces extend into the style. Several (2-4) of the lateral bundles fuse to form a large bundle which supplies the ovule. In certain other palms (Morrow, 1965) the ovules are supplied by several ventrally situated bundles which do not fuse but extend into the ovule The carpel margins (Fic. 14, centripetal to locules) are very wide. The word “margin” is arbitrarily chosen for the descriptive purposes of this paper. Whether this thick area represents wide carpel margins, Puri’s “lateral face” (Puri, 1961), or a closely appressed region of the ventral surface, as would be postulated by the conduplicate concept (Swamy and Periasamy, 1964), may become clarified as more palms are studied. The single, pendulous ovule (Fic. 16) is attached near the top of the locule on one of the carpel margins. (Note the difference in the length of the slits beside the funiculus in Ficure 13.) The attachment of the ovule between closely appressed carpel margins is unusual. With the exception of the Cocoideae, where the ovules are sessile, and sometimes (£/aeis) , embedded in the ventral carpel walls, most palm ovules appear to be submarginal. A better understanding of the ovule attachment in Aristeyera should re- sult from work in progress on other palm genera (Moore, 1966). SUMMARY 1. Flowers in Arvisteyera occur in “triads” and are deeply sunken in pits, which are regarded as invaginations of the axis. 2. The relation of the three bracteoles to the three flowers in each pit suggests that each triad represents a monochasial branching unit. 3. The vascular supplies of the three axes, each terminating in a flower, are enclosed in the axis below each pit 4. Anatomical evidence of additional aborted flowers or branches was found on the branches terminating in each staminate flower and possibly also on the branch ending with the pistillate flower. This suggests that the Aristeyera triad has originated from a more ramified type of inflorescence. 5. Perianth parts of both staminate and pistillate flowers are supplied by a variable number of traces which arise in a spiral pattern as branches from bundles of the floral receptacles. Fic. 19. Diagram showing the origin of the traces to a young Sperges flower. Details: S1 and S2, bundles to the first two als; third sepal of this flower received no traces. P petal bundles; asterisks, traces to scadtedes: C1, lowest carpel; C2, middle carpel; C3, upper carpel. 22 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 6. In the petals of the staminate flowers and the sepals of the pistillate flowers some of the lateral veins appear to differentiate basipetally and never become connected to the bundles of the floral receptacles. 7. Fibrous bundles having no connection to floral receptacles occur in definite patterns in perianth parts. Such bundles extend the entire length of the sepals and petals in staminate flowers, and of the sepals in pistillate flowers, but occur only in the distal, valvate part of the petals in the pistillate flowers. 8. The usual vascular supply to stamens, staminodes, and rudimentary carpels consists of one small vascular bundle. he gynoecium is composed of three spirally inserted carpels. At the base they are closed and congenitally adnate to the extended central axis. In the upper part of the ovary, above the extension of the floral recep- tacle, the carpels are free and ventral sutures can be distinguished. Through the long stylar region, each carpel is open and connate at its margins with the other two carpels forming a 3-pronged stylar canal. 10. A single, pendulous, hemitropous, bitegmic ovule is attached to one margin of each carpel. The ovular supply is a large trace formed by the fusion of 2 or 3 bundles. ACKNOWLEDGEMENTS I wish to express my appreciation to Dr. H. E. Moore, Jr., Cornell University, for directing this study; to Doctors H. W. Blaser, University of Washington, and P. B. Tomlinson, Fairchild Tropical Garden, for re- viewing the manuscript critically; and to my husband, Dr. C. H. Uhl, Cornell University, for assisting in all possible ways. LITERATURE CITED CHEADLE, V. I., & N. W. Unt. 1948. Types of vascular bundles in the Mono- cotyledoneae and their ia to the late metaxylem conducting elements. Sig Jour. Bot. 35: 486-4 Moore, H. E., Jr. 1966. oem a new genus of Geonomoid palms. Jour. held Arb. 47: 1-8. Morrow, L. O. 1965. Floral morphology and anatomy of certain Coryphoideae (Palmae). Ph.D. Thesis, Cornell Univ. Purr, V. 1961. The classical concept of Angiosperm carpel: a reassessment. Jour. Indian Bot. Soc. 40: 511-523 RicketTt, H. W. 1955. Materials for a rane igs of botanical terms — III. Inflorescences. eo Torrey Club 82: 419-455. SHARMAN, B. C. 1942. Developmental ie of the shoot of Zea Mays L. Ann. Bot. N.S. 6: 245-282. Swamy, B. G. L., & K. Pertasamy. 1964. The concept of the conduplicate carpel. Phytomorphology 14: 319-327. Tom Linson, P. B. 1966. Notes on the vegetative anatomy of Aristeyera spicata (Palmae). Jour. Arnold Arb. 47: 23-29. 3 pls L. H. BarLey Hortortum, CoRNELL UNIVERSITY IrHaca, New YorkK 1966] TOMLINSON, ANATOMY OF ARISTEYERA 23 NOTES ON THE VEGETATIVE ANATOMY OF ARISTEYERA SPICATA (PALMAE) 1 P. B. ToMLINSON THE FOLLOWING ACCOUNT OF THE ANATOMY of leaf and stem of Aris- teyera spicata H. E. Moore forms an appendix to the taxonomic descrip- tion by Dr. Moore and the study of the morphology and anatomy of the inflorescence axis and flowers by Mrs. Uhl being published simultaneously with this paper.” MATERIAL AND METHODS Small dried fragments of leaf-blade, petiole, and stem, without precise indication of their position, from the type collection (Steyvermark 90010) were available for study. Material of the lamina was revived, and prepara- tions in which the surface layers and internal tissues could be studied in various planes were made according to techniques outlined in Tomlinson, 1961 (p. 4). Dried fragments of petiole and stem were infiltrated directly with celloidin solvent (ethanol : ether in equal volumes) under a slight vacuum. Celloidin impregnation involved transference through a graded series of celloidin solutions (2-12%) in a 60° C oven in tightly stoppered bottles. Longitudinal and transverse sections of the hardened celloidin blocks were cut on a Reichert sliding microtome. Maceration of tissues with 10% KOH and 20% chromic acid provided isolated elements, par- ticularly of the xylem. Most observations and drawings were made from sections mounted in glycerine; some permanent coh peuae stained in safranin and Delafield’s haematoxylin were also examine ANATOMY Lamina. Dorsiventral. Adaxial epidermis smooth; abaxial epidermis prominently ribbed, ribs irregularly spaced and with a graded series from largest, prominent ribs to large veins of lamina not producing ribs (Fics. 12 and 18-21). Margin ribbed (Fic. 11). Hatrs (Fics. 3, 9, 10, 24, 25) restricted to costal regions, most frequent below large ribs, least frequent in adaxial costal regions; solitary, irregularly scattered or rarely i in pairs. * This study has been carried out with support from the National Science Founda- tion, ena GB-506. E. Aristeyera, a new genus of Geonomoid Palms. Jour. Arnold Arb. 47: con 1966. Unt, N. W. Morphology and anatomy of the oe axis and flowers of a new palm, sllbtevonh » spicata. Jour. Arnold Arb. 47: 9-22. : 24 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 Each hair a more or less uniseriate adpressed filament, always pointing distally, not or scarcely sunken, arising from a single epidermal cell; basal cells conspicuously pitted, slightly lignified with a transition io thin-walled, delicate distal cells; filament probably much longer in im- mature leaves, but shortened in mature leaves by disappearance of terminal ephemeral cells. Adaxial somewhat thicker-walled than abaxial hairs, tending to become multiseriate by transverse segmentation of fila- ment cells. ADAXIAL EPIDERMIS (Fic. 1) uniform, cells above veins not or scarcely differentiated from those elsewhere; cells in longitudinal files but markedly obliquely extended, more or less rhombohedral in surface view, without sinuous walls. Outer wall somewhat thicker than remaining walls with a thin cutinized layer penetrating somewhat between anticlinal walls. ABAXIAL EPIDERMIS somewhat shallower and thinner-walled than adaxial (14-15 compared with 20-22); subdivided irregularly into costal bands of varying width below veins (Fic. 3) and wider intercostal surface view, smaller than intercostal; intercostal cells obliquely extended but more irregular than adaxial cells. Stomata (Fics. 2 and 6-8) re- stricted to abaxial surface, diffusely distributed in intercostal regions, not in distinct longitudinal files. Terminal subsidiary cells somewhat shorter than, but otherwise not differentiated from normal epidermal cells. Lateral subsidiary cells wide, thin-walled and more conspicuously distin- guished from remaining epidermal cells. Guard cells not sunken, equal in depth to abaxial epidermis, each with two prominent cutinized ledges (Fic. 8). Wall beneath ledges somewhat thickened; cell lumen wide and otherwise unrestricted (Fic. 7). EXPANSION CELLS developed as longitu- dinal bands of enlarged cells at intervals below each surface (Fic. 21 exp.). COLORLESS HYPODERMIS absent adaxially; represented beneath abaxial surface by a compact layer of thin-walled, colorless, more or less cubical cells (Fic. 13), those below largest veins smaller than cells else- where. Hypodermis interrupted by wide substomatal chambers each sur- rounded by 4 cells appearing more or less L-shaped in surface view (Fic. 4). ASSIMILATING MESOPHYLL without distinct palisade layers; adaxial cells compact (Fics. 5, 14), somewhat transversely extended; middle mesophyll layers looser, aggregated into indistinct transverse diaphragms, forming a well-developed intercellular space system obvious only in longi- tudinal sections (cf. Fics. 13 and 14). Freers up to 18» wide, with wide septate lumina, unlignified walls; abundant throughout mesophyll, most common in adaxial layers; solitary or 2~3 forming narrow fibrous strands (Fic. 13). Adaxial fibers commonly in contact with adaxial epidermis, abaxial fibers rarely interrupting hypodermis and in contact with abaxial epidermis. VASCULAR BUNDLES (veins) either compound and situated in prominent ribs (Fic. 21) or smaller, simple and not producing pronounced ribbing of lamina (Fic. 18), with a complete transition between infre- quent large ribs and numerous small vascular bundles (Fics. 18-21). Vascular bundles independent of and equidistant from surface layers ex- cept largest bundles continuous with abaxial epidermis. Bundles each ST 1966 | TOMLINSON, ANATOMY OF ARISTEYERA 25 with a continuous uniseriate outer parenchymatous sheath of colorless, tabular cells elongated somewhat along the vein; outer sheath incomplete below larger veins. Inner sheath of somewhat thicker-walled cells, uni- seriate and indistinct in smaller veins, more pronounced in larger bundles and including lignified abaxial fibers forming a distinct phloem sheath. Fibrous phloem sheath best developed in largest ribs. Largest veins in- cluding adaxial sheathing fibers (Fics. 20, 21). Vascular tissues reduced in smaller veins (Fic. 13); collateral with a wide metaxylem vessel in larger veins (Fics. 18, 22), large veins with several separate phloem strands (Fics. 19, 20); largest veins compound with additional lateral vascular bundles resembling isolated small veins of lamina (Fics. 21, 23). Protoxylem well developed only in larger veins; phloem of smaller veins often divided into two incipiently separate strands by a narrow medium sclerotic isthmus. TRANSVERSE COMMISSURES at irregular intervals con- necting longitudinal veins; each with very reduced vascular tissues sheathed by uniseriate slightly sclerotic parenchyma (Fic. 15). MARGINAL RIB including a broad, compound vascular bundle representing fusion product of major longitudinal veins (Fic. 11). CELL CONTENTS not well preserved, no crystalline deposits observed. STEGMATA (silica cells) abun- dant next to mesophyll and vascular fibers, each with a massively thickened basal wall enclosing an irregularly spherical, spinulose silica body (Fics. 16 and 17). Bodies 9-12 wide in stegmata next to veins; smaller, 5-7 wide next to mesophyll fibers. Leaf axis (petiole) (Fics. 26-29). Harrs common abaxially, resem- bling those of lamina. Ep1perMts shallow, outer wall somewhat thickened, wholly cutinized; cells more or less rectangular in surface view, slightly elongated. Abaxial epidermis somewhat papillose towards margin. Sto- MATA occasional. GROUND PARENCHYMA uniform, cells compact, some- what elongated; only hypodermal layers slightly modified. Abaxial hypo- dermal ground parenchyma somewhat sclerotic; adaxial ground paren- chyma cells somewhat anticlinally extended, thin-walled, their walls plicately collapsed in a concertina-like manner. Hypodermal ground parenchyma including small fibrous or fibro-vascular strands (e.g. Fic. 27). VASCULAR BUNDLES in two distinct series: (1) Single ADAXIAL SERIES of large INVERTED mostly compound bundles (Fics. 26, 27), each including a large vascular strand towards the center of the petiole, the vascular tissues with inverted orientation, and one or more small strands embedded in the most fibrous part of the bundle towards the outer surface of the petiole; the vascular tissues of these supplementary strands with oblique or inverse orientation and often represented by phloem alone. (2) Re- maining bundles with normal orientation occupying bulk of petiole, divisi- ble into: (a) ABAXIAL sERtEs of alternately large and somewhat smaller bundles each with a massive fibrous phloem sheath, the largest bundles compound and including one or more supplementary strands of vascular tissue, often phloem alone, within the fibrous sheath (Fic. 28); (b) CEN- TRAL SERIES of uniformly constructed bundles with normal orientation, 26 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 more or less irregularly arranged but forming two distinct horizontal series in adaxial part of petiole (Fic. 26). Each bundle (Fic. 29) simple with a smaller fibrous sheath than in superficial bundles, including well de- veloped protoxylem, usually two wide metaxylem vessels; the phloem subdivided into at least two separate strands by a medium sclerotic isthmus. Bundles of adaxial and remaining series partially fused in leaf margin to form a large commissure. TRANSVERSE COMMISSURES connecting longitudinal bundles occasional throughout petiole. Stegmata, crystals as in lamina. Stem.* PrERImDERM formed by suberisation of outer layers augmented by etagen meristematic activity in outer cortex. CorTEX narrow, up to 2 mm. wide including numerous narrow fibrous strands transitional internally to larger strands including either phloem or both xylem and phloem tissues. No distinct limiting layer for inner cortex. CENTRAL CYLINDER with periph- eral congested vascular bundles, each bundle with well developed fibrous phloem sheath, the bundles often somewhat rhombohedral in transverse section. Obvious leaf traces frequent. Central vascular bundles some- what less congested but with scarcely reduced fibrous sheaths. VASCULAR BUNDLES (Fic. 32) almost invariably with one wide metaxylem vessel, with or without protoxylem; two vessels appearing in bundles close below their exsertion as leaf traces. GROUND PARENCHYMA Cells slightly thick- walled, conspicuously pitted, including abundant starch. Stegmata abun- dant next to vascular fibers (Fic. 34). Secretory and conducting cells. RApHtpE-sAcs infrequent in all parts. StecMATA (silica cells) abundant next to fibers in all parts, silica bodies smallest (5p diameter ) in cells next to mesophyll fibers of lamina, largest (12 diameter) in cells next to vascular fibers of stem. STARCH abundant in ground parenchyma of stem; grains simple, more or less spherical, up to 15» diameter. VESSELS present in metaxylem of stem, rachis and large ribs of lamina; larger elements in leaf mostly more than 2 mm. long, 70” wide with scalariform perforation plates with many thickening bars (Fics. 30, 31) on oblique or very oblique end walls; elements in stem (Fic. 33) average 860 long, 120, wide with scalariform perforation plates with 4-10 thickening bars on slightly oblique or oblique end walls. SIEVE-TUBES in stem with slightly oblique compound sieve- plates coinciding with vessel-element end walls; diameter 15—20,. DISCUSSION A limited amount of information on vegetative anatomy of some other genera of Geonomeae: Calyptrogyne, Calyptronoma, Geonoma, Pholido- stachys and Welfia is available in Tomlinson (1961), and suggests that the group shares a few diagnostic anatomical features. Most distinctive is the filamentous uniseriate (rarely biseriate) hair, commonly restricted to * General stem construction corresponds to that described for Rhapis by Zimmer- mann and Tomlinson, 1965. 1966 | TOMLINSON, ANATOMY OF ARISTEYERA 27 costal regions and most frequent on the abaxial surface of the lamina. Those illustrated for Aristeyera (Fics. 9, 10, 24 and 25) are quite typical. Hairs are more uniformly dispersed in Calyptrogyne and Welfia. Those in Calyptronoma and Welfia are somewhat distinctive in being sunken and this is associated with enlargement of the basal cell. Elsewhere in palms uniseriate filamentous hairs are rare. They occur in the bactroid palms which on morphological and anatomical grounds, however, show no special affinity with the Geonomeae. Reinhardtia and Sclerosperma may have hairs somewhat similar to those in Calyptronoma and Welfia. Other anatomical features which are common to those geonomoid palms which have so far been examined and which are illustrated for Aristeyera above are: (1) shape of epidermal cells which are always thin-walled except in Calyptronoma, (2) structure and distribution of stomata, (3) structure and distribution of stegmata and to a certain extent the struc- ture of assimilating mesophyll, veins, and ribs. The degree of differentia- tion of hypodermal mesophyll layers as specialized colorless layers varies from: (1) Welfia with a distinct 2-layered colorless hypodermis below each surface, to (2) Calyptronoma and, less conspicuously, Pholidostachys with a 1-layered hypodermis below each surface, to (3) Aristeyera with a 1-layered abaxial hypodermis, to (4) Calyptrogyne and Geonoma with no colorless hypodermis. Non-vascular fibers vary in their distribution. They are (1) absent from some Geonoma species, (2) solitary or in small strands scattered more or less throughout the mesophyll as in Aristeyera, Calyp- trogyne, Calyptronoma, and some Geonoma species, (3) infrequent as small strands in a single series in Pholidostachys, and (4) distinctive in Welfia as large strands pectinating with the vascular bundles. Mesophyll fibers throughout the alliance tend to be septate, unlignified and with a wide cell lumen. It is quite clear, therefore, even from the limited information available, that although the members of the Geonomeae share some common fea- tures they are anatomically quite diverse. Welfia seems rather isolated. Aristeyera shows no special anatomical affinity with Welfia although these two genera are most alike in floral morphology. Without further informa- tion on many more species of this tribe further speculation about generic affinity on an anatomical basis is pointless. Nevertheless, the above account has demonstrated that there are ana- tomical features of diagnostic value in this tribe and continued investiga- tion should reveal more. In view of the size of this assemblage, in the region of 200 species according to Dr. Moore, and considering the small sample for which anatomical information is available (about 10 species and for these mostly fragments of the lamina), one feels justified in stating that the anatomy of the Geonomeae is largely a terra incognita. SUMMARY A description of the anatomy of lamina; petiole and stem of Arvisteyera spicata based on small fragments from the collection Steyermark 90010 28 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 supplements the formal taxonomic treatment and the account of floral anatomy. Aristeyera shows anatomical features which hitherto have appeared to be diagnostic for the Geonomeae. Without information about many more species from this tribe it is impossible to comment on the affinities of Aristeyera from an anatomical point of view except that it shows less resemblance to We/fia than to other geonomoid genera. LITERATURE CITED Tomitnson, P. B. 1961. Anatomy of Monocotyledons, ed. C. R. Metcalfe, Vol. Il. Palmae. Oxford. aananrniny M. H., & P. B. Tomiinson. 1965. Anatomy of the palm, Rhapis xcelsa 1. Mature vegetative axis. Jour. Arnold Arb. 46: 160-180. 1965. FAIRCHILD TROPICAL GARDEN Mramt, FLorwa 33156 AND CaBotT FOUNDATION HARVARD UNIVERSITY EXPLANATION OF PLATES (The abbreviations used on the figures are cited here with their equivalents: m : , c xylem; phl. (dotted) = phloem; pxy. (open circles) = protoxylem; sc. par. (lined) = sclerotic parenchyma of bundle sheath; scl. = sclerenchyma; xy. = xylem; xy. par. (white) = xylem parenchyma.) PLATE I Ficures 1-11, Lamina of Aristeyera spicata. Fics. 1-5, Surface views, * 220. Fic. 1, Adaxial epidermis. Fic. 2, Abaxial epidermis, intercostal region. Fic. 3, Abaxial epidermis, costal region. Fic. 4, Abaxial hypodermis. Fic. 5, Adaxial mesophyll, hypodermal layer. Fics. 6-8, Stomata, x 440. Fic. 6, Surface view. Fic. 7, Longitudinal section. Fic. 8, Transverse section. Fics. 9 and 10, Abaxial hairs, X 440. Fic. 9, Transverse section. Fic. 10, Longitudinal section. Fic. 11, Lamina margin, transverse section, X 32 (cf. FIGs. 18-21). PLATE II Figures 12-25, Lamina of Aristeyera spicata (continued). Fic. 12, Trans- verse section, 32, diagrammatic representation of the distribution of veins section of lamina. Fics. 16 and 17, Stegmata, * 440, next to isolated mesophyll fibers. Fics. 18-21, Transverse section, 32, of veins of increasing size; vascular and mesophyll fibers shown in solid black, xylem lined, phloem dotted, chlorenchyma as short lines. Fic. 22, Ege section of vein, 220, en- largement of vein shown in FIG. 18. Fic. Transverse section of small vascular bundle, < 220, enlargement of small wee bundle from Fic. 21. Fics. 24 and 25, Hairs shown in surface view, X 440. Fic. 24, Abaxial hair. Fic. 25, Adaxial hair. 1966 | TOMLINSON, ANATOMY OF ARISTEYERA 29 PLATE III Ficures 26-34, Petiole and stem ea Aristeyera spicata. Fics. 26-31. Petiole. Fic. 26, Transverse section, < 6, incomplete, complete outline indicated by dotted line. Inverted bundles es in solid black, normally oriented bundles lined. Fics. 27-29, Enlarged details, X 60, of vascular bundles from Fic. 26. Fic. 27, Adaxial, inverted bundles. Fic. 28, Abaxial bundle. Fic. 29, Central bundle. Fics. 30 and 31, Metaxylem vessel element. Fic. 30, Outline of end wall, X 80. Fic. 31, Details of perforation plate, xX 440. Fics. 32-34, Stem. Fic. 32, Transverse section of peripheral vascular bundle, * 60. Fic. 33, Metaxylem vessel elements from stem, X 80, perforations seen directly represented by solid black; perforations seen through an imperforate wall represented by lines. Fic. 34, Stegmata next to stem fiber, < 440. PLATE I Jour. ARNOLD Arp, VOL. 47 ToMLINSON, ARISTEYERA SPICATA Jour. ARNOLD Ars. VoL. 47 PLATE II =e ~e : a he BESSY LB OG Sage ee ie, TOMLINSON, ARISTEYERA SPICATA Jour. ARNOLD Ars. VOL. 47 Prate III Be TOMLINSON, ARISTEYERA SPICATA 1966] FERGUSON, GENERA OF CAPRIFOLIACEAE 33 THE GENERA OF CAPRIFOLIACEAE IN THE SOUTHEASTERN UNITED STATES ! I. K. FERGUSON CAPRIFOLIACEAE A. L. de oe ia Pl. 210. 1789, “Caprifolia,” - con a. ak Deciduous [or evergreen] shrubs, sometimes small trees or vines, rarely herbs, with opposite [or alternate], entire or divided, exstipulate or stipu- late leaves. Inflorescence a corymb, cyme, thyrse, or spike of whorls, or flowers solitary or in pairs, usually bracteate and bracteolate. Flowers regular or irregular, epigynous, bisexual, rarely some sterile. Floral tube adnate to the ovary, with a distinct constriction at the level of the 3—5 usually small calyx lobes. Corolla sympetalous, 3—5-lobed, salverform, rotate, campanulate, funnelform, or tubular, often bilabiate, sometimes gibbous at or near the base; aestivation usually imbricate. Stamens 5 [4], epipetalous, alternate with the corolla lobes; filaments long or short; anthers 2-locular, longitudinally dehiscent, oblong or mained versatile dorsifixed, introrse or rarely extrorse; pollen tricolpate. Gyn 1—5- carpellate, syncarpous; style | or wanting; stigmas as many as rhe quran distinct or united; ovary inferior, 2—5-locular; locules with solitary to Fruit a berry or drupe, rarely a capsule [or achene], 1—5-locular, with l-co seeds or as many stones (pyrenes, nutlets) as locules. Endospe copious, solid or ruminate; embryo straight, small, situated near the "Prepared for a generic flora of the southeastern United States, . beara project of the Arnold Arboretum and the Gray Herbarium of Harvard University made possible i ee yee and Reed C. Rollins. This treatment follows the pattern established in the first paper in the series (Jour. Arnold Arb. 39 296-346. 1958) and continued through those in volumes 40-46 (1959-1965). The es d South Carolina, Georgia, Florida, Tennessee, Ala- bama, Mississippi, Arkansas, and Louisiana. The Poca iein are based primarily on the plants of this area, with supplementary material in brackets. References which the author has not seen are marked by an asterisk. The author is indebted to Dr. Wood for his aid and valuable criticisms; to Dr George K. Brizicky for his guidance and suggestions; to Dr. Rolla M. Tryon advice on some of the nomenclature; to Dr. Ric hard A his unpublished data on petiolar anatomy in Viburnum; a were prepared by Arnold D. Clapman. Parts of materials used for illustration were collected by R. J. Eaton, P. S. Green, M. else and C. E. Wood, Jr. 34 JOURNAL OF THE ARNOLD ARBORETUM [voL, 47 hilum, radicle terete, cotyledons ovoid or oblong. (Including Sambucaceae Link, Handb. 1: 662. 1829, “Sambucinae.”) Type cENus: Caprifolium Mill. = Lonicera L. A family of about 18 genera and 500 species, primarily of the North Temperate Zone but extending to the mountains of the Tropics, South America, Australia, and New Zealand; seven genera native in North America; six in our area. Caprifoliaceae are distinguished from most other families by the syn- carpous, inferior ovary and opposite leaves. The presence of stipules is probably the most useful character for distinguishing the very closely re- lated Rubiaceae, but there are exceptions, and there seems to be no single character which separates the two families consistently. Airy-Shaw has at- tempted to differentiate them on the basis of the scaly buds of the Capri- foliaceae versus the naked buds of Rubiaceae, but there are numerous exceptions in Viburnum. Some authors favor uniting the two families, but Kern & Van Steenis do not think this desirable without a full investigation, since many of the genera of the Caprifoliaceae would be placed in different tribes of the Rubiaceae and the identity of the family lost. These authors have drawn attention to small but important characters common to all Caprifoliaceae: the presence of enlarging calyx lobes, a distinct constric- tion below the limb of the calyx, dorsifixed anthers, and the absence of colleters (glandular trichomes) which occur on the inner surface of the stipules of many rubiaceous genera. The Asiatic genera Carlemannia Benth. and Silvianthus Hook. have been assigned to both the Rubiaceae and the Caprifoliaceae, but doubtfully belong to either family (see Airy- Shaw, 1965). It is problematical whether the New Zealand Alseuosmia Cunn. and the New Caledonian Pachydiscus Gilg & Schlechter and Memecylanthus Gilg & Schlechter belong to the Caprifoliaceae where they have been placed. Viburnum and, to a lesser extent, Sambucus show affinities with the Cornaceae, a relationship supported by pollen mor- phology. These genera also appear to be related to the Valerianaceae, valerianic acid occurring in both. In addition, the leaves of Sambucus closely resemble those of many of the Valerianaceae. The tribe Linnaeeae may also show affinities with the Valerianaceae, for Abelia resembles the Asiatic genus Nardostachys in some anatomical characters. The Caprifoliaceae have been subdivided variously, Bentham & Hooker placed Viburnum and Sambucus together in the tribe Sambuceae HBK. ex DC. and the remainder of the genera in the tribe Lonicereae R. Br. ex - DC.; and pollen morphology and biochemical studies do indeed indicate a close relationship between Sambucus and Viburnum. A better treatment, however, appears to be that of Fritsch, who separated the two, recognizing the tribes Sambuceae, Lonicereae bee Caprifolieae), Linnaeeae Fritsch, and Viburneae Fritsch. Meyer had previously segregated Diervilla and Weigela as tribe Diervilleae, and Wagenitz followed this course. Reduction in the number of carpels and ovules seems to have been a feature in the evolution of the genera of the Caprifoliaceae. The basic 1966 | FERGUSON, GENERA OF CAPRIFOLIACEAE 35 chromosome numbers appear to be eight and nine, and changes in chromo- some number presumably have played only a minor role in differentiation of the genera. The cytogenetics of Lonicera and Viburnum have been investigated quite extensively, but other genera have been neglected. No chromosome counts are available for Triosteum, Leycesteria, and Alseu- osmia and few for Symphoricarpos (two of 16 species) and Abelia (two of 30 species). Pollination is primarily by insects which are attracted by the conspicuous corollas, the flowers aggregated into heads, and the usually abundant nectar. The family is of little economic importance, aside from the many species of Lonicera, Weigela, Viburnum, Leycesteria, Abelia, Symphoricarpos, and Sambucus cultivated as ornamentals, Abelia & grandiflora (André) Rehd. (A. chinensis & A. uniflora) is one of the most widely cultivated orna- mental shrubs of the southeastern United States. REFERENCES: Arry-SHaw, H. K. A revision of the genus Leycesteria. Bull. Misc. Inf. Kew 1932: 161-— sa 1932 . Onan wr Species of the genus Silvianthus Hook. f. and on the family Elaine Kew Bull. 19: 507-512. 1965. ARTIUSHENKO, Z. T. Morphological nature of the lower ovary in some repre- sentatives of the Caprifoliaceae family. (In Russian.) Bot. Zhur. 33: 202- 212. 1948. [Sambucus, Symphoricar pos. | . Development of blossom and fruit in the honeysuckle family (Capri- foliaceae). (In iain Acta Inst. Bot. Acad. Sci. URSS 7. Morphol. Anat. Pl. 2: 131-169. BaILiton, H. Recherches sur |’ en ee le développement et l’anatomie des Caprifoliacées. Adansonia 1: 353-380. pl. 12. 1861. Rubiacées. Hist. Pl. 7: 257-503. 1880. [Diervilleae, Lonicereae, Sumnbuceae, 497-502. ] BENTHAM, G., & J. D. Hooker. Caprifoliaceae. Gen. Pl. 2: 1-7. 1873. BREMEKAMP, C. E. B. On the position of the genera Carlemannia Benth. and Sylviantus [Sylvianthus] Hook. f. Rec. Trav. Bot. Néerl. 36: 372. 1939. CanpDoLte, A. P. ve. Caprifoliaceae. Prodr. 4: 321-340. 1830. Crarkson, R. B. Twinflower in West Virginia. Castanea 23: 110. 1958. [Linneea borealis L. var. americana (Forbes) Rehd. Cooper, T. B. A study of the has in the Cudhaliass: Trans. Proc. Bot. Soc. Edinb. 32: 548-555. 1939 CreteE, P. Origine et structure ms sac embryonnaire, de l’albumen et du tégu- ment séminal chez le Leycesteria formosa Wall. (Caprifoliacées). Bull. Soc. Bot. France 101: 130-133. 1954 CRUTCHFIELD, P. J. Taxa collected from Roanoke Island new to the flora of North Carolina. Castanea 29: 129-137. 1964. [Abelia grandiflora per- sisting after cultivation, 136.] Demi, L. On the germinal nature of the glandular cells presenting meiotic phenomena in the basal stylar canal in the genera Viburnum, Sambucus, and Adoxa Poae geen (In Italian; English summary.) Caryologia 5: 378-403. pls. 26-30. 1952/1953. Dovctas, G. E. The inferior ovary. II. Bot. Rev. 23: 1-46. 1957. [Capri- foliaceae, 18, 19, 44, 45.] 36 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 Fenc, Y. A. Recherches cytologiques sur la caryocinése la spermatogénése et la fécondation chez les Caprifoliacées. Botaniste 26: 1-85. pls. 1-8. 1934. [Centrosomes in Lonicera and Viburnum. See also ibid. 24: 335-352. pl. 30. 1932; Jour. Agr. Assoc. China 128: 127-131. 1934 *.] FritscH, K. Caprifoliaceae. Nat. Pflanzenfam. IV. 4: 156-169. 1891. —— ie Gattungen der Caprifoliaceen und demonstrirte Vertreter dieser Gattungen in Herbar-Exemplaren. Bot. Centralbl. 50: 137-139, 168-170. 1892. [See also Akad. Wiss. Wien Sitz-ber. 42: 7-10. 1892.] Garcin, A. G. Recherches sur l’histogénése des péricarpes charnus. Ann. Sci. Nat. eo VII. 12: 175-401. pls. 23-26. 1890. [Caprifoliaceae, 333-349. pls. 2 singe R. 4 je chemistry and phylogeny of flowering plants. Trans. y. Soc. Canada III. Biol. Sci. Sect. 5. 48: 1-47. 1954. [Caprifoliaceae, pea GUNDERSEN, A. L. Recherches anatomiques sur les Caprifoliacées. Thése. 79 p. pls. 1-4. Paris. 1910 Hara, H. Contributions to the study of variations in the Japanese plants closely related to those of Europe or North America. Part 2. Jour. Fac. Sci. Univ. Tokyo Bot. 6: 343-391. 1956. [Caprifoliaceae, 381-. -385.] wee E. J. P. The Caprifoliaceae of Ohio. Ohio Jour. Sci. 65: 118-129. bee N. P. A contribution to a knowledge of the organogeny of the flower and of the embryology of the Caprifoliaceae. Proc. AAAS 49: 280, 281. 1900. [ Viburnum. | Horne, A. S. A contribution to the study of the evolution of the flower, bab special reference to the Hamamelidaceae, Caprifoliaceae, and Cornace Trans, Linn. Soc. II. Bot. 8: 239-309. pls. 28-30. 1914. Kern, J. H., & C. G. G. J. vAN SteEENIs. Caprifoliaceae. Jn: VAN STEENIS, Fl. Males. I. 4: 175-194. 1951. re . Beitrage zur vergleichenden Anatomie der Caprifoliaceen. Verh. Zoo t. Ges. Wien 45: 43-68. pl. 6. : LovELL, 4 i The visitors of the Caprifoliaceae. Am. Nat. 34: 37-51. 1900. MicuHaeEL, P. O. Vergleichende Untersuchungen iiber den Bau des Holzes der Compositen, Caprifoliaceen und Rubiaceen. Bot. Jahrb. 7(Lit.): 50. 1885. Morsst, E. Vergleichende embryologische Studien iiber die Familie der Capri- foliaceae. Osterr. Bot. Zeitschr. 90: 153-212. 1941. Nakal, T. Tentamen Sie ee Japonicarum. Jour. Coll. Sci. Univ. Tokyo 42(2): 1-139. 193 Persipsky, D. The formation of ee sterile loculi and rudimentary ovules in specimens of the family Caprifoliaceae. (In Ukrainian; summaries in Russian & English.) Jour. Inst. Bot. Acad. Sci. RSS Ukr. 12: 107-112. 1937. Gynoeceum evolution in the family Caprifoliaceae. (In Ukrainian; summaries in Russian & English.) /bid. 21/22: 45-105. 1939. Prxier, V. A. The Caprifoliaceae of West Virginia. Castanea 15: 80-91. 1950. Piouvier, V. Sur l'étude biochimique comparée de quelques Caprifoliacées. Compt. Rend. Acad. Sci. Paris 232: 1013-1015. 1951. Poucgues, M. L. pe. Recherches caryologiques sur les Rubiales. Revue Gén. Bot. 56: 5-27, 74-138, 172-188. pls. 1-10. 1949. [Caprifoliaceae, 77-101, pls. 3-7.] Rypserc, P. A. The North American twinflowers. Torreya 1: 52-54. 1901. [ Linnaea. | 1966 | FERGUSON, GENERA OF CAPRIFOLIACEAE 37 Sax, K., & D. A. Kriss. aera = phylogeny in Caprifoliaceae. Jour. Arnold Arb. 11: 147-153. pl. 24. SCHERMERHORN, J. W., & M. W. Qu ise eds. Lynn Index. Monograph V. 67 pp. Boston. 1962. pee 11-19 SCHLECHTER, R., & E. Gric. Caprifoliaceae. se R. SCHLECHTER, Beitrige zur Kenntnis der Flora von Neu-Kaledonien. Bot. Jahrb. 39: 268-270. 1907. [Notes on Memecylanthus and Pachydiscus and their relationships with Alseuosmia. STEYERMARK, J. A. Flora of Missouri. Ixxiii + 1725 pp. Ames, Iowa. 1963. [Caprifoliaceae, 1405-1419. ] TrEGHEM, P. van. Sur le réseau sus-endodermique de la racine des Capri- foliacées. Bull. Soc. Bot. France 34: 251-253. 1887. . Remarque sur mp it age de l’embryon des Caprifoliacées. Ann. Sci. Nat. Bot. IX. 7: 128. 1908. Vipat, L. Sur la oa et le développement du pistil et du fruit des Capri- foliacées. Ann. Univ. Grenoble 9: 583-601. Vos, F. pe. The stem anatomy of some species of the Caprifoliaceae with reference to phylogeny and identification of the species. Ph.D. Thesis, 81 pp. Cornell Univ. 1951.* WaceEnITz, G. Die systematische Stellung der Rubiaceae. Bot. Jahrb. 79: 17- 35. 1959. [Gentianales = Contortae + - i Dipsacales = Capri- foliaceae, Adoxaceae, Valerianaceae, Dipsac . Caprifoliaceae. Jn: H. MELCHIOR, Engler’s Syllabus der Pflanzen- familien. ed. 12. 2: 473-475. 1964. WEBERLING, F. Morphologische Untersuchungen zur Systematik der Capri- foliaceen. Akad. Wiss. Lit. Abh. Math.-Naturw. Mainz 1957; 1-50. 1957. [See also Bot. Jahrb, 77: 458-468. 1958.] Wivxkrnson, A. M. Floral anatomy and morphology of some species of the tribe Lonicereae of the Caprifoliaceae. Am. Jour. Bot. 35: 261-271. 1948. . Floral anatomy and morphology of some sage of the tribes Linnaeeae and Sambuceae of the Caprifoliaceae. [bid. 3 fe Floral anatomy and morphology . . and of the Capri- foliaceae in general. Jbid. 36: 481-489. 1949. Key TO THE GENERA OF CAPRIFOLIACEAE General characters: shrubs or trees, rarely herbs, mostly with opposite, ex- floral carpous, with a 2—5-locular, inferior ovary; endosperm copious, embryo small, Straight. A. Corolla rotate or subrotate, deeply 5-lobed; style very short or absent; stigmas 1-5. B. “otha pinnately compound; drupe with 3-5 stones. .... 1. Sambucus. simple (sometimes lobed); drupe with 1 stone. .. 2. Viburnum. A. Cosulla patent tubular, funnelform, or bilabiate; style elongate; stigma capitate. ; C. Plants coarse, erect, herbaceous perennials; fruit a dry ge with 3 SLT t eee Weg eee Steer MMSE a, cts Cl eset pg ea MC Cb auto Sas A Triosteum. C. Plants shrubs or woody vines; fruit a capsule, fleshy Fake or berry. 38 JOURNAL OF THE ARNOLD ARBORETUM i [voL. 47 D. Corolla companulate; ovary 4-locular; fruit a drupe with 2 stones. PEST OPO eer Cette eels cee TS 4. Symphoricarpos. D. Corolla funnelform, — or bilabiate; ovary 2- or 3(—5)-locular; fruit several- or many-see E. Ovary slender; calyx ‘ies linear-lanceolate; fruit a capsule. Di Uae eh © Seeley cee eee PPA e eae ne eee rr 5. Diervilla. E. Ovary globular to ovoid; calyx lobes short-triangular; fruit a CN sr ss ah eh ed So eee doe are oeneres 6. Lonicera. Tribe SampuceaE HBK. ex DC. 1. Sambucus Linnaeus Sp. Pl. 1: 269. 1753; Gen. Pl. ed. 5. 130. 1754. Rank-smelling deciduous shrubs or small trees [rarely perennial herbs]. Stems with a large pith. Winter buds with several pairs of scales. Leaves pinnately or bipinnately compound and usually with serrate leaflets, ex- stipulate or stipules small [or conspicuous] or reduced to glands. In- florescence a corymb, cyme, or thyrse, usually terminal, bracteolate, the pedicels articulate at apex. Flowers small, white or yellowish, regular, bisexual. Calyx lobes 5 [3], small. Corolla rotate, 5[3 or 4|-lobed, lobes imbricate [rarely valvate], ovate [to oblong-lanceolate]. Stamens 5, inserted toward the base of the corolla; filaments slender, subulate; an- sem ry) stamens removed to aor solitary, pendulous, aanedece ovule in each locule, X 10; e, fruit, X 3; f, stone, 9. 1966 | FERGUSON, GENERA OF CAPRIFOLIACEAE 39 thers short, oblong, extrorse; pollen small to medium sized, reticulate. Stigma sessile, red or white, 3—5-lobed; ovary 3—5-locular with a solitary, pendulous ovule in each locule. Fruit a small ovoid or globular drupe with 3—5 cartilaginous, 1-seeded stones; stones oblong, compressed, rugu- lose or smooth. Seeds with a membranaceous testa and fleshy endosperm. Embryo sac development of the Adoxa type. LectoTyPeE sPEctrs: S. nigra L.; see Britton & Brown, Illus. Fl. No. U. S. ed. 2.3: 268. 1913. (Classical Latin name for the plant; believed to be from Greek, sambuké, a stringed musical instrument supposed to be made of the wood of the elder, S. nigra.) — ELDER. A widespread genus of about 20 species, centered chiefly in the temperate regions of the Northern Hemisphere but extending south to the mountains of the Tropics and to East Africa, South America, Australia, and New Zealand; about ten species in North America, including the introduced Sambucus nigra and S. Ebulus L.; two or three species in our area. Sambucus racemosa L. is widespread in cool-temperate regions of the Northern Hemisphere. In North America it is represented by subsp. pubens (Michx.) Hultén (S. pubens Michx.), 2n = 36, which occurs throughout much of Canada and the northern United States and south- ward in the mountains to North Carolina, Tennessee, and northern Georgia. The paniculate inflorescences, red fruits, brown pith, and early spring flowering are characteristic. It is closely related to the European subsp. racemosa, 2n = 36, separated by its denser inflorescences and glabrous leaves, and to the Japanese subsp. Sieboldiana (Miq.) Hara, distinguished by its reddish stigmas and small fruits. Some authors have treated these taxa as distinct species; their status needs further investiga- tion. Sambucus canadensis L., 2n = 36, widespread in our area and extending throughout the greater part of North America, contrasts with S. racemosa in its umbelliform inflorescences, black fruits, white pith, and later flower- ing time in areas where both occur. Sambucus Simpsonii Rehd., a variant which occurs from the West Indies to Florida and north to southern Georgia, Alabama, Mississippi, and Louisiana, is very doubtfully distinct from S. canadensis. Rehder distinguished this species by its treelike habit, smooth bark, narrow pith, and usually five-locular ovary, in contrast with S. canadensis with shrubby habit, rough bark, wide pith, and usually four- locular ovary. Investigation of some of Rehder’s material shows that the number of locules in the ovary may vary from three to five but is pre- dominantly four. The other differences are perhaps a result of more favorable climatic conditions. Later authors have called attention to the pinnately divided lower leaflets of S. Simpsonii, a characteristic which is common in peninsular Florida, although northward numerous intermediates with a variable number of leaflets occur. Brown reports the frequent occurrence in Louisiana of specimens which have occasional leaves with the lower leaflets divided as in S. Simpsonii but with most of the leaves on the same branch like those of S. canadensis. 40 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 The pinnate leaves and extrorse anthers, as well as numerous small anatomical characters, make Sambucus a very distinct genus, but its taxonomic position is not clear. It shows affinities with Viburnum in floral characters, and some authors (e.g., Bentham & Hooker and Fer- nald) associate these genera as the tribe Sambuceae. Others (eg., Fritsch and Wagenitz) regard Sambucus as constituting a separate tribe, while Hock has segregated it as the family Sambucaceae. Van Steenis has drawn attention to a marked structural resemblance between Sambucus and Turpinia Vent. and suggested that a fuller investigation should be made of the relationships between Sambucus and the Staphyleaceae. Pollination is probably by insects, although few pollinators have been reported for some species and floral nectaries are absent. Both proterandry and proterogyny have been observed, and both self- and cross-pollination may occur. Chromosome numbers of 2n = 32, 36, and 38 have been Several species are grown as ornamentals, and ripe fruits of some, including Sambucus canadensis, are used for making jellies and wines. The hard wood of S. nigra has been used in making musical instruments, as a source of pegwood and charcoal in Europe, and in various other ways. The large pith has been used both as a cleaning agent for the pivots of delicate machinery and for holding botanical specimens in the preparation of freehand sections. REFERENCES: Under family references see ARTIUSHENKO, Demi, Hara, Motsst, PERSIDSKY (1937, 1939), WILKINSON (1948, pp. 365-371) Brown, C. A. Louisiana trees and shrubs. La. Forestry Commiss. Bull. 1. x + 262 pp., front. Baton Rouge. 1945. [S. Simpsonii, 232, 233.] CAMMERLOHER, H. Bliitenbiologie I. Sammlung Borntraeger 15. 199 pp. pls. 1, 2. Berlin. 1931. [Extrafloral nectaries in S. javanica, 46-48.] CHAMPAGNAT, P. L’autonomie de croissance des entre- eiaeiias chez le sureau (Sambucus nigra L.). Compt. Rend. Acad. Sci. Paris 228: 1965-1967. 1949. DamMer, U. Die extrafloralen Nectarien an Sambucus nigra. Osterr. Bot. Zeitschr. 40: 261-264. 1890. Davis, O. H. Germination and early growth of Cornus florida, Sambucus cana- densis, and Berberis Thunbergii. Bot. Gaz. 84: 225-263. 1927. Dotcuer, T, La meiosi e lo sviluppo del gametofito nell’ovulo di Sambucus nigra L. (Caprifoliaceae). (English summary.) Caryologia 4: 58-76. 1951. sia K. Vascularization of the vegetative shoots of Helianthus and Sambucus. Am. Jour. Bot. 32: 18-29. 1945. [S. glauca Nutt.] FarMER, J. B. On the development of the endocarp in Sambucus nigra. Ann. Bot. 2: 389-392. 1888. FERNALD, M. L. Recent discoveries in the Newfoundland flora. Rhodora 35: 298-315. 1933. [Distinguishes S. pubens from S. racemosa, 310. GrERARD, W. R. Origin of the name Sambucus. Garden Forest 8: 368. 1895. Hocx, F. Zur systematischen Stellung von Sambucus. Bot. Centralbl 51: 233, 234. 1892. [Proposes a separate family, Sambucaceae, between Capri- foliaceae and Valerianaceae. Horm, T. Medicinal plants of North America. 32. Sambucus canadensis L. 1966] FERGUSON, GENERA OF CAPRIFOLIACEAE 41 oo Rep. 18: 259-262. 1909.* [See also Bot. Jahresb. 37(1): 993. 12.] fae B. L. VAN DER. Staphyleaceae. Ju: C. G. G. J. van STEENIs, FI. Males. I. 6: "49-59. 1960. [Note by VAN STEENIS on a possible relation- ship between Sambucus and Staphyleaceae, 51.] MetcaLre, C. R. The elder tree (Sambucus nigra L.) as a source of pith, peg- wood and charcoal, so some notes on the structure of the wood. Kew Bull. 3: 163-169. pl. 3. 1948. MULLER, C. Ueber ee der Bliithen von Sambucus australis Cham. et Schltdl. Ber. Deutsch. Bot. Ges. 2: 452-456. 1884 REIcHEL, L., H. H. Strou, & W. ReicHwaLp. Uber die Farbstoffe der schwar- zen Holunderbeere. ye aia 44: 468. 1957. [Notes on Ritter, C. M., & G. W. McKee. The elderberry. History, classification, and culture. Bull. Pa. State Coll. Agr. Exp. Sta. 709. 24 pp. 1964. [S. canaden- sis. Sanpo, C. S. The isolation and identification of rutin from the flowers a elder (Sambucus canadensis L.). Jour. Biol. Chem. 58: 737-745. SARGENT, C. S. Sambucus. Silva N. Am. 5: 85-92. pls. 221, 222. 1893. . Sambucus leiosperma. ne os 10: 174. 1897 ScuueTTe, H. A., H. A. Vocet, & J. A. Batn. The saturated fatty acids of elderberry seed oil. Oil Soap ‘20: 46. oe * [S. canadensis. | ScHtruHorr, P. N. Uber regelmassiges Vorkommen zweikerniger Zellen an den Griffelkanalen von Sambucus. Biol. Centralbl. 36: 433-439. 1916. SCHWERIN, F. von. Monographie der Gattung Sambucus. Mitt. Deutsch. Dendrol. Ges. 18: 1-56. 1909. Revisio generis Sambucus. Ibid. 29: 194-231. 1920. SovEcEs, R. Embryogénie des Caprifoliacées. Développement de l’embryon chez le Sambucus nigra L. Compt. Rend. Acad. Sci. Paris 248: 1072-1078. Warnes, ~ C. Two more Sierran plants in southern California. Madrofio 2: 158. 1934. [S. racemosa. ]} YARDIN, Sur la présence d’alcaloides dans les omni Compt. Rend. Soc. Biol. Paris 122: 155, 156. 1936. [.S. migra, S. racemo Yates, W. Natural history notes. Jour. Proc. Hamilton pi 25 /26: 122-124. 1910. [Notes on S. canadensis and S. pubens and fossil representatives of the genus. ] Tribe VIBURNEAE Fritsch 2. Viburnum Linnaeus, Sp. Pl. 1: 267. 1753; Gen. Pl. ed. 5. 129. 1754. Deciduous [or evergreen], erect or decumbent shrubs or small trees. Winter buds naked or with scales. Leaves membranaceous to subcoria- ceous, opposite or rarely whorled, entire, crenate, dentate, or lobed, gla- brous or hairy (often with stellate tomentum), petiolate, exstipulate or stipulate, or the stipules partially adnate to the petiole [or sometimes re- duced to glands]. Inflorescences terminal or axillary, compound, umbel-like or paniculate, cymose; bracts and bracteoles usually small, caducous. Flowers regular, bisexual, sometimes the marginal flowers of the inflores- cence radiate, slightly irregular, and neutral. Calyx lobes 5, persistent. 42 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 Corolla rotate or subrotate [to campanulate], 5-lobed, white (rarely cream or pink), imbricate in bud. Stamens 5, inserted at the base of the corolla; filaments long or short, slender; anthers oblong, introrse, 4-lobed; with a single pendulous ovule. Fruit an ellipsoid, ovoid, or globose, 1-seeded drupe crowned by the persistent calyx lobes and style; mesocarp with bitter or sweet flesh; stone globose or laterally compressed, oblong-elliptic to ovate, often grooved or furrowed, the edges inflexed or undulate. Seed with copious, fleshy, solid or ruminate endosperm; embryo minute, linear, situated near the micropyle. Embryo sac development normal (Polygonum type). Lecrotypr species: V. Lantana L.; see Hitchcock & Green, Int. Bot. Congr. Cambridge 1930. Nomencl. Propos. Brit. Bot. 142. 1929.7 (Classical Latin name for the genus, thought to refer to V. Lantana; also said to be from Latin, viere, to bind, because of the pliability of the branches of some species.) — ARROW-WOOD. A complex genus of about 250 species, mainly of the North Temperate Zone, but occurring south to Central and South America, Java, and Timor; approximately 25 species in North America, including three naturalized from Asia; about 13 (in four sections) native in our area. Rehder distinguished nine sections based primarily on the shape and furrowing of the stone, number of scales on the winter buds, and leaf venation. Unpublished data from studies of petiolar anatomy by R. A. Howard support the separation of some of these sections. Rehder’s treat- ment, however, does not include the species of Central ae South America and Malesia, and a full investigation of the genus is n Section PsEUDOTINUS Clarke (‘“Alnifolia” Small), aistingusd by the presence of radiate flowers, naked winter buds, and ruminate endo- sperm, includes three species of eastern Asia and one of eastern North America. Viburnum lantanoides Michx. (V. alnifolium Marsh.),°> 2n = * Britton (N. Am, Trees 853. 1908) sarge! was the oe author to select a lectotype species for Viburnum. His ag e of V. Tinus L., also adopted by Abrams & Ferris (Illus. Fl. ae States 4: 46. 1960), cannot be a seal for V. Tinus is a constituent of the segregate genus yeac Miller (Gard. “Dict. Abr. ed. 4. 1754). Rehder (Bibliogr. Cult Trees Shrubs 602. 1949) also gives V. Lantana L. as the lectotype species. * Viburnum alnifolium Marshall (Arbustrum Americanum 162. 1785) must be typified by the description, for there is no known Marshall specimen. This description Mackenzie, Torreya 27: 81-83. 1927). The Miller specimen in the British Museum (Natural History) labeled V. a americanum is Hydrangea arborescens L, it early fruit- ing condition. Miller i ovatis acum rr petiolis lo laevibus. Wayfaring tree w rt- shaped, oval, acute-pointed, sawed leaves, growi: nm very oth foot-stalks] can a tir to Hydrangea arborescens, but further informa can 1 account only be interpreted as a mixture of V. lantanoides Michx. and Hydrangea arborescens. In the latter Miller refers to the leaves as having long, slender petioles and “deeply sawed” edges, characters of Hydrangea arborescens, not V. lantanoides, while he 1966 | FERGUSON, GENERA OF CAPRIFOLIACEAE 43 18, is widespread in cool forests of the northeastern United States and adjacent Canada and extends southward in the higher mountains to North Carolina, Georgia, and Tennessee Section Lentaco DC. (‘‘Prunifolia” Small, “Obovata” Small), repre- sented by some six species in North America, is characterized by winter buds with one pair of scales and leaves with anastomosing veins. Viburnum cassinoides L., 2n = 18, is widespread from Newfoundland to Ontario but southward is confined mainly to the mountains in the Carolinas, Tennessee, northern Alabama, and Georgia. Characteristically with obscurely dentate to denticulate leaves, grooved stones, and shortly exserted anthers, it is usually distinguishable from the closely related and more polymorphic V. nudum L., 2n = 18, which has entire or only crenulate leaves, flat stones, and long sot Vilrwinn nudum occurs at lower altitudes throughout much of the southeastern United States, north to Kentucky and southern Connecticut, and west to Texas. The interrelationships of these species may merit further study. Viburnum rufidulum Raf. (V.rufotomentosum Small), 2n = 18, usually easily recognized by its rusty tomentum and serrate leaves and widespread throughout our area (although absent from peninsular Florida), ranges to hie Indiana, Illinois, Missouri, and Ohio. Viburnum prunifolium L., 18, which lacks the rufescent tomentum, occurs from the Carolinas, re "Tennessee, and Arkansas, north to New York and Connecticut, and west to Kansas and Texas. Some authors have regarded the popula- tions from Arkansas and Missouri as a separate species (V. Bushii Ashe) or variety (var. Bushii (Ashe) Palmer & Steyerm.), but Steyermark has shown that the leaf characters involved are so variable that these popula- tions should not be treated as a distinct taxon. Some specimens seen from g L., 2n = 18, has been reported from North Carolina and upland Georgia, but no Stet specimens have been found. The remaining species of the section is V. obovatum Walt. (“Obovata” Small), 2n = 18, which has included stamens and usually small leaves. It occurs throughout most of Florida (apparently absent from the Keys), north to southern Georgia describes the plant as eight or ten feet tall and the fruits as oval red berries. The other details of the description can apply equally well to wean ne If the Viburnum element in Miller’s description is use typify V. americanum, this unfamiliar and long unused name must be taken sy in preference to either This procedure hep that proposed by Mackenzie (Joc. cit.), Suringar (Mitt. Deutsch. Dendrol. Ges. 43: 216, 217. 1931), and Rehder (Jour. Arnold Arb. 28: 446. 1947) = rian another change in the name of this very characteristic species of Viburn Iam aes to Dr. N. K. B. Robson, of = British Museum (Natural History) for examining and commenting upon Miller’s specimen of Viburnum americanum, 44 JOURNAL OF THE ARNOLD ARBORETUM [VvoL. 47 and southwestern South pena Viburnum Nashii Small appears to be only a robust form of this specie Section ODONTOTINUS ela: ¢Dentata” Small), the species of which have coarsely dentate or lobed leaves and winter buds with two pairs of outer scales, has about six North American representatives, all of which occur in our area. A careful biosystematic study of this taxonomically difficult and confused group is needed. There is no general agreement on the delimitation of the species; the treatment followed here is that pro- posed by Fernald. Viburnum Rafinesquianum Schultes var. Rafinesquia- num (V. affine var. hypomalacum Blake), 2n = 36, distinguished by small, short-petioled or almost sessile stipulate leaves, is widely distributed from Quebec to Manitoba and south to upland Georgia and North Carolina. The glabrous var. affine (Bush ex Schneid.) House (V. affine Bush ex Schneid.) reaches our area in Arkansas. Closely related, V. molle Michx., 2n = 18 and 36, with its long-petioled, deeply cordate leaves and flat stones, ranges from Indiana to Missouri, south to northern Arkan- sas. Viburnum dentatum L., 2n = 36, 54, and 72, including V. scabrel- lum Chapm. (V. semitomentosum (Michx.) Rehd.) and V. Ashkei Bush, a highly polymorphic complex characterized by rounded leaf bases, coarsely dentate leaf margins, long, usually exstipulate petioles, and hairy styles, is widespread throughout our area (except peninsular Florida) and beyond, reaching Texas and Massachusetts. Viburnum bracteatum Rehd., dis- tinguished from V. dentatum by its long bracts and bracteoles and larger corolla, is known only from northwestern Georgia. Viburnum recognitum Fern., 2n = 36, very closely related to V. dentatum, is considered by some authors as var. lucidum Ait. It is separated by its glabrous young shoots and leaves, and ranges from New Brunswick to Ontario south to the Carolinas, Tennessee, northern Alabama, and Arkansas in our area. In contrast to this difficult group, V. acerifolium L., 2n = 18, throughout its wide range from Minnesota to Quebec, south to the Carolinas, northern Georgia, Tennessee, Alabama, northern Florida, and Louisiana, is im- mediately recognized by its palmately lobed leaves. Viburnum densiflorum Chapm., described from northern Florida and reported to occur in southern Georgia, Alabama, and Louisiana, appears at best to be only a variety (var. seo ome (Chapm. ) McAtee) of this species. Viburnum is a well-defined genus distinguished by its single-stoned eae sie regular flowers, nearly sessile stigma, and simple leaves. It shares with Sambucus regular flowers, uniovulate locules, and the occur- rence of stipules in some species. Although various authors have as- sociated the two, others have placed Viburnum and Triosteum together. A very close relationship with either seems doubtful, however, and the genus should probably by regarded as the only member of a separate tribe. Viburnum also shows affinities with Abelia and Symphoricar pos. Seed germination, anatomy, and the pharmaceutical properties of the wood and bark of many species have been extensively investigated, but there is a need for further ecological information. Volatile oils and as- 1966} FERGUSON, GENERA OF CAPRIFOLIACEAE 45 sociated unpleasant odors resembling those found in Valeriana are charac- teristic. Pollination is by insects, especially Hymenoptera and Diptera. Chromo- some numbers of 2” = 16, 18, 20, 22, 27, 32, 36, 40, 48, 54, and 72 have been reported. Many species are widely cultivated as ornamentals. An extract from the bark of Viburnum prunifolium has been used in medicine for its antispas- modic properties. REFERENCES: Under family references see DEmiI, FENG, Hara, — MolssL, PERsID- SKY (1937, 1939), STEYERMARK, and WILKINSON (1949). Barton, L. V. Germination and seedling production of species of Viburnum. Proc. Pl. Propag. Soc. 8: 126-136. 1958 [1959]. BiakeE, 5S. F. On the names of some species of Viburnum. Rhodora 20: 11-15. 1918. [V. affine and V. pubescens (Ait.) Pursh = V. dentatum L.] Boynton, K. R. Viburnum rufidulum. Addisonia 12: 61, 62. pl. 415, 1927. Britton, N. L. Viburnum molle Michx. Torreya 4: 69. 1904. BRUMBAUGH, J. H., & A. T. Guarp. A study of evidences of introgression among Viburn stdin Lentago, V. prunifolium and V. rufidulum based on leaf ¢ is eran (Abstr.) Proc. Indiana Acad. Sci. 66: 300. 1957. Busy, B. F. A new Viburnum. Am. Midl. Nat. 9: 192-194. 1924. [V. Ashei ] ———. Some species of Viburnum. Ibid. 10: 225-243. 1927. [An account of the genus in the central U. S.] Cross, G. L. The morphology of the bud and the development of the leaves of Viburnum rufidulum, Am. Jour. Bot. 24: 266-276. 1937 . A comparative eres study of the bud scales and foliage leaves of Viburnum Opulus. Ibid. 25: 246-258. 1938. DENNIsToN, R. H. Contributions from the pharmacognostical laboratory of the University of Wisconsin. IV. The comparative structure of the bark of certain sey viburnums. Pharm. Arch. 1: 137-148. pls. 7-13. 1898. Downs, R. J., & A. A. Prrtncer. Growth and flowering responses of five Viburnum species to various photoperiods. Proc. Am. Soc. Hort. Sci. 72: 511-513, 1958. Duncan, W. H. Synonymy in Viburnum obovatum and V. cassinoides. Rho- dora 52: 179-183. 1950. Eccer, K. Astragalin and paeonosid, the principal glycosides of the snowball ese Opulus L.). (In German.) Zeitschr. Naturf. 17b: 139-141. uae oe R. Cytological and interspecific hybridization studies in the genus iburnum. Diss. Abstr. 17: 22, 23. ; A cytological study of the genus Viburnum. Jour. Arnold Arb. 43: 132-172. 1962. [An extensive investigation, including a summary of known chromosome numbers. FERNALD, M. L. Viburnum edule and its nomenclature. Rhodora 43: 481-483. . Another century of additions to the flora of Virginia. /bid. 635-657. [Notes on nomenclature of V. recognitum, V. dentatum, V. nudum, 647— Deel 46 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 GIERSBACH, J. Germination and seedling i ae of species of Viburnum. Contr. Boyce Thompson Inst. 9: 79-90. HILLEBRAND, G. R. D. E. =e cting iesuidial correspondence among selected genera of the Cornales, Garryales, Rosales, Rubiales, and Umbellales as an indication of the taxonomic position of the genus Vibur- num, (Abstr.) Am. Jour. Bot. 52: 648. 1965. Hoim, T. Medicinal plants of North America. 69. Viburnum prunifolium L. tel Rep. 22: 35, 36. 1913.* [See also Bot. Jahresb. 41(2): 1075. te ie D. eae Rafinesquianum Schultes. Torreya 35: 126. 1935 JANAKI Amat, E. K. J. Chromosomes and the species problem in the genus Viburnum. Curr. Sci. Bangalore 22: 4-6. 1953.* Kern, J. H. The genus Viburnum (Caprifoliaceae) in Malaysia. Reinwardtia 1: 107-170. 1951. Kiri, E. P., & A. C. SmirH. The South American species of Viburnum. Bull. Torrey Bot. Club 57: 245-258. 1931. Know es, R. H., & S. Zak. Effects of temperature treatment and of a native inhibitor on seed dormancy and of cotyledon removal on epicotyl growth in Viburnum trilobum Marsh. Canad. Jour. Bot. 36: 561-566. 1958. Kren, W. Die Besiedlung des mittleren Neckarlandes durch den Wolligen Schneeball (Viburnum Lantana). Jahresh. Ver. Vaterl. Naturk. Wurttem- berg 107: 161-172. 1952. Lussock, J. On the form of the leaf of Viburnum Opulus and V. Lantana. Jour. Linn. Soc. Bot. 28: 244-247. 1891. McAtTEE, W. L. Notes on Viburnum and the assemblage Caprifoliaceae. Bull. Torrey Bot. Club 48: 149-154. 1921. . A review of the sgh Viburnum. iv + 125 pp. 9 pls. Published privately, Chapel Hill, N. C. 1956. Mackenzie, K. K. Viburnum cassinoides. Addisonia 7: 17. pl. 233. 1922. A botanical riddle. Torreya 27: 81-83. 1927. [V. almifolium vs. V. J Maximowicz, C. J. Diagnoses plantarum novarum asiaticarum. II. Bull. Acad. Sci. St.-Pétersb. III. 26: 420-542. 1880. [Synopsis of Viburnum, Miine-Repueap, E. Viburnum alnifolium. Bot. Mag. 157: pl. 9373. 1934. Morton, C. V. The Mexican and Central American species of Viburnum. Contr. U. S. Natl. Herb. 26: 339-366. 1933. Orstep, A. S. Til belysning af slaegten Viburnum. Vid. Medd. Naturh. For. Kj@benhavn 13: 267-305. pls. 6,7. 1861. [ Monograph. | Poucgues, M. L. pe. Etude ea gt oa de quelques Viburnum. Compt. Rend. Soc. Biol. Paris 141: 183-185. 1947. Rewper, A. Viburnum i een Rehd. nov. sp. Jn: C. S. SARGENT, Trees and shrubs 1: 135. pl. 68. 1903. Notes on some cultivated trees and shrubs. VI. Jour. Arnold Arb. 28: 445, 446. 1947. [V. alnifolium vs. V. lantanoides. } Russy, H. H. Viburnum prunifolium. Addisonia 3: 59. pl. 110. 1918. SARGENT, C.S. Viburnum lantanoides. Garden Forest 2: 531, 532. fig. 141. 1889. [See also V. soa eg! ibid. 3: 124. fig. 26. 1890; V. cassinoides, ibid. 9: 304. fig. 41. 1896.] ——. Viburnum. Silva N. Am. 5: 93-100. pls. 223-225. 1893; 14: 23, 24. pl. 710. 1902 1966 | FERGUSON, GENERA OF CAPRIFOLIACEAE 47 SUNESON, S. Zur Embryologie der Gattung Viburnum. Bot. Not. 1933: 181-194. 1933 SURINGAR, J. V. Nomenklaturalia. Mitt. porn Dendrol. Ges. 43: 199-217. 1931. [V. alnifolium vs. V.lantanoides, 216, 217.] SveNSON, H. K. Plants of agerv United States. I. Viburnum dentatum L. Rhodora 42: 1-6. pl. 585. Tuomas, J. L. The cytology - some cultivated species of Viburnum. Jour. Amold Arb. 42: 157-164. 1961. Wicut, R. On the structure of the ovarium and the generic character of Vibur- num, Calcutta Jour. Nat. Hist. 6: 357-359. 1846 WILKINSON, A. M. Floral anatomy and morphology of some species of the genus Viburnum of the Caprifoliaceae. Am. Jour. Bot. 35: 455-465. 1948 Younc, R. A. Key to the Ohio viburnums in the winter condition. Ohio Nat. 6: — H. W. The pharmacognosy, chemistry and pharmacology of Vibur- m. III. History, botany and pharmacognosy of Viburnum Opulus L. var. americana (Miller) Ait. Jour. Am. Pharm. Assoc. 21: 444-462. 1932. [See also V. cassinoides, ibid. 18: 1225-1228. 1929 *; anatomy of the bark, bid. 19: 680-704. 1930*; V. Lentago, ibid. 20: 315-328. 1931.] . Observations on the leaves of several species of Viburnum. Jour. Am. Phann. Assoc. Sci. Ed. 37: 349-353. 1948.* [See also ibid. 38: 549-553. 1949, & “A C. Muncu. Studies on Viburnum. The pharmacognosy and pharmacology of Viburnum alnifolium. Jour. Am. Pharm. Assoc. Sci. Ed. 29: 439-447. 1940.* [See also Quart. Jour. Pharm. 13: 358. 1940.*] ZABOROVSKY, E. P., & N. N. Varasova. Germination of seeds of the water 4 (Viburnum Opulus L.) and the wayfaring tree (V. Lantana L.). Russian.) Bot. Zhur. 46: 1169-1171. 1961. 3. Triosteum Linnaeus, Sp. Pl. 1: 176. 1753; Gen. Pl. ed. 5. 81. 1754. Perennial herbs with coarse, erect, glabrous or hairy stems. Leaves opposite, entire [divided], obovate to oblanceolate to panduriform, con- nate or united by a ridge around the stem, sessile, exstipulate, usually hairy. Flowers bisexual, irregular, sessile, solitary or in clusters of 2-4 in the axils of leaves [or in short, terminal, whorled spikes], bracteate and sometimes bracteolate. Calyx lobes 5, linear-lanceolate, foliaceous, per- sistent. Corolla 5-lobed, imbricate, a little longer than the calyx, greenish yellow to dull red, tubular, usually hairy within, slightly gibbous at the base, with a nectary of sessile or very short-stalked glands. Stamens 5, inserted about halfway up the corolla tube, usually included; filaments short, hairy; anthers oblong or linear, introrse, united for most of their length: pollen medium to large, echinate. Stigma capitate, 3—5-lobed; style usually hairy below, included or shortly exserted; ovary 4-locular, 3 locules with solitary ovules, 1 empty. F ruit a dry penne containing 3 ribbed oe with a thick, bony endocarp. Seeds with well-developed mbryo minute, situated near the micropyle. Embryo sac development ka (Polygonum type). LECTOTYPE sPEcIEs: T. per- 48 JOURNAL OF THE ARNOLD ARBORETUM [ VoL. 47 Greek, treis, three, osteon, bone, and sperma, seed, in allusion to the three stones; see Linnaeus’s comments about sesquipedalian generic names in Crit. Bot. 133-135. 1737 [pp. 104-107 in English ed., A. Hort., transl., 1938].) —- FEVERWORT, HORSE-GENTIAN. A genus of about eleven species of the Himalaya, eastern Asia, and eastern North America; three species in North America, all occurring in our area. Triosteum angustifolium L., recognized by its hispid-setose in- dumentum, solitary flowers, and paired bracts, ranges from Connecticut to Illinois and south to the Carolinas, Tennessee, Georgia, and Alabama, and is reported to occur in Louisiana. Triosteum perfoliatum L. extends from Massachusetts to Wisconsin and Minnesota south to the Carolinas, Ten- nessee, Georgia, and Alabama. It is characterized by the lower leaves with broad connate leaf bases, soft indumentum, and 1—4-flowered in- florescences, with a single bract and paired bracteoles. Both this and the preceding seem to be associated with basic soils. Triosteum aurantiacum Bickn. is a polymorphic species considered to be doubtfully distinct from T. perfoliatum by some authors and treated as var. aurantiacum (Bickn.) Wiegand. Separated from the other species by the lower leaves tapering to the base, soft indumentum, shorter style, and earlier flowering time, it occurs throughout the northeastern United isan southward to the mountains of North Carolina, Tennessee, and Georgi Triosteum is a well- defined genus readily distinguished by its tubular corolla, herbaceous habit, and dry drupaceous fruit. Its uncertain affinities with the Caprifolieae than with Viburnum. Fritsch thought that it prob- ably lies between the tribes Viburneae and Linnaeeae. The mode of pol- lination is unknown, the cytology of the genus has not been investigated, and ecological information is needed. A study of dormancy, weathering of the bony endocarp, and germination might yield interesting data. Triosteum is of no economic importance. The dried and roasted fruits have been used as coffee. The roots of T. perfoliatum are a mild cathartic. REFERENCES: Under family references see WILKINSON (1949). BICKNELL, E. P. A new horse gentian (Triosteum) common in the Eastern States. Torreya 1: 25-28. 1901. [T. aurantiacum. Crark, A. G. Triosteum perfoliatum in Massachusetts. Rhodora 6: 179, 180. 1904. Dempsey, E. The biological anatomy of Triosteum perfoliatum. Trans. Kan. Acad. Sci. 34: 217-219. 1931. Lang, F. i A new color form of Triosteum angustifolium. Rhodora 55: 346, 347. 1953. The genus Triosteum yet mies seine Diss. Abstr. 15: 17. 1955. [Abstract of a es saeiehace ab Maximowicz, C. J. Diagnoses Haptic novarum asiaticarum. IV. Bull. Acad. Sci. St.-Pétersb. III. 27: 426-560. pl. 1. 1881. [Synopsis of Triosteum, 476, 477.] 1966 | FERGUSON, GENERA OF CAPRIFOLIACEAE 49 WIEGAND, K. M. A new variety of Triosteum aurantiacum. Rhodora 20: 116. 1918 . Notes on Triosteum perfoliatum and related species. Jbid. 25: 199-203. 923. Tribe LINNAEEAE Fritsch 4. Symphoricarpos Duhamel, Traité Arbres Arbustes 2: 295. pl. 82. Deciduous, erect [or decumbent] shrubs. Winter buds with 2 pairs of outer scales. Leaves opposite, usually entire but sometimes toothed or lobed, glabrous or hairy, exstipulate, short petiolate. Inflorescences ter- minal or axillary racemes or spikes of paired, opposite, sessile or short- pediceled flowers [or flowers solitary]. Flowers bisexual, slightly irregular, subtended by bracts and paired bracteoles. Calyx lobes 5 [4], persistent. Corolla 5{4]-lobed, campanulate [salverform or tubular-funnelform], sometimes slightly gibbous at the base, with a nectary of sessile glands [or with 5 distinct nectaries]. Stamens 5 [4], inserted near top of corolla tube, included or exserted; filaments short, sometimes villous; anthers oblong or linear, introrse, united for most of their length; pollen echinate. Style glabrous or hairy, shorter than the corolla; stigma capitate or slightly 2-lobed; ovary 4-locular, usually with 2 fertile 1-ovulate locules and 2 sterile several-ovulate locules, the placentation axile. Fruit a globose, ovoid or ellipsoid drupe with 2 more or less compressed stones with a bony endocarp. Seeds oblong; endosperm copious, fleshy; embryo minute. Embryo sac development normal (Polygonum type). TYPE sPECIEs: Lonicera Symphoricarpos L. = 8S. orbiculatus Moench. (Name from Greek, symphorein, to bear together, and karpos, fruit, in allusion to the clustered fruit.) — SNOWBERRY. A genus of about 15 species of North America, two occurring in our area, and Symphoricarpos sinensis Rehd. in central China. The species fall into two groups treated either as sections or subgenera: sect. Sym- PHORICARPOS (subg. Eusymphoricarpos G. N. Jones), in which the lobes of the shortly campanulate corolla are as long as or longer than the tube, and sect. MERIDIONALES Gray ex Schneid. (subg. Anisanthus (Willd. ex Roem. & Schult.) G. N. Jones), in which the lobes of the tubular or salverform corolla are shorter than the tube. These differences hardly seem to be of subgeneric value. Both species in our area belong to sect. SYMPHORICARPOS. Symphoricarpos orbiculatus Moench (S. Symphoricarpos (L.) Mac- Millan), 2n = 18, ranges from New York south to Florida, west to Texas, Colorado, and northern Mexico. The species seems to be a calciphile. The short, hairy styles and abundant small, purplish-red fruit contrast with the longer, glabrous styles and white fruit of S. albus (L.) Blake (S. racemosus Michx.). The var. albus is of wide range in the northern United States; the western American var. /aevigatus (Fern.) Blake (S. rivudaris 50 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 Suksdorf), widely cultivated as an ornamental shrub, is sporadically naturalized in our area. The characters of glabrous leaves and young shoots, large berries, and erect habit used to separate var. /aevigatus from var. albus do not seem to be constant. Experimental studies are desirable. Pollination appears to be chiefly by Hymenoptera. There seems to be no mechanism to limit self-pollination. Cleistogamous flowers have been reported in cultivated specimens of Symphoricarpos orbiculatus. The structure of the pericarp and dormancy and after-ripening of the seeds have been extensively investigated. It is reported that the root develop- ment of some species suggests a mycorrhizal association, The cytology of the genus is not well known; chromosome numbers have been reported only for S. albus, 2n = ca. 54, and S. orbiculatus. Symphoricarpos is usually placed in the tribe Linnaeeae, which is dis- tinguished by its four or three carpels with one-seeded fertile locules and two sterile locules. It is probably closely related to Abelia, in which the ovary is apne gah the fruit an achene. It also shows affinities with Kolkwitzia and Linn A number of oe Gactitae both of ours, are cultivated as orna- mentals. REFERENCES: Under family references see ARTIUSHENKO, MorssL, Persipsky (1937, 1939), WILKINSON (1948, pp. 365-371; 1949). AnprE, H. Neue Untersuchungen iiber Bliitenverlaubungen von Symphoricar- pus racemosus Michx. Ber. Deutsch. Bot. Ges. 57: 285-298. 1939. avous. BLAKE, S. F. The earliest name of the snowberry. Rhodora 16: 117-119. 1914. [S. albus (L.) Blake, comb. nov. ] Burtt Davy, J. The origin of Symphoricarpus. Science II. 72: 220. 1930. Cocker, W., & A. L. Canton. Extractions from woods. IV. Extractions from os Ft a rivularis. Perfum. Essent. Oil Rec. 54: 171-174. 1963.* FERNALD, M. L. Symphoricarpos racemosus and its varieties in eastern America. Rhodora 7: 164-167. 1905. FLemion, F. Physiological and chemical changes preceding and during the after-ripening of Symphoricarpos racemosus seeds. Contr. Boyce Thomp- son Inst. 6: 91-102. 1934. [S. albus.] . Effect of the addition of nitrogen upon erent - seeds of Sym- phoricarpos racemosus. Ibid. 12: 485-489. 1942. [S. a — &E. Parker. Germination studies of seeds of Symploiaro orbicula- tus. Contr. Boyce Thompson Inst. 12: 301-307. GANZINGER, K. oe auf laces hei Protoplasma 30: 488-508. 1938. [S. albus. GRAENICHER, S. The fertilization of oo and Lonicera. Bull. Wis. Nat. Hist. Soc. II. 1: 141-156. front. 1900. Gray, A. Revision of the genus Symphoricarpus. Jour. Linn. Soc. Bot. 14: L875: 9-12. Hutt, E. D. Notes on the coral-berry (Symphoricarpos orbiculatus). Rhodora 49: 117, 118. 1947 Jones, G. N. A monograph of the genus Symphoricarpos. Jour. Arnold Arb. 21: 201-252. 1940. 1966 | FERGUSON, GENERA OF CAPRIFOLIACEAE 51 Keck, D. D. Lonicera and Symphoricarpos in southern California. Bull. So. Calif. Acad. Sci. 25: 66-73. 1926. McMuran, C., & B. F. Pacer. Phenological variation within a population of Symphoricarpos occidentalis. Ecology 39: 766-770. 1958. Nasu, G. V. Symphoricarpos albus laevigatus. Addisonia 3: 27. pl. 94. 1918. PeLton, J. Studies on the life-history of Symphoricarpos occidentalis Hook. in ere g Ecol. Monogr. 23: 17-39. 1953. [Extensive biological treat- ent. J Pes, F. W. Symphoricarpos Symphoricarpos. Addisonia 3: 61, 62. pl. 111. 918. [S. orbiculatus.] a. D. Che mention cells in petals and sepals of Symphoricarpus racemo- Michx. (In Ukrainian; English summary.) Bot. Zhur. Kiev 1: 139- 142. 1940. [.S. albus. ]} Preirrer, N. E. Morphology of the seed of Symphoricarpos racemosus and the relation of fungal invasion of the coat to souls capacity. Contr. Boyce Thompson Inst. 6: 103-122. 1934. [S. albus.] Sims, J. Symphoria racemosa. Bot. Mag. 48: pl. po 1821. [S. albus var. landipatus 4 SMITH, F. D. Identification of Symphoricarpus occidentalis. Proc. Iowa Acad. Sci. 37: 127-130. SovuEcgs, R. Embryogénie des Caprifoliacées. Développement de l’embryon chez le Symphoricarpos occidentalis Hook. Compt. Rend. Acad. Sci. Paris 277: 1066-1068. 48. Svenson, H. K. Did Symphoricarpos albus come originally from Canada? Rhodora 39: 461, 462. 1937. URL, W. Phasenoptische Untersuchungen an Fruchtfleischzellen von Symphori- carpus racemosus Hooker. Protoplasma 51: 399-414. 1959. [S. albus var. laevigatus. | Tribe DrerviLLEAE C. A. Meyer 5. Diervilla Miller, Gard. Dict. Abr. ed. 4. 1754. Deciduous, erect, often stoloniferous shrubs. Winter buds with several pairs of pointed scales. Leaves opposite, entire or serrate, glabrous or ew sessile or short petiolate, exstipulate. Inflorescence of terminal or xillary 3- to several-flowered, short-pedunculate, bracteate cymes some- pos forming a terminal thyrse. Flowers bisexual, slightly irregular, short pediceled, bracteolate. Floral tube slender; calyx lobes 5, linear- lanceolate, persistent or deciduous late in fruit. Corolla 5-lobed, bilabiate, with a 4-lobed upper lip, the lower lip and inner surface of the funnelform tube densely hairy; tube slightly gibbous at the base, with a large sub- globular nectary. Stamens 5, inserted near the top of the corolla tube; filaments pubescent; anthers linear, usually exserted, introrse, dorsifixed, anther-halves free at the base, united above; pollen medium to large, echi- nate. Stigma capitate; style long, slender, densely pubescent below, equal in length or a little longer than the tube; ovary elongate, 2-locular, with numerous ovules, placentation parietal. Fruit a thin-walled septicidal capsule very weakly or tardily dehiscent, bivalvate, crowned with the per- sistent calyx. Seeds small, ovoid, with a reticulate testa; endosperm oe JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 calyx, ig 4; f, see essi hd. rivularis: dehisced pol to show two valves and persistent ieee x 4; fleshy; cotyledons large, superior. Embryo sac development normal Polygonum type). Type species: Lonicera Diervilla L. = D. Lonicera Mill. (Named after Diéreville or Dierville, a French surgeon who traveled in eastern Canada (Acadia), 1699-1 700, and introduced the plant into France.) — BUSH-HONEYSUCKLE. A genus including either three species of eastern North America or about 1966 | FERGUSON, GENERA OF CAPRIFOLIACEAE 53 12 more of eastern Asia, depending upon the generic concept. Diervilla Lonicera, 2n = 18, characterized by its petiolate rather than sessile leaves, occurs from Newfoundland to Saskatchewan, southward in the mountains to North Carolina and Tennessee. Diervilla sessilifolia Buckl., 2n = 36, of mountain woods in North Carolina, Tennessee, Georgia, and Alabama, is distinguished from the closely related D. rivularis Gattinger (D. sessili- folia var. rivularis (Gattinger) Ahles), 2m = 36, which occurs over about the same range, by its longer capsules and glabrous leaves. The flowers are reported to be proterogynous and are adapted for pol- lination by Hymenoptera. The color of the corolla changes in the Nor American species from yellow to orange or red after anthesis. The generic concept of Diervilla has been varied. A distinctive group of Asiatic species has been treated as a separate genus, Weigela Thunb., or as either a section (Weigela (Thunb.) Koehne) or subgenus (Weigela (Thunb.) K. Koch) of Diervilla. The Asiatic species are characterized by large, regular, pink to red corollas, woody capsules with a well-developed sistent columella, irregular bilabiate calyx, pilose anthers connected by hairs in a ring about the style, and long-winged seeds. This last species has been variously treated as a separate genus, Macrodiervilla Nakai, or as sect. Calyptrostigma Koehne of Diervilla. Another Japanese species, W. Maximowiczii (S. Moore) Rehd., intermediate in its yellow flowers, treated as a separate genus (Weigelastrum (Nakai) Nakai) or as a sec- tion (Weigelastrum Nakai) of Diervilla. Further investigation is needed to determine the status of these groups. Both a and Weigela have affinities with Lonicera, Leycesteria, and Abel Seni: en Weigela are important as ornamental shrubs; the small- flowered North American species of Diervilla are rarely cultivated. REFERENCES: Under family references see WILKINSON (1948, pp. 261-271; 1949). Battey, L. H. The case of Diervilla and Weigela. A discussion of bush honey- suckles and weigelas in the interest of the cultivator. Gent. Herb. 2: 39- 2 54. 19 BaILton, H. Sur les graines des Diervilla. Bull. Soc. Linn. Paris 1: 202, 203. [Seed morphology of Diervilla, Weigela, Calyptrostigma. | Bartiett, H. H. Flower color of the American diervillas. Rhodora 9: 147, 148. 1907 Fassett, N. C. Mass collections: Diervilla Lonicera, Bull. Torrey Bot. Club 69: 317-322. 1942. Maximowicz, C. J. Diagnoses plantarum novarum asiaticarum. VI. Insunt stirpes quaedam nuper in Japonia detectae. Bull. Acad. Sci. St.-Pétersb. III. 31: 12-121. 1887. [Diervilla = Weigela, 59-61.] 54 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 Meyer, C. A. Einige Bemerkungen iiber Diervilla, Weigela, Calysphyrum und eine neue, mit ihnen verwandte Gattung. Bull. Phys.-Math. Acad. Sci. St.- Pétersb. II. 13: 216-220. Naxkar, T. Weigela and its akins j in Japan proper and ante Jour. Jap. Bot. . 1936. [Weigela, Weigelastrum, Macrodier eines, C. Die Gattungen Diervilla und Weigela. ss pued. Dendrol. Ges. 42: 13-23. 1930. Tribe CAPRIFOLIEAE [Lonicereae R. Br. ex DC.| 6. Lonicera Linnaeus, Sp. Pl. 1: 173. 1753; Gen. Pl. ed. 5. 80. 1754. Deciduous or, more rarely, evergreen shrubs or woody climbers. Winter buds with several or 2 outer scales. Leaves usually entire, sometimes lobed persistent [or deciduous]. Corolla 5-lobed, bilabiate, with a 4-lobed upper lip, or regular, imbricate, tubular or funnelform with a long or short tube, a ments long, slender; anthers oblong or linear, usually exserted, introrse; pollen medium to large, echinate. Stigma capitate; style slender, equaling or a little longer than the corolla tube; ovary 2—3[rarely to 5|-locular with 3-8 ovules [the walls of the ovaries of 2-flowered cymes sometimes united], placentation axile, the septa becoming free toward the top of the ovary. Fruit a few-seeded fleshy berry. Seeds ovoid, with a smooth or reticulate [or granular] testa; endosperm fleshy. Embryo sac development normal (Polygonum type). Lectotype species: L. Xylosteum L.; see Rehder, Bibliogr. Cult. Trees Shrubs 614. 1949.4 (Named in honor of Johann *The lect ~— species of Lonicera has been incorrectly cited as L. Caprifolium L. by Britton & Brown (Illus. Fl. No. U. S. ed. 2. 3: 277. 1913), Britton (FI. Ber- muda 372. 1918), and Hitchcock & Green (Int. Bot. Congr. Cambridge 1930. ica Propos. Brit. Bot. 131. 1929). Rehder’s choice of L. Xylosteum appears to be corre Early authors treated Lonicera in various ways, recognizing one or more genera. Linnaeus (Sp. ii ae 175. 1753) placed all the species known to him in one genus, Lonicera, with two informal groups, “Periclymenum” and “Chamaecerasus.” In con- trast, Miller (Gard. Dict. Abr. ed. 4. 1754) distinguished the genera Caprifolium, Lonicera, Periclymenum, and Xylosteon. The species of Miller’s Caprifolium cor- respond closely with those of Linnaeus’s informal group “Periclymenum,” and Miller’s species 7, C. Italicum Dod., the Italian honeysuckle, is cited by Linnaeus in the Periclymenum italicum, as it is ed, corresponds exactly with the description of L. Caprifolium in Species Plantarum. Thus, w hile maintaining Lonicera as a genus, Miller effectively segregated L. Caprifolium from Lonicera by placing it in his genus Caprifolium. This species alee cannot be used to typify Lonicera. In L. japonica: i, flowering branchlet, show- — ~~ e ing flowers in pairs, x 1/2: . flower, x 2: contrast, L. Xylosteum was retained by Miller in Lonicera throughout. His species 3, “Louicere pedunculus bifloris ovatis acutis integris Lin. Hort. Cliff. The Fly Honey- ckle, vulgo,” (Gard. Dict. Abr. ed. 4. 1754) fits Linnaeus’s description of i ther clarified in 1768 - Gard. Dict. ed. su i as fur 8), when Miller used the binomial Lonicera aX ylosteum for this species 56 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 Lonitzer, latinized Lonicerus, 1499-1569, a German herbalist and phy- sician, not Adam Lonitzer, 1528-1586; see Linnaeus, Crit. Bot. 93. 1737.) — HONEYSUCKLE. A genus of about 180 species of temperate and tropical regions of the Northern Hemisphere. About 25 species occur in North America; about seven are native in our area and five are introduced. Although Small recognized Phenianthus Raf., Xylosteon Mill. (X ylosteon Juss.), and Nintooa Sweet as genera separate from Lonicera, there do not seem to be sufficient grounds for adopting this treatment. The species fall into two well-marked subgenera: Lonicera (subg. Chamaecerasus Rehd.*), distinguished by axillary inflorescences with the flowers in pairs and the leaves free, and CapriroLium (Mill.) Pers. (subg. a p genera may be further divided into sections by the veuilat | or irregular corolla, the length of the corolla tube, the climbing or erect habit, and the free or connate bracteoles.® One indigenous and about four introduced taxa of subg. LONICERA occur in the southeastern United States. Lonicera canadensis Marsh. (Xylosteon ciliatum (Muhl.) Pursh), 2n = 18, a northeastern species extending southward in the mountains to North Carolina, Tennessee, and eorgia, is distinguished by a short, nearly regular earulla aria gib- bous at the base and by red fruit. Lonicera japonica Thunb. (Nintooa japonica (Thunb.) Sweet), 2x = 18, introduced from Asia now wide- spread in eastern North America, has become a pernicious and dangerous pest in woodlands, on roadbanks, and in hedgerows, overwhelming and strangling the native flora. An exuberant woody vine, it is readily dis- tinguished by its bilabiate corolla, long-toothed calyx, and black fruits. Lonicera fragrantissima Lindl. & Paxton (X ylosteon fragrantissima (Lindl. & Paxton) Small), 2” = 18, a Chinese shrub widely cultivated in the south- *Rehder (Missouri Bot. Gard. Rep. 14: 27-232. 1903) recognized two sub- genera: Chamaecerasus L. and Periclymenum L. As shown in the preceding footnote, os. The species onion in subg. x ylosteum were very e subg. Lonice e speci Persoon’s subg. Caprifolium correspond closely with those in Rehder’s Periclymenum and include L. Caprifolium. Caprifolium is the earliest name used in the correct sense at the rank of subgenus and should be adopted. Rehder recognized five sections, four within subg. Lonicera, bigs some author. r both subgenera and sections. The sections then usually recognized are Lonicera, ” Caprifoli ium ons ee Isles, ed. 2. 790, 791. 1962) mistakenly regarding L. ‘Ca pecies of Lonicera, gives the sections as Xylosteon (Mill.) DC Maxim. .. Lonicera, and Nintoos (Spach) 1966] FERGUSON, GENERA OF CAPRIFOLIACEAE s7 eastern United States; L. Maackii (Rupr.) Maxim., 2” = 18, fro eastern Asia; and L. Y, bella Zabel (L. Morrowii Gey < L. tatarica L. ), 2n = 18, are reported to spread from cultivation in our area. The remaining species in the Southeast are members of subg. Capri- FOLIUM. Lonicera sempervirens L. (Phenianthus sempervirens (L.) Raf.), 2n = 36, of wide range throughout our area, extending north to Maine and west to Nebraska, Ohio, Iowa, and Texas, is distinctive in its slender trumpet-shaped, regular, bright-red corolla. Lonicera dioica L., a poly- morphic species occurring from Quebec and Maine to British Columbia, Kansas, and Missouri and extending south to North Carolina, Tennessee, and Georgia, is readily recognized by its bilabiate corolla gibbous at the base and uppermost connate leaves green above. The very doubtfully distinct L. glaucescens Rydb. is treated as a variety * of L. dioica by most authors. Lonicera flava Sims, a variable species occurring in the Carolinas, northern Georgia and Alabama, and Tennessee, Kentucky, Missouri, Arkansas, and Oklahoma, is distinguished by its bright orange-yellow bi- labiate corolla and elliptic or rhombic, connate uppermost leaves. The paler, somewhat smaller-flowered L. flavida Cockerell ex Rehd. is probably best treated as L. flava var. flavescens Gleason (cf. Steyermark). Lonicera prolifera (Kirchner) Rehd. (L. Sullivantii Gray), 2n = 18, closely re- lated to both ZL. flava and L. dioica, occurs from Ontario to Wisconsin, Kansas, and Missouri, south to Arkansas and Tennessee. Its nearly orbicular connate uppermost leaves with a glaucous dorsal surface, its extended spicate inflorescence, and its bilabiate pale-yellow corolla gib- bous at the base are characteristic. Lonicera albiflora Torr. & Gray was described from Fort Towson, Arkansas. Small (in both editions of Flora of the Southeastern United States) and Rehder recorded this species from Arkansas, but Fort Towson is today in southeastern Oklahoma, and there appears to be no other record of this species in Arkansas. The European L. Caprifolium, 2n = 18, is reported as an escape from cultivation. Many of the species are proterandrous. Species with long corolla tubes are pollinated by long-tongued insects or by hummingbirds, while those with short tubes are visited by various Diptera and Hymenoptera. Chromo- some numbers of 2” = 18, 36, and 54 have been reported; most species appear to be diploids (2n = 18). A large number of species are grown as ornamentals, and many hybrids (some of them complex) have arisen spontaneously in cultivation. Lonicera is closely related to Diervilla and Leycesteria, the three genera showing their affinities in their two- or three-flowered cymes and ovaries with several ovules. Lonicera is distinguished from Leycesteria by its two- or three-locular vs. five-locular ovary and from Diervilla by its fleshy fruit. 7 Gleason (New Britt. & Brown. 1952) recognized both var. glaucescens (Rydb.) Butters and var. dasygyna (Rehd.) Gleason. In the event that these varieties are combined (as in nage de Manual, ed. 8. 1950), the latter name is correct, for it is th older name in the of variety, having been described in 1903 under Lonicera glaucescens, which itself was not reduced to varietal status until 1913 58 JOURNAL OF THE ARNOLD ARBORETUM [voL, 47 REFERENCES: Under family references see Fenc, Hara, Morsst, Perstpsky (1937, 1939), STEYERMARK, and WILKINSON (1948, pp. 261-271; 1949); under Symphori- carpos see GRAENICHER and Kec ALEXANDER, E. J. Lonicera canadensis. Addisonia 21: 9. pl. 677. 1939. AnpbREWwS, E. F. The Japanese honeysuckle in the eastern United States. Torreya 19: 37-43. 1919. [L. japonica. ] Arser, E. A. N. On the synanthy in the genus Lonicera. Jour. Linn. Soc. Bot. 35: 463-474. 1903 ARTIUSHENKO, Z. T. On the nature of the giant cells in the ovary of Lonicera L. (In Russian.) Bot. Zhur. 36: 532-534. BEGuIN, C. Recherches biochimiques sur quelques espéces de chévrefeuille (Lonicera). Pharm. Acta Helvet. 9: 233-241. 1934.* [See also ibid. 10: 109-114. 1935 *; ibid. 11: 202-208, 361-367. 1936 *; and ibid. 13: 77-82. BRENDER, E. V. Progress report on control of posit datas and kudzu. So. Weed Conf. Proc. 13: 187-193. 1960.* [L. japo BroapHurst, J. Fasciation in the Japanese Sea ancoii Torreya 9: 208, 209. 1909. CL. japonica. | Cormack, R. G. H., & A. L. Gornam. Effects of exposure to direct sunlight upon the development of leaf structure of two deciduous shrub species. Canad. Jour. Bot. 31: 537-541. 1953. [Menziesia glabella, Lonicera glaucescens. | DancEaArD, P. A. Remarques sur la présence d’un appareil centrosomique chez les espéces du genre Lonicera. Compt. Rend. Acad. Sci. Paris 196: 1260, 1261. 1933. alpigena. | FERNALD, M. L. The American representatives of Lonicera caerulea. Rhodora 27: 1-11. 1925. [= L. villosa (Michx.) R. & S. Goopwin, T. W. The carotenoids of the berries of Lonicera japonica. Biochem. Jour. 51: 458-463. 1952. GREEN, P. S. Identification of the species and hybrids in the Lonicera tatarica complex. Jour. Arnold Arb. 47: 75-88. 1966. Hooker, he J. Lonicera chinensis. Bot. Mag. 61: pl. 3316. 1834. [L. japon Huremsson, .. Seek of the honeysuckle. Gard. Chron. III. 116: 24. (L. Periclyme ie ee. a re D. SAUNDERS. Chromosome numbers in species of Lonicera. Kew Bull. 7: 539-541. 1952. Jennincs, O. E. A new species of Lonicera from Pennsylvania. Ann. Carnegie Mus. 4: 73-77. pl. 20. 1906. [L. altissima, sp. nov.; key to spp. of ne. U5 Kemp, E. L. Some observations on Lonicera japonica Thunb. (Japanese honey- suckle). Torreya 24: 103, 104. 1924. LEATHERMAN, A. D. Ecological life-history of Lonicera japonica Thunb. Diss. Abstr. 15: 1987. 1955.* McVaucH, R. Suggested gg of Prunus serotina and other wide-rangin phylads in North America. Brittonia 7: 317-346. 1952. Lonicera subsect. Phenianthi, 343, 344.] Maxtimowicz, C. J. Diagnoses plantarum novarum asiaticarum. II. Bull. Acad. Sci. St.-Pétersb. III. 24: 26-89, 1878. [ Lonicera, 35-50. ] [Phylogeny of 1966 | FERGUSON, GENERA OF CAPRIFOLIACEAE 59 Mayserry, M. W. Comparative anatomical research within the genus Lonicera. Univ. Kan. Sci. Bull. 22: 27-53. pls. 5-9. 1935 Nakal, T. A new classification of the genus Lonicera in the Japanese Empire, together with the diagnoses of new species and new varieties. Jour. Jap. Bot. 14: 359-376. 1938. ae A. oat - the genus Lonicera. Missouri Bot. Gard. Rep. 14: 27- S. 1-20. 1 ; a on rs eaieiisiies of the fruit of Lonicera caerulea. Rhodora 11: 209-211. 1909. J ele prolifera and L. flavida. Ibid. 12: 166, 167. 1910. [Nomen- clatu SARGENT, °C. S. Two American honeysuckles. Garden Forest 3: 187. figs. 33, 34. 1890. [L. flava, L. Sullivantii (L. prolifera . An interesting rediscovery. Garden spleen 4: 253, 254. 1891. [L. flava in S. C. Sims, J. Lonicera sempervirens. Bot. Mag. 20: pl. 781. 1804. [See also L. flava. Ibid. 32: pl. 1318. 1810; L. sempervirens (B.) minor. Ibid. 41: pl. 1753, 1815.] Tyter, A. A. An examination of the pubescence of the styles and filaments of Lonicera hirsuta Eaton, L. Sullivantii Gray and L. glauca Hill. Bull. Torrey Bot. Club 21: 123- ret pls. 181, 182. 1894. [L. Sullivantii = L. prolifera; L. glauca = L. dioic ZAITSEV, G. N. The uae of seed characters for the taxonomy of the genus Lonicera L. (In Russian.) Bot. Zhur. 43: 836-840. pis. 1, 2. 1958. 60 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 THE GENERA OF STERCULIACEAE IN THE SOUTHEASTERN UNITED STATES * GEORGE K. BrIzICKY STERCULIACEAE Bartling, Ord. Nat. Pl. 255, 340. 1830, nom. cons. (STERCULIA FaMILy) Trees, shrubs, or herbs [rarely vines]; pubescence of stellate and/or simple hairs: inner bark usually fibrous; mucilage cells and/or cavities mole ducts in the axis sig leaves and een organs. Leaves sexual by abortion, regular, usually hypogynous, ia oi to sessile, usually bracteolate, sometimes with an involucel 3 or 4 distinct bract(let)s. Sepals usually 5, connate [rarely distinct} often nectarifer- trorse, usually dorsifixed, 2- or rarely 3-locular, the locules parallel or divergent [rarely separated, or exceptionally superposed], sometimes of unequal size, longitudinally dehiscent; pollen usually small to medium sized, oblate to prolate, often nearly spherical, 3—8-colp(oidor)ate, mostl reticulate. Gynoecium 5-carpellate [rarely up to 60-carpellate], syn- carpous, rarely apocarpous, or 1-carpellate, rudimentary in 4 flowers; *Prepared for a generic flora of the Seenengee United States, a joint project of the Arnold Arboretum and the Gray Herbari of Harvard University made possible through the support of George R. Cooley an the Bios os cience Foundation and under the direction of Carroll E. Wood, Jr., and Reed C. Rollins. This treatment follows the pattern established in the first paper in the series phen Arnold Arb. 39: 296-346. 1958) and continued through those in volumes 40-46 (1959-1965). The area covered includes North and South Carolina, Georgia, Florida, Tennessee, Ala- bama, Mississippi, Arkansas, and Louisiana. The descriptions are based primarily on the plants of this area, with any supplementary material in brackets. References which ne author has not seen are marked by an asterisk. author is indebted to Dr. Wood for his many editorial suggestions pur helpful ‘islam in the course of this work; to Dr. R. K. Godfrey for flowering material of Firmiana; and to Mrs. Gordon W. Dillon for her help in the preparation of on type- script. 1966 | BRIZICKY, GENERA OF STERCULIACEAE 61 stigmata 1—5; styles as many as carpels, distinct or variously connate; ovary superior, usually 5-locular with axile placentae, or rarely 1-locular with a submarginal placenta; ovules |1] 2 to many in a locule, ascendent to pendulous, hemianatropous to anatropous, epitropous to apotropous, 2-integumented, with a thick nucellus. Fruit various, often capsules, folli- cles, or schizocarps, with many- to 1-seeded locules. Seed [sometimes arillate or winged| with [or without] endosperm; embryo axial, straight or curved, the cotyledons plane or plicate and/or coiled around the radicle. Germination epigeous [or hypogeous]. (Including Byttneriaceae R. Br.). Type Genus: Sterculia L. An almost exclusively pantropical family of 65 or 70 genera (about 1000 species), with a few extensions into warm-temperate areas; six of 12 genera (one naturalized) in our area; three additional eng extending northward from Mexico to Texas or California and Arizon The family has usually been subdivided into eight to rl tribes. Evi- dence from wood and petiole anatomy favors the establishment of two subfamilies: Sterculioideae (including only Sterculieae) and Byttnerioi- deae (including the remaining tribes), with at least one tribe intermediate in petiolar anatomy. Floral anatomy emphasizes the great uniformity of the family in regard to the development and structure of the androecium. In contrast, ‘Pollen morphology is in favour of referring Erioleneae and Dombeyeae to Malvaceae, Fremontodendreae |Fremontieae] to Bom- bacaceae and parts of the rest of Sterculiaceae to Tiliaceae”’ (Erdtman, p. 421). Thus emended, the Sterculiaceae would be very homogeneous paly- nologically, but the limits between the families of the Malvalves would become even weaker and more uncertain than they are at present. It seems desirable, at least for the present, to treat the family within the generally accepted limits, with its subdivision into tribes Sterculiaceae are closely related to Tiliaceae, Malvaceae, and Bomba- caceae. Evidence from floral anatomy and embryology suggests that the family is the most primitive in the Malvales. On the same basis the Ster- culieae (with pentacyclic flowers, numerous stamens fertile in both series, apocarpous gynoecia, and many-ovulate locules) are regarded as the most primitive, and Hermannieae (with tetracyclic flowers, a reduced num- ber of stamens and ovules, and a tendency to perigyny) the most advanced tribe of Sterculiaceae. In respect to perianth structure the Helictereae DC. (with irregular almost two-lipped flowers) and Byttnerieae DC. (with hooded and terminally or + abaxially ‘“‘appendaged” petals of complex structure) appear to be the most highly specialized tribes of Sterculiaceae. Data on floral biology are few, but the family in general appears to be adapted to insect-pollination. Nectaries, when present, seem to be located * In addition, Guazuma ulmifolia Lam. (as G. tomentosa HBK.) has been recorded aks one locality in Dade County, Florida (Moldenke). This record is in need of verification and bmuens as a basis for the inclusion of Guazuma Adans. in our ra 62 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 at the base of the sepals (Brown) where they sometimes form an extrastam- inal disc lining the calyx base and the receptacle around the ovary or the androgynophore. Foetid, indol- or scatol-scented flowers, usually dark purple or chocolate in color, are apparently pollinated by flies. Some of these flowers are of the “lantern type,” with fly-traps for small flies. The nectariferous, irregular flowers of Helicteres Isora L. are supposedly pol- linated by large Apidae or by small honey birds. In Theobroma Cacao L. pollination by thrips, ants, midges, and aphids has been recorded. Pro- terandry and the relative positions of the extrorse anthers and stigmata, characteristic of the family, seem to secure cross-pollination. Heterostyly is known in Melochia and Waltheria (q. v.). Cleistogamous flowers occur in a few species of Ayenia (Cristobal). In Theobroma Cacao, individual trees are either self-incompatible, or cross-incompatible, or self-incom- patible but cross-compatible only with the pollen of self-compatible trees (cf. Cope in Cuatrecasas). The androecium in Sterculiaceae is basically obdiplostemonous. The traces for the antipetalous stamens, arising conjointly with those of the petals, divide first into three bundles each, a median and two laterals, which sometimes undergo a further secondary division (multiplication). Consequently, there are usually three (sometimes more or fewer) stamens opposite each petal (alternate with the sepals in apetalous flowers). The traces for the antisepalous stamens usually remain unbranched (single), sometimes becoming nonfunctional. Correspondingly, there characteristi- cally are only five stamens (usually sterile), one opposite each sepal. All members of the androecium usually are connate into a staminal tube of various lengths. Thus, the androecium in Sterculiaceae seems to have evolved along the lines of cohesion and reduction (Rao, 1952). The most significant embryological features of the family include the secretory type of anther tapetum; the multicellular archesporium of the ovule, with only one functional cell; Polygonum-type embryo sac; hooked synergids and usually ephemeral antipodals; nuclear endosperm becoming cellular at a late stage of seed development; and embryo development of the Urtica (exceptionally Polygonum) variation of the Asterad type. Polyembryony has been recorded in Cola nitida. Apotropous ovules occur at least in Melochia and Waltheria. Chromosome counts recorded for about 20 genera (40 species) are 2” = 14, 16, 20, 24, 32, 36, 38, 40, 46, ca. 50, and 120, suggesting that dis- ploidy, polyploidy, and aneuploidy have been of importance in the evolu- tion of Sterculiaceae. It is notable that chromosome numbers of 20 or 40 occur in 60 per cent of the genera investigated. The lowest number, 2” = 14, has been found in Melochia and Waltheria. Theobroma Cacao, native to tropical America and widely cultivated there and in the Tropics of the Old World, especially in West Africa, yields seeds which are the source of cocoa, chocolate, and cocoa butter. The seeds contain the alkaloid theobromine, as well as traces of caffeine. Species of the tropical African Cola Schott & Endl., mainly C. nitida (Vent.) Schott & Endl. and C. acuminata (Beauv.) Schott & Endl., furnish edible seeds, 1966 | BRIZICKY, GENERA OF STERCULIACEAE 63 kola nuts, which contain caffeine and a heart-stimulating glucoside, kolan- in, and which are used in medicine as a cerebral, cardiac, and nerve stimulant and in industry for beverage purposes. Some species of several genera yield bast fibers and/or lumber of local importance (e.g., Guazuma, Sterculia L., Cola). A few species of Dombeya Cav., Fremontia Torr. (Fremontodendron Cov.), Brachychiton Schott & Endl., Sterculia, and some other genera are grown as ornamentals in southern Florida. REFERENCES: ALEXANDER, E. J. Dombeya Cayeuxii of hybrid origin. Addisonia 23: 1, 2. pl. 737. 1954. ARENES, J. Sterculiacées. Jam: H. Humpert, Fl. Madagascar Comores. 131¢ Famille. 539 pp. Paris. 1959. BaILLon, H. Malvacées. Hist. Pl. 4: 57-160. 1873. [Sterculiaceae included in Malvaceae, 57-82, 103, 104, 121-137.] BENTHAM, G. Notes on Malvaceae and Sterculiaceae. Jour. Linn. Soc. Bot. 6: 97-123. 1862 (15 May). [Sterculiaceae, 110-123.]| & J. D. Hooker. Sterculiaceae. Gen. Pl. 1: 214-228. 1862 (July); 982-984. 1867. Boparp, M. Po age pg chez Cola nitida. VIII€ Congr. Int. Bot. Rapp. Commun. Sects. 7, 8: 259-261. — [Two or 3 separate or completely fused seedlings ay a seed: origin u wn. nares oie sur la natin des Sterculiées. Univ. Dakar Ann. Fac. Sci. ; Brown, W. H. ti ade of nectaries on the phylogeny of flowering plants. Proc. Am. Philos. Soc. 79: 549-595. 1938. [Includes Dombeya, Helicteres.] CAMMERLOHER, H. Biliitenbiologie I. Samm lung Borntraeger 15. 199 pp pls. 1, 2. Berlin. 1931. [Sterculia javanica R. Br. and S. laevis Wall., multicellular glandular “sugar-hairs” on adaxial side of sepals or forming a ring around base of androgynophore. | CaNDOLLE, A. P. pe. Byttneriaceae. Prodr. 1: 481-502. 1824. { = Sterculia- ceae. Cuattaway, M. M. The wood of Sterculiaceae. I. Specialization of the vertical wood parenchyma within the sub-family Sterculieae. New Phytol. 31: 119-132. pls. 4,5. 1932. . Tile-cells in the rays of the Malvales. bid. 32: 261-273. pl. 9. 1933. =, Bay Da ar in the Sterculiaceae. Forestry Great Britain 7: 93- 108. 1933 The wood anatomy of the Sterculiaceae. Philos. Trans. Roy. Soc. London B. 228: 313-366. 3 pls. 1937.* [Author’s summary in Trop. Woods 53: 62, 63. 1937.] Cuatrecasas, J. Cacao and its allies. A taxonomic revision of the genus Theobroma. Contr. U. S. Natl. Herb. 35: 379-614. pls. 1-12. 1964. [In- cludes: Anatomy of the wood, by W. L. Stern, 439-442; Pollen morphology, by G. Erdtman, 442-446; Cytology, 446, and Pollen incompatibility, 447- 449, by F. W. Cope; extensive bibliography, 599-605, supplemented by numerous references in footnotes, 439, 446, 447. Denay, C. L’appareil libero-ligneux foliaire des Sterculiacées. Ann. Sci. Nat. Bot. XI. 2: 45-132. 1941. [Thirty-one genera, 86 spp. Dumont, A. Recherches sur l’anatomie comparée des Malvacées, Bombacées, 64 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 Tiliacées, Sterculiacées. Ann. Sci. Nat. Bot. VII. 6: 129-242. pls. 4-7. 1887. {Sterculiaceae, 198-224, 240. ] Eputn, H. L. A critical revision of certain taxonomic groups of the Malvales. New Phytol. 34: 1-20, 122-143. 1935. [Sterculiaceae and Byttneriaceae distinct families, both derived from Tiliaceae, 124, 125, 136, 137; Scyto- petalaceae sui primitive within Malvales, 136, 137; Fremontieae belong in Bombacace FAIRCHILD, D. Florida plant immigrants. A tree that is famous for its smell. Occ. Pap. Fairchild Trop. Gard. 12: 4-6. 1943. [Sterculia foetida L., odor of tio very offensive. Freytac, G. F. A revision of the genus Guazuma. Ceiba 1: 193-225. 1951. [ Four spp. in i _ G. tomentosa HBK. a distinct sp. ] GazET DU CHATEL oe Recherches sur les Sterculiacées. Revue Gén. Bot. 52: 174-191, oi oas 257-284, 305-332. 1940. [Comparative morphology of the flowers and petiolar anatomy; for critical notes see C. DeHay, Bull. Soc. Bot. France 89: 76-78. 1942. Geuric, M. Beitrage zur peg ee gens re Malvales. Anatomie des Laub- blattes. Diss. 84 pp. 5 pls. Basel. 193 GERMAIN, R. Sterculiaceae. Fl. Congo 2 err, 1963. [Dombeya Cav., Melhania Forsk., by P. BAMPs. Gray, A. Byttneriacese. Gen. Pl. U. S. 2: 82-88. pls. 134, 135. 1849. [Melo- chia, Hermannia L. | GriseBACH, A. H. R. Flora of the British West Indian islands. i-xvi + 789 pp. London. 1859-1864. [Ayenia, sects., 91, 92; Melochia, sects., 93, 94; Waltheria, 94, 95.] Harvey, M. A revision of the genus Fremontia. Madrono 7: 100-110. 1943. LakeLa, O., & F. C. CratcHeap. Annotated checklist of the vascular plants of Collier, Dade, and Monroe counties, Florida. 95 pp. Fairchild Trop. Gard. & Univ. Miami Press, Coral Gables. 1965. [Sterculiaceae, 61. ] LEINFELLNER, W. Zur Entwicklungsgeschichte der Kronblatter der Sterculia- lee ines Osterr. Bot. Zeitschr. 107: 153-176. 1960. [Ten genera, 25 5 tsi i C. The comparative internal morphology of seeds. Am. Mid]. Nat. 36: 513-647. 1946. [Endosperm fleshy to firm-fleshy; embryo usually “spatulate,” “folded” in Ayenia, “investing” in Firmiana simplex Mo.penkE, H. N. A contribution to our knowledge of the wild and cultivated flora of Florida—1. Am. Midl. Nat. 32: 529-590. shes rhs gerne 543; Guazuma tomentosa, Moldenke 3781, “Dade Co.: edge of canal along Tamiami Trail, 25 miles west of Miami.’ | Pry, L. vAN DER. On the flower biology of some plants from Java with general remarks on fly-traps (species of Annona, Artocarpus, Typhonium, Gnetum, Arisaema and Abroma). Ann. Bogor. 1: 77-99. 1953. [Abroma augustum (L.) L. f., 90-95; also Ayenia glabrescens K. Schum. and Sterculia rubigi- nosa Vent., 95.] Raprorp, A. E., H. E. Antes, & C. R. BELL. Guide to the vascular flora of the Carolinas. 383 pp. Chapel Hill, N. C. 1964. [Sterculiaceae, 229, 2 Rao, C. V. Contributions to the embryology of Sterculiaceae —I. Jour. “Indian Bot. Soc. ea 180-197. 1949. [Five genera, 6 spp. ] Fl anatomy of some Sterculiaceae with sabia references to the —. of stamens. Ibid. 237-245. 1949. [Three gen en grains of Sterculiaceae. /bid. 29: 130-137. “19800. [Nine genera in 5 tithes: key to the genera, based on pollen. } 1966} BRIZICKY, GENERA OF STERCULIACEAE 65 —. Contributions to the embryology of Sterculiaceae—II. Wealtheria indica Linn. Ibid. 163-176. 1950b. . Contributions to the embryology of Sterculiaceae. III. Melochia corchorifolia L. Ibid. 30: 122-131. 1951. Floral anatomy of some Malvales and its bearing on the affinities of families included in the order. /bid. 31: 171-203. 1952. [Sterculiaceae, 171-179, 196, 197. Contributions to the embryology of Sterculiaceae. IV. Development of the gametophytes in Pterospermum suberifolium Lam. Ibid. 251-260. 1953. [Primitive embryological characters: multicellular archesporium and the development of 2 or 3 embryo sacs in an ovule. ———. Contributions to the embryology of Sterculiaceae — V. Jbid. 32: 208- 238. 1954. [Eight genera, 10 spp.; seed structure in Pterospermum.] Rosyns, A. Sterculiaceae. Jn: R. E. Woopson, Jr., & R. W. Scuery, Flora of Panama. Ann. Missouri Bot. Gard. 51: 69-107. 1964. Rurr, O. Phylogenie des Columniferen-Astes der Dicotylen. Bot. Arch. 31: 1-139. 1930. [Sterculiaceae, 17, 18, 45-47, 137, 138.] SAUNDERS, E. R. On carpel polymorphism. IV. Ann. Bot. 45: 91-110. 1931. [ = Floral aged of Sterculiaceae, 10 genera. | SCHULTZE-MorTEL, W. Malvales. Jn: H. Metcuior, A. Engler’s Syllabus der Piuiwieteaice ed. 12. 2: 304-316. 1964. [Sterculiaceae, 313-31 SCHUMANN, € cian age In: C. F. P. Martius, FI. Brasil. 12(3): 1-114. pls. 1-2 6a. ——_.. a aE ae pe der ayes ep oa Jahrb. Bot. Gart. Berlin 4: 286-332. pls. 3,4. 1886b. [Byttneri ———. Sterculiaceae. Nat. Pflanzenfam. ILL. 6: ab 1890. SHENSTONE, F. S., & J. R. Vickery. Occurrence of cyclo-propene acids in some plants of the ‘order Malvales. Nature 190: 168, 169. 1961. [ Brachychiton, 2 spp., seeds; Sterculia foetida, leaves and seeds 7 WINKLER, H. Buti zur Morphologie und Biologie tropischer Bliiten und Friichte. Bot. Jahrb. 38: 233-271. 1907. [Sterculiaceae, 248-258. ] “ KEY TO THE GENERA OF STERCULIACEAE General characters: trees to herbs; hairs simple and/or stellate; leaves alter- nate, simple, petioled, stipulate; inflorescences cymose; flowers small, regular, hypogynous, bisexual or rarely unisexual; perianth 5-merous, sepals and petals distinct, the latter sometimes wanting; stamens 5 to many, monadelphous, the antisepalous 5 sterile or wanting, rarely fertile; gynoecium usually 5-carpellate, syncarpous, rarely apocarpous, or 1-carpellate, ovules 2 to many in a locule; fruit simple or rarely aggregate. A. Plants arse leaves large, the blade commonly 15-20 cm. long and 20-30 cm. broad, usually deeply 3-5-lobed, long petioled; is apni ample ae 20-50 cm. long; flowers apetalous, unisexual; stamens (10-)15, on a conspicuous androgynophore; fruit of 5 stalked follicles 65 cm. long, leaflike after dehiscence, with 2 seeds on each margin. 1. Firmiana. Plants herbaceous, suffruticose, or ar fruticose; leaves not more than 8 cm. long usually unlobed; inflorescences small or flowers solitary; flowers og aicting bisexual; rr ti 5; fruit simple, capsular. B. Flowers axillary, solitary or in fascicles of 2 or 3; petals long clawed, aa hooded: Pay alee ne attached by the tips to the apex of staminal 66 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 tube, the corolla appearing parachute-like; androgynophore present; anthers 3-locular; capsule covered with short subcylindrical prickle-like protuberances; seeds coarsely reticulate. ................ 2. Ayenia. Flowers in dense headlike or umbel-like inflorescences; petals short clawed, adnate at base to the staminal tube, + spre reba androgyno- phore wanting; anthers 2-locular; is and seed smoot C. Corolla orange-yellow; gynoecium 1-carpe ellate, stigma brushlike, style 1, ovary 1-locular; ie inclosed in persistent calyx, 1(2)- ne Sn asia Vine k he ee hee ee 4. Waltheria. Corolla pink to sissies, or white, very rarely gunn: gynoecium 5-carpellate, stigmata 5, subulate to subclavate, styles 5, connate at base to half their length, ovary 5-locular; capsule ne say by per- sistent calyx, 5-10-seeded, 5—10-valved. .......... Melochia. w QO Tribe STERCULIEAE 1. Firmiana Marsili, Saggi Sci. Lett. Accad. Padova 1: 106. 1786. Deciduous trees [rarely shrubs], with minute stellate-tomentose pubes- cence. Leaves large, membranaceous, palmately 5—7-veined and -lobed, deeply cordate [or truncate to rounded| at base, long petioled; stipules caducous. Inflorescences large axillary and terminal thyrses; bracts cadu- cous. Flowers regular, hypogynous, apetalous, unisexual by abortion (the plants monoecious), pedicels articulate in the upper half. Calyx with a short [rarely rather long], campanulate to urceolate tube with a dense transverse band of long hairs in the throat and a nectariferous disc around the androgynophore, the 5 lobes longer [or shorter| than the tube, some- what fleshy, yellow-green [to yellow-orange, or white], reflexed. Stamens usually 15; anthers subsessile to sessile on margin of a +campanulate to subcylindrical (in ¢ ) or saucer-like (in @ flowers) staminal tube at the apex of a conspicuous androgynophore, forming a globular clump con- cealing the pistillodium in ¢, and encircling the ovary base and indehis- cent in @ flowers, 2-locular, the locules contiguous, often unequal; pollen medium sized, prolate-spheroid, 3-colpate, reticulate. Gynoecium nearly apocarpous, but appearing syncarpous, the densely pubescent, closely ap- pressed carpels usually connate by the upper part of the styles; rudimen- tary in 4 flowers; stigmata 5, subpeltate, spreading, closely appressed adaxially and appearing as a single 5-lobed stigma; styles 5, connate for a short distance near their apices; ovaries 5, each with 4 submarginal, sessile, anatropous, nearly horizontal ovules. Fruit of 5 stalked membranaceous follicles, each opening before maturity into a leaflike body exposing 2—4 on its margins. Seeds subglobular; the outer seed coat brown, sub- coriaceous, subreticulately wrinkled, the inner crustaceous; endosperm copious, firm, fleshy; embryo straight, axial; cotyledons thin, suborbicular, cordate at base; radicle short, stoutish, inferior. Type species: Sterculia platanifolia L.f. = F. simplex (L.) W. F. Wight. (Named in commemora- tion of Count K. J. von Firmian, 1716-1782, Austrian statesman, Governor of Lombardy under Maria Theresa.) 1966 | BRIZICKY, GENERA OF STERCULIACEAE 67 An Asiatic genus of about ten species, primarily tropical, ranging from the western Himalaya to eastern China, the Philippines, Malesia, and Papuasia. Firmiana simplex (F. platanifolia (L. f.) Schott & Endl.), Chinese parasol tree or phoenix tree, 2” = 40, indigenous to subtropical and warm-temperate China (Yunnan to Kwantung, north to Shensi and Hopeh) is grown as an ornamental in our area and has become naturalized locally, mostly in the Coastal Plain, from northern Florida to North Carolina, southern Alabama, and Louisiana. It is a fast-growing tree with smooth gray-green bark, occurring on roadsides, in city lots, thickets, and mixed deciduous woods, The inflorescences are either predominantly staminate or carpellate with a few flowers of the opposite type appearing by the end of blooming; both kinds of inflorescences occur on the same tree. Flowers intermediate between staminate and carpellate have also been observed occasionally. The staminate inflorescences seem to flower before the carpellate. Although insect-pollination may safely be assumed for the genus (presence of a nectariferous disc, petaloid sepals, etc.), no data on pollinators are avail- able. Cross-pollination is favored by the unisexual flowers and the some- what earlier expansion of the staminate ones, but some self-pollination is very probable. In both floral anatomy and embryology Firmiana seems to be very similar to Sterculia L. Chromosome counts for three species are 2n = 40. The seeds, which remain firmly attached to the margins of the opened, leaflike, apparently deciduous follicles, are probably dispersed by win The gets is closely related to Scaphium Schott & Endl. (follicles two- seeded), Pterocymbium R. Br. (follicles with a single basal seed), and Sterculia (follicles leathery to woody, seeds dropping off the carpels, radicle opposite the hilum). The caffeine-containing seeds of Firmiana simplex are sometimes roasted and used to make a tea in China. The timber finds application in the paper industry, and the phloem fibers are utilized in making cordage. REFERENCES: Under family references see Dumont (pp. 198-201, under Sterculia), GazET DU CHATELIER (pp. 211-259), Raprorp et al. (p. 229), Rao (1954, as to Sterculia colorata), SAUNDERS (pp. 118, 119, under Sterculia), and SCHUMANN (1890, p masciicixins, A. J.G. H. A note on some African Sterculiaceae. Bull. Jard. Bot. Bruxelles 24: 335-338. 1954. [Erythropsis Lindl. ex Schott & Endl. con- generic with Firmiana; some African spp. of Sterculia and Firmiana trans- ferred to Hildegardia Schott & Endl. ] e genus Firmiana Marsili (Sterculiaceae). (Revised.) peng 4: 281-310. 1957. [For an earlier publication see Commun. Forest R Inst. Indon. 54: 3-33. 1956.*] Miscellaneous botanical notes 2. Reinwardtia 2 375-418. 1961. [Firmiana, 2 new names and 1 new near 383-390. ] Kozima, H., T. Araki, K. Waka, T. TAKAHASHI, & Y. Matsuo. Chemical constituents of a young leaves of Firmiana platanifolia Schott et Endl. (In Japanese.) Jour. Chem. Soc. Jap. Pure Chem. Sect. 82(1): 134. 1961.* 68 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 Morrison, B. Y. Firmiana simplex. Natl. Hort. Mag. 38: 58, 59. 1959. [An estimate of horticultural value. | Ny1 [YEN], T. K. Observations on the sex differentiation and flower anatomy of Firmiana simplex. Contr. Biol. Lab. Sci. Soc. China 5(3): 1-25. 1929. Yen, T. K. Carpel dehiscence in Firmiana simplex. Bot. Gaz. 93: 205-212. 1932. Tribe BYTTNERIEAE DC. 2. Ayenia Linnaeus, Sv. Vet.-Akad. Handl. 17: 23. pl. 2. 1756. Usually decumbent [prostrate or erect] subshrubs [or deciduous shrubs, rarely small trees], + pubescent with simple and/or stellate hairs [or glabrous]. Leaves small (to 1 x 0.7 cm.) [to rather large, 21 x 10 cm.]}, membranaceous [rarely coriaceous], 3—5-veined at base, toothed [or en- tire], petioled; stipules small, subulate, persistent at least in part [or cadu- ous]. Flowers pediceled, axillary, solitary or in fascicles of 2 or 3 [to 12, or arranged in axillary or terminal, simple, peduncled, sometimes fascicled or panicled cymes], small 2—3[-7] mm. long, pink to red, bisexual [rarely unisexual, or sometimes cleistogamous]|. Sepals 5, oblong-lanceolate, con- nate at base. Petals 5, slender clawed, + rhombic (in ours), -- hooded in the upper part, cleft and 2-dentate [or entire] at apex, with [or without] a clavate [to linear] abaxial appendage below the cleft, inflexed toward center of flower and attached by means of teeth to apex of staminal tube, the whole corolla resembling a parachute with the style protruding through its center. Androgynophore usually conspicuous [rarely wanting]. Sta- mens 5, antipetalous, staminodia 5, antisepalous, all connate into a + campanulate staminal tube; upper parts of filaments of fertile stamens distinct, reflexed; upper parts of staminodia distinct, + triangular [some- times hooded], recurved; anthers with 3 contiguous locules; pollen small to medium sized, oblate-spheroid, 3-aperturate, reticulate [rarely spinulif- erous|. Gynoecium usually 5-carpellate, syncarpous; stigma subcapitate, 5-lobed; style single; ovary 5-locular, minutely tuberculate, with axile placentae; ovu 2, hemianatropous, somewhat curved toward the placenta, superposed in each locule, the lower one rudimentary (in ours). Fruit a small, subglobular, usually 5-locular capsule covered with short, subcylindrical [or wartlike to trigonous| protuberances, both septicidally and loculicidally dehiscent, 10-valved. Seed solitary in each locule, nar- rowly ovoid, slightly curved, abruptly narrowed into a subcylindric-coni- cal beak at apex; seed coat crustaceous, dark brown, with a few broad longitudinal and transverse ribs forming a few-meshed reticulum [or corrugated to smooth], minutely black-verruculose; chalazal knot round, subbasal, close to hilum; endosperm wanting; embryo straight; cotyledons bisect [or bilobed? |, plicate, spirally rolled around an elongated superior radicle. Type species: A. pusilla L. (Name commemorating Louis de Noailles, 1739-1777, Duc d’Ayen, later Maréchal de Noailles, promoter of sciences, amateur gardener, and owner of a famous garden in St. Germain; see Linnaeus, Schwed. Akad. Wiss. Abh. 18: 25. 1757.) 1966 | BRIZICKY, GENERA OF STERCULIACEAE 69 A tropical American genus of about 70 species, extending north to Cali- fornia, Arizona, New Mexico, Texas, and Florida, and south to Argentina and Uruguay, centered in Mexico and northern Central America, with a secondary center in Bolivia, Paraguay, and central and eastern Brazil; one species of sect. AYENIA in our area. Grisebach subdivided the genus into three sections differing mainly in the structure of the petals, the length of the distinct parts of filaments, the characters of the surface of fruits and seeds, and in the sculpture of the pollen. Cristébal’s species concept, although perhaps a little too narrow, is followed here. The Cuban Ayenia euphrasiifolia Griseb. occurs in pinelands on the Florida Keys and in southernmost peninsular Florida (Dade and Monroe counties).* It is closely related to and perhaps not specifically distinct from the Cuban-Bahamian A. tenuicaulis Urb. Four Mexican species range into Texas, New Mexico, and Arizona, and one reaches California. The pink to red, sometimes foetid flowers are pollinated by small in- sects. The petals perfectly separate the stigma from the anthers, prevent- ing self-pollination. Unisexual flowers occur at least in the Brazilian A. and androgynophore and mostly introrse anthers have been recorded in four species (Cristébal) The comparative morphology and ontogeny of the cucullate petals in Byttnerieae (Schumann, Leinfellner) suggest that in Ayenia (§ Ayenia) the petal blade is reduced to a clavate [or linear] appendage, while the hooded rhombic portion of the petal (‘‘face” or ‘“‘mask”) is homologous with the upper part of the claw. Chromosome counts made for eleven species are 2” = 20 and 40, the latter number occurring in three species. Fruits in Ayenia, which dehisce first septicidally and then loculicidally, as well as along the adaxial (ven- tral) suture, the valves detaching themselves from a persistent central column, may be regarded as a capsule type transitional to the schizocarp. There seem to be no special adaptations for seed dispersal, and seed dis- persing agents are unknown. The genus is closely related to the pantropical Byttneria, which differs mainly in petals with an evident apical portion (corresponding to the petal blade) not attached to the staminal tube. REFERENCES: ae family references see DuMont (p. 217), GRISEBACH (pp. 91, 92), LAKELA & CRAIGHEAD (p. 61), LEINFELLNER (pp. 158-162), and SCHUMANN (1886a, pp. 101-108, pls. 23, 24; 1886b, pp. 294-297; 1890, pp. 83-85 CristopaL, C. L. Revision del género Ayenia L. (Sterculiaceae). (English ummary.) Op. Lilloana 4: 1-230. 1960 other species seems to occur in Florida. Small’s record of A. pusilla from ions ich Ecuador and northern Peru (Cristébal, p. 191). All specimens from Florida labeled as A. pusilla were annotated by Cristébal as A. euphrasiifolia. 70 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. 47 Nueva contribucién al estudio del género Ayenia L. a Anales Instituto Biol. México 32: 191-200. 1961. [One new Mexican sp. and several new records from Mexico, Central America, a "Colombia: first record of cleistogamous flowers in A. fruticosa Brandegee. | Notas sobre el género Ayenia. Lilloa 31: 207-212. 1962. [A new sp from Colombia described; 1 sp. from Cuba redescribed and illustrated. ] Tribe HERMANNIEAE Schott & Endl. emend. Benth. 3. Melochia Linnaeus, Sp. Pl. 2: 774. 1753; Gen. Pl. ed. 5. 304. 1754. Herbs to shrubs [or rarely trees]; pubescence of simple and/or stellate hairs. Leaves narrow to broad, palmately 3—5-veined at base, serrate, petioled; stipules small, persistent at least in part. Inflorescences few- flowered umbel-like peduncled cymes or many-flowered dense headlike thyrses composed of 2- or 3-flowered dichasia or monochasia, or thyrses sometimes secondarily arranged into interrupted spikelike panicles, axillary and/or terminal (rarely opposite the leaves); bracts persistent or cadu- cous. Flowers small, bisexual, heterostylous, often with an involucel of 3 or 4 distinet bract(let)s at base, pediceled or sessile. Calyx + campanulate, 5-dentate to -cleft, persistent, nectariferous at base. Petals 5, pink to purple, usually yellow or white at base, or white, or rarely yellowish, spatulate to oblong, clawed, adnate at base or higher up to staminal tube, longer than the sepals, marcescent. Stamens 5, antipetalous, the staminal tube cylindrical, inflated after fertilization; free part of filaments elongated in short-styled, very short in long-styled flowers; 5 obsolete, toothlike staminodia sometimes present in short-styled flowers; anthers 2-locular, the locules contiguous before and + divaricate after fertilization; pollen medium sized, spheroid to oblate-spheroid, 3-colpate, finely reticulate. Gynoecium 5-carpellate, syncarpous; stigmata 5, subulate or filiform to + clavate, papillose, or shortly and racemosely branched; styles 5, distinct or connate to half their length; ovary sessile to short stipitate, 5-locular, the placentation axile; ovules 2 in each locule, superposed, ascendent, anatropous, apotropous. Fruit a 5-locular, loculicidal and/or septicidal capsule [rarely a schizocarp], the locules 1- or 2-seeded. Seed obovoid, nearly round in cross-section or + angular on adaxial (ventral) surface, with a prominent humplike chalazal knot and thin threadlike raphe; outer seed coat thin, membranaceous, usually obliterated, the inner crustaceous; endosperm fleshy; embryo straight, spatulate, cotyledons large, foliaceous; radicle short, inferior. LEcrotypr species: M. corchorifolia L.; see Medi- cus, Kiinstl. Geschl. Malven-Fam. 10. 1787, and Britton & Millspaugh, Bahama FI. 277. 1920.4 (Name from melokhiya or melukhiye, an Arabic * Hitchcock & Green (Int. Bot. Congr. Cambridge 1930. Nomencl. Propos. Brit. i tances. his Genera Plantarum (ed. 5. 1754, p. 304) Lin _ Regge ie Melochia to Dillenius, referring to plate 176 of Hortus Elthamensis on which Dillenius’s Melochia corchori folio (M. corchorifolia L.) was figured. Thus ‘eames himself seems to 1966 | BRIZICKY, GENERA OF STERCULIACEAE 71 name for Corchorus olitorius L., widely cultivated as a potherb in the east- ern Mediterranean, particularly in Egypt, since ancient times.) A pantropical, predominantly American genus of about 60 species, with one extending into warm-temperate regions; four species (one naturalized) in our area. The three predominantly American sections of the genus established by Grisebach are delimitable by the shape and/or the kind of dehiscence of fruits and sometimes by the type of inflorescence and presence or absence of a floral involucel. The fourth, the Indo-Malayan-Pacific sect. VISENIA (Houtt.) Benth. & Hook. differs in the arborescent habit and usually apically winged seeds of its species. Section Metocuia (§ Riedlea (Vent.) Griseb.; Riedlea Vent.). Cap- sules sessile, subglobular, slightly 5-lobed, the lobes rounded; inflorescences dense, many-flowered, peduncled to sessile headlike thyrses; flowers sessile, with an involucel. About 30 (or more) species, centered in tropical America, a few in the Old World, two endemic on Madagascar; M. melissifolia Benth. in America and Afric The paleotropical Melochia bpehisar chocolate weed, with peduncu- late, usually terminal inflorescences and long-petioled leaves glabrous or with scattered hairs on the veins beneath, has become naturalized in old, especially sandy, fields, cultivated grounds, and waste places in the Coastal Plain, from southern peninsular Florida north to southeastern North Carolina and west to southeastern Texas. In the New World it is also recorded from Panama, suggesting that it may become a pantropical weed. Melochia villosa (Mill.) Fawc. & Rendle (M. hirsuta Cav., Riedlea hirsuta (Cav.) DC.), with sessile axillary and terminal inflorescences forming interrupted spikelike thyrses and short-petioled leaves silky- pubescent or glabrescent beneath, is widely distributed in tropical America. It reaches southern Florida, where it occurs on moist sands in marshy lands, flatwoods, and pinelands. In the Old World this species has become naturalized at least on Madagascar. Section Morucutia (Medic.) Brizicky (§ Eumelochia Griseb.; Moluchia Medic.). Capsules + stipitate, pyramidal, deeply 5-lobed, the lobes wing- have sonar the type of this genus, and the generic description does not napus folia as the lectotype species, Britton and Millspau typification of the genus, which had already been indicated by Linnaeus and effected by Medicus. ; With this typification the following nomenclatural changes in two of three sections recognized by Grisebach are necessary elochia sect. Melochia (Melochia sect. Riedlea (Vent.) Griseb. Fl. Brit. W. Indian Is. 93. 1859.) Type species: M. corchorifolia L. Melochia sect. Ee A gga Brizicky, stat. nov. (Melochia sect. Eumelo- chia Griseb. FI. Is. 93. 1859. Moluchia Medic. Kiinst hl. Malven-Fam. 10. oer) en: species: M. pyramidata L. (Moluchia hevbacsn rfi.). Medic., nom. supe 72 JOURNAL OF THE ARNOLD ARBORETUM [VOL. 47 like; inflorescences 3-10-flowered umbel-like peduncled [sometimes panicled| cymes; flowers pediceled, without an involucel. About nine species, eight American, one endemic on Madagascar. Melochia tomentosa L. (Moluchia tomentosa (L.) Britt.), with axillary cymes and tomentose leaves and capsules, occurs in pinelands in southern- most Florida (Dade County) and ranges widely in tropical America. Melochia pyramidata L., with inflorescences opposite the glabrous leaves and capsules with scattered minute stellate hairs, has also been reported from Dade County, Florida (Lakela & Craighead). Widely distributed in tropical America (north to Mexico and Texas), it has been introduced into the Old World, where it seems to have become a pantropical weed. In Texas the flowers of this species open at noon and close in late afternoon {from a collector’s field note]. The nectaries and the bright-colored petals indicate insect-pollination, but pollinators have not been observed. Heterostyly seems to be common within the genus. Floral anatomy and embryology have been investigated only in M. corchorifolia. The single chromosome count recorded is 2n = 14 (in M. bracteosa F. Hofim. = M. melissifolia Benth. var. bracteosa (F. Hoffm.) K. Schum., from West Africa). Although fruits in Melochia mostly are septicidal and/or loculicidal capsules, types intermediate to schizocarps (“pentacoccous capsules’) also occur (sect. MOUuGEOTIA (HBK.) Griseb.). Seed-dispersal agents are unknown, except in the species of sect. VisENIA, the usually winged seeds of which are said to be dispersed both by wind and ocean currents. The very wide distribution of M. pyra- midata and M. corchorifolia is due mainly to human agenc Melochia is closely related to Waltheria L. The genus is in need of a modern revision. REFERENCES: Under family references see GERMAIN (pp. 255-258), Gray (pp. 85, 86, pl. 134), GRISEBACH (pp. 93, 94), LaKELA & CRAIGHEAD (p. 61), RADFORD ef al. (p- 230), Rao (1950a; 1951; 1952, p. 179), Rospyns (pp. 77-85), and SCHUMANN (1886a, pp. 27-49, pls. 7-10; 1890, pp. 80, 81). GARCKE, “A ee einige Arten von Melochia. Bot. Jahrb. 12(Beibl. 27): 29- ve 0. JoNngEs, .. Jr. Additional Alabama weeds. Castanea 26: 136-138. 1961. sue it (M. corchorifolia) is one of the most troublesome weeds in os fields in Mobile and Baldwin counties, but was not observed rth of those counties.” : mel F. K. Ueber einige kiinstliche Geschlechter aus der Malven-Familie, denn der Klasse der Monadelphien. 158 pp. Mannheim. 1787. [Moluchia Medic., 10-12.] 4. Waltheria Linnaeus, Sp. Pl. 2: 673. 1753; Gen. Pl. ed. 5. 304. 1754. Herbs or subshrubs [or shrubs], erect, single stemmed or much branched from the base, rarely diffuse and prostrate; pubescence mainly of stellate hairs. Leaves usually oblong to ovate, 3—5-veined at base, toothed, peti- oled; stipules small, lanceolate-subulate, caducous. Inflorescences axillary 1966 | BRIZICKY, GENERA OF STERCULIACEAE 73 and terminal, peduncled to sessile, dense headlike compound cymes (or thyrses?) composed of 2- or 3-flowered sessile or subsessile dichasia or monochasia, sometimes forming short-branched or spikelike thyrses. Flowers small, bisexual, homostylous |or often heterostylous], sessile, with 1 or more bractlets at base. Calyx persistent, the tube campanulate-tur- binate, slightly 5-angled, 10-veined, nectariferous at base, equaling or ex- ceeding the 5 lanceolate-subulate lobes. Petals usually orange-yellow, spatulate, clawed, adnate at base to the staminal tube, longer than the sepals, marcescent. Stamens 5, connate into a tubular staminal tube equal- ing [or shorter than] the pistil; free upper parts of filaments very short [or elongated]; anthers 2-locular, the locules contiguous; pollen medium sized, usually oblate-spheroid, (3)4(—8)-colpate, reticulate. Gynoecium l-carpellate; stigma brushlike [or clavate?|; style single, solid, excentric (sublateral) ; ovary sessile, 1-locular, subtruncate and hairy at apex; ovules 2, submarginal, superposed, ascendent, anatropous, apotropous. Capsule obliquely obovoid, hairy at apex, 1(2)-seeded, 2-valved. Seed obliquely obovoid, slightly compressed laterally, with a shallow adaxial raphe-bearing groove, dark brown to black except the rusty-colored round or elliptic chalazal knot; endosperm fleshy; embryo straight; cotyledons large, foliaceous; radicle short, inferior. LecrotyPE spEcirs: W. indica L (W. americana L.); see R. Brown in Tuckey, Narr. Exped. Congo 484. 1818.5 (Name commemorating Augustin Friedrich Walther, 1688-1746, professor of botany in Leipzig; see Linnaeus, Crit. Bot. 95. 1735.) An almost exclusively tropical American genus of about 30 (or fewer) species, one pantropical and two African (one endemic on Madagascar) ; one species in our area The polymorphic, pantropical Waltheria indica (W. americana of al- most all American authors), a usually + canescent herb or subshrub, occurs in hammocks, pinelands, and on sand dunes on the Florida Keys and in southern peninsular Florida, north to Highlands and Lee counties. It was recorded as adventive on the coast of Alabama (Small, Man. SE. Fl. 863. 1933) but does not seem to occur there at present Heterostyly is fairly frequent within the genus. Pollination by wasps and bees has been recorded for W. indica and W. viscosissima St. Hil. in Brazil. The floral anatomy and embryology of W. indica appear to be very similar to those of Melochia corchorifolia, but embryo development con- forms to the Polygonum variation of the Asterad type (in contrast with the Urtica variation in Melochia and other genera of Sterculiaceae thus far investigated).6 The single recorded chromosome count is 2m = 14. Seed *“Robert Brown (in Tuckey, Narr. Exp. Congo 484. 1818) appears to _ been the first to unite Waltheria americana L. (Sp. Pl. 673. 1753) and W. indica L. (Sp. Pl. 673. 1753). He adopted the name Waltheria indica for the combined species, and this name must accordingly be retained” (Art. 57, Int. Code Bot. Nomencl. 47. 1961). e following comment by Rao (1950b, p. 175) is enigmatic: “Embryological u n Prantl.” In the references to Rao’s article “Engler, A gor Prantl, K. (1928). — Natiir~ liche Pflanzenfamilien, 2. Auflage” is cited. This incomplete and apparently erroneous 74 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 dispersal by wild goats and perhaps by some granivorous birds (e.g., doves) has been recorded for W. indica. The genus is closely related to Melochia. A modern revision is desir- able. REFERENCES: Under family references see GERMAIN (pp. 258, 259), Rao (1950a & b; 1952, p. 179), Rospyns (pp. 74-77), and SCHUMANN (1886a, pp. 50-68. pls. 11-14; 1890, p. 81). Diers, L. Der Anteil an Polyploiden in den Vegetationsgiirteln der West- kordillere Perus. Zeitschr. Bot. 49: 437-488. 1961. [2m = 14 in W. ovata Cav., 450.] reference provides no clue to Rao’s statement. The order Malvales (or Columniferae) has not appeared in the second edition of the Natiirlichen Pflanzenfamilien, and the only volumes of this series published in 1928 are Bd. 2 and Bd. 5, with treatments of Algae and Fungi respectively. UW 1966 | GREEN, LONICERA TATARICA COMPLEX 7 IDENTIFICATION OF THE SPECIES AND HYBRIDS IN THE LONICERA TATARICA COMPLEX P. S. GREEN THE NEED TO VERIFY THE IDENTITY of the numerous plants of Lonicera growing in the Arnold Arboretum became an urgent matter with the start of a karyological survey of the genus which is at present being under- taken by Mrs. Lily Riidenberg. Alfred Rehder, whose revision of the genus, published in 1903 (Annual Rep. Missouri Bot. Gard. 14: 27-232), is still the latest comprehensive account, identified the older plants in the collection but, as might be expected, there have been many new accessions. The determination of most species presents no difficulties, but under cultivation hybridization has become intensive in the group around L. tatarica. Keys already exist for the identification of species but not for these very confusing hybrids. Furthermore, some of the bush honey- suckles most highly prized in horticulture belong to this difficult group, so it was decided, after having studied them for some time, that an attempt should be made to construct a key and provide aids to their identification. together it is probable that much of the seed set is of hybrid origin. In- trogressive hybridization and the resultant hybrid swarms are well known in the wild but Lonicera provides a good example among cultivated plants. The first records of hybrids occur towards the end of the last century, when Zabel, in particular, was cultivating and studying the genus. Ap- parently the hybrids arose spontaneously and without artificial aid, and their hybrid nature was perceived through their intermediate characters. At first they were propagated vegetatively but later plants were grown from seed collected from these hybrids and, where either or both parental species were growing nearby, it is clear that back-crossing had often taken place. The result today is that collections of shrubby honeysuckles in cultivation invariably include some hybrid plants which to varying degrees resemble one parent more than the other. Furthermore, triple hybrids occur as well, for the hybrids themselves are interfertile with the other species in the group. FicurE 1 portrays, diagrammatically, the main complex with which Lonicera tatarica is involved. As will be seen, three other species are in- cluded and four interspecific hybrids. Of these latter, three have further crossed with other species and it is quite probable that if an attempt were made it might prove possible to create every conceivable cross involving these species. 76 JOURNAL OF THE ARNOLD ARBORETUM L. morrowil [voL. 47 L. X muscaviensis \ \ L.X bella - Pia \u ay ar L. X muendeniensis ea ae L. KX minutiflora \ L at ruprechtiana ox, L. X notha ey = me ss i \ \ 7 L. tatarica \ 2S eee L. X salicifolia \ — = ~ \ ? i \ Me L. X xylosteoides L. xylosteum os Lonicera tatarica. Diagram representing the main complex of hybrids involved with Hybridization has brid being called L. & amoena. In FicuRE 3 typical leaf outlines and ovary Lonicera tatarica also hybridizes with L. korolkowii, the resultant hy- types for these two species and their hybrid are illustrated. also n reported between Lonicera tatarica and L. nigra. Zabel (Gartenflora 38: 524. 1889) incorrectly referred the plant to 2 [ee ereaen 1966 | GREEN, LONICERA TATARICA COMPLEX 77 L. micrantha, but no recent material of it has been seen in this investiga- tion and it is doubtful whether it is still in cultivation. For these reasons it is not included either in the illustrations or the key. In the identification of these plants the characters of the leaves, and of the ovary with its bracts and bracteoles have proved of particular help with both living and dried material. For this reason they are illustrated in FIGURES 2 and 4 respectively. The arrangement of FIGURE 2 is the same as that of FicuRE | in order to illustrate most readily the interrelation- ship of the hybrids and double hybrids. It proved impracticable, how- ever, to arrange the drawings of ovary types in the same manner, instead, they are set out in simple alphabetical order. Other characters of value not portrayed in all the figures are the degree of hairiness or glaucousness of the leaf and, in fresh flowering material, where it may be easily ob- served, the extent to which the corolla lobes are reflexed or recurved and the position, in the fully open flower, in which they are displayed in relation one to another. For some of the species this is also shown in FIGURES 5, 6, and 7 by means of close-up photographs of living flowers. Rehder, in his revision, classified Lonicera tatarica and L. korolkowii in a separate subsection from the other species involved in this complex of hybrids, but in view of the ease with which fertile hybrids are produced between the species separated by this classification it is very doubtful whether subsectional recognition is justified. The main distinction used was whether or not the flowers turned yellow on fading, and certainly those of LZ. tatarica and L. korolkowii do not appear to turn yellow or orange-yellow as is strongly characteristic of the other species and of their hybrids. Where fresh material is available for observation this yellowing of the corolla seems to be a useful character and appears in L. tatarica backcrosses, even when the hybrid nature of the plant is otherwise barely perceptible. It is not always easy to tell just how much backcrossing has taken place, for the parental species are themselves very variable. Lonicera tatarica, for instance, has many variants, coming as it does from a wide geographi- cal area. Nor have many of the cultivated plants of this species or its hy- brids been considered distinct and distinguished enough to be named as cultivars, the same hybrid name is applied, therefore, to any plants which involve the same two parental species, whether or not backcrossing has taken place. Pose In the key that follows it is hoped that plants of hybrid origin will key out under the appropriate name without any difficulty. However, as many of the plants in cultivation arose from seed collected in botanic gardens, nurseries, etc., where more than one species are often grown together, it should be realized that a high proportion may be crosses or backcrosses. For this reason the key should be used in close association with the figures. Furthermore, it should be borne in mind that species like Lonicera tatarica exhibit considerable variation even without hybridization. Fortunately perhaps, the species show less variation in cultivation, due to their relatively few introductions, than they do throughout their total distributional range 78 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 K=> L. X muscaviensis L. X bella er {— KS L, X muendeniensis ES x mee rene iy CESS X notha L. K=> Ss as ™~ “ies < ” ifoli Cee ee L. X xylosteoides L,. xylosteum Fic. 2. Leaf-outlines of species and hybrids of Lonicera arranged to be com- parable with FIGURE 1 in the wild, but it is important to realize that the leaf shapes and ovary types portrayed in the figures were selected as being ‘“‘average’’ for the species and, it is hoped, typical of the taxa as best known in cultivation or, in the case of hybrids, of plants of the first generation. 1966] GREEN, LONICERA TATARICA COMPLEX 79 Key To LonIceRA ghee ITS ag AND Hysrips ULTIVATI 1. Leaves usually broadest at or below the middle. 2. Plant glabrous throughout (rarely a few hairs on a mee Moca sionally a few glands on the calyx lobes and bractlets. .... L. tatarica. 2. Plant + hairy, especially the leaves, pedicels, and bra 3. Bracts usually shorter than the ovary, sails a slightly lon- ger (in L. X amoena sometimes all longer); leaves elliptic or ovate; whole plant + glaucous. 4. Back of bracts and calyx lobes hairy. .......... . korolkowii. 4. Back of bracts and calyx lobes glabrous (only aint). Seaatees a WwW moena, . Bracts as long as, or longer than the ovary, rarely peace ones shorter; leaves lanceolate, oblanceolate or a (or even ovate in ell in Xx a and occasionally in LZ. X morrowit). Ss. rae (from base of tube to i a oe lobe) usually not more mm. long, rarely up to 2 6. foes of corolla tube 2.5—-4 % ei from base of ie a tip of upper lobes 10-13 mm . : ot LS a eg An gate ruprechtiana. 7; Corolla from base aes to tip of upper lip 7-9 mm. long; most flowers abortive. ...............- L. X salicifolia. 6. Length of corolla tube sen 8. Bractlets 3/4 to ia re length of the es (not includ- ing the calyx Inbes). = .-..-+..24+.-: . X minutiflora. 8. Bractlets up to VS length of the ovary, een 2/3 as long. L. X muendeniensis. 5. Corolla (from base of tube to tip of upper lobe) 12 mm. long or more. 9. Bractlets between 3/4 and the full length of the ovary, rarely 1/2 its length (not including the calyx teeth). 10. Calyx, bractlets, and bracts glandular and hairy. 11. Bracts only sparsely hairy and glandular; margins of bractlets sparsely hairy and glandular. cl pea tah tenn ok oy eee ata ha nV L. minutifilora. 11. Bracts + densely hairy (and bern margins of resi usually with numerous oar’ airs (and MUNN ee for RAs oa ts uscaviensis. 10. oa Tawa and bracts hairy but without poe er SIN ee Olt eh be Mant See eee ort er re ree . morrowii. Bractlets usually not more than 1/2 the length of ovary, occa- sionally up to 2/3 its leng 12. Corolla tube about 3 mm. long, not more than 4 hor glabrous. L. otha. . Corolla tube 4-6 mm. long, usually with at iene a few scattered hairs. ; f apex somewhat acuminate or, at least, acute; eg from base of tube to tip of wit lobes, 9-12 . long. L, uendeniensis. 13, teu apex usually somewhat obtuse, or sometimes iy Nm 80 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 acute; corolla, from base of tube to tip of upper Jobes, 14-18 mm. long. .............. L. X bella. 1. Leaves usually broadest above the middle, tip often + acuminate. 14. Length of leaf not more than twice the breadth. 15. Bractlets and calyx lobes usually hairy on the back (rarely = We WEES TO TIE io ein v5 a ewe neoedd weeKns . xylos 15. Bractlets and calyx lobes ciliate and glandular. L. « Gate 14. mes is leaf usually more ‘a twice the breadth. a tube up to 4 mm 7. coi with numerous i oe hairs (also glands); corolla tube 2-3 mm. long; leaf apex distinctly acuminate. L. ruprechtiana. 17. Bracts without or with few scattered short hairs, only glandu- lar; corolla tube 3-4 mm. long; leaf apex acute or subacumi- 18. Flowers very small, 7-9 mm. long, usually most flowers TE ns Sheds oh aeebic 6460) <_ salicifoha. 18. Flowers larger, 12-16 mm. long, usually ‘igchenne, 7 LRA Ep eNO POO eR Pee tae L. X notha. 16. Corolla tube 4-5 mm. lon 19. Bractlets 3/4 the length of the ovary (not including the calyx or more; corolla 12-14 mm. long. L. X muscaviensis. 19. Bractlets 1/3 to 1/2 the length of the ovary; corolla 9-12 mm long. | ied nuainbeiontie NOTES ON THE SPECIES AND Hyprips Lonicera < amoena Zabel, Mitt. Deutsch. Dendr. Ges. 1901: 96. 1901 (L. korolkowii X tatarica). This is a very graceful shrub up to eight feet or more high and noted for its slender growth, abundance of blossom, and the bluish-green or glaucous color of its foliage. Zabel (l.c.) divided the hybrid into two forms, f. alba and f. rosea, based mainly on the possession of white and rose-colored flowers respectively, but such a division is an over simplification for in- dividual plants differ in the amount of color in their flowers. Another plant, var. arnoldiana Rehd. (Mitt. Deutsch. Dendr. Ges. 1908: 163. 1908), has a particularly delicate growth habit, with narrower leaves and somewhat larger flowers, which start as pale pink and fade until they are almost white. This plant arose from seed collected in 1899 from a plant of Lonicera korolkowii growing in the Arnold Arboretum. Lonicera < bella Zabel, Gartenflora 38: 525. 1889 (L. morrowit X tatarica). Apparently first raised from seed collected from Lonicera morrowii and distributed by the botanic garden at Leningrad (when it was called St. Petersburg). This is one of the hybrids which can readily arise when the two parent species are grown together. Similar in many ways to L. tatarica, the flowers are not so attractive as those of some forms of that species. They are not quite as large and, in fading, turn quite rapidly to a dull 1966] GREEN, LONICERA TATARICA COMPLEX 81 L. tatarica L. X amoena 1G. 3. Sketches of ovaries and leaf-outlines of Lonicera tatarica, L. korol- Rowii, and their hybrid, L. x amoena yellowish color, a character inherited from the other parent, L. morrowii, and one which tends to spoil the effect of the pink or white flowers still in bud or opening. However, Lonicera >< bella shows SaeSeAETY hybrid vigor ra as might be expected, when one of the parents, L. tatarica, has so many color variants, the hybrid is variable too. Gaunt have been patna 82 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 first as varieties by Zabel (Gartenflora 38: 525. 1889) and then as forms by Rehder (Annual Rep. Missouri Bot. Gard. 14: 206. 1903) with pure white flowers (f. candida), white with a tinge of pink in the bud (f. albida), light flesh-colored with a darker middle stripe (f. imcarnata), rose-pink, usually clearly bordered (f. rosea) and dark rose, also distinctly margined (f. atro-rosea). The cultivar ‘Dropmore’ belongs to this hybrid. A very hardy plant, it was raised at the Skinner Nurseries at Dropmore, Manitoba, and on careful examination shows traces of a possible back- cross to L. tatarica. Lonicera korolkowii Stapf, Garden & Forest, 7: 34, fig. 4. 1894. A handsome species, native to Turkestan, which was at first mistakenly named Lonicera floribunda by Zabel (Gartenflora 38: 525. 1889) but despite the fact that the name was particularly appropriate for this plant because of its floriferous habit, it was a misidentification for another, un- related honeysuckle from Persia, named almost thirty years previously. L. korolkowii is an attractive species, notable for its bluish-green foliage and graceful twiggy habit, two characters it has passed on to the hybrid L. X amoena. Three varieties are known in cultivation: var. aurora Koehne (Repert. Sp. Nov. 8: 31. 1910), var. floribunda Nicholson (Ill. Dict. Gard. Suppl. 503. 1901) and var. zabelii (Rehd.) Rehd. (Annual Rep. Missouri Bot. Gard. 14: 131. 1903), varieties differing mainly in their leaf shapes and the intensity of the pink coloration in their flowers. Lonicera minutiflora Zabel, Gartenflora 38: 523. 1889 (L. mor- row * L. * xylosteoides). This shrub is grown for its botanical interest rather than its horticul- tural value, for, as its name suggests, the flowers are smaller than those of the parental species, all of which are superior to this triple hybrid. It ap- parently originated from seed collected on a plant of Lonicera morrowit in the botanic garden at Miinden in Germany. Lonicera morrowii A. Gray in Perry, Narr. Exp. Chin. Jap. 2: 313. 1857. A distinctive species, native of Japan, which may grow to about 7 or 8 feet but does not reach this height as quickly as the other species in the complex. It tends to have a more wide-spreading habit and, in a fully grown bush, presents a more or less rounded outline of dense twigs. Its leaves are distinctive too, being softly pubescent, especially below. The flowers, which have a deeply divided corolla (Fic. 5), are white when they first open but turn yellowish on fading. Like several of these species it is most attractive in fruit, producing large, deep red and juicy berries which are highly attractive to birds, who have been responsible for the fact that Lonicera morrowii is now dispersed and naturalized in numerous places throughout the eastern United States. In addition there is a f. xantho- 1966 | GREEN, LONICERA TATARICA COMPLEX 83 L. X notha L. x bella ruprechtiana L. X salicifolia ; : L. X xylosteoides L.X muendeniensis | : ae Fic. 4. Sketches of the ovaries, bracts, and bractlets of the species and hybrids in the main complex involved with Lonicera tatarica. carpa Nash ex Teuscher (Addisonia 17: 37, t. 563. 1932) with yellow fruits Lonicera muendeniensis Rehd. Gartenflora 42: 102. fig. 18, 4-6. 1893 (L. x bella < ruprechtiana) Mainly of interest as a triple hybrid, this plant arose in the botanic 84 JOURNAL OF THE ARNOLD ARBORETUM [VoL. 47 Fic. 5. Photograph of flowers of Lonicera morrowii from a plant growing in the Arnold Arboretum (accession number 1232.53), X 7. garden at Miinden from seed collected on Lonicera « bella sometime be- fore 1883. It forms a handsome shrub with red, juicy fruits and, like the species from which it arose, also has a yellow-fruited form. Lonicera muscaviensis Rehd. Gartenflora 42: 100, fig. 18, 1-3. 1893 (L. morrowtt * ruprechtiana). This hybrid apparently arose before 1888 from seed sent from the botanic garden at St. Petersburg, collected from a plant of Lonicera mor- row. Its name commemorates the fact that it was raised at the arbore- tum at Muskau, in what was then called Prussian Silesia. Interesting as 4 hybrid, floriferous, and of dense shrubby habit, it is nevertheless in no way superior, from the horticultural point of view, to either of its parental species. Lonicera Xnotha Zabel, Gartenflora 38: 525. 1889 (L. ruprechtiana X tatarica). This hybrid, also tracing its origin to seed distributed from the St. 1966 | GREEN, LONICERA TATARICA COMPLEX 85 Petersburg botanic garden under the name of one of the parent species, and presumably open pollinated, arose in 1878. As might be expected, considering the variability of Lonicera tatarica, one of the parents, individual plants vary considerably in their flower color. In fact, five forms were described by Zabel (Gartenflora 38: 525. 1889): alba with white flowers turning yellowish-white on fading; gilva with yellowish-white flowers bordered with yellow; carneo-rosea with light flesh-colored flowers margined with rose; grandiflora with large white to yellowish-rose flowers margined with light or dark rose; and ochroleuca with yellowish-white flowers fading to dark yellow, and followed by yel- low fruit. In the same year Dippel (Handb. Laubh. 1: 237. 1889) de- scribed three forms under the name L. ruprechtiana which, from their descriptions, clearly apply to this hybrid: f. carnea with light red flowers, f. striata with them striped white and rose, and f. rosea with flowers tinged pale rose. It is tempting to try and equate Dippel’s forms with those of Zabel but considering the variability of the parental species, and there- fore of the hybrids, this would be foolish. It is wiser to consider them as names given to some of the first segregates to be grown, for all intermediate shades of flower color may be expected from further hybridization. As a hybrid, Lonicera notha shows considerable vigor and florif- erousness. It is attractive, especially in the forms with slightly larger flowers, but as the flowers turn yellow on fading it is a matter of taste whether, as an ornamental, it is preferable to the species, especially L. tatarica. Lonicera ruprechtiana Regel, Gartenflora 19: 68, t. 645. 1870. This native of northeastern Asia is the hardiest member of the group. With its upright and then spreading branches, its habit is very similar to that of Lonicera tatarica. It is mainly distinguished by the shape of its leaves (see FIG. 2), which are pubescent below in the typical plant, in the color of the flowers, which are creamy-white turning to yellow, and in the manner in which the narrow petals are curled back at the margins (FIG. As in Lonicera tatarica and L. morrowiti there is a yellow-fruited f. xan- thocarpa (Zabel) Zabel (in Beissner et al., Handb. Laubh.-Benenn. 458. 1903) which, in this case, is also said to have somewhat smaller, more yel- lowish flowers. The species shows considerable variations in the wild and its glabrous extreme with leaves only sparingly hairy on the veins below, has been named var. calvescens Rehd. (Mitt. Deutsch. Dendr. Ges. 1912: 194. 1913), a plant also notable in cultivation for bearing fruits which are a dull purplish-red, instead of the typical, light orange-red. Lonicera x salicifolia Dieck ex Zabel, Gartenflora 38: 524. 1889 (L. ruprechtiana < L. * xylosteoides). This triple hybrid is included here only for completeness. With its very small, almost minute, flowers, a large proportion of which abort and 86 JOURNAL OF THE ARNOLD ARBORETUM [VOL. 47 Fic. 6. Photograph of flowers of Lonicera a Sam from a plant growing in the Arnold Arboretum (accession number 352.62), * 6. never develop, it has very little merit to horticultural recognition. The eaves, too, are narrow and small and, although first raised before 1880, it is grown only as a botanical curiosity Lonicera tatarica L. Sp. Pl. 173. 1753. This is the most important and valuable species in the group, occurring in the wild from the Altai and Ural Mountains in central Asia to the area of the Volga River in southern Russia. It has been cultivated for over 200 years and has escaped and become naturalized in many places throughout the northern and eastern United States. The showiness and beauty of this species depends in part on the breadth of the corolla lobes and the way in which they are borne, not curling back as in related species (FIG. 7), and also because they tend to retain their color as they fade and shrivel, rather than become yellowish. With a wide range in character, it is not surprising that many of the variants in cultivation have been given names. There are some twenty forms listed either in Rehder’s revision of the genus (Annual Rep. Mis- souri Bot. Gard. 14: 27-232. 1903) or his Manual of Cultivated Trees and Shrubs (ed. 2. 1940) which today would be assigned to the rank of cul- 1966 | GREEN, LONICERA TATARICA COMPLEX 87 Fic. 7. Photograph of flowers of Lonicera tatarica from a plant growing in the Arnold Arboretum (accession number 304.5 tivar. Other names have been treated in synonymy by Rehder, although it is quite possible that, in modern terms, each represented a different clone. For the most part, differences rest upon variations in the size and color of the flowers, to a lesser degree on the size and breadth of the leaves, and even, in the case of the dwarf cultivar ‘Nana’, on the habit. There would be little point in listing all the names in this context but a few notes on one or two growing in the Arnold Arboretum may be of value. Lonicera tatarica ‘Arnold Red’ is noteworthy as bearing the darkest red, and therefore showiest, flowers of all the cultivars. Also handsome in fruit, it is one of the most desirable clones in the whole group. Two other cultivars were selected in relatively recent times for their red flowers: ‘Hack’s Red’ and ‘Sheridan Red’ but neither compares with ‘Arnold Red’ in richness and depth of color. ‘Morden Orange’ might at first be thought to have orange flowers but, in fact, it has been selected for the most attractive, light orange fruits. ‘Lutea’ has bright yellow berries (tinged almost orange) and has been in cultivation for well over a century. Throughout the species the berries are 88 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 covered with an attractive, slight glaucous bloom and the clones, grown mainly for their flower colors, show considerable variation in the light- ness or depth of color in their red fruit. There is a dwarf form in cultivation today with pure white flowers and although grown under the name ‘Nana’ it is not this original form. If considered worthy of distinction it needs a new name. The original dwarf var. nana Hort. ex Alphand (Descr. Arb. Arbriss. 42. 1875), which trans- posed into modern terms becomes cv. Nana, was described as bearing rose-colored flowers. Lonicera < xylosteoides Tausch, Flora 21: 736. 1838 (L. tatarica X xylosteum). A botanically interesting hybrid of no particular distinction from the horticultural point of view, except that the blue-green glaucousness of the undersides of its leaves is somewhat unusual. It is, in fact, the reason why this hybrid was also, but later, given the name Lonicera coerulescens Dippel (Handb. Laubh. 1: 233, fig. 149. 1889). Lonicera xylosteum L. Sp. Pl. 174. 1753. This species, of long cultivation, is native from Europe through Turkey and European USSR to the Altai Mountains, and has occasionally escaped from cultivation in eastern North America. It is worth growing for its soft, abundant, light green foliage and attractive deep red fruits, although its yellowish-white flowers are undistinguished. As in any species with such a wide natural range and long history as a cultivated plant, many variants have been described and named, both from the wild and the garden. Perhaps the best known is f. /utea (Veil- lard) Rehd. (Annual Rep. Missouri Bot. Grad. 14: 138. 1903), dis- tinguished by its yellow fruits, but f. glabrescens Zabel (Mitt. Deutsch. species: the former having leaves glabrous or almost so, above, and the latter having them densely pubescent on both surfaces. A form with a dwarf compact habit which has attracted attention of late is Lonicera xylosteum ‘Claveyi’ (also referred to as ‘Clavey’s Dwarf’). Although a different clone, it is not certain just how it differs from f. compacta which Zabel described from a plant received from Dessau in Germany in 1899 (Mitt. Deutsch. Dendr. Ges, 1901: 92. 1901) and which is possibly extinct now. Because of its habit ‘Claveyi’ is a valuable plant in a small garden or for forming low, dense hedges. VoLuME 47 NuMBER 2 JOURNAL OF THE ARNOLD ARBORETUM HARVARD UNIVERSITY B. G. SCHUBERT EDITOR L. I. NEVLING, JR. C. E. WOOD, JR. LAZELLA SCHWARTEN CIRCULATION THE JOURNAL OF THE ARNOLD ARBORETUM Published quarterly by the Arnold Arboretum of Harvard University. Subscription price $10.00 per year. Volumes I-XX, reprinted, and back issues of volumes XXI-XLV are part on the Kraus Reprint Corporation, 16 East 46TH STREET, New York, N.Y. 10017. Subscriptions and remittances should be addressed to Mrs. LazeLia ScHWARTEN, ARNOLD ARBORETUM, 22 Divinity AVENUE, CAMBRIDGE, Mass4- CHUSETTS 02138. CONTENTS OF NO. 2 On THE IpENTITY oF DrosERA BREVIFOLIA. Carroll E. Wood, Jr. 89° NortEes ON THE NOMENCLATURE OF Cornus. I. K. Ferguson ............ 100° Tue CoRNACEAE IN THE SOUTHEASTERN Untrep Srarss. I. K. Ferguson 106: THE NYSSACEAE IN THE SOUTHEASTERN UNITED STATES. Richard H. Eyde 117° THe GENERA OF ARALIACEAE IN THE SOUTHEASTERN UNITED Stratus. Shirley A. Graham 126° Notes on Some Piants or Pusrto Rico. Richard A. Howard .... 137 Stupres in Dioscorna, I: A CotLecTIon rroM British HONDURAS. Bernice G. Schubert . 147 ADDITAMENTA 4D Fioram Jorpanican. Baki Kasapligil ................ 160 Volume 47, No. 1, including pages 1-88, was issued January 14, 1966. Second-class postage paid at Boston, Massachusetts JOURNAL OF THE ARNOLD ARBORETUM VoL. 47 APRIL 1966 Zz 2 ,UMBER 2 ON THE IDENTITY OF DROSERA BREVIFOLIA C. E. Woop, Jr. THE PRINCIPLE OF TYPIFICATION of the names of plants or of groups of plants as now embodied in the International Code of Botanical Nomen- clature offers one of the best pathways to nomenclatural stability, for only by typification can the application of names be fixed precisely. Once all plant names have been typified accurately, a major source of nomen- clatural instability will have been eliminated, since many of the changes which still have to be made are the result of either lack of typification or inaccurate typification. Hasty or careless typification, however, can in- troduce confusion and provide justification for the accusation that taxon- omists are continually changing the names of plants (presumably just for the joy of changing). This paper is an outgrowth of both a long-standing interest in Drosera and work on a generic flora of the southeastern United States.1 It seems to me to have grown to be very long for its limited title. However, a number of well-known names, of which Drosera brevifolia is one, have been upset recently on more or less capricious grounds. I have therefore gone in some detail into the background and typification of this name in the hope of settling its status finally. One can have no objections to changing a name if this is clearly necessary, but it is irresponsible and not in the interests of taxonomy to overthrow current usage of a name when the full evidence is not available. For more than a hundred years, the name Drosera brevifolia Pursh was applied consistently to a distinctive American species characterized by its annual habit; cuneate, short-petiolate, exstipulate leaves; gland- -tipped hairs on scapes, pedicels, and calyces; relatively large corollas (for sect. Drosera) mostly 1-1.6 cm. across; and generally obovoid seeds with 10-12 longitudinal rows of shallow pits (formed by the collapse of the outer wall of individual large cells of the outer seed coat). In the United States it occurs primarily in moist, acid, sandy soils on the Atlantic and Gulf *T am grateful to — directors and curators of the institutions which I have visited and to Caroline K. Allen, Mary T. Dillon, Frances M. Jarrett, Alicia Lourteig, CE. B: Bonner, H. R. De nog Toseph Ewan, and W. R. Ernst for their help in various ways. The work on the generic flora has been supported Sara the generosity of George R. Cooley and grants from the National Science Foundatio 90 JOURNAL OF THE ARNOLD ARBORETUM [ VoL. 47 coastal plains from southeastern Virginia to southern Florida, westward through Alabama, Mississippi, and Louisiana, to Arkansas, southeastern Oklahoma, and eastern Texas. Disjunct stations occur in northern Alabama and Tennessee. In the eastern part of this range the petals are white; in at least Tennessee and Louisiana populations with both white-flowered and pinkish- to rose-purple-flowered plants are known; and in Texas, Arkansas, and Oklahoma white-flowered plants apparently are unknown. Many of the plants from this westernmost part have flowers of reduced size which may open only partly or hardly at all, strongly suggesting self- pollination, or even cleistogamy, linked with the less favorable rainfall conditions of this area. These western populations are thought by some to represent a distinct species, D. annua Reed (Torreya 15: 246. 1915). Drosera brevifolia also occurs disjunctly in southern Brazil, Paraguay, Uruguay, and northeastern Argentina, where both the white and rose- purple forms occur (cf. G. Dawson, Revista Argent. Agron. 5: 231-239. 1938, and notes on herbarium specimens). Originally described as dD. maritima St. Hil. (Hist. Pl. Remarg. Brés. Parag, 264. 1825), these populations have long been recognized as being conspecific with the plant of the southeastern United States. White-flowered plants of this species are known from Oaxaca, Mexico (Galeotti 7246 |c]), and, to judge from the description, D. chiapasensis Matuda (Anal. Inst. Biol. Méx. 27: 362. 1957), described from the state of Chiapas, Mexico, most probably repre- sents additional disjunct populations of the same plant. (Such wide disjunctions also occur in the ranges of D. intermedia Hayne |Europe, from the Caucasus, northward to Finland and Sweden, and westward to Portugal and Ireland; eastern North America; Cuba and Hispaniola; Suriname, British Guiana and Trinidad; and southern Brazil and Para- guay] and in D. capillaris Poir. [see below]. It is also notable that in various parts of the ranges of these species petal-color varies in the former from white to pink or purple, and in the latter from pink to white or vice versa, paralleling the situation in D. brevifolia.) I , in an unnecessarily ill-humored paper on Drosera in the south- eastern United States (Sida 1: 53-59), Lloyd Shinners argued that the plant described by Pursh is not the species which has been known for so long as Drosera brevifolia but is instead D. capillaris Poiret, another well- marked species described from the southeastern United States. Since Shinners recognized both D. maritima (of which he saw only five indi- viduals, one sterile) and D. annua as distinct species, and there is no other name available (unless it be D. chiapasensis, of which he seemingly was unaware), he redescribed the plant of the Atlantic Coastal Plain of the southeastern United States as D. leucantha Shinners. The holotype was given as Cronquist 5255 (smu) (apparently an error for 4255, the num- ber on isotypes in GH and ny), from Glynn County, on the coast of Georgia. Unfortunately, his conclusion was too hastily arrived at and was n inadequate evidence. It is quite unnecessary to abandon the well-established D. brevifolia Pursh. Since Drosera capillaris Poir. is also involved, the account which follows deals first with the typification of that species to eliminate any question 1966] WOOD, IDENTITY OF DROSERA BREVIFOLIA 91 of its identity, and then with the selection of a lectotype for D. brevifolia Pursh. Drosera capillaris Poiret in Lamarck, Encycl. Méth. Bot. 6: 299. 1804. Poiret’s Latin diagnosis reads, “Drosera scapis radicatis, capillaribus, paucifloris, glabris; foliis spathulatis, petiolatis. (N.).” In the discussion he says that this species should be distinguished from Drosera longifolia L., which in his sense included both D. longifolia (D. anglica Huds.) and D. intermedia Hayne (1800). He notes the spatulate leaves, rather long petioles “enlargis insensiblement a leur partie supérieure,” and the small, nearly sessile, very glabrous flowers. He further notes, “Cette plant croit 4 la Caroline; elle m’a été communiquée par M. Bosc, qui en a fait la découverte. (V. s.).”’ Nothing else in the description can be regarded as of diagnostic value. The slender, glabrous scape and small, glabrous flowers are characteristic features of this species as usually understood. Spatulate leaves in combination with glabrous scapes may also suggest D. intermedia, which occurs in the same area of the Coastal Plain of the southeastern United States. The leaf blades of D. capillaris are char- acteristically almost orbicular, tapering abruptly into the petiole, but some individuals, presumably in response to varying ecological conditions, have distinctly spatulate blades gradually tapering into the long petiole. Plants with leaves of the latter kind are sometimes misidentified as D. intermedia, capillaris) provide ready characteristics for distinguishing between the two. Fortunately, the spatulate leaves of the description need introduce no doubts as to the application of the name, for there is a type specimen. In early July, 1964, through the courtesy of Prof. A. Aubréville, director of the Laboratoire de Phanérogamie, Muséum National d’Histoire Nat- urelle, Paris, and with the kind help of Mlle Alicia Lourteig, I was able to locate the holotype of Drosera capillaris, According to the labels on the sheet, the specimen passed in turn from Poiret to Moquin-Tandon, to Cosson, to [Ernest] Durand, and finally to its present location in the general phanerogamic herbarium of the Muséum. It is a single skimpy plant bearing seven leaves and a scape 15 cm. long with four mature or nearly mature fruits, the seeds from one of which have been removed by a previous worker and placed in a small packet on the sheet. The con- spicuous stipules; the long, slender, straight, glabrous scape; the glabrous calyces; and the seeds with longitudinal rows of papilliform cells all testify that this name is being applied correctly. The label reads: Drosera capillaris. (n.) dict an Dr. longifolia? Mich. non Lin. bosc roline a. Poiret {in another hand] 92 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 This species is currently known to occur from southeastern Virginia, southward along the Atlantic Coastal Plain to southern Florida and west- ward along the Gulf Coastal Plain to eastern Texas; in the states of Veracruz and Tabasco, Mexico, and in British Honduras; in eastern and western Cuba and in Jamaica; and in Suriname, British Guiana, and Trinidad. Reaffirming the importance of seed characteristics as the most constant and reliable specific characters in this vegetatively plastic genus, C. H. Brummer-Dinger has shown that the later-described Drosera tenella Willd. is conspecific with D. capillaris (Acta Bot. Neerl. 4: 136-138. 1955). The petal color of D. capillaris is rose-pink to almost white in the United States, and has been recorded as white or pink in British Honduras and in Cuba. Drosera brevifolia Pursh, Fl. Am. Sept.1: 211. 1814. Pursh’s description reads: 3. D. pusilla; scapis radicatis simplicibus, foliis brevibus brevifolia. cuneatis vix petiolatis, petalis ovalibus. In sandy swamps of Georgia. Enslen. June. v. s. in Herb. Enslen. The smallest of all the species known; flowers rose-coloured. All of the Latin diagnosis applies to the plant with which this name has long been associated. It can be argued that it can also apply to Drosera capillaris Poir., and, indeed, a part of it can. However, the short, cuneate, scarcely petiolate leaves described are not at all those of D. capillaris, in which the petiole is quite distinct and usually very much longer than the almost orbicular to obovate-spatulate blade. (Cf. D. capillaris above.) Of all the species of the southeastern United States, the leaves described by Pursh are most nearly those of the plant long known as D. brevifolia. The petal color of this same plant, however, is generally white in popu- lations from the Atlantic Coastal Plain, whereas that of Drosera capillaris is generally pale pink to rose. (A collection of D. capillaris from Louisiana, Shinners 23514, carries the comment “petals nearly white, with pink tinge,” and both pink- and white-flowered plants occur elsewhere.) I have seen pink-tinged petals on the plant in question (D. brevifolia) in North Carolina (Wood 8500, Channell & Rock |[cu]), an occasional collector notes “petals white, turning pink on drying” (cf. Cooley & Eaton 5777 [cH]), and the petals of many herbarium specimens are pink, white with pink tips (cf. Kral 6428 [cu]|), or brown with suggestions of pink. If herbarium specimens were entirely without notes on flower color (and the majority are), one might conclude, depending upon the age of the speci- men and conditions of drying, that the plant is pink, white, or yellow flowered. On these grounds, and especially since Pursh saw only dried specimens, one is justified in questioning whether the pink color noted in Pursh’s description is necessarily associated with D. capillaris, rather than with the cuneate-leaved plant of the Latin description. 1966 | WOOD, IDENTITY OF DROSERA BREVIFOLIA 93 After considering these possibilities in the course of preparing an ac- count of the Sarraceniaceae and Droseraceae for a generic flora of the southeastern United States (Jour. Arnold Arb. 41: 152-163. 1960), I had concluded that the entire description best fitted herbarium specimens of the cuneate-leaved, exstipulate, white-flowered, glandular-scaped plant. In the event that a mixture with D. capillaris was involved, it seemed that greater weight should be given to the Latin diagnosis than to the single comment on flower color, and that this procedure was consistent with the recommendations on typification in the International Code of Botanical Nomenclature. It seems particularly in the spirit of Recommendation 7B, then in Appendix IV (ICBN, ed. 1956, p. 294), but now included in both the “Guide for the determination of types” (ICBN, ed. 1961, p. 65) and the main body of the Code (p. 20): “Whenever the original material is heterogeneous, the lectotype should be so selected as to preserve current usage unless another element agrees better with the original description and (or) figure.” Thus I continued to apply the name Drosera brevifolia in its usual sense and, because of the press of other work on the generic flora, did not attempt to typify it further. Shinners (loc. cit. 55) came to quite a different conclusion: “Without seeing the type (location unknown; possibly not in existence; Diels cites another Enslen specimen bearing no locality beyond Southern U. S.) there may be a little doubt as to what Pursh had, since the description says nothing about pubescence on the scape (though failure to mention it sug- gests there was none) or presence of stipules. Pursh listed only four species, the other three being D. rotundifolia, D, longifolia, and D. fili- formis; he makes no mention of D. capillaris. But his statements ‘smallest of the species known; flowers rose-coloured’ apply exactly to D. capillaris among the Southeastern species. The words cannot possibly refer to the plant with large, white flowers which Chapman and later authors mis- takenly have called D. brevifolia. The phrase ‘foliis brevibus cuneatis vix petiolatis’ of course excludes D. filiformis, and makes both the very long- petioled D. intermedia and the round-leaved D. rotundifolia very unlikely candidates. Enslen collected in Lower Georgia, which I take to mean the Coastal Plain. The only species in this area to which Pursh’s description reasonably applies is D. capillaris.” This conclusion does not seem to be very well founded. “Smallest of the species known” applies to plants, not flowers, as Shinners implies, and fits better the glandular-scaped plant than Drosera capillaris, although either may flower at small size. ‘“Foliis brevibus cuneatis vix petiolatis” does not apply to D. capillaris with its well-developed petioles, but to the species which has been known for so long as D. brevifolia; further- more, the argument based on flower color does not necessarily hold. The description can apply in its entirety to the white-flowered, cuneate-leaved, exstipulate, glandular-scaped plant, but not to D. capillaris or any other species of the southeastern United States. On the basis of the description alone there is little justification for assigning the name D. brevifolia to the synonymy of D. capillaris. 94 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 In the hope of resolving the question, I have sought to locate the Enslen collections on which Pursh based his description and to typify Drosera brevifolia by a specimen. In the preface to his Flora Americae Septentrionalis Pursh discusses some of the sources of his specimens and information. On page xii he eel “While I was thus engaged in describing and figuring those new acquisitions to the American Flora [the specimens from the Lewis and Clark Expedition], another opportunity offered to augment my resources. Mr. Aloysius Enslen, who had been sent to America by Prince Li[e]chten- stein of Austria, as a collector of new and interesting subjects of natural history, returned to Philadelphia from his extensive travels through the Western Territories and Southern States. This gentleman, with whom I had previously been on terms of intimate friendship, was now in posses- sion of an extremely valuable collection of living and dried plants, to which I had unrestrained access. To his liberality I am indebted for many new and scarce specimens, which filled up a desideratum in my collection, particularly in the plants of Lower Louisiana and Georgia. Those species cepa racenin from his collection I have distinguished by ‘v. s. im Herb. Ens Little is ane about Aloysius Enslen and his activities in the United States, but a few comments and clues are scattered through Pursh’s Flora, in the correspondence or publications of Enslen’s contemporaries, and in various other places. Professor Joseph Ewan, of Tulane University, whose scholarly and fascinating studies of the Pennsylvania botanists of this period have cleared up many botanical mysteries, has most generously shared all the bits of information about Enslen he has accumulated in the course of his extensive research. The information in the next three paragraphs is entirely from his notes or published papers; most of the documentation is omitted here, for he will present this material later in another context. The data about Pursh, Barton, Lambert, and their collections are drawn from either Prof. Ewan’s studies (Frederick Pursh, 1774-1820, and his botanical associates. Proc. Am. Philos. Soc. 96: 599-628. 1952) or those of the late F. W. Pennell (The elder Barton — his plant-collection and the mystery of his floras. Bartonia 9: 17-34. 1926; Benjamin Smith Barton as naturalist. Proc. Am. Phil. Soc. 86: 108-122. 1942; Historic botanical collections of the American Philosophical Society and the Academy of Natural Sciences of Philadelphia. /bid. 94: 137-151. 1950). Enslen apparently arrived in America at the expense of Prince Liechten- stein about 1800 (1799-1801). He set up a garden in Philadelphia to grow plants which would either be collected by himself or acquired from others (see especially John Lyon; cf. Ewan & Ewan, Trans. Am. Philos. Soc. 53(2): 10. 1963). These plants would be for reshipment to Vienna, not for public sale. It is known that he was in Savannah in at least 1804 and that he collected between Savannah and Louisville, Georgia. He traveled to New Orleans, evidently following the Coweta Falls trail from Augusta, Georgia, via Mitchell, Macon, and Cowetatown (near Columbus). 1966 | WOOD, IDENTITY OF DROSERA BREVIFOLIA 95 He also met the entomologist and artist John Abbott in Georgia, and Alexander Wilson cites him in his American Ornithology. In 1806, he traveled to St. Louis via Pittsburgh, and he at least intended to travel beyond with the Osage Indians. In 1809, Henry Muhlenberg wrote to Stephen Elliott of his visit to Enslen’s growing garden where he was culti- vating a number of Georgia plants. On December 17 of the following year Muhlenberg wrote to Elliott: “Mr. Enslin [sic] is allmost gone & past recovery of a consumption” (quoted by Ewan & Ewan, loc. cit. 14). “There is evidence that |John] Lyon and Enslen . . . traded nursery stocks during the years of their activities in Philadelphia. Lyon is credited with the introduction of the copper-colored Jris fulva into English gardens in 1812. This plant Enslen had discovered in the marshes near New Orleans. [See Pursh, Fl. Am. Sept. 1: 30, under J. cuprea.] Pursh ac- quired specimens from both Lyon and Enslen; and in at least one instance, Sida hispida, he credited to Lyon a collection almost certainly taken by Enslen” (Ewan & Ewan, loc. cit. 10). There is nothing to show that Enslen ever returned to Vienna after com- ing to America, but a large number of living plants (ca. 1000 species) and herbarium specimens (representing ca. 670 species) reached there. The latter are incorporated in the herbarium of the Naturhistorisches Museum, Vienna. All are without data and bear labels wholly in the handwriting of Leopold Trattinick (1764-1849), who presumably discarded any original labels. Trattinick was curator of the herbarium in Vienna at the time Enslen’s plants were being studied by Pursh. In addition to these specimens, there are 69 Enslen collections, also without data, in the herbarium of Benjamin Smith Barton, which is now at the Academy of Natural Sciences of Philadelphia (Pennell, 1950, p. 141). This herbarium, built up almost entirely between 1797 and 1807, is directly connected with Pursh, who was employed by Barton about 1805 and who made two collecting trips (1806, 1807) for him. Pursh was largely responsible for Barton’s herbarium of about 2000 sheets, and 1149 of the 1674 sheets now in it were collected by him. Little or nothing was added to the herbarium after Pursh left Philadelphia and went to New York in the late autumn or early winter of 1807 (Pennell, 1942). In the winter of 1811-1812, Pursh went to England, where, under the patronage of Aylmer Bourke Lambert (1761-1842), he prepared his flora for publication. The specimens he took with him (including, among others, many of his own collections from the trip of 1807 and some of those of the Lewis and Clark Expedition which had come to Barton for study in November, 1805) went to Lambert before Pursh left England in 1816 for Canada, where he died in 1820. There appear, then, to be three possible sources for Enslen collections which Pursh might have used in drawing up his description of Drosera brevifolia: Lambert’s herbarium; the herbarium of the Naturhistorisches naga, Vienna; and the Benjamin Smith Barton herbarium in Philadel- ‘ ie Lambert’s death in 1842, his large collection was broken up, sold 96 JOURNAL OF THE ARNOLD ARBORETUM [ VoL. 47 at auction, and widely dispersed. Most of the North American materials were bought by Edward Tuckerman (collections of Fraser, Bradbury, Nuttall, Lewis and Clark, etc.) and by Pamphlin (Pursh’s own collections, largely from his trip of 1807). Tuckerman also bought a bundle of “Mis- cellaneous Plants, N. America, &c.” which proved to include only Pursh collections (cf. Ewan, 1952). The specimens bought by Tuckerman are now at the Academy of Natural Sciences of Philadelphia, while those bought by Pamphlin are in the herbarium of the Royal Botanic Gardens, Kew (fide Ewan). Other parts of Lambert’s collections, including the important Ruiz and Pavon specimens which went to the British Museum (Natural History), are preserved in various herbaria, but a large part seems to have disappeared. There are no Enslen specimens among the materials from the Lambert herbarium now in Philadelphia, nor was I able to locate any at either Kew or the British Museum (Natural History) in the summer of 1964. I also checked hopefully, but without success, the collections of the Conservatoire Botanique, Geneva; those of the Botanical Museum and Herbarium, Copenhagen; and, insofar as possible, those of the Muséum d’Histoire Naturelle, Paris. Enslen collections apparently are known only in Vienna and Philadelphia. If, however, an Enslen specimen of Drosera brevifolia from Lambert’s (hence Pursh’s) herbarium should be found, it would take precedence over others in the typification of the name. Professor K. H. Rechinger, director of the Naturhistorisches Museum, Vienna, has kindly sent on loan Enslen’s specimens of Drosera preserved there. There are four large, gray sheets ca. 32 48 cm. which dwarf even more the small plants on them. The labels written by Trattinick bear only a name and “Amer. bor. Enslen.” The first sheet, labeled “Drosera brevifolia,” has six specimens of the cuneate-leaved, glandular-scaped plant which is usually known by this name. It was annotated by Diels as D. brevifolia during the preparation of his treatment of the Droseraceae for Das Pflanzenreich (IV. 122(Heft 26): 1-136. map. 1906) and is presumably the collection cited by him (p. 90) and mentioned by Shinners. The second sheet, with two specimens of this same species, is labeled “Drosera brevifolia a. flor[i]b[us] albis Tratt[inick].” The third, also bearing two specimens of this species, is labeled “Drosera brevifolia ~. florib[us| roseis Tratt.” The fourth, labeled “Drosera cuneifolia glabrata Tratt.,” was annotated by Diels as D. cuneifolia Thunb., certainly a lapse on his part, since he undoubtedly knew better and later cited only South African specimens in Das Pflanzenreich. The two specimens on it are D. capillaris, neither with open flowers. Seven of the ten plants of the other species have open but badly pressed flowers. The petals of all are brown with age. The Barton herbarium in Philadelphia also contains an Enslen collec- tion of Drosera which was located for me by Dr. W. R. Ernst. The single sheet, which I have since seen in Philadelphia, bears two small specimens: to the left, a plant of Drosera capillaris, and to the right a flowering plant of D. brevifolia. On the label is written only “Drosera cuneifolia” (which ———oor 1966 | WOOD, IDENTITY OF DROSERA BREVIFOLIA 97 one suspects may have been the name on the original labels of the Enslen specimens in Vienna) and “Mr Enslin” in two different handwritings which I have not identified. Again, the flower color is impossible to deter- mine, but the size of the flower clearly shows that it was not one of the short-petaled forms from the western part of the range of the species. From these specimens in Vienna and Philadelphia it seems that Enslen’s material was a mixture of Drosera capillaris and the cuneate-leaved plant with glandular scapes, with a great predominance of the latter. To judge from the size of the three specimens of D. capillaris (scapes 15—20 cm. long), it is most likely that Pursh’s “D. pusilla” and “smallest of all the species known” are derived from the other species, for the ten plants in Vienna are only three to five centimeters tall, while that in Philadelphia is even. Trattinick’s sorting of Enslen’s collections shows that he thought some plants to be white, others to be pink flowered. His third category without indication of color suggests that he was uncertain of the flower color of these specimens. Pursh’s ‘flowers rose-coloured” may once again be ex- plained as applying either to plants of D. capillaris, to white-flowered plants of the second species, the petals of which turned pink in drying, or to pink- to rose-purple-flowered plants of the same species from the western part of the range. Since Enslen is known to have collected in southeastern Georgia where both species occur and since Pursh attributed the specimens to “sandy swamps of Georgia,” it would seem that the last possibility is the least likely of the three, especially when the flower size is also considered, Of the two species involved, one, Drosera capillaris, had already been named when Pursh described D. brevifolia. Pursh’s description best fits the new element, and the name has been applied in that sense for more than a hundred years. A specimen of that element, therefore, should be used in typifying the name, in accordance with Recommendation 7B of the International Code. There is no reason for interpreting D. brevifolia as a synonym of D. capillaris, and there is no obstacle to the continued use of the former name. In the absence of a specimen from Pursh’s herbarium, it seems best to select as lectotype the right-hand plant on the sheet in Barton’s her- barium in the Academy of Natural Sciences of Philadelphia, for this is more clearly associated with Pursh than is any part of the material sorted by Trattinick in Vienna, although, as Pursh’s comments suggest, he may well have seen much of that material. It is certain that Pursh saw all of the specimens that went into Barton’s herbarium, for he was largely responsible for the care and development of it, and little or nothing was added to it after he left Philadelphia in 1807. The specimen in the Barton Herbarium is linked with Pursh and will serve adequately as a lectotype. Since it was said to have come from Georgia and since Enslen is known to have collected in southeastern Georgia, this plant should be taken as the white-flowered form which occurs there. Since I have not yet seen material of Drosera chiapasensis, the detailed 98 JOURNAL OF THE ARNOLD ARBORETUM [VOL. 47 taxonomy of D. brevifolia (including D. maritima and D. annua) will be dealt with at another time, but it should be noted again that in what I regard as a single well-marked species Dr. Shinners sees three. To show the basis for his conclusion, his comparisons of D. brevifolia (D. leu- cantha) with D. annua and of D. maritima with D. annua are summarized in tabular form below. He did not give any comparison of D. maritima with D. brevifolia, with which it has long been considered conspecific. D. BrRevIFOLIA (D. leucantha) D. ANNUA Petals 5, pure white, 5-10 mm. long Petals 1-5, light to deep lavender- pink, 2.5-8 mm. long Sepals 3-4.5 mm. long Sepals 2.5-4 mm. long D. MARITIMA (based on 4 plants) D. ANNUA Naked portion of scape 1.5-2.5 (rarely Naked portion of scape 2.5—7 times as —4) times as long as leaves long as leaves Sepals obtuse or subacute Sepals subacute or acute Lowest pedicel 1-3.5 mm. long Lowest pedicel 1-5 mm. long Writing of Drosera annua and D. maritima, Shinners says (loc. cit. 56), “These differences may appear slight. But considering how closely herbarium specimens of D. annua and D. leucantha may resemble each other when well-opened flowers and color data are lacking, while live plants could not possibly be confused, I prefer to treat the North and South American plants as two species. Some rather robust specimens collected by Dr. B. C. Tharp on Padre Island, Texas, greatly resemble the South American species. At the other extreme, collections made by Dr. H. K. Svenson in Coffee and Franklin Counties, Tennessee [where both white- and purple-flowered plants occur in the same populations; cf. Svenson 9974]|, are exceptionally small.” These differences do indeed appear to be slight, and measurements based on larger numbers of specimens make them appear even slighter, par- ticularly when the specific differences between the other American species of Drosera are considered. Petal color (essentially presence or absence of anthocyanin) as a primary specific difference is not very convincing, especially when populations with both white- and purple-flowered forms are known (Tennessee; Louisiana), when both white and rose-purple forms are known in South America, and when similar color shifts occur within such well-marked species as Drosera intermedia and D. capillaris. Neither is it very convincing when, in a genus notable for its morphological plasticity, minor and almost completely overlapping size-differences in calyces, petals, and pedicels are used to distinguish species. Most lamen- table of all is the lack of mention anywhere in his paper of the most dependable specific characters in the genus: the shape and markings of the seeds. This seems inexplicable, since Dr. Shinners makes many refer- ences to Frances Wynne’s paper on the North American species of Drosera (Bull. Torrey Bot. Club 71: 166-174. 1944) in which she stresses the usefulness of seed characters and illustrates the seeds of the North American species. Many botanists, beginning with Hayne (1800) and 1966 | WOOD, IDENTITY OF DROSERA BREVIFOLIA 99 continuing to such recent workers as Wynne, C. H. Brummer-Dinger (Acta Bot. Neerl. 4: 136-138. 1955), and Maguire and Wurdack (Mem. N. Y. Bot. Gard. 9: 331-336. 1957) have realized that seed characteris- tics are among the most useful and stable in this genus. It is unfortunate that seeds were not studied, for those of the three supposed species match one another well. Altogether it would seem that species based on such trivial differences as those used by Dr. Shinners have very little to do with the concept of the species as a natural biological unit. 100 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 NOTES ON THE NOMENCLATURE OF CORNUS? I. K. FERGUSON DURING THE PREPARATION of an account of the Cornaceae for a generic flora of the southeastern United States it was found that the genus C ornus has been delimited in different ways by various workers, some accepting the genus in a broad sense and dividing it into subgenera and sections, others segregating these groups as distinct genera. As a result of these quite varied taxonomic treatments the nomenclature of the sections, sub- genera, and segregate genera has become very confused. The purpose of this paper was to try to find the correct names at different ranks for the various categories that have been recognized and to bring these names together with some of their relevant synonymy. ; The characters used by most workers for subdividing Cornus are mainly those of the inflorescence, and there appears to be little other supporting evidence. These characters do not seem to be sufficient for recognizing segregate genera, and in this work the subdivisions of Cornus are treated as subgenera. No taxonomic revision is proposed here, and the treatment of Wangerin (Pflanzenreich IV. 229(Heft 41). 1910), who recognized seven subgenera within Cornus, is closely followed. However, C. alternt- folia L. and C. controversa Hemsl., with alternate leaves and pitted stones, are separated as an additional subgenus, Mesomora Raf. The genus Cornus as established by Linnaeus comprised five species: C. mas, C. suecica, C. canadensis, C. sanguinea, and C. florida. Hill (Brit. Herbal 331. 1756) established the genus Chamaepericlymenum, and, though he did not describe any species, he refers to only one, Cornus suecica L. Ch pericylmenum must be based on this species. Some recent workers, including Warburg (in Clapham, Tutin & Warburg, Fi, Brit. Isles, ed. 2. 493. 1962) and Pojarkova (in Komarov, Fl. URSS 17: 324. 1951), maintain this genus for the herbaceous species of Cornus. Dumortier (Florula Belgica 83. 1827) recognized two sections in Cornus: Thelycrania, based on C. alba L. and C. sanguinea L., and Corno- typus, based on Cornus “nudiflora = Mas, L.” This latter division may be considered as effectively typifying the genus. In 1830, Lindley (Nat. Syst. Bot., ed. 1. 241) mentioned Benthamia incidentally and without a description as a genus in the Boraginaceae. Later (Bot. Reg. tab. 1579. 1833) he formally established Benthamia, * Continuing a series of miscellaneous notes and papers on the flora of the scuth- eastern United States made possible through the support of George R. Cooley and a r., Dr. George K. Brizicky, and Dr. Rolla M. Tryon for aid and valuable advice; and to Dr. P. W. Ball, of Flora Europaea, for drawing my attention to the place of publication of Swida Opiz. 1966 | FERGUSON, NOMENCLATURE OF CORNUS 101 based on Cornus capitata Wall., as a genus in the Cornaceae. Unfortunate- ly, this name had been validly published in 1828 by Richard as the name for a genus of orchids. Benthamia Lindl. is thus a later homonym. Later, Rafinesque (Alsograph. Am. 58. 1838) recognized a number of categories within Cornus. The status of some of these has been discussed by various workers (O. A. Farwell, Rhodora 34: 29, 30. 1932, and Torreya 42: 130. 1942; H. W. Rickett, Torreya 42: 11-14, 131. 1942), but without agreement. Rafinesque, on page 58 of his Alsographia, recognizes a genus Cornus with a typical element, Cornus, and two subgenera, Mesomora 2 and Kraniopsis. Under the typical element he includes “most of the American species,”’ as well as C. sanguinea and C. alba. Mesomora includes only C. alternifolia, and in Kraniopsis Rafinesque places C. paniculata L’Héer. (= C. racemosa Lam.) and C. comosa Raf. On the following page he lists three ped her en without clearly designating their rank: Eukrania, with C. S prised of C. cies an Benthamia Lindl. The names Mesomora and Kraniopsis must clearly be adopted for subgenera, while it appears that Eukrania, Cynoxylon, and Benthamia may be considered to be genera or subgenera The evidence available suggests that Eukrania should be regarded as a refers to “Cornus (Eukrania) cyananthus Raf.,” commenting that it is “a very curious herbaceous Cornus very near to e . canadensis.” Since he included Cornus mas, the type of Dumortier’s section Cornotypus, in Eukrania, this name must be treated as a synonym of the typical subgenus. On the other hand, Benthamia Lindl. would appear to have been in- tended to be a distinct genus, for Rafinesque cites his own earlier work, Sylva Telluriana (1836), under Benthamia, and in this work he clearly refers to Benthamia Lindl. as a genus and also to his ‘“‘subgen. Cynoxylon of Med. Fl. of 1828.” This latter statement illustrates the uncertainty surrounding the rank of Cynoxylon, since Rafinesque had clearly applied this name to a section of Cornus in the Medical Flora (1828). However, there does not seem to be any evidence to indicate that he intended Cynoxy- fon to be a distinct genus, and for this reason it is accepted as a subgeneric name here. Merrill (Index Rafinesquianus 184. 1949) has indicated that Rafinesque recognized Cynoxylon as a subgenus of Cornus and Eukrania as a separate genus, but the reasons for this treatment are quite obscure. * Rafinesque (Alsograph. Am. 58. 1838) in his formal description of subgenera uses the spelling “Mesomera,” but paw ie A his discussion on page 62, he refers to “Mesomora” and “the Mesomoras.” Rudbeck (Acta Lit. Suec. 1: 98. 1720) proposed the e Mesomora for the plant oe was later known as Cornus suecica L Oo and he uses the epithet in Philosophia Botanica (201, 202. 1763) for an informal category in Cornus. It seems that Rafinesque’s spelling “Mesomera” should be con- sidered a typographical error. 102 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 Opiz (in Berchtold & Opiz, Oekon.-tech. Fl. Béhmens 2: 174-180. 1838) segregated Cornus alba and C. sanguinea as a separate genus Swida.* Some later authors (e.g., Pojarkova, Not. Syst. Leningrad 12: 170. 1950), overlooking this valid publication, have considered Swida Opiz as a nomen nudum, citing the place of publication as Seznam Rostlin Kveteny Ceské (1852), where the name is mentioned without description. The next author to add to the confusion was Endlicher (Gen. Pl. 798. March, 1839), who recognized three categories of undesignated rank in Cornus: a. Arctocrania, b. Tanycrania, and c. Thelycrania. Since End- licher makes no statement either in the preface or elsewhere as to what he intended these subdivisions to be, they cannot be taken into consideration. Later in the same year, Spach (Hist. Nat. Vég. 8. October, 1939) described three sections in Cornus (Macrocarpium, Microcarpium, and Cornion) and segregated Cornus florida as the genus Benthamidia. Cornion Spach is the earliest sectional name for the herbaceous species of Cornus. Two years later, Reichenbach (Repert. Herb. 143. 1841) validated Endlicher’s Tanycrania and Arctocrania as subgenera, while still later, Ledebour (FI. Rossica 2: 377, 378. 1844) validated Arctocrania and Tanycrania as sec- tions. Thelycrania, originally published by Dumortier as a section (1827), was raised to generic rank by Fourreau (Ann. Soc. Linn. Lyon IT. 16: 394. 1868), who cited only one species, T. sanguinea. He thus selected a lecto- type for both the section and the genus. As a genus, however, Thelycrania is antedated by Swida Opiz. Schneider (Illus. Handb. Laubholz. 2: 437. 1909) attributed subg. Thelycrania to Endlicher and subg. Benthamidia to K. Koch, but neither author treated these as formal categories. Instead, Schneider, himself, appears to be the first author to validate the names at the rank of subgenus. Harms (Nat. Pflanzenfam. III. 8: 266. 1898) described the African Cornus Volkensii and placed it in a separate section Afrocrania. At the same time, he also segregated the Mexican C. disciflora and C. grandis as sect. Discocrania. Wangerin (Pflanzenreich IV. 229(Heft 41). 1910) later raised both to subgeneric sie and Hutchinson (Ann. Bot. II. 6: 89. 1942) treated Afrocrania as a gen In 1906, Rydberg (Bull. ie Bot. Club 33: 147) proposed the ge- neric name Cornella for the herbaceous species of Cornus, but this is a superfluous name for Chamaepericlymenum Hill. Nakai later (Bot. Mag. Tokyo 23: 39. 1909) raised Arctocrania Endl. to generic rank, providing * Various spellings have been used for the generic name Swida. The author of the that language. Following the serch Code, the original spelling, Swjda, is the form which must be retained. However, Art. 73. Note 6 of the Code states, “The letters ‘j? and ‘v’ must be changed rie @ and ‘u’ respectively when they represent vowels.” It is proposed that Swjda be changed to Swida which makes pronunciation ier. 1966 | FERGUSON, NOMENCLATURE OF CORNUS 103 yet another synonym for the same group of herbaceous species. He also raised sect. Macrocarpium Spach to generic rank, restricting Cornus to the species of Dumortier’s section Thelycrania, in direct contradiction of Dumortier’s typification of Cornus. Adding still further to the already formidable nomenclatural problems, Nieuwland & Lunell (Am. Midl. Nat. 4: 487. 1916) adopted two new names for segregate genera of Cornus, attributing both to pre-Linnaean authors without giving descriptions for either genus, a practice neither outlawed nor recommended by the International Code. These authors recognize the genus Mesomora [Rudbeck] Nieuwland & Lunell, which they apply to Cornus canadensis L., as still another synonym for Cham- aepericlymenum Hill. Under Ossea Lonicer, both a pre-Linnaean and a post-Linnaean reference are given. The post-Linnaean reference (Adams Lonicers Vollstandiges Krauter-Buch, B. Ehrhart, ed., with indices by P. Uffenbach. 1783) is an unaltered reprinting of earlier editions of Lonicer, but there is also an Anhang by Ehrhart, published in the same year, in which he comments (p. 23) under Ossea, ““Dieses Gewiichses rechter latein- ischer Name is nur gar zu wohl bekannt, nemlich Cornus foemina, oder Virga sanguinea.” Thus it seems that Ehrhart cannot be considered to have accepted Ossea. Since the International Code states “a name is not validly published when it is not accepted by the author who published it,” Lonicer’s work (1783) cannot be taken as a place of valid publication of the genus Ossea. It must be then attributed to Nieuwland and Lunell (1916). However, Lonicer’s description and illustrations are inadequate and do not seem to refer clearly to any species of Cornus, no references are cited, and, as a result, the validity of the name seems doubtful. In the sense of Nieuwland & Lunell, Ossea is a synonym of the genus Swida Opiz. Other historical items and nomenclatural combinations will be evident from the summary of the nomenclature of the eight subgenera given below. In order to fix the application of the names involved, the type or lectotype species is given for each name. When lectotype species are cited without an author these have been selected here. Cornus Linnaeus, Sp. Pl. 1: 117. 1753; Gen. Pl. ed. 5. 1754. (Lectotype species: C. mas L.; see Dumortier, Florula Belgica 83. 1827, and P. A. Rydberg, Bull. Torrey Bot. Club 33: 147. 1906.) Cornus subg. Kraniopsis Rafinesque, Alsograph. Am. 58. 1838. (Lecto- type species: C. paniculata L’Héritier = C. racemosa Lamarck.) Cornus subg. Cornus Rafinesque, ibid. (Lectotype species: C. sanguinea L.) Cornus sect. Thelycrania Dumortier, Florula Belgica 83. 1827. (Lectotype species: C. sanguinea L.; see Fourreau, Ann. Soc. Linn. Lyon II. 16: 394. 1868.) Cornus c. Thelycrania Endlicher, Gen. Pl. 798. March, 183 Thelycrania (Dumortier) peazest Ann. tet ne ok TI. 16: 394. 1868. (Lectotype species: T. sanguinea (L.) F Cornus subg. Thelycrania (Dumortier) Endlicher ex Schneider, Illus. Handb. Laubholz. 2: 437. 1909. 104 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 Cornus sect. Microcarpium Spach, Hist. Nat. Vég. 8: 92. 1839. (Lectotype species: C. sanguinea L. Cornus subsect. Amblycaryum Koehne, Gartenflora 45: 286. 1896. (Lectotype species: C. corynostylis Koehne. Cornus sect. Amblycaryum (Koehne) Nakai, Bot. Mag. Tokyo 23: 37. 1909. Swida Opiz in Berchtold & Opiz, Oekon.-tech. FI. Bohmens 2: 174-180. 1838, “Swjda.” (Lectotype species: S. sanguinea (L.) Opiz.) Ossea [Lonicer | Nieuwland & Lunell in Lunell, Am Mid. Nat. 4: 487. 19 Type species: O. instolonea (A. Nelson) Nieuwland & Lunell in Canell Am. Midl. Nat. 4: 487. 1916.) Cornus subg. Mesomora Rafinesque, Alsograph. Am. 58. 1838. (Type species: C. alternifolia L.) Cornus sect. Mesomora (Rafinesque) Nakai, Fl. Sylv. Koreana 16: 81. 1927. Cornus sect. Microcarpium subsect. Bothrocaryum Koehne, Gartenflora 45: C. controversa Hemsley; see Pojarkova, Not. Syst. Leningrad 12: 169. 1950.) Cornus sect. Bothrocaryum (Koehne) Harms ex Nakai, Bot. Mag. Tokyo 23: 36. 1909, ‘‘Brothrocaryum. Bothrocaryum (Koehne) Pojarkova, Not. Syst. Leningrad 12: 169. 1950. Type species: B. controversum (Hemsley) Pojarkova.) Cornus subg. Afrocrania (Harms) Wangerin, Pflanzenreich IV. 229 (Heft 41): 76. 1910. Cornus sect. Afrocrania maine Nat. Pflanzenfam. III. 8: 266. 1898. (Type species: C. Volkensii Harm Afrocrania (Harms) vedic Ann. Bot. II. 6: 89. 1942. (Type species: A. Volkensii (Harms) Hutchinson.) Cornus subg. Cornus. Cornus sect. Cornotypus Dumortier, Florula Belgica sus 1827. (Type species: C. nudiflora Dumortier, nomen superfluum = C. m Cornus subg. Eukrania Rafinesque, Alsograph. pe 59, 1838. (Lectotype species: C. mas L. Cornus b. T. Sasa Endlicher, Gen. Pl. 798. March, 1 Cornus subg. Tanycrania Endlicher ex Reichenbach, ial Herb. 143. 1841. (Type species: C. mas L.; see Endlicher.) Cornus a 2 anycrania (Endlicher ex Reichenbach) Ledebour, Fl. Rossica 23378. Cornus oe “Macrocorpim Spach, Hist. Nat. Vég. 8: 101. October, 1839. (Type species: C. mascula L. = C. mas L. Cornus hig Ma divbeareean (Spach) Casa Illus. Handb. Laubholz. 2 450. 9. Macrocarinm (Spach) Nakai, Bot. Mag. Tokyo 23: 38. 1909. (Type species: . mas (L.) Nakai.) Cornus subg. Arctocrania Endlicher ex Reichenbach, Repert. Herb. 143. 1841. (Lectotype species: C. suecica L.) Cornus a. Arctocrania Endlicher, Gen. Pl. 798. March, 1839. 1966] FERGUSON, NOMENCLATURE OF CORNUS 105 Cornus sect. g Pousonbinan enue ex Reichenbach) Ledebour, Fl. Rossica 2: 377. 1844, “Orctocr. Arctocrania (Endlicher ex Reichenbach) prem Bot. Mag. Tokyo 23: 39. 1909. (Lectotype species: A. swecica (L.) Nakai.) gestae oh adawnsc Hill, Brit. Herb. 331. gs fe species: Cornus suecica L. = = Chamaepericlymenum suecicum (L.) Gra Cornus nect. Cornion Spach, Hist. Nat. Vég. 8: 103. Shee 1839. (Lecto- type species: C. suecica L. Cornella Rydberg, Bull. Torrey Bot. Club 33: 147. 1906. (Lectotype species: Cornella suecica (L.) Rydberg.) Mesomora [Rudbeck] Nieuwland & Lunell in Lunell, Am. Midl. Nat. 4: 487. (Type species: M. canadensis (L.) Nieuwland ex Lunell.) Cornus subg. Discocrania (Harms) Wangerin, Pflanzenreich IV. 229 (Heft 41): 84. 1910. Cornus sect. Discocrania Harms, Nat. Pflanzenfam. III. 8: 266. 1898. (Lectotype species: C. disciflora Mocifio & Sessé ex De Candolle.) Cornus subg. Cynoxylon (Rafinesque) Rafinesque, Alsograph. Am. 59. 1838. Cornus sect. Cynoxylon Rafinesque, Med. Fl. 132. 1828. (Type species: C. florida L. Cynoxylon (Rafinesque) oa Fl. SE. U. S. 854. 1903. (Type species: Cynoxylon floridum (L.) Small. Cornus b. Tanycrania Endlicher, Gen. Pl. 798. Cornus subg. Tanycrania Endlicher ex pecilnack, Repert. Herb. 143. 1841. (Type species: C. florida L.; see Endlicher Benthamidia Spach, Hist. Nat. Vég. 8: 92. “1839. (Type species: B. florida (L.) Cornus ae Benthamidia (Spach) K. Koch, Dendrologie 1: 694. 1869. Cornus subg. Benthamidia (Spach) K. Koch ex Schneider, Illus. Handb. Laub- holz. 2: 453. 1909. (Type species: C. florida L.) Cornus subg. Benthamia (C. B. Clarke) Schneider, Illus. Handb. Laub- holz. 2: 454. 1909. Cornus sect. Benthamia C. B. Clarke in Hooker f. Fl. Brit. India 2: 745. 1879. (Lectotype species: C. capitata Wallich in Roxburgh.) Benthamia Lindley, Bot. Reg. 19: tab. 1579. 1833, not Benthamia Richard, Mém. Soc. Hist. Nat. Paris 4: 37. 1828. (Type species: B. fragifera Lindley = Cornus capitata Wallich in ae Dendrobenthamia Hutchinson, Ann. Bot. II. 6: 92. 1942. (Type species: D. capitata (Wallich) Hutchinson.) ARNOLD ARBORETUM Present address: FiorA EvROpAEA RESEARCH PROJECT THE UNIVERSITY LIVERPOOL, ENGLAND HARVARD UNIVERSITY 106 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 THE CORNACEAE IN THE SOUTHEASTERN UNITED STATES ? I. K. FERGUSON CORNACEAE Dumortier, Anal. Fam. Pl. 33, 34. 1829, “Corneae,’’ nom. cons. (Docwoop FaMILy) Trees or shrubs [rarely perennial herbs]. Flowers small, bisexual | uni- sexual in dioecious species], regular. Floral tube adnate to the ovary; calyx lobes small to obsolete. Petals 4 [5], [rarely absent], inserted on an epigynous disc. Stamens equal in number to the petals and alternate with them. Ovary inferior, 1—4-locular, with a single pendulous, anatropous ovule in each locule. Fruit a drupe [rarely a berry]. Embryo straight; endosperm copious. (Excluding Alangiaceae DC., Garryaceae Lindl., and Nyssaceae Dumort.) TyPE GENUS: Cornus L. A rather heterogeneous family of seven to sixteen genera, depending upon the generic and familial concepts adopted, with about 100 species, mainly in north-temperate regions, a few in the Tropics, in Africa, Mada- gascar, South America, and also in New Zealand; only Cornus occurs in eastern North America. The family has been placed in the Umbellales on the basis of its few- carpellate inferior ovary, reduced calyx, and tendency toward a many- flowered umbellate inflorescence. The anatomy of Cornaceae differs in many respects from that of the Umbelliferae and Araliaceae, but the secre- tory ducts and extrorse micropyle of Mastixia Blume are thought by some authors to link Cornaceae with Araliaceae. Biochemical studies have shown aucubin to be present in Cornaceae but absent from Umbelliferae and Araliaceae. Garryaceae, Alangiaceae, and Nyssaceae are very closely re- * Prepared for a generic flora of the southeastern United States, a joint project of the Arnold Arboretum and the Gray Herbarium of Harvard University made possible through the support of George R. Cooley and the National Science Foundation and under the direction of Carroll E. Wood, Jr., and Reed C. Rollins. The treatment follows the pattern established in the first paper in the series (Jour. Arnold Arb. 39 296-346. 1958) and continued to the present paper. The area covered includes North and South Carolina, Georgia, Florida, Tennessee, Alabama, Mississi sippi, Arkansas, and Louisiana. The descriptions apply primarily to the plants of this area, with supple- mentary information in brackets. References marked with an asterisk have not been seen by the author I am grateful to Dr. Wood for his generous suggestions and valuable criticisms and to Dr. George K. Brizicky for his patient guidance and discussion. I am also grateful to Dr. R. H. Eyde for his comments on the manuscript. The typescript was prepared by Mrs. Gordon W. Dillon, and the plate is the work of Arnold D. Clapman 1966 | FERGUSON, CORNACEAE 107 lated families more or less recently segregated from the Cornaceae. Al- though studies by Moseley and Beeks and by Eyde (1964) confirm the affinity of Garryaceae with Cornaceae, the highly specialized amentiferous structure of its inflorescence appears to justify treating the group as a separate family. Nyssaceae and Cornaceae have been grouped together as the Cornales, but on the basis of their valvate petals Hutchinson places Cornaceae and Araliaceae (one tribe of which has imbricate petals) in an order Araliales next to the Cunoniales and distinct from the Umbelliferae, which have imbricate petals. Serological studies on Nyssa L., Davidia Baill., and Cornus may be interpreted to support the separation of Nyssa- ceae from Cornaceae. In anatomy and morphology the Cornaceae re- semble the Caprifoliaceae in many ways, and the pollen, in particular, is similar to that of Viburnum and Sambucus. Metcalfe and Chalk draw attention to a similarity between Cornaceae and Hydrangeaceae in wood anatomy. thin the family, Harms separated the subfamilies Curtisioideae and divided the Cornoideae into four tribes: Toricellieae, Helwingieae, Corneae, and Griselineae. The Toricellieae and Helwingieae are considered dis- tinct families by some workers, and studies of pollen morphology (Chao) and wood anatomy (Li & Chao) support this treatment. On the basis of wood anatomy, Adams has proposed the removal of Aucuba Thunb. and Kaliphora Hook. f. to separate tribes. The position of Corokia A, Cun- ningh. is doubtful (cf. Eyde). Engler places it near Argophyllum J. R. & G. Forst. in the Saxifragaceae subfam. Escallonioideae. Eyde, on the basis of floral anatomy, rejects the assignment of Corokia to Cornaceae, but questions certain points in Engler’s alternative taxonomic treatment. Pollination is probably by insects; Griselinia Gmel. is believed to be wind pollinated, but at least one species produces nectar (Percival, New Phytol. 60: 235. 1961). Some species of Helwingia Willd., Corokia, Griselinia, Aucuba, and many species of Cornus are cultivated for their ornamental value; and some also possess medicinal properties. Curtisia dentata (Burm. f.) C. A. Sm. (C. faginea Ait.) is a useful African hard- wood tree. Chromosome numbers of 2” = 16, 18, 20, 22, 27, 32, 36, 44, ca. 72, 120, and 144 have been reported. REFERENCES: Apams, J. E. Studies in the comparative anatomy of the Cornaceae. Jour. Eli- sha Mitchell Sci. Soc. 65: 218-244. 1949. BarLton, H. Sur les affinités des Helwingia. Bull. Soc. Linn. Paris 1: 137-139. 1877. [See also J. Decatsne, Ann. Sci. Nat. Bot. II. 6: 65-76. pls. 6, 7. 1836.] Cornacées. Hist. Pl. 7: 66-83. 1879. [Includes Garrya. ] A BENTHAM, G., & J. D. Hooxer. Cornaceae. Gen. oe 1: 947-952. 1867. [In- cludes Nyssa, Garrya Dougl. ex Lindl., and Alangium. | Brown, F. B. H. Cornaceae and allies in the micsauent and neighboring islands. 108 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 Bishop Mus. Bull. 52: 1-22. 1928. [A rather speculative account of the relationships between Corokia, Lautea F. Br., and Cuao, C. Y. Comparative pollen morphology of the Cornaceae and allies. Taiwania 5: 93-106. 1954. [Alangiaceae, Garryaceae, Helwingiaceae, Nys- saceae, Toricelliaceae. | Coutter, J. M. Geographical distribution of North American Cornaceae. Proc. AAAS 39: 319-322. 1890 . H. Evans. A revision of North American Cornaceae. Bot. Gaz. 15: 30-38, 86-97. 1890. [See also Bot. Jahrb. 15(Lit.): 28. 1893. DaLirmorE, W. Useful woods of Cornaceae. Bull. Misc. Inf. Kew 1916: 96-99. 1916. sapere A. Saxifragaceae. Nat. Pflanzenfam. ed. 2. 18a: 74-226. 1930. [Coro- ia placed in subfam. Escallonioideae, 215. a R. H. Inferior ovary and generic affinities of Garrya. Am. Jour. Bot. 51: 1083-1092. 1964. [See also F, A. Hattocx, Ann. Bot. 44: 771-812. pl. 36. . Floral anatomy and systematic position of Corokia. (Abstr.) Ibid. 52: sas aaa D. E., & M. A. Jounson. Comparative serological studies within milies Cornaceae (dogwood) and Nyssaceae (sour gum). Pp. 305-318 in C. A. Leone, ed., Taxonomic biochemistry and serology. New York. 1964. [See also Am. Jour. Bot. 48: 534. 1961.| Faure, A. L. Etude organographique, anatomique et pharmacologique de la famille des Cornacées. Thése, Univ. Lille. 221 pp. 1924 GopinaTH, D. M. A contribution to the embryology of Alangidm Lamarckit Thw. with a discussion of the systematic position of the a Proc. Indian Acad. Sci. B. 22: 225-231. 1945. [See also Mirra, J. N., .M asi Bull. Bot. Soc. Bengal 3: 11-15. 1949,* A. chinense ee Harm arm HALLIER, H. Ueber Gaertner’sche Gattungen und Arten unsicherer Stellung, einige Rubiaceen, Sapotaceen, Cornaceen und iiber versunkene Querver- bindungen der Tropenlander. Rec. Trav. Bot. Néerl. 15: 27-122. 1918. Harms, H. Die Gattungen der Cornaceen. Ber. Deutsch. Bot. Ges. 15: 21-29. 1897. ——. Cornaceae. Nat. ine III. 8: 250-270. 1898. [Includes Alangium, creer Garrya, and N HILLEBRAND, G. D. E. biter: Phytoserological correspondence among fonts genera of the Cornales, Garryales, Rosales, Rubiales, and Umbellales as an indication of the taxonomic position of the genus Vibur- num. (Abstr.) Am. Jour. Bot. 52: 648. 1965 Hoar, C. S. A comparison of the stem anatomy of the cohort Umbelliflorae. peti Bot. 29: 55-63. pls. 4, 5. 1915. [Cornaceae, Araliaceae, and Umbellif- | Advanc. Sci. 1911: 585. ———. A contribution to the study of the evolution of the flower, with special reference to the Hamamelidaceae, Caprifoliaceae, and Cornaceae. Trans. Linn. Soc. II. Bot. 8: 239-309. pls. 28-30. 1914. [Cornaceae, 261-296. | Hu, H. H. Notulae systematicae ad florem sinensium V. Bull. Fan Memorial Inst. Biol. 5(Bot.): 305-318. 1934. [Toricelliaceae, fam. nov., 311-313; closely related to Araliaceae and Cornaceae. ] rae. fae A. S. The polyphyletic origin of the Cornaceae. Rep. Brit. Assoc. 1912. 1966 | FERGUSON, CORNACEAE 109 Hurcuinson, J. porn generic characters in the family Cornaceae. Ann. Bot. II. 6: 83-93. 194 KIRCHHEIMER, F. Uinbelliflare: Cornaceae. Jn: W. Jonomans, ed., Fossil. tal. mg 23: I-xxu, 1-188. 1938. [Fossil records for the family : includes Hoss ae. | Paes XK Zur Kenntnis des unilokularen Cornaceen- ee (Corna- ceen-Studien I). Ber. Deutsch. Bot. Ges. 76: 33-39. 1963 Li, H. L., & C. Y. Cao. Comparative anatomy of the woods of the Cornaceae and allies, Quart. Jour. Taiwan Mus. 7: 119-136. pls. 1-8. 1954. [Alangia- ceae, Garryaceae, Helwingiaceae, Nyssaceae, Toricelliaceae. | Maekawa, F. Aucuba and its allies; the phylogenetic consideration on the Cornacese, (In Japanese, Jour. Jap. Bot. 40: AA, 1965.* Miki, S. Endocarp remains of Alangiaceae, Cornaceae and Nyssaceae in Japa Jour. Inst. Polytech. Osaka Univ. D. 7: 275-297. 1956. [Stone nea of fossil and living representatives. | Mosetey, M. F., & R. M. Beexs. Studies of the Garryaceae —I. The compara- tive morphology and phylogeny. Phytomorphology 5: 314-346. 1955. [Re- lationships with Cornaceae and other families discussed. | Nakat, T. Cornaceae in Japan. Bot. Mag. Tokyo 23: 35-45. 1909, [Discussion of subgenera and sections of several genera. | Pam, B., & A. A. L. Rutcers. The embryology of Aucuba japonica. Rec. Trav. Bot. Néerl. 14: 119-126. 1917 Paris, R. The distribution of plant elycosides. Pp. 337-358 in T. Swain, ed., Chemical plant taxonomy. London & New York. 1963. [Distribution of aucubin, 35 Bi H. W. Cornales, N. Am. FI. 28B: 297-316. 1945. [Cornaceae, Nyssa- hicaatis A. Beitrage zur Kenntnis der Anatomie der Cornaceae. Bull. Herb. Boiss. 1: 469-484, 496-512, 551-570, 614-639. 1893. TAKHTAJAN, A. Die Evolution der Angi iospermen. viii + 344 pp. Jena. 1959. TauBERT, P. Revision der Gattung Griselinia. Bot. Jahrb. 16: 386-392. 1892. WANGERIN, W. Die Umgrenzung und Gliederung der Familie der Cornaceae. Bot. Jahrb. 38(Beibl. 86): 1-88. 1906. —. Cornaceae. Pflanzenreich 4: 1-110. 1910. 1. Cornus Linnaeus, Sp. Pl. 1: 117. 1753; Gen. Pl. ed. 5. 54. 1754. Deciduous trees or shrubs [rarely perennial herbs]. Leaves opposite, rarely alternate, simple, usually petiolate, exstipulate; blade entire, or obscurely serrate, pinnately veined, often pubescent, the lower surface occasionally covered with a fine network of short hairs or papillae and sometimes with appressed, erect, or curling, simple or characteristic bifid hairs. Inflorescences terminal or axillary, dichotomously branched, ebrac- teate cymes or panicles, [or bracteate umbelliform cymes,] or heads (re- duced cymes) surrounded by petal-like involucral bracts and an inner ring of small, obtuse, membranaceous bracteoles. Flowers bisexual [uni- sexual in dioecious species], epigynous, sessile or short pedicellate. Floral tube turbinate, urceolate or campanulate, adnate to the ovary; calyx lobes 4, distinct or obsolete. Petals 4, valvate in bud, inserted on the margin of a fleshy nectariferous disc, oblong or ovate, white, purple, or yellow. 110 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 Stamens 4, exserted; filaments filiform or subulate; anthers oblong, 2- locular, dorsifixed, versatile, introrse, longitudinally dehiscent; pollen small or medium, prola te, tricolporate, reticulate or granular [occasionally united in tetrads]. Gynoecium (1)2(3)-carpellate, syncarpous gle, filiform or columnar, sometimes dilated below the stigma; stigma capitate or truncate; ovary inferior, adnate to the floral tube and to the disc, (1)2(3)-locular, 1 locule sometimes aborting; ovules solitary in each locule, pendulous, anatropous, with abaxial raphe, micropyle superior. Fruit a globose, ellipsoid, or ovoid drupe; stone 2(rarely 1)-seeded, fur- rowed, rarely with a distinct pit at the apex. Seeds oblong, compressed; endosperm copious; embryo as long as the endosperm, straight, the cotyle- dons foliaceous, radicle terete and near the micropyle. Embryo sac de- velopment of the Fritillaria type in some species, reported to be of the normal (Polygonum) type in others. (Including Afrocrania (Harms) Hutchins., Benthamidia Spach, Bothrocaryum (Koehne) Pojark., Chamae- periclymenum Hill, Dendrobenthamia Hutchins., Swida Opiz.*) Lrcrto- TYPE SPECIES: C. mas L.; see B. C. Dumortier, Florula Belgica 83. 1827, and P. A. Rydberg, Bull. Torrey Bot. Club 33: 147. 1906.* (Name from Latin, cornu, a horn, presumably because of the hardness of the wood.) — OGWOOD. A genus of about 60 species, mainly of temperate regions of the Northern Hemisphere, but extending to Central and South America, with one species in Africa; some 17 species in North America and about six in our area. The treatment of the genus has varied widely, some authors recognizing a number of segregate genera, others treating these divisions as either sub- genera or sections, Eight subgenera compose the genus as delimited here (cf. Ferguson, Jour. Arnold Arb. 47: 100-105. 1966); four are repre- sented in the United States, and three in our area. Only the correct sec- tional or generic names are given here as synonyms of the subgenera. All but two of the species in the southeastern United States are members of subg. Kraniopsis Raf. (sect. Thelycrania Dumort.; Swida Opiz), characterized by ebracteate cymose or paniculate inflorescences appearing * Although usually given as Svida and originally published as Swjda, the crthography used here seems to be most in accord with the International Code (Art. 73, Note 6), 38. 1909). He incorrectly cited Britton asi Brown (Illus. Fl. No. U. S. 2: 660. 1913) as the earliest authors to select C. mas as lectotype species. Rehder (Bibliogr. Cult. Trees Shrubs 496. 1949) also adopted C. sanguinea as l\ectotype. However, Furthermore, C. sanguinea is a component of both Swida Opiz (Berchtold & Opiz, Oekon.-tech. Fl. BOhmens 2: 174. 1838) and Thelycrania (Dumort.) Fourreau (Ann. Soc. Linn. Lyon II. 16: 394. 1868). 1966] FERGUSON, CORNACEAE 111 stp . Cornus. a~e, he Amomum subsp. Amomum: a, mcapcdaad branchlet, x Fic. aa: wer, x 6; flower (with petals and stamens rem ) in semi- diagrammatic vertical posal to show nectariferous disc en two- vi ar Ovary ure dru stone in latera with solitary, pendulous ovules, X 6; d, matur pe, X 2; e, view and from above, pa: © f-h, C. flori i abov. 1/2: flow 4; stone from above and in lateral view, b g, flo iC altenafolie i; stone in lateral view and from above to show apical pit. 112 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 with or after the foliage. Cornus Amomum Mill, (Swida Amomum (Mill.) Small), distinguished by its usually blue fruit, smooth leaves, dilated style, brown pith, and long calyx lobes, is represented with us by two subspecies treated as distinct species by some authors. Subspecies Amomum, 2n = 22, with leaves rounded at the base and green on the lower surface, extends from southern New England, west to southern Illinois and Kentucky, and south to Tennessee, the Carolinas, Georgia, Alabama, and northern Missis- sippi; subsp. obliqua (Raf.) J. S. Wilson, 2n = 22, characterized by cuneate leaf bases and white papillae on the lower surfaces of the leaves, occurs from northwestern New England and adjacent Canada, westward to North Dakota, Kansas, and Nebraska, and south to Arkansas and western Tennessee. Cornus asperifolia Michx. and C. Drummondii C. A. Meyer (Swida asperifolia sensu Small, S. Priceae Small) are distinguished from other species in the subgenus by their scabrid upper leaf surfaces. These two species are closely related, although Wilson has treated the former as a subspecies (microcarpa (Nash) J. S. Wilson) of C. foemina. Cornus microcarpa Nash (Swida microcarpa (Nash) Small) appears to be only a variant of C. asperifolia. Cornus Drummondii, occurring from Alabama and western Tennessee to Texas, Arkansas, Nebraska, and Ontario, is separated by its white drupes and white papillose indumentum on the lower surface of the leaves from C. asperifolia, a species of the southeastern Coastal Plain from Florida to Georgia, South Carolina, and Alabama with blue drupes and nonpapillose leaves. The complex Cornus foemina Mill. is distinctive in its smooth twigs and the smooth upper surface of the leaves. Subspecies foemina (Swida stricta (Lam.) Small), characterized by blue fruits, usually flat-topped inflores- cences, and red young branches, occurs throughout the southeastern United States and extends to Delaware, Indiana, and Texas. Some workers regard the closely allied C. racemosa Lam. (white fruit, usually convex in- florescences, and brown young branches) as occurring in the southeastern United States, but the treatment followed here is that of Wilson, who regards this taxon as C. foemina subsp. racemosa (Lam.) J. S. Wilson with a distribution to the north of our area. The frequent hybridization between species of subg. KRANropsIs often makes identification difficult. The color of the pith and of the fruit, im- portant taxonomic characters for distinguishing between some of the sub- species, should be recorded by collectors. The similar subg. Mrsomora Raf. (sect. Bothrocaryum (Koehne) Harms ex Nakai; Bothrocaryum (Koehne) Pojark.) is distinctive in the alternate leaves, stones with a characteristic deep apical pit, and chromo- some number. It includes only Cornus controversa Hemsl., 2n = 20, of eastern Asia, and C. alternifolia L. (Swida alternifolia (L.) Small), 2n = 20, widespread in the northeastern United States and adjacent Canada and extending southward through eastern Tennessee, the Carolinas, Geor- gia, and Alabama to northwestern Florida (Gadsden County), and west- ward to Minnesota and northern Arkansas. 1966 | FERGUSON, CORNACEAE 113 Subgenus CyNoxYLon (Raf.) Raf. (sect. Cynoxylon Raf.; Benthamidia Spach; Cynoxylon (Raf.) Small), characterized by woody habit, flowers aggregated into sessile heads surrounded by large, white, petaloid in- volucral bracts, and free, thin-fleshed drupes, is represented by Cornus florida L. (Cynoxylon floridum (L.) Small), 2n = 22, widespread in our area (but probably absent from southern Florida) and extending north to southern Maine and Ontario, and west to eastern Texas and Kansas, with the disjunct but doubtfully distinct var. Urbiniana (Rose) Wangerin in Mexico (Nuevo Léon and Veracruz). A calciphile, it is one of the most characteristic and conspicuous spring-flowering trees of the eastern United States. The herbaceous subg. ARCTOCRANIA Endl. ex Reichenb. (sect. Cornion Spach; Chamaepericlymenum Hill) is represented in the eastern United States by Cornus canadensis L., bunchberry, 2n = 44, of transcontinental distribution northward and sporadic occurrence southward to California and New Mexico, in the West, and to Pocahontas County, West Virginia, in the East. Wherry has suggested that the migration of C. canadensis southward “is restricted by its inability to establish itself in soils which are heated above 65° [F.] [18° C.] during the summer.” The morphological relationships between the subgenera appear to be reticulate. (See also Rickett, 1950.) A number of interspecific hybrids are known, but only a single hybrid between subgenera, Cornus < acadien- sis Fern. (C. alternifolia < C. stricta [C. stolonifera]), has been recorded. Basic chromosome numbers of 9, 10, and 11 have been reported: 2” = 18 in subg. Cornus; 20 in MresomoraA; 22 in BENTHAMIA, CYNOXYLON, and Kraniopsis; and 22 or 44 in ArctocraniA. (A triploid count of 27 has also been recorded for C. mas, 2n = 18.) Serological studies support the distinctness of subg. ARCTOCRANIA, CYNOXYLON, and Kraniopsis but show a close relationship between subg. BENTHAMIA (C. Kousa Buerger ex Miq., 2n = 22) and subg. CyNoxyton (C. florida and C. Nuttallit Audubon). Flowers are usually homogamous, but may be proterandrous in subg. ARCTOCRANIA (e.g., Cornus suecica L., 2n = 22), with exposed nectar secreted by the epigynous disc surrounding the style. Pollination is probably accomplished by insects, including small creeping Coleoptera and some Diptera and Hymenoptera. Cross-pollination is favored by the dif- fering lengths of style and stamens, but self-pollination may also occur by the spreading of the stamens to touch neighboring flowers. Fossils of Cornus have been reported from late Cretaceous, Tertiary, and Pleistocene strata. The remains include leaf impressions, stones, wood, and pollen. Many species, especially Cornus florida, are cultivated as ornamentals. Cornus florida contains a bitter principle, cornine, which has been used for its medicinal properties. The wood of some species is valuable for its hard, smooth, fine texture and is used for tool handles, cotton reels, meat skewers, and charcoal. 114 JOURNAL OF THE ARNOLD ARBORETUM [VOL. 47 REFERENCES: Under family references see HuTCHINSON, NAKal, RIcKETT, and WANGERIN. Ammons, N., & E. Core. The dogwoods of West Virginia. Castanea 10: 88-91. 1945. eh also STRAUSBAUGH & Coreg, Fl. W. Va. 3: 690-693. 1958. BEAN, W. J. Cornus florida, var. rubra. Bot. Mag. 136: pl. 8315. 1910. [Forma rubra (Weston) Palmer & Steyerm . Cornus Nuttallii and its allies. Bull. ii Inf. Kew 1915: 177-179. 2 pls. 1915. [C. florida, C. Kousa, C. capitat BortHwick, H. A. Light effects on tree plein and seed germination. Ohio Jour. Sci. 57: 357-364. 1957. [C. florida. BoyNTON, K. R. Cornus stricta. Addisonia 8: 13, 14. pl. 263. 1923. [C. foemina. | BRITTON, E. G. Wild plants needing protection. 9. “Flowering dogwood” (Cynoxylon floridum). Jour. N. Y. Bot. Gard. 14: 133, 134. pl. 120. 1913. [C. florida. ] BucaLa, W. A new system of the extent [sic] genus Cornus L. (In Polish; Russian & English summaries.) Ann. Sect. Dendrol. Soc. Bot. Pologne 9 205-210. 1953. [Discussion of system proposed by PoyarKova, 1950.] Bunt, C. A. The nomenclature of some species of Cornus. Rhodora 37: 222, 223. 1935. [Favors C. obliqua Raf. over C. Purpusii Koehne and C. stricta Lam. over C. candidissima Mill. | CuHopra, R. N., & H. Kaur. Some aspects of the embryology of Cornus. Phytomorphology 15: 353-359. 1965. [C. alternifolia, C. Amomum, C. macrophylla, C. oblonga, C. stolonifera Cuno, J. B. ae of dogwood and persimmon. U. S. Dep. Agr. Bull. 1436: 1-43. 1926 DALLIMORE, Ww. ote uses of Cornus wood. Bull. Misc. Inf. Kew 1915: 179-181. 1915, D’Amato, F. Osservazioni cito-embriologiche su Cornus mas L. con particolare riguardo alla sterilita di un biotipo triploide. (English summary.) Nuovo Gior. Bot. Ital. II. 53: 170-210. 1946. Danby, J. E. Some new names in the British flora. Watsonia 4: 47. 1957. [Notes on nomenclature of Thelycrania; adopts T. sericea (L.) Dandy for Cornus stolonifera Michx. Also see FosBErc. | Davis, O. H. Germination and early growth of Cornus igen Sambucus canadensis and Berberis Thunbergii. Bot. Gaz. 84: 225-263. 1927 DERMEN, . Cytological studies of Cornus. Jour. Arnold Arb. 13: 410-416. pl. 53. 1932. [Undocumented chromosome counts of 23 spp. | FARWELL, O. A. Concerning some species of Cornus of ve preps Rhodora 33: 68-72. 1931. [C. Amomum, C. candidissima, C. foemi nia and Cynoxylon not genera of Rafinesque. bid. 34: 29, 30. 1932. [Subgenera of Cornus. | Cornus, a reply. Torreya 42: 130. an [Further maintains Cynoxylon and Eukrania 4 Rafinesque are subgene FERNALD, M. L. A hybrid Cornus from — Breton. Rhodora 43: 411, 412. 1661. fC. ar ie Fosserc, F. R. Co big pelea L. (C. stolonifera Michx.) Bull. Torrey Bot. Club 69: aaa Hann, R. M., & C. , i. Scyllitol from flowering dogwood (Cornus florida). Jour. Biol. Chem. 68: 399-402. 1926.* 1966 | FERGUSON, CORNACEAE 115 Hara, H. The nomenclature of the flowering dogwood and its allies. Jour. Arnold Arb, 29: 111-115. 1948, [Maintains Benthamidia as a genus.] Hoim, T. Medicinal plants of North America. 34. Cornus florida L. Merck’s Rep. 18: 318-321. 1909.* [See also Bot. Jahresb. 37(1): 994. 1912.] Howarp, fe Registration lists of cultivar names in Cornus L. Arnoldia 21: 18. Kano, Dir Uber ein neues mechanisches hag in den Blattern der Cornus- Bot. Mag. Tokyo 51: 926-930. ake F. Uber die ah bs der Friichte von Cornus L. und verwandter Gattungen. Planta 36: 85-102. 8. ee. E. Die Sektion Microcarpium der Gattung Cornus. Mitt. Deutsch. Dendrol. Ges. 12: 27-50. 1903. Krart, H. D. The natural habitat of Cornus mas L. with some remarks about classification of plant ee, hn Dutch; English summary.) Nederl. Dendrol. Ver. Jaarb. 20: 169-201. urz, H., & R. K. Goprrey. Trees olga Florida. xxxiv + 311 pp. Gainesville, Fla. 1962. [ Cornus, 246-250. ] Meyer, C. A. Sur quelques espéces de Cornus appartenant au sous-genre Thelycrania. Ann. Sci. Nat. Bot. III. 4: 58-74. 1845. [See also Mém. Acad. Sci. St.-Pétersb. Sci. Nat. VI. 5(Bot.): 191-223. 1846.* Morse, W. C. Contribution to the life history of Cornus florida. Ohio Nat. 8: 197-204. ~ Narr, G. V., & E. von Rupuirr. Isolation of hyperin from red-osier dogwood (Cornus olor Michx.) Canad. Jour. Chem. 38: 2531-2533. 1960.* [C. seri OLSEN, C. The stracture and biology of Arctic flowering plants, od 9. Corna- ceae. Medd. Grgnl. 37: 127-150. 1914. [C. canadensis, C. sue OswaLp, F. W. A new color form of cornel from New Teriey. ioe x 337, 338. 1956. [C. racemosa f. caeruleocarpa Oswald. ] POYARKOVA A. De systemate generis Linneani Cornus L. (In Russian & Latin.) t. Syst. Leningrad 12: 164-180. 1950. [Divides Cornus into 6 genera. ] . De sectione Benthamia (Lindl.) Nakai generis Cynoxylon Raf. (In Russian & Latin.) Jbid, 181-195 Raymonp, M., & J. KucynraK. Cornus racemosa in Quebec. Rhodora 49: 23, 24. 1947 Rickett, H. W. Cornus Amomum and Cornus candidissima, Rhodora 36: 269- 274. 1934. nomenclature in Cornus. Ibid. 38: 51. 1936. [See also C. Drum- mondit, ibid. 99, 100.]| ornus asperifolia and its relatives. Am. Midl. Nat. 27: 259-261. The n i of Cornus. Torreya 42: 11-14. 1942. [See also “Cornus again,” ibid. 131, concerning the generic status of Cynoxylon and apes J Cornus sone es and Cornus occidentalis. Brittonia 5: 149-1 1944. ; . New combinations in Cornus. Bull. Torrey Bot. Club 72: 223. 1945. Cornus in Mexico, with notes on the evolution of the genus. Anal. Inst. Biol. México 21: 83-94 : Sanpo, C. E. Inositol from blackberry (Rubus argutus Link) and Pees dog- wood (Cornus florida). Jour. Biol. Chem. 68: 403-406. 192 SARGENT, C. S. Cornus. Silva N. 63-72. pls. 212-216. 1893. [C. florida, C. Nuttallii, C. alternifolia.] 116 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 SCHAFFNER, J. H. ec to the Ohio dogwoods in the winter condition. Ohio 19. 1905 SHANK, C. K. The leaf and stem anatomy of Cornus florida in relation to light exposure. Am. Midl. Nat. 19: 419-426. 1938. Sims, J. Cornus florida. Bot. Mag. 15: pl. 526. 1801. SmaLL, J. K. A Kentucky cornel. Torreya 1: 54, 55. 1901. [C. Priceae.] SoyAK, J. Swida hungarica (Karp.) in der Tschechoslowakei. Acta Dendrol. Cechosl. 3: 105-107. 1962. (See also Novit. Bot. Del..Sem. Hort. Bot. Univ. Carol. Prag. 1960: 9-11. 1960.) [Notes on the nomenclature of ornus. STANDLEY, P. C. The genus Cornus in South America. Trop. Woods 43: 16, 17. 1935. StrAKA, H. Die Pollenmorphologie der europaischen Cornus-Arten und einiger ahnlicher Pollentypen. Flora 141: 101-109. 1954. (See ae reprint in a.| TIEGHEM, P. vAN. Structure de quelques ovules et parti qu’on en ai tirer pour améliorer la classification. Jour. Bot. Morot 12: 197-220. 1898. [Single thick integument in Cornus. | TomaszewskA, E., & M. Tomaszewski. The chemical composition of fruits of the edible Cornus mas L. and Cornus officinalis cesar et Zucc. (In Polish; English summary.) Arb. Kornickie 2: 239-241. J TraLavu, H. The recent and fossil distribution of some ake and arctic montane plants in Europe. Ark. Bot. II. 5: 533-582. pls. 1-8. 1963. [Cornus suecica, 545-547. | Wherry, E. T. Temperature relations of the bunchberry, Cornus canadensis L. Ecology 15: 440-443. 1934. Witkinson, A. M. Floral anatomy of some species of Cornus. Bull. Torrey Bot. Club 71: 276-301. 1944. [Sixteen spp. in 5 subg. | . The floral anatomy and morphology of some species of Cornus and of the Caprifoliaceae. Abstr. Theses Cornell Univ. 1945: 184-187. 1946.* Witson, J. S. Variation of three taxonomic complexes of the genus Cornus in eastern United States. Trans. Kan. Acad. Sci. 67: 747-817. 1965. [C. Amomum, C. Drummondii, C. foemina. | ARNOLD ARBORETUM Present address HARVARD UNIVERSITY FLORA Eumoraka RESEARCH PROJECT THE UNIVERSITY LIVERPOOL, ENGLAND 1966 | EYDE, NYSSACEAE 117 THE NYSSACEAE IN THE SOUTHEASTERN UNITED STATES ! RICHARD H. Eypr NYSSACEAE Dumortier, Anal. Fam. Pl. 13. 1829 (A. L. de Jussieu, Dict. Sci. Nat. 35: 267. 1825, ‘““Nysées”), nom. cons. (TUPELO FAMILy) Trees, mostly [polygamous or] androdioecious, with alternate, exstipu- late leaves. Flowers small, + regular, subtended by 1-5 bracts. Floral inferior 1(rarely 2)-locular [or often 2-locular, or 6—8-locular], with 1 pendulous, apotropous ovule in each locule. Fruit a drupe; each locule of en opening at apical part by a + triangular abaxial valve at germina- tion. Embryo straight; endosperm copious. (Cornaceae, in part, of many pa including Davidiaceae H. L. Li.) Type Genus: Vyssa L. Three genera and seven or eight species, distributed in eastern Asia and eastern North America, mostly in warm-temperate areas. Camptotheca Decaisne and Davidia Baillon are monotypic genera of China, while Nyssa is eastern American-eastern Asian in distribution. Harms treated Camptotheca and Nyssa as subfam. Nyssoideae and Davidia oe as subfam. Davidioideae of the Cornaceae. Wangerin re- tained these as the two subfamilies of a separate family, Nyssaceae. A close relationship between Cornaceae and Nyssaceae, supported by evi- Prepared for a generic flora of the peerieay NORE States, a joint project of the Arnold Arboretum and the Gray Herbar of Harvard University made possible through the support of George R. ‘Coo ley rey the ional Sagi Foundation and under the direction of Carroll E. Wood, Jr., and Reed C. Rollins. The treatment follows the pattern established in the first paper in the series (Jour. Arnold Arb. 39: 296-346. 1958) and continued to the present paper. — area covered includes North and South Carolina, Georgia, Florida, Tennessee, Alabama, Mississippi, Arkansas, and Louisiana. The descriptions apply primarily to the Sa of this area, with supple- a brackets. Preserved flowers and fruits were obtained with the aid of Mr. W. W. es Superintendent, Seashore Natural Area, Virginia Beach, Virginia; Mr. Be on L. Tarbox, Jr., Director, Brookgreen Gardens, Murrells Inlet, South Carolina f. Elso S. Barghoorn, Harvard University; and Prof. R. B. Channell, Vander bilt aie Rancho Santa Ana Botanic Garden, for his thoughts on the infraspecific variants of Nyssa sylvatica. The drawings ira ting Nyssa were made by Arnold D. Clapman under the supervision of Dr. Woo 118 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 dence from palynology, floral and wood anatomy, serology, and cytology, has not been challenged in recent years; however, the Nyssaceae are here tentatively treated as a separate family because of their greater number of petals and stamens and their unique triangular germination valves. The chromosome number of Davidia has been reported as 2n = + 40 and that of Nyssa sylvatica as 2n = 44. The latter count may provide further evidence that Cornaceae and Nyssaceae are related, since 2n = 22 for most species of Cornus. Considering the present small size of the group and its rather limited modern distribution, the Nyssaceae have one of the most remarkable fossil records of all angiosperm families. Excellently preserved fruits have been found in England, continental Europe, Siberia, Japan, and the western United States; nyssaceous pollen is a frequently encountered microfossil in continental deposits of Tertiary age throughout the Northern Hemi- sphere; and leaf impressions resembling foliage of Nyssa are often found in association with the pollen and fruits. Moreover, silicified woods from Japan and from the state of Washington and a lignitized wood from the Netherlands have been assigned to the family. Fossil Nyssaceae are common in all major time zones from the Eocene through the Pliocene; the present restricted distribution of the family is clearly a relatively re- confirmed by fossils. Many of the oldest fruits are three- or four-carpellate, and vestiges of major vascular bundles provide evidence for the greater number of petals and stamens in these extinct forms. Some of the ancient fruits are also much larger than their present-day counterparts. The affinities of certain fossils from North America are with modern Asiatic Nyssaceae. Aside from Nyssa, only Davidia involucrata Baill., dove tree, cultivated to a limited extent for its showy, bracteate inflorescences, is of any eco- nomic importance. REFERENCES: For complete fpsiiygecnt see EypE and Eype & BARGHOORN. Burcu, J. vAN DER. Holzer der niederrheinischen Braunkohlenformation I. Hilzer a. Braunkohlengrube “Anna” zu Haanrade (niederlandisch Lim- burg). Acta Bot. Neerl. 13: 250-301. 1964. CHANDLER, M. E. J. The Lower Tertiary floras of southern England. I-IV. Brit. Mus. (Nat. Hist.), London. 1961-1964. [Reports of fossil Nyssaceae in all 4 vols. | DeERMEN, H. Cytological studies of Cornus. Jour. Arnold Arb. 13: 410-415. 1932. [Includes chromosome counts for Nyssa sylvatica and Davidia in- Bee aif Eype, R. H. Morphological and paleobotanical studies of the Nyssaceae, I. . survey of the modern species and their fruits. Jour. Arnold Arb. 44: 1-59. 1963. A peo distribution maps. E. S. BarcHoorn. Morphological and paleobotanical studies of the re \ 1966 | EYDE, NYSSACEAE 119 Nyssaceae, II. The fossil record. Jour. Arnold Arb. 44: 328-376. 1963. FaIRBROTHERS, D. E., OHNSON. Comparative serological studies within the families Cornaceae (dogwood) and Nyssaceae (sour gum). P 305-318 in C. A. Leone, ed., Taxonomic biochemistry and serology. New York. 1 1964, Harms, H. Die Gattungen der Cornaceen. Ber. Deutsch. Bot. Ges. 15: 21-29. 1897. [Includes Nyssaceae. Cornaceae. Nat. Pflanzenfam. III. 8: 250-270. 1898. [Includes Nyssaceae. | 5 geome F. Cornaceae. Jn: W. Jongmans, ed., Fossil. Catal. II. 23: ,1- 188. 938. : cage Laubgewachse der Braunkohlenzeit. x + 783 pp. Halle a.d. Saale. 1 Lr, H. ‘. & J. R. Scoramm. Davidia in the Philadelphia region. Morr Bull. & 31-33. 1954. [Includes brief review of discovery and i lodutiion into cultivation. ] MarxcraF, F. Die phylogenetische Stellung der Gattung Davidia. Ber. Deutsch. Bot. Ges. 76(1. Generalversammlungsheft): (63)-(69). 1964 [“‘1963”]. [Suggests a phylogenetic relationship between Davidia and Actinidiaceae; sae opinion contradicts that of other authors that Davidia is related to Nys yss MENNINGER, E. A. Davidia, the dove tree. Am. Hort. Mag. 44: 142-149. 1965. Mrkr, S. Endocarp remains of Alangiaceae, Cornaceae, and Nyssaceae in Japan. Jour. Inst. Polytech. Osaka Univ. D. 7: 275-297. 1956. Souma, K. Pollen morphology of the Nyssaceae, I. Nyssa and Camptotheca. Sci. Rep. Téhoku Univ. Biol. 29: 389-392. 1963. Turercart, F., & U. Frantz. Some spores and pollen grains from the Tertiary browncoal of Neyveli. Palaeobotanist 11: 43-45. 1963 [“1962”]. [In- cludes photographs of a new sp. of Tertiary (Miocene?) ame Lucknowia Sahnii, of southern India. “Affinity to Nyssaceae seems proba TITMAN, P. W. Studies in the woody anatomy of the family Nyssaceae. Jour. Elisha Mitchell Sci. Soc. 65: 245-261. 1949. WancerIn, W. Die Umgrenzung und Gliederung der Familie der Cornaceae. Bot. Jahrb. 38(Beibl. 86): 1-88. r Nyssaceae. Pflanzenreich IV. oe 41): 1-20. 1910 Wasscuer, J. Nyssaceae. Jn: C. G. G. J. vAN STEENIS, FI. Males. I. 4: 29-31. 1. Nyssa Linnaeus, Sp. Pl. 2: 1058. 1753; Gen. Pl. ed. 5. 1028. 1754. Deciduous trees, becoming shrubby under adverse conditions of growth; branching markedly excurrent; bark gray, divided into segments by deep fissures; base of trunk enlarged and roots sometimes forming arches when growing in water. Leaves alternate, often crowded near ends of branches, simple, petiolate, exstipulate; blade membranaceous or subcoriaceous, el- liptic or ovate to narrowly or broadly obovate, entire (or irregularly den- tate, especially on seedlings and on sprouts arising from roots). Inflores- cences (or solitary flowers) axillary, pedunculate, produced with the leaves from mixed imbricate buds. Staminate flowers in short racemes or heads; calyx lobes minute; petals 5-10; stamens 8-15, much longer than petals, usually in 2 + distinct whorls, the outer whorl antipetalous, the 120 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 WV; 1. Nyssa. a-i, N. ogeche: a, branchlet with immature fruit, x 1/2; b, ehsesin unisexual flower — note scars of fallen bracts at base of ovary, X 4; c, disc and style with rudimentary second style, ves 4; d, flower in diagrammatic vertical section to show locule and ovule, « 4; e, diagrammatic cross section ovary and ovule, X 4; f, mature fruit, abaxial surfac e, X 1; g, immature fru from “a” in diagrammatic cross section to show radicle of embryo, endospes Wine.” 1966 | EYDE, NYSSACEAE 121 inner alternipetalous; anthers on short, green, filiform filaments, introrse, longitudinally dehiscent, basifixed; pollen tricolporate; gynoecium absent, except for occasional presence of stylar rudiment at center of pulvinate disc. Bisexual, ¢ -sterile, or 9 flowers solitary or in inflorescences of 2—4 [or more] and sessile (in ours) on slender peduncles; floral tube urceolate or campanulate, densely tomentose to glabrous, adnate to the ovary; calyx lobes, when present, usually 5; petals 5-8, tomentose or glabrous, ovate or oblong; stamens often visibly abortive, fewer than in ¢ flowers or lack- ing, about equal in length to petals; nectariferous disc pulvinate or coni- cal; gynoecium syncarpous, 1 (rarely 2)-carpellate; style 1 (rarely 2, more often with only the rudiment of a second style); stigma reflexed or revo- lute, abaxially sulcate; ovary inferior, adnate to floral tube and disc, usually 1-locular (or 2-locular when 2 styles are present) and pseudo- monomeric, with 1 apotropous, 1-integumented ovule pendent from the axile placenta near apex of (each) locule. Drupe ovoid or ellipsoid, blue- black, purple, or red, with corky epidermal spots, crowned by the persistent disc; outer tissues fleshy; woody endocarp ovate, obovate, or elliptic in outline, sometimes flattened abaxially, bearing broad longitudinal ridges, sharp ridges, or papery wings. Seed conforming to locule; seed coat mem- branaceous. Type species: N. aquatica L. emend. Du Roi; see Eyde, Taxon 13: 129. 1964. (Nyssa or Nysa, the name of a nymph, applied by Linnaeus in the belief that all members of the genus are aquatic; cf. Hortus Cliffortianus 462. 1738.) — TUPELO, GUM. Three or four species in North America and two in southeastern Asia. Nyssa sylvatica Marsh. var. sylvatica (including vars. dilatata Fern. and caroliniana (Poir.) Fern.), black gum, sour gum, or pepperidge, 2” = 44, is a common constituent of moist forests from Maine to southern Florida, west to Texas (Brazos River), Oklahoma, eastern Kansas, Illinois, and Michigan, with disjunct Mexican stations in Hidalgo, Puebla, and Chiapas. A tree frequently 20 m. tall (occasionally more than 25 m.), it is charac- terized by entire-margined obovate to elliptic leaves; pediceled staminate flowers; two, three, or four fertile flowers on each peduncle; blue-black ovoid (to globular) drupes 1—1.5 cm. long with five distinct calyx lobes; and stones with low, rounded ridges. The var. a ee Sarg. (in- cluding NV. ursina Small), swamp black-gum, 2” = 44, differs in its narrower leaves more gradually tapered at base, an atin peduncles, and a predominance of paired fertile flowers (although one or three may occur). It is largely confined to very wet soils or standing water in swamps (stippled), developing endocarp (note germination valve, a a >» and pap wings in soca 2; h, mature stone with papery wings — note lp tte bundle between wings x - i, same, wings removed, eal onilive with germination valve, « -n, N. aquatica: j, staminate flower seen from above showing petals, stamens, and disc, X 4; ellate flower before anthesis stigma and style not fully developed, three petals removed, x 4; 1, mature fruit 1; m, stone, abaxial surface, showing germination valve, X 2; n, same, adaxia s surface, X 2. 0, N. sylvatica: stone, abaxial surface, showing germination valve, x 2 122 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 of the Coastal Plain from eastern Texas to eastern Virginia. This variant is treated as a species, NV. biflora Walt., by some authors, but the morpho- logical differences that separate it from var. sylvatica are not as great as those that distinguish undisputed species of Nyssa, and plants are found that are intermediate in character between the two taxa. Since var. biflora can be roughly delimited ecologically, geographically, and morphologically, some authors may prefer to treat it asa subspecies. The closest relative of Nyssa sylvatica is N. sinensis Oliver, of meso- phytic forests in southern China and Burma. Nyssa sinensis is notable for having pediceled fertile flowers borne mostly four in a cluster and in commonly having two styles and a bilocular ovary. Of the two, V. sinensis is morphologically more primitive and somewhat less winter hardy. Fossil materials show that the subgeneric stock which gave rise to this pair of species has been in existence since early Tertiary times and that it was distributed around the Northern Hemisphere before the Pleistocene glacia- n. Although bearing no close relationship either to each other or to the Nyssa sylvatica alliance, N. aquatica L. and N. ogeche Bartr. ex Marsh. share in common capitate staminate inflorescences, solitary fertile flowers with obsolete calyx lobes, and drupes longer than 2 cm. Nyssa aquatica has leaves with coarse, irregularly distributed mucronate teeth; purple, oblong drupes with the floral tube constricted below the flat disc; and mucronate stones with eight to ten sharp longitudinal ridges. It is common in backwaters of the Mississippi Embayment and Gulf and At- lantic coastal plains, reaching its limits in southern Illinois, eastern Texas, northern Florida, and southeastern Virginia. In contrast, N. ogeche (in- cluding NV. acuminata Small) has leaves with variable margins; hairy- peduncled red drupes without a constriction beneath the conical disc; and sharp-pointed stones with about 12 ridges, their crests extended as wide, papery wings. Reaching only about 10 m. (vs. 33 m. in NV. aquatica), it grows only in swamps of southern Alabama, northern Florida, southern Georgia, and southernmost South Carolina. Anatomical features of the fruit suggest an affinity between NV. aquatica and N. javanica (Blume) Wangerin, of southeastern Asia. Locally, some nyssas develop a shrubby form when subjected to adverse growing conditions, such as periodic burning; the segregation of these forms as separate species (Nyssa acuminata, N. ursina) seems unsupport- able. The tolerance of Nyssa species for saturated soil and standing water is remarkable. Of our species, only NV. sylvatica var. sylvatica is found in upland situations; the others are characteristic of deep fresh-water swamps, where they occur in dense stands, with Taxodium the most fre- quent associate. The genus contributes a substantial proportion to the pollen-rain in fresh-water swamps; it is not surprising, therefore, that Nyssa is an important entity in palynological studies of fresh-water peats and brown coals. It is clear that many extinct species of the genus occupied aquatic habitats like their modern counterparts. Some of the 1966 | EYDE, NYSSACEAE 123 brown coals that have yielded nyssaceous pollen and fruits originated as long ago as the Eocene epoch. Although aquatic nyssas can grow in sites where water rises annually to depths of five feet or more and seeds can remain viable after months of submergence, periodic drought is required for germination and perpetuation of the species. Fruits float well, and water currents undoubtedly play a major role in distributing the aquatic species. The fruits of N. sylvatica, on the other hand, are relished by birds, which apparently aid in maintaining the wider distribution of this species. Pollination is by insects, but Wodehouse reports the collection of air- borne pollen about 150 m. from a tree of Nyssa sylvatica var. sylvatica. The leaves of Nyssa sylvatica var. biflora have the capacity for accumu- lating unusually large amounts of cobalt, a phenomenon which has been used in evaluating the cobalt status of soils. In the United States, Nyssa species annually provide more than 200 million board feet of lumber. Although a tendency to warp because of interlocking grain makes the wood unsuitable for many purposes, it is used extensively for boxes and crates, as well as for furniture, flooring, and paper pulp. The light wood from roots and stumps of V. aquatica was formerly used for floats and for a kind of surgical sponge called “tupelo tent.” Nyssa sylvatica is cultivated for its colorful autumn foliage. Plant- ings of V. ogeche have been undertaken in western Florida in an effort to increase the production of the highly regarded tupelo honey; fruits of this species can be made into a preserve, ‘“‘ogeechee limes.” REFERENCES: ALLEN, P. H. Black willow dominates baldcypress-tupelo swamp eight years after clear cutting. U. S. Dep. Agr. Forest Serv. SE. Exp. Sta. Res. Notes 177. 2 pp. 1962. AppLequist, M. B. Longevity of submerged tupelogum and baldcypress seed. La. Univ. Forestry Notes 27. 2 pp. 1959 Briscor, C. B. Diameter growth and effects of flooding on certain bottomland forest trees. 103 pp. D. F. Thesis. School of Forestry, Duke Univ. 1957. Brusu, W. D. Water tupelo, Nyssa aquatica L. Am. Forests 55(6): 28, 29. 1949. | Includes photographs. | Dororeev, P. I. Tretichnye flory zapadnoi Sibiri. 346 pp. Akad. Nauk SSSR, Bot. Inst. Komarova Leningrad. 1963. [Includes fruits of Nyssa from Tertiary deposits of w. Siberia; author assigns these to 4 fossil spp. ] EccLer, W. A., & W. G. Moore. The vegetation of Lake Chicot, Louisiana, after eighteen years impoundment. SW. Nat. 6: 175-183. 1961. [Survival of N. aquatica in deep water through production of sprouts from broken or decayed trunks. ] Eype, R. H. The discovery and naming of the genus Nyssa. Rhodora 61: 209~218. 1959. ———. Typification of Nyssa aquatica L. Taxon 13: 129-132. 1964. i Haaxman, P. J. Over lignum-Nyssae, Tupelo-hout, Tupelo-stiften. Nieuw Tijd. Pharm. Nederl. 12: 145-147. 1879. [Describes absorbent properties of tupelo tents. } 124 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 Hatt, T. F., & W. T. Penrounp. A phytosociological . bg a Nyssa biflora consocies in — Louisiana. Am. Midl. Nat : 369-375. 1939. & : ess-gum communities in the Blue a Swamp near Selma, Alabama. Plan 24: 208-217. Hom, T. Nyssa sylvatica Marsh. Am. Midl. Nat. 1: 128-137. 1909. [ Anatomy of root, stem, leaf; emphasis on seedling. | Hotroyp, H. B. The utilization of tupelo. U. S. Dep. Agr. Forest Serv. Circ. Ketter, E. L., R. M. Kinespury, & D. J. Faney. Boards and papers from shortleaf pine, black tupelo and southern white oak neutral sulfite semi- chemical pulps. U. S. Dep. Agr. Forest Serv. Forest Prod. Lab. Rep. 2141. 9 pp. + 8 tables. 1959. [Paper prepared from pulped wood of N. sylvatica has unusually high resistance to tearing. Keyarts, E. Tupelo honey. Am. Bee Jour. 104: 90, 91. 1964. [Illustrated description of tupelo honey harvest: botanical name of source incorrectly given as NV. aquatica. | Ktawitter, R. A. Sweetgum, swamp tupelo, and water tupelo sites in a South Caroling bottomland forest. 191 pp. D. F. Thesis. School of Forestry, Duke Univ. 1962. [See Diss. Abstr. 24(1): 12. 1963.] Kusora, J., V. A. Lazar, & K. C. Beeson. The study of cobalt status of soils in Arkansas and Louisiana using the black gum as the indicator plant. Soil Sci. Soc. Am. Proc. 24: 527, 528. 1960. Kurz, H., & R. K. Goprrey. Trees of northern Florida. xxxiv + 311 pp. Gainesville, Fla. 1962. [Nyssa, 233-240; periodic burning cited as cause mig shrubby habit of NV. biflora on Apalachicola River delta Latrp, C. R. Tupelo honey, a $75,000 industry in Northwest Florida. Glean. =. Cult. 91: 678-681. 1963. [Adapted from J. A. WurTForp. Tupelo Honey. Fla. Dep. Agr. Spec. Ser. 66: 75-84. 1948.] Mancit, E. Pullboatin’ on the Blind. Forests People 10(4): 12-16. 1960. Feces methods used to transport Nyssa logs in remote areas along Blind River, NEELY, D. N tupelo ee caused by Botryosphaeria ribis. (Abstr.) Phyto- pathology 49: 547. 19 Rickett, H. W. Sag N. Am. Fl. 28B: 313-316. 1945. [Nyssa.] SARGENT, C.S. Nyssa. Silva N. Am. 5: 73-84. pls. 217-220. 1893. [Excellent illustrations, detailed descriptions. | Sears, P. B., & G. C. CoucH. Microfossils in an Arkansas peat and their significance. Ohio Jour. Sci. 32: 63-68. 1932. [Pollen of Nyssa reported as 51% of total pollen at 36-inch depth in peat of recent origin; this is perhaps the highest percentage ever recorded for the genus. SHuUNK, I. V. Oxygen requirements for germination of Nyssa aquatica. Science 90: 565, 566. 1939. SOUTHERN Harpwoop Propucers, Inc. Tupelo and black gum from southern hardwood forests. So. Hardwood Inf. Ser. 4. 8 pp. 1937 or 1938. [Illus- trated pamphlet, excellent source on the production of lumber from JN. aquatica and N. sylvatica var. biflora. Toote, E. R. Tupelo lesion caused by Fusarium solani. Pl. Disease Rep. 46: 132, 133. 3962, Unirep STATES DEPARTMENT OF AGRICULTURE, ForEST SERVICE. Woody-plant seed manual. U. S. Dep. Agr. Forest Serv. Misc. Publ. 654. vi + 416 pp. 1948. [Endocarps of Nyssa require additional year for germination if allowed to dry out. | + 1966 | EYDE, NYSSACEAE 125 UNITED STATES DEPARTMENT OF AGRICULTURE, FOREST SERVICE, DIvIsIoN OF . Agr. Handb. 271. vi + 762 pp. 1965. [Nyssa, 277-286; includes distribution maps. | Upuor, J. C. T. Die Ges. 43: 2-16. 1931. var. biflora. WuitTTaker, R. H. Vegetation of the Great Smoky Mountains. Ecol. Monogr. 26: 1-80. 1956. [Includes statistical evidence that N. sylvatica var sylvatica grows predominantly on subxeric sites in the Smoky Mountains; interesting in view of the wet habitat of most spp. of Nyssa. ] amerikanischen Nyssa-Arten. Mitt. Deutsch. Dendrol. {Includes observations on pollination in NV. sylvatica DIVISION OF PLANT ANATOMY SMITHSONIAN INSTITUTION 126 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 THE GENERA OF ARALIACEAE IN THE SOUTHEASTERN UNITED STATES? SHIRLEY A. GRAHAM ARALIACEAE A. L. de Jussieu, Gen. Pl. 217. 1789, “Araliae,” nom. cons. (GINSENG FAMILY) Perennial herbs, vines, shrubs [or trees, rarely scandent epiphytes]. Leaves alternate on the stem or basal from the rhizome, petiolate, [simple,] palmately compound or pinnately or ternately compound or decompound, the stipules either adnate (in varying degrees) to the base of the slightly sheathing petiole or wanting. Inflorescence basically umbel- late, the umbels many flowered, solitary or arranged in panicles, racemes [corymbs, or rarely in compound umbels], the peduncles and pedicels subtended by lanceolate, deciduous or persistent bracts. Flowers regular, bisexual and/or unisexual, the plants polygamomonoecious, polygamodioe- cious, or dioecious. Floral tube adnate to the ovary, cupuliform, obconic, or cylindrical, articulated or continuous with the pedicel; calyx lobes 5, small, deltoid [or the apex of the floral tube merely undulate], persistent. Petals 5{-10], valvate or imbricate in bud, broad [or narrow] at base, greenish white to pink, arising from the margin of a fleshy, epigynous, flattened or cupuliform nectariferous disc, free and separately deciduous many as the petals and alternate with them, inserted epigynously ments filiform [or ligulate]; anthers oblong [to subglobose], Adie versatile, 2-locular, longitudinally dehiscent; pollen usually 30—40 p long, oblate-spheroid [to prolate], 3-colporate, the pores often transversely elongated, the exine reticulate. Gynoecium 2—5(6)[—15]|-carpellate, syn- carpous; styles persistent, generally as many as the carpels, distinct or *Prepared for a generic flora of the pupae vise States, a joint project of the Arnold Arboretum and the Gray Herbar rvard University made possible through the support of George R. Cooley ne oa se Science Foundation and under the direction of Carroll E. Wood, Jr., and Reed C. Rollins. The treatment follows the pattern established in the first paper in the series (Jour. Arnold Arb. 39: phil Ste 1958) and continued to the present paper. The area covered includes North d South Carolina, Georgia, Florida, Tennessee, Alabama, Mississippi, Arkansas, and panes na. The descriptions apply primarily to the plants of this area, with supple- mentary information in brackets. References not seen by the author are marked with an asterisk. The author is indebted to Dr. Lily M. Perry for references to Chinese medicinal uses of Panax and to Dr. Wood for his generous assistance and suggestions. Harry E. Ahles and Wilbur H. Duncan have kindly contributed information on Hedera in the Carolinas and Georgia, respectively. 1966 | GRAHAM, GENERA OF ARALIACEAE 127 united for a short distance above the disc [or completely united into a cone or column, or rarely wanting and the stigmata sessile]; stigmatic surfaces terminal and/or along the adaxial surface of each style; ovary inferior, 2-5(6)[—15]-locular, the wall adnate to the floral tube, the apex fused to the disc; ovules 2— are 15], one in each locule, anatropous, epitropous, pendulous, the raphe ventral, the placentation axile. Fruit a subglobose [or ellipsoid], sulcate [or smooth] drupe, bearing the persistent styles and calyx lobes; exocarp thin, this and the fleshy mesocarp drying to a membranaceous covering holding together the 2—5(6)[—15] oblong, laterally flattened [or globose] stones (pyrenes) [in one genus the fruit dry and the carpels partly separating]. Seeds with thin seed coat; endo- sperm abundant, uniform [or ruminate]; embryo straight, in the micro- pylar end of seed, minute at maturity of seed, the radicle superior. TyPE GENUS: Aralia L. A predominantly woody family of about 60 genera and 750 species, with the two major centers of speciation, tropical America and Indo- Malaya, sharing few genera; represented in the southeastern United States by Aralia and Panax, both of eastern American-eastern Asiatic distribu- tion, and possibly by the introduced Hedera Helix L.2 The familial classification generally followed is that of Harms, who di- vided the group on the basis of somewhat unsatisfactory characters of aestivation and shape of petals into three tribes: Schefflereae (petals with a broad base, valvate in bud), Aralieae (petals with a broad base, + imbricate in bud), and Mackinlayeae (petals narrowed at the base, valvate in bud). Viguier later recognized ten tribes (invalidly published) based mainly on anatomical characters. No modern treatment is available for the whole family, and few of the genera have been monographed, making it necessary to rely almost solely on regional treatments for determination of species and information regarding their biolo The Araliaceae are closely allied to the Umbelliferae, but the latter differ in being chiefly herbaceous and in possessing consistently bicarpel- late dry fruits, the carpels of which separate at maturity. Anatomically, the two families are very similar. The Araliaceae are notable for secretory canals in the pith, phloem, and cortex, the distribution of which is valuable in identifying the species and, to a more limited extent, in recog- nizing the genera. The pollen of Araliaceae is similar to that of the Um- belliferae and Cornaceae. In a study of some woody Umbellales, Rodriguez (p. 276) has sum- * Oplopanax horridus (J. E. Sm.) Miq., which ranges from Alaska, south to Cali- fornia, and east to Montana, with qo crags in the Siar Bay District, : nee horridus, as spinosa is oem in South Carolina, ina the esieromot ena for Oplopanax is more th nd miles away. It seems likely that a confusion in common names (“devil’s ine vs. “devil’s walking stick”) “led to the misidentifica- tion. — C. E. W. 128 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 marized what he believes is the relationship of the Araliaceae to related families: “Although the Umbelliferae form highly specialized expressions of donia|, whose fruit approximates most that of the Umbelliferae, resembles rather the Cornaceae in its wood anatomy. Both families, Umbelliferae and Araliaceae, apparently diverged from a single ancestral source. Of the tribes of Araliaceae, the Mackinlayineae | Mackinlayeae Harms| appear to be closest to the Umbelliferae on anatomical grounds. The information available on Cornaceae, Nyssaceae, and Garryaceae does not conflict with the assumption that they too may be divergent offshoots of the same source. Within the complex of families, the Umbelliferae are the most advanced in habit, fruit type, and secondary-xylem characteristics; the Garryaceae are the most advanced in flower morphology; the Araliaceae occupy a wide central position bridging the distance between those ad- vanced families and the more primitive Cornaceae and Nyssaceae.”’ The fossil record of Araliaceae, represented by leaves and pollen of several genera, is extensive. Leaves of Aralia and Hedera are reported from the Upper Cretaceous and Palaeocene of Alaska, far north of their present ranges. Panax is reported from the Oligocene of Colorado, and the tropical American Oreopanax Dec. & Planch. has been recorded from the Eocene of Tennessee, Mississippi, Arkansas, and Texas, and the Miocene of Oregon. Another tropical genus, Schefflera J. F. & G. Forst., has been reported from the Eocene of Tennessee and Kentucky. The Cre- taceous occurrence of members of the Araliaceae places it among the oldest known angiosperm families. Chromosome numbers of 2m = 22, 24, 44, 48, ca. 54, 96, ca. 127, and 192 have been reported, the most frequent numbers being 2m = 24 and 28. The family is of little economic importance. Rice paper, a soft, velvety material which finds limited use in the manufacture of artificial flowers, is the thinly sliced pith of T etoaetien sein (Hook.) K. Koch, a monotypic genus from Formosa and China. Roots of Panax are valued in China for medicinal use. About 28 sees are cultivated as ornamentals in the United States. The European ivy, Hedera Helix L., widely cultivated in a North America, is known to reproduce vegetatively to some degree as a waif tere naturalized, but it is uncertain whether this is so elsewhere and whether it ever spreads by seeds. Both the vars. Helix, 2n = 48, and hibernica (Kirchn.) Jaeger, 2m = 96, are said to be “naturalized in wooded areas of Virginia and North Carolina” (Lawrence). Hedera Helix is easily distinguished from other species by its 4-10-rayed stellate hairs; in the four other species the hairs are scaly or scalelike. Hedera is of special morphological interest because of its distinctive juvenile and adult forms, the unbranched juvenile stems climbing by adventitious rootlets and bear- ing lobed leaves lying in a single plane, in contrast with the nonclimbing, ” 1966 | GRAHAM, GENERA OF ARALIACEAE 129 branched and shrubby adult flowering stems with their unlobed leaves circling the stem. The leaves of juvenile stems are necessary for the determination of varieties, for within a given species of Hedera the leaves of adult stems do not show similar variation. The juvenile leaves of H. Helix var. Helix are rarely more than 6 cm. long, dark green with promi- nently whitish veins above; those of var. hibernica are glossier and larger (to 13 cm.) with green venation above. A number of references dealing with Hedera have been included with the family references. REFERENCES: BaILton, H. Ombelliféres. Hist. Pl. 7: 84-256. 1879. [Araliaceae, as tribe Aralieae, 151-170, 175, 244-256. ] BaumaANN, M. G. Myodocarpus und die Phylogenie der Umbelliferen-Frucht. Umbeliforen-Studien I. Bull. Soc. Bot. Suisse 56: 13-112. 6 pls. 1946. BAUMANN-BopENHEIM, M. G. Ableitung und Bau bicarpellat-monospermer und pseudomonocarpellater Araliaceen- und Umbelliferen-Friichte. (English summary.) Jbid. 65: 481-510. 1955. [‘‘The majority of these fruits (of Araliaceae) are either pentacarpellate or bicarpellate, homocarpellate, iso- spermous, clausicarpous stratocapsulae. . . . with the Umbelliferae in the form of bicarpellate and homocarpellate, isospermous, carpelli- — acon ai i aaibana ers with isolation of a protomericarpophor o mericarpophore eiieone G., & J. D. A en Araliaceae. Gen. PI. 1: 931-947. 1867. CANDOLLE, A. P. pe. Araliaceae. Prodr. 4: 251-266. 1830. Ducamp, L. Recherches sur l’embryogénie des Araliacées. Ann. Sci. Nat. Bot. VIII. 15: 311-402. pls. 6-13. 1902. Ser G F. Beitrige zur i creators Anatomie der Araliaceae. Inaug.-diss. p. pl. 1. Breslau. hee L Araliaceae. Nat. De ea III. 8: 1-62. 1894/1897 . Zur Kenntnis der Gattungen Aralia und Panax. Bot. Jahrb. 23: 1-23. 18 Hom, T. Medicinal plants of North America no. 97. Merck’s Rep. 25: 11-15, 62-65, 126-130, 177-180. 1916; 26: 6-8. 1917.* [Aralia and Panax. | JAcoBsEN, P. Chromosome numbers in the genus Hedera L. (Abstr.) Hereditas JoHanseN, D. A. Plant embryology. Embryogeny of the spermatophyta. xvi + 305 pp. Waltham, Mass. 1950. [Araliaceae, 204, 205. LaMorte, R. S. Catalogue of the Cenozoic fen of North America through 1950. Geol. Soc. Am. Mem. 51. 381 pp. 1952. [Aralia, 71, 72; Hedera, 182; Oreopanax, 236; Pusat: 242; Scheffiera, 323.| LAWRENCE, G. H. M. The cultivated ivies. agit Arb. Bull. 7: 19-31. 1956. [Hedera: includes keys to spp. and cultiva . The cultivated genera of Araliaceae. Bales 7: 133-140. 1959. [In- cludes key to 28 genera with leaf of each genus illustrated. | & A. E. Scuutze. The cultivated hederas. Gent. Herb. 6: 106-173. 1942, [Key to and taxonomy of the 5 cultivated spp.; H. Helix with 38 vars. | Li, H. L. The Araliaceae of China. Seas 2: 1-134. 1942. [Hedera, 49-51; Aralia, 101-116; Panax, 116-11 MEEHAN, T. On the flowers of dros spinosa L. and Hedera Helix L. Ann. Mag. Nat. Hist. IV. 7: 315, 316. 1871. Metcatre, C. R., & L. CHarx. Araliaceae. Anat. Dicot. 2: 725-735. 1950. 130 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 Mirra, S. P. Studies in the Umbellales. II. The vegetative anatomy. Jour. Indian Bot. Soc. 40: 424-443. 1961. [Nodal anatomy in some genera of Araliaceae, Umbelliferae, and Cornaceae, and consideration of the relation- ship of these families Mutter, C. Ueber phloémstindige aes genre der Umbelliferen und Aralia- ceen. Ber. Deutsch. Bot. Ges. 6: 2 Naxkar, T. Araliaceae Imperii Japonici. ae “Arnold Arb. 5: 1-36. 1924. [Hedera, 24-27; Aralia, 27-32; Panax, 32-36.] PerbuE, R. E., & C. J. KRAEBEL. The rice-paper plant — Tetrapanax papyrif- erum (Hook.) Koch. Econ. Bot. 15: 165-179. 1961. Rossins, W. J. Further observations on juvenile and adult Hedera. Am. Jour. Bot. 47: 485-491. 1960. [Juvenile and adult foliage. | RopricuEz, R. L. Systematic anatomical studies on Myrrhidendron and other woody Umbellales. Univ. Calif. Publ. Bot. 29: 145-318. pls. 36-47. 1957. SEEMANN, B. Revision of the natural order Hederaceae. 107 pp. 7 pls. London. 1868. [See also Jour. Bot. 2-6. 1864-1868. ] capa ise C. Araliaceae. N. Am. Fl. 28B: 3-41. 1944. [Nine genera; Aralia, ; Panax, 9, 10. See also notes in Brittonia 2: 247-261. 1936. ee R. Recherches anatomiques sur la classification des Araliacées. Ann. Sci. Nat. Bot. IX. 4: 1-210. 1906. Nouvelles recherches sur les wT Ibid. 9: 305-405. 1909. [Aralia, Acanthopanax, Schefflera, Dizygotheca. | WelssE, A. Blattstellungsstudien an Hedera Helix. I. Plagiotrope Sprosse und Samlinge. Ber. Deutsch. Bot. Ges. 42: 391-396. 1924; II. Orthotrope Sprosse und Bliitenstand. /bid. 43: 11-15. a Youne, R. A. Fatshedera. Natl. Hort. Mag. 17: 81-83. 1938. [Supposedly an intergeneric hybrid between Fatsia japonica var. Moseri and Hedera Helix var. hibernica. | KEY TO THE GENERA OF ARALIACEAE General characters: perennial herbs, shrubs, trees, or vines with alternate, often compound or decompound leaves; inflorescence basically an umbel; flowers regular, generally 5-merous; ovary inferior, the ovules solitary and pendulous in each locule ; fruit a drupe, with 2-5 or more oblong stones. A. Leaves simple, lobed or entire; vines climbing by adventitious roots. . bial yg ok ees Ee OT aa Re ee a etary ge i ee Ee [ Hede ra.| A. Leaves compound; perennial herbs or shrubs to small trees. B. Leaves ternately or pinnately compound or decompound, alternate or basal; inflorescence of 2 to many umbels; fruit purple or black m 1. Ara Leaves palmately compound, whorled at the summit of the stem; in- florescence a solitary umbel; fruit red or yellow 2. Panax Tribe ARALIEAE 1. Aralia Linnaeus, Sp. Pl. 1: 273. 1753; Gen. Pl. ed. 5. 134. 1754. Glabrous or spinescent perennial herbs, shrubs, or small trees, with stout, elongate rhizomes, our species generally inhabiting rich, moist woods. Leaves alternate on stem or (in acaulescent species) arising directly 1966 | GRAHAM, GENERA OF ARALIACEAE 131 from rhizome; petiole broad at base and partially fused with stipules, glabrous or with prickles; blade pinnately or ternately compound or de- compound, the leaflets composing each pinna 3 to many, ovate to lanceo- late-elliptic [or orbicular], obliquely cordate to acute or rounded at base, acute to acuminate at tip, glabrous [to densely pubescent], the margin finely to coarsely serrate or biserrate. Inflorescence axillary, terminal, or basal, the umbels solitary or paniculate [rarely compound]. Flowers 5- merous, bisexual and/or unisexual, the plants reportedly polygamomonoe- cious or dioecious [or polygamodioecious]. Floral tube cupuliform, ob- conic, or cylindrical, articulated with pedicel; pedicel slightly swollen at apex beneath flower and bearing small involucral bracts. Petals greenish white, oblong, the apex inflexed, valvate to slightly imbricate in bud, arising from the margin of the flattened disc, separately deciduous. Stamens introrse in bud; ents long in ¢ flowers, short in 9 flowers. Gynoecium (2—)5(6)-carpellate; styles (2—)5 (6), completely free or con- nate for a short distance above the disc, much shorter than stamens in 4 flowers, much longer in @ flowers; stigmatic surfaces terminal; ovary inferior, (2—)5(6)-locular, each locule with a single ovule. Fruit a sub- globose purple to black drupe with (2—)5 (6) oblong, laterally flattened stones. LECTOTYPE SPECIES: A. racemosa L.; see N. L. Britton, N. Am. Trees 733. 1908. (Etymology of name obscure; adopted by Linnaeus from Tournefort, who first received seeds of a plant bearing this name from Holland [Elemens Bot. 1: 249. 1694]. According to Fr. Marie- Victorin [Fl]. Laurent. 411. 1947], the name is of American Indian origin [perhaps Iroquoian, fide Webster’s New Int. Dict. ed. 2. 1959] and was communicated to Tournefort by the Quebec physician Sarrazin.) A genus of about 30 species in six sections, with eight species ounce nous to North America, the remainder in Asia and Malesia. Three, haps four, species range into the Southeast, the most widespread a Aralia spinosa L. (sect. ARBORESCENTES Harms), 2m = 24, a shrub or small tree with stout, often curved prickles, bipinnate leaves up to 75 cm. long, and a showy, compound terminal panicle of umbels. Popularly known as Hercules’ club or devil’s walking stick, it occurs locally throughout our area, except in southern Florida, in and along edges of damp woods. It s Harms), spikenard, widely distributed outside the southeastern United States, is known or reported from scattered counties in Tennessee, North and South Carolina, and northern Georgia, Alabama, and Mississippi. The plant is an unarmed-perennial herb with bipinnate leaves, the lower pinnae having more leaflets than the upper ones, and a compound, panic- ulate, terminal inflorescence. The dried rhizomes have been used, with- out medical basis, as a home remedy in the treatment of rheumatic fever, syphilis, and dermatoses. Aralia nudicaulis L. (sect. NANAE Harms), wild Sarsaparilla, 2n = 24, of northern, almost transcontinental, distribution, but with a southern extension through the Blue Ridge physiographic prov- 132 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 ince to northern Georgia, is easily recognized by its glabrous, acaulescent habit; ternately divided leaves, the pinnae of which are 3—5-foliolate; and scapose inflorescence. A fourth species, A. hispida Vent. (sect. HUMILES Harms), is reported in various floras as extending south in the moun- tains to North Carolina, but its presence there has not been confirmed. It is known with certainty as far south as Augusta County, Virginia, and Pocahontas County, West Virginia. A perennial herb, it is densely bristly, especially near the base, and produces several short, clustered stems from a stout horizontal rhizome. Twenty-five fossil species of Aralia have been described from the Terti- ary of North America, several from areas no longer within the range of the genus. Leaves have also been reported from the Upper Cretaceous (see family discussion). REFERENCES: Under family references see Harms (1896), Horm, LaMorte, Li, MEEHAN, NAKAI, SMITH, and ViGUIER (1906, 1909). Avpers, W. C. The oil and terpenes of Aralia nudicaulis, Am. Jour. Pharm. 71: 370-378. 1899. Berry, E. W. Aralia in American palaeobotany. Bot. Gaz. 36: 421-428. 1903. [Known fossil record of Aralia leaves; see also Bull. Torrey Bot. Club 37: 27, 28. 1910.] Davies, P. A. Origin and dk phiecrociam of prickles of Aralia spinosa. Trans. Ky. Acad. Sci. 10: 29. 194 Haar, A. W. VAN DER. sal zur Anatomie der Araliaceae. Die Blatter und Stengel von Aralia montana Bl. Rec. Trav. Bot. Néerl. 19: 277-280. pl. 9. 1922. a T. Contributions to the knowledge of the germination of some North American plants. Mem. Torrey Bot. Club 2: 57-108. pls. 5-19. 1891. [A. spinosa, 75, pl. 11, fig. 75.] SarcenT, C. S. Aralia. Silva N. Am. 5: 57-62. pl. 211. 1893. [A. spinosa.] 2. Panax Linnaeus, Sp. Pl. 2: 1058. 1753; Gen. Pl. ed. 5. 481. 1754. Perennial herbs of moist, rich woods; stem unbranched, erect, arising from a short rhizome and an elongate, often branched, or globose, un- branched tuberous root; base of stem subtended by a few membranaceous [or fleshy|, deciduous [or persistent] scales. Leaves 1-5, generally 3, in a single whorl at summit of stem, palmately compound, membranaceous, exstipulate, subtended by a narrow, lanceolate bract; leaflets 3-5, mostly 5, petiolulate or sessile, glabrous or with sparse [to dense] setae along the veins on the upper (adaxial) surface, narrowly elliptic to obovate, the base acute to rounded, the apex acute to long acuminate or obtuse, the margin coarsely to finely serrate or biserrate. Inflorescence a single long- peduncled, terminal umbel of many flowers [rarely 2-6 umbels arising at irregular intervals along the peduncle], the pedicels subtended by lance- olate bracts. Flowers bisexual and/or unisexual, the plants hermaphro- ditic or polygamodioecious (reportedly also polygamomonoecious). Floral tube obconical or cylindrical, articulated with the pedicel in bisexual and 1966 | GRAHAM, GENERA OF ARALIACEAE 133 @ flowers, continuous in ¢ flowers; pedicel slightly swollen at the apex be- neath the flower in all but the ¢ flowers; lobes of the calyx 5, broadly del- toid, small to minute. Petals greenish white to white or pink, oblong or broadly ovate, slightly imbricate in bud, arising from the margin of the cupu- liform nectariferous disc, generally separately deciduous, occasionally co- herent and falling together. Stamens wanting in 9 flowers. Gynoecium 2- or 3-carpellate, reduced to a single column in ¢ flowers; styles 2 or 3 (in bisexual and @ flowers only), arising from the center of the disc, free to the base or slightly united, the apices and adaxial surfaces stigmatic; ovary 2- or 3-locular, generally wanting in ¢ flowers; ovules 2 or 3. Fruit a red or yellow, subglobose, 2- or 3-sulcate drupe with 2 or 3 oblong, laterally flattened [or globose] stones. Lectotype species: P. eye L.; see Britton & Brown, Illus. Fl. No. U. S. ed. 2. 2: 618. (Name Greek and Latin, panax [| = panacea], an herb to which was pee the power to heal all diseases, from Greek, panakes, all-healing.* ) An eastern American- eastern Asiatic genus of about seven species, the northern Georgia. Panax quinquefolius, ginseng, 20-60 cm. tall, with a fusiform root, petiolulate leaflets, mostly bisexual flowers, two styles, and red fruits, ranges from Quebec to Minnesota, south to Oklahoma and Georgia, but has become rare in the wild because of exportation of its tuberous roots to China, where they are esteemed for their reputed cura- tive powers. The closely related Asiatic Panax Ginseng C. A. Meyer (P. Schin-seng T. F. L. Nees) ,4 2n = 44, has been used for centuries by the Chinese, who believe the resemblance of the roots to the human figure signifies their power to cure illnesses. The common name in China, “‘schin-seng,” “‘jen- seng,” or “ginseng,” means “essence of the earth in the form of a man.” The first description of ginseng by a European was published in 1714 by Pére Jartoux, a missionary in China. Soon afterward, Pere Lafitau, a missionary among the Iroquois in Quebec, found a similar plant (Panax quinquefolius) growing at Montreal. Exports of the dried roots to China mber of recent authors have treated Panax and its compounds (e.g., Den- of m * The classical gender of Panax is masculine in both Greek and Latin; it is aE that pear treated it as neuter. See A. C. Smith’s pertinent comments in N. : 944. Fl. “yn fconpect pl. 70.) 1833) has been ‘widely ae for this species, this name was supe mate name available is P. Ginseng C. A. Meyer (Bull. Phys.- -Math. Acad. Sci. St.- Pétersb. IT. 1: 340. 1843). 134 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 soon began, probably at first by way of France or England, but later directly from northeastern American ports. In 1860, alone, 600,000 Ibs. of dried roots of P. quinquefolius were exported to China; in 1886, the value of exported roots was $998,332 (Butz). The export of ginseng has greatly declined since World War II, but even in 1957 approximately 100,000 Ibs. were shipped, mainly to Hong Kong, with collectors or grow- ers receiving 20-24 dollars a pound for wild roots and some 40 per cent less for cultivated roots (Williams). Populations of wild ginseng have been greatly depleted in the eastern United States by “‘sang” (a contrac- tion of ginseng) diggers who dig the roots in the spring and summer before the plant produces seeds. Cultivation of the species has been attempted, but Alternaria blight and other fungal diseases have discouraged many growers. Today there are relatively few growers; the largest is Fromm Bros., Inc., with about 100 acres under cultivation in Hamburg, Wisconsin. Originally, the most valuable roots were those of Panax Ginseng grown in Chinese imperial gardens. Now the most valuable are those from Man- churia and Korea, the roots of the American a cori and the Japanese P. japonicus (T. F. L. Nees) C. A. r (P. Ginseng var. japonicus (T. F. L. Nees) Makino) being Sree he effective. Now rare in the wild as a result of collecting, ginseng is currently being culti- vated in both Korea and Manchuria. Panax Ginseng is grown from seeds which are collected in the fall when the embryo is not yet completely developed. Seeds must be stored for a year under moist conditions, with a three- to four-month cold period during which the embryo lengthens from 0.3 to 4.5 mm., and the surrounding endocarp splits along one suture. The stratified seeds are sown the following fall for germination in the spring. Five or six years’ growth is required before flowering begins. The germination and maturation of Panax quinquefolius is apparently similar to that of P. Ginseng. Characters used to distinguish the species pair are minor: the slope of the base of the leaflets and the degree of persistence of the fleshy or membranaceous scales at the base of the stem. The root, generally ground to a powder and made into a tea, is said to have a stimulating effect, enhancing digestion, blood circulation, and sexual desires, and is believed by users to be valuable in treating fatigue, nervous disorders, and blood diseases. At present there is little sound evidence that the constituents of the roots (which seem to be approxi- mately the same in the American and Asian species) act as more than a mild stimulant. The major components are a glucoside, panaquilon; a saponin, panaxine; and a volatile oil, panacen. In addition, the roots contain mucilage, resin, a hormone of the steroid group, and vitamins B, and Bz». REFERENCES: Under family references see HARMS (1896), LAMotrTeE, Li, NAKAI, SMITH, and VicurerR (1906). ANonyMous. American —— (Aralia quinquefolia, A. Gray). Bull. Misc. Inf. Kew 1893: 71-75. 1893. 1966] GRAHAM, GENERA OF ARALIACEAE 135 BaLaNnDIN, D. A. The chemical composition of Panax Ginseng. A survey of the literature. (In Russian.) Mater. Izuch. Zhen’-Shenia Limonnika 2: 77-96. 1 pl. 1955.* [Monographic survey. | Bryant, G. The wildcrafters ginseng manual; a guide to American ginseng. 40 pp. Terre Haute, Indiana. 1949.* Bucu, T. G. The physiology of seed germination in ginseng. (In Russian.) Bull. Glavn. Bot. Sada Moskva 20: 109-114. 1955 Borz,.G. CC. -The serge of American ginseng in Pennsylvania. Pa. Dep. Agr. Bull. 27: 1-24. 1897. Cuanc, Y. L. The cultivation of Panax Schin-seng Nees v. Esenb. at the foot of Changpaishan Mountain (southeastern Manchuria). Acta Soc. Harbin. Invest. Nat. Ethnogr. 12: 39-42. 1954. [P. Ginseng.] Esporn, I. Die Ginsengwurzel auf dem heutigen Drogenmarkt. Pharmazie 15 (2 ): 75-81. a [Includes morphology and anatomy of root. GRUNDULS, J. seng, the root of life. (In Norwegian; English summary. ) Blyttia 15: part 1957. [An account of ginseng diggers in Manchuria. | GrusuHvitzky, I. V. “Contractile roots’ — an important biological characteris- tic of Panax Ginseng C. A. M. (In Russian.) Bot. Zhur. 37: 682-685. 1952. . Panax Ginseng C. A. Mey. —the real name of ginseng. (In Russian.) Mater. Izuch. Zhen’-Shenia tenec: 13-18. 1955.* . The diseases of dase es A survey of the literature. (In Rus- sian.) Ibid. 35-70. figs. Instructions for arate measures of accelerated germination of ginseng seeds. (In Russian.) Bot. Zhur. 41: 1021-1023. 1956. [P. A contribution to the knowledge of biology of the wild-growing ginseng (Panax Ginseng C. A. M.). (In Russian; English summary. id. 44: 1694-1703. 1959. [Brief life-histo . M. Novicuxkova. A contribution to the knowledge of aie nthesis in the ginseng. (In Russian.) Bot. Zhur. 42: 751-756. 1957 HarpDInc, x R. Ginseng and other medicinal plants. 367 pp. Columbus, Ohio. 1908. [P. quinquefolius, 1- oe Horwammer, L., H. Wacner, & B. Lay. Zur Kenntnis der Inhaltsstoffe von Radix Ponax Ginseng C. A. ahs Laat Mitteilung.) Pharm. Zei Berlin 106: 1307, 1308, 1310, 1311. IcHmmura, T. On the anatomy of the bo a Aralia quinquefolia var. Ginseng. (In : eee Bot. Mag. Tokyo 9: 131-134. Jounson, M. C. Our most valuable native plant, Panax Specs: Cas- tanea 25: 132-134. 1960. [Popular account i the sp. and terminology applied to its parts by the mountaineers who di Maraxkvey, A. V. Ginseng in scientific and hae Air (1596-1932). (In Russian; English summary.) Vestn. Dal’nevost. Akad. Nauk SSSR (Bull. Far East. Branch Acad. Sci. USSR) 1932: ai "1932. [A digest of the literature. ] Mosic, A., & G. Scoramm. Der Arzneipflanzen- und Drogenschatz Chinas und die Bedevdune des Pén-Ts’ao Kang-Mu. Beih. Pharmazie. Heft 4. 71 pp. Berlin. 1955. [Lists constituents of roots of P. Ginseng, 53, 54 4,] PETROVSKAYA-BaRANova, T. P. An embryological investigation of ginseng. (In Russian; English summary.) Trudy Glavn. Bot. Sada Moskva 6: 211-225. 1959.* 136 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 Poyarkova, A. De Ginseng nomine botanico legitimo. (In Russian.) Not. Syst. Leningrad 12: 196, 197. 1950. [Maintains P. Schin-seng Nees is correct. Sims, J. Panax quinquefolia. Bot. Mag. 33: pl. 1333. 1811. . Panax pusilla, Ibid. pl. 1334. |P. trifolius.| peng G. A. Chinese Materia Medica. Vegetable kingdom. iv + 558 pp. + i. Shanghai. 1911. [Chinese uses of P. Ginseng, 301-304. Taxes N., & T. Osumr. On the inheritance of the colour of stem and fruit anax Ginseng, C. A. Meyer. (In Japanese.) Jap. Jour. Genet. 1 oh 276. 1940. VAN — J. S. Diseases of ginseng. Cornell Agr. Exp. Sta. Bull. 219: 163-186. Fone I. T. On the germination of Panax Ginseng C. A. M. seeds. In Russian; English summary.) Bot. Zhur. 20: 242-244. 1935. oe H. H., & J. RosenspAum. The diseases of esr and their control. Dep. Agr. Bur. Pl. Ind. Bull. 250: 1-44. pls. 1-12. 1912 a es: Ginseng. Econ. Bot. 11: 344-348. oe P. Ginseng, P. quinquefolius s.] >. . Acr. Ginseng (Panax aoe! eps culture. (Revised.) U. S. Dep. Agr. Farmers’ Bull. 1184. 13 p Wonc, Y. C. The constituents of on . of American ginseng. Jour. Am. Pharm. Assoc. 10: 431-437. 1921.* [P. quinquefolius.] pss! ig ah Present address: DEPARTMENT OF BIOLOGY VERSITY OF AKRON GRAY peexsantete AKRON, OHIO 1966] HOWARD, PLANTS OF PUERTO RICO 137 NOTES ON SOME PLANTS OF PUERTO RICO RicuHarp A. Howarp AMONG SOME GENERAL BOTANICAL COLLECTIONS made during the initial stages of a research program on the mossy or elfin forests of eastern Puerto Rico (a forest type found at the summit of several peaks in the Luquillo Mountains) are several plants worthy of special mention or requiring new names to permit their proper use in other publications. The research program on the mossy or elfin forest of higher elevations is supported by grant GB-3975 from the National Science Foundation. Special acknowledgement is also made of the hospitality and the physical assistance rendered in this program by Mr. Joseph B. Martinson. The illustrations were prepared by Miss Judy Appenzeller. Cytological infor- mation has been supplied by my colleague Dr. Lorin Nevling. Cordia wagnerorum sp. nov. Pr. I. Frutex scandens, ad 4 m. altus; folia lanceolata, irregulariter obtuseque dentata, iridescentia, scabrida cum umbonibus parvis solitariis, umbonibus singulis apice uno pilo brevi rigidoque coronatis; inflorescentia sessilis, axillaris, 3—5-flora; calyx et corolla quadripartita; stamina 3-4; fructus ruber, drupaceus. A weak shrub with trailing or scrambling branches reaching 4 m. in length. Young stems brown and densely covered with arching, ascending, stiff hairs. Petioles 8-12 mm. long, pubescent with conspicuous appressed ascending hairs and globular glandular hairs. Leaf blades lanceolate, 5.5 xX 2.5,8 & 3, to 10 & 3.5 cm. long and wide, apex short and acute, base acute to cuneate, margin undulate in the lower portion, bluntly irregularly dentate in the upper half or two-thirds; primary veins arcuate, ascending, 8-10 pairs, not conspicuously anastomosing; upper surface grayish green and iridescent when fresh, with distinct, isolated umbos each topped with a short, stout, curved and rigid trichome producing a scabrous condition, smaller curved trichomes scattered on the surface or marginal; lower sur- face lighter in color with abundant glandular exudates when dry. Inflores- cence an axillary sessile cluster of 3 to 4 flowers. Calyx ovoid, tube 2.5 mm. long, the lobes 4, greenish white, broadly triangular, 2.5 mm. long, with simple elongate thin-walled hairs and globular resinous hairs; corolla white, the tube 5~7 mm. long, the lobes 4, narrowly triangular, 1-1.5 mm. long, glabrous outside; stamens 3 or rarely 4, attached above the middle of the tube in the few flowers examined, the free portion of the filament 1 mm. long, the attached portion bearing long white-pilose pubescence, anthers oblong, 1 mm. long; ovary ovoid, 2-3 mm. in diameter, glabrous, 138 JOURNAL OF THE ARNOLD ARBORETUM [ VoL. 47 the style 2—2.5 mm. long, slightly flattened, glabrous, bifid, each portion divided nearly to the base, the stigmatic areas 1-2 mm. long. Fruit ovoid, 4—6 mm. in diameter, slightly fleshy, red, endocarp warty, style persistent, the persistent calyx surrounding basal portion of fruit; fertile seed un- known. Puerto Rico. South side of Luquillo Mountains along the El Toro trail 0.5 km. from Route 191. Collected in flower October 4, 1964, Richard J. Wagner 695 (holotype, A); collected in fruit in November, 1964, R. A. Howard 15735 (A). This species was found by Dr. and Mrs. Richard J. Wagner and is named in their honor. It is one of the many botanical collections made by the Wagners which have greatly increased our knowledge of the flora of eastern Puerto Rico. The closest relative of Cordia wagnerorum is C. bellonis Urban, a species of the Maricao area of western Puerto Rico. Cordia bellonis, described by its recent collectors as a small compact shrub, is smaller in all its parts than C. wagnerorum and is distinguished from it by the color of its leaves and the uniform distribution of its dense indument. In discussing Cordia beilonis, Britton noted that it is “a peculiar species, not closely related to the others, perhaps generically distinct” and trans- ferred it to the genus Varronia. This species is distinguished partly by its glomerulate axillary inflorescence, a character shared by Cordia wag- nerorum and not known in other species of Puerto Rico or the West Indies. Although only a very few plants of Cordia wagnerorum have been found in the single location cited, the search continues for additional material. It is hoped that cytological data and additional information about the morphology of the fruit may also be obtained. All fruits examined up to the present contain only a single aborted seed in a single large locule, a condition similar to that noted by Urban in the original description of Cordia bellonis. Not one of the few fruits planted germinated; however, a small number of seedlings were found near the type locality of Cordia wagnerorum. It should be mentioned that one wall of the endocarp is abnormally thicker than the others which may result from the inclusion of a vascular canal or a reduced cavity representing locules. The characteris- tics of the fruit and the nature of the inflorescence seem to be significant characters to distinguish the two species discussed here from others assigned to the genus Cordia. Solanum woodburyi sp. nov. Pr. EL. Frutex virgatus, ad 3 m. altus; caules, axes inflorescentiarum, perian- thium et folia abaxialiter copiose stellato-pubescentia; inflorescentia cymosa; corolla purpurea; stamina subsessilia, antherae elongatae. Wand-like shrub to 3 m. tall, stems 1-2 cm. in diameter at breast height, the young stems, petioles, lower leaf surface, the midrib above, inflores- cence, and outer surfaces of the perianth densely covered with golden multi-armed, spherical, stellate hair clusters. Petioles 1.5 cm. long, taper- ing to the base of the blade. Leaf blades oblong to elliptic, 5.5 2.5 to 1966] HOWARD, PLANTS OF PUERTO RICO 139 10 4 cm. long and wide, apex obtuse and often short mucronate, base rounded, margin entire, slightly recurved, primary veins 7 pairs, arcuate and conspicuously anastomosing within the margin, upper surface dark green, glabrous except for the depressed midrib which is stellate-pubes- cent; prickles 8-10 mm. long at maturity, golden, few, mostly 2 at apex of petiole and occasionally 1 or 2 on the midrib near the base of the blade on the lower surface only. Inflorescence axillary, cymose, 2-4 cm. long, usually less than 10-flowered, axes densely stellate-pubescent. Calyx densely stellate, 6 mm. in diameter, shallowly lobed; corolla rich purple when fresh, drying bluish purple or blue, tube 1 mm. long or less, the lobes 5, ovate-lanceolate, to 8 mm. long and 3.5 mm. wide at the base, bluntly acute at the apex, stellate-pubescent outside, glabrous inside; stamens 5, subsessile or the free portion of the filament less than 0.5 mm. long, glabrous, anthers oblong-ovate, broadest at the base but only slightly tapering above, 4.5—5 mm. long, obliquely poricidal and slightly cleft towards the outside at the apex; ovary glabrous, ovoid, 2 mm. in diameter, style terminal, stout, glabrous, 5 mm. long, stigma capitate; fruiting in- florescence erect, stout, berry depressed-globose, green when immature, 2 cm. in diameter and 1.5 cm. long, glabrous. erto Rico. South side of Luquillo Mountains along the El Toro trail 2 kms. from Route 191. Collected in flower and fruit, October 13, 1964, R. A. Howard & George Taylor 15692 (holotype A). This handsome plant is named for Mr. Roy Woodbury who has so gen- erously shared his knowledge of the flora of Puerto Rico. Although Wood- bury showed us the small and isolated population of this new species in 1964, an unidentified specimen of it is in the herbarium of the New York Botanical Garden, having been collected by W. R. Barbour along the El Toro trail on March 18, 1935 This species is clearly distinct from any previously described Solanum of the West Indies in its virgate habit, woody structure, density of stellate pubescence, rich purple flowers, and golden prickles. The type locality of Solanum woodburyi has been visited frequently by Woodbury and others yet only one fruit has been found. This fruit is associated with the holotype and, regrettably, it is immature. The lack of fruit suggests some type of developmental anomaly as yet undetermined. Pollen stainability, with cotton blue in lactophenol, was 88.5 per cent based on 614 grains. Meiotic cell division appears normal and Nevling has obtained a chromosome count of m = 12 from pollen mother cells of buds collected on April 1, 1965. The voucher specimen for this count is Howard & Nevling 15747 (a). Psychotria guadalupensis (DC.) comb. nov. Loranthus guadalupensis DC. Prodr. 4: 294. 1830 Viscoides pendulum Jacq. Select. Amer. 73. t. 51. f. 1. 1763. Psych te parasitica Sw. Prodr. 44. 1788; Fl. Ind. ‘one 1: 408. 1797, nomen illeg: 140 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 Loranthus portoricensis DC. Prodr. 4: 293. 1830, not Psychotria portoricensis DC. Prodr. 4: 515. 1830 Uragoga “ap ait Baill. Adansonia 12: 227, 328. 1876-9; Urb. Engl. Bot. Jahrb. 24: 7 Mapouria a K. Sch. in Engler, Pflanzenfam. 4: 112. en Uragoga parasitica Maza, Anal. Soc. Esp. Nat. II. ser. 3: 293. 18 Psychotria pendula Urb. Symb. Antill. 1: 445. 1900, non ae Fl. Brit. Ind. 3: 164. Psychotria pendula subsp. tetrapyrena Urb. Symb. Antill. 1: 445. 1900. Psychotria pendula subsp. grosourdyana Urb. ibid. Psychotria pendula subsp. genuina Urb. ibid. 446. Psychotria pendula subsp. pachyphylla Urb. ibid. 447. Psychotria pendula subsp. grenadensis Urb. ibid. Psychotria pendula subsp. trinitensis Urb. ibid. Psychotria grosourdyana (Baill.) Urb. op. cit. 4: 596. 1911. Psychotria tetrapyrena (Urb.) Urb. op. cit. 8: 677. 1921. This epiphytic but weak shrubby plant assigned to the genus Psychotria, sensu latu, is common in the Greater Antilles and abundant in the Lesser Antilles. The earliest name applied to the plant is Viscoides pendulum Jacquin. In 1788 Olaf Swartz described Psychotria parasitica and cited in synonymy Viscoides pendulum. The Swartz name, although used in many floras, must be rejected as illegitimate. In 1900 Urban published the name Psychotria pendula presumably basing it on the earlier Jacquin epithet (see subsp. genuina) but he stated clearly “non Hook. f. Fl. Brit. Ind. III p. 164”. The name Psychotria pendula (Jacq.) Urb. is also widely used in floras of tropical America but it is illegitimate, being a later homonym of the Hooker name Psychotria pendula which was applied to a plant from the Andaman Islands. The earliest name available for this species is Loranthus guadalupensis DC. which is based on a Bertero specimen. The type is in the Prodromus herbarium in Geneva, Switzerland. The correct name is Psychotria guada- lupensis (DC.) Howard. In establishing Psychotria pendula Urban noted it was polymorphic and he described six subspecies including a subspecies called genuina. These minor taxa are based primarily on differences in leaf size and shape and on the number of pyrenes in each fruit. Three of the subspecies are found on single islands of the West Indies and three on groups of islands. Urban’s subsp. genuina had a geographic range from Cuba to Grenada. In subsequent issues of his Symbolae Antillanae Urban elevated two of the subspecies to specific rank as Psychotria tetrapyrena and P. grosourdyana without explanation or comment. I have seen this plant on every island of the West Indies from which Urban recorded it and on a few for which he had no records. There appear to be two growth forms represented by herbarium specimens which can be recognized in the field and which are often closely associated. The plants may have stout upright branches in their epiphytic habitat and 1966 | HOWARD, PLANTS OF PUERTO RICO 141 leaves with thick blades and short petioles. Other plants have a few to all branches weak and drooping, tending to be thinner and more delicate and leaves thinner in texture with narrowed bases and more elongate petioles. It is difficult to conclude that two species would be growing in such close proximity often intermingled in the crotch of a single tree, yet it is impossible to be certain that only one plant is involved in many such collections. The variation in leaf thickness is great in fresh condi- tion and is often accentuated in drying. The thicker leaves tend to dry more slowly and therefore turn dark brown or black in the process. Variation in peduncle color is mentioned by Urban as an accessory characteristic. In living specimens seen in the field this color varies from deep red through bright red to pale green. A collection from the summit of Cerro de la Punta in the center of Puerto Rico (Howard & Nevling 15417) has peduncles clear white in color. This albino phase is the extreme variation Urban places —— _emphasis on the number of pyrenes produced by throughout the range revealed quite unexpectedly that these plants are heterostylous, a fact not previously reported for the species. Further, the pollen proved to be of three types which will be described in a later paper. There was a correlation between the number of stigma lobes and the number of locules in the ovary; however, a variation was noted in the number of sepals. Recently it has been possible to examine a large number of plants in Puerto Rico, for the species is common in the Luquillo Mountains, occurring in wet but open forests and in the mossy or elfin forest zone. One flower was taken from each plant, and the number of sepals and stigmas recorded. In the following tabulation the first number is of calyx points (sepals); the second, the number of stigma lobes; and the third, the number of examples seen. 7-3-1 5-3-41 6—4~-2 5—2-9 6-3-3 4-4-2 6—2~1 4-3-6 5—4—7 4-2-2 Britton and Wilson (Flora Puerto Rico and Virgin Islands 6: 244. 1925) use the name Psychotria grosourdyana for this plant saying it is “parasitic on forest trees” and describe it as having a fruit containing “3 pyrenes.” In an actual check of the mature fruit the calyx of the soft black multistoned drupe was examined and then the drupe squashed be- tween fingers to separate and count the number of fertile (the hard brown) Ppyrenes. White pyrenes were regarded as sterile since they did not contain seeds. Again, using one mature fruit from each plant the following count was obtained representing, respectively, number of calyx lobes, number of 142 JOURNAL OF THE ARNOLD ARBORETUM [VoL. 47 pyrenes, the number of mature and fertile pyrenes, and the number of examples seen. 5—4—3-2 5—2-1-1 6—4-3-3 5—4—2-2 4—4-2-1 6—3—3-1 5—3—3-14 4—3-3-2 6—3-1-1 5—3—2-19 4—3-2-1 5—3-1-15 5—3—-0-2 It is clear from this tabulation of plants growing in eastern Puerto Rico that the variation in number of pyrenes removes any value the character might have for purposes of classification. Representative recent collections of Psychotria guadalupensis (DC.) Howard from eastern Puerto Rico are Wagner 81, Howard & Nevling 15749 and 15963, all of which have been distributed widely to major herbaria. Curculigo capitulata (Lour.) O. Ktze. Rev. Gen. Pl. 2: 703. 1891. Leucojum capitulatum Loureiro, Fl. Cochinchin. 199. 1790. Curculigo recurvata | Dryander in] Aiton, Hort. Kew. ed. 2. 2: 253. Molineria recurvata Herbert, Amaryll. 84. 1837; Brackett, eeaue ce 161. 23 Molineria hortensis Britton in Fl. Puerto Rico and Virgin Islands 5: 161. 1924. Britton described Molineria hortensis as a new species similar to Molineria recurvata and based it on a specimen cultivated in a garden near La Muda. The collection, Britton 7892, is preserved at the New York Botanical Garden. It was collected in flowering condition in March, 1923. Britton stated the species was cultivated in “Porto Rico gardens” and that its origin was unknown but it probably came from tropical India or Malaya. In the New York Botanical Garden herbarium there is also an older specimen collected by J. A. Stevenson (1891) at Rio Piedras (Agri- cultural Experiment Station) on April 30, 1914. A hand written annota- tion label by N. E. Brown is attached to this sheet and states “Curculigo sp. (not C. recurvata). I have looked through the whole of the Kew amet twice and through the collection of drawings and cannot match this with anything.” This annotation appears to be the sole basis for Britton’s conclusion that the plant represented a new species. The pub- lished description in no way differs from one of “Curculigo recurvata.” Regrettably, the illustrations which have been published and were available at the time of Brown’s search vary greatly in their accuracy and a comparison of Stevenson’s specimen or that collected by Britton with certain illustrations could easily lead to the conclusion drawn. Today even the status of the genus Molineria is uncertain. It was recognized by Herbert and by Baker, early authorities on the Amaryllidaceae, and by Pax and Hoffman in their treatment of the family for Die Pflanzenfamilien 1966 | HOWARD, PLANTS OF PUERTO RICO 143 (ed. 2). More recently Brackett indicated Molineria as a distinct genus and so did Neal in her book, Jn Hawaiian Gardens, published in 1965. By contrast most recent floras, horticultural encyclopedias and taxonomic works place Molineria in the synonymy of Curculigo. In describing the genus Molineria Colla (Hortus Ripulensis, Append. 2. 331. ¢. 18. 1825) noted that the anthers in the species he considered were fused. Later Herbert (Amaryll. 84. 1837) questioned the monadelphous character of the anthers and, surprisingly, this condition has not been further discussed in any recent study yet encountered. The several authors who have accepted Molineria as distinct from Curculigo all adopt different characteristics as generically significant and a further study of the broad- leaved and petioled species is needed. If the fusion of the anthers is a character of generic value in the Amaryllidaceae, and it may be in related families, then the material Britton described is properly placed in Molineria. The collection, Britton 7892, does have the anthers fused. Collections made in the Luquillo Mountains by Dr. and Mrs. Wagner, e.g. 486, 567, also have fused anthers and the consistency of this character has been checked with an abundance of living materials from this population. No mention is made of the fusion of anthers in the description of Curculigo recurvata (by Dryander in Hortus Kewensis), although the character is clearly shown and is men- tioned in Edwards Botanical Register (9: pl. 770. 1823) where the description is credited to Roxburgh “Mss.” Merrill appears to be the first to accept the application of the Loureiro name to this complex and has repeated the synonymy given above. The syngenesious character is not mentioned by Loureiro in his description of the basionym Leucojum capitulatum, yet if Molineria is to be recognized as the proper genus a new combination is needed for the specific name. In any case Molineria hortensis Britton is not a new species but a synonym of Curculigo capitu- lata (Lour.) Ktze. as currently accepted. Conostegia hotteana Urban & Ekman Ark. Bot. 22A. 17: 29. 1928. A single plant identified as this species and verified by Dr. John Wurdack was collected by Dr. and Mrs. Wagner, numbers 678 and 843. It was growing at km. 21.3 on route 191 on the south side of the Luquillo Experi- mental Forest. The plant is a small tree, 5 meters tall with a trunk diameter of 15 cm. at breast height. The flowers are 2.5 cm. in diameter and the white petals are characteristically broadly cuneate. As the specific name indicates, Conostegia hotteana was described from material collected July 12, 1928, by Erik Ekman, between Dame-Marie and Montagnac in the Massif de la Hotte in Haiti. The exact location is at the extreme western end of the southern peninsula of Hispaniola. The present collection is not only the first record of the species since the original collection but also the first record of the genus in Puerto Rico. These disjunct localities are not represented in any other genus or species occurring in either Puerto Rico or Haiti. Although the Puerto Rican plant 144 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 is but a few yards from the roadside, on the steep slopes of a river bank and within an experimental forest it is not a planted specimen. There is neither a reason nor a record of the deliberate introduction of this non- economic plant to Puerto Rico. The location of the original collection in Haiti is not one commonly visited by tourists and so does not suggest a chance introduction. The occurrence of this species in Puerto Rico and in Haiti defies immediate explanation. EXPLANATION OF PLATES PLATE I Cordia wagnerorum Howard. Fics. a-e from Wagner 695 (holotype); FIG. e ioe Howard 15735, a, habit, X 1/2; b, portion of upper leaf surface, showing umbo, * 50; ¢, seedling, > 1/4; s external view of flower, X 4; e, longitudinal section of flower, x 4; f, fruit, PLATE II Solanum woodburyi Howard. All Fics. from Howard & Taylor 15692 (holo- type). a, habit, X 1/2; b, trichome showing stellate character, X 20; view of flower, ba 2: , portion of inflorescence showing buds and mature flower, X 2; e, anther, xX 5; f, immature fruit, x 1; g, incase at first meiotic division, m= 12. Se Jour. ARNOLD Arp. VOL. 47 Pirate I Corda Wagner orum Howard 2 RS 3 8 LR 3°, % 4 Howarb, PLANts OF PUERTO Rico Jour. ARNOLD Ars. VOL. 47 Pirate II Solanum Woodbury! Howard Howarb, Pants or Puerto Rico 1966 | SCHUBERT, STUDIES IN DIOSCOREA, I 147 STUDIES IN DIOSCOREA, I: A COLLECTION FROM BRITISH HONDURAS BERNICE G. SCHUBERT ! IN 1924, WHEN KNUTH PUBLISHED his monograph of the Dioscoreaceae, he maintained in the family nine genera and 654 species, of which 614 were members of the genus Dioscorea. It was not until about 20 years after the publication of Knuth’s work that serious interest in Dioscorea ex- tended beyond the purely botanical aspects or the cultivation of the few species used in tropical areas for food. However, with the discovery of the presence of steroidal sapogenins in the underground parts of many the world. The botanical results of this sudden acquisition of both mate- rials and data for study have been very rewarding and should, with suffi- cient time, provide a much better understanding of this interesting genus than has been possible previously. The difficulties met in studying species of Dioscorea are to be expected in a genus composed almost entirely of dioecious twining plants with extensive, sometimes fragile, usually subterranean rhizomes or tubers. Collection of material is difficult because the stems are most often en- tangled with stems of other genera and are difficult to trace to the proper base; the stem may break off at the base or the rhizome itself may break while it is being dug. Plants of both sexes usually grow rather close to- gether, but both sexes cannot always be found; as a result, species are often described from one sex only, Although it is possible to determine the relationship of a plant from flowers of only one sex (especially the stam‘nate) there are at least several, if not many, examples of two species being described, one from staminate the other from pistillate material, which were, in fact, the same. In spite of these difficulties, however, many recent collections made with great skill and care (and including both My taxonomic studies on the genus Dioscorea were begun na I was esi . &. Depa 0 J leaving there at the Arnold Arboretum - Har vard Unive rsity. I am sa to be able kn of my former and present colleagues. The illustrations were drawn with great skill and care by Mrs. Regina O. Hughes, Crops Research Division, U. S. Department of Agriculture. 148 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 photographs and detailed notes) have helped or aaa in clarifying this collection is of such interest that it seems appropriate to discuss some of the significant numbers, to provide more complete descriptions for two previously inadequately known species, and to describe one which is new. I shall conclude the paper with a list of determinations of all the identifi- able Bampton numbers in the hope that this may provide the basis for an accurate account of the species of Dioscorea in British Honduras. I appreciate the interest and consideration of the officers of many insti- tutions who have made available collections in their care for my study, either on loan or when I visited their herbaria. Full representation of any one species may be so inadequate in one herbarium that only a combined collection in a genus such as Dioscorea can help to give a clear understand- ing of the material. The institutions concerned are cited in the paper by the symbols given in Index Herbariorum (Lanjouw and Stafleu, Reg. Veg. 31. 1964) B Botanisches Museum, Berlin-Dahlem, Germany F Chicago Natural History Museum, Chicago, Illinois, U.S. A. GH Gray Herbarium of Harvard University, Cambridge, Massachusetts, U.S. As K The Herbarium, Royal Botanic Gardens, Kew, Richmond, Surrey, Great Britain LE Herbarium of the Komarov Botanical Institute of the Academy of Sciences of the U.S. S. R. LL Lundell Herbarium, Texas Research Foundation, Renner, Texas, U. S.A. MEXU Herbario Nacional del scorvibae de Biologia, UNAM, Ciudad Uni- versitaria, México, D. F., Méx MICH University Herbarium, pee of Michigan, Ann Arbor, Michigan, Uz S.A, MO Missouri Botanical Garden, St. Louis, Missouri, U. S. A. NA U.S. National Arboretum, Washington, D. C., U.S. A. NY New York Botanical Garden, Bronx Park, New York, U. S. A. US Department of Botany, U. S. National Museum, Smithsonian Institu- tion, Washington, D. C., U.S. A. Dioscorea hondurensis Knuth, Repert. Sp. Nov. 38: 120. 1935. Holo- type, Peck 549 9 (B); isotypes (GH, K). Fics. 1-3. D. belizensis Lundell, Contr. Univ. Mich. Herb. 6: 5. 1941. Holotype, Gentle 2998 3 (MICH); isotypes (F, GH, US). 1966 | SCHUBERT, STUDIES IN DIOSCOREA, I 149 D. tabascana Matuda, Bol. Soc. Bot. Méx. 21: 1. 2 figs. 1957. Holotype, Gilly, Hernandez & Berlin 44a 6 (us); photo. and fragment (MExv). Dextrorse vine. Rhizome conical at apex, becoming horizontal about 3 dm. below surface, cylindrical, little branched, with creamy to yellow or almost orange flesh, 1-4 m. long, 1-5 cm. in diameter, 5.9-65.9 kg. in weight. Stem stout, reddish brown (in age becoming dark brown or gray at base, the bark sloughing off), cylindrical, striate, somewhat pilose or with only glandular trichome bases remaining, to 5 mm. in diameter. Leaves alternate, mostly coriaceous, cordate or ovate to ovate-suborbicu- lar, abruptly acuminate into a short-attenuate tip, nearly truncate at base or more often with a rather deep inverted U-shaped to inverted V-shaped sinus, or sinus very narrow and the basal lobes overlapping; 9-nerved, the two outer nerves bifurcate, glabrous above with veins impressed, below with veins prominent and abundantly pilose with multicellular, branched or fasciculate trichomes (0.6—-1 mm. long), gland-dotted throughout, to 26 cm. long, 23 cm. broad; petioles slightly winged above with decurrent leaf tissue, sulcate, somewhat angulate, striate, swollen at apex and base (in age becoming contorted at base), 4% to scarcely ™% length of leaf racemes up to 1.5 m. long, rachis essentially glabrous, angulate, minutely serrulate-alate on the angles, secondary rachises similar; flowers solitary, bibracteate, pedicellate, the outer bract ovate-acuminate, to 2 mm. long, m. wide, the inner ovate, abruptly attenuate, 1.5 mm. long, 1 mm. wide, pedicels 1.5 mm. long, scarcely differentiated from the 1.5-2 mm. long perianth tube; tepals fleshy to papyraceous (in age), ovate-obtuse, ca. 1.5 mm. long and wide, narrowed to a claw 0.8 mm. wide at base; stamens 6, introrse, inserted at base of tube and slightly exceeding it in length, becoming conspicuous as tepals become reflexed, 0.5—0.6 mm long; stylar rudiment central, 3-parted, rather conspicuous, to 0.5 mm. high. Inflorescence of 2 plant a solitary, axillary, densely flowered raceme to 4.5 dm. long, rachis gland dotted, angulate, striate, essentially glabrous; flowers bibracteate, bracts gland dotted, the outer ovate-attenuate, 3 mm. long, 1.5 mm. wide, inner ovate-acuminate, 2 mm. long, 1 mm. wide; tepals fleshy, somewhat gland dotted, ovate-obtuse, early reflexed, 2.5 mm. long, 1.5~2 mm. wide, tube to 0.6 mm. long; staminodia 6, inserted at base of tube, introrse and somewhat antheriferous; styles 3, fused below, becoming distinct and bifid above with the stigmatic surface extrorse; Ovary (excluding neck) to 7 mm. long. Capsule broadest above the basal third, its wings obtusely oblong, surface glabrous and remotely gland dotted at maturity, 4.8-5.5 cm. wide, 3.6—4 cm. high, on a pedicel 6 mm. long; seeds suborbicular, or throughout, to 2.2 cm. long, 1.8 cm. wide (including wing). DISTRIBUTION: Southern Mexico (Tabasco), British Honduras, Guate- mala es Rica, where probably cultivated] to Panama. Mexico. Tabasco: between the Escuela Agricola and Cuevas de Cocoma, Teapa, Gilly, Herndndez & Berlin 44a 8 (mMEXxU, US); between Escuela Agricola 150 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. Museum botanicum Berolinense Acenent (3 Ho. 5 ¥7 Britiseh-Henduras. iog_M. E. Peck. 1905-07 Lec EE Peck, OE | Fic. 1. Dioscorea hondurensis Knuth. Photograph of type (B). and Gruta de Cocomé, along trail to Gruta, Gilly, Hernéndez & Berlin 7129 é (NA). British Honduras. Stann Creek District, in acahual. Carib Reserve, Gentle 2998 3 (¥, GH, MICH, US). Toledo District, Monkey River, Gentle 3728 & (¥, GH) unta Gorda, five-year old Pinus caribaea plantation at Machaca Forest Station, alt. 100 ft., Bampton 22 [sterile] (x). Upper Moho River, “Yr 1966 | SCHUBERT, STUDIES IN DIOSCOREA, I 151 Peck 549 2 (B, GH, K). Guatemala. Dept. Alta Verapaz: Panzas, alt. + 400 ft., Gentry 12468 [sterile] (NA); 4-6 miles west of Panzas, alt. + 500 ft., Gentry 12476, 12477 [sterile] (NA); 12 miles west of Estor, alt. + 200 ft., Gentry 12475 [sterile] (NA). Dept. Izabal: brushy slope, alt. 75-225 m., Standley 24112 [sterile] (GH, Ny, US). Honduras. Dept. Atlantida: wet forest, Lancetilla Val- ley, near Tela, alt. 20-600 m., Standley 53857 2 (¥, us); idem, Standley 55218 é (, us); Lancetilla Field Station, alt. + 100 ft., reddish porous clay soil, Gentry 12493 [sterile] (NA); about 3 miles w. of Ceiba, in sandy loam, alt. 20 ft., Gentry 12502 [sterile] (NA). Costa Rica [probably cultivated]. Prov. Alajuela: llanuras de San Carlos, alt. 500 ft. (received from Experimental Plan- tations Inc., Chicacao, Guatemala, where probably also cultivated), Krukoff 2012 (¥ [sterile], us [1 coll. sterile, 2024529 9]). Panama. Canal Zone: Barro Colorado Island, Silvestre Aviles 12 2 In 1935 Knuth described Dioscorea hondurensis, of British Honduras, from pistillate material only (Fic. 1), assigning it to a new section, TrI- ANGULARES. He did not indicate to which subgenus these new taxa might Fic. 2. Pan ee aie: Portions - nthe the lowermost showing base of leafy stem emerging, from Bampton 22, X 3 be referred. Morton, in 1936, suggested that the section might “belong to Eudioscorea,” but material was not available to confirm this. Recent examination of the holotype from Berlin and the isotype from the Gray Herbarium confirms Morton’s suggestion. From 1935 until now no stami- nate material has been associated with Knuth’s species. However, the pistillate flowers bear six equal see ag indicating that the staminate flowers probably have six equal s n 1941, Lundell described Vhs belizensis, of British Honduras, from staminate material, suggesting no relationship, but describing the 152 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 AM so Le a-j, Dioscorea hondurensis Knuth. a, portion of stem with leaf and bracts, from same specimen, ; d, staminate flower opened to show the 6 introrse stamens and 3-parted rudimentary style, from same specimen, X 3; e, 1966 | SCHUBERT, STUDIES IN DIOSCOREA, I 153 material very fully. The species was characterized by its large leaves pubescent below, and by its ion long (to 150 cm.), branched, stami- nate inflorescences. The staminate flowers are solitary, dark red, and with six stamens. In 1952, yeni identified Krukoff 2012 from oh Rica (where it was probably cultivated) as D. belizensis. This was the first material other than the type, and the first pistillate collection, as- signed to the species. In 1957, Matuda described Dioscorea tabascana, from the state of Tabasco, Mexico, from staminate material. This species was also char- genus Euptoscorra, § MacrocyNopium, distinguishing it from the other members by its solitary flowers and extrorse stamens. The form of the inflorescence cannot be reconciled, however, in a section characterized by densely flowered scorpioid or cymosulous branchlets in a compound raceme (§$ Macrocynopium). The chief difference between the staminate flowers in the type collection of Dioscorea belizensis and those of the type of D. tabascana is in the texture of the tepals. The tepals of D. belizensis are thick, fleshy, and with glands only remotely visible; those of D. tabascana are papyriferous and abundantly gland dotted. This variation seems to me to be a function of age rather than a morphological differentiation. With more collections the problem can be clarified. From close and careful examination of the type material of Dioscorea hondurensis Knuth, D. belizensis Lundell, and D. tabascana Matuda, I have come to the conclusion that they are all one species (Fic. 3). There is not even sufficient variation in the collections to distinguish geographical varieties. The name to be maintained, therefore, is the oldest one, Dios- corea hondurensis Knuth, For the present it seems expedient to maintain also Knuth’s § TrraNGULARES, but I think it will be possible to show, after examination of more material of other species, that D. hondurensis is a perfectly characteristic member of § SARCANTHA Uline, up to now thought to be limited to South America. In the thirty years it has taken to associate staminate with pistillate elements of Dioscorea hondurensis a good representation of material from a rather wide geographical range has accumulated. It may be seen in the case of D. hondurensis, as will be shown for other species also, that with staminate flower opened to show the 6 introrse stamens and 3-parted rudimen- tary eit < 3, from Gilly, Herndéndez & Berlin 44a vs), holotype of D. tabascana; f, portion of staminate inflorescence to show so litary, ante Le pei a flower s, X 3, from — . (us 1808657); g, portion of leafy stem and pistillate nada ith 1 mature eee ale, xX 1/3, from M. E. P 549 (B); h, portion of — chacoenee «2, ah same specimen; i, pistillate flower to show and 6 sta minodia, x 3, from same specimen; j, mature seed, X 3, rie Krehof 2012 (us 2024529). 154 JOURNAL OF THE ARNOLD ARBORETUM [VOL. 47 intensive collecting many species formerly considered very local are, in act, rather widespread. Although some material of D. hondurensis is sterile, it can be placed with relatively little doubt because of the char- acteristic aise Spe on the lower leaf surface (Fic. 3b). Such pubescence is not known to occur on plants of other species with leaves of similar size, seek and outline. Ww hen collectors have given notes on underground parts of Dzoscorea hondurensis their information is remarkably uniform. For photographs (Fic. 2) indicating the general aspect of the rhizomes I am indebted to Dr. Bampton. Sketches sent to me by Dr. Gentry are similar. Dioscorea (§ Higinbothamia) gaumeri Knuth, Notizbl. Bot. Gart. Ber- lin 7: 199. 1917 [type cited from Colombia]; Pflanzenreich IV. 43 (heft 87): 122. 1924 [same type cited from Yucatan]. gee Gaumer, Pl. Yucat., sub Higinbothamia 2?! (B). Fics. 4, 5a—d. se So ohne synandra Uline, Publ. Field Mus. Bot. 1: 415. pl. 22. 1899. Lectotype, Gaumer 794 p.p. 6 (F 36597). Popa: synandra (Uline) Stand]. Publ. Field Mus. Bot. 3: 231. 1930, non Dioscorea synandra Uline, Nat. Pflanzenfam. Nachtrage zu II: 86. 1899, men nudum; Uline ex Knuth, Pflanzenreich IV. 43(heft 87): 219. 1924. Lectotype to be chosen from among the syntypes, since Knuth cited both a $ anda @ type and neither collection has been examined. Sinistrorse vine.” Rhizome more or less cylindrical ca. 3% cm. long and 1 cm. in diameter, becoming subconical at apex [not branched in photo- graph of specimen seen], growing near surface, flesh whitish brown. Stem Fic. 4. Dioscorea gaumeri Knuth. Portion of rhizome with base of emerging stem, from Bampton 10, X 1/2 (photo. Bampton). slender, not ridged when fresh, becoming finely ridged and grooved, to 10 feet or more long (or high) [fide Bampton], tawny to yellowish, glabrous, 1-1.5 mm. in diameter. Leaves alternate, thin, glabrous, cordate-ovate, narrowed to a long-attenuate apex, basal sinus broad, inverted V- or U-shaped, 5-11.5 cm. long, 3.5-11 cm. wide, 5-nerved, the two outer * Dioscorea gaumeri was described as dextrorse by Knuth, but observations on material at hand agree with those of Uline and Bampton that the plants are sinistrorse. 1966 | SCHUBERT, STUDIES IN DIOSCOREA, I iss nerves bifurcate; petioles sulcate on adaxial surface, ridged and grooved otherwise, glabrous throughout, 1.5-3 cm. long, contorted at base. In- florescence of ¢ plant composed of solitary axillary racemes or a basally branched panicle of racemes 10-20 cm. long, rachis minutely and often densely serrulate-alate on the fine ridges; flowers usually solitary (but sometimes two on a short peduncle (Lundell & Lundell 7918), bracteate, pedicellate, each (if flowers solitary) subtended by two lance- to oblong- attenuate bracts moderately to densely papillate and somewhat glandular on the adaxial surface, to about 1.5 mm. long and 0.5 mm. wide (if 2 flowers borne together then each unibracteate); pedicels also serrulate-papillate, glandular and + angulate, to about 3 mm. long; the 6 tepals purple even when fresh | Bampton], free from slightly above base, in 2 series, elliptic- ovate, acutish to rounded at apex, the inner 3 mm. long, 1.5—2 mm. wide, the outer similar or slightly longer and narrower, but not sharply differenti- ated; stamens 3, filaments fused, anthers extrorse at the apex of the stami- nal column (to 1 mm. high), staminodia alternating with stamens in an inner whorl and usually at least slightly exceeding them. Inflorescence of @ plant lax, elongate, simple, axillary and usually solitary racemes, often becoming more than 7 dm. long; rachis + angulate, serrulate- papillate to remotely alate; flowers solitary, bibracteate, bracts papillate- granular, the inner ca. 1.6 mm. long, the outer a little longer, lance-attenu- ate from a clasping base; tube very shallow, ca. 1 mm., tepals 6, the 3 inner ovate-subacuminate, the 3 outer ovate-obtuse, or all subacute, essen- tially equal, somewhat gland dotted on adaxial surface, papillate all over abaxial surface and somewhat puberulent, ca. 3 mm. long and 2 mm. wide; staminodia 3, borne at base of stylar column opposite the outer tepals; styles fused, column gland dotted, ca. 1 mm. long, stigmas 3, slightly bifid, outwardly curved; ovary, excluding neck ca. 6 mm. long. Capsule on a short stout pedicel, exalate, oblong to narrowly elliptical, tapering to acutish base and apex, 3-locular, each locule [3-]4-seeded; seeds 6 mm. long, 2.5 mm. wide (including wing), alate all around. DistriBuTIon: Yucatan and British Honduras. Mexico. Yucatan: Kaua road, Chichen Itza, Lundell & Lundell 7918 & (ws); Merida-Uxmal road, km. 40, Lundell & Lundell 8092 8 (GH, LL, US). Without locality, Gaumer & sons 1580 $ & 2 (G, LE, MO, US). British Honduras. San Andres, Gentle 1084 2 [3 seeds in cach locule} (cH); Pembroke Hall near Corozal, alt. 100 ft., Bampton 14 2 (kK); 3 miles south of Maskall, O’N eill 8385 8 (us); limestone outcrop on Mountain Pine Ridge near Augustine, Bampton 10 (xk). It may be seen from both the nomenclatural and taxonomic history of Dioscorea gaumeri that much confusion has resulted from inadequate material, mixed collections, and some misjudgements. Inadequate ma- terial, as well as misjudgement, presumably led Knuth (1917, p. 199; 1924, pp. 120, 122) to place D. gaumeri in subgenus HELMIA, section MonavdELPpHA. Morton (1936, p. 244) was the first to recognize the con- fusion and correctly identify Higinbothamia synandra Uline (Dioscorea [VOL. 47 JOURNAL OF THE ARNOLD ARBORETUM 156 Ds aes, Les: poss a Srey Se HS > es esl Se \ he ae se” cM, HANA \ 1966] SCHUBERT, STUDIES IN DIOSCOREA, I 157 synandra (Uline) Standl., not Knuth) with D. gaumeri, which he then reassigned to subgenus EuproscorEA (now, according to the Code, to be known as subg. DioscorEA). At the same time Morton united the genus Higinbothamia with Dioscorea as a section, under the correct subgenus. This seems to be a proper disposition of Higinbothamia, although some few botanists still propose to maintain it as a distinct genus. The larger number of ovules in each locule of the ovary is scarcely a character of sufficient significance for generic distinction (particularly since the number varies from three to four) when all characters of flowers and fruit are so clearly those of Dioscorea. After Uline’s publication of Higinbothamia synandra in 1899, it was discovered that several Gaumer collections were mixed and some of the numbers cited as representing Higinbothamia synandra actually repre- sented at least two other species. Knuth noted (1924, p. 327) that Gaumer 794 the “staminate type” was a mixed fe and illustrated portions of both elements of the collection in his fig. 63 (p. 326). Standley, in his Flora of Yucatan (1930, p. 158), discussed the Gaumer collections and some of the difficulty in organizing them. Surely the difficulties must have been greater with specimens of Dioscorea than with material of many other genera because Dioscorea species cannot be easily distin- guished by superficial characters, and the dioecism of the plants leaves greater margin for error. Caution should be used in annotating any collec- tion bearing the same number as that cited here. Although Uline expressed some uncertainty about which were generic and which specific characters when he described the genus Higinbothamia and its one species, H. synandra, it now seems reasonably clear that possession of a staminal column is a sectional character, the position of stamens on it and their relation to the staminodia are specific, as are the shape and differentiation of the tepals. On the basis of these conclusions the material collected by Dr. Bampton as his number 25 seems to be a new species in section Higinbothamia. It may be distinguished from Dioscorea gaumeri by the characters noted in the key which follows. Leaves with broad basal sinus, inflorescence a simple raceme, flowers pedicel- late; tepals all alike, obtuse; stamens exceeded by staminodia.......... Pena re cyee se Beanie Maun Ge earn Mc evan ort aceite Seer s y's tines gaumeri. Leaves with narrow basal sinus or the lobes overlapping; inflores cence a some- what zigzag raceme, flowers almost sessile; tepals differentiated into outer Fic a—d, Dioscorea gaumeri cena (a-c from Bampton 14 (xk); d from Lunda & Lundell 8092 (GH).) a, portion of 2 plant with both young flowers and mature capsules, x 1/3; b, cisulate Gone and sims mature ovary opened to show loculi and seeds, X 3; c, mature seed, X 3; d, portion of androecium to show stamens and staminodi ae 4 Fic. 5. e and f, Dioscorea " sandwithii Schubert, from pana nese grown from rhizome of Bampton 25 (xk); e, leafy stem and staminate inflor cences, X 1/3; f, portion of staminate bniscubseseiee to show digetetitiation “of 2 series of tepals and androecial structure, X 3 158 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 and inner series, acuminate; stamens exceeding staminodia............... Date ee AN rc ek iyie spi x4, sonia ie pipe Ny Son Dioscorea sandwithii. Dioscorea sandwithii Schubert, sp. nov. Fic. 5e and f. Volubilis, sinistrorsum torta, caulibus angulatis subtiliter canaliculatis papillatis canis vel flaventibus; foliis basi sinu angusto vel lobis attingenti- bus, 9-nerviis, nervis extremis bifurcatis. Inflorescentiae ¢ axillares soli- sparsim glandularibus; pedicellis ca. 0.8 mm. longis, tubo perianthii similibus + equilongis. Perianthium 6-partitum, carnosum atro-pur- pureum, segmentis exterioribus lanceo-obtusis vel -acutis, interioribus ovato-obtusis. Stamina 3, filamentis in columnam centralem fusis, antheris extrorsis apicalibus, staminodiis 3, cum antheris longioribus alternantibus. Planta @ incognita Sinistrorse vine, rhizome similar to that of D. gaumeri, stem angulate, finely grooved and papillate, grayish to yellowish, 1.5 mm. in diameter. Leaves alternate, almost orbicular, apex narrowly acuminate, basal sinus narrow or the lobes overlapping, 11 (or more) cm. long and 9 cm. wide, 9-nerved, the two outermost nerves bifurcate, surfaces essentially glabrous, somewhat papillate below; petioles eilions when fresh “becoming 3-4 times diameter of stem” to 4 mm. in diameter (when dry), 4-8 cm. long, sulcate, finely ridged and grooved and papillate-serrulate. Inflorescences of ¢ plant axillary, solitary, zigzag racemes to 1.5 dm. long, rachis angulate, papillate-serrulate, remotely alate; flowers solitary, bibracteate, pedicellate, bracts glabrous and sparsely gland dotted, outer bract lance- ovate, acute, 1.5 mm. long, 0.5 mm. wide, inner bract lanceolate, ca. 1 mm. long and 0.4 mm. wide; pedicels ca. 0.8 mm. long, not especially dis- tinguished from the 1 mm. long perianth tube; corolla of 2 whorls of + fleshy dark-purple tepals, the outer lance-obtuse to -acute, to 3.5 mm. long, 1.3 mm. wide, the inner ovate-obtuse, 3 mm. long and 1.5 mm. wide. Stamens 3, the filaments fused in a central column 1 mm. long, the ex- trorse anthers apical, staminodia 3, in a whorl within and between the stamens, usually a little shorter than the anthers but often folded over and seeming much exceeded by them. No @ plants known. Hotorype: rocky limestone hill, Gracie Rock near Belize, British Hon- duras, February 23, 1959, Bompton 25. Sh). Cuvrivaren: Royal Bot. Gard., Kew, December 16, 1959, from Bampton 25 4 (a). It seems appropriate to name this new species of Dioscorea for the late Noe, Yvri SANDWiTH (8 September 1901-7 May 1965), through whose interest the collection of Dr. Bampton came to me for study. Mr. Sand- with’s interest in and knowledge of tropical American plants and _ his generous helpfulness to all American botanists was great indeed. His sudden death means the loss of a good friend to many of us and creates a great gap in the understanding of tropical American botany neither soon nor easily filled. > sm 1966] SCHUBERT, STUDIES IN DIOSCOREA, I Collection number . 1 1 2 4. 5 8 DIOSCOREAE BAMPTONIANAE BritisH Honpuras — 1959 Species Dioscorea bartlettii Morton Dioscorea bartlettii Morton Dioscorea cf. bernoulliana Prain & Burkill Dioscorea spiculiflora Hemsley Dioscorea alata L Dioscorea convolvulacea Cham. & Schlecht. Dioscorea pilosiuscula ae ex Spreng. vel aff. Dioscorea gaumeri Dioscorea spiculiflora Hemsley Dioscorea densiflora Hemsley Dioscorea densiflora Hemsley Dioscorea floribunda Mart. & Gal. Dioscorea gaumeri Dioscorea spiculiflora Hemsley [very immature] Dioscorea spiculiflora Hemsley Dioscorea cf. cayenensis Lam. [sterile] not identifiable Dioscorea pilosiuscula Bertero ex Spreng. y Dioscorea convolvulacea Cham. & Schlecht. r. glabra (Hemsl.) Uline Dioscorea sandwithii Schubert Dioscorea convolvulacea Cham. & Schlecht. var. glabra (Hemsl.) Uline Dioscorea floribunda Mart. & Gal. Dioscorea sp. Dioscorea sp. 160 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 ADDITAMENTA AD FLORAM JORDANICAE + BAKI KASAPLIGIL THE PLANTS DESCRIBED IN THIS PAPER were collected by the author during an assignment in the Hashemite Kingdom of Jordan, during the years 1954 to 1956, as a Forest Ecologist of the Food and Agriculture Organization of the United Nations. Specimens of the new taxa proposed were distributed among the following herbaria: Jubeicha Agricultural College, Jordan; Université Saint Joseph, Beiruth, Lebanon; Conserva- toire et Jardin Botaniques, Geneva, Switzerland (c), the University of California, Berkeley, California (uc), and the Arnold Arboretum of Har- vard University, Cambridge, Massachusetts (a). Atriplex asphaltitis sp. nov. Fics. 1 and 2. A. halimus var. argutidens Bornm. rane Thiir. Bot. Ver. 30: 82. 1913; Post & Dinsmore, Fl. Syr. Palest. & Sinai, ed. 2. 2: 436. 1933; Aellen, Bot. Jahrb. 70: 13. 1940; Eig, Palest. Jour. Soe 3(3): 124. 1945. Frutex 1—1.5 m. altus, glaucus, dioicus; folia inferiora 1.5—-2 cm. longa, 1-1.5 cm. lata, rhomboidea vel late trullata vel elliptica, apice acuta vel obtusa, basi cuneata vel truncata, margine integra vel sparse dentata, folia superiora triangularis mucronata; margine subintegra, basin versus sparse dentata; glomeruli laxe spiciformes, distincte interrupti, fasciculi diametro 6-8 mm.; bracteae fructiferae 4 mm. longae, 7 mm. latae, de- presse ovales vel reniformes, basi auriculatae, margine distincte dentatae, basi solide coniunctae et fructus cingentes, nervi prominuli, sine appendi- cibus; stylus brevis, 0.25—0.5 mm. longus, stigmata dua 1—1.5 mm. longa, in statu fructifero e bractea cingente leviter exserta; semina diametro 2—2.5 mm., brunnea, radicula sursum curvata; grana pollinis sphaeroidea, multi- forata, numerus foraminum 60 (50-72), foramina in superficie granae * The present study was carried on meaey 5 a sabbatical poe from Mills College, Oakland, California, in the academic year 1962-63. I wish t ank Monsieur René Gombault, Attaché au Muséum National a'Histoie Naturelle, pte who gave gen- 1 Royal Botanic Gardens, Kew, Great Britain (kK), and Botanischer Garten und Institut fur Systematische Botanik der Universitat Ziirich (z). Re ae ee eee eee 1966 | KASAPLIGIL, FLORA JORDANICAE 161 ) Igso1 W6idesey SC rete Ss a é Fic. 1. Atriplex asphaltitis Kasapligil. Fruiting branch, X 4/10. pollinis regulariter dispersa, textura granularis, 24.7 (21.1-29.9) micra diametro. Hororype: Kale (or Kallia), south of Jericho and at the north end of the 162 JOURNAL OF THE ARNOLD ARBORETUM [ VoL. 47 ¢ (VLAN ZY S TN ES Fic, 2. Atriplex asphaltitis er ee ao variation in the lower 8.2K i ; E,a rop [note protruding stigmas and upwardly pointed radicle], x 1 , pistil at time of pollination [dotted lines showing locule, campylotropous stat and funiculus attached to ovary base], X 70; H, spheroidal, polyforate pollen ve resembling golf ball [note + equally spaced pores and thick ectexine], < 1600; I, ground plan of pistillate flower show — prophylls and ae ovary; K- dissected embryos showing stee ural and size variation See the presence of three cotyledons in L, and fol ded i inner cotyledon in M], x ~e 1966 | KASAPLIGIL, FLORA JORDANICAE 163 Dead Sea (= Lacus poe hence the specific epithet), on dry cal- careous salines, alt. ca. elow sea level, Dec. 8, 1954, B. Kasapligil 1656 (uc 1,083,535; ae Fi Arabic names: Kataf, Rughat. In its native habitat Atriplex asphaltitis is associated with the follow- ing species, Anabasis haussknechtii Bunge, Arthrocnemum glaucum (Del.) var. stenophylla Boiss., Nitraria retusa (Forsk.) Aschers., Tamarix pen- tandra Pall., T. amplexicaulis Ehrenb., and T. jordanis Boiss. e new species, Atriplex asphaltitis, is a member of section FRUTICU- LosAE Aellen (1940) and is related to A. griffithii Moq., which is indig- enous to regions from the Persian deserts to Afghanistan. The compari- FTE j “kasapi j dahil ud G. 3. Diplotaxis kerakensis Kasapligil. Habit of fruiting plants [note pro- fuse esha of stems at base]. 164 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 Fic. 4. Diplotaxis kerakensis Kasapligil. A, stamen, < 8; B, adaxial view of sepal with involute margins, x 5; C, petal with single trace, X 5; D, floral recep- tacle bearing pistil and horn-shaped nectaries, & 5; E, longitudinal section of “~~? 1966 | KASAPLIGIL, FLORA JORDANICAE 165 son of characters of A. asphaltitis and A. griffithii is summarized in the table which follows Hasit LOWER LEAVES UPPER LEAVES SEXUALITY INFLORESCENCE STIGMA FRUITING BRACTS SEEDS Empryo POLLEN Atriplex asphaltitis Shrub, 1-1.5 m. high, white- glaucous. Alternate, entire or sparsely dentate, rhombic to broadly trullate or elliptic; acute or retuse at apex; cuneate or truncate at gs 1.5-2 cm. long, 1-1.5 cm Triangular, sire the margins entire, but dentate at base. Dioecious. oe, loosely spicate, prominently interrupted Soa 6-8 mm. thick. “ 2 stigmas from a short style, slightly protruding from bracts at maturity, 1-1.5 m. long. Depressed-ovate to reni- and around fruit, vascula- ture egy eee: appenda- ges none, 4 mm. long, 7 mm. a ad haigle 2-2.5 mm. in diam Cotyledons 2 or 3, radicle directed upwardly oF (21.12-29.92) mi- ons in diameter, germ a 60 (50-72). Atriplex griffithii ee to 1 m. high, white- glauco posit or siteronte, en- base; 2.5 cm. long, 2 cm wide. Narrowly ovate, margins en- tire or sparsely dentate. Monoecious. Glomerules densely spicate, unisexual, clusters 10-11 mm. thick 2-3 stigmas from a longer style, completely hidden within ee at maturity, ea, 1:51 ng. Roundish oval or triangu- lar-cordate, rounded at base, long, 9 mm. wide. rains hed 2-2.5 mm. in diam Cot ee ons always 2, radi- cle directed laterally. ha ie (19.36-24.64) mi- in ne a germ ane 33 (24-44) Unfortunately, staminate specimens of Atriplex asphaltitis were not The pollen grains described here were obtained from available for study. the stigmatic surfaces of the pistillate flowers. A. asphaltitis is remarkably distinct from A. griffithii. Its relationship to A. halimus L. which belongs to section Cortacea Aellen is remote. seed showing folded embryo Sve that i hee 8 are longer than the radicle], X 45; F, ground plan o -H, rial and aad views of tricolpate pollen grains showing ‘icinestie reticulate aera x 120 166 JOURNAL OF THE ARNOLD ARBORETUM [VOL. 47 Zilla spinosa (Turra) Prantl var. emarginata Kasapligil. Portion of Hig. 5) intricately branching stem Diplotaxis kerakensis sp. nov. Fics. 3 and 4. Planta basi lignescens, ramosa; pedicelli et flores aequilongi; petala intense lutea; antherae acumine cuspidatae; nectariae corniformes; coty- ledones quam radiculae longiores; grana pollinis prolata vel subprolata, 1966 | KASAPLIGIL, FLORA JORDANICAE 167 tricolpata, exina 2.5 micra crassa, polygonali-reticulata, axis polaris 36.7 (33-40) micra longa, diameter aequatorialis 26.2 (24-28.5) micra longus. Perennial herb 20-30 cm. high, branching only at the woody base, the lower parts of stems sparingly hairy, the upper parts glabrescent, rarely bearing leaves; leaves spatulate-oblanceolate, 2—4.5 cm. long, 0.5—1 cm. broad, glaucous, coarsely dentate, hispid; pedicels equaling flowers in length; sepals 6-7 mm. long, 1.5—-2 mm. broad, ovate, the margins mem- branous, involute, the dorsal surface velutinous, with 3-5 vascular traces from base; petals spatulate, twice as long as sepals, bright yellow, with a single vascular trace from base; short stamens 7 mm. long, long stamens 9-9.5 mm. long, anthers 2—2.5 mm. long, pointed at tip, filaments without appendages, with a single trace from base; pistil 6—-6.5 mm. long, 1 mm broad on flat side, gynophore filiform, 2—2.5 mm. long, stigma obscurely bilobed, style very short; nectaries horn-shaped, 0.5 mm. long, persistent at fruiting stage; silique 3.5—4 cm. long, 2.5—3 mm. broad, erect-ascending when young, drooping at maturity; seeds 76-98 per silique, 1-1.5 mm. long, 0.75-1 mm. broad, oval, flat, the seed coat brownish yellow, the embryo folded, cotyledons longer than radicle; pollen grains prolate to subprolate, tricolpate, exine 2.5 microns thick, polygonal-reticulate, polar axis 36.7 (33-40.5) microns, equatorial diameter 26.2 (24—28.5) microns. Ho.otype: Kerak district, Kerak, Castle of Crusaders, on walls, alt. ca. 1000 m. above sea level, Dec. 16, 1954, B. Kasapligil 1773 (uc 1,083,543; isotype, A). This new species of Diplotaxis, D. kerakensis, is a member of section Catocarpum DC. (Syst. 2: 629. 1821; Prodr. 1: 221. 1824), in which the silique is characterized by having a filiform gynophore and 50-150 ovules. D. kerakensis is allied to D. harra (Forsk.) Boiss. (cf. Schulz in Engler, Pflanzenfam. 17b: 343. fig. 188 A, B. 1936), but it differs from that species as shown in the summary which follows. Diplotaxis kerakensis Diplotaxis harra Hasit Stems branching at woody Stems branching above. ase. FLOWER Large, bright yellow. Small, pale yellow. PEDICEL Equaling the flowers. Longer than the flowers. SEPALS 6-7 mm. long. 4—5 mm. long. PETALS 12-13 mm. long. 7-9 mm. long. ANTHERS Pointed at tip. Blunt at tip. STIGMA Capitate or slightly bilobed. Distinctly bilobed. NECTARIES Horn shaped Club shaped. CoTYLEDONS Longer than radicle. Shorter than radicle. Zilla spinosa (Turra) Prantl var. emarginata var. nov. Fics. 5 and 6 A, B. Folia 2-4.5 cm. longa, 3-14 mm. lata, lanceolata; spini graciles, usque 168 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 @. re y RY 2 4 ma Pox Oye: “ a CT on Ry a XS 4 ° ¥, 4: Vy se. % | oO 8, ay, TA +, bt) Q CY . ‘ cy | 8 qy oO n <5 3 a erst eee ele ie a 13s 19. Leaflet bases oo to rounded . Z. pluviatile. Zanthoxylum scandens Bl. Bijdr. Natuurk. Wetens. 249. 1825. Type: Blume 1603, Java Xanthoxylum cuspidatum Champ. ex Benth. Jour. Bot. Kew Misc. 3: 329. 1851. Type: Champion 86, Hong Kon Zanthoxylum khasianum Hook. : nig Brit. Ind. 1: 494. 1875. Type: Hooker f. & Thomson 1517, India, Ass Fagara scandens (Bl.) Engl. Nat. nie oe III. 4: 118. 1896. j . Ibid. Fagara khasyana (Hook. f.) Engl. [bi Fagara laxifoliolata ac Te.:Pi. Formosa 3: 50. 1913. Type: Kawakami & Shimada, 1911, Tai Fagara cyrtorhachia Ba Ibid. 6: 8. and Hayata, 1914, Taiwan Fagara leiorhachia Hay. Loc. cit. 10. Type: Hayata, 1913, Taiwan. : Fagara chinensis Merr. Philip. Jour. Sci. Bot. 13: 141. 1918. Type: Merr ill 10660, China, Kwangtung Province. Zanthoxylum chinensis (Merr.) Chun ng, Mem. Sci. Soc. China 1: 123. 1924 (fide Huang, Acta Phytotax. Sinica 6 (1): 72. 1957). Fagara kwangsiensis Hand.-Mazz. Sinensia 3: 186. 1933. Type: Ching 5401, China, Kwangsi Province. an Zanthoxylum yunnanense Huang, Acta ape Sinica 6 (1): 59. 1957. pe: Tsai 55960, China, Yunnan Pro Zanthoxylum laxifoliolatum (Hay.) ice. 1 bid. 81. 1916. Syntypes: Jto & Hayata, 1914 178 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 Zanthoxylum leiorhachium (Hay.) Huang, bid. Zanthoxylum cyrtorhachium (Hay.) Huang, /bid. Scandent and generally climbing or occasionally suberect or erect shrubs; dioecious; evergreen; branchlets and leaf rachises generally armed, the prickles scattered and retrorse. Leaves imparipinnate, 8-37 cm. long; rachis glabrous to short pubescent; petiolules 1-5 mm. long; leaflets (1) 2-12 pairs, subopposite to alternate, chartaceous to Se dark green and glabrous above, paler and glabrous or sparsely pubescent below, generally blackening slightly with drying, ovate to elliptic-lanceolate, 3-9 cm. long, 1-4 cm. wide, base cuneate to obtuse, main veins 4—14 on each side of the midr rib, oe peri in the apical half of the blade, th ‘ terminal, paniculate, to 12 cm. long and generally about one-half as wide, the axes and branches glabrous to short pubescent. Staminate flowers 3-6 mm. long; pedicels obsolete to 5 mm. long; sepals 4, triangular to rounded, about 1 mm. long; petals 4, white to pale yellow, occasionally with a purple margin, elliptic-ovate, about 3 mm. long; stamens 5-6 mm. long, anthers about 1 mm. long; disc flat, about 0.3 mm. high; rudimentary carpels 1-4, to 1.5 mm. high. Carpellate flowers about 3 mm. long; pedicels, sepals, petals and disc as in staminate flowers; rudimentary stamens obso- lete to 1 mm. long; gynoecium 4-carpellate, 1—-1.5 mm. high, styles diver- gent at anthesis, stigmas globose. Fruiting pedicels 1-6 mm. long; folli- cles subglobose, 4-5 mm. in diameter, in 4’s, 3’s, 2’s or single with 0, 1, 2, or 3 undeveloped carpels, respectively. India. Assam: Khasia Hills, Chand 4908 2 (micw), Hooker f. & Thomson 1517 2 (K —holotype of Zanthoxylum ap eIO GH — isotype), Ruse 62 (A), 62A 2 (a). China. YUNNAN PRovINCE: Sungkw S acae rite 21554 Q (A, US) ; Mengtsz, Henry 10421A 6 (a), 11163 Q (A), Tsai 52369 2 (A); Wen-shan Hsien, Tsai 51703 2 (a), 51764 2 (a); O-shan Hsien, Tsai 53494 8 (a); Lan-ping Hsien, Tsai 54018 2 (a); Ping-pien Hsien, Tsai 55108 2 (a), 60072 2 (A) 60451 2 (a); Lung-ling Hsien, Tsai 55610 @ (a), 55695 8 (a); A-tun- tze, Huann-fu-ping, Wang 69107 2 (a): Fo Hai, Wang 74145 2 (a); Kiukiang Valley, Yu 19463 2 (a), 20409 @ (a); without definite locality, Forrest 11922 2 (A), 15715 6 (a) 17712 & (a), Tsai 55960 2 (A— isotype of Zanthoxylum yunnanense), 60269 2 (a). KwerIcHow Province: Fan Ching Shan, Steward, Chiao & Cheo 757 @ (a); Tsingchen, Mou-po, Teng 90156 2 (a), 90174 & (A). Kwancst Province: Luchen, Moo Shan, Ching 5401 2 ( A— isotype of Fagara kwangsiensis); Lau Ha Tung, Ching 6584 Q (a, us); N. Hin Yen, Yeo Mar Shan, Ching 7130 2 (a, us); Tzu Yuen District, Chung 83584 2 (a); Waitsap District, Se Tze Shan, Tsang 23310 2 (a); Shang- -sze District, Shap Man Taai Shan, Na Wai Village, Tsang 23924 2 (a); Yao Shan, Wang 39369 2 (A), 40057 @ (A); without definite locality, Wang 41125 2 (a). Kwanocrunc Prov- INCE: Road to Mahang, Chun 5952 2 (a); Lung Tau Shan, Kang, W. T. Tsang & U. K. Tsang CCC 12266 2 (us), CCC 12770 2 (us), To & Tsang 12770 2 (aA); Wung Yuen District, Tsing Wan Shan, Lau 2463 2 (a); Loh Fau Mt., Levine, McClure & To, August 31-September 4, 1921 2 (a), Merrill 10660 2 (us — isotype of Fagara chinensis; A, photo. of holotype from Manila, specimen 1966 | HARTLEY, MALESIAN ZANTHOXYLUM 179 destroyed); Canton Delta, Heung Shan, Tung sige To CCG 22972): Sin-fung District, Ngok Shing Shan, Taam 426 6 (A); Tseungshing District, Tsang 20149 (a): Tsung-fa District, Sam Kok Shan, Tsang 25017 ¢ (a), 25193 2 (A); Hwei-yang District, Lin Fa Shan, Tsang 25698 2 (a); Jen-hua District, Man Chi Shan, Tsang 26248 6 (a); Ho-yuen District, Kwai Shan, Tsang 28508 & (a); Ying Tak, Tai Tsan, Chung Tung, Tsang & Wong 3110 9 (aA); Wai-yeung District, Pak-Wan Cheung, Tsui 137 2 (A, us). HAINAN IsLAND: Po-ting, How 73154 (a). Honc Kone: Champion 86 8 (Kk — holotype of Xanthoxylum cuspidatum), Hance 1145 6 Sea Taam 1111 6 (a, us), 1825 2 (A, US), Wright 82 $ & @ (cH), sm (us). HUNAN PROVINCE: Yun-schan prope Wukeng, Handel- Maszetti oe | (a), Hui Handel-Maz- zetti Expedition 12321 6 (A); Yi Chang District, Ping Tou Shan, Pai Mu Village, Tsang 23387 8 (A, us). KiANncst Province: Kaling, Chun 4278 @ (a); Swe-chuen Hsien, Hu 840 2 (a); Chung Yih Hsien, Hu 934 2 (a); circa carbonis minas Pinghsiang, Hui 158 2 (A); Lungnan District, Oo Chi Shan, near Lam Uk Tung Village, Lau 4529 2 (a, US). FUKIEN Province: Hinghwa District, Chung 1030 8 (a); Kuliang, Chung 6647 2 (a), Norton 1393 2 (us). CHEKIANG PRovINcE: Kwanying Tung, Yen Tang Shan, Chiao NU 14760 @ (a, US). Taiwan. Mt. Lee-shan, Taichung Hsien, Feung & Kao 4583 2 (A); Mt. Arisan, between Heishana and Funkiko, Hayata, April 24, 1913 2 (11— holo- type of Fagara leiorhachia; a, photo.); Mt. Arisan, between Heishana and Ni- mandaira, Hayata, April 28, 1914 2 (tTI—syntype of Fagara cyrtorhachia; A, photo.) ; South Cape, Henry 343 2 (us), 1969 2 (us); between Funkiko and Taroyen, Ito & Hayata, March 29, 1914 &é (t1—syntype of Fagara cyrtor- hachia; Aa, photo.) ; Mt. Shichiseitonzan, Kawakami & Shimada, April 2, 1911 ° TI — holotype of Fagara laxifoliata; A, photo.); Tamsu, Oldham 59 @ (GH); Batakan Tabito, Tanaka, October 22, 1930 (a); Nanto Province, Lake Candi- pa Wilson 9973 2 (a, us); Taihoki Province, near Hesinbi, Wilson 10229 é A, US); without os peseyet Faurie 8216 (A). Ryukyu Islands. Ishigaki Island, Hatusima 18923 2 (us), Masamune, March 31, 1935 6 (us). Sumatra. Sumatra occidentale nel Pandangesche bovenlanden (alto Padang), Beccari 132 (K, L, MEL); Westkust, G. Koerinajr, Bunnemeyer 9574 6 (L); Bukit G. Sembilang, Mt. Sago, near Ladang Lawas, Meijer 4043 Q (1). Java. WEST Java: ad montem Tjerimai Provinciae Cheribon, Blume 1603 2 (1 — holotype of Zanthoxylum scandens; A— isotype); Tjibodes, Ooststroom 13255 ¢ ( L). CENTRAL Java: Dieng (Dijeng Plateau), Junghuhn 9 (L). SouTHEAST JAVA: Forbes 1001 2 (BM), s.m., 1880-82 2 (BM). dems definite locality: Hors- field 66 2 (MEL), 1093 9 (x), sm. [cH (2), K $)]. British North Borneo (Sabah). Mt. Kinabalu, Tenompok, J. & M. S. eines 29213 [a (8), K (¢), L (2), NY (6)], 29243 8 (a, K, L, NY). Distripution, NE India east to Taiwan and Ryukyu Islands; south in Sumatra, Java and British North Borneo; on dry, rocky or sandy slopes and in thickets and mountain forests at altitudes from 1000 to 2000 meters in Malesia and from 300 to 3400 meters in continental Asia. See Map 1. Ittustrations. Huan, C. C., Acta Phytotax. Sinica 6(1): ¢. XJ. 1957, as Zanthoxylum yunnanense. L1v, T. S., Illustrations of native and introduced ligneous plants of Taiwan 2: ¢. 1962, as Zanthoxylum cuspidatum. Acceptance of the extremely variable characters of pubescence, number 180 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 5 , a = ‘ PES : “> os j 3 ; Eo4 ne, a . oo : f 1 ba . TATY pts Mert t--- pest Map 1. Distribution of Zanthoxylum scandens B\. The broken line sur- ounding the southern stations indicates the geographic area of Malesia as delimited i in this paper of leaflets, and presence or absence of prickles as being of taxonomic sig- nificance would necessitate recognition of an even larger number of taxa than has already been described in this complex. The more conservative delimitation presented here is based on floral characters, arrangement of leaflets, and the type of crenation of the leaflet margin, all of which are relatively constant and, I feel, serve best to show the natural relationship of the various morphological forms involved. A number of species from continental Asia, including Zanthoxylum dissitiodes Huang, Z. laetum Drake, Z. calcicolum Huang, Z. chaffanjoni Levl. and Z. oxyphyllum Edgw., show a close relationship to Z. scandens. The following have previously been listed as synonyms of Zanthoxylum cuspidatum or Fagara cuspidata (= Zanthoxylum scandens): Zanthoxy- lum khasianum, by Guillaumin in Flore générale de l’Indo-Chine 1: 640. 1911; Fagara laxifoliolata, F. cyrtorhachia and F. leiorhachia, by Li, Woody flora of Taiwan, 373. 1963; and F. chinensis, by Huang, Acta Phytotax. Sinica 6 (1): 72. 1957. 2. Zanthoxylum nitidum (Roxb.) DC. Prodr. 1: 727. 1824. Fagara nitida Roxb. Fl. Ind. ed. Carey & byes 1: 439. 1820. Lectotype: Roxburgh Icones 2430, Cult. Bot. Gard. Calcutt Nugae sylvarum silvestris Rumph. Herb. Amb. 5: Piper pinnatum Lour. Fl. Cochinch. 31. 1790 yee el Ps 1966 | HARTLEY, MALESIAN ZANTHOXYLUM 181 Zanthoxylum torvum F. Muell. Frag. Phyt. Austral. 7: 140, 1871. Type: Dallachy, August 21, 1869, Queensland. Zanthoxylum hamiltonianum Wall. ex Hook. f. Fl. Brit. Ind. 1: 494. 1875. Type: Hamilton, February 12, 1809 (Wallich cat. n. 7117), India. Fagara torva (F. Muell.) Engl. Nat. Pflanzenfam. III. 4: 119. 1896. Fagara warburgii ting Fragm. Fl. Philip. 160. 1905. Type: Warburg 11784, Philippines, not see Zanthoxylum hirtellum Ridl. oo Fed. Malay States 10: 131. 1920. Type: Ridley 11291, Malay Penins Zanthoxylum collinsae Craib, Kew Bull. 1926: 165. 1926. Type: Collins 705, Thailan Fagara hamiltoniana (Wall. ex. Hook. f.) Engl. Nat. Pflanzenfam. ed. 2. 19a: 931. Fagara hirtella (Ridl.) Engl. Jbid. Zanthoxylum scabrum Guill. Bull. Soc. Bot. Fr. 91: 215. 1944. Type: Poilane 28200, North Vietnam. Fagara oblongifolia Bakh. f. Blumea 6: 366. 1950. Syntypes: Blume 1814 and $.n., Java. Fagara pendjaluensis Bakh. f. Ibid. Type: Koorders & Koorders-Schumacher 479638, Java. Zanthoxylum asperum Huang, Acta Phytotax. Sinica 6 (1): 75. 1957. Type: Wang 75548, China, Yunnan Province. Zanthoxylum asperum var. glabrum Huang, Ibid. 76. Type: Wang 77912, China, Yunnan Province. Scandent and generally climbing or occasionally suberect or erect shrubs; dioecious or rarely monoecious; evergreen; old stems occasionally with thick, corky spines to 2 cm. long; branchlets, leaf rachises and midribs of leaflets generally armed, the prickles scattered and retrorse. Leaves imparipinnate, 5—40 cm. long; rachis glabrous to short pubescent; petio- lules 2-4 mm. long; leaflets (1) 2—4 pairs, opposite, chartaceous to coria- ceous, with or without scattered and/or marginal pellucid dots, glabrous to slightly scabrous above, glabrous to short pubescent below, ovate to elliptic, (1.3) 5-12 (16) cm. long, (0.7) 2.5-6 (8.2) cm. wide, base cuneate to truncate, main veins 5-15 on each side of the midrib, margins entire to glandular crenate with as many as 4 crenations per cm., apex acuminate to obtuse. Inflorescences axillary or axillary and terminal, racemose to paniculate, to 15 cm. long and generally less than one-half as broad, the axes and branches glabrous to short pubescent. Staminate flowers 4—5 mm. long; pedicels 1-3 mm. long; sepals 4, broadly triangular, about 1 mm. long; petals 4, white to pale yellow or rarely reddish, elliptic-ovate, 2-3 mm. long; stamens 4, 4-5 mm. long, anthers about 1 mm. long; disc flat, about 0.3 mm. high; rudimentary carpels 2 or 4, about 1 mm. high, Carpellate flowers 2-3 mm. long; pedicels, sepals and petals as in staminate flowers; disc flat to pulvinate, 0.1-0.5 mm. high; gynoecium 4-carpellate, 1.5—2 mm. high, stigmas coherent at anthesis, the combined stigmatic struc- ture peltate. Fruiting pedicels 1-7 mm. long; follicles subglobose, 5—7 : in diameter, in 4’s, 3’s, 2’s or single with 0, 1, 2 or 3 undeveloped carpels, respectively. 182 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. 47 India. Bruar: Borybari (near Bhagalpur), Hamilton, February 12, 1809 (xk —holotype of Zanthoxylum hamiltonianum, not seen; A, photo.). NORTHERN BENGAL: Biswas 1707 (A). Assam: Naga Hills, Prain’s Collector 758 ( A), 883 definite locality: Kamphovener 169H & (A). Thailand. Sriracha, Tapibut, Lape- boot-tong Kan Forest, Collins 705 6 (kK —holotype of Zanthoxylum collinsae) ; Ponpissay, Kerr 8573 6 (sinc). North Vietnam. Si Ninh Binh Province, Pete- lot 875 & (a, us); Lontoy Province, Petelot 1309 (a); Bhai Nginfen Province, Petelot 2691 2 (A); Than-hoa Province, La-han, Poilane 1687 (A, photo.) ; inter Suyut et Cho Bo, Poilane 28200 (p — holotype of Zanthoxylum kali not seen; A, photo.). South Vietnam. Mt. Bana, ca. 25 km. W of Tourane (Da Nang), ia & M.S. Clemens 3996 ‘9 (a, us); Hue, Squires 170 2 (A, SING, US). China. YUNNAN PrRovINcE: Si-chour-hsien, Faa-doon, Feng 11735 2 (a); Fo- hai, Wang 74643 2 (A); Che-li Hsien, Wang 75548 2 (A— isotype of Zanthoxy- lum asperum), 77546 2 (A), 77912 2 (A), 79173 9 (A). KWwancsi PROVINCE: Lar Pan, W of Hochih, Ching 6510 2 (a); W. Poseh, Bako Shan, Ching 7685 @ (A, us); Pin-lam, Ko 55674 2 (a); Lungchow, Morse 332 & (us), 382 6 (Us); Sui-luk, SW of Nanning, Tsang 21919 8 (a); Sun-to District, Po Yam Shan, near Tai Chung Village, Tsang 22973 @ (a). Kwanotunc Province: Lofou- shan, Chun 40435 8 (a, sinc), Tsiang 1619 & (A), 1651 2 (A, US); vicinity of Yeungkong, Ferris 11934 2 (GH); Canton and vicinity, Hai 33 ¢ (aA), Levine CCC 207 (A, ws), CCC 236 2 (4, O8),CCC #07 ¢ (a, us), CCC 1730 & (A), CCC 1625 2 (a, us), Merrill 9940 2 (a), To CCC 407 8 (a), CCC 1130 2 (A, us), Tsiang 17 8 (a), 31 2 (a); Chung Shan District, Tong Ka Wa Village, Hom 7-79 & (a); SE of Shangsze, Tsang 23924A 2 (a); Ho-yuen District, Nam Shan, Tsung-shue Village, Tsang 28864 2 (A); Yingtak District, Tsiang 1927 3 (A, stnG); Sun-wui District, Tso & Tsaing 2008 2 (a, sinc); Yun Fou District, West River, Lou Lu Tsun, Wang 382 6 (a); without definite locality, Tsoong 1472 & (a). Hanan Istanp: Dung Ka, Chun & Tso 43581 2 (A); Fan Maan Tsuen & vicinity, Chim Shan, Fung 20202 2 (a, us); Ta Hau, Gressitt 965 2 (aA); Yaichow, How 70306 3 (a, us), 70675 2 (a), Liang 62340 2 (us); vicin- ity of Manning, How 71366 & (A); Manyun, Tun Shan Ling, Ko 52130 2 (A), Liang 61515 2 (A); Ngai District, Yeung Ling Shan, Lau 62 @ (a, us); Tan District, I Kap Shan & vicinity, Low 1079 (a); Kan-en District, Chim Fung Ling, Lau 3605 9 (a), 5523 & (a); Ching Mai District, Lei 342 (a), 363 2 (SING), baer definite locality, Liang 65076 & (A), 65207 6 (us), 65216 % (a), 65318 8 (A). Hone Kone: Chun 4809 @ (a), Taam 1159 8 (a, us), 2024 2 (US), Tsang LU 16697 & (a, us), Wright 81 6 & 2 (cH), 83 $ & 2 (GH), sm. @ us). FuxKreN Province: Hinghwa District, Chung 936 2 (a, stnc); Inghok Hsien, Fang-quang-yen, Chung 2652 2 (a), 7977 2 (A); Amoy & vicinity Chung 4770 (a), 5607 2 (a), 5889 8 (A), 5892 2 (A), 5935 8 (A), 6014 Q (A), 6252 2 (A); Foochow vicinity, Chung 8417 (a), Ging 13051 (a), 13531 8 (a), Norton 1394 (us); Kushan, Dunn 72 2 (a); Hok Chiang vicinity, Ging 15333 2 (A); without definite locality, Chung 6951 2 (a), Dunn’s expedition, April to June, 1905 2 (a). Taiwan. Tamnii, Faurie 23 2 (a); Hokuto, Faurie 1268 (4); Ku- raru, Gressitt 478 2 (a); South Cape, Henry 205 two sheets, 6 & 2 and 6 (A); Tchow, Henry 205A 2 (us), 205G 2 (us), 1782 [a (4), us (two sheets, ¢ and 2)]; Taitung Mts., Keng & Kao K2674 Q (a); Taihoku and vicinity, Kou & Kae 4496 & (a), Odashima, June 22, 1934 2 (A, GH, sING, Us), Suzuki, June 18, 1933 2 (a), Wilson 10161 [a (two sheets, 6 and 2), us (@)]; Tamsuy, 1966 | HARTLEY, MALESIAN ZANTHOXYLUM 183 Oldham 60 8 (us); Mt. Daiton and vicinity, Tanaka 1784 8 (a, GH); leg definite locality, Faurie 23 2 (a), Henry 464 $ (a), Oldham, 1864 2 (GH). Ryukyu Islands. SaxisH1ma GunT0: Iriomote Island, Hatusima 18833 2 (us); Ishigaki Island, Smith 85 (us); Yonaguni Island, Walker & Tawada 6801 (vs). Sumatra. Middle Patani Valley, Lorzing 15257 2 (L); Sibolangit, Galoengi 231 2 (L). Malay Peninsula. PERAK: Padang Rengas, Haniff SF 14978 (stn); Din- dings, Ridley 10281 2 (sING). PAHANG: Manchio, Haniff SF 16778 (stnc); Temerloh, Ridley, 1891 2 (sinc). SIncAPoRE: Ridley 11291 2 (sinc — holo- type of Zanthoxylum hirtellum; K, fragments). Java. West Java: Batavia vicin- ity, Backer 73 2 (1); Lengkong, Bacher 16980 & (1); Res. Batavia, bee Goea si Gadjah, Backer 31164 2 (1); Preanger Regencies, Mt. Megamendoeng, Blume 1814 2 (1—syntype of Fagara oblongifolia); Mt. Sapaetinn Blume 2 (L—syntype of Fagara oblongifolia); Depok, Hallier, August 11, 1896 (L); vicinity of Buitenzorg, Hochreutiner 28 (L); Natuurmonument- eilandje Noes- agede in het meer van Pendjaloe, Koorders & Koorders-Schumacher 479638 @ (L— holotype of Fagara pendjaluensis). East Java: Kediri, Backer 11717 @ (L). Anambas Islands. Siantan Island, Henderson SF 20132 2 (stnc), Steenis 748 2 (1, SING). Philippines. PALAWAN IsLAND: vicinity of Puerto Princesa, Ebalo 448 (a), Sulit PNH 12282 9 (a, L). Luzon Istanp: Cagayan Province, IsLanp: Palo, Elmer 7278 [a, K (4), L]; without definite locality, Wenzel 1232 9 (GH, NY). TawrtTawi IsLaAnp: Ramos & Edano BS 44007 2 (A, BM, BRI, K, NY, SING, US). Celebes. Kabaena Island, Elbert 3445 2 (A, L, SING). Moluccas. Soela Islands, Sanana Island, Atje 2 (1); Amboina Island, Robinson Plantae Rumph. Amb. 250 @ (BM, GH, K, L, NY, SING, US). Papua. Western District, Lower Fly River, east bank opposite Stuart Island, Brass 7983 2 (A, BRI, L); Central District, Dieni, Ononge Road, Brass 3833 2 (A, BM, BRI, NY). Territory of New Guinea. Morose ee Wampit River 34 mile from Lae-Wau Road, Millar NGF 9997 @ (prt, CANB, K); Bulolo Valley, Floyd NGF 7471 2 (t). Solomon Islands. cece ie Island, Karngu, Bbuin, Kajewski 2304 @ (A). Australia. QUEENSLAND: Daintree River, Brass & White 180 2 (A); Rocking- ham Bay, Herbert’s River, Dallachy, October 4, 1866 @ (BRI, MEL), August 21, 1869 [MEL (two sheets, 2 and ¢ 2) —syntypes of Zanthoxylum torvum; GH (8)]; Mossman, Jones 2060 (canB); Johnstone River, Ludbrook 58 ° (prt): Cairns, Nugent 2 (prt); Tully, Jarra Creek, Webb 2352 (CANB). Cultivated. India: Calcutta, Roxburgh Icones 2430, drawn from a oe from Canton, China (kK — lectotype of Fagara nitida, not seen; A, photo. DistriBuTIoNn, NE India east to Taiwan and the Ryukyu Islands; south in South Vietnam, Thailand, Sumatra, Malay Peninsula, Java, Anambas Islands, Philippines, Celebes, Moluccas, E New Guinea, Solomon Islands and NE Queensland; in rain forests and thickets at altitudes up to 1100 meters in Malesia and up to 1400 meters in continental Asia. See MaP 2. ILLUSTRATIONS. Huance, C. C., Acta Phytotax. Sinica 6(1): ¢. XV (as Zanthoxylum nitidum) and t, X VI (as Zanthoxylum asperum). 1957. Lt, H. L., Woody flora of Taiwan 374. 1963. Liv, T. S., Illustrations of native and introduced ligneous plants of Taiwan 2: ¢. 729, 1962. PrerrE, L., Flore forestiére de la Cochinchine 4: ¢. 291. 1893. Occasionally cultivated as a hedge plant in China. In Taiwan the root 184 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 Map 2. Distribution of Zanthoxylum nitidum (Roxb.) DC. is used in making pipe stems. In both the Philippines and China (Hainan Island) collectors have noted that the plant was pounded and placed in pools to stupefy fish. In the Malay Peninsula the bark is used for treat- ment of toothache. According to Merrill, Comm. Lour. Fl. Cochinch. 218. 1935, the culti- vated material at Calcutta that provided the type for Roxburgh’s Fagara nitida was originally from Canton, China. Sealy, Kew Bull. 11: 339. 1956, points out that another Roxburgh draw- ing at Kew, that of Elaeagnus arborea Roxb., bears the same number as Fagara nitida Roxb. Merrill, Interp. Rumph. Herb. Amb. 288. 1917, lists Nugae sylvarum silvestis Rumph. as a synonym of Fagara torva (F. Muell.) Engl. (= Zanthoxylum nitidum). The name Piper pinnatum Lour. is listed as a synonym of Zanthoxylum nitidum by Merrill, Comm. Lour. Fl. Cochinch. 218. 1935, who reasoned that Loureiro must have been misled by the peppery taste of the fruit. The epithet pinnatum was used earlier to describe a Norfolk Island species, Blackburnia pinnata J. R. & G. Forst. [Zanthoxylum pinnatum (J. R. & G. Forst.) Druce}. I have not seen the type of Fagara warburgii Perk. (Warburg 11 ae which was apparently lost at Berlin. Merrill, Enum. Philip. Fl. Pl. 2: 327 1923, cites this Warburg number (apparently from an isotype at Manila, also presumably lost) and lists the name as a synonym of Zanthoxylum torvum F. Muell. (= Zanthoxylum nitidum). Vegetatively, Zanthoxylum nitidum exhibits a wide range of variation that appears to correspond with ecological differences within its range. 1966 | HARTLEY, MALESIAN ZANTHOXYLUM 185 In general, the collections fall into two groups: large-leaved plants of rain forests and small-leaved plants of dryer situations, generally described as thickets. Plants of the first group are high climbing lianas with leaves over 25 cm. (up to 40 cm.) long, leaflets chartaceous or subcoriaceous, dull above and generally with some pubescence on the rachis and lower surface. This form predominates in the Malesian area and is otherwise scattered throughout most of the range of the species. Plants of the second, smaller- leaved group are climbing or semi-erect shrubs with leaves less than 25 cm. (down to 5 cm.) in length, leaflets tending to be coriaceous, shiny above, and glabrous. For the most part this form is restricted to the Asiatic main- land, though there are a few outliers at higher elevations in the islands of Malesia (Ebalo 448, from the Philippines; Elbert 3445, from the Celebes; and Galoengi 231, from Sumatra). There is a great deal of intermingling of the various vegetative features of these two groups, especially on the Asiatic mainland, and while this further complicates the problem by giving rise to several local races that at first appear distinct from either of the two main morphological groups, intergradation is so complete that I have found it impossible to recognize but a single species. The closest relative of Zanthoxylum nitidum appears to be Z. tetra- spermum Wight & Arn., of southern India and Ceylon, which differs from Z. nitidum in having primarily terminal inflorescences, rather fleshy petals, sepals connate basally and leaflets with 15—20 main veins on each side of the midrib. Z. caudatum Alston, of Ceylon, is also closely related and is otherwise interesting since it seems to combine some vegetative characters of Z. scandens. An isotype examined [Silva 128 (K)] has leaflets resem- bling Z. scandens in size, shape and color, but with the arrangement and margins of Z. nitidum 3. Zanthoxylum myriacanthum Wall. ex Hook. f. Fl. Brit. Ind. 1: 496. 1875. Type: Porter (Wallich cat. n. 1214), Malay Peninsula. Zanthoxylum rhetsoides Drake, Jour. Bot. Paris 6: 275. 1892. Syntypes: Balansa 4044 (not seen) and 4045, North Vietnam. Fagara myriacantha (Wall. ex Hook. f.) Engl. Nat. Pflanzenfam. HI. 4: 118. 1896. Zanthoxylum diabolicum Elmer, Leafl. Philip. Bot. 2: 477. 1908. Type: Elmer 10217, Philippines. ; Evodia odorata Levi. Repert. Sp. Nov. 9: 458. 1911. Type: Cavalerie 2978, hina, Kweichow Province. Zonthoxylum odoratum (Levl.) Lev]. /bid. 13: 266. 1914 Fagara gigantea Hand.-Mazz. Anzeig. Akad. Wiss. Wien. 58: 64. 1921. Type: Handel-Mazzetti 12327, China, Hunan Province : Zanthoxylum giganteum (Hand.-Mazz.) Rehd. Jour. Arnold Arb. 8: 151. 1927. Fagara diabolica (Elmer) Engl. Nat. Pflanzenfam. ed. 2. 19a: 220. 1931. Fagara odorata (Levl.) Hand.-Mazz. Symb. Sinica 7: 623. 1933. 5 Fagara rhetsoides (Drake) Reeder & Cheo, Jour. Arnold Arb. 32: 69. 1951. 186 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 Small to large trees to 25 m.; dioecious; evergreen; main stems prickly or with thick, conical spines to 3 cm. long; branchlets with numerous straight prickles, generally swollen and hollow and apparently housing ants. Leaves imparipinnate, 25-45 (60) cm. long, glabrous or (rarely) short pubescent on the lower surface; petiolules obsolete to 5 mm. long; leaf- lets 4-9 (11) pairs, opposite or subopposite, chartaceous to coriaceous, dark green and lustrous above, pale green below, pellucid dotted, elliptic, 8-18 cm. long, 2.5—-8 cm. wide, base obtuse to subcordate, slightly oblique, main veins 8-18 on each side of the midrib, margins finely (often indis- tinct) glandular crenate with 6-8 crenations per cm., apex acuminate. In- florescences terminal and from the upper leaf axils, paniculate, 15—25 cm. long and generally about as wide, the axes and branches glabrous to (rarely) short pubescent. Staminate flowers 3-4 mm. long; pedicels obso- lete to 2 mm. long; sepals 5, broadly triangular, about 0.75 mm. long; petals 5, white, pale yellow or occasionally violet, elliptic, 1.5—-2.5 mm. long; stamens 5, 3—4 mm. long, anthers 0.5—1 mm. long; disc flat; rudimen- tary carpels 3, about 1 mm. high. Carpellate flowers 2-3 mm. long; pedi- cels, sepals, petals and disc as in staminate flowers; rudimentary stamens absent; gynoecium 3-carpellate (occasional flowers 4-carpellate), 1.5—2 mm. high, styles and stigmas coherent, the combined stigmatic structure peltate. Fruiting pedicels obsolete to 2.5 mm. long; follicles subglobose, 3-6 mm. in diameter, in 3’s, 2’s or single with 0, 1 or 2 undeveloped carpels, respectively. India. Assam, Naga Hills, Koelz 25762 2 (micH). North Vietnam. Vallei de Lomkok, Balansa 4045 2 (» —syntype of Zanthoxylum rhetsoides, not seen; A, hoto.; L); Chapa, Petelot 3569 $ (us), 7911 2 (a); Tien-yen, Kau Nga Shan, Tsang 30511 2 (A). China. YUNNAN PrRovINce: Che-li Hsien, Wang 77979 2 (A— isotype of Zanthoxylum rhetsoides var. pubescens), 78566 & (A). KweEI- cHOW Province: Pin-fa, Cavalerie 1771 & (a, photo. and fragment); Ma-ho (Ma-jo), Cavalerie 2978 2 (e—holotype of Evodia odorata, not seen; A, photo., mae fragment) ; Bios Hsiang Ping, Fan Ching Shan, Steward, Chiao & pee 708 2 (A, US). AN Province: Wukang, Vuu-shan, Handel-Mazzetti 12327 8 (a—isotype ne is gigantea). KwANGsI Province: S Nanning, Seh-feng Dar Shan, Ching 8128 2 (a); Tzu Yuen District, Chung 83485 6 (A); Pin-lam, Ko 55597 & (a); San Chiang Hsien, Steward & Cheo 969 2 (A, SING). KWANGTUNG Province: Wung Yuen District, Fan Shiu Shan, Lau 2770 2 (A); Ying Tak, Tsang & Wong 2453 6 (a). Hatnan IsLanp: Po-ting, Lau 28166 9 (a); Taam-Chau District, Hoi Ta Shan, Tsang 837 2 (a); without definite locality, Liang 64222 2 (A). Sumatra. CENTRAL SuMaTRA: Padang and Bedagei, Lorzing 1519 (Lz); Gouv. O. kust. ond afd. Simeloengoen, N/FS bb. 2 (L); vicinity of Aek Moente (Aer Moette), Asahan (NE of Tomoean Dolok and W of Salabet, Rahmat si Boeea 9383 (A, L, MICH), SOUTHERN SU- MATRA: Res. Benkoelen ond. afd. Redjang, Endert NIFS E 1078 3 (1). Malay Peninsula. Perak: Goping District, Tohoflow Hills, King’s Collector 8157 2 BM); without definite locality, Scortechini 275 (L, sinc). PAHANG: Fraser’s © Henderson SF 21706 (stnc); Cameron Highlands, Sg. Ikon, Jaamat SF 27628 ¢ (stnc); Telom, Ridley 13541 2 (BM, K, SING). SELANGOR: Kepong, So Tachou SF 16893 6 (stnc). Matacca: Sg. Udang, Alvins 36 2 (stNnc), Derry 1966 | HARTLEY, MALESIAN ZANTHOXYLUM 187 1121 2 (sING); Bukit Bruang, Holmberg 857 8 (sinc); without definite local- ity, Maingay 279 & (A, K, L), 1098 2 (K). PENANG: Penang Hill, Corner, July (SING); without definite locality, Porter 2 (k—holotype of Zanthoxylum myriacanthum ; BM — isotype). SINGAPORE: Ridley 8 (Bm). Borneo. Sarawak. Ulu Lawas, Omar Sar 00107 & (stNc); British North Borneo (Sabah), Mt. Kinabalu, J. & M. S. Clemens 26344 Q (A, BM, L, NY), 26344A @ (tL, Ny). Philippines. Luzon Island, Laguna Province, Mt. Makiling, Canicosa FB 31197 . (NY); Samar Island, Ramos BS 1680 4 (BM, BRI, GH, L, NY, SING, US), Rosen- bluth FB 12866 & (us); Negros Island, Dumaguete, Cuernos Mts., Elmer 10217 3 (isotypes of Zanthoxylum diabolicum: A, BM, K, L, NY, US). DistriBuTIoN. Assam, North Vietnam, SW China, Sumatra, Malay Peninsula, Borneo and the Philippines; in forests and thickets at altitudes from 100 to 1230 meters in Malesia and from 200 to 2145 meters in conti- nental Asia. See MAP 3. Map 3. Distribution of Zanthoxylum myriacanthum Wall. ex Hook. f. The fruit is said to be used as a condiment in Assam . This species is very closely related to Zanthoxylum ailanthoides, which has identical flowers and fruits. Z. molle Rehd., a species from Hupeh and Anhwei Provinces, China, is also closely related, differing mainly in having leaves that are white tomentose below. ; Reeder & Cheo, Jour. Arnold Arb. 32: 69. 1951, cited Evodia odorata and Fagara gigantea as synonyms of Fagara rhetsoides (= Zanthoxylum myriacanthum). Wandel-Mazzetti, Symb. Sinica 7: 623. 1933, listed Fagara gigantea as a synonym of F. odorata (= Zanthoxylum myriacan- thum). 188 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 4. Zanthoxylum ailanthoides Sieb. & Zucc. Abh. Akad. Miinchen 4 (2): 138. 1846 (“Zanthoxylon”). Type: Siebold, Japan. Zanthoxylum emarginellum Miq. Ann. Mus. Lugd.-Bat. 3: 22. 1867. Type: apan. Fagara ailanthoides (Sieb. & Zucc.) Engl. Nat. Pflanzenfam. III. 4: 118. 1896. Zanthoxylum hemsleyanum Makino, Bot. Mag. Tokyo 21: 86. 1907 (Zan- thoxylum emarginellum sensu Hemsley, Ann. Bot. 9: 149. 1895). Syntypes: Cunningham, Chushan Archipelago (not seen); Ford, June, 1884, Taiwan. Fagara hemsleyana (Makino) Makino, Bot. Mag. Tokyo 21: 161. 1907. Zanthoxylum ailanthoides Sieb. & Zucc. var. inerme Rehd. & Wils. Jour. Arnold Arb. 1: 118. 1919. Type: Wilson 8265, Bonin Islands. Zanthoxylon inerme (Rehd. & Wils.) aici Bot. Mag. Tokyo 33: 218. 1919 Fagara boninshimae Koidz. ex H ‘au Spermat. Jap. 3: 25. 1954 (based on Zanthoxylum pi sere eb var. " inerme). Medium to large trees to 25 m.; dioecious; apparently evergreen; main stems prickly or with thick, conical spines or rarely unarmed; branchlets generally armed, the prickles straight and scattered. Leaves imparipin- nate, 25-80 cm. long, glabrous; petiolules obsolete to 3 mm. long; leaflets 7-15 pairs, opposite, chartaceous to subcoriaceous, medium to dark green above, glaucous below, pellucid dotted, elliptic-lanceolate to lanceolate, broadest near the base, 9-14 cm. long, 2.3-4.7 cm. wide, base rounded to cordate, slightly oblique, main veins 12-20 on each side of the midrib, margins glandular crenate with 4-8 crenations per cm., apex long acumi- nate to attenuate. Inflorescences terminal and from the upper leaf axils, paniculate, 10-30 cm. long and generally about as wide, the axes and branches glabrous. Staminate flowers 3-4 mm. long; pedicels 1-2 mm. long; sepals 5, broadly triangular, about 0.75 mm. long; petals 5, white or pale yellow, elliptic, 1.5-2.5 mm. long; stamens 5, 3-4 mm. long, anthers 0.5—1 mm. long; disc flat; rudimentary carpels 3, about 1 mm. high. Car- pellate flowers 2-3 mm. long; pedicels, sepals, petals and disc as in stami- nate flowers; rudimentary stamens absent; gynoecium 3-carpellate, 1.5—2 mm. high, styles and stigmas coherent, the combined stigmatic structure peltate. Fruiting pedicels 1-3 (4.5) mm. long; follicles subglobose, about 3.5 mm. in diameter, in 3’s, 2’s or single with 0, 1 or 2 undeveloped carpels, respectively. China. Kwancst Province: N Hin Yen, Yeo Mar Shan, Ching 7102 6 (A, us), CHEKIANG Province: Shihpu, Chiao NU 14130 (a); without definite local- 4142 8 (a), Wilson 9440 2 (a). Japan. Honsuv IsLanp: Nakatsugawa, Nojiri, Shinano, Jack, September 6, 1905 2 (a, GH); Yokohama, Maximowicz, 1862 [cH (6), us (2)]; Shima, Toba, Sakurai, August 2, 1908 (a); Mino, Shiota 4459 (a); Oshima Island, Mizushima 573 9 (a). Kyusuu Istanp: Naga- saki, Moximowicz, 1863 ¢ (us), Wilson 6313 2 (a). Without definite locality: Sargent, August 26, 1892 ¢ (a), Siebold (xk — sketch of holotype of Zanthoxy- 1966] HARTLEY, MALESIAN ZANTHOXYLUM 189 lum emarginellum), sist (m—isotype of Zanthoxylum ailanthoides, not seen; A, photo.). Bonin Islands. a i Gonzales, 1917 (a), Otomo, May 4, 1917 é (A); Hales hon. Wilson 8265 (a—holotype of Zanthoxylum ailanthoides Eo inerme; US — isotype) ; ae definite locality, Wright 41 2 GH, US). Ryukyu Islands. Yaeyama Gunto, Iriomote Island, Funauki Bay, Walker & Fle 6762 (US). Taiwan. Kelung, Ford, June, 1884 (k — syntype of Zanthoxylum hemsleyanum, not seen; A, photo.), June 14, 1884 (a, photo.) ; South Cape, Henry 1353 8 (a); Tamsuy, Oldham 58 (cx); Taihoku vicinity, Shimada 39 & (a), Wilson 11220 2 (a); without definite locality, seat 1630 2 (A, US). Philippines. BaTAN eee Mt. Iraya, Ramos BS 80273 8 (kx, NY); Basco, Ramos BS 80381 & (Ny). Cultivated. Japan, hig pl July 25, 1958 4 (us), Terakawa, June 23, 1946 (us). DIsTRIBUTION. China (Kwangsi and Chekiang Provinces), Korea Japan, Bonin and the Ryukyu Islands, Taiwan and the oo (Batan Island) ; in lowland forests at altitudes up to 300 meters. See MA - 30 5 0 a Map 4. Distribution of Zanthoxylum ailanthoides Sieb. & Zucc. ILtustrations. Liv, T. S., Illustrations of native and introduced ligneous plants of Taiwan 2: f. 727. 1962. MAKINo, T., Illustrated flora of Japan, ed. 2. ¢. 1173. 1940, as Fagara ailanthoides; revised ed. t. 1173. 1949; New illustrated flora of Japan ¢. 1328. 1961. The wood is used in the Bonin Islands to make canoes Zanthoxylum ailanthoides is closely related to both Z. myriacanthum and Z. micranthum Hems]. The latter species, known from Hupeh, Hunan, and Szechuan Provinces, China, differs in having more slender branchlets and smaller leaves. 190 JOURNAL OF THE ARNOLD ARBORETUM [VOL. 47 Li, Woody flora of Taiwan 372. 1963, has cited Zanthoxylum emarginel- lum and Z. hemsleyanum, both based on collections from immature plants, as synonyms of Fagara ailanthoides. 5. Zanthoxylum avicennae (Lam.) DC. Prodr. 1: 726. 1824. Fagara avicennae Lam. Encycl. 2: 445. 1788. Type: d’Incarville 179, China (pro pa Kwangtung Province, fide Merrill, Comm. Lour. FI. Cochinch. 218. 1935). ster is wioieuleaaeg Champ. ex Benth. Jour. Bot. Kew Misc. 3: 329. 851. Type: Champion, Hong Kong (not seen). Zanthoxylum atid Mig. Ann. Mus. Lugd.-Bat. 3: 246. 1867. Type: Teys- mann HB 5167, Moluccas, Tidore Islan Zanthoxylum ie Warb. in Engl. “Bot. Jahrb. 13: 339. 1891. Type: Warburg 20145, Moluccas, Kai Islan Fagara diversifolia (Warb.) Engl. Nat. PHanzentan. III. 4: 118. 1896. Zanthoxylum iwahigense Elmer, Leafl. Philip. Bot. 5: 1833. 1913. Type: Elmer 12751, Philippines, Palawan Island. fcancent (rarely) or erect shrubs or small trees to 13 m.; dioecious; evergreen; branchlets generally armed, the prickles straight and scattered mm. on either side; petiolules 2-5 mm. long; leaflets 2-11 pairs, suboppo- site, chartaceous to subcoriaceous, dark green above, pale below, generally blackening somewhat in drying, pellucid dotted, ovate to elliptic- lanceolate 1-8 cm. long, 0.75—3 cm. wide, base obtuse to cuneate, main veins 4—11 o each side of the midrib, margins subentire to finely glandular crenate nih as many as 8 crenations per cm., apex obtuse to acuminate. Inflorescences terminal and occasionally from the upper leaf axils, cymose, the primary branches generally whorled, 5—21 cm. long, flat-topped, glabrous. Staminate flowers 1.5~3 mm. long; pedicels 0.5—-3 mm. long; sepals 5, broadly tri- angular or rounded, 0.5—0.8 mm. long; petals 5, white to greenish yellow, elliptic, 1-2.5 mm. long; stamens 5, about 2 mm. long, anthers about 0.75 mm. long; disc flat; rudimentary carpels 2, about 0.75 mm. high. Carpel- late flowers 1.5-3 mm. long; pedicels, sepals and petals as in staminate flowers; rudimentary stamens about 0.5 mm. long; disc pulvinate, about 0.3 mm. high; gynoecium 2-carpellate, about 1 mm. high, styles and stigmas coherent, the combined stigmatic structure peltate. Fruiting pedicels 1-3 mm. long; follicles subglobose, about 4.5 mm. in diameter, in pairs or occasionally single and paired with an undeveloped carpel. Thailand. Dauang-Quinom, Smitinand & Abbe 24842 2 (L). North Vietnam. Ninh Binh Province, Cho Ganh, Petelot 822 2 (a), 1676 2 (a); Lontoy Prov- ince, Petelot 2591 6 (a); Ha-coi, Taai Wong Mo Shan, Tsang 27016 @ (A), 29377 8 (A, SING), 29549 8 (a, SING); pt bas Kau Nga Shan, Tsang 27392 2 (A), 30494 8 (a, snc); Dam-ha, Sai Wong Mo Shan, Tsang 30448 8 (A, SING). South Vietnam. Mt. Bana, 25 km, 7 of roaniae (Da Nang), J. & M.S. Clemens 4067 & (a, vs), 4445 2 (a, us). China. Kwanocst Province: S$ Nan- ning, Seh-feng Dar Shan, Ching 7827 2 (a, oy Sun-to District, Po Yam Shan, 1966] HARTLEY, MALESIAN ZANTHOXYLUM 191 (A). Kwanctunc Province: Canton and vicinity, gine - Q (a), Levine CCC 170 @ (a, GH, us), CCC 563 2 (a, cu, us), CCC 8 é (A, GH, us), CCC 1814 2 (A, GH, US), CCC 3134 8 (a, cH), Merrill os Ee 2 (a); Wung-yuen oe ib Yung-yun, Lau 632 2 (a); Wong Chuk I, Tsing Wan Shan, Lan 2495 g ; Kao-yao District, Ting Woo Shan, Lau 20292 4 (sinc); Wah Shau Toi, pei & McClure CCC 6913 & (us); Pakhoi, Playfair, April, 1883 @ (cH); Hwei-yang District, Sam Hang Shek Tau, Lin Fa Shan, Tsang 25872 9 (A); Fang Cheng District, Taan Faan, Kung Ping Shan, Tsang 26868 6 (a); Lo- ee Phd 1645 2 (A, SING, US); Yan Fon District, West River, Wang 322 2 (A). HaINaN IsLanp: Nodoa, Chun NU 5825 2 (us); Dung Ka, Chun & Tso 43401 2 (A); Tai Pin, Gressiti 1117 é (A); Yaichow, How 70939 & (a); Nagai District, Yeung Ling Shan, Lau 133 2 (a, us); Chang-kiang District, Ka Chik Shan, Law 3085 2 (a); Kan-en District, Chim Fung Mt., Law 4932 2 (a); Loktung, Lau 27140 8 (a); Pak Shik Ling vicinity, Ku Tung Village, Lei 55 9 pa US), 876 2 (A, SING, US); San Hu, Tin Si, McClure CCC 7731 9 (a); m Ko District, Lin Fa Shan, Tsang 213 6 (a, us); Hung Mo Shan, Tsang 787, é (A); without definite locality, Liang 62183 6 (us), 63604 2 (us). Hok-chang, Ling-soik Temple, Tang 16055 & (A). Without definite orig Father d@’Incarville 179 @ (p—holotype of Fagara avicennae, not seen photo.). Java. East Java, Madiun Province, Ngebel, Koarders 232088 ay. Lesser Sunda Islands. Lombok Island, Rindjani-vulkangebirge, Elbert 1582 [A (3), cans (2), L (2)]; Soemba Island, de Voogd 1991 @ (L). British North Borneo (Sabah). Mt. Kinabalu, Penibukan, J. & M.S. Clemens 31036 & ( A, K, 77626 (Ny); Mt. Manalsal, Edano BS 77760 4 (xy, sING); Mt. Pulgar, Elmer 12751 2 (isotypes of Zanthoxylum iwahigense: A, BM, GH, K, L, NY, US); Victoria Mts., Panacan, Aborlan, Sulit PNH 12396 (a, L, SING); without definite oe Curran FB 3860 2 (Ny), Foxworthy BS 905 2 (GH, K, NY, US). BATAN Istanp. Mt. Iraya, Ramos BS 80214 2 (NY, SING). BaBUYAN IsLaNbs. Dalupiri Island, ‘Bartlett 15109 [a (8 & 2), MicH (@)], 15133 3 (a, MICH), 15827 2 (A, MICH). Luzon Istanp. Benguet Subprovince: Pauai, Santos BS 32025 ¢ BRI, NY, SING). hd Ecija Province: Mt. Umingan, Ramos & Edano BS 26303 2 (A, BM, US). Zambales Province: _. BS 5090 2 (Ny, US). Rizal Province: Antipolo, Ramos BS 20988 @ (A, GH, K, L, NY, SING, US); with- out definite locality, Loher 14840 4 (a), Merrill ‘SB 1002 (A, BM, GH, L, NY, US), SB 1060 M, GH, L, NY, US). Camarines Province: Simeon FB 28756 2 (A, K, US). Without definite locality : Loher 241 (xk), Vidal 1445 & (k). a DANAO IsLAND. Misamis Province: Caster FB 29749 @ (ny), Cuming 1622 4 M Cotabato Province: Ferraris FB 23044 @ (a, BM, us), Ramos & Edano B 84843 8 (a). Moluccas. Taulaud Island, Karakelong, Lam 3254 2 (A, L); Morotai Island, Kostermans 1481 & (BRI, K, L, SING); Kahatola one off Halmaheira, Fairchild 3484 6 (a); Tidore ‘Island, Teysmann HB 5 3} (isotypes of Zanthoxylum tidorense: K, L, MEL), de Vriese & Teysmann, seed 192 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 1860 (L); Ternate Island, de Vriese, 1857-1861 3 (L), de Vriese & Teysmann, 1859-1860 [a (2), cans (2)]; Kai Island, Beccari, August, 1873 [a (¢ & 9), K (2)], Jaheri HB 430 2 (1), Warburg 20145 (a—isotype of Zanthoxylum diversifolium). DistRIBUTION. Thailand east to Fukien Province, China; south in Java, Lesser Sunda Islands, British North Borneo, the Philippines and the Moluccas; in dry forests, thickets and on open slopes at altitudes up to 1630 meters. See MAP 5. ac. < | abs Ges’ ee hee Map 5. Distribution of Zanthoxylum avicennae (Lam.) DC. ILLUSTRATIONS. PrerReE, L., Flore forestiére de la Cochinchine 4: ¢. 289. 1893, as Zanthoxylum avicennae vars. touranense Pierre and tonkinense Pierre. I have not seen the type of Xanthoxylum lentiscifolium which is listed as a synonym of Xanthoxylum avicennae by Bentham in Flora Hong- Rongensis 58. 1861. Plants from dry, open situations and high mountain areas tend to become dwarfed or scandent with considerably reduced leaves and in- florescences, and more or less recurved, predominantly pseudostipular prickles. Some of the extremes of this type, such as Smitinand & Abbe 24842, from Thailand, Koorders 232088, from Java, J. & M. S. Clemens 40530, from Sabah and Elmer 12751, from Palawan Island, look strik- ingly different from the arborescent form of this species. They are not geographically isolated, however, and there are numerous intergradations. The Elmer collection cited here is the type of Zanthoxylum iwahigense, which has already been reduced to synonomy of Zanthoxylum avicennae 1966 | HARTLEY, MALESIAN ZANTHOXYLUM 193 by Merrill, Enum. Philip. Fl. Pl. 2: 326. 1923. Fagara anisata Back. ex Bakh. f., listed as a provisional name by Bakhuizen, Fl. Java 2: 97. 1965, also seems to represent this scandent form. I have not seen the collection cited (Backer 11431), but the description fits reasonably well and the author does not include Zanthoxylum avicennae 6. Zanthoxylum megistophyllum (Burtt) Hartley, comb. nov. Fagara megistophylla Burtt, Kew Bull. 1935: 301. 1935. Type: Waterhouse B344, Solomon Islands, Bougainville Island. Small trees to 8 m.; dioecious. Leaves imparipinnate, 80-175 cm. long, glabrous; petiolules of lateral leaflets obsolete to 5 mm. long; leaflets 4—5 pairs, opposite, chartaceous, dark green above, pale below, pellucid dotted, elliptic, 20-35 cm. long, 9-16.5 cm. wide, base obtuse to rounded, main veins 13-16 on each side of the midrib, margins entire, apex short acuminate. Inflorescences terminal or terminal and from upper leaf axils, paniculate, to 25 cm. long, broadly spreading, the axes and branches glabrate. Staminate flowers 4-5 mm. long; pedicels 2-3 mm. long; sepals 4, connate in the lower half, triangular, about 1 mm. long; petals 4, white, elliptic, 3-4 mm. long; stamens 4, 4-5 mm. long, anthers 0.5 mm. long; disc flat; rudimentary carpel 1, globose, 0.5 mm. high. Car- pellate flowers 3-4.5 mm. long; pedicels, sepals, petals, and disc as in staminate flowers; rudimentary stamens about 0.3 mm. long; gynoecium 2-carpellate, about 2 mm. high, styles and stigmas coherent, the combined stigmatic structure Lao Fruiting pedicels 2-5 mm. long; follicles subglobose, 9-12 mm. in diameter, in pairs or occasionally single and paired with an Weeicliaan carpel. Papua. CENTRAL District: Kairuku Subdistrict, near Maipa Village, Darby- shire 934 2 (CANB, L), 998 & (CANB, x); Sogeri Region, central position, Forbes 2 (BM). Pay cogs Islands. BoUGAINVILLE IsLAND: Kugumaru, Buin, Ka- jewski 1835 Q (a); Siwai District, Waterhouse Y167 2 (BRI, NY); without definite locality, Waterhouse B344 2 (isotypes: A, BRI). CHOISEUL ISLAND northern part, Whitmore’s Collectors BSIP 5661 2 (L). DistR1BUTION. Solomon Islands and Papua; in well drained, primary rain forests at altitudes up to 600 meters A very distinctive species with probably the largest leaves of any Zanthoxylum in the Australasian-Pacific region. Additional collections are desirable since at present we lack flowering material from the Solo- mons and fruiting material from Papua. 7. Zanthoxylum forbesii Hartley sp. nov. Arbor magna; foliis paripinnatis, 40-50 cm. longis, glabris; foliolis in paribus 3-4, oppositis, chartaceis, ellipticis, 15-25 cm. longis, 5-9 cm. latis, basi obtusis, apice breviter acuminatis, margine gunnnes venis pri mariis utrinsecus 12— 15, petiolulis 0.7-1 cm. longis; infructescentiis 194 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 terminalibus et axillis superioribus orientibus, paniculatis, 8-14 cm. longis, patentibus, glabris; = ellis 2-3 mm. longis; sepalis persistentibus 4, late triangularibus, 1 mm. hae folliculis subglobosis, ca. 1 cm. diametro, in ae: ‘vel interdum folliculo singulo et cum carpello abortivo; floribus non visis. Holotypus: Forbes 1707 (GH). Fic. 1. Fic. 1. Zanthoxylum forbesii Hartley. Photograph of type (GH), x 3/8. 1966 | HARTLEY, MALESIAN ZANTHOXYLUM 195 Sumatra. potent Lampongs, altitude 138 m., Forbes 1707 2 (cH — holotype; BM, SING — isotypes). Apparently allied to Zanthoxylum megistophyllum from which it differs in having smaller leaves, fewer leaflets, and longer petiolules. 8. Zanthoxylum backeri (Bakh. f.) Hartley, comb. nov. Fagara backeri Bakh. f. Blumea 6: 366. 1950. Type: Backer 17017, Java. Liana 15 m.; branchlets sparsely armed with scattered, retrorse prickles. Leaves imparipinnate, to 30 cm. long, glabrous; petiolules about 3 mm. long; leaflets 6-8 pairs, opposite or subopposite, subcoriaceous, elliptic- ovate, 6-8 cm. long, 2.6—3.2 cm. wide, base oblique, main veins 10-12 on each side of the midrib, margins essentially entire, apex acuminate, the acumen about one-fourth the length of the blade. Infructescences terminal and from upper leaf axils, paniculate with opposite branching, 10 cm. (to 30 cm., fide Bakhuizen f.) long and broadly sahiionag 8 se axes and branches glabrate. Fruiting pedicels about 1 m persistent sepals 4, triangular, about 1 mm. long; follicles eran about 1 cm. in diameter, i in pairs or single and paired with an undeveloped carpel. Flowers not known. Java. West Java, Preanger Regencies, Lengkong, in forest at 700 m., Backer 17017 2 (L— isotype). Although known from only a single collection, this species seems to be closely related to Z anthoxylum retroflexum of Sumatra. 9. Zanthoxylum retrofiexum Hartley sp. nov. Frutex scandens; ramulis et foliorum rhachibus parce aculeatis, aculeis retrorsis; foliis imparipinnatis, 15-30 cm. longis, glabris, acute reflexis, ad basin rhachibus pulvinatis: foliolis in paribus 4-5, oppositis, subcoriaceis, pellucido-punctatis, elliptico-oblongis, 4-7.5 cm. longis, 1.7—2.5 cm. latis, basi acutis vel leviter obliquis, apice acuminatis, acumine 6-9 mm. longo, ad apicem plerumque leviter emarginato, margine integris vel parce glandu- loso-crenatis, venis primariis utrinsecus 12~15, petiolulis 3-5 mm. longis; paniculis fructiferis terminalibus et axillis superioribus orientibus, 10-12 cm. longis, 5—6 cm. latis, ramulis oppositis, axi et ramulis glabratis; sa cellis ca. 1 mm. longis; sepalis persistentibus 4, triangularibus, ca. 1 m longis; folliculis subglobosis, 6-7 mm. diametro, in paribus vel eadin singulis cum carpello abortivo; floribus non visis. Holotypus: Yates 1850 (mMicH). Fic. 2. matra. Asahan, Bandar Poelo, in forest at 250 m., Yates 1850 (mIcH — ce BRI, K — isotypes). The sharply reflexed leaves of this species are not found in the other 196 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 FLORA OF. : BERS ROOVERGEE Satta, Rist . Fic. 2. Zanthoxylum retroflexcum Hartley. Photograph of type (miIcH), X 0/3. Malesian species of Zanthoxylum. Z. backeri, of Java, appears to be closely related but also differs in having longer acuminate leaflets with fewer main veins on either side of the midrib and larger follicles. 1966] HARTLEY, MALESIAN ZANTHOXYLUM 197 10. Zanthoxylum limonella (Dennst.) Alston, in Trimen, Handb. FI. Ceylon suppl. 6. 37. 1931. Tipalia limonella Dennst. Schliiss. Hort. pee) 31. 1818. Type: plate and description, Rheede, Hort. Malab. 5: 67. t. Fagara budrunga Roxb. Fl. Ind. ed. Carey & a 1: 437, 1820. Lectotype: Roxburgh Icones 2113, cult. Bot. Gard. Calcutta. Fagara rhetsa Roxb. Fl. Ind. ed. Carey & i. 1: 438. 1820. Lectotype: Roxburgh Icones 185, cult. Bot. Gard. Calcu Zanthoxylum budrunga (Roxb.) DC. Prodr. i Oe 1824, sub “species non satis notae.’ Zanthoxylum rhetsa (Roxb.) DC. Prodr. 1: 728. 1824 Zanthoxylum rhetsa var. budrunga (Roxb.) Pierre, Fl. F orestiére Cochinch. 4: t. 290. Zanthoxylum budrunga var. paucijuga Koords. & Val. Booms. Java 4: 224. 1896; Exk. Fl. Java 2: 418. 1912, pro syn. sub Zanthoxylum rhetsa (Roxb.) D . Syn ntypes: Koorders 69838 & 153538, Java (not seen). Zanthoxylum? minahassae Koords. Fl. N.O. Celebes 639. 1898; Koorders- Schumacher, Syst. Verzeich. 3: 59. 1914. Syntypes: Koorders 162588 (not seen), 187738, 187748 (not seen) and 187758, Celebes. Trees to 35 m.; dioecious; deciduous; rather long bole and spreading crown; main stem generally with broad, conical spines 2-3 cm. long; branchlets usually sparsely armed with straight or ascending prickles, often swollen and hollow and apparently housing ants. Leaves paripinnate or imparipinnate, 30-40 cm. long, glabrous; petiolules of lateral leaflets 3-5 mm. long; leaflets 5-8 pairs, opposite or subopposite, chartaceous, occasionally with scattered pellucid dots, ovate to elliptic, 7-13 cm. long, m. wide, base oblique, main veins 10~15 on each side of the midrib, margins entire to glandular crenate with as many as 3 crenations per cm., apex acuminate, the slender acumen at least one-eighth the length of the blade. Inflorescences terminal and from upper leaf axils, paniculate, 8-14 cm. long and generally about as broad, the axes and branches glabrous or glabrate. Staminate flowers 1.5-2.5 mm. long; pedicels 1-2 mm. long; sepals 4, green, rounded to triangular, 0.5~1 mm. long, connate at about half their length or free to the base, valvate, margins subentire or fim- briate; petals 4, white or yellowish white, elliptic, 1-2 mm. long; stamens 4, about as long as the petals at anthesis, anthers about 1 mm. long; disc pulvinate, irregularly lobed, about 0.5 mm. high; rudimentary carpel about 0.5 mm. high. Carpellate flowers 1.5-2.5 mm. long; pedicels, sepals and petals as in staminate flowers; rudimentary stamens absent; disc pulvinate, about 0.25 mm. high; gynoecium 1-carpellate, about 1 mm. high, style excentric, stigma flattened. Fruiting pedicels 2-4 mm. long; follicles single, globose, 6~7 mm. in diameter. India. Bombay PresIDENCY: region E of Goa boundary, Fernandes 691 & (A). Assam: Garo Hills, Phulbari, Chand 3113 6 (aticH); Rani, Koelz 29225 2 (uicH); without definite locality, King’s Collector 189 3 (A, us). Without definite locality: Wight 353 9 (GH). Ceylon. Thwaites CP 3490 2 (GH). Thailand. Hue Pang Huang, Kerr 5473 4 (SING). South Vietnam. Bien Hoa 198 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 Province, Bao Chiang, Bois 97 2 (a), Pierre, July, 1877 & (a). Malay Peninsula. Kelantan, Kota Bharu, Corner SF 33458 [a (6), stNc (2 sheets, 6 and @)]. Java. West JAVA: Samaranx Residency, Koorders, February 28, 1892 2 (1), 1161C & (1), March 20, 1892 2 (1); Cheribon Residency, Kromong Range, Mt. Djaja, Steenis 12880 4 (1). CENTRAL JAVA: yagi: NIFS Ja 2933 (a, L). Eastern Java: G. Baloeran, Backer 24802 K, L); Poeger, Koorders 60958 (L); Djalipapak, Koorders 69978 (L), 69988 aa Besoeki, Koorders 118128 (L); Banjoemas, Koorders, December 8, 1891 6 (L); Djember, sires March, 1921 (L); Bodjonegoro, NJFS Ja 2229 (Lt). Peutyanc Istanp: Udjun Kulon Reserve, Kostermans, December 3, 1960 6 (A, L). KANGEAN eee Tjangkramaan, Backer 27639 2 (L). Philippines. PALAWAN ISLAND. Curran FB 3822 2 (K, Ny). Luzon IsLanp. Ilocos Norte Province: Addurn FB 22989 134 8 (a, K, L). Zambales Province: Curran FB 6920 & (wus). Bataan pipet Mt. Mariveles, Lamao River, Borden FB 1565 2 (BM, NY), FB 3051 é (ny, us); without definite locality, Ahern’s Collector FB 1455 @ (Ny, US). Rizal Province: Morong, Loher 5145 2 (Kk, us), Ramos BS 1421 (xy, Us), Vidal 2269 2 (xk); Antipolo, Merrill SB 872 (a, BM, GH, K, L, Ny, US), SB 961 6 (A, GH, K, L, NY, US); Bosoboso, Merrill 2706 & (Kk, Ny, US); San Mateo, Vidal 1015 Q (K, L); without pe locality, Ahern’s Giese FB 3105 @ (xy, us). Laguna Province: Los Bafios (Mt. Maquiling), Elmer 18231 ? (A). Batangas Province: Curran FB ree 2 (ny), Curran & Merritt FB 7715 2 (NY, US uezon Province: Tayabas, Ware FB 16 @ (BM, K, NY, US). Camarines Province: Curran FB 10695 & (us), FB 10726 6 & 9 (us). Albay Province: np ie 1218 6 (BM, K, L), 1356 6 (BM, L). MASBATE ISLAND. Clark FB 1710 2 (x, NY, US). NEcROS IstaAND. Curran FB 17437 @ (ws). TAWITAWI ISLAND. Ramos & Edano BS 44110 2 (A, BM, NY, US). BASILAN Istanp. Miranda FB 18971 2 (BM, K, US). MINDANAO IsLAND. Davao Province, Mt. Pagdaugan, Ramos & Edano BS 49631 & (A, NY, SING); Cotabato Province, Buayan, Ramos & Edano BS 85175 8 (a); Zamboanga, Whitford & Hutchinson FB 9067 @ (ny). Celebes. Minahasa Province Koorders 187738 (L), 187758 (L)—syntypes of Zanthoxylum minahassae; Paria van Longkang naar Giliraeng, Noerkas 154 a Ae Moluccas. Wetar Island, Tara, NIJFS bb 27284 92 (kK, L, SING). Western District, Mabaduan, Brass 6491 2 (A, BRI, L). Cultivated. a “Calcutta, Roxburgh Icones 185, drawn from a plant from Sikar, India (k — lectotype of Fagara rhetsa, not seen; A, photo.), Roxburgh Icones 2113, drawn from a plant from Silhet, India (k —lectotype of Fagara budrunga, not seen; A, photo.) DistripuTion. India, Ceylon, Thailand, South Vietnam, Malay Penin- sula, Java, Philippines, Celebes, Moluccas (Wetar Island) and southern Papua; in rather dry, often monsoonal, forests and thickets at altitudes up to 500 meters. See MAP ILLUSTRATIONS. Koorpers, S. H., & TH. VALETON, Atlas der baumarten von Java 2: t. 352. 1914, as Fagara rhetsa. Pierre, L., Flore forestiére de la Cochinchine 4: t. 290 & 291. 1893, as Zanthoxylum rhetsa var. bu- drunga. As was pointed out in the discussion of the genus, the leaflets of this species are quite variable in the occurrence of oil dots and marginal crena- tions. This has been a source of confusion between Zanthoxylum budrunga Map 6. Distribution of Zanthoxylum limonella (Dennst.) Alston. and Z. rhetsa (see Hooker, f., Fl. Brit. Ind. 1: 495. 1875), although the main difference Roxburgh had in mind was the number of leaflets, also a variable character. Koorders’ Zanthoxylum? minahassae, based on sterile and apparently juvenile material with prickly leaf rachises and pellucid-dotted, crenate leaflets, is placed here with considerable uncertainty. It could easily be young material of the closely related Z. celebicum, which occurs in the Same area in the northeast Celebes as Z. limonella. 11. Zanthoxylum parviflorum Benth. Fl. Austral. 1: 363. 1863. Lectotype: Armstrong 569, Australia, Northern Territory. Fagara parviflora (Benth.) Engl. Nat. Pflanzenfam. III. 4: 119. 1896. Trees to 20 m.; dioecious; deciduous; main stem with a few scattered prickles; branchlets unarmed or sparsely prickly. Leaves imparipinnate, 22-28 cm. long, glabrous; petiolules of lateral leaflets 1.5-3 mm. long; leaflets 5-8 pairs, opposite, chartaceous, with scattered pellucid dots, elliptic-oblong, 5-8 cm. long, 2—3.5 cm. wide, broadest below the middle, base rounded, main veins 8-11 on each side of the midrib, margins glan- dular crenate with 2-3 crenations per cm., apex acuminate, the acumen one-fourth to one-fifth the length of the blade. Inflorescences terminal and from upper leaf axils, paniculate, 6-8 cm. long and about as broad, the axes and branches glabrate, Staminate flowers 2-3 mm. long; pedicels 2-4 mm. long; sepals 4, green, rounded to triangular, about 1 mm. long, free to about the base, valvate, margins subentire; petals 4, white, elliptic, 2 mm. 200 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 long; stamens 2~3 mm. long, anthers about 1 mm. long; disc pulvinate, 0.5-1 mm. high; rudimentary carpel 0.25—0.5 mm. high, poorly differen- tiated. Fruiting pedicels 3-5 mm. long; follicles single, globose, about 7 mm. in diameter. Carpellate flowers not seen. West New Guinea (West Irian). Merauke Area, Versteeg 1910 4 (XK, L). Australia. NORTHERN TERRITORY: Port Essington, Armstrong 569 6 (MEL — lectotype); Darwin, Bleeser 532 8 (MEL), Holtze 160 6 (MeL); islands in the Gulf of Carpentaria, Brown 5334 9 rn MEL); Golburn Island, Cunning- ham 181 2 (BM, MEL); Arnhem Land, Liverpool River, Gulliver, October, 1867 & (MEL); without definite locality, Pri igseiae (us DisTRIBUTION. West New Guinea and Northern Territory of Australia; in monsoon forests at low elevations. Bentham cited three collections in the original description — Armstrong 569, Brown 5334, and Cunningham 181. The most complete of these, Armstrong 569, is designated here as the lectotype. This species is very similar to Zanthoxylum limonella, differing mainly in the smaller leaflets with rounded rather than oblique bases and smaller inflorescences. The New Guinea collection, Versteeg 1910, is somewhat intermediate between Z. parviflorum and Z. limonella. 12. Zanthoxylum celebicum Koords. Fl. N.O. Celebes 639. 1898; Koorders-Schumacher, Syst. Verzeich. 3: 59. 1914. Lectotype: Koorders 187468, Celebes. Trees to 35 m., dioecious; evergreen; branchlets with numerous poner or ascending prickles, Leaves es paripinnate or imparipinnate, 40-50 c long, glabrous or puberulent on the rachises, petiolules and lower eh ae petiolules of lateral leaflets 3-5 mm. long; leaflets 6-10 pairs, opposite, subcoriaceous, elliptic-oblong, 10-20 cm. long, 4—6 cm. wide, base oblique, main veins 14—20 on each side of the midrib, margins entire or with a few glandular crenations near the apex, apex abruptly acuminate, the acumen about one-twelfth the length of the blade. Inflorescences terminal and from upper leaf axils, paniculate, 15-18 cm. long and about two-thirds as broad, the axes and branches densely puberulent. Staminate flowers (seen only in bud) about 2.5 mm. long; pedicels 1-2 mm. long; sepals 4, broadly rounded, about 1 mm. long, slightly imbricate basally, margins ciliate; petals 4, elliptic-ovate, about 2 mm. long; stamens 4, 1.5 mm. long, anthers about 1 mm. long; disc pulvinate, about 0.5 mm. high; rudimentary carpel about 0.3 mm. high. Fruiting pedicels 0.2-1.2 mm. long; follicles (premature) single, ovoid, about 7 mm. long and 4 mm. wide. Carpellate flowers not seen. Celebes. Near Manado, Minahasa Province, vibe pet vith fae sen8 type), 187478 (L), 187728 (L), NIFS bb 19450 8 (a, L). Moluc Mororal IsLanD: Subdistrict Tobelo, N Tjao, Tamngkilisan NIFS bb 33848. (a, SING) ; Sangowo River, Kostermans 1426 2 (1, sING). 1966 | HARTLEY, MALESIAN ZANTHOXYLUM 201 DistrIBuTION, NE Celebes and Morotai Island in the Moluccas; in primary rain forests at elevations from 60 to 500 meters. Koorders’ description of this species was based on sterile material and no collections were cited. Later, Koorders-Schumacher (Syst. Verzeich. 3: 59. 1914) cited four Koorders collections for this species, all made in 1895 from the same general locality in the Celebes. I have chosen the lecto- type (Koorders 187468) from the three duplicates of these collections at Leiden, each of which bears the name Zanthoxylum celebicum in Koorders’ handwriting. It consists of a packet of leaflets, a section of a rachis with two leaflets attached and a vegetative stem apex. The two collections from Morotai Island have a more pronounced pubescence of the inflorescences and more clearly ciliate sepals than the only fertile collection (NJFS bb 19450) I have seen from the Celebes. The leaves, however, are a reasonably good match for Koorders’ type material of Zanthoxylum celebicum and I think there is little doubt that they belong here. 13. Zanthoxylum vinkii Hartley sp. nov. Arbor dioica sempervirens; ramulis parce aculeatis, aculeis rectis, inflatis et cavis et, ut videtur, a formicis inhabitatis; foliis paripinnatis, ca. 60 cm. longis, glabris; foliolis in paribus 10, oppositis, subcoriaceis, elliptico- oblongis, 14-19 cm. longis, 4-6 cm. latis, basi subcordatis, apice abrupte et breviter acuminatis, acumine 6-15 mm. longo, margine integris vel ad apicem parce glanduloso-crenatis, petiolulis 2-3 mm. longis; inflorescentiis paniculatis, terminalibus vel in axillis superioribus orientibus, 15-30 cm. longis latisque, glabris; floribus @ 3.2 mm. longis, pedicellis 0-2 mm. longis; sepalis 4, virentibus, late rotundatis, 0.6 mm. longis, margine ciliatis; petalis 4, flavis, ellipticis, 2.7 mm. longis; staminibus rudimen- tariis 0.1 mm, longis: disco pulvinato, 0.5 mm. alto; gynoeceo 1-carpellato, 1.3 mm. alto, stylo excentrico, stigmate 0.75 mm. lato; floribus ¢ et fructibus non visis. Holotypus: Vink BW 12065 (a). Fic. 3. New Guinea (West Irian). Schouten Islands. Brak ISLAND: Sues tree 12 m., common, Moll BW 7025 (L); Korim Road, old secondary pease forest on rocky soil, altitude 60 m., young tree 12 m., scarce, Vink BW 12065 (A— holotype; cans, Lt — isotypes). The subcordate leaflet bases of this species distinguish it from the closely related Zanthoxylum celebicum (which also differs in having puberulent rather than glabrate inflorescences) and Z. pluviatile. 14. Zanthoxylum pluviatile Hartley sp. nov. Arbor usque ad 35 m. alta, dioica, sempervirens; trunco spinoso vel exarmato, spinis magnis conicis, ramulis exarmatis vel parce aculeatis, aculeis saepe inflatis et cavis et, ut videtur, a formicis inhabitatis; foliis 202 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 MET RLAK Cia ae Sw Fem te LE ke Oe FH. Wi ldebrand 201962 We Pree Chenery iy ie: EP ma giens la yell stim, po re pon es TERBART ie on es ae : ug otigua ovnre ? pcthebglan. inh. Marthe, ARNOLD ARBO RETUM = red ¥ meses Pinswe sett em abe nti erce tee oo we wet Vw isi eee my Farner, 0 } “ Perecmived by Thamar G Wartey Now. Coated COMMENICAVIT HERR LOGD. SATAVOREM neni Fic. 3. Zanthoxylum vinkii Hartley. Photograph of type (a), X 3/8. impari- vel paripinnatis, 30-60 cm. longis, glabris; foliolis in paribus 5—12, oppositis, chartaceis, elliptico-oblongis vel anguste oblongis, 9-15 cm. longis, 2—4.5 cm. latis, basi obliquis vel rotundatis, margine integris vel glanduloso- crenatis, crenis 1 cm. distantibus, apice breviter acuminatis, acumine 5- 5 1966 | HARTLEY, MALESIAN ZANTHOXYLUM 203 13 mm. longo, petiolulis 2-5 mm. longis; inflorescentiis terminalibus vel in axillis superioribus orientibus, paniculatis, usque 15 cm. longis, patentibus, axi et ramulis glabratis; floribus ¢ 4-7 mm. longis, pedicellis m. longis vel nullis; sepalis 4, viridibus, late rotundatis, 1—2 usque petalis 4, albis, mm, eee ad basin leviter imbricatis, margine ciliatis; Fires; pa oe S wiite. Presetwad £5: Thomet G Hartley Fic. 4. Zanthoxylum pluviatile Hartley. Photograph of type 204 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 ellipticis, 3-5 mm. longis, plerumque ad anthesin erectis; staminibus 4-6 mm. longis, antheris 1.5—2 mm. longis inclusis; disco pulvinato tenuiter 4-lobato, 0.75-1.5 mm. alto; carpello rudimentario ca. 0.5 mm. alto; pedicellis usque ad 1 mm. longis; folliculo singulo, globoso, 7— 8 mm. diametro; floribus ¢ non visis. Holotypus: Hartley 11349 (a), Territory of New Guinea. Fic. 4. Papua. NorTHERN District: Kakoda, Carr 16422 2 (1). CENTRAL DISTRICT: Mt. Obree, Lane-Poole 392 (prt); Kanosia, Carr 11582 & (1); Boridi, Carr 13256 2 (CANB, K, L). MILNE Bay District: Milne Bay Area, Smith NGF 1287 & (1). Territory of New Guinea. Morose District: 10 miles N of Lae, Hartley 11349 4 (a—holotype; L— isotype); Lae Area, Vickery NGF 1405 (L). Solomon Islands. SANTA IsABEL IsLAND: Sigana, Brass 3449 8 (A, BRI, L, SING); Tatamba, Whitmore BSIP 2719 (L). GUADALCANAL: near Point Cruz, Walker & White BSIP 126 2 (A, BRI, CANB, L, US). DisTRIBUTION. Eastern New Guinea and the Solomon Islands; in well- drained, primary rain forests at altitudes up to 2100 meters A flowering collection from the Solomon Islands (Brass 3449) seems to indicate a tendency of this species to grade into Zanthoxylum limonella. In this collection the flowers are smaller and have less exserted stamens and more spreading petals than the New Guinea material of Z. pluviatile. Collections from elevations of 1500 and 2100 meters in Papua (Lane- Poole 392 and Carr 13256) differ from the lowland material in having smaller, more narrowly oblong leaflets and sparsely prickly branchlets. They may indicate a close relationship with Zanthoxylum cons persipuncta- tum of which there is a similar narrow-leaved form (Brass 22553 and 22919 from Mt. Dayman, Milne Bay District, Papua). 15. Zanthoxylum conspersipunctatum Merr. & Perry, Jour. Arnold Arb. 22: 33. 1941. Type: Brass 11579, West New Guinea. Trees to 25 m.; dioecious; evergreen; rather long bole and spreading crown; main stem prickly, becoming smooth with age; outer bark gray- brown, longitudinally fissured; inner bark yellow or yellowish brown; wood yellow; branchlets with straight or ascending prickles, often swollen and hollow and apparently housing ants. Leaves paripinnate or imparipinnate, 10-30 cm. long, glabrous; petiolules 3-5 mm. long; leaflets 2—8 pairs, opposite, subcoriaceous to coriaceous, generally lustrous above, with scat- tered and/or marginal pellucid dots, ovate to elliptic, 5-13 cm. long, 2-5 cm. wide, base oblique to obtuse, main veins 12—18 on each side of the midrib, margins subentire to glandular crenate with as many as 2 crena- tions per cm., apex acuminate, the acumen about one-tenth the length of the blade. Inflorescences terminal and from upper leaf axils, paniculate, 6-10 cm. long and generally about as wide, the axes and branches glabrate. Staminate flowers 3.5-7 mm. long; pedicels obsolete to 1 mm. long; sepals 4, green to pink, broadly rounded, 1—1.3 mm. long, free to about the base, valvate or slightly imbricate basally, margins entire or slightly fimbriate; 1966} HARTLEY, MALESIAN ZANTHOXYLUM 205 petals 4, white to pink, elliptic, 3-5 mm. long; stamens 4, about as long as the petals at anthesis, anthers 1.3—-2.5 mm. long; disc pulvinate, 4—lobed, 0.5—1 mm. high; rudimentary carpel 0.2-2 mm. high, well differentiated. Carpellate flowers 3.5—7 mm. long; pedicels, sepals and petals as in sta- minate flowers; rudimentary stamens 0.2-2 mm. long; disc pulvinate, slightly lobed, about 0.5 mm. high; gynoecium 1-carpellate, 1.5—2.5 mm. high, style excentric, stigma flattened, 0.5—1 mm. broad. Fruiting pedicels 1—2 mm. long; follicles single, globose to ovoid, 6-10 mm. in diameter. West New Guinea (West Irian). Bele River, 18 km. NE of Lake Habbema, Brass 11579 8 (a—holotype; L—isotype). Papua. Milne Bay District, Ma- Mt. Hagen, Womersley NGF 9418 6 (A, BRI, K, L). EASTERN HIGHLANDs D1s- TRIcT: Mt. Wilhelm, E slopes, Brass 30345 2 (CANB, K, L, US), 30496 $ (A, K, L, US), 30614 8 (A, K, L, NY, US); about 1 mile S of Akuna, Hartley 12027 2 (A, L); Marafunga, about 20 miles NW of Goroka, Hartley 13270 2 (a), Womersley & Sleumer NGF 14011 6 & 2 (cans); Asaro-Mairi Divide, near Daulo camp, Hoogland & Pullen 5471 2 (CANB), 5475 4 (A, BRI, K, L, US), Saunders 892 (CANB, L); ridge above Toromambuno Mission, Pullen 309 4 (A, BRI, K, L, US); about 5 miles W of Keglsugl airstrip, Saunders 770 (a, L); Wonatabe, 15 miles S of Okapa, Womersley NGF 17629 3 (BRI, CANB). Morose District: Yunzaing, J. & M. S. Clemens 3574 2 (A), 4060 2 (aA); Sambanga, Clemens 6813 & (A, L). DistriBuTIon. New Guinea; in mountain forests from 1500 to 2000 meters. A highly variable species with four morphological extremes possibly rep- resenting incipient species. Apparently closely related to Zanthoxylum pluviatile and also to Z. veneficum F. M. Bailey and Z. brachyacanthum F. Muell. (distinct?) of eastern Australia. The type collection of Zanthoxylum conspersipunctatum, Brass 11379, has much more conspicuous pellucid dots in the leaflets than any of the other collections cited for this species. Examination of cleared leaflets of this collection revealed that the dots are actually cavities in the mesophyll not traversed by the veins. The oil dots in the mesophyll of the other col- lections of this species examined proved to be small groups of cells not altering the venation pattern, though marginal dots in some of these were of the cavity type. As I have noted in the discussion of the genus, the presence and type of oil dots seem to be variable characters in the Malesian species of Zanthoxylum, and, without other morphological differences, I prefer not to make taxonomic distinctions. Of the species of Zanthoxylum in New Guinea, Z conspersipunctatum is by far the most frequent, being especially common in the mountain forests of the Eastern Highlands District. 16. Zanthoxylum integrifoliolum (Merr.) Merr. Enum. Philip. Fl. FL 23'327...1923. 206 JOURNAL OF THE ARNOLD ARBORETUM [ VoL. 47 Fagara integrifoliola Merr. Philip. Jour. Sci. Bot. 1(suppl.): 68. 1906. Lectotype: Barnes FB 88, Philippines. Trees to 20 m.; dioecious; evergreen; older, apparently exposed roots sometimes (always?) covered with dense, soft, yellowish fibers that appear to be phelloderm tissue; main stems with short, scattered prickles when young, becoming smooth with age; branches relatively few, widely spaced, horizontal with ascending tips; branchlets unarmed. Leaves paripinnate or imparipinnate, 20-60 cm. long, glabrous, crowded toward tips of branch- lets; petiolules of lateral leaflets 7-13 mm. long; leaflets 4-12 pairs, op- posite, chartaceous to coriaceous, bright green above, much paler below, obovate to elliptic, generally broadest above the middle, 6-15 cm. long, 4-6 cm. wide, base obtuse to cuneate, often rather oblique, main veins 9-15 on each side of the midrib, margins entire, becoming cartilaginous with age, apex abruptly acuminate, the acumen usually about one-twelfth the length of the blade. Inflorescences terminal and from the upper leaf axils, pani- culate, 12-46 cm. long and one-third to one-half as broad, the axes and branches glabrous or glabrate. Staminate flowers 4-5 mm. long; pedicels 2-4 mm. long; sepals 4, green, broadly triangular, about 1 mm. long, the margins entire; petals 4, white, elliptic-ovate, 2.5-3.5 mm. long; stamens 4, slightly exceeding the petals at anthesis, anthers about 1 mm. long; disc pulvinate, about 0.5 mm. high; rudimentary carpel conical, to 0.2 mm. high. Carpellate flowers 3-4 mm. long; pedicels, sepals, petals and disc as in staminate flowers; rudimentary stamens absent; gynoecium 1-carpellate, about 1 mm. high, style excentric, stigma flattened, about 0.5 mm. broad. Fruiting pedicels 3-5 mm. long; follicles single, subglobose, 6—9 mm. in diameter, with numerous, large oil glands. Philippines. Minporo Istanp. Mt. Yagaw, S slope, Conklin PNH 19019 (A, L); Bongabong River, Merritt FB 3659 2 (ny), Whitford 1464 2 (ny, US); Paluan, Ramos BS 39551 2 (A, BM, GH). BATANES IsLANDs. Batan Island, Mt. Iraya, Ramos BS 79957 2 (kK, NY); without definite locality, Fenix BS 3584 2 (xy). BapuyANn Istanps. Calayan Island, Penas FB 26698 9 (us). Luzon Istanp. Bataan Province: Mt. Mariveles, Lamao River, Barnes FB 88 4 (us —lectotype of Fagara integrifoliola; ing = ae sinc), Borden FB 2351 ¢ (NY, SING, US), FB 2484 [ny, . (2)], FB 2740 2 (ny, stnc, us), Foxworthy BS 1567 (us), Meyer FB 2307 & (k, NY, SING, ey Whitford 1044 9 (Ny, US), 1295 2 (kK, NY, us), Williams 575 92 (cH, NY, US); Mamac River, Barnes FB 336 2 (BM, K, NY, SING, US); without definite locality, Curran FB 17578 6 (1). Ramos BS 22465 2 (A, BM, BRI, L, US). Camarines Norte Province: Paracale, Ramos & Edano BS — is (A, x). Sorsogon Province: Irosin (Mt. Bulusan), Elmer 15566 2 (A, BM, GH, NY, L, US); without definite locality, Ramos BS 23421 6 (us). SAMAR istavo, Catubig River, Ramos BS 24171 @ (A, K, US), BS 24522 8 (1, us); Oquendo, Mt. Mahagna, Sulit PNH 14512 @ (A, K, L); without definite a id Lasquety FB 23569 8 (a, us), Ramos BS 1629 ? (BM, GH, L, SING). Leyte Istanp. Cabalian, Ramos BS 41586 2 (A, L). Panay IsLanp. Capiz Province: Magallanes (Mt. Giting-Giting), Elmer 12063 2 (A, BM, GH, K, L, NY, US); without definite locality, Edano BS 46165 8 (A, Ny). MrnpaNnao IsLanp. Surigao Province, Mt. Cantugas, Ramos & Convocar 1966] HARTLEY, MALESIAN ZANTHOXYLUM 207 BS 83490 & (ny); Davao Province, de Mesa FB 27604 2 (sinc); Zamboanga Province, Quezon FB 30252 (prt). Dinacat Istanp. Ramos & Pascasio BS 35248 2 (BM, 1L), Ramos & Convocar BS 83892 2 (ny). DiIstRIBUTION. Widespread in the Philippines in well-drained rain forests and thickets up to 640 meters elevation, Otherwise known from Lanyu Island (Botel Tobago) off the SE coast of Taiwan, fide Li, H. L., Woody flora of Taiwan, 373. 1963. ILLUSTRATION. IrTo, T., Illustrated flora of Formosa, ¢. 13. 1928. Closely related to Zanthoxylum limonella, Z. integrifoliolum is distin- guished by its obovate leaflets with entire, cartilaginous margins and un- armed branchlets. The soft, fibrous material from the roots has reportedly been used by the natives of Lanyu Island to calk canoes. 17. Zanthoxylum ovalifolium Wight, Illus. Ind. Bot. 1: 169. 1839 (“Zanthoxylon”). Type: Wight 356, India, Madras. Zanthoxylon sepearium Wight, Illus. Ind. Bot. 1: 169. 1839. Type: Wight 355, India, Madras. , Zanthoxylum ovalifolium var. sepearium (Wight) Hook. f. Fl. Brit. Ind. 1: 3. 1875. Fagara ovalifolia (Wight) Engl. Nat. Pflanzenfam. III. 4: 118. 1896. xylum inerme White & Francis Bot. Bull. (Dept. Agr. Brisbane) 22: 6. 1920 (nomen illegit.) non Sessé & Mocifio, nec Koidz. Type: White, January, 1918, Australia, Queensland. : Fagara varians Domin Bibliot. Bot. 22 (894): 846. 1927 (nomen illegit.) non enth. Type: Domin 5657, Australia, Queensland. Zanthoxylum dominianum Merr. & Perry, Jour. Arnold Arb. 22: 32. 1941 (based on Fagara varians Domin). Zanthoxylum suberosum White, Proc. Roy. Soc. Queensland, 53: 208. 1942 (based on Z. inerme White & Francis). Erect shrubs or small trees to 8 m.; branchlets unarmed or rarely with a few straight, scattered prickles. Leaves trifoliolate (occasional leaves bi- or unifoliolate), 8-28 cm. long, glabrous; leaflets chartaceous to subcoria- ceous, with scattered pellucid dots, ovate to elliptic, 3-19 cm. long, sessile or subsessile, base cuneate, main veins 7-20 on each side of the midrib, margins subentire to doubly crenate with about 5 crenations per cm., apex acute to short acuminate. Inflorescences axillary and terminal, paniculate, 3-12 cm. long and about half as broad, more or less flexuous, the axes and branches glabrate. Staminate flowers 2-3 mm. long; pedicels 1-3 mm. long; sepals 4, green, narrowly triangular, about 1 mm. long; petals 4, white, elliptic, about 2.5 mm. long; stamens 4 (in occasional pentamerous flowers, 5), 2-3 mm. long, anthers about 0.75 mm. long; disc pulvinate, about 0.75 mm. high; rudimentary carpel narrowly conical, about 0.5 mm. high. Carpellate flowers about 3 mm. long; pedicels, sepals and petals as in sta- minate flowers; rudimentary stamens 4, about 1 mm. long; disc pulvinate, about 0.5 mm. high; gynoecium 1-carpellate, about 1.5 mm. high, style 208 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 excentric, stigma globose. = pedicels 2-4 mm. long; follicles single, subglobose, 6—7 mm. in diamet Sikkim. Brandis 856 (A). India. Assam: Khasi Hills, Chand 6328 2 (micH), Hooker f. & Thomson 2 (cu), Koelz 30380 2 (micH), 30809 2 (micH), 31105 : (micH), Ruse 122 $ (a). Mapras: Nilghiri Mts., Hooker f. & Thomson & 2 (cH); Pulicat Hills, Wight 355 3 (x — holotype of Zanthoxylon bee GH); Shevagherry Hills, Wight 356 & & 2 (k—holotype of pir hairy segs arog GH). bho ao locality: Blinkworth (Wall Cat. n. 7469) 2 (Kk), Griffith 6 (cH). ma. Kachin Hills, Mokim, 1897 4 (A), s.m., 1898 2 (A). Andaman ie ” Heinig, December 12, 1900 2 (A), King, December 13, 1890 6 (sING), King’s Collector 102 9 (sInc), October 1, 1892 2 (stnc), Prain’s Collector, October, 1900 2 (GH). Java. WESTERN JAVA: Goenoeng Poetjah Tjatjing, Forbes 838 $ (8M, GH, sING); Bandoeng Tjigenteng, Koorders 262848 2 (A, L), 263338 2 (L); Kjaja, Lam 299 6 (1L). CENTRAL Java: Semarang, Ambarawa Telomojo, Koorders 279298 2 (1), 279388 2 (1). KANGEAN IsLANDS: Tambajangan, Backer 27465 2 (1), 27590 Q (1), 27892 (L), 27921 9 (1), 27968 2 (L), Beguin, May 24, 1919 (L); Kajoe Waroe, Backer 28122 2 (1). Lesser Sunda Islands. SOEMBAWA IsLAND: Batuduhang, Batu la nteh Mt., Kostermans 18075 2 (CANB, L). Trmor: Plateau of Fuiloro, Muapitine, Steenis 18153 6 (1); Mt. Perdido, Steenis 18317 & (1). Papua. WEsTERN District: Lower Fly River, east bank opposite Sturt Island, oe 8060 & (A, BRI, L); Wassi Kussi River, Tarara, Brass 8507 8 (A, L). CenTRAL District: Kanosia, Carr 11183 2 (a); Koitaki, Carr 11913 ; (a), 12594 2 (a); Boridi, Carr 14925 2 (a); Sogeri, Mt. Sawada, Forbes 875 2 (sm); tributary of Laloki River 2 miles E of Rouna, Hartley 10704 & (A). Without definite locality: Chalmers, 1885 2 (MEL). Australia. QUEENSLAND. Atherton Tableland: Bailey, June 27, 1899 92 (prt); Forest Reserve 45, Curry, September 20, 1922 2 (A); in hygrodrymio apun lacum Lake Eacham, Domin Map 7. Distribution of Sentilles ok. Wight. 1966 | HARTLEY, MALESIAN ZANTHOXYLUM 209 Ravenshoe, Millet Farm, Samundsett 22 2 (BR 1); near Atherton, White, January, 1918 9 (prt — ho lotype of Zanthoxylum inerme White & Francis; MEL). Cairns: Trinity Bay, Fitzalan, 1882 2 (MEL). DistripuTION, Sikkim and India southeast to Queensland; in monsoon forests and thickets at elevations up to 1750 meters in Malesia and from 1000 to 2000 meters in continental Asia. See MAP ILLUSTRATION. Wuite, C. T., & W. D. Francis (descr. of Zanthoxylum inerme), Bot. Bull. (Dept. Agr. Brisbane) 2a°6, 1920; Zanthoxylum ovalifolium is markedly different from the other Malesian species and is apparently most closely related to two species of southwestern China, Z. dimorphophylium Hemsl. and Z. robiginosum (Reeder & Cheo) Huang. The first of these, as I have pointed out in the introductory part of this paper, appears to be transitional between Fagara and Zanthoxylum, sensu stricto. 18. Zanthoxylum acanthopodium DC. Prodr. 1: 727. 1824. Type: Wallich, 1821, Nepal. Scandent or erect shrubs or small trees to 6 m.; dioecious or (in Su- matra) with perfect flowers; apparently both deciduous and evergreen; branchlets villous with ferruginous hairs to glabrate, generally armed, the prickles flattened, predominantly pseudostipular, to 1.2 cm. long. Leaves trifoliolate or imparipinnate, 2-25 cm. long; rachis villous to glabrate, often with flattened prickles, narrowly to broadly (to 3 mm. on either side) winged; petiolules obsolete to 3 mm. long; leaflets 1-6 pairs, opposite, chartaceous, villous to sparsely hairy on the main veins below, with ap- pressed hairs or glabrous above, often with flattened prickles along the midrib above and below, ovate to elliptic-lanceolate, 1-12 cm. long, 0.5—4.5 cm. wide, base obtuse, main veins generally conspicuous, 10— 28 on each side of the midrib, margins subentire to glandular crenate with as many as 8 crenations per cm., apex acute to acuminate. Inflore- scences axillary, paniculate to racemose, 0.5—2 cm. long and generally glomerate. Staminate flowers about 3 mm. long; pedicels obsolete to 2 mm. long; perianth segments 6-8, undifferentiated although occasionally slightly unequal in size, uniseriate to irregularly biseriate, green or yellowish green elliptic to ligulate, 1-2 mm. long, sparsely hairy to glabrous; stamens 6, about 2 mm. long, anthers about 1 mm. long and reddish purple prior to anthesis; disc pulvinate, about 0.75 mm. high; rudimentary carpels 2-5. Carpellate flowers about 2 mm. long; pedicels and perianth segments as in staminate flowers; rudimentary stamens none; disc pulvinate, 0.5-0.75 mm. igh; gynoecium 2-5-carpellate, sparsely hirsute to glabrous, about 1.5 mm. high, styles about 0.75 mm. long, divergent, articulating about 0.3 mm. below the globose stigma. Perfect flowers (only in Sumatran speci- mens) about 3 mm. long; pedicels and perianth segments as in staminate 210 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 flowers; stamens 3—6, about 3 mm. long, otherwise as in staminate flowers; gynoecium 2-4-carpellate, ce, hirsute, otherwise as in carpellate flowers. Fruiting pedicels 0.5-1. m. long; follicles generally reddish, subglobose, about 4 mm. in ee in 2’s to 5’s, the undeveloped carpels caducous. Popp Urrar PrapEsH: Kumaun, Jagesar, Strachey & Winterbottom 2 2 (cH). West Bengal: Darjeeling District, Smith, April 20, 1908 2 (A). pine Naga Hills, Chand 3376 2 (micH); Khasia, Hooker f. & Thomson & GH, us); Mawphlang, Koelz 31303 @ (micH); without definite locality, Simons 620 & (1). Nepal. Doti District, Ram, May 15, 1929 2 (a); without definite locality, Wallich, 1821 (G-pc — holotype of Zanthoxylum acanthopodium, not seen; A, photo. and fragment). Sikkim. Sureil, Smith 702 (a); without definite locality, Hooker f. & Thomson 2 (GH). East Pakistan (East Bengal). Griffith 1186/2 8 (cH). Burma. Thandaung, Dickason 5273 & (a), 6718 4 (A); Haka, Dickason 7563 2 (a); North Triangle, Kingdon-Ward 20736 °& (A). Thailand. Doi Sootep Mt. range, summit of Doi Chom Cheng, Rock 1166 . in PROVINCE: Likiang & vicinity, Ching 21045 4 (A), 21882 2 (a), Schneider 2877 8 (cu, us); S. Chungtien, Chiao-tou on Yangtze, Feng 3067 2 (A); Si-Chour Hsien, Faa-doon, Feng 11687 @ (a), 11807 6 (a); S of Red River from Manmei, Henry 9758 6 (A); Mengtsz, ow: 10150 [a (4), us (2 sheets, ¢ & 2 and 6)], 10150B 2 (a, us), 11413 2 (A); Szemao, Henry 12451 6 (A); Yunnanfu, Schoch 268 & (a, us); Peyen-tsien, Ten 502 ¢ (a, us); Wen-shan Hsien, Tsai 51691 2 (A); Kien-shuei Hsien, Tsai 53113 2 (a); Pin-chuan Hsien, Tsai 53686 & (A); Lung-ling Hsien, Tsai 55526 2 (a); Lan-ping Hsien, Tsai 56103 2 (A); Lu-se, Tsai 56423 9 (a); Che-tse-lo, Tsai 58566 2 (a); Wei-si Hsien, Tsai 59989 2 (a); Ping-pien Hsien, Tsai 60761 2 (A), 61922 4 (A); W of Champutung, Wang 67388 ¢ (a); Shunning, Yu 16576 6 (a); Mienning, Poshang, Yu 17849 & (a); Kuikiang Valley, N of Mon-ting, Yu 20391 2 (a); Muchietu, Yu 21009 ¢ (a); without definite locality, Forrest 7881 & (A), 8846 & (A), 11790 & (a), 15914 3 (a), Tsai 57118 2 (a), 57245 & (A), 60092 2 (A). SzECHWAN PRovINCE: Yalung, Handel-Mazzetti 482 @ (a); zwischen Ruapie und Takaioko, Schneider 1353 2 (A), 1373 (A); Huei-li Hsien, Yu 1607 2 (A). Sumatra. Gajo Loeas, Wenareh, Pringgo Atmodjo 55 % (L); Karo Highlands, Berastagi, Hamel & Rahmat Si Boeea 630 % (A, MIcH), Lorzing 6100 % (L), Yates 1548 % (A, BM, MicH); Karo Highlands, Kabandjahe, Lorzing 13687 % (1); Asahan, Rahmat Si Boeea 10558 % (a, 1); Atjeh, Boer ni Telong, Steenis 6342 3 (a L) DiIstRIBUTION. N India, Nepal, Sikkim, East Pakistan, Burma, Thai- land, SW China and Sumatra; in mountain forests and thickets at eleva- tions of 1200 to 1400 meters in Sumatra and up to 2900 meters in SW China. See map 8. iia: PIERRE, L., Flore Forestiére de la Cochinchine 4: ¢. 290. Similar to Zanthoxylum armatum but with relatively constant differences in the size and position of the inflorescence, number of lateral veins in the leaflets and the color of the anthers prior to anthesis. These are maintained even though the two species apparently grow together in Kumaun, Nepal, 1966 | HARTLEY, MALESIAN ZANTHOXYLUM 211 es ; - 5. ~ ou ie ae) 34 Ge ty yy, Oe. ‘ if G c ji oA ° pa ames al cae e Map 8. Distribution of Zanthoxylum acanthopodium DC. Assam, and Yunnan. Perfect flowers were noted only in material from Sumatra where the unisexual condition seems to be lacking entirely. This may indicate that the population of that island is a relict, particularly since there appears to be a considerable geographic discontinuity with the uni- sexual material of the Asiatic mainland. Aside from this, however, the Sumatran material matches the Asian very closely Two collections from Szechuan Province, China (Schneider 1353 and Handel-Mazzetti 2524), with dwarf leaves 2-3 cm. long and very short, compact inflorescences have been designated in error as Zanthoxylum acanthopodium var. diminutum (Rehd.) Reeder & Cheo, Jour. Arnold Arb. 32: 71. 1951. I have seen the holotype of Z. dimor phophyllum Hemsl. var. diminutum Rehd. (Jour. Arnold Arb, 22: 577. 1941) which clearly is most closely allied to Z. dimorphophyllum, a distinctive species restricted to China. This misjudgement by Reeder and Cheo apparently occurred because Rehder cited Schneider 1353 as a paratype of Z dimorphophyllum var. diminutum. 19. Zanthoxylum armatum DC. Prodr. 1: 727. 1824. Lectotype: Lambert, 1816, India, non Z. armatum ean Druce, 1917. car ae da 2! Roxb. Hort. Beng. 7 nomen nudum; Roxb. Fl. Ind. 8. 1832. Lectotype: pen bia pian 1916, cult. Bot. Gard. Calcut Lausicuie planispinum Sieb. & Zucc. Abh. Akad. Miinchen 4 (2): 138. 1846. Type: Siebold, Japan Zanthoxylum alatum Roxb. var. planispinum (Sieb. & Zucc.) Rehd. & Wils. Pl. Wils. 2: 125. 1914 212 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 Scandent or erect shrubs or small trees to 6 m.; dioecious; apparently deciduous or evergreen; branchlets glabrous to short pubescent, generally armed, the prickles flattened, predominantly pseudostipular, to 2 cm. long. Leaves trifoliolate or imparipinnate, 5—23 cm. long; rachis glabrous, often with flattened prickles, narrowly to broadly winged (to 6 mm. on either side); petiolules obsolete to 3 mm. long; leaflets 1—5 pairs, opposite, chartaceous to subcoriaceous, glabrous or occasionally with a few hairs along the midrib below, often with flattened prickles along the midrib above and below, ovate to lanceolate, 1.5—13 cm. long, 0.5—4.7 cm. wide, base cuneate to acute, main veins generally faint, 7-15 on each side of the mid- rib, margins entire to glandular crenate with as many as 8 crenations per cm., apex acute to acuminate. Inflorescences terminal on short lateral branchlets and occasionally axillary, paniculate, 1-7 cm. long and gener- ally about as wide. Staminate flowers about 2 mm. long; pedicels 0.2—1.5 mm. long; perianth segments 6—7, undifferentiated, uniseriate to irregularly biseriate, green to yellow, elliptic to ligulate, 0.3-1 mm. long, glabrous, stamens 4—6, 1-2 mm. long, anthers about 1 mm. long and yellow prior to anthesis; disc pulvinate, about 0.5 mm. high; rudimentary carpels none. Carpellate flowers about 2 mm. long; pedicels and perianth segments as in staminate flowers; rudimentary stamens none; disc pulvinate, 0.5—0.75 mm. high; gynoecium 1—3-carpellate, glabrous, about 1.5 mm. high, styles about 0.75 mm. long, divergent, articulating about 0.3 mm. below the globose stigma. Fruiting pedicels 1-2.5 mm. long; follicles generally reddish, sub- globose, 4-5 mm. in diameter, single or in 2’s or 3’s, the undeveloped carpels caducous. West Pakistan. Hazara (Abbottabad) District, Massar Reserve, Sirau Range, Parker, May, 1910 4 (A). India. Punyas: Kangra, Baijnath, Koelz 4516 9 us); Kulu Kulu, Koelz 8315 2 (a), 10274 & (a); Shahpur, Dharmsola Road, Stewart 1823 2 (a), Hrmacuat PrapesH: Simla District, potas 3019 @ (A); Jaranda Range, between Sarahair & Gohoilhur, Selhi, July, 1953 2 (a). UTTAR & Winterbottom 3 2 (GH), 4 2 (GH). Orissa: Koraput District, near Pottangi Sirimanda Parbat, Mooney, July 10, 1950 (a). Assam: Khasia & Jaintia Hills, Bor 16083 8 (a), Hooker f. & Thomson Q (cu), Koelz 22785 2 (micu), Ruse 8 (A). Without definite locality: Lambert, 1816 2 (c-pc—lectotype of Zanthoxylum armatum, not seen; A, photo.), Thomson 9 (cH). Nepal. West Nepal, Doti District, Rikhar, Ram 398 2 (a); without poem locality, Ram 181 2 (A), Wallich, 1821 (a, photo.), — ? (sinc). S . Hooker (GH). Burma. Kalewa, Dickason 6148 2 (aA); Southern ae State, Khalil, 1893 2 (A). North Vietnam. Chapa, Petelot 5815 2 (A, sinc). Tibet. Sikang Province, Dzer-nar, Tsa-wa-rung, Wang 66378 2 (a). China. YUNNAN PRov- INCE: Likiang City, Ching 21062 2 (a); Tapintze, Delavay 777 2 (a, photo. and fragment); NE. Likiang, between Min-ing and Hung-mun-kow, Feng 2549 2 (a); W. Likiang, Kai-tze on Yangtze bank, Feng 2647 2 (a); S. Chungtien, Chiao-tou on Yangtze bank, Feng 3125 (a); Wen-shan-hsien, Maa-luh-tarng, Feng 11245 2 (a); Mengtsz, Henry 9366 2 (a, Us), 9366A [a (2 sheets, $ and 2), us (3)], 9366C 2 (a), 9366D 2 (A, US), 9366F 8 (a); Szemao, Henry 11908 $ (a), 12249 2 (a, us); Kun-ming, Miao-kao-szu, Ta-pu-chi, Ko 56310 1966 | HARTLEY, MALESIAN ZANTHOXYLUM 213 2 (A); W of Mekong en route from Pingpo to Youngchang and Tengyueh, Rock 6994 2 (A, Us); Shweli River drainage basin & environs of Tengyueh, Rock 8031 8 (A, Us); zwischen Yunnanfu und Sah-ying-pan, Schneider 281 (a); Talifu, Schneider 2803 2 (A, GH, US); prope Yang Lin, Schoch 351 @ (a, us); Kienshuei Hsien, Tsai 53132 8 & 2 (a); Shang-pa Hsien, Tsai 54638 2 (a), 54984 2 (a); Ping-pien Hsien, Tsai 55163 2 (a); Chenkang, near Feng-tai-szu, Tsiang & Wang 16258 2 (A); Wei-si Hsien, Kang-pu, Wang 64191 2 (a); Chen-kang Hsien, Wang 72187 2 (a); Fo-hai, Wang 74490 9 (a); Che-li Hsien, Kuen-ger, Wang 79257 2 (a); Saluen Valley, NW of Champutong, Yu 19171 2 (A); without definite locality, Forrest 7535 2 (a), Schneider 3887 @ (A), Tsai 57272 2 (a). SzEcHWAN Province: Mt. Omei, Chien 5565 8 (a), Fang 3913 2 (A), 12666 2 (a), Lee 2715 2 (us), Liu 1032 2 (a), 1650 @ (a), Wang 23192 2 (a), Wilson 4769 & (A), Vu 254 2 (a), 325 8 (GH), 341 3 (a); Kuan Hsien, Chien 5596 2 (a), Fang 2026 2 (a), Wang 20542 8 (a); Huili, Handel-Mazzetti 248 2 (a); Mili Kingdom, Mili & Litang River Valley, near Mili Gomba, Rock 16129 2 (a, us); Ningyuanfu, Lu-shan, Smith 1790 2 (a); zwischen Mo-so-ying und Kung-mu-ying, Schneider 645 2 (a); Yun-ching-hsien, Sun 1287 2 (us); Mao Hsien, Wang 21856 2 (A); Ping-shan Hsien, Wang 22697 2 (A); Lo-shan Hsien, Wang 23556C 2 (a); Wen Chuan Hsien, Wilson 2693, July, 1908 2 (a); He-chang Hsien, Yu 1120 2 (A). KwEeIcHow PROVINCE: Tsunyi Hsien, Liang Feng Yah, Steward, Chiao & Cheo 34 2 (a, us); Ta Ho Yen, Fan Ching Shan, Steward, Chiao & Cheo 696 ° (us); Tsingchen, Ju-long-san, Teng 90201 8 (a). Kwancst Province: Luchen, Tia Lian Shan, Ching 5295 9 (A); E. Hin Yen, Bian Chen, Ching 6710 2 (A, us); Hin Yen, Spring Cave, Ching 6741 2 (a, us); Shuen-yuen, Chung 81695 2 (a); Hing On District, Sai On Vil- lage, Chung 83615 2 (A); Yang-so, Ching-o-ken, Fung 21106 9 (a); Ling Yun District, Lau 28557 2 (a); Ling Yun Hsien, Loh Hoh Tsuen, Steward & Cheo 56 $ (A, SING); Yung Hsien, Ta Tseh Tsuen, Steward & Cheo 795 2 (A, SING); Yung Hsien, Chang An, Steward & Cheo 1188 2 (A); Mts. surrounding Pa Lau Village near Sui-luk, Tsang 21871 8 (a); Ling-chuan District, Ling-chai-miao, Hai-yang Shan, Tsang 27829 2 (a, us); Kwei-lin District, San-min Village, Tsang 28021 2 (A, us); Nam Tanyuen, Wang 40895 2 (A); without definite locality, Wang 41127 2 (a). Kwanctunc Province: Nan Shung, Chun 5704 2 (a); Yim Na San, Gressitt 1395 9 (a); Lung Tau Mt., Kang, W. T. Tsang & U. K. Tsang CCC 12709 2 (us); North River Region, near Fung-wan, Kang, W. T. Tsang & U. K. Tsang CCC 12857 2 (us); Sin-fung District, Hau Tong Shan, Fuk Lung Monastery, Taam 757 2 (a); Jen-hua District, Man Chi Shan, Shek Pik Ha Village, Tsang 26136 2 (a); Ying Tak District, Wan Tong Shan, Tsui 302 ? (a, us); Yang Shan District, Yang Shan, Tsui 458 @ (A, Us), 665 @ (us). Hunan Province: Sinning Hsien, Ma-ling-tung, Fan & Li 481 2 (A); Tschang- scha, Handel-Mazzetti 2347 6 (a); Yi Chang District, Ping Tou Shan, Pai Mu Village, Tsang 23406 8 (a, us); Wukang, Yun-shan, Wang 15 4 (a). HureH PROVINCE: Yangtze-Hwaiho Divide on Hupeh-Honan border, Chikungshan, Bailey, June 29, 1917 (a); Ichang, Chien NU 8034 2 (us), Chun 3831 9 (us), Tsung 3486 2 (a); Gran Gra Kon, Ching 3560 2 (a); Patung Hsien, Chow 349 @ (a); Enshih Hsien, Chow, June 14, 1934 2 (a); Kwan Ying Tong, Chun 3686 2 (a), NU 4028 2 (us); Liang-sun-yon, Chun 3849 @ (A) ; Wuchang, Lokiashan, Chung 9176 (a), Sun 19 2 (A), 775 @ (A), 848 2 (A); U-tan Hsien, Silvestri 4849 $ (a); Hoing-shan Hsien, Wilson 138 [a (2 sheets, ¢ & 9 and ?), cH (S$ & 2), us (¢ & Y)]; Chang-to Hsien, Wilson 2693, April, 1907 [A (2 sheets, $ & 2 and 2), us (2)]; without definite locality, Henry 3584 Q (us), 7687 9 (a, GH), Wilson 338 2 (us). HoNaAN Province: Sung Hsien, 214 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 Shih Tze Miao, Hers 594 2 (a); Yungning, Tieh-lu-ping, Hers 1001 2 (A). Krancst Province: Lu Shan, Chiao NU 18767 2 (us); Nanchang, Chung 10 2 (a); Kiennan District, Sai Hang Cheung, near Tung Lei, Lau 4183 @ (us); Lungnan District, Oo Chi Shan, Lam Uk Tung Village, Lau 4480 2 (A, US); arr-yung, ing NU 7892 2 (A). FUKIEN Province: Foochow: Nantai Island, Chang 4250 2 (A); Nanshan Monastery, Chung 2802 2 (a, stNc); Kushan, Chung 8355 2 (a); Kuliang Hills, Norton 1395 2 (us). Diong Loh Hsien, Chung 1238 2 (A, SING); Diong Loh, Muoi Hua, on the seacoast, Ging 13651 9 (a); Yeuping, Buong Kang, Chung 3390 2 (a); N. Mountain, Chung 8597 2 (a); Baekliang, Ging 15596 @ (a), 15652 2 (A); without definite locality, Dunn 845 2 (A). CHEKIANG PRovINcE: Shihpu, Chiao NU 14139 Q (a, US), NU 14152 @ (aA, us); Sizchu, Ching 1659 Q (A, GH, US); Tsing-tien, Keng 56 Q (A); vicinity of Ningpo, Macgregor, 1908 2 (a); Hangchow, Tang & Hsia 79 2 (a); without definite locality, Barchet 85 2 (us), sm. 2 (us). ANHWEI Province: Tien Chu Shan, Chien Shan Hsien, Fan & Li 115 2 (a); Chuchow, Zee NU 1298 $ (wus). Ki IANGSU Province: Nanking and vicinity, Chen 8664 Q (A), Merrill 11394 2 (a), Steward NU 2169 2 (A), Tso 1021 Q (a); Ih-shing, Ching 4883 2 (A); Chiu-yong, Ling NU 12192 2 (GH); Chankiang, Chin Shan, Tso 254 2 (a); Ma An Shan, Quin Shan, Tso 1633 9 (A). Taiwan. Maru- yama, Faurie 22 2 (A). Korea. Fusan, Faurie 465 2 (A); Namtijoukak, Taquet 820 2 (A); Hokeum, Jaquet 2707 2 (a); Nokan, Taquet 2708 9 (A); Quelpart Island, Faurie 1627 2 (a), Wilson 9397 2 (aA). Japan. TsusHiMA IsLanp: Wil- (A), 11437 (A); Mt. Setsuhiko, Muroi 20 2 (a); Mt. Ruridera, Muroi 914 2 (a); Owari Province, Shiota 4462 2 (A); Kariyose-yama, Suzuki 106 2 (us), 336-1 2 (A), 336-2 2 (A); Mt. Kogane, Uno 2599 9 (a). Kyusnu ISLAND: Fukuoka, /chikawa 138 2 (A); Mt. Kirishima, Tashiro, September 23, 1917 2 (A); Kiushiu, Shiraishi, Uno 2611 2 (A). Without definite locality: Siebold 2 (a1 — isotype of Zanthoxylum planispinum, not seen; A, photo.). Ryukyu Islands. Kume-jima, Amano 7074 (us). Lesser Sunda Islands. SoEMBA: ney i Iboet 559 @ (1). Trmor: near Kapan, Bloomborgon 3478 @ (A, L); en Ond. Z. midden Timor, Steenis NIFS bb 17584 (a, L); Mt. Perdido, ascent of Ossu, Steenis 18324 8 (L); Mt. Moetis, DeVoogd 2283 2 (1); Kot Olim, Walsh 252 2 (BM); without regret locality, Forbes 3653 2 (BM, GH, L). Philippines. Luzon Istanp: Benguet Subprovince, Loher 243 (xk), Merrill 774 2 (vs), alatum, not seen; A, photo.). Japan: Fukuoka, Jchikawa 247 2 (cH); Tokyo, Mizushima TNSM 1398 2 (us), Sargent, August, 1892 @ (A). DiIsTRIBUTION. West Pakistan and northern India east to Japan and Taiwan; south only in the Philippines and Lesser Sunda Islands; in rain forests, thickets and, at higher elevations, often on open slopes and rock ledges; at elevations from 50-1750 meters in Malesia and up to 2400 meters in continental Asia. See map 9. ILLustRATIONS. Huane, C. C., Acta Phytotax. Sinica 6 (1): ¢. 3. 1957, as Zanthoxylum planispinum. Liv, T. S., Iustrations of native and intro- duced plants of Taiwan 2: t. 730. 1962, as Z. planispinum. MAxIno, T., Illustrated flora of Japan, 2nd. ed. t. 1175. 1940, as Xanthoxylum 1966] HARTLEY, MALESIAN ZANTHOXYLUM zi5 ieee ; mani ew - k & be Map 9. Distribution of Zanthoxylum armatum DC. planispinum; revised ed. t. 1175. 1949, as Zanthoxylum alatum var. pblanispinum; New illustrated flora of Japan ¢. 1326. 1961, as Z. alatum var. planispinum. Occasionally cultivated as a hedge plant in China. The bark is pungent and several collectors noted that it is used to clean teeth. It also has been used, in China, to repel insects from furs. Indian collectors noted that the leaves, bark, and fruits were used in treatment of smallpox and cholera. Various parts of the plant are used to season food in China and India. This is apparently the first use of the epithet armatum for this species since DeCandolle substituted it for Roxburgh’s alatum in 1824. That Zan- thoxylum alatum has so long been the accepted name may probably be accounted for by the fact that the DeCandolle name is listed as a mis- spelling in Index Kewensis. It is clear, however, that the spelling was in- tended since the Lambert collection (which is cited by DeCandolle) in the Prodromus Herbarium bears the annotation “Zanthoxylum armatum Roxb.” in A. P. DeCandolle’s handwriting. This collection is mounted on a sheet with a Wallich collection from Nepal and consists of two branchlets (both collected in 1816). I have chosen the larger of the two, in the lower right corner of the sheet, as the lectotype. “Zanthoxylum planispinum” differs in having fewer leaflets — generally 3-5, compared to 5~11 in typical Indian material of Zanthoxylum arma- tum. The number of leaflets is extremely variable in this species, especially in specimens from China. 216 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. 47 EXCLUDED SPECIES The following species, described from Malesia, apparently belong to genera other than Zanthoxylum. FAGARA DECANDRA Blanco, Fl. Filip. (ed. 1) 66. 1837 = Buchanania arborescens Bl. (Anacardiaceae) according to Merrill, Enum. Philip. Fl. Pl. 2: 465. 1923. FAGARA GLABRA BI, Cat. Gew. Buitenz. 40. 1823 = Evodia glabra (Bl.) Bl. Bijdr. Natuurk. Wetens. 245. 1825. FAGARA MONTANA (BI.) Engl. Nat. Pflanzenfam. III 4: 118. 1896 = Turpinia montana (Bl.) Kurz (Staphyleaceae), Jour. Asiat. Soc. Beng. 44: 182. 1875. ?FAGARA PAPUANA Laut. Nova Guinea 8 (4): 823. 1912 = Melicope papuana (Laut.) Laut. in Engl. Bot. Jahrb. 55: 241. 1918. FAGARA SERRULATA (BI.) Engl. Nat. Pflanzenfam. III. 4: 118. 1896 = Turpinia montana (Bl.) Kurz (Staphyleaceae) according to Koorders, Exk. Fl. Java 2: $23.0 912; FAGARA TRIPHYLLA Lam. Encycl. 2: 447. 1788 = Melicope triphylla (Lam.) Merr. Philip. Jour. Sci. Bot. 7: 375. 1912. ZANTHOXYLON ACCEDENS (Bl.) Mig. Fl. Ind. Bat. 1 (2): 671. 1859 (based on Evodia accedens Bl. Bijdr. Natuurk. Wetens. 246. 1825) = Evodia accedens Bl. according to Miquel, Ann. Mus. Lugd.-Bat. 3: 242. 1867. ZANTHOXYLON AROMATICUM (BI.) Migq. Fl. Ind. Bat. 1 (2): 670. 1859 (based on Evodia aromatica Bl. Bijdr. Natuurk. Wetens. 246. 1825) = Evodia glabra (Bl.) Bl. according to Miquel, Ann. Mus. Lugd.-Bat. 3: 243. 1867. ZANTHOXYLON EUNEURON Miq. FI. Ind. Bat. (suppl.) 532. 1860 = Evodia euneura (Miq.) Mig. Ann. Mus. Lugd.-Bat. 3: 243. 1867. ZANTHOXYLON MACROPHYLLUM (BI.) Miq. Fl. Ind. Bat. 1 (2): 670. 1859 (based on Evodia macrophylla Bl. Bijdr. Natuurk. Wetens. 246. 1825) = Evodia accedens Bl. according to Miquel, Ann. Mus, Lugd.-Bat. 3: 242. 1867. ZANTHOXYLON MARAMBONG Mig. FI. Ind. Bat. (suppl.) 533. 1860 = Evodia marambong (Miq.) Mig. Ann. Mus. Lugd.-Bat. 3: 244. 1867. ZANTHOXYLUM HORSFIELDII bhi Bull. Soc. Nat. Moscou 36 (2): 597. 1863. I have not seen type material but from the description—the leaves are described as opposite — iti is sprees not a Zanthoxylum. ZANTHOXYLUM JAVANICUM (Bl.) Walp. Repert. Bot. Syst. 1: 519. 1842 = Picrasma javanica Bl. (Simaroubaceae) Bijdr. Natuurk. Wetens. 248. 1825. ZANTHOXYLUM LAMARKIANUM Cham. Linnaea 5: 58. 1830 = Melicope triphylla (Lam.) Merr. according to Merrill, Philip. Jour. Sci. Bot. 7: 375. 1912. ZANTHOXYLUM LATIFOLIUM (DC.) G. Don, Gen. Syst. 1: 804. 1831 (based on Ampacus latifolia Rumph. Herb. Amb. 2: 186. 1741) = Evodia latifolia DC. according to Merrill, Interpret. Rumph. Herb. Amb. 289. 1917. ZANTHOXYLUM LucipuM Mia. FI. Ind. Bat. (suppl.) 532. 1860 = Evodia lucida (Miq.) Miq. Ann. Mus. Lugd.-Bat. 3: 244. 1867. ZANTHOXYLUM MONTANUM BI. Bijdr. Natuurk. Wetens. 248. 1825 = Turpinia montana (Bl.) Kurz (Staphyleaceae), Jour. Asiat. Soc. Beng. 44: 182. 1875. ZANTHOXYLUM RHIZINOIWES Bl. Bijdr. Natuurk. Wetens. 248. 1825. According to the description the leaves are simple and the calyx 3-parted, a combination not found in any of the known Malesian species of Zanthoxylum. Miquel (F. Ind. Bat. 1 (2): 669. 1859) and Koorders (Exk. Fl. Java 2: 418. 1912) also express the opinion that the generic designation may be incorrect. 1966 | HARTLEY, MALESIAN ZANTHOXYLUM 217 ZANTHOXYLUM ROXBURGHIANUM Cham. Linnaea 5: 58. 1830 = Evodia lunur- ankenda (Gaertn.) Merr. according to Merrill, Philip. ee Sci. Bot. 7: 378. 1912. ZANTHOXYLUM RUMPHIANUM Cham. Linnaea 5: 58. 1830 = Evodia latifolia DC. according to Merrill, Interpret. Rumph. Herb. Amb. 289. 1917 ZANTHOXYLUM SERRULATUM BI. Bijdr, Natuurk. Wetens. 249. 1825 = = Turpinia cs agen is Kurz (Staphyleaceae) according to Koorders, Exk. Fl. Java 2: 528. 19 as TIMORIENSE Span. Linnaea 15: 185. 1841. I have not seen type material of this but from the description it seems closer to Toddalia than Zanthoxylum. ZANTHOXYLUM? TRIPLINERVE Turcz. Bull. Soc. Nat. Moscou 36: 597. 1863. Reported from “Insula Lucon, in provincia North Hows.” The leaves are described as simple, so this is probably not a Zanthoxylum. Perkins, Frag. Fl. Philip. 161. 1905, states that the type has not been accounted for and that the locality given is not clearly a Philippine one. LITERATURE CITED Brizicky, G. K. Taxonomic and nomenclatural notes on Zanthoxylum and Glycosmis pangs Jour. Arnold Arb. 43: 80-93. 1962. ENcLer, A. Rut Nat. Pflanzenfam. III. 4: 95-201. 1896. FOsBERG, FL. Typification of Zanthoxylum L. Taxon 8: 103-105. 1959. Huanec, C. C. Preliminary study on Chinese Rutaceae (1). (In Chinese.) Acta Phytotax. Sinica 6 (1): 1-143. 1957. REepeER, J. R., & S. CHEo. Notes on Xanthoxylum and Fagara in China. Jour. Renoer, A., & E. H. Wrson. Rutaceae. Pl. Wils. 2: 121-151. 1914 a C. G. G. J. van. Phytogeography of Malaysia. Fl. Males. Ser. 1. 1: XX-LXXV. 1950. a W. L., & G. K. Brizicxy. The morphology and relationships of Diomma, Gen. Inc. Sed. Mem. N. Y. Bot. Gard. 10: 38-57. 1960. INDEX TO EXSICCATAE The numbers in parentheses refer to the corresponding species in the ext. Addurn FB 22989 (10) Barnes FB 88, FB 336 (16) Afalla 9 (2) Bartlett 15109, 15133, 15827 (5) Ahern’s Collector FB 1455, FB 3105 Beccari 132 10) Beguin (L) (17) Alvins 36 (3) Biswas 1707 (2) Amano 7074 (19) Bleeser 532 (11) Armstrong 569 (11) Blinkworth Wall. cat. n. 7469 (17) Backer 73, 11717, 16980 (2); 17017 Bloomborgon 3478 (19) (8); 24802 (10); 27465, 27590 Blume 1603 (1); 1814 (2) (17); 27639 (10); 27892, 27921, Bois 97 (10) 27968, 28122 (17); 31164 (2) Bor 16083 (19) Balansa 4045 (3 Borden FB 1565 (10); FB 2351, FB Barchet 83, 84, 85 (19) 2484, FB 2740 (16); FB 3051 (10) 218 Brandis 856 (17) Brass 3449 (14); 3833 (2); 6491 (10); 7983 (2); 8060, 8507 (17); 11579, 22553, 22919, 30345, 30496, 30614 (15 Brass & White 180 (2) Brown 5334 (11) Bunnemeyer 9574 (1) Burkill SF 16330 (18) Canicosa FB 31197 (3) Carr 11183 (17); 11582 (14); 11913, 12594 (17); 13256 (14); 14925 (17); 16422 (14) Caster FB 29749 (5) Cavalerie 1771, aoe (3) Champion 86 Chand 3113 aoe 3376 (18): 4908 (1); 6328 (17) Chiao NU 14130 (4); NU 14139, NU 14152 (19); NU 14760 (1); NU 18767 (19) Chien 5565, 5596, NU 8034 (19) Ching 1659, 3560, 4883, 5295 (19); 5401 (1); 6510 (2); 6584 (1); 6710, 6741 (19); 7102 (4); 7130 (1); 7685 (2): 7827 (5); 8128 (3); 21045 (18); 21062 (19); 21882 (18) Chow 349 (19) Chun, N. K., 40435 (2) Chun, N. K., & Tso 43401 (5); 43581 (2 ) Chun, W. Y., 3153 (5); 3686, 3831, 3849, NU 4028 (19): 4278 (1): 4809 (2); 5052, 5704 (19); NU 5825 (5); F952 (1): 6806 (5) Chung, H. H., 70 (19); 936 (2); 1030 (1); 1238 (19); 2652 (2): 2802, 3390 (19); 4770, 5607, 5889, 5892 5935, 6014, 6252 (2); 6647 (1); 6951 (2); 7583 (5); 7977 (2) ae (19); 8417 (2); 8597, 9176 19 wt we a Chung, Z. S., 81695 out 83485 (3); 83584 (1); 83615 (19 Clark FB 1710 (10) Clemens, J., & M. S., 3574 (15); 3996 (2); 4060 (15); 4067, 4445 (5); 26344, 26344a (3); 29213, 29243 (1); 31036, 40530 (5) JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 Clemens, M. S., 6813 (15) Collins 705 (2 Conklin PNH 19019 (16 Corner SF 33166 (3); SF 33458 (10) Cuming 1218, 1356 (10); 1622 (5) Cunningham 181 (11 Curran FB 3822 (10); FB 3860 (5); B 6920, FB 7688, FB 10695, FB 10726, FB 17437 (10); FB 17578 16 ) Curran & Merritt FB 7715 (10) Curtis 1076 (3 ee 934, 998 (6) Darnton 1 Delavay os Sag Derry 1121 penne sas (18); 6148 (19); 6718, Domin fe (17) Dunn Ae 845 (19) Ebalo 448 (2 Edano es 46165 (16); BS 77626, BS 7760 (5 Elbert 1582 (5); 3445 (2) Elmer 7278 (2); 10217 (3); 12063 (16); — (5); 15566 (16); 18231 Endert fee E 1078 (3) Fairchild 3484 (5) Fan & Li 115, 481 (19) Fang 2026, 3913, 12666 (19) Faurie 22 (19); 23 (2); 465 (19); 467 (4); 1268 (2); 1627 (19); 1632 (4); 2427 (19); 8216 (1) Feng 2549, 2647 (19); 3067 (18); 3125, 11245 (19); 11687 (18); 11735 (2); 11807 (18) Fenix BS 3584 (16) Fernandes 691 (10) Ferraris FB 23044 (5) Ferris 11934 (2) Feung & Kao 4583 (1) Floyd NGF 7471 (2) Forbes 96 (6); 838, 875 (17); 1001 (1); 1707 (7); 3653 (19) Forrest 7535 (19); 7881, 8846, 11790 (46). .19002 . £8715 ..(1); 15914 (18): 17712, 21554 (1 Foxworthy BS 905 (5); BS 1567 (16) Fung 7-79, 20202 (2); 21106 (19) Galoengi 231 (2) 1966] Gressitt 478, 965 (2); 1117 (5); 1395 19 Griffith 1186/2 (18) Hai 33 (2); 43 (5) Hamel & spieys si Toroes (Rahmat si Boeea) 630 Hamilton Wall. Je n.7117 (2) Hance 1145 (1) Handel-Mazzetti 248 (19): 482 (18); 2347 (19); 2563 (1); 12327 (3) Haniff SF 14978, SF 16778 (2) Hartley 10704 ( 37); 11349 (14); 12027, 13270 (15 Hatusima 18833 (2); 18923 (1) ge nderson SF 20132 (2); SF 21706 3) Henry 205, 2054, 205G (2); 343 (1); 464 (2); 1353, 1630 (4); 1655, 1782 (2); 1969 (1); 3584, 7687, 9366, 9366A, 9366C, 9366D, 9366F (19); 9758, 10150, 10150B (18); 10421A, 11908, Hoogland & Pullen 5471, 5475, 6116 15 Hooker f. & Thomson 1517 (1) Horsfield 66, 1093 (1) How 70306, 70675 See 70939 (5); 71366 (2); 73154 (1) Hu 840, 934 (1 Hui 158, 12321 (1) Tboet 559 (19) Ichikawa 138, 247 (19) d'Incarville 179 (5) Jaamat SF 27628 (3) Jaheri HB 430 (5) Jones 2060 (2) er ae 1358 (17); 1835 (6); 2304 2) Kamphovener 169h (2) Kang, Tsang & Tsang CCC 709 (19); COC tree CCC. Jar70.. (1): CCC 12857 (19) Keng, H., & Kao K 2674 (2) Keng, Y. L., 56 (19) Kerr 5473 (10); 8573 (2) Kingdon-Ward 20736 (18) HARTLEY, MALESIAN ZANTHOXYLUM 219 es Collector 102 (17); 189 (10); ) Bae FB 22464 (5) Ko 52130 (2); pi (3); 55674 (2); 56310 (1 Koelz 4516, 8315, 10274, 22785 (19); 5762 (3); 29225 (10): 30380, 30809, 31105 (17); 31303 (18) Koorders 1161C, 69958, 69978, 69988, 118128 (10); 187468, 187478, 187728 (12); 187738, 187758 (10); 232088 (5); 262848, 263338, 279298, 279388 (17) Koorders & Koorders-Schumacher 479638 (2) Kostermans 1426 (12); 18075 (17 Kou & Kae 4496 (2) Lam 299 (17); 3254 (5) Lane-Poole 392 (14) Lasquety FB poe: See Lau, S.. Bip 6212) 2 133, 632.45); 1079 (2): (1); poe (5); 2770 1481 (5); 4480 (19); 5523 (2); 27140 (5); 28166 (3); 28307 (1); 28557 (19) Lau, S. Y., 20292 (5) Lee 2715 (19) Lei 55 (5); 342, 363, 800 (2); 876 5 Levine CCC 170 (5); CCC 201, CCC 236, CCC 407 (2); CCC 563 (5); CCG 11320 (2); CCC 1298 (5): CCE a (2); (CCC. tai4, CCC 3134 Levine & hee COC.6913-(5) Liang 61515, 61631 (2); 62183 (5); 62340 (2); 63604 (5); 64222 (3); 65076, 65207, 65216, 65318 (2) Ling NU 12192 ~~ Liu 1032, 1650 (19) Loher 241 (5); 243 (9); 6058 (16); 14840 Lorzing 6100, 13687 7 ); 15199 (3); 15257 (2) Ludbrook 58 (2) Maingay 279, 1098 (3) McKee 9296 (17) Meijer 4043 (1) Merrill 774 (19); SB 872, SB 961 (10); SB 1002, SB 1060 (5); 2706 5145 (10); 220 (10); 9940 (2); BS 10052 (5); 10660 (1); 11394 (19) Merritt FB 3659 (16) de Mesa FB 27604 (16) Meyer FB 2307 (16) Millar NGF 9997 (2) Miranda FB 18971 (10) Mizushima 573 ); TNSM 1080, TNSM 1398, 11094, 11437 (19) Moll BW 7025 (13 Indies Forest Service (NIFS), the following by anony- mous collectors: Ja 2229 (10); bb 910 (3); Ja 2933 (10); bb 19450 (12); bb 27284 (10) Noerkas 154 (10) Norton es Ds 1394 (2): 1395 (19) Nur SF 1 Oldham = a): 7 (1); 60 (2) Omar Sar 00107 (3) van Ooststroom 13255 (1) Parker 3019 (19) Penas FB 26698 (16) Petelot 822 (5); 875, 1309 (2); 1676, 2591 (5); 2691 (2); 3569 (3); 5815 (19); 7911 Poilane 1687, 28200 (2) Porter Wall. cat. n. 1214 (3 Prain’s Collector 758, 6 987 (2) Pringo Atmodjo 55 (18) Pullen 309 (15) Quezon FB 30252 (16) Quimpo FB 30149 (5) Rahmat si Boeea 9383 (3); 10558 (18) Ram 181, 398 (19 Ramos BS 1421 (10); BS 1629 (16); BS 1680 (3); BS 5090 (5): BS 13261 (2); BS 20988 (5); BS 22465, BS 23421, BS 24171, BS 24522 (16); BS 32711 (10); BS 39551, BS 41586, BS 79957 (16); BS 80214 (5); BS 80273, BS 80381 (4) Ramos & Convocar BS 83490, BS 83892 (16) Ramos & Edano BS 26303 (5); BS 33770 (16); BS 44007 (2); BS 44110 (10); BS 46617 (2): BS JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 49631 (10); 85175 (10) Ramos & Pascasio BS 35248 (16) Ridley 10281, 11291 (2); 13541 (3) Rock 1166, 1726 (18); 6994, 8031, 16129 Rosenbluth FB 12866 (3) Roxburgh Icones 185 (10); 1916 (19); 2113 (10); 2430 (2) Ruse 62, 62A (1); 73 (19); 122 (17) Samundsett 22 (17) Santos BS 32025 (5) Saunders 770, 892 (15) Schneider 281, 645 (19); BS 84843 (5); BS 1353, 1373 (18); 2803 (19); 2877 (18); 3887 (19 Schoch 268 So 351 (19) Scortechini 275 (3 Shiota 4459 () 4462 (19) Silvestri ene 19) Shimada 39 (4) Simeon ee 28756 (5) Simons 620 (18) Sinclair & Edano 9765 (19) Singh 22 (19) Smith, H., 1790 (19) Smith, L. S., NGF 1287 (14) Smith, R., 85 si Smith, W. W., 702 (18) Smitinand & Abbe 24842 (5) Sou & Tachou SF 16893 (3) Squires 170 (2) van bees 748 (2); 6342 (18); 12880 - NIFS bb 17584 (19); 18153, oe (17); 18324 (19) Steward NU 2169 (19) Steward & Cheo 56, 795 (19); 969 (3); 1188 (19) Steward, Chiao & Cheo 34, 696 (19); 708 (3); 757 (1) Stewart 7823 (19) Strachey & Winterbottom 2 (18); 3, 4 (19) Sulit PNH 12282 (2); PNH 12396 (5); PNH 14512 (16) Sun 19, 775, 848, 1287 py Suzuki 106, 336-1, 336-2 ) Taam 426 (1); 757 ao): P71 414); 1159 (2): 1617, 17918 (5); 1825 (1); 2024 ( Tanaka 1784 (2) ang, S. G., 6751 (5); 13051 (2); 1966 | HARTLEY, MALESIAN ZANTHOXYLUM 221 13119 (5); 13531 (2); 13651 (19); 15333 (2); 15596, 15652 (19): 16040, 16055 (5) Tang, T., & Hsia 79 (19) Tangkilisan NJFS bb 33848 (12) Taquet 619 (4); 820, 2707, 2708 (19): 2713, 2714, 4142 (4 Ten 502 (18) Teng 90156, 90174 (1); 90201 (19) Thwaites CP 3490 To CCC 407, CCC 1130 (2); CCC 2237 (1) To & Tsang 12770 (1) Tsai 51691 (18); 51703, 51764, 52369 (1); 53113 (18); 53132 (19); 53494 (1); 53686 (18); 54018 (1); 54638, 54984 (19): 55108 (1); 55163 (19); 55526 (18): 55610, 55695, 55960 (1); 56103, 56423, 57118, 57245 (18); 57272 (19); 58566, 59989 (18); 60072 (1); 60092 (18); 60269, 60451 (1); 60761, 61922 18 Tsang 98, 115 (2); 213 (5); 422, 601, 667 (2); 787 (5); 837 (3); 16529 (5); LU 16697 (2); 20149 (1); 21871 (19); 21919, 22973 (2); 23022, 23291 (5): 23310, 23387 (1); 23406 (19); 23924 (1); 23924a (2); 24326 (5); 25017, 25193, 25698 (1); 25872 (5); 26136 (19); 26248 (1); 26868, 27016, 27392 (5); 27829, 28021 (19); 28508 (1); 28864 (2); 29377, 29549, 30448, 30494 (5); 30511 (3) Tsang & Wong “ Mer 3110 (1) ytd, LOI (2); 1645 (5); 1651, 1927 ee Tsiang & Wang 16258 (19) Tso 254, 1021, 1633 (19) Tso & Tsaing 2008 (2) Tsoong 1472 Tsui 137 (1); 302, 458, 665 (19) Tsung 3486 (19) Uno 2599, 2611 (19) Versteeg 1910 (11) Vickery NGF 1405 (14) Vidal 134, 1015 (10); 1445 (5); 2269 (10 Vink BW 12065 de Voogd 1991 a 2283 (19) Walker & Tawada 6762 (4); 6801 (2) Walker & White BS/P 126 (14) Walsh 252 (19) Wang, C. W., 15 (19), 322 (5); 382 (2): 39369, 40057 (1); 40895 (19); 41125 (1); 41127, 64191, 66378 (19); 67388 (18); 69107 (1); 72187 (19); 74145 (1); 74490 (19); 74643, 75548, 77546, 77912 (2); 77979, 78566 (3); 79173 (2); 79257 (19) ang, F. T., 20542, 21856, 22697, 23192, 23556c (19) Warburg 20145 (5) Ware FB 16 (10 Waterhouse Y 167, B 344 (6) Watt 10441 Webb 2352 (2) Wenzel 1232 (2) Whitford 1044, 1295, 1464 (16) Whitford & Hutchinson FB 9067 (10) Whitmore BS/P 2719 (14) Whitmore’s Collectors BSIP 5661 (6) Wight 353 (10); 355, 356 (17) Williams 575 (16) Wilson 138, 338, 2693, 4769 (19); 6313, 8265, 8514 (4); 9397 (19); 9440 (4); 9973 (1); 10161 (2); 10229 (1); 11220 (4) Womersley NGF 9418, NGF 17629 (15 bie ois & Sleumer NGF 14011 wo ai (4); 82 (2); 82 (2); 82 ; 84 (5 Yates 1548 (18); 1850 (9) u 254, 325, 341, 1120 (19); 1607, ‘10576 17849 (18); 19171 (19); oe (1); 20391 (18); 20409 (1); 21009 (18) Zee NU 1298 (19) 222 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 A KARYOLOGICAL SURVEY OF LONICERA, I Lity RUDENBERG AND PETER S. GREEN * RECORDS OF PLANT CHROMOSOME NUMBERS have been accumulating for about sixty years and counts now exist for a very great number of genera (Darlington & Wylie, 1955) of the families of temperate regions at least. For tropical, and for temperate plants from the southern hemisphere, the need for random investigations still exists, but for plants from the northern hemisphere the stage has been reached where the further counting of chromosomes will be most effective and useful when carried out either in conjunction with an intensive biosystematic investigation, or where an attempt is made to deal as comprehensively as possible with a whole genus or similar taxon. The present survey is just such an attempt. The genus Lonicera was selected, not because of its particular cytologi- lected, and identified many of the accessions still in cultivation there. CYTOLOGICAL MATERIALS AND METHODS For chromosome counts at meiosis, and for the observation of chromo- alcohol: glacial acetic acid, and kept under refrigeration. The fixative was changed to 70 per cent alcohol for long storage in cases where material had to be preserved for some time before examination. The pollen mother cells were smeared on a slide in a drop of aceto-carmine, and after slight over- staining, a drop of Hoyer’s medium (Alexopoulos & Beneke, 1952) was added, gently mixed with the stain, then covered by the cover slip. When carefully warmed and pressed, the slides were satisfactory for counting and have since kept well. For mitotic counts, root-tip meristems and dividing cells in young leaves were generally used. This material was treated for one hour in a 0.0 Mol. aqueous solution of oxyquinoline (Tjio & Levan, 1950), a then fixed for ca. 12 hours in 3:1 alcohol; acetic acid. After hydrolysis in N HCl at 60°C for 25 or 15 minutes, for root-tips and leaves respectively, the material was washed in distilled water and sometimes hardened for 10 * In this survey, the cytological investigations have been carried out by one of us (L.R.), and the complementary taxonomy by the other (P.S.G.). gy 1966 | RUDENBERG & GREEN, LONICERA, I 223 minutes in 70 per cent alcohol. Best results were obtained by staining in leuco-basic fuchsin, followed by maceration on a slide in diluted aceto- carmine. After judicial warming and pressing the slides were made per- manent, using the freezing technique of Conger and Fairchild (1953). Because all counts have been documented, no meristematic cells from the roots of germinating seeds have been used in this investigation. Cuttings of the plants to be studied were rooted instead, and we wish to record our thanks to Alfred J. Fordham for carrying this out. However, favorably squashed cells from anther walls sometimes per- mitted the counting of diploid cells while the PMC’s were still in early prophase; tapetal cells were excluded because of the frequency of irregular or polyploid divisions that occur in them. All observations were made with a Zeiss phase microscope. DOCUMENTATION Chromosome numbers without adequate documentation are almost with- out value, and in this investigation herbarium specimens have been pre- served from each of the individual plants which has been examined cy- tologically. Either new material has been collected or the plants have been checked against specimens which already exist in the herbarium, i.e., speci- mens taken from the same plants on previous occasions. In the Arnold Arboretum each individual plant has its own accession number, and is recorded on a detailed map, so that even if labels are lost or transposed, the various plants can be accurately and positively related, each to its accession number and, through this number, to its identification. An at- tempt has also been made to ensure that the specimen growing under a particular number is the original plant and not, either the understock of a weak graft, or a bird-sown seedling, perhaps of hybrid origin and exhibit- ing heterosis, which has taken over and smothered the individual growing in that position in the first instance. This has happened, or was found to be taking place, in one or two cases, and Lonicera tatarica and its hybrid, L-. x bella, are particular culprits. Documentation of the counts recorded below has been achieved by the preservation of herbarium specimens in the herbarium of cultivated plants at Jamaica Plain (Aan) and by the recording of chromosome numbers on the Arboretum’s master set of accession cards. The collectors who have been responsible for making specimens during the two years since this sur- vey commenced are: Sandra Callaway, Madelon Dodd, Susan Kreps, Theo- dore R. Dudley, and Peter S. Green, whose invaluable help is gratefully acknowledged. Prior to this investigation, specimens of the honeysuckles growing in the Arboretum had been collected by Caroline K. Allen, C. E. Kobuski, F. P. Metcalfe, E. J. Palmer, A. Rehder, and C. S. Sargent and, where their collections coincide with plants which have been examined cytologically, their specimens are cited below. In a few cases the collec- or’s name is not actually given on the herbarium sheet, but, from the hand- writing or date, it is assumed that the collector was A. Rehder. The oldest 224 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 specimen cited is one collected by C. S. Sargent in 1891 from the same clone of Lonicera korolkowii from which the actual type specimen was originally gathered. The clone is still living and its chromosomes have been counted. In addition to the specimens incorporated in the herbarium at Jamaica Plain, duplicate collections have been made in a large number of cases. These are being distributed to other botanic gardens and arboreta, by way of exchange. In every case great care has been taken to check the identity of the plants reported upon, including a critical investigation of the complex of hybrids surrounding Lonicera tatarica and its allies (Green, 1966). CYTOLOGY The basic chromosome number of the genus Lonicera is nine and the majority of the species studied are diploid. Polyploidy seems to have played a minor role in speciation within the genus, for only a limited number of wild species are known to be tetraploids or hexaploids. How- ever, some varieties and cultivars of diploid species have been found to be polyploid, but not among the many cultivars of Section Lonicera. No triploids have been reported for the genus, although L. henryi has been observed as being hexaploid, a condition also recorded for L. periclymenum by Gadella and Kliphuis (1963). The chromosomes are small (from about 1.5,), but preliminary karyo- gram studies of mitotic chromosomes show that the nine chromosomes have a definite morphology which enables the observer to distinguish some of them individually, while others are present in small groups of similar size these differences in size or morphology coincide with the infrageneric group- ings proposed by Rehder. The results will be recorded in a subsequent paper but, as a preliminary observation, it may be said that in at least one group of related species a pair of chromosomes may be found which bear satellites consistently, while in others they are apparently absent. The satellites may be seen in FicurE 1 which shows a root tip cell of Lonicera morrowi at metaphase and exemplifies the typical morphology of the complement found in L. tatarica and its related species. As indicated below, hybridization between Lonicera tatarica and related species takes place very readily, yet, in these cases, no disturbance in the pairing relationships of the chromosomes was generally observed at meiosis, and the hybrids produced are abundantly fertile. NOMENCLATURE No attempt has been made to revise the genus Lonicera. The nomen- clature and sequence of species is that of Rehder (1903), and as modified by him many years later (1949). If chromosome counts have already been recorded under names not recognized in these two publications, the names 1966 | RUDENBERG & GREEN, LONICERA, I 225 Fic 1. Mitotic eet in root-tip cell of spunea morrowit (AA 1283-65) note metacentric chromosomes and a pair with satellite are cited in parentheses, otherwise no attempt is made to provide synonymy. Backcrossed hybrids are listed under their respective hybrid epithet, which is taken to cover all hybrids between the species involved, whether intermediate or backcrossed to either parent. In accordance with the present International Code of Botanical Nomen- clature the names of infrageneric divisions have been modified slightly from those of Rehder. The correct names at subgeneric and sectional level are discussed by Ferguson (1966) whose treatment is followed. In the case of Lonicera modesta and its variety lushanensis, both de- assessment and differentiation is required. Rehder based the differences between the type variety and var. ushanensis upon pubescence and leaf- shape. Leaf-shape now appears to be valueless for distinguishing the two, and degree of pubescence is never a very reliable character alone. How- ever, one of the two plants of this species whose chromosomes have been examined in this investigation (both raised from seed sent in 1936 from Lushan Botanic Garden in China) has proved to be diploid and the other tetraploid. The former has densely pubescent stems, and leaves with scattered hairs all over the lower leaf surface, and most nearly approaches the type variety, whereas, the latter has stems glabrous or with only a few hairs, and the leaves almost glabrous. The amount of flowering material available i is still inadequate but it is suspected that more reliable morpho- logical characters for separating these varieties will be found in the flower size (1 or 2 mm. larger in var. modesta) and flower color (a touch of anthocyanin in var. /ushanensis). It is interesting to note that the sizes 226 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 of mitotic chromosomes in the polyploid var. ushanensis were observed to be smaller than those in var. modesta HYBRIDS Spontaneous hybridization in the wild is not common in Lonicera, even though it can take place very readily under cultivation when certain species which otherwise occupy separate geographical areas are brought together. Fertile progeny and frequent backcrossing may result, and it appears that in parts of the genus, at least, speciation has not been dependent upon the existence of genetical or cytological barriers. Hybridization has been particularly prevalent in three groups: Section LonicerA and Subsections CyPHEOLAE (Raf.) Rehd. and EucAPRIFOLIUM (Spach) Rehd. of Subgenus Caprirottum. Hybridization between differ- ent subsections has frequently taken place, with the production of fertile offspring, and it would seem that they cannot be natural groups. In order that the possible effects of hybridization on the behavior of the chromosomes could be borne in mind during this investigation, the follow- ing list of hybrids has been compiled from the literature. The parental species are arranged alphabetically, and the sequence is that of the first parental species in the classifications of Rehder (1903, 1949). HYBRIDS OF WILD ORIGIN LONICERA ALTAICA Pall. & L. pALLAstI Ledeb. Recorded by Pojarkova (1958, p. 490); although these species are separated by her in separate series, Ser. Stenanthae Pojark. and Ser. Pallasianae Pojark. respectively, they are treated by Rehder (1903) as parts of the one variable species, L. caerulea L. Pojarkova reports that this hybrid, which is intermediate, occurs near Samed parent, and also in European USSR, beyond the range of L. alta pennants HYPOLEUCA Decne. &* L. QuUINQUELOCULARIS Hardw. Ac- cording to Wendelbo (1965, p. 15) one of the herbarium sheets of Lace 4024 collected at Ziaret, West Pakistan, is a possible hybrid of these two species. If this is so, it is very interesting as representing a cross between species of different sections, Sect. IstrkaA and Sect. Lonicera respectively. LONICERA FLORIBUNDA Boiss. & Buhse & L. NUMMULARIIFOLIA Jaub. & Spach. Recorded by Pojarkova (1958, p. 552) from the western part of Kopet Dag in Turkmenistan; the second parent cited above is treated by Rehder (1903) as a synonym of the wide ranging L. arborea Boiss LoNICERA NUMMULARIIFOLIA Jaub. & Spach & L. QUINQUELOCULARIS Hardw. Wendelbo (1965, p. 15) suggests that the mixture of characters in some plants, where the ranges of distribution of these two species over- lap, indicates that introgressive hybridization = be taking place. LoNICERA NIGRA L. & L. xYLosteuM L., or L. & HELVETICA Briigger (Jahresber. Naturf. Ges. Graubiindens IT. 29: 95. 1885 [Mitt. Neue & Krit. Formen der Biinder- & Nachbar-Floren, 50. 1886], Rehder 1903, p. re 1966 | RUDENBERG & GREEN, LONICERA, I 227 204). However, Rehder (1927, p. 832) suggests that this “supposed hy- brid” is probably a form of L. xylosteum. LONICERA CAPRIFOLIUM L. & L. ETRUSCA Santi, or L. & AMERICANA (Mill.) K. Koch, Wochenschr. Girtnerei & Pflanzenkunde 10: 279. 1867 (L. X italica (Schmidt) Tausch, Flora, 21: 735. 1838). LONICERA CAPRIFOLIUM L. & L. ImpLExA Ait. (Rehder 1903, p. 211). LONICERA CAPRIFOLIUM L. & L. PERICLYMENUM L. (Rehder 1903, p. 212): LONICERA ETRUSCA Santi & L. impLexa Ait. (Rehder 1903, p. 212), HYBRIDS OF ORIGIN IN CULTIVATION LONICERA FRAGRANTISSIMA Lindl. & Paxt. X L. sTANDISHII Jacques, or L. X purpusm Rehder, Jour. Arnold Arb. 4: 252. 1923. LONICERA ALPIGENA L. & L. LEDEBOURII Eschsch., or L. & PROPINQUA Zabel, Gartenflora 38: 580. 1889. LONICERA TATARICA L. & L. xyLosTEUM L., or L. X XYLOSTEOIDES Tausch, Flora 21: 736. 1838 (L. < coerulescens Dippel, Handb. Laubh. 1: 233. 1889; Zabel, Mitt. Deutsch. Dendr. Ges. 1901: 94. 1901). LONICERA KOROLKOWI!I Stapf < L. TATARICA L., or L. X AMOENA Zabel, Mitt. Deutsch. Dendr. Ges. 1901: 96. 1901. LONICERA RUPRECHTIANA Regel & L. TATARICA L., or L. X NOTHA Zabel, Gartenflora 38: 525. 1889 LONICERA MoRROWII A. Gray * L. TATARICA L., or L. X BELLA Zabel, loc. cit. LoNICERA MORROW! A. Gray x L. RUPRECHTIANA Regel, or L. X MUSCAVIENSIS Rehder, Gartenflora 42: 100. 1893. LONICERA MOoRROWII A. Gray * L. X xyLosTEOIDES Tausch, or L. MINUTIFLORA Zabel, Gartenflora 38: 523. 1889. LONICERA & BELLA Zabel & L. RUPRECHTIANA Regel, or L. & MUEN- DENIENSIS Rehder, Gartenflora 42: 102. 1893. LONICERA RUPRECHTIANA Regel & L. & XYLOSTEOIDES Tausch, or L. x SALICIFOLIA Dieck ex Zabel, Gartenflora 38: 524. 188 LoONICERA CHRYSANTHA Turcz. < L. RUPRECHTIANA Regel, or L. x GIBBIFLORA Dippel, Handb. Laubh. 1: 237. 1889, non Maximowicz. LONICERA CHRYSANTHA Turcz. * L. xyLosTEUM L., or L. & PSEUDO- CHRYSANTHA Braun ex Rehder, Man. Cult. Trees & Shrubs 833. 1927. LONICERA DEFLEXICALYX Batal. L. QUINQUELOCULARIS Hardw., or L. > VILMORINiIr Rehder, Mitt. Deutsch. Dendr. Ges. 1912: 194. 1913. LoNIcERA FLAVA Sims & L. HIRSUTA Eaton (Rehder 1903, p. 211). LONICERA FLAVA Sims L. PROLIFERA (Kirchn.) Rehd. (Zabel in Beis- sner, Schelle & Zabel, Handb. Laubh.-Benenn. 450. 1903, as L. flava x sullivantii). LONICERA HIRSUTA Eaton & L. PROLIFERA (Kirchn.) Rehd., or L. x SARGENTII Rehder, Jour. Arnold Arb. 7: 37. 1926. LONICERA HIRSUTA Eaton « L. SEMPERVIRENS L., or L. X BROWNII (Regel) Carr. Fl. des Serres, 11: 123. 1856. 228 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 LONICERA ETRUSCA Santi X L. IMPLEXA Ait. (Rehder 1903, p. 212). LONICERA ETRUSCA Santi *& L. PERICLYMENUM L. (Mottet, Revue Hort. 1900: 695. 1900). LonicerRA DIoIca L. * L. FLAVA Sims (Rehder 1903, p. 211 Lonicera pioica L. X L. HirsuTA Eaton (Zabel in Beissner, Schelle & Zabel, Handb. Laub.-Benenn. 450. 1903 as L. glauca Hill & L. hirsuta Eaton). LonIcerRA pioica L. & L. PROLIFERA (Kirchn.) Rehd., or L. & EPSOMI- ENsIs Hort. ex K. Koch, Dendr. 2(1): 12. 1872, nomen. LONICERA SEMPERVIRENS L, & L. TRAGOPHYLLA Hemsl., or L. & TELL- MANNIANA Magyar ex Spath, Hauptverz. no. 241. 1927-28 (not seen). LONICERA CAPRIFOLIUM L. & L. ETRUSCA Santi, or L. X AMERICANA K. Koch, Wochenschr. Giartnerei & Pflanzenkunde 10: 279. 1867. (L. X italica Tausch, Flora 21: 725. 1838). LONICERA AMERICANA K. KocHu L. SEMPERVIRENS L., or L. X HECKROTTII Hort. ex Rehder, in Bailey, Cycl. Amer. Hort. 2: 942. 1900. In addition, it has been suggested that Lonicera myrtilloides Purpus (Mitt. Deutsch. Dendr. Ges. 1907: 255. 1908) might be a hybrid of L. angustifolium Wall. « L. myrtillus Hook. f. & Thoms. (see Rehder, Man. Cult. Trees & Shrubs, 822. 1927) and that L. micranthoides Zabel (Mitt. Deutsch. Dendr. Ges. 1901: 94. 1901) might be the hybrid L. nigra L. X L, tatarica L. (see Rehder, Annual Rep. Missouri Bot. Gard. 14: 204. 1903), but L. micranthoides was later treated by Rehder (1927, p. 833) as a probable synonym of L. x xylosteoides Tausch. Likewise, and in the same works, Rehder treated L. segreziensis Lavallé ex Dippel (Handb. Laubh, 1: 221. 1889) first as a hybrid of L. quinquelocularis Hardw. and L. xylosteum L. and later as a probable synonym of L. xylosteum L. Rehd- er (1927, p. 832) also has a note that the supposed hybrid of L. nigra L. x L. orientalis Lam. is probably a form of L. & xylosteoides. LITERATURE CITED ALExopouLos, C. J., & E. S. BENEKE. 1952. pes ined Manual for Introduc- tory Mycology. pp. 177. Minneapolis, Minnesot Cuuane, T.-I., C. Y. Cuao, W. W. L. Hu, & S. C. ead, 1962. Chromosome numbers of the vascular plants of T aiwan, I. Taiwania 8: 51-66. [Lonicera, Sz. p. Concer, A. D., & L. M. Farrcutip. 1953. A quick-freeze method for making smear slides permanent. Stain Technology 28: 281 DaruinctTon, C. D., & A. P. Wyte. 1955. Pe ceincane Atlas of Flowering Plants. pp. 519. London. Fercuson, I. K. 1966. The genera of Caprifoliaceae in the Southeastern United States. Jour. Arnold Arb. 47: 33-59. [Lonicera, pp. 54-59. Gavetta, T. W. J., & K. Kureuvuts. 1963. Chromosome numbers of flowering plants in the Netherlands. Act. Bot. Neerl. 12: 195-230. [Lonicera, p. 2 GREEN, > s. 1966. Identification of the —— and hybrids in the Lonicera tatarica complex. Jour. Arnold Arb. 47: 75-8 1966 | RUDENBERG & GREEN, LONICERA, I 229 Hacerup, O. 1941. Nordiske Kromosom-Tal. I. Bot. Tidsskr. 45: 385-395. ] JANAKI Amal, E. K., & B. SAUNDERS. 1952. Chromosome numbers in species of Lonicers. Kew. Bull. 1952: 539-541. Matsuura, H., & T. Suto. 1935. Contributions to the idiogram study in phanerogamous plants, I. Jour. Fac. Sci. Hokkaido Univ. V. 5: 33-76. | Lonicera, p. 52.] PoJARKOVA, a I. 1958. Jn: V. L. Komarov, Fl. U.S.S.R. 23. pp. 776. Lenin- grad. | Lonicera, pp. 467-573. Poucgues, M.-L. de. 1949. Recherches caryologiques sur les Rubiales. Revue Gén. Bot. 56: 5-27, 74-138, 172-188. [Lonicera, pp. 84-95, 129 & 186.] REHDER, A. 1903. Synopsis of the genus Lonicera. Ann. Rep. Missouri Bot. Gar d. 14: 27-232 1927, Manual of Cultivated Trees and Shrubs. pp. 930. New York. . 1949. Bibliography of Cultivated Trees and Shrubs. pp. 825. Jamaica Plain, Massachusetts. [Lonicera, pp. 614-630. RuvENBERG, L. 1964. In: A. Love & O. T. Sotsric. IOPB Chromosome number reports, II. Taxon 13: 201-209. [ Lonicera, p. 207. Sax, K., & D. A. Krrps. 1930. Chromosomes and eae in Caprifoliaceae. Jour. Arnold Arb. 11: 147-153. [Lonicera, p. 150 Sorsa, V. 1962. grant armeaes maser’ Kormophyten, I. Ann. Acad. Sci. Fenn. A. IV. 58: 1-14. [Lonicera, p 1963. Chromsomenalen celal th Kormophyten, II. op. cit. 68: 1-14. [Lonicera, p. 1 Suciura, T. 1936. A list a chromosome numbers in Angiospermous plants, II. Proc. Imp. Acad. Japan 12: 144-146. [Lonicera, p. 144.] Tyro, J. H., & A. Levan. 1950. The use of oxyquinoline in chromosome analy- sis. An. Estac. Exp. Aula Dei 2: 21-64. [Not seen. ] Vitmortn, R. pe, & M. Srmonet. 1927. Nombres des chromosomes dans les genres Lobelia, Linum et chez quelques autres espéces végétales. Compt. Rend. Soc. Biol. Paris 96: 166-168. [Lonicera, p. 1 WENDELBO, P. 1965. Jn: K. H. RECHINGER, Flora aes 10 (Caprifoliaceae). pp. 16 + tab. 4. Graz. Table. CHROMOSOME NUMBERS IN LONICERA SPECIES n 2n AUTHORITY OCUMENTATION Doc GENERAL AND COLLECTOR DISTRIBUTION Subgenus Lonicera (Subgen. Chamaecerasus (L.) Rehd.) Sec Rehd. Subsect. ‘napa dec Rehd. L. myrti (18) > 7c Pakistan to Sikkim args var. depressa (36) 7 Himala L. myrtilloides (9) (18) 6 ? areiey L. angustifolia L. tomentella (18) 7 silekim ‘mee L. thibetica (9, 18) 2 Western China (18, 36) 7 L. rupicola (18) 7 Himalaya L. syringantha (18) 7 Northwest China Subsect. PURPURASCENTES Rehd. t. IsoxYLOSTEUM Sect. Istka (Adans.) Rehd. L. microphylla Willd. ex Roem. & Schult. (18) 2 Central Asia (36) 7 18 12* AA‘ 1008—-52-B, Dodd & Dudley, : 12 May 1965 L. canadensis (18) 7 Northeast America e. % - » 4 4 ’ L. tenuipes Nakai (18) 2 Japan (18) 7 18 36 12* AA 11147 (Wilson 6713, Japan, 1914), Dudley & Dodd, 10 May 1965 Subsect. CAERULEAE Rehd. L. caerulea L (9, 1 2 Eurasia (9) (18) 6 (18, 36) 7 18 36 12* AA 215-35, Kreps, 25 May 1 18 36 12* AA 126-37, Kreps, 25 May 1964 18 36 12* AA 20-43, Kreps, 25 May 1964 18 12* AA 308-51, Kobuski & Metcalfe, 9 M 1930 (as AA 5300) var. edulis Regel 18 12* AA 6714, Rehder, Siberia and Tibet 7 May 1924 f. graciliflora Dippel 18 12* AA 15727, Palmer, Turkestan 7 May 1921 [o96r Lb “10A] WOALAAXOTUV GIONYV AHL AO TVNUNOL O¢2 I ‘VYAOINOT ‘NAAUD ® OVAANAGAYA “Species names cited without authority are those used in undocumented counts; the sequence of species is that of Rehder (1903, 1949). ” Numbers in parentheses are those given for undocumented counts. oe Vion rin & Simonet (1927); 2, Sax & Kribs (1930); 3, Matsuura & Suto (1935); 4, Sugiura (1936); 5, Hagerup (1941); 6, eae (194 7, Janaki Ammal & Saunders (1952); 8, Chuang et ws — 9, Sorsa (1962); 10, Sorsa (1963) ; 11, Gadella & Kliphuis (1963) ; Ridenber (1964) ; 12*, Riidenberg, reported here for the first t n the Documentation column indicates that the ars ‘which follows is an Arnold Arboretum accession number. Tez SPECIES : oe DocUMENTATION ENERAL UTHORITY AND COLL Subsect. PILEATAE Rehd. eat caissiimash sive L. gymnochlamyd. — if 12* AA 15032-A (Wilson Western China ie W. Hupeh, 1908), L. pileata reen, 27 May 1965 (9) (18) 6 Central and western i L. nitida (9) (18) 7 — (18) Western China Subsect. VestcartaE (Komar.) Rehd. L. vesicaria L. ferdinandii Franch. (9) (18) 7 Korea (9) (18) : Northern China 9 12* AA 18360, Palmer, var. induta Rehd. 9 Sept. 1936 9 18 12* AA 18169-A (Purdom 554, Shensi, 1910), Dodd & Callaway, 28 var. leycesterioides (Gr 965 (Graebn.) Rehd. 18 12* AA oo -A, — Northern China a Subsect. CHLAMypocarpt Jaub. & Spach J L. iberica M. Bieb. : 12* AA 955-1, Rehder ?, Transcaucasia and 29 June 1916 Persia Subsect. FRAGRANTISSIMAE Rehd. ; L. standishii Jacques (9) (18) 6 China (18) 7 9 12* AA 15118-2, Kobuski, 30 Apr. 1932 f. lancifolia Rehd. 9 18 12* AA 6669-1-A, Palmer, 31 Mar. 1936 (as AA 6669-1) cv. Rosea 9 12* AA 448-43, Green, 30 Apr. 1964 L. fragrantissima Lindl. & Paxt. (9) 2 Eastern China (9) (18) 6 (18) 7 9 18 12, 12* AA 2245-1, Rehder, 1 Dec: 1927 L. X purpusii Rehd. 9 18 12* AA 397-36-B, Green, Cultivation 30 Apr. 1 18 12* AA 486-43, Green, 30 Apr. 1964 Subsect. —— (Hk. f. & Thoms.) Rehd. L. altman (9) 2 Turkestan (18) 7 L. praeflorens Batal. 18 12* AA 10103-B (Wilson Manchuria, Korea 8441, Japan, 1917), and Japan Green, 30 Apr. 1964 18 12* AA 657- ae oe 25 Mar L. hispida (9) (18) 6 Turkestan var. bracteata (18) 7 Himalaya to western rae 4 WOLAXOTAV GCIONUV AHL AO TVNUNOL Ly “I0A| I ‘VYAOINOT ‘NAAUD ¥ DAAINAGONA [9961 £ez DOCUMENTATION SPECIES n 2n AUTHORITY AND COLLECTOR L. chaetocarpa (18) 7 Subsect. gaia oes Rehd. L. pyrena (18) 7 Subsect. Disrectar (Raf.) Reh L. involucrata (Richards. ) ele ex Spreng. (18) 7 9 18 12* AA 16-44, Palmer, 16 July 1936 (as AA 221-33) f. humilis (18) 7 f. serotina Koehne 9 12* AA 817-61, Green, 21 June 1965 L. ledebourii (18) 7 Subsect. ALPIGENAE Rehd. L. alpigena (9) (18) 6 (36) 7 f. mana (Carr.) Zabel 18 12* AA ya oo 13 Aug L. glehnii (9) 6 (36) 7 L. webbiana Wall. ex DC. 18 36 12* AA 1735-4-B, oo 20 May 196 L. tatsiensis (9) 6 — abana (Maxim.) Rehd. a 2 iczit es 18 12. AA 11732 (Wilson ‘s coat Rehd. 9275, Korea, 1917), Rehder, 29 May 1923 L. orientalis Lam. (9) 2 (18) 7 18 12* AA 417-38-A, Dudley, 4 June 1965 9 12* AA 798-41, Dudley, 4 June 1965 var. caucasica (Pallas) Rehd. (18) 7 9 12* AA gots: cal é, 23 June 1 9 12* AA 364-29, a. 6 June 1940 var. longifolia (18) : L. discolor (18) 9 18 12* AA 24-36, Green, L. modesta Rehd. 31 May 1965 ; 18 36 12* AA 765- we iia var. lushanensis Rehd. 31 May L. nervosa (18) : L. nigra L. (9) (18) 18 12* AA 771-35, Kreps, 25 May 1964 GENERAL DISTRIBUTION Western China Pyrenees Northern America and south into Rocky Mts. Colorado Colorado California Central and southern European Mts Japan Afghanistan to the Himalaya Western China Saghalin and Korea Korea Asia Minor to western China Transcaucasia, Asia Minor and Persia ? Kamchatka Afghanistan to Central China Central China Northwest China Central and southern European Mts. Lp “TOA ] WOLAXYOMUY CIONUV AHL JO TYNUNOL I ‘VYAOINOT ‘NAAAD ® OUAANAAGNA [9961 A SPECIES nN 2n AUTHORITY DOCUMENTATION GENERAL AND COLLECTOR DISTRIBUTION py aae Foes Rehd. a2. Section Lontcera (Sect. Coeloxylosteum Rehd.) L. tatar (9) (9) (18) (18) 18 18 9 9 18 9 18 f. angustifolia (Kirchn.) Rehd. 9 f. elegans Carr. 9 f. fenzlii Dippel 9 f. grandiflora (18) f. latifolia (18) f. leroyana (Zabel) Rehd. 9 18 ~ : f. lutea (Loud.) Rehd. 9 f. punicea (18) f. rosea Regel (18) 9 f. roseo-alba Regel 18 f. sibirica (Pers.) Rehd. (18) 18 9 f. virginalis Jaeger 9 cv. Arnold Red 9 cv. Hack’s Red 18 cv. Morden Orange 9 cv. Sheridan Red 9 var. 9 cv. (unnamed dwarf) 18 9 18 Eastern Europe to Turkestan AA 328-39- te igliae 28 May AA 287-41- A, ke 25 May 1964 AA 280-41. A, Green, 31 May AA lige tine 20 May AA 305- st, on 26 May 1964 AA 326-39, Kreps, 25 May 1964 AA 306-32, Rehder ?, 21 May & 12 July 1915 (as AA 7337) AA 566-57-A, Kreps, 25 May 1964 AA 18293, Kobuski & Metcalfe, 9 May 1930 AA 7335, Rehder, _ May & 12 July 91 5 Aa le Rehder, 21 y & 12 July 1915 AA 3771-1-A, Kobuski & Metcalfe, 9 May 1930 (as AA 3771-1) AA 716-45-A, Kreps, 26 May 1964 AA 199-53, Kreps, 25 May 1964 AA 243-49 (Type clone), Kreps, 25 May 4 AA 475-56, Kreps, AA 476-56, Kreps, 25 May 1964 AA 791 a oe 25 May 1 AA 304— is oe 20 May 1965 [9961 Ly “TOA ] WOLAYOTUV GIONUV AHL AO IVNWNOL 9¢z I ‘VYROINOT ‘NAAAD ¥ OUAANAACONA : DOCUMENTATION GENERAL SPECIES n 2n AUTHORITY AND COLLECTOR DISTRIBUTION L. X xylosteoides Tausch 9 12* AA 5406, Palmer, Cultivation : 7 July 1936 L. floribunda Boiss. & Buhse 18 12* AA 341-44, Kreps, Northern Persia ie 1 June 1964 L, korolkowii Stapf. (9) 2 Turkestan, Se and Paki (18) 9 18 12* AA 976-6-A (Type clone), sa 31 May 1891 (as AA 976), an Dudley, 10 July 1 9 18 12* AA ees : Dudley, 10 June 1965 var. aurora Koehne 18 12 AA 20444, Palmer, 8 Sept. 1936 var. floribunda Nicholson 9 12* AA 10083-2-A, Palmer, 27 May 1936 L. X amoena Zabel (18) 7 Cultivation (backcross to L. tatarica) 9 12* AA 818-35, Green, 26 May 1965 18 i? AA sis i 25 May var. arnoldiana Rehd. 9 18 12* AA 180-44. a mes 25 May 1964 f. alba Zabel 9 12* AA 793-49-A & B, Kreps, 25 May 1964 —— ia _ ay. a , v 77) Subsect. dimers (Zabel) Rehd. L. ruprechtiana Rege (9) (18) 6 Manchuria to northern China (18) 7 9 12* AA 15111, Rehder, 19 May 1910 var. calvescens Rehd. (18) 7 9 12* AA 15112, Kobuski & Metcalfe, 16 May 1930 f. xanthocarpa (Zabel) Zabel 9 12* AA 15109-1-B, Rehder, 6 Aug. 1919 (as AA 15109-1) 9 18 12* AA352-62, Kreps, 25 May 1964 L. X notha Zabel 9 12* AA 15134-A, Green, Cultivation (backcross to L. tatarica) 24 May 1965 9 12* AA 572-1-A, Palmer, 15 May & 7 July 1936 (as AA 572 f. alba Zabel 9 12* AA 969-4, Palmer, 7 July 193 9 12* AA 348-32, Kobuski & Metcalfe, 16 May 1930 (as AA 3744-1) f. carnea Dippel 9 12* AA 5403, Kobuski & Metcalfe, 16 May 1930 f. carneo-rosea Zabel 9 12* AA 3743, Palmer, 7 July 1936 BET Ly “ToA] WOLAXOPUVY GIONYV AHL AO IWNUMNOL [9961 I ‘VYAOINOT ‘NAqYD X YJadHdNAGllda 6¢Z SPECIES L. demissa L. morrowii A. Gray L. X bella Zabel ee wy: Ord DOCUMENTATION GENERAL n 2n AUTHORITY AND COLLECTOR DISTRIBUTION L. X salicifolia (Dippel) Zabel 18 izg*. AA oo ig Cultivation 25 May (9) 2 Japan nent (9) 4 Japan = (9) (18) 6 ys) (18) 7 = 9 18 12* AA 1283-65, Kobuski & SS Metcalfe, May = 1930 (as AA 15086) fe f. xanthocarpa Nash ex Teuscher 9 12* AA 18902-A, Kreps, ae 25 May 1964 a (9) (18) 6 Cultivation a 9 12* AA 10087-A, Green, D 27 May 1965 18 12 AA 22206, Palmer, ws) 17 May & 8 Aug. > 1936 oo (backcross to L. morrowii) 9 12* AA 181-38-B, Green © 20 May 196 e 9 18 12* AA 48-42-A & B, ne. Kreps, 25 May 1964 J (backcross to L. morrowii) 18 12* AA 600-58, ie 24 May 196 f. atrorosea (Zabel) Rehd. 9 12* AA 346-32, oe & Metcalfe, 16 May 3 1930 (as AA 6290) iy = a ip ~ ~~ 0 f. candida (Zabel) Rehd. 9 18 12* AA 3475, Palmer, g 18 May 1938 fi 18 12* AA 3475-1, Palmer, 15 May & 7 July 1936 9 12* AA 22843, Green, f. rosea (Zabel) Rehd. 31 May 1963 9 12* AA 727-49, Green, = cv. Dropmore 26 May 1965 a, G ae 9 18 6 ultivation = L. X minutiflora Zabel _ (18) ae) Aa eee Palmer, Zz 15 May (as I AA 2-32 a i 18 12* AA 290-41-B, Green, cv. Valencia 31 May 1965 | nl 1ensi. 18 7 Cultivation ee L. X muendeniensis Rehd. : (18) 12* AA 4996, Rehder, & 27 May 1916 2 . ‘4 12* AA 15110, Green, (possible backcross to L. morrowii) 9 ees ea cr S 9 12* AA 23152-A, Green, = 21 July 1 2 9 18 12* AA 795-49-A & B, re) Kreps, 25 May 1964 P _ f. xanthocarpa Hort. 9 12* oe ae -A, Kreps, L. X muscaviensis Rehd. 9 18 12* ar ae Kobi & Cultivation 1930 18 12* AA 355-53, Green, 22 May 1965 = DOCUMENTATION AND COLLECTOR GENERAL DISTRIBUTION SPECIES n 2n AUTHORITY L. xylosteum L. (9) (18) 6 Europe to Altai Mts. (9) 9 (18) 7,10 9 12* AA 15142-A, Dodd & Callaway, 25 June 1965 9 12* AA 422-39, Green, 31 May 1965 9 12* AA 57-42, Kreps 26 May 18 [2° AA eae aie 26 May 1 9 12* AA 281-53, Green, 31 May 1965 f. mollis (18) 7 cv. Claveyi 18 12* AA 349-55, Kreps, 25 May 1964 L. chrysantha Turcz. (9) 2 Northeast Asia and Japan (18) 7 9 12* AA 61-29, Kreps, 25 May 1964 var. longipes Maxim. (18) 7 Northwest China 18 12* AA 21592 (Rock, W. Kansu, 1925), Kreps, 25 May 1964 9 12* AA 558-51, Kreps, 26 May 19 -- : - ie cn Maile ’ ¥ t f. latifolia Korshinsky 9 12* AA _ 15008-2, Kobuski & Allen, 17 May 1933 f. regeliana (Kirchn.) Rehd. 18 12* AA 15015, Palmer, 7 May 1921 f. villosa Hort. ex Rehd. 9 12* AA 5297-B, Green, 21 July 1965 L. X pseudo-chrysantha (18) 7 Cultivation L. koehneana Rehd. (18) : Western China 9 12* AA 5976, Palmer, 10 July 1936 9 18 12* AA 15043 (Wilson 93a, W. Hupeh, 1907) oe ?, 29 May 1 L. maackii (Rupr.) Maxim. (9) 2 Manchuria and China (9) (18) 6 (18) 7 9 12* AA 15047, Kobuski & Metcalfe, 2 June 1930 9 12* AA 1026-37-A, Green, 21 July 1965 f. erubesce (18) 7 f. se peti Franch. ex Rehd. 18 12* AA 10085-B, Kreps, China 25 May 1964 L. deflexicalyx Batal. (18) 7 Western China 18 12* AA 6674 (Wilson 4179, Szechuan, 1910), Rehder, 3 June 1918 ) 18 12* AA 4801-2-B, Green, Cultivation L. X vilmorinii Rehd. 21 June 1965 che Ly “I0A] WOALAXOFUAV GIONAV AHL JO TYWNUAOL I ‘VYAOINOT ‘NAAYD ¥ ONAANAGAY [9961 fez DOCUMENTATION GENERAL SPECIES n 2n AUTHORITY AND COLLECTOR DISTRIBUTION L. trichosantha (18) 7 Western China f. acutiuscula Rehd. 12 AA 15106-1, Palmer, 19 May & 18 Sept. 1936 L. pros (9) 2 Western China L. qu ie oacatind (9) 2 Afghanistan to the Himalaya (9) (18) 6 f. translucens (Carr.) Zabel 12* = Beet Pinamal 2s 9 iZ* ee 2 ot oe Subsect. BREVIFLORAE Rehd. L. alseuosmoides 18 June 1940 Sect. Nrntooa (Sweet) Maxim. adA WOLAYOFTAV GATIONUV AHL AO TYNUNOL (18) 2 Western China (36) i “coriacea”’ € (54) is ‘ basi Hemsl. (27) 2 Western China (54) 7 54 12* AA 5968-A (Wilson 254, W. pay eee Green, 3 Nov. Subsect. LoNGIFLoRAE Rehd. L. biflora (18) 7 Southern Europe ‘3 and northeast- Ee ern Africa = be é ~- ~ a i é a . Py L. affinis var. pubescens (8) 8 ? Chuang & Liu, TIC China and Japan a 4945 [not seen] s L. japonica (9) 2 Eastern Asia (18) 7 L. hildebrandiana (18) 7 Burma and south- Subgenus CapriroLium (Adans.) Dippe Subsect. PHENIANTHI (Raf.) Rehd. L. sempervirens L. cv. Dreer’s Everblooming cv. Magnifica L. X tellmanniana Magyar ex Spath Subsect. CYPHEOLAE (Raf.) Rehd. 7 — ensis L. dioi L. prolifera (9) (18) 6 (18) 3 36 12* 18 i7f 18 12* (ca. 45) 7 (18) 7 (9) 2 18 12* (G) Z (18) 7 western China 1 (Subgen. Periclymenum ( Mill.) Rehd.) Eastern and central North America AA 771-57-B, Dodd & Callaway, 4 June 1 965 AA ae oi 30 July AA 543- a Duley 30 July Cultivation Southwestern China Eastern Nort America AA 130-49, Kreps, 1 June 1964 Central North America * No such variety has been traced, it is suspected that this may have been an error for L. henryi Hemsl. var. subcoriacea Rehd. ‘VUAOINOT ‘NAAAD ¥ OUAANACOAA I Svc DoCUMENTATION GENERAL SPECIES n 2n AUTHORITY AND COLLECTOR DISTRIBUTION L. flava Sims 9 12* C. E. Wood, Jr. 8919, Eastern North Walker Co., Georgia merica (A); AA 557-59, a clonal propagation from this collection 18 12* i od, Jif. 8963, DeKalb Co., Alabama (A); AA 556-59, a clonal propagation from this collection, Green, 31 May 1965 Subsect. EUCAPRIFOLIA (Spach) Rehd. L. implexa (18) 7 Mediterranean L. caprifolium (9) (18) 6 urope and western Asia (18) 7 L. X americana (Mill.) K. Koch (18) 7 Cultivation 9 12* AA 18-43-B, Dodd & Callaway, 4 June 1965 L. stabiana (9) 1 Italy L. splendida (18) 7 Spain L. etrusca Santi (18) 7 Mediterranean 9 18 12* AA 231-46, Dodd & Callaway, 4 June 1965 a ci ds - ¥ : / é » L. periclymenum L. (18?) 5 Europe, western Asia and northern Africa (18) (36) 6 (18, 36) 7 54 11 Gadella & Kliphuis 534, 623, & 715 (v) [not seen] f. belgica (36) cv. Wisleyensis (36) 7 Unidentified Lonicera sp. (36) ARNOLD ARBORETUM HARVARD UNIVERSITY Present address (for P.S.G.): OYAL BOTANIC GARDENS Kew, RICHMOND, SURREY GREAT BRITAIN 97 Le “I0A] WOLAYOTUVY GIONUV AHL AO TWNUNOL I ‘VUADINOT ‘NAAWD ® OMAANAGAY [9961 Lez 248 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 ANATOMY OF THE PALM RHAPIS EXCELSA Il. RHIZOME * P. B. ToMLINSON AND MartTIN H. ZIMMERMANN Tue PALM Rhapis excelsa develops horizontal, subterranean rhizomes which bear only non-assimilating scales. This kind of axis may be con- trasted with that of some other palms like Nypa or Sabal minor in which the foliage leaves are borne directly by the underground stem. Scale-leaf bearing rhizomes of the Rhapis type are relatively rare in the Palmae (Tomlinson, 1961) although common in other monocotyledonous orders (e.g., Scitamineae, Liliiflorae). Anatomical comparison of the scale-leaf bearing rhizome with the aérial axis which bears the assimilating foliage is of considerable interest because the two types are morphologically equivalent but their development must be quite different from a nutritional point of view. The present paper is an account of the vascular structure of the mature rhizome. The develop- ment of the two types of axes will be described in a later paper in this series. Palms show two major growth forms. In one kind the vegetative axis is unbranched, producing solitary columnar stems, as in the coconut, in the other it is caespitose by the production of axillary vegetative axes, as in the date. Branching of the second type is always restricted to the base of the parent axis for simple adaptive reasons which are well known (Tomlinson, 1964). Normally these basal suckers rapidly turn erect. The first leaves produced by lateral branches are reduced to scales, but there is a rapid transition to normal assimilating foliage leaves as the branch assumes a vertical position. Thus the construction of the parent axis is repeated and by further branching of all axes a tufted and congested sympodial system results. Rhapis represents a growth form in which mutual competition of successive segments of the sympodium is eliminated by long-continued A which has become much protracted. In some ways it is a ‘persistently juvenile” stage of axis development. MORPHOLOGY OF THE RHAPIS RHIZOME Axillary shoots which grow out as rhizomes appear in the seedling * of Rhapis. After a variable period of horizontal growth the rhizome apex turns * Research on the anatomy of palms by P. B. Tomlinson is supported by National Science Foundation Grant GB 299 * Seedling morphology and anatomy will be described in a subsequent paper. 1966 | TOMLINSON & ZIMMERMANN, RHAPIS EXCELSA 249 erect and there is a rapid period of transitional growth leading to the establishment of an erect leafy shoot. Features of the construction and growth of rhizome segments are shown in Fics. 1 and 2. In the transitional period of growth the axis is somewhat swollen (Fics. 2B, C, D, 1B) and successive leaves are elaborated so that a transition between the scale leaves of the rhizome (Fic. 1C) and the foliage leaves of the aérial stem occurs (Fics. 1D-F). Transitional leaves show a gradual development of the blade which may be initially either compound or simple depending on the vigor of the shoot. Only compound leaves are illustrated in Fic. 1D-F. This transitional sequence clearly illustrates that the scale leaf is morpho- logically equivalent to the base of the adult foliage leaf. The sympodium is continued by new rhizomes originating as buds in the axils of scale or transitional leaves at the base of the newly-erect shoot q UU ¥ w py: a Sis: Boge Ficure 1. Relation of aérial to underground axes in Rhapis excelsa. A. One segment of sympodium, foliage leaves removed from aérial shoot at left, < 1/8. B. Base of right-hand aérial shoot of Fic. 1A, all but distal leaves removed, oldest proximal, number 4 is on the remote side of the axis. C. a hr from rhizome, X 1/2. D-F. Successive leaves from transitional sequen t base erect shoot, X 1/4, showing gradual deve stoeonstt 6 of blade. Adult foliage oes have blades ts times wider than that of leaf illustrated in Fic. 1 250 JOURNAL OF THE ARNOLD ARBORETUM [ VoL. 47 (Fic. 2D). Several buds may grow out, in the order of their age (Fic. 1B) and in this way the sympodium multiplies. With frequent production of laterals sympodia may become very complex (Fic. 2E). Clonal spread of Rhapis under favorable conditions is therefore rapid. Erection of rhizome FIGURE 2. Development ck rhizome system in Rhapis excelsa. x 1/4. underground creeping rhizom transitional type of foliage inated x 1/4. C. Later stage, transitional leaves wit x : . Base of erect shoot at a stage . 2C; leaves removed to — buds of renewal o Fic lea shoots, X 1/2. E. Old, much- aces oe thizome complex, X 1 deve elopment corresponding t A. Apex of en 1966 | TOMLINSON & ZIMMERMANN, RHAPIS EXCELSA 251 apices and the morphological responses it induces seem independent of the season in South Florida since all stages of development can be found at any time of year. Horizontal, underground parts of the rhizome, with which this article is concerned, may be up to 1 meter long and 1 to 1.5 cm. in diameter; the apex is protected by the rigid overlapping scales (Fic. 2A). Scales, like the foliage leaves of the aérial stem, are arranged in a 2/5 phyllotactic spiral. Each originates as a closed tube with a narrow distal opening which is widened by enlargement of enclosed organs (Fic. 1C). Internodes are 1.5 to 2 cm. long in older parts of the rhizome. The covering of scale leaves is retained for a long time (Fic. 2B). Very old rhizomes still bear the fibrous remains of decayed scales. Neither buds nor their vestiges are associated with scale leaves; they are restricted to the transition region where the rhizome turns erect (Fic. 2D). Roots are borne all around the rhizome at irregular intervals. We have not recorded this quantitatively but it seems to us that root insertions are associated (a) with each other and (b) less obviously, with the nodes. Root production is also vigorous on the swollen regions at the junction of successive sympodium segments. Rhizomes are always colorless and non-assimilating. MATERIALS AND METHODS Rhizome segments used in this investigation are all from specimens culti- vated at Fairchild Tropical Garden. Sequential sections 30-50 » thick were cut at 0.5 mm. intervals, stained and mounted by procedures pre- viously described (Zimmermann & Tomlinson, 1965). In order to clarify the vascular system, starch was eliminated from the tissue by immersing sections for 5 to 10 minutes in 50 per cent hydrochloric acid prior to wash- ing and bleaching. For details of nodal anatomy a continuous series of sec- tions from a short length of rhizome was prepared. Cinematographic analysis was carried out with the drawing method de- scribed before (Zimmermann & Tomlinson, 1965) as well as with the was also very useful in making quantitative plots of single vascular bundles of the types illustrated in Fic. 4. In addition to microscopic analysis we have studied the rhizome struc- ture extensively by macro-cinematography of cut surfaces on the micro- tome. A special clamp for the ‘Reichert’ sliding microtome was designed and constructed which allows continuous advance of long specimens. Such motion picture sequences give an excellent over-all view of the vascular system. GENERAL RHIZOME ANATOMY OF RHAPIS An over-all view of a transverse section of the rhizome is given in G. 3 Rhizome 1-1.5 cm. in diameter, divided into cortex 1-1.5 mm., central [ VOL. 47 JOURNAL OF THE ARNOLD ARBORETUM ee + Fy ie Ot Oe, a2 x, J ws aM a, 2S, -> cx sy Se: aes ae) Kc 8 LS ou * vans, ae La Ls % od yf. és" * 3 ie,» © é “tol ig % & Jth ' At § ; ‘ete, or #06 +2. ae oe 28% ) &e es hud hae 6 ’ ig -~ Yk i rig ® x ae. ae —* %e% one © Sous E25 * M ~“8~o Begs 2Ee0°0 aug = ea Mt oe Se" a ow hs os 5.64 phew | Kase eo Or Ena oe Severe Neo — 4 oo. soi a ee ot SEL» — 2 5258 oA EO On oo 4 Aem & oe | 09. > OE ene od o £ OGG eS me fae ee a seu~ mH» Vow , S..f& Ss © j=) SSerE bad So Bu Hqo™ Oe &o Oo to's Scar ss Song? Rone © RA cS 3 §< 2's Sn awn io ao Bo fe” 5 8:5 vj [zy Sei | nges “ea SB OS ao “oe mn S44 og 43 wo wm 3 a= 0 aU Se on ooo Rae 3 ~~ oo OOS » S sr Ss e2o 35% o” ¢ o. & wn 2 bd rat Oe eee oss ae ELS oe oe g 53 2 & 2 °3 wv ae Le 33° — tee a° o aes ey ae a ee ton! eG Sas ua “pe z. ee Se eS ifs = Me a ge q sis On B 1966 | TOMLINSON & ZIMMERMANN, RHAPIS EXCELSA 253 occasional cells with pronounced pitted thick secondary walls becoming ligni- fied. Stomata not infrequent, resembling those of aérial stem but without chloro- plasts and presumably non-functional. Periderm not developed except locally around wounds and leaf scars by ligno- suberisation of ground parenchyma ac- companied by divisions of “etagen” type. Cortex with 3-4 outermost layers of narrow cells with truncate end walls; walls thickened and becoming ligno-suberized to form a narrow hypodermal sclerotic layer interrupted below stomata by loose, wide unlignified cells con- tinuous with middle cortex. Middle and inner cortical ground parenchyma with a well-developed intercellular _ Space system, cells frequently becoming slightly 120 » wide) including a narrow central strand of vascular tissue, often only phloem. Innermost cortical layers ipa free of fibrous bundles. Central cylinder not delimited from cortex by endodermis or other specialized layer, but by abrupt transition to narrow compact ground parenchyma with slightly thickened walls and zone of narrow, somewhat congested peripheral vascular bundles. Compact peripheral zone only interrupted by parenchyma at exit of leaf traces to form distinct “leaftrace gaps” (Fic. 6). Central vascular bundles wider, more diffuse. Central ground parenchyma very uniform, cells somewhat wider and looser than those at periphery. Individual vascular bundles of central cylinder more or less completely sheathed by fibrous tissue, but fibers least well developed around xylem. Fibers narrowest and most heavily lignified next to phloem. Peripheral vascular bundles somewhat more fibrous than central, but difference much less pronounced than in aérial stem. Vascular tissues includ- ing usually one wide metaxylem vessel and often protoxylem in varying amounts (see below). Phloem strand narrow, uniform, never with a ome median sclerotic isthmus as is common in bundles of aérial stem. Metaxylem vessel elements average 60 » wide, 1-1.5 mm. long with oblique or slightly oblique scalariformly perforated end walls with 6-10 thickening bars. Sieve-tubes tanniniferous deposits giving a strong positive lignin reaction with phloroglu- cinol and concentrated HCI; tannin occlusions not especially restricted to traces to old leaves. Stegmata (silica a only observed next to cortical fibrous bundles. Raphide sacs not obse In summary, the general anatomy of the rhizome, as seen in a single transverse section, differs from the aérial stem as follows: relatively greater development of cortex and cortical fibrous system; central and peripheral vascular bundles of central cylinder less conspicuously different; less obvious variation in the numbers of metaxylem elements. COURSE OF VASCULAR BUNDLES IN THE RHIZOME Central system. In spite of frequent root insertions the over-all course of vascular bundles in the Rhapis rhizome remains quite clear. It resembles 254 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 -CORTEX CENTRAL CYLINDER CM. | \ LEAF = | , Se 3F- LEAF 24 ; MAJOR i LEAF \ CORTICAL INTERMEDIATE L foot eat a 1 MM. j 2 3 4 RHIZOME AERIAL STEM FicuRE 4. ABoveE: plots of vascular bundles supplying scale leaf of rhizome. Measurements taken from microtome sections. Course of major, intermediate, and minor bundles from central cylinder, with continuing vertical bundles, and a cortical bundle without a continuing vertical bundle. The figure is diagram- 1966 | TOMLINSON & ZIMMERMANN, RHAPIS EXCELSA 255 the aérial stem and may be described in the same terms. Vascular bundles behave essentially alike, maintaining their individuality throughout the stem. At irregular intervals each of these vertical bundles * forks to give off a leaf trace which passes sharply outward into a scale leaf, the vertical bundle continuing distally, eventually to repeat the cycle. Each leaf re- ceives a few branches from the central bundles (major bundles) which pass primarily to the dorsal side of the leaf, a larger number of branches from sub-peripheral bundles (intermediate bundles), and the largest number of branches from peripheral bundles (minor bundles) which pass to all parts of the leaf. There is a continuous series from major (dorsal) to minor (ventral) leaf traces around the stem circumference at each node. Major bundles pass into the leaf much more abruptly than minor bundles (Fic. 4); they may sometimes be horizontal or even pass below the horizontal and are commonly sinuous so that different parts of the same leaf trace occur in a single section. Part of this unevenness is due to the oblique in- sertion of most scale leaves. The most significant way in which rhizome bundles differ from those of the aérial stem in over-all distribution is that rhizome bundles deviate from the vertical at points of leaf contact much less than in the aérial stem. Von Mohl’s classic ‘double curve” which is so characteristic of the vascular bundles of aérial stems of palms is therefore not so obvious in the rhizome (Fic. 4 below). Likewise the central bundles of the rhizome do not describe a regular helical path as is common in the aérial stem.* A slight movement does, however, tend to occur in one direction, suggesting an incipient helix. As in the aérial stem, departure of the leaf trace is complicated by the presence of bridges which link it with adjacent vertical bundles, always in an upward direction (Fic. 5D). One to four (usually two) bridges are de- veloped by each leaf trace. Narrow bridges may contain only phloem; the narrowest are sometimes fibrous strands without vascular tissue. Most bridges are short, 2 to 6 millimeters long, but occasionally there are much longer ones resembling vertical bundles until they fuse distally. The significance of this is discussed below. Satellites, which in the aérial stem irrigate inflorescences (or their aborted vestiges), are absent from the rhizome which wholly lacks this lateral type of appendage. Brief mention may be made of irregular girdling traces (Fic. 6 GT) ® Vertical bundles of the rhizome run horizontally, of course, but the term is used necatse they are homologous with the vertical bundles of the aérial axis. *More recent observation of the aérial stem of Rhapis has revealed that internal helix is correlated with phyllotaxis; stems with a right-handed ate spiral have a right-handed internal helix and vice versa matic in that the bundles are all represented in one plane, in fact they enter the leaf along different, widely divergent radii. Dotted vertical line represents limit between cortex and central cylinder, arbitrarily chosen as a base line in making pdboenaeniy Horizontal exaggeration 12. 5 times BELow: diagrammatic comparison of course of vascular bundles in rhizome and aérial seen (latter figure taken from Zimmermann & Tomlinson, 1965). For ee purposes the fanaa: rhizome is represented in an erect position 256 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. 47 : eg a4 c 3 gy Be Cent Loon Ee Ey i. Se) {A 2 % oe O08 J “eo FicurE 5. Successive transverse sections of the same leaf trace complex from rhizome of Rhapis excelsa. In all figures the cortex is towards the upper, the central cylinder towards the lower margin of the te. Scale is 250u. A. Vascular bundle in uncrowded central part of stem, below level of forking and with well-developed protoxylem. B. Vertical bundle (VB) to right splitting from leaf trace (LT) at left; 1.85 cm. above A. C. Vertical bundle (VB) at right including one wide metaxylem vessel, no protoxylem, free of leaf trace (LT) at left with no wide metaxylem elements; 2.25 cm. above A. D. Leaf trace further towards cortex splitting off bridges (B) to left and right; vertical bundle is now left behind and is no longer associated with leaf trace; 2.9 cm. above A. - 1966 | TOMLINSON & ZIMMERMANN, RHAPIS EXCELSA rsa | among the peripheral bundles of the central cylinder. These appear to be the ultimate extensions of the root trace system which will be described in a subsequent article. Cortical system. Fibrous (less commonly narrow vascular) bundles of the cortex are continuous with the same system in the leaf. Traced downwards from the leaf insertion they are seen to form an irregularly anastomosing system which remains quite independent of the central vascular system. Cortical bundles all ultimately seem to end blindly below; occasional cortical strands which end blindly in each direction have been observed. Endings are, however, always near other fibrous bundles and suggest branching during early stages of development. Larger bundles with distinct vascular tissue connect the leaf with the inner part of the cortex (e.g., Fic. 6 CBL and cortical bundle plotted in Fic. 4). Smaller bundles (Fic. 6 CBS), usually without vascular tissue, remain in the peripheral part of the cortex. DETAILS OF VASCULAR BUNDLE ANATOMY Variation in the anatomy of a single bundle throughout its length re- sembles that described for the aérial stem (Fic. 5). A vertical bundle at its point of divergence from a leaf trace includes one wide metaxylem vessel, no protoxylem (VB in Fic. 5C). At some higher level narrow protoxylem elements appear and increase to reach a maximum number of elements just below the next higher leaf contact (Fic. 5A). At the forking of the vascular bundle the leaf trace loses its metaxylem to the vertical bundle (Fic. 5B, C), so retaining only protoxylem. Scale leaves are thus irrigated solely by protoxylem (Fic. 5E, F). Unlike the aérial stem, however, there is little or no proliferation of metaxylem elements just below a leaf contact because there are fewer branches (no satellites). Proliferation of metaxy- lem represents vessel overlap. Overlapping vessels occur regularly where vascular bundles branch, for example in the leaf-trace complex and where bridges unite with vertical bundles (in the vertical bundle immediately to the right of the leaf trace in Fic. 5E), and at the periphery of the central cylinder under the influence of root insertions (GT in Fics. 5 and 6). COMPARISON OF THE RHIZOME WITH THE AERIAL STEM Described in qualitative terms, the vascular anatomy of the Rhapis rhizome seems to differ little from that of its aérial stem. Important differ- ences become obvious, however, when quantitative aspects are considered. These show that the over-all plan is less regular than in the aérial stem. We are content at the moment merely to draw attention to this plasticity, an explanation of which must be deferred until development of the vascular Leaf trace passing through ele periphery of central cylinder; 3.15 c ahi ove A. (Configuration of metaxylem of adjacent vertical bundle to right Suggests a recent bridge linkage ig haere leaf trace.) F. Leaf trace in cortex passing one hhesisonially into leaf; 3.65 cm. above A. 258 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 &e® Pad ¥ . = Ficure 6. Rhapis excelsa, transverse section of rhizome including cortex and ery of central cylinder immediately below a node. Sclerotic base of leat j f fi cylinder, t cortical bundle ( ) passing into the leaf from the inner cortex and the system of many small cortical bundles (CBS) diverging from the outer cortex. Girdling traces (GT) are ultimate ends of root traces. Scale equals 1 mm. system in the rhizome apex is considered. Because relatively few leaf traces enter the scale leaves in proportion to the total number of central cylinder bundles, it is clear that on average any bundle must make a leaf contact 1966 | TOMLINSON & ZIMMERMANN, RHAPIS EXCELSA 259 TABLE I. Essential Differences Between Rhizomes and Aérial Stems RHIZOME AERIAL STEM MorPHOLOGY Horizontal Vertical Subterranean Aérial Roots present Roots absent Scale leaves, protective, non-assimilat- Foliage leaves, assimilating Axillary buds praia distally as lateral inflorescence Axillary buds absent Vegetative buds abundant in the transition region between rhizome and aérial stem. ANATOMY Cortex wide, with conspicuous system Cortex narrow, with few fibrous bun- of fibrous bundles dles Peripheral congested bundles of cen- tral cylinder mechanically insignifi- cant Leaf contacts irregular, peripheral bun- Peripheral congested bundles of cen- tral cylinder mechanically significant Leaf contacts relatively uniform dles often of “indefinite length” Internal helix not obvious Internal helix obvious Vascular bundles deviate relatively lit- Vascular bundles deviate considerably tle at leaf contact at leaf contact Branching at leaf contact: Branching at leaf contact: Leaf trace Leaf trace Vertical bundle Vertical bundle Bridges Bridges Satellites Metaxylem vessel overlap not pro- Metaxylem vessel overlap pronounced nounced in leaf trace complex (few in leaf trace complex (many branches) branches ) much less frequently than in the aérial stem. Measurements confirm this. Peripheral bundles without protoxylem have been traced up to distances 20 cm. without change. This implies vascular bundles of “indefinite length.”” However, once protoxylem appears in such a bundle, the bundle will make a leaf contact by splitting off a leaf trace. Protoxylem-bearing lengths of vascular bundles from 3 to 14 cm. were measured in the rhizome. A measured leaf contact distance of 25 cm. (10 cm. bearing protoxylem, 15 cm. protoxylem-free) over 15 internodes compares favorably with central bundles in the aérial stem; values of 12 cm. (6 protoxylem-free, 6 protoxylem-bearing) over 6 internodes for a peripheral bundle may be cited 260 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 to show that these may make frequent leaf contacts. However, we do not wish to emphasize these few isolated figures in the absence of develop- mental information Plasticity of the vascular system of the rhizome is also demonstrated by occasional vertical bundles split. As in the aérial axis, leaf traces occasion- ally give off either two or no vertical bundles. Once we recorded a single vertical bundle, originating as two separate strands one from each side of a leaf trace, which fused distally. Vertical bundles of the central cylinder never end blindl Essential differences between rhizomes and aérial stems are summarized in Tasie I. Structural differences between rhizome and aérial stem are based on the difference in the physiology of their respective development. In the aérial stem, leaf contacts are established in the apex; we assume therefore that mature leaves supply developing leaf primordia and young leaves. The rhizome, on the other hand, is never an assimilating organ; movement of nutrients must, therefore, be along the axis toward the apex. In this light, the presence of bundles of “indefinite length’? seems explic- able. However, differences between the aérial stem and the rhizome can only be fully understood when the anatomy and development of their respective apices are known. Such studies are under way; a report about them will be published in a later paper of this series. SUMMARY Analysis of the vascular system of the subterranean, scale-bearing rhizome of Rhapis excelsa by cinematographic methods demonstrates a sys- tem with the essential features of the aérial stem. Satellite bundles are absent since the rhizome bears no lateral inflorescences. The vascular sys- tem is less regular than that of the aérial stem and deviation of vertical bundles at points of leaf contact are less pronounced. Leaf contact inter- vals in peripheral bundles may be very long. These differences must be related to the non-assimilating nature of the rhizome. ACKNOWLEDGMENTS Miss Lesley Jackson carried out the routine serial sectioning on which this study is based. Her reliable help is very much appreciated. We are indebted to Miss Priscilla Fawcett, Botanical Illustrator at Fairchild Tropical Garden, for Ficurrs 1 and 2 LITERATURE CITED TomLtnson, P. B. 1961. Palmae. Jn: C. R. MeETCALFE, ed. Anatomy of the Monocotyledons 2. xv + 453 pp. Clarendon Press, Oxfor d. 1966] TOMLINSON & ZIMMERMANN, RHAPIS EXCELSA 261 . 1964. Stem structure in arborescent monocotyledons. Jn: M. H. Z1m- MERMANN, ed. The formation of wood in forest trees. pp. 65-86. Academic ZIMMERMANN, M. H., & P. B. ToMLINSON. 1965. Anatomy of ee es Rhapis excelsa I. pees vegetative axis. Jour. Arnold Arb. 46: 1 ———————. 1966. Analysis of a vascular systems in ce Optical shuttle ere Saas 152: 72-7 HarvVARD UNIVERSITY Casot FouNDATION, PETERSHAM MaAssACHUSETTS 013 AND FAIRCHILD TROPICAL GARDEN MIAMI, Fiorina 33156 262 JOURNAL OF THE ARNOLD ARBORETUM [VOL. 47 REDISCOVERY OF DAPHNOPSIS HELLERIANA Lorin I. NEVLING, JR., AND Roy WoopBURY Amos ArTHUR HELLER (1867-1944) first discovered and collected plants of the Puerto Rican species Daphnopsis helleriana Urban. Several biographical sketches of Heller are to be found but they emphasize his botanical contributions to the study of plants of the United States. In 1898, “Through the kind liberality of Mr. Cornelius Vanderbilt, Mr. and Mrs. A. A. Heller have been sent to the island of Porto Rico to collect specimens illustrating the flora and natural vegetable resources of the new colony.” ! They returned to New York in June of 1899 “. . . with nearly 8,000 specimens, and copious notes referring to them.” This collecting excursion sometimes is mentioned in the biographical sketches but is not important in terms of the present note. In 1900, however, Heller ‘. . . returned to Puerto Rico to extend his collections made in 1899. He will keep the field during January and February, with headquarters at Mayaguez in the western part of the island.” * This trip may have been financed personally by Heller for the published herbarium accession record of the Garden shows ‘450 herbarium specimens given by Mr. A. A. Heller.” * If financed by the Garden, or special funds through the Garden, the specimens would not have been accessioned under the gift category. Further, the following notice ap- peared in Muhlenbergia (1: fly leaf following page 30. August 9, 1900): “Several sets of my Porto Rican collection of 1900 may still be had at the rate of 10 cents per specimen, carriage prepaid. They contain about 100 species of flowering plants, among which are a number of species confined exclusively to Porto Rico, and one new species, Daphnopsis Helleriana Urban.” In the same issue of Muhlenbergia (1: 10-17) an article by F. S. Earle included identifications of Heller’s collection of fungi from this trip. The numbers and dates show that Heller did not restrict his collect- ing to the western part of the island but collected also in the vicinity of Santurce and Bayamon. It was while collecting on the “calcareous hills near Bayamon” on February 23 that he found (no. 4689) D. helleriana. Thereafter, this plant was not recollected for nearly sixty years. lack of subsequent collections was particularly frustrating to Nevling in the preparation of a taxonomic revision of Daphnopsis. Al- *N. L. Britton. Report of the Secretary and Director-in-Chief, Bull. N. Y. Bot. Gard. 1: 182. 1899. *N. L. Britton. Report of the Secretary and Director-in-Chief. Bull. N. Y. Bot. Gard. 1: 309. 1900. * Anon. News, Notes and Comments. Jour. N. Y. Bot. sah 1: 16. 1900. * Anon. Accession. Jour. N. Y. Bot. Gard. 1: 108. July, 19 1966 | NEVLING & WOODBURY, DAPHNOPSIS 263 though Heller’s “type collection” had received relatively wide distribution it consisted of immature staminate specimens exclusively. An attempt to recollect the species, in January of 1963, by Nevling and Richard A. Howard was not successful.® Many of the limestone hills were (and are) in the process of being leveled to provide raw materials for the building industry. It was feared that the locality might have been destroyed. Howard discussed the Daphnopsis helleriana mystery later in the same year with Woodbury who told of rediscovering it in 1958 during the course of a detailed examination of the flora of the chain of mogotes which parallel the north coast of Puerto Rico. The mogotes, in this instance, are low, steeply eroded, calcareous hills. The populations of D. hiliotane are to be found at their summits except for the population at kilometer 21.2 which is found on the first ledge below the summit. In the fall of 1963 Woodbury supplied a few seedlings for transport to Cambridge. Through an unfortunate accident these plants were lost. A second attempt, in 1964, was successful. Thymelaeaceous plants are, in general, difficult to propagate or transplant and it was necessary to keep the seedlings under mist propagation and polyethylene where they remain nearly two years later. Nevling was able to see living plants in the field in April of 1965, thanks to Woodbury’s guidance. Although this species has many of the typical thymelaeaceous features such as low stature, flexible branches, and a well-developed unlignified fiber structure within the bark, one feature was lacking. Missing was the nearly typical odor of stems or crushed leaves. This fragrance, from coumarin derivative compounds, or lack of it, may prove to be indicative of the presence or absence of certain compounds and may eventually serve as an accessory character of taxonomic value The scientific description published by Nevling (Ann. Missouri Bot. Gard. 46: 287-289. 1959) has been recast to include additional informa- tion derived from more recent collections and from field observations. Daphnopsis helleriana Urb. Symb. Antill. 2: 453. Oct. 1901. [Type: Heller 4689 (8) !] Dioecious shrubs or small trees to 4.5 m. tall, dichotomously branching, the bark thick, fibrous, the wood white, soft, the young branches with few leaves, flexible, golden-tomentose and glabrescent; tap root long. Leaves simple, alternate, the blade elliptic, oblong, or obovate, 3-16 cm. long, 2-7 cm. broad, mucronulate to obtuse at the apex, cuneate at the base, thin- coriaceous, glabrous, golden-tomentose to villous but soon glabrescent beneath; both surfaces light green but on drying the upper sometimes becoming reddish-brown, the lower glaucescent, developing reddish-brown pigment bordering the veinlets, the costa plane above, elevated beneath, primary lateral veins prominulous above, prominent beneath, somewhat *I am especially grateful to Mr. Robert B. Martinson for providing facilities and generous hospitality on numerous plea ae occasions in Puerto Rico. Acknowledge- ment also is made to the National Seite Foundation for a grant | ichard A. Howard (GB-4975) which stiri supported a portion of this study. L.I.N. 264 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 arcuate-ascending and extremely hia petiole 2-6 mm. long, scarcely canaliculate. Inflorescences borne terminally on the very young stems, umbelliform, 3—5-flowered, golden- arent to villous throughout, the primary peduncle 5-10 mm. long, nodding, the rachis ca. 1 mm. long, secondary peduncles 1-3 mm. long. Staminate flowers: pedicel to 1 mm. long; calyx tube tubular, 5-11 mm. long, 2-3 mm. in diameter at the orifice, golden-tomentose without, glabrous and somewhat red-pigmented within: calyx lobes unequal, greenish white and indefinitely papillate within, the outer 2—3 mm. long, ca. 2.5 mm. broad, the inner 1.5 mm. long and broad, spreading; petals 4, squamelliform, as long as broad, inserted at the orifice, minute; antisepalous stamens inserted below the orifice, at most subexserted, the alternisepalous inserted about two anthers’ lengths below the orifice, included, the anthers oblong, 1.25—1.5 mm. long, 0.25—0.5 mm. broad, sessile, the pollen yellow; disc of a few irregular lobes nearly as tall as the pistillode, free, glabrous; pistillode fusiform, ca. 11 mm. long, glabrous. Pistillate flowers: pedicel to 1 mm. long; calyx tube sub- campanulate to urceolate, ca. 5 mm. long, ca. 2-3 mm. in diameter at the orifice, golden-tomentose without, glabrous within; calyx lobes unequal, spreading, indefinitely papillate and pale green within, ca. 2-3 mm. long and broad; petals 4, papilliform, inserted at the orifice, minute; staminodia 8 or sometimes absent, in two whorls below the orifice, minute, with brownish a cap oanggna disc minute or absent (?); pistil conical, tapering to the stigm . 5 mm. long, 1.5 mm. in diameter, glabrous, the stigma capitate, sae ae exserted. Fruit a white berry {pseudodiupe), elliptic- ovoid, 10-15 mm. long, 8-10 mm. in diameter. Flowering irregularly from February through April. Found near the summit and at the summit of calcareous hills at an altitude of 75-300 feet. Puerto Rico: San Juan, near — Heller 4689 (A, E, GH, L, MICH, NY, Sant w. of Bayamon, Route 2, kilometer 23.7, Woodbury 6711 (A, ARP 6) - of Bayamon, Route 2, kilometer 21.2, ‘Woodbury 6716 (A, ARP), Howard, Nevling & Woodbury 15686 (a — chromosome voucher), seedlings under cultivation, Nevling 118 (AaH); s. of Dorado, Higuillar Bro. gue os 10680 (GH, NY); Guajataca Gorge, Quebradillas, Woodbury 3196 (A, A The holotype of this species was destroyed in Berlin during World War II. Because of its relative completeness I am choosing the specimen of Heller 4689 deposited in the Gray Herbarium as the lectotype. Knowledge of the cytology and genetics of the Thymelaeaceae is limited severely except in isolated cases. Chromosome numbers have been reported for only a few genera and they, unfortunately, are all members of the same subfamily. Apomixis has been demonstrated in Wikstroemia and a polyploid series is known in Pimelea. The latter situation is of more than passing interest as all of the species of Pimelea, which have been examined in detail, have been found to be gynodioecious. All of the species of Daphnopsis, which are known in sufficient detail, are dioecious. ° Agricultural Experiment Station, University of Puerto Rico, Rio Piedras. 1966 | NEVLING & WOODBURY, DAPHNOPSIS 265 To date, chromosome numbers of only two species have been reported: D. americana (Mill.) Johnst. (n = 9) and D. philippiana Krug & Urb. (n = 9). It was, therefore, somewhat of a surprise to find D. helleriana with a haploid number of 18. This count is based on observations of meiotic tissue, metaphase I through anaphase II, in developing pollen (Voucher: Howard, Nevling & Woodbury 15785). No configuration ex- cept bivalent formation was observed and all divisions appeared normal and orderly. Mature pollen is abnormal in that there are many collapsed grains and some size variation. Pollen stainability, with cotton blue in lactophenol, is not high but we are not reporting a percentage because of the small sample size. Fruit is set relatively frequently and seedlings are seen in nature. There appear to be no special modifications for vegetative reproduction and it is believed that the species reproduces solely by fruit. There is no evidence to support or reject apomixis in this case although the small size of the total population might be regarded as negative evi- dence. At present we do not wish to postulate the nature of this tetraploid species. Whether an allotetraploid or autotetraploid, it is of interest because it has arisen in a dioecious group. ARNOLD ARBORETUM AND GRAY HERBARIUM ARVARD UNIVERSITY, CAMBRIDGE AND AGRICULTURAL EXPERIMENT STATION NIVERSITY OF PuerTO Rico, R10 PIEDRAS 266 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. 47 SCIAPHILA DOLICHOSTYLA (TRIURIDACEAE) PETER S. GREEN AND OTTo T. SOLBRIG ONE OF THE BEST WAYS to find small and insignificant plants is to look around as one sits down to eat lunch in the field. In just this way one of us (P.S.G.) noticed a small, red saprophyte, which was later identified as Sciaphila dolichostyla Schitr., growing in rich humus on the bank of a stream in New Caledonia.* Unfortunately, full ecological notes were not made at the time, but bud material was fixed in acetic-alcohol and flown, by mail to Cambridge, Massachusetts, where one of us (O.T.S.) carried out the cytological ex- amination reported upon below. The Triuridaceae is a small family of saprophytic, monocotyledonous herbs, containing 7 genera and 70 or more species. It is confined to the tropics of both the New and Old Worlds, with the greatest concentration re with valvate perianth lobes usually in multiples of three. In some species the flowers are unisexual, in others hermaphrodite. As saprophytes, the plants contain no chlorophyll and are generally reported as being reddish, purplish or yellow in color. Endogenous mycorrhizae have been described by Larsen (1963), together with references to earlier reports. The Sciaphila reported upon in this note (P. S. Green 1329), was grow- ing on a fairly steep, earthy bank with patches rich in humus, and below a relatively light canopy of low to medium-sized trees fringing the Nekoroya, a rocky tributary stream of the Riviére Téné, near Bourail on the southern side of the central mountain chain and somewhat less than 150 km. north- east of Nouméa. Except for the woodland fringing the streams, the general area was one of ‘“‘maquis,” at an approximate altitude of 150 meters, where lateritic soils overlay the serpentines and peridotites so characteristic of much of New Caledonia. The shoots of Sciaphila, approximately 5 to 12 centimeters high above the surface of the humus and leaf litter (Fic. 1, A & B), were red through- out with no sign of chlorophyll. The underground rhizome was very thin and brittle and appeared to branch and ramify throughout the humus, giving off numerous small, twisted roots which were presumed, from their appearance, to be mycotrophic. Examination showed the plant to be mo- noecious with actinomorphic, unisexual flowers, each with six strongly re- flexed petals (Fic. 1, C & D), each tipped with a beard of abundant, mar- is field work was supported by the National Science Foundation (Grant Pe 1545) for which peat acknowledgment is expressed. 1966 | GREEN & SOLBRIG, SCIAPHILA DOLICHOSTYLA 267 Figure 1. pseu cago er A & B, habit, natural size; C, male flower, Xx 5; D, female flow 5; E & F, anterior and posterior view of stamen, x 20; G &H, single ovaries, x 25. ginally borne, pale purplish hairs. The pedicels were arcuate and some- what secund, as originally described by Schlechter (1907, p. 19). The male flowers clearly possessed six stamens (Fic. 1, C) in contrast to the condition in the only other species of Sciaphila dexcribed from New Caledonia, S$. neo-caledonica Schltr. which has three. The stamens each had three nastier sacs and were borne on distinct filaments about one-fourth millimeter long (Fic. 1, E & F). In the female flowers (Fic. 1, D) the numerous ovaries, each capable of producing a single seed, were carried on a rounded receptacle and each 268 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 possessed a simple basal style with stigmatic tuft (Fic. 1, G & H). The length of the style, to which allusion is made in the specific epithet, varied somewhat, perhaps depending upon the maturity of the flower, from about 0.3 to 0.5 millimeter in length. The gametic complement was found to consist of 22 chromosomes, three of which were large, six medium and two small and which varied in length from approximately 17 to approximately 3 micra. All the large chromo- somes were subterminal, while the smaller, and most of the intermediate sized ones, were median (Fics. 2 & 3). “s a “AVWitiee. Ficures 2 and 3. Sciaphila dolichostyla. Lert, F1c. 2. Chromosomes at prometaphase, General habit of 1-year-old seedling, x 1 B. Base of same axis, leaves removed, 2; early stages in development of (as suckers. er, more ‘vigorous. seedling, X 1/4; rhizomatous development of axillary suckers. In B and C, 1-7 indicate successive axillary suckers which develop in order of their age, 1 youngest, 7 oldest. 304 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 For microscopic analysis, continuous serial sections 30 » thick of two seedling axes were cut on the sliding microtome and stained in safranin and Delafield’s haematoxylin, as described previously. A complete series of sections was mounted, not every tenth as in previous studies, because the congested vascular system could only be understood by photographing each section. Each section was mounted on a separate slide, as is no our standard practice, in order that the optical shuttle method of align- ing successive sections in the camera frame could be used (Zimmermann & Tomlinson, 1966). Photography of serial microtome sections was again supplemented by direct photography of successive surface cuts of seedlings clamped in the sliding microtome. GENERAL ANATOMY OF THE JUVENILE AXIS Over-all increase in diameter of the axis in the juvenile phase of growth of Rhapis is illustrated in Fics. 2-4, representing sections at successively higher levels from one of the microtome series. Increase in certain dimen- sions in the same series is given in TABLE Taste I. Increase of Certain Dimensions in the Seedling Axis of Rhapis excelsa | Central Cylinder Slide Level | Average : Approx. no. | Ay, diam. vascular number (mm.) cortex Av.diam.| vascular bundles diam. (mm.)| (mm.) bundles Periphery Center ue m 1 (Fic. 2A) 6) 0.8 0.7 8 ores rs 50 (Fic. 2B) 1.5 iZ 1.4 38 150 150 101 (Fic. 2C) 3.0 2:0 2.4 82 150 180 200 (Fic. 3) 6.0 1.6 535 160 300 240 349 (Fic. 4) 10.5 1.3 8.0 400 450 330 Axis outline irregular owing to distortion imposed by numerous root and branch insertions. Surface layers either resembling those of rhizome, or more usually with irregular divisions, but without a regular phellogen, cells ligno- suberized, abundantly tanniniferous and forming an irregular periderm. Cortex very wide in lowest levels and exceeding central cylinder, but becoming propor- te I). Co branch insertions. Cortical fibrous strands well developed at all except very lowest levels, fibrous strands up to 120 u wide, largest including narrow central traces recognized by oblique, often more or less horizontal passage from central cylinder through cortex. Diameter of bundles increasing upwards (TaBLE I), 1966 | TOMLINSON & ZIMMERMANN, RHAPIS EXCELSA 305 largely due to increase in amount of fibrous sheathing tissue, especially in peripheral bundles (cf. Fics. 2~4). etaxylem elements narrow, irregular, ieee parenchyma except at periphery of cortex. Tannin cells mostly restricted to outer cortex. Stegmata (silica cells) common next to fibers of bundles in both cortex and central cylinder. COURSE OF VASCULAR BUNDLES Although complex, the distribution of vascular bundles can be inter- branches,” indicative of a close developmental relation, does not obscure the vascular system of the axis itself. Vascular systems of central cylinder and cortex remain distinct, although less so than in rhizome and aérial stem 1, CENTRAL CYLINDER. Cinematographic analysis reveals frequent leaf traces passing more or less horizontally into leaves, and readily distin- guished from the numerous vertical bundles cut more or less transversely in transverse sections. Vertical bundles are transformed into leaf traces by change from a vertical to a horizontal course. The constant relation between leaf traces and continuing vertical bundles which was found in the aérial stem does not exist in the seedling axis. This is discussed later. In addition to this main vascular system there is an interlinking network of usually narrow bundles which correspond to the bridges of the adult stem. They are sufficiently extensive to be referred to as an anastomosing bridge system. Vertical bundles. Traced upwards, the distance of vertical bundles from the periphery of the central cylinder increases so that the bundles come to occupy the uncrowded center of the stem, prior to their con- version into leaf traces. This apparent displacement is largely the result of the widening of the central cylinder in the obconical axis. Vertical bundles may even diverge from the stem center, but at a shallower angle than the limit of the central cylinder. The helical pathway of the central bundles which is characteristic of the aérial stem is established early and again it corresponds in direction (but not in magnitude) to the phyl- lotactic spiral. Leaf traces. Vertical bundles become leaf traces by an abrupt change to an approximately horizontal direction. Unlike the bundles of the adult axis, however, there is no fixed relation between leaf traces and continu- ing vertical bundles. In the lowest internodes where there are few bundles, the vertical part of each bundle is short and all bundles diverge from a ada of root and branch traces will be described in detail in a later paper in this serie 306 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 FiGuRE 2. Transverse sections of a juvenile axis of Rhapis excelsa at suc- cessively higher levels. A. At approximate cae .. cotyledonary node. B. 1.5 above A. C. 3 mm. above A. Scale equals 1966 } TOMLINSON & ZIMMERMANN, RHAPIS EXCELSA 307 about the same depth in the stem center. Distally, as the total number of bundles increases, the length of the vertical part of the bundles’ course rapidly increases and differences between major, intermediate, and minor leaf traces become evident as the trace complement to each leaf becomes arger. SS. i MM. , ae cB = h—————_—— Ficure 3. Rhapis excelsa, same axis as illustrated in: Fic: 2, 6 mm. above 2A, same magnification. Two leaf traces (LT) indicated, the one in the center of the figure with two recently derived bridges. 308 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. 47 Anastomosing bridge system. As in the adult axis departure of a leaf trace is associated with the development of bridges which are very pronounced in the juvenile phase. The first bridge originates as the through the crowded periphery. Bridges may even be given off by the leaf trace as it passes through the cortex, such bridges becoming in- corporated into the cortical system. Bridges in the juvenile axis resemble those of the adult stem in diverging upwards from a leaf trace and often linking immediately with a nearby vertical bundle. Otherwise they may differ in three important respects: (a) they are often very long, extending irregularly without interruption through several internodes; (b) they commonly split; (c) they, or their branches, ultimately join vertical bundles or reach the periphery of the central cylinder where they become vertical bundles. Bridge bundles are narrow, with inconspicuous vascular elements although the largest resemble peripheral vertical bundles in the base of the stem. Small bridges may lack vascular tissue and appear as narrow fibrous strands, From the manner in which they diverge, upwards from leaf traces, these bridge bundles seem developmentally homologous with those of the aérial stem but their ramifications and extensive develop- ment justify referring to them as an anastomosing system. Root and branch traces connect extensively with these bridges or their branches. 2. Cortex. The cortical system of fibrous bundles remains largely in- dependent of the central system. Cortical bundles traced downwards from the leaves anastomose freely to form a reticulum but many bundles end blindly below. The reticulum is much distorted around branch and root insertions. Interconnection between vascular systems of cortex and central cylinder is somewhat more pronounced than in the rhizome and aérial stem, either because of direct interchange of vascular bundles between the innermost region of the cortex and the periphery of the central cylinder or because leaf traces, as they pass across the cortex, link with the cortical reticulum via bridges. MULTIPLICATION OF VASCULAR BUNDLES We come now to the most interesting anatomical aspect of the juvenile axis, namely the question of how the increase in number of vascular bundles is accomplished during establishment growth. In Taste I it can be seen that during a height increase of 10.5 mm. the seedling axis had increased from 0.7 to 8 mm. in diameter sips cylinder) and from 8 to 400 vascular bundles. Thorough study of numerous motion picture sequences have revealed the following processes. Ultimate branches of the anastomosing bridge system reach the periph- ery of the central cylinder, turn erect there and become vertical bundles. Each leaf trace thus contributes (via the anastomosing system) on the average more than one vertical bundle. In the higher levels of the seed- 1966 | TOMLINSON & ZIMMERMANN, RHAPIS EXCELSA 309 Ficure 4. Rhapis excelsa, se axis as illustrated in Fic. 2, 10.5 above : 2A, same magnification. This closely ange adult stem, pomtenaks vertical bundles (VB) ‘with well Paebelooed fibrous sheaths. Bundles of the anastomosing bridge system (AS), root traces (R T). 310 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. 47 ling axis the anatomy resembles more and more the mature vegetative stem (Fic. 4), leaf traces giving off several bridges and vertical bundles. In principle then, the increase in number of vascular bundles in the post- seedling stage is accomplished by the same mechanism by which the num- ber of vascular bundles in the mature stem can be varied, i.e., by the variable number of vertical bundles which is given off by each leaf trace. It must be emphasized that production of additional vertical bundles is a property of the periphery of the central cylinder. No vertical bundles are developed in the stem center. We have also seen occasional vertical bundles split in the peripheral region. It is quite clear that the system of vertical bundles becomes “directed” and ‘stabilized’ in the peripheral region. This must be based upon the way in which the vascular system is produced in the developing apex, a problem which we shall have to take up in a later paper. DISCUSSION The most important finding of the present study is the fact that the vascular anatomy of the post-seedling stage is not fundamentally different from the anatomy of the mature vegetative axis. Leaf traces give off numer- ous bridges (‘‘the anastomosing bridge system”) which turn into vertical bundles at the periphery of the central cylinder. The structure of the post-seedling stage is quite confusing at first and it is very doubtful if conventional anatomical methods could have demonstrated it correctly. Only innumerable projections with the data analyzer during which a great number of bundles were followed in both directions finally led to an understanding of the anatomical principles. The anatomy of the seedling axis of Rhapis may shed light on a botani- cal misconception to which we have drawn attention. Elsewhere (Tomlin- son & Zimmermann, 1966a) we have shown that modern ideas about the distribution of vascular bundles in palm stems are erroneous. Diagrams by early investigators implied either a blind ending of vascular bundles at the periphery of the central cylinder when traced downwards, or possibly basal fusion between peripheral bundles. These early workers made it clear that they were undecided on this point. Later writers, often without first-hand familiarity with palm stems, produced diagrams which have become “standard” in modern textbooks in which peripheral fusion of bundles is regarded as a normal feature. But we have shown for Rhapis in the first paper of this series, and have much unpublished information about many larger palms, that in the adult aérial axis the continuing vertical bundle is derived from the leaf trace at its point of departure into the leaf. Splitting of vertical bundles in an upward direction (or their fusion if traced downwards) is not a normal property of the adult palm stem al- though it may occur as a rare abnormality in the center of the rhizome axis. We have now shown that the inverted cone of the juvenile axis may contain splitting vertical bundles in the periphery of the central cylinder. 1966 | TOMLINSON & ZIMMERMANN, RHAPIS EXCELSA 311 This may possibly be a rational explanation for previous botanical mis- conception of the vascular anatomy of the palm stem. Early workers who studied the juvenile axes might have seen these rare cases in the seedling without appreciating that this is not a property of the adult axis. Among these early workers Falkenberg (1876) and Nageli (1858) both seem to have studied young plants of Chamaedorea. We have reason to believe that Nageli in particular, who described downward fusion, may have influenced subsequent writers who constructed increasingly hypo- thetical diagrams of palm stems (Tomlinson & Zimmermann, 1966a). SUMMARY In the post-seedling establishment of the adult axis of Rhapis excelsa the obconical development at the basal internodes may result in multipli- cation (in a structural sense) of eight vascular bundles at the cotyledonary node to 400 vascular bundles 10.5 mm. higher. The leaf trace system corresponds closely to that of the aérial stem. Leaf traces give off num- erous bridges (the “‘anastomosing bridge system’’) which are often quite long, branch, and join neighboring vertical bundles. Branches of the anastomosing system that reach the periphery of the central cylinder become vertical bundles. Since the number of vertical bundles thus pro- duced by each leaf trace is greater than one, the number of vascular bundles in the stem increases. At higher levels the complicated anastomo- sing system is more and more reduced and the structure approaches that of the mature axis. ACKNOWLEDGEMENTS We are indebted to Miss Lesley Jackson for routine microtome section- ing on which this study is based, and to Miss Priscilla Fawcett for FicureE 1. LITERATURE CITED FALKENBERG, P. 1876. Vergleichende Untersuchungen iiber den Bau der hen core ame der Monocotyledonen. 202 pp. 3 pls. Ferdinand Enke, GATIN, rg oe 1906. Recherches anatomiques et “cai sur la germination des palmiers. Ann. Sci. Nat. sér. 9, 3: 191-315 2. Les Palmiers. cata naturelles et horticoles des différents cen res. 338 pp. Doin et Fils, GEHRKE, O. 1887. Beitrage zur F anhis der Anatomie von Palmenkeimlingen. Inaug. Diss. 29 pp. Bading, Berlin. Giniers, C. 1952. Contribution a l’étude anatomique des plantules de palmiers IV. La plantule de Washingtonia gracilis Parish. Bull. Mus. Hist. Nat. Paris, sér. 2, 24: 392-399. ye Heim, J. 1937. Das Erstarkungswachstum der Palmen und einiger anderer Monokotylen, zugleich ein Beitrag zur Frage des Erstarkungswachstums der Monokotylen iiberhaupt. Planta 26: 319-364. 342 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 Karsten, H. 1856. Ueber die Bewurzelung der Palmen. Linnaea 28: 601-608. Micueets, H. 1889. Recherches sur les jeunes palmiers. Mém. Acad. Roy. Sci. Belg. 51: 126 pp. 4 pls Mout, H. von. 1824. De palmarum structura. Jw: K. F. P. von Martius. Historia Naturalis Palmarum I: pp. I-LII. 16 pls. NAcELt, C. W. 1858. Ueber das Wachsthum des Stammes und der Wurzel bei den Gefasspflanzen. Beitrage zur wissenschaftlichen Botanik Heft 1, pp. 1-156. pls. 1-19. a A. 1896. Morphologischer und anatomischer Vergleich der Koty- nen und ersten ope eae der Keimpflanzen der Monokotyledonen. Bibliot, Bot. 35: 88 p STRASBURGER, E. 1906. Uber die Verdickungsweise der Stdmme von Palmen nd chraubenbaumen. Jahrb. wiss. Bot. 43: 580-628. SUESSENGUTH, K. 1921. Beitrage zur hea — systematischen Anschlusses der Monokotylen. Beih. Bot. Centr. 38: TomLinson, P. B. 1960. Seedling ae in oe and their morphological significance. Jour. Arnold Arb. 41: 414-428. 1964. Stem structure in arborescent monocotyledons. Ju: M. H. ZIMMERMANN, ed. The formation of wood in forest trees. pp. 65-86. Academic Press, New k. M. IMMERMANN. 1966a. Vascular bundles in palm stems — their bibliographic evolution. Proc. Am. Philos. Soc. 110: 174-181. ———————.. 1966b. Anatomy of the palm Rhapis excelsa. Il. Rhizome. Jour. Arnold Arb. 47: 248-261. TROLL, W. 1936. Vergleichende Morphologie der hoéheren Pflanzen. Bd. 1. Vegetationsorgane. Gebr. Borntraeger Berlin. pinnae NN, M. H., & P. B. Tomiinson. 1965. Anatomy of the palm Rhapis xcelsa. i Mature vegetative axis. Jour. Arnold Arb. 46: 166-178. 6. Analysis of complex vascular systems in plants: Optical shuttle Brot Science 152: 72-73. HARVARD UNIVERSITY CasoT FOUNDATION, PETERSHAM MASSACHUSETTS 0136 AND FAIRCHILD TROPICAL GARDEN Miamt, FLoripa 33156 ee eee oo armrests 1966 | CRITCHFIELD, PINUS AND ABIES 313 PHENOLOGICAL NOTES ON LATIN AMERICAN PINUS AND ABIES WitrttAmM B. CRITCHFIELD VERY LITTLE INFORMATION has been reported on the time of pollen shedding of the many pines and firs native to Mexico and northern Central America. This kind of information is necessary for any breeding operation, and is often of considerable taxonomic and evolutionary inter- est as well. An example is the difference in pollination time of a month to six weeks between Pinus radiata D. Don and P. attenuata Lemmon, two closely related closed-cone pines native to California. This difference is a reliable means of distinguishing the two species, and is also the principal mechanism that limits natural hybridization between them. The reproductive phenology of many conifers, especially the pines, is well documented (summarized by Bingham & Squillace 1957), but I have encountered only a few references to the time of pollen shedding and pollination in Latin American pines and firs growing in their native habitats. Shaw (1909) reported that Pinus pringlei Shaw flowers in November or early December, before the associated P. oocarpa Schiede. Martinez (1948) noted that P. durangensis Martinez, native to northwestern Mexico, flowers in April. Little (1962) reported that occasional trees of a few Mexican pines shed pollen in September and October, and noted pollen and receptive female strobili of P. ayacahuite Ehrenb. in October. Mirov (1962) observed in mid-February that P. oocarpa had recently completed pollination in the mountains of Nicaragua. Meiosis occurs in the pollen cones of P. pinceana Gord. in the middle of March in northeastern Mexico, according to Diaz Luna (1962). Since meiosis in pines generally precedes pollen shedding by about 3 weeks, pollination probably takes place in early April in this pinyon pine. The only mention of the time of pollen shedding in the Latin American firs is Martinez’ report (1963) that the pollen cones of Abies religiosa (H.B.K.) Schlecht. & Cham. mature be- tween March and May in central Mexico (Hidalgo and the Distrito Federal). The phenological observations summarized in Taste 1 were made by three people: John W. Andresen, in southern Mexico and Guatemala during December, 1962; Elbert L. Little, Jr., in northern and central Mexico during March, 1963; and the author on an assignment in Hon- duras with the Food and Agriculture Organization of the United Nations during January, 1965, and in Guatemala and southern Mexico the fol- lowing month. Many of these observations are the dates on which pollen collections were made. They often pertain to single trees, and are not always representative of the prevalent stage in the surrounding stand. 314 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 Because of the dearth of information on the reproductive phenology of these species, I have also included observations of species that either had not yet begun to shed pollen or had completed shedding when they were observed. The places listed in TaBLe 1 are accurately located by latitude and longitude in most instances, but the elevations are much less reliable, especially in Mexico. The devations of Mexican localities are in most cases either altimeter readings or estimates from the topographic maps published by the Mexican government, at a scale of 1:500,000. Speci- mens corresponding to most of the tabulated observations are deposited in the conifer herbarium of the Institute of Forest Genetics at Placerville, California. Some of them are duplicated in the U.S. National Herbarium and in the herbaria of Michigan State University and the Instituto Na- cional de Investigaciones Forestales in Mexico City. The pines shed their pollen earlier in the season in Latin America than in the southern and western United States. The earliest of the southern pines, Pinus clausa (Chapm.) Vasey, begins shedding pollen in late December (Dorman & Barber 1956). P. radiata, the earliest of the western pines, sometimes begins shedding as early as late January (Duf- field 1953). P. caribaea, probably the earliest of the Central American pines, was shedding pollen in late October and early November, 1964, in eastern Nicaragua, near the southern end of its range.!. Farther north, in British Honduras, neither P. caribaea nor P. oocarpa had begun to shed on November 2-7, 1960.2 In Honduras P. caribaea had completed pollen shedding by late January, 1965, except for a few possibly aberrant trees in one locality (TaBLE 1: Agalteca). The condition of the pollen cones in Honduran stands suggested that this species had shed its pollen earlier than the associated P. oocarpa. Central American P. caribaea is much earlier in time of pollen shedding than the closely related P. elliottit Engelm. This pine of the southeastern United States starts shedding at the end of January and continues for about a month (Dorman & Barber, 1956). Other Latin American pines also begin to shed pollen much earlier than their U. S. relatives. Pollen shedding of P. lawsonii Roezl was at its peak on December 13, 1962, near Lake Patzcuaro in Michoacan, Mexico (elev. 2,000 m.).*? E. L. Little, Jr., observed a few trees of P. oocarpa shedding pollen on October 17, 1960, in western Chiapas, Mexico (N. lat. 16°45’, W. long. 99°30’, elev. 1,100 m.).4 In Guatemala and Oaxaca, P. oocarpa and P. teocote Schiede & Deppe had begun to shed in mid-December, 1962, when J. W. Andresen collected their pollen (TABLE 1). In Honduras P. oocarpa had completed shedding nearly everywhere that I saw it during the last half of January, 1965 (TaBLe 1). * Personal communication from L. C. Saylor. onal communication from L. C. Savior. Branches bearing mature pollen cones cand ovulate strobili are illustrated in North Carolina State College 1963 (Fig. ‘ Posical communication. 1966] CRITCHFIELD, PINUS AND ABIES 315 Another widespread Honduran pine, Pinus tenuifolia Benth. (often identified as the closely related but more northern P. pseudostrobus LindI.), exhibited, in 1965, an unexpected inversion of the usual relation- ship between greater elevation and later pollen shedding (see Duffield, 1953, for examples). On El Picacho, a peak of about 1,300 meters (4,360 ft. ) elevation in the suburbs of Tegucigalpa, the pollen cones of this species had not yet begun to emerge from the bud by February 3, but two days earlier pollen shedding was general in a stand only a few miles distant but 440 meters (1,440 ft.) higher (TaBLe 1: Jutiapa). I observed the same unexplainable phenomenon two weeks later in the Sierra Madre del Sur, south of the city of Oaxaca, Mexico. In the foot- hills south of this range at an elevation of 760 meters, P. tenuifolia had not yet begun to shed, but at the crest of the mountains, 1,200 meters (3,900 ft.) or more higher, shedding was nearly complete (TABLE 1: San Gabriel Mixtepec). Although the date of pollination in conifers is very much affected by the TABLE | suggest that the Latin American conifers may not all exhibit this degree of constancy in relative pollination time. In the following list of pollen-shedding sequences taken from TaBLeE 1, simultaneous shedding is indicated by = , successive shedding by /. PLACE CHRONOLOGICAL ORDER OF POLLEN SHEDDING Uruapan, Michoacan P. douglasiana/P. pseudostrobus x P. oocarpa/P. pseudostrobus Ciudad Hidalgo, ra ora P. montezumae = P. teocote Paso de Cortés, Méxic A, atte montezumae/P. hartwe Zacatepec, Puebla - ie = teocote Perote, Puebla A. religiosa = Ixtlan, Oaxaca P. lawsonii/P. pseudostrobus var. oaxacana/P. patula var. longipedunculata San Gabriel Mixtepec, Oaxaca PP. oocarpa/P. tenuifolia Honduras (general) ie es One notable inconsistency appears in these sequences. Abies religiosa was just beginning to shed pollen in mid-February, 1965, at 3,400 meters (about 11,000 ft.) in the State of México, long before its associate, Pinus hartwegii Lindl. I estimated that this pine would not begin to flower for another month or six weeks. E. L. Little, Jr., collected pollen of the same species in this vicinity near the end of March, 1963 (TABLE 1: Paso de Cortés). But pollen of A. religiosa and P. hartwegii was collected on the same date near Perote, Puebla, in 1963 (TABLE 1: Cofre de Perote). Two of the observations listed in TABLE 1 do not agree with earlier reports. I collected pollen of Abies religiosa in the State of México (TABLE 1: Paso de Cortés) a month and a half before the April-May date reported TABLE 1, — Observations on the time of pollen shedding in Latin American pines and firs 1NS = not yet shedding; St = noted. N. W. ELEva- STATE NEARB Lar. LONG. TION POLLEN SPECIES DATE AND/OR COUNTRY PLACE NAME ace vd ed METERS STAGE * ABIES religiosa (H.B.K.) March 24, 1963 México, Mex. Bosencheve 19 26 100 11 2950 [Sh] Schlecht. & Cham. i mega 28,1963 Puebla, Mex. Cofre de Perote 19 30 97 11 3250 [Sh | . Feb. 16, 1965 México, Mex. Paso de Cortés 19 06 98 41 400 St vejari Martinez se 30, 1963 Nuevo Leon, Mex. Cerro Potosi 24 51 100 13 2950 Sh INUS caribaea Morelet Jan. 20, 1965 Honduras Agalteca 14 28 87 12 770 MS = Jan. 25, 1965 a San Esteban 15 10 85 55 800 SO a an, 27, 1965 “ Miravalles 14 40 87 14 750 SO montezumae Lamb. March 25, 1963 Michoacan, Mex. Ciudad mea 19 41 100 40 2700 [Sh | . ig éxico, Mex. Bosenchev 19 24 100 04 3150 [Sh] “i March 26, 1963 4 Mexico City 19 19 98 50 3100 [Sh] a March 29, 1963 Hidalgo, Mex. acala 20 51 99 14 2250 [Sh | : Feb. 5, 1965 Guatemala Chimaltenango 14 38 90 54 2100 NS* F Feb. 16, 1965 México, Mex. Paso de Cortés 19 06 98 41 3400 NS* hartwegii Lindl. March 26, 1963 “ 19 05 98 39 3250 [Sh] : March 28,1963 Puebla, Mex. Cofre de a 19 30 97 11 3250 [Sh] ss Feb. 16, 1965 México, Mex. Paso de Corté 19 06 98 41 3400 NS* i Feb. 11, 1965 Oaxaca, Mex. Sola de Vega 16 31 96 58 1400 NS* preudostrobu Lindl. March 24,1963 Michoacan, Mex. Mil Cumbres 19 39 100 46 2400 MS March 25, 1963 * Zitacuaro 19 27 100 19 2650 [Sh] ’ March 28,1963 Puebla, Mex. Zacatepec 19 19 OT 33 2650 [Sh] = March 29,1963 Hidalgo, Mex. cala 20 48 99 16 1860 [Sh] e March 30, 1963 Nuevo Ledn, Mex. Iturbide 24 45 99 57 1600 MS Feb. 5, 1965 Guatemala olala 14 49 91 12 2300 Sh - Feb. 18, 1965 Michoacan, Mex. Uruapan 19 26 102 04 1700 NS’? ‘et aa a 7 pseudostrobus var. ‘ : 800 Sh oaxacana Martinez An 10, 1965 Oaxaca, Mex. Ixtlan de Juarez 17 21 96 27 0 er tenuifolia Benth. n. 15 to Honduras Los Limones 14 10 87 08 1440 NSto 2 th a8 : Mateo 14 02 87 21 1570 NS 2 Feb. 1, 1965 " Ms 14.12 87 07 £730 = 2 Feb. 3, 1965 - Teguci 14 08 87 p so ee ” : San Gabriel Mixtepec 16 03 97 7 ‘ Feb. 12, 1965 Oaxaca, Mex P ‘ oF 0K 1950 MS douglasiana Martinez Feb. 18, 1965 Michoacan, Mex. Uruapan 19 26 102 04 1700 SO tan. a March 25, 1963 ‘s Ciudad Hidalgo 19 41 100 40 ena st z March 28,1963 Puebla, Mex. Zacate 9 19 97 33 650 St és March 29, 1963 _ Cofre de Perote 19 32 . 12 cen ist “ Mex. Oaxaca de Juarez 17 10 96 40 : Dec. 13, 1962 Oaxaca, ‘ J rll as sank St Feb. 10, 1965 Oaxaca, Mex. Ixtlan de Juarez 17-21 96 27 2800 SO atula Schlecht. . & Cham. March 29, 1963 Veracruz, Mex. Jalapa 19 35 96 58 2100 [Sh] Be Feb. 9, 1965 Hidalgo, Mex. Lake Tejocotal 20 09 98 10 2200 NS patula longed i tees pi SE feiihe de yukees 17 21 96 27 2800 NS oocarpa Schiede Dec. 18, 1962 Guatemala Quezaltenango 15 08 91 36 2360 St ‘i Jan. 26, 1965 Honduras ateo 14 02 87 21 1570 MS 4 Jan. 16, 1965 d Tegucigalpa 14 08 87 11 1310 sO ” Feb. 5, 1965 Guatemala Guatemala Cit 14 44 90 22 1000 SO o Feb. 12, 1965 Oaxaca, Mex San Gabriel Mixtepec 16 03 97 06 760 SO Rg Feb. 18, 1965 Michoacan, Mex. 26 102 04 1700 MS pringlei Shaw Feb, 11, 1965 Oaxaca, Mex. Sola de Vega 16 29 97 O1 2050 Sh starting to shed; Sh = shedding; MS = mostly shed; SO = 2 Pollen cones emerging from bud; estimated 2 to 3 weeks until shedding begins. ’ Pollen cones still covered by bud scales; at least a month until shedding begins. shed out; [Sh] = pollen collected but stage not [9961 Lp “T0A] WOLANOdTUV GIONUV AHL JO TWNMNOL 9T¢ SaIdV GNV SONId ‘A TAISHOLIYO 318 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 by Martinez (1963) in the nearby Distrito Federal. And pollen shedding was general in a stand of Pinus pringlei in the Sierra Madre del Sur (Tasie 1: Sola de Vega) in mid-February, two to three months after the flowering time reported by Shaw (1909) for this species. These discrepancies and some of the other observations reported here suggest that conifer pollen shedding may be under less rigid enviromental control in Mexico and Central America than in more northern regions. LITERATURE CITED BrncHam, R, T., & A. E. SqurLLace. 1957. Phenology and other features of the flowering of pines, with special reference to Pinus monticola Dougl. U.S. Forest Serv., Intermountain Forest & Range Exp. Sta. Res. Pap. 53. 26 pp. dlus. Diaz Luna, C. L. 1962. Estudio cromosomico de Pinus pinceana Gordon. Mex. Inst. Nac. Invest. Forest., Bol. Tec. 4, 27 pp. illus. Dorman, K. W., & J. C. BARBER. 1956. Time of flowering and seed errs in southern pines. US. Forest Serv., SE. Forest Exp. Sta. Pap. 72. 15 pp. ill illus. DUuFFIELD, J. W. 1953. Pine pollen collection dates — annual and geographic variation. U.S. Forest Serv., Calif. Forest & Range Exp. Sta. Forest Res. Notes 85. 9 pp. illus. Littte, E. L., Jr. 1962. Variation and evolution in Mexican pines. Seminar and study tour of Latin American conifers. pp. 83-98. Secretar. Agr. y Ganad., Subsecretar. Recursos Forest. y Caza, Inst. Nac. Invest. Forest., Mexico City, Mexico. (English ed.) Martinez, M. 1948. Los pinos Mexicanos. Ed. 2. 361 pp. illus. México, D.F., México; Ediciones Botas. 1963. Las pinaceas Mexicanas. Ed. 3. 400 pp. illus. México: Univ. Nac. Autonoma de Méx. Mrirov, N. T. 1962. Phenology of tropical pines. Jour. Arnold Arb. 43: 218, 2 19. NortH CAROLINA STATE COLLEGE. 1963. Collecting pine material in Mexico for provenance trials and wood studies. N.C. “ ate Coll. Sch. Forest., Forest Tree Impr. Program, Tech. Rep. 18. 23 pp. SHAw, G. R. 1909. The pines of Mexico. ese om Publ. 1. 29 pp. dlus. Paciric SOUTHWEST ForEST AND RANGE EXPERIMENT STATION ForEST SERVICE, DA: BERKELEY, CALIFORNIA 94701 —_—7y 1966 | GREEN, THE GENUS LINOCIERA 319 THE GENUS LINOCIERA (OLEACEAE) IN NEW CALEDONIA * P. S. GREEN THE POSSIBILITY THAT THE GENUS Linociera, might occur in New Cale- donia, first became apparent when type material of Notelaea brachystachys Schltr. was examined. This suspicion was later brought to certainty in the field, on the lower slopes of Mt. Ignambi, toward the northern end of New Caledonia, when two small oleaceous trees were found with the flowers and fruit of a Linociera. That the phytogeographic distribution of this genus should include New Caledonia is not unexpected, for, with many species in the East Indies and New Guinea, others are known from Australia, Lord Howe Island, Fiji and Niue Island Furthermore, there have been reasons for doubting the distinctness of the genus Sarlina, described from New Caledonia in 1951, and examination of the type material shows that it, too, is a Linociera, as are other species previously treated as Notelaea or Osmanthus and not accounted for in the relatively recent revision of Osmanthus from New Caledonia (Green, Jour. Arnold Arb. 43: 268-283. 1963). Type material for all the Oleaceae de- scribed from New Caledonia has now been examined and I should like to express my gratitude to the directors and curators of the herbaria cited be- low, either for facilities to study their material or for its loan. All the ma- terial cited has been examined unless otherwise stated, and the respective herbaria are indicated by the abbreviations published in the Jndex Herbari- orum, supplemented by those in Kent (British Herbaria, 1957). Linociera brachystachys (Schltr.) P. S. Green, comb, nov. Notelaea brachystachys Schlechter, Bot. Jahrb. 39: 228. 1906; Guillaumin, Ann. Mus. Col. Marseille II. 9: 191. 1911; Bull. Mus. Hist. Nat. Paris 28: 199. 1922; Bull. Soc. Bot. France 89: 232, 233. 1942, et Fl. Nouv.-Caléd. 283. 1948. Notelaea — Guillaumin, Bull. Mus. Hist. Nat. Paris 28: 198. 1922; ibid. TI. 10: 520. 1938; ibid. 14: 456. peal Bull. Soc. Bot. France 89: 232, 233. HO et. FI. Nouv. -Caléd. 283 Osmanthus brachystachys (Schltr. ) hia Repert. Sp. Nov. 41: 152. 1936. Osmanthus francii (Guillaum.) Knoblauch, Joc. cit Sarlina cylindrocarpa Guillaumin, Bull. Mus. Hist. Nat. Paris II. 23: 539. 1951. Evergreen tree 4-15 m. tall; branches glabrous. Leaves glabrous; petiole 10-25(—30) mm. long, glabrous; lamina thickish, slightly coriaceous, The support of field work in New Caledonia by the National Science Foundation (Grant no. GB-1545) is gratefully acknowledged. 320 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 oblanceolate to broadly obovate, or more or less elliptic, (4.5—)7—14(-19) cm, long by (2.5—)3-5.5(—7) cm. broad; margin entire, scarcely thickened, more or less flat; apex obtuse (or acute) often almost rounded, sometimes very slightly and very shortly acuminate, tip blunt; base obtuse to acute, attenuate into the petiole; venation more or less obscure, except for slightly raised primary veins above and below, 6-8 per side. /nflorescence axillary, with or below the leaves, decussate, 1-3 cm. long, 7—13-flowered with a single terminal flower, pubescent, with somewhat flattened hairs; bracts broadly triangular, 1-1.5 mm. long, pubescent, deciduous. Flowers herma- phrodite, pale yellow, inodorous (?); pedicels 0.5-2 mm. long. Calyx 1-1.5 mm. long, pubescent, with somewhat flattened hairs, lobes 4, bluntly triangular, 0.5—0.75 mm. long, pubescent, especially toward the apex. Corolla thickish, lobes 4, joined in pairs by the filaments, 2—2.5 mm. long, strongly valvate, cucullate, enclosing the stamens in bud. Stamens 2, 1.5-2 mm. long; filament broad, 0.5 mm. long. Ovary 1.5 mm. long, more or less conical, tapering to style and stigma 0.5 mm. long, shallowly bifid; bilo- cular with four axile ovules. Drupe club-shaped, 4.2 cm. long, 2 cm. broad, obscurely ridged, mesocarp thick, fleshy, “rosé” orem & Baumann- Bodenheim 10424), endocarp very hard, 1.5-2 mm. thic Bois des montagnes, Balade, 1861, Vieillard 97 (vp); humid forest, about 700 m. alt., Mt. Ignambi (about 10 km. south of Pouébo), 7 Dec. 1963, Green 1764 (A); margin of humid montane forest, about 600 m. alt., Mt. Ignambi, 7 Dec. 1963, Green 1798 (aA); in den Waldern der Berge .. ia ca. 800 m. alt., 2 Jan. 1903, Schlechter 15600 (isotypes: BM, E, K, LE, P, Z); forét ivorouhile sur schistes, 500 m. alt., Mé Aoui, 7 Feb. 1951, Gulllaumin & Baumann-Boden- heim 10270 (Pp); ibid., Guillowmnin co Baumann- Bodenheim 10271 (P, not seen) ; ibid., 8 Feb. 1951, Guillaumin & Baumann-Bodenheim 10424 (a, Pp); Ourai, 31 Oct. 187 6, 500 m. alt., Lécard sn. (P); vallée de la Thi, 1951, Sarlin 76 (holotype of Sarlina cylindrocarpa: P); forét rocheuse, Prony, 15 Sept. 1914, Franc 1867 (holotype of Notelaea francii: P; isotypes: A, BRI, K, P); ibid., Dec. 1914, Franc 1867a (A, z). The divided corolla clearly distinguishes Linociera from Osmanthus, and its relatively massive fruit from both Osmanthus and Notelaea. The decus- sate inflorescence is common to all three genera, although in one or two places on the type material of L. brachystachys the inflorescence is slightly compound with three flowers borne where a single one is normally found (see the discussion under L. paniculata below). Bud material was fixed in the field and flown to Dr. Barbara Briggs of the Royal Botanic Garden at Sydney, Australia, who examined it cytol- ogically. I would like to take this opportunity to express my sincere grati- tude to Dr. Briggs for her interest and willing help. Unfortunately, the material was not in proper condition for an accurate count but it was found that there were about 46 chromosomes in the somatic cells. This is close to a count of 2m = 46 obtained by Dr. Briggs on material of Linociera quadristaminea (F. Muell.) Knobl. which I fixed on Lord Howe Island, and agrees with the number almost uniform throughout the whole sub- family Oleoideae (see Taylor, Brittonia 5: 363. 1945) 1966 | GREEN, THE GENUS LINOCIERA 321 Linociera paniculata (Guillaum.) P. S. Green, comb. nov. Notelaea ? paniculata a Bull. Soc. Bot. France 89: 233. 1943, et Fl. Nouv.-Caléd. 283. Evergreen tree 10 m. high, branches glabrous. Leaves glabrous; petioles 8-12 mm. long, glabrous; lamina thickish, elliptic-oblanceolate, (4.2-) 6-9 cm. long by (2—)2.5-3 cm. broad; margin entire, scarcely thickened, flat; apex acute, tip blunt; base acute, long attenuate into the petiole; venation obscure above, only primary veins visible, slightly raised below, 7-8 per side. Inflorescence axillary, paniculate-decussate (? abnormal), 3.5—4.5 cm. long with up to 50 flowers, glabrous; bracts triangular, acute, very small, early deciduous or absent, 0.2-0.5 mm. long. Flowers herma- phrodite, pedicels 1-2 mm. long. Calyx glabrous, 0.75—1 mm. long, with 4 very shallowly triangular, more or less rounded lobes 0.25—0.5 mm. long. Corolla thickish, lobes 4, joined in pairs by the filaments, 2.25—2.5 mm. long strongly valvate, cucullate, more or less enclosing the stamens. Stamens 2, 2-2.25 mm. long; filaments flat, broad, 0.5 mm. long and 0.5 mm. wide. Ovary 1.5 mm. long, broadly conical with barely discernable style and two small stigmatic lobes 0.2—0.3 mm. long. Fruit unknown Sommet du Nékou, au-dessus de Bourail, vers 600 m. alt., 3 Apr. 1862, Balansa 1224 (holotype and isotype, P). It is strongly suspected that the paniculate condition of the inflores- cences on these specimens is abnormal. The tree from which the branches were taken was obviously stunted in growth and a pair of leaves on the isotype are fasciated at the base. Knoblauch, has determined the speci- mens as Notelaea brachystachys and it is interesting to note that a careful examination of the type of this latter species (Schlechter 15600) shows slight subsidiary branching of the inflorescence in one or two places, where three flowers are borne in the place of a normal single one. When more collecting has been done, and more material is available for study, this may possibly prove to be the correct disposition, and, were it not for the speci- mens mentioned immediately below, I would be inclined to agree with Knoblauch. In 1950 three specimens were collected on Mt. Koniambo in central New Caledonia, which bear immature fruit and appear to belong to Lino- ciera (sur serpentine, Mt. Koniambo, 21 Dec. 1950), Guillaumin & Baumann-Bodenheim 9458, 9479, 9487 (a, z). They appear to represent a different species from L. brachystachys, for the fruits, which are probably slightly immature, are shorter and broader in proportion (12-20 mm. long, 7-13 mm. broad), and the leaves are somewhat different (elliptic to slightly oblanceolate, (3.5—)5-8(—9) cm. long and (1—)2—2.5(-3) cm. broad, with the veins more obscure). Another gathering which is difficult to place, with the limited material at present available, is one collected in the Loyalty Islands by Daniker (vereinzelt im Walde am Weg von Donkin nach Nathalo beobachtet, Lifou, 2 Nov. 1923) Déniker 2366 (z). Judging by the size and narrow- 322 JOURNAL OF THE ARNOLD ARBORETUM [ VoL. 47 ness of the leaves this may prove to be conspecific with the Guillaumin and Baumann-Bodenheim material from Mt. Koniambo but it may also be a small-leaved expression of Linociera brachystachys. Linociera sp. Tree 12 m.; branches glabrous. Leaves wees petioles 4-5.5 cm. long, ieee: lamina thickish, oblanceolate, 15— ca. 23 cm. long and 5—9 cm. broad; margin entire, scarcely thickened, more or less flat; apex ? acute; base acute, long attenuate into the petiole; venation raised and more or less reticulate towards the margins above and below, 10-11 primary veins per side. Inflorescence ? decussate, ? ca. 5 cm. long. Flowers un- known. Drupe asymmetrically ellipsoid, obscurely ridged, 4.7 cm. long and 2 cm. broad (? immature), with blunt apex. Col d’ Ignambi, above Oubatche, 900 m. alt., 1 Oct. 1956, McKee 5379 HLU). This specimen, with its large leaves, bears a very close similarity to some of the species of Linociera from the East Indies and New Guinea. How- ever, until at least the Asiatic members of this genus are revised, it is im- possible to say whether it is different from all the previously described species. The material is inadequate for description as a new species. The opposite, exstipulate leaves, covered on both surfaces by small, sunken, peltate scales, and the type of fruit, all place the specimen in the Oleaceae. The type of fruit indicates Linociera and the size of the leaves and petioles distinguish it from L. brachystachys. It is apparent that much collecting needs to be done in the more remote areas Of New Caledonia. The flora in the northern part of the island and the mountainous hinterland is still inadequately known. It is to be hoped that future collectors will endeavor to obtain material from these relatively rien areas at times of the year when they have not, as yet, been visited. ARNOLD ARBORETUM Present address: HARVARD UNIVERSITY RoysaL BotANic GARDENS Kew, RICHMOND, SURREY GREAT BRITAIN 1966 | THE DIRECTOR’S REPORT 323 THE DIRECTOR’S REPORT THE ARNOLD ARBORETUM DURING THE FISCAL YEAR ENDED JuNE 30, 1966 THE EVENT WHICH DOMINATED the past year at the Arnold Arboretum was the decision rendered by the Supreme Judicial Court of Massachusetts in the case of the “Attorney General vs. President and Fellows of Harvard Arnold Arboretum in Cambridge, thereby forcing their return to Jamaica Plain. In a long opinion submitted by a majority of the Court (three justices) with two justices dissenting, the Court allowed for the continued close codperation between the Arboretum and the University. The Court also found that “the effects of the move do not show that inadequate consideration was given to probable results or that bad judgment was exercised.” The Court noted that “What we have said disposes of the contention that the move itself, so far as carried out as voted, was a breach of trust. We disagree with the view of our dissenting colleagues that it is implicit in the Indenture that the place of such implied related activities as the library and herbarium, if established, be the place speci- fied for the growing of the trees, shrubs and plants. The dominant con- sideration is that those things and only those things be done that will, in the good judgment of the Corporation, promote the express purposes of the Indenture.” ; The Court did question whether “the implementation of the resolution of January 19, 1953, does not, or may not, conform to its terms.” That resolution (cf. The Director’s Report on the Arnold Arboretum, Jour. Arnold Arb. 35: 367-381. 1954) “calls for the retention at Jamaica Plain of ‘such books and specimens as may be required to provide there a working library and herbarium’.” The “resolution required that the books and specimens to be housed at Cambridge be appropriately identi- fied.” The Court found that an “implication from the master’s observa- tion . . . may be that the intention of the Corporation’s vote has not been met.” The Court also ordered that “The rescript shall provide that the county court retain jurisdiction of the cause and that the defendant shall file with the clerk within six months of rescript (a) a report to it by the Arnold Professor or other appropriate person that in the Corpora- tion’s judgment shows either that an adequate working library and herbarium exist at Jamaica Plain or recommends the steps to cause them 324 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 to exist; (b) a report of its proposals or action to amplify the catalogue and general listing references; and (c) a report of its proposal or action in respect of a general designation on or within the Harvard University Herbarium.” The required report showing the nature of the herbarium and library for horticultural research which are maintained in Jamaica Plain has been approved by votes of the President and Fellows and transmitted to the single justice for the entry of the final decree. Since the decision of the Court in this litigation did not require the removal from Cambridge and the return to Jamaica Plain of all the books and specimens transferred in 1954, it is now possible for the staff to plan for permanent changes and improvements in the existing facilities in Jamaica Plain. Consequently, contracts were placed in the spring for (1) an additional unit at the Dana Greenhouses, constituting a one- third expansion; (2) air-conditioning of the laboratory for cytology and plant morphology; (3) new lighting in the tases of the Administration Building; (4) changes in office arrangement In the Harvard University Herbaria in Cambridge equally important consideration must be given because of growth of collections and increase in personnel and visitors for whom facilities are now limited by the physical plant. The decision of the Supreme Judicial Court of Massa- chusetts was necessary before any new programs could be proposed. The staff believes that the present organization and locations will permit the goals of the Arnold Arboretum to be extended and is hopeful that the hesitation and uncertainty of the past are, indeed, over. Two small sections of land have been officially incorporated into the Arnold Arboretum during the year. One piece of approximately five acres, including a house at 383 South Street, was in the title of the Bussey Institution. With the taking of the major part of the Bussey Institution property by the Commonwealth of Massachusetts in 1963, this area, adjacent to the collections of the Arboretum, remained. It has been used by the Arboretum staff and the house has been occupied by a staff member as an essential guard post for one corner of the grounds. The Corporation in a vote of June 6th approved that “the land shall be held and used for the general purposes of the Arnold Arboretum” and “in consideration the fair market value of $71,000.00 be transferred from the Martha Dana Mercer Trust gift held for the - ah soite of the Arnold Arboretum to the Bussey Institution Endowmen A second piece of land in the northwest corner of the Arboretum was isolated from the main collections and could not be developed because of the use and retention of ownership of an old roadway by the Depart- ment of Public Works. After many years of effort the title to this isolated right-of-way was transferred to the owners of the abutting property, Harvard University and the City of Boston Department of Parks and Recreation, respectively, and its use given to the Arnold Arboretum. We deeply appreciate the efforts of the Commissioner, Mr. William J. Devine, and of his associates, Mr. Arthur J. O’Keefe and Mr. Daniel Byrnes, ree ee HOR Pgs, ia — 2 ae ‘ Edwin Shit of the National Park Service presenting certificate designating the Keassa iba: ink retum a Naloead Historic rete ick Left to nant, Mr. Henry Scagnoli, Mr. L. Gard Wiggins, Mr. A. Parker, Jr., RicH Mr. Small ree (far right) Dr. ard, The bronze slate at the main nates cs to the Arnold Arboretum. [9961 LYOdaa S AOLIAAIG AHL oe) Nn 326 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 in making this land available. The expansion of the conifer collection and the development of additional plantings of ornamental apple trees in this area, along United States Route 1, will greatly enhance the site for visitors approaching Boston. Much time was devoted by the staff to securing aid to defeat a bill submitted to the Massachusetts legislature ordering the construction of a ski tow within the Arnold Arboretum, on Peters Hill. Although the bill was filed by a representative living in Jamaica Plain, the legislature properly realized that passage and implementation of it would destroy a major collection of apples and crab apples, and defeated the bill by voice vote. The staff is grateful to the many Friends of the Arboretum whose aid in contacting their representatives and expressing their disap- proval was essential in defeating the bill. A pleasant occasion to record occurred on May 17 during the meeting of the Committee to Visit the Arnold Arboretum. At that time Mr. Edwin Small, of the National Park Service, representing Stewart L. Udall, Secretary of the Interior, presented the certificate officially recog- nizing the Arnold Arboretum as a National Historic Landmark. Mr. A. H. Parker, Jr., Chairman of the Committee, accepted the certificate for the Arnold Arboretum. Also present were Mr. Henry Scagnoli, repre- senting John F. Collins, Mayor of the City of Boston; Mr. L. Gard Wig- gins, representing Mr. Pusey and the Harvard Corporation; and Mr. Daniel Byrnes, representing Mr. Devine and the Department of Parks and Recreation. Earlier in the year a bronze plaque was placed on a gate post at the main entrance to the Arboretum. Staff: The retirement of one staff member and the resignation of four others took place during the past year. Mrs. Lilian Burian, who had mounted herbarium specimens for ten years, retired on January 30th. Dr. Theodore Dudley, Assistant Horticultural Taxonomist, resigned to accept a posi- tion at the U. S. National Arboretum. Mr. Peter Green, Horticultural Taxonomist, resigned to join the staff of the Royal Botanic Gardens at Kew, and Dr. Keith Ferguson resigned to work as a Fellow on the Staff of the Flora Europaea Research Project. Dr. Mary Sanders accepted an appointment at the Mount Alto Campus of the Pennsylvania State University, and resigned her appointment as Research Associate. The appointment of Dr. Harrison Leigh Flint as Associate Horti- culturist was approved by the Corporation to become effective near the end of the present fiscal year. Dr. Flint, whose research interests in- clude problems of winter hardiness of woody plants, joins us from the staff of the University of Vermont. Miss Dulcie Alicia Powell, formerly of the staff of the Institute of Jamaica, Kingston, Jamaica, was appointed Research Fellow, and will continue her work on the cultivated ornamental plants of tropical areas. Miss Stephanne Barry Sutton was also appointed Results of damage caused by severe snow storm of January 23, 1966. Lerr: Trees in the Quercus collection with broken peg RicHt: Damage to branches of this 40-year-old Corcidiphylinn necessitated complete removal of the LYOdaU SMOLOAYIG AHL [9961 Lee 328 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 a Research Fellow and continues her work toward a biography of Charles Sprague Sargent and a history of the Arnold Arboretum. The promotion of Dr. Thomas G. Hartley to be Associate Curator of Pacific Botany was approved by the Corporation, to become effective during the next academic year. Five Mercer Research Fellows were appointed during the year for varying lengths of time. Dr. Charles Tseng, of Windham College, Putney, ermont, is carrying on morphological studies in the Araliaceae. Mr. Pablo Legname, of the Instituto Lillo, Tucuman, Argentina, has under- taken studies on the unworked collections from Argentina in the her- barium of the Arnold Arboretum. Appointments as Mercer Fellows in the area of horticulture, for work with the living and herbarium collec- tions, were made to Miss Sandra Shannon, of the University of Massa- chusetts; Mr. Frank Wolfe, of the University of Vermont; and Mr. William Gensel, of the University of Rhode Island. Among honors to staff members was the award of the Jackson Dawson Medal of the Massachusetts Horticultural Society to Mr. Alfred Fordham. It was accompanied by the citation, “To Alfred James Fordham, who following in the footsteps of famous propagators at the Arnold Arboretum, is doing much to maintain the high standards set by them years ago in this often difficult and sometimes most perplexing art.” r. Wyman was presented the “Silver Alder Leaf” by the Scandinavian Horticultural Congress, at Aulanko, Finland, in July, 1965. Horticulture: The drought which has plagued New England and which has been mentioned in several recent reports continued during the fall and winter of this past fiscal year. The records show that the calendar year 1963 ended with the total rainfall 7.93 inches below average, 1964 was 6.30 inches below normal, while 1965 had a deficiency of 16.76 inches from the 44 inches normally expected at the Arnold Arboretum. The first four months of 1966 were 3.26 inches below normal, but the drought appeared to be broken, at least temporarily, by occasional heavy showers during the early part of June. The effects of a long continuing drought, however, are very evident in the resultant loss of plants. The few ponds in the Arboretum remained dry again during the summer of 1965, indicating how low the water table had dropped. In addition to the dry spell, the grounds were also affected by surface fires caused by carelessness and vandalism. Twenty-one separate fires were recorded during the past year. Fires beginning at the base of Bussey Hill have severely damaged plantings of azaleas so that many = plants have had to be pruned to the base in an attempt to save em. Adding to the damage caused by the continued drought was a major storm on January 23rd with a heavy wet snow which remained on the branches of shrubs and trees. Colder weather following the storm caused 1966 | THE DIRECTOR’S REPORT 329 The planting program, drastically limited in the last two years, had to be resumed during the past spring because of the accumulation of plants in the nurseries. Over 1000 plants were added to the living col- lections. New azalea plantings numbered 265. About 50 new Cytisus plants were added to the collection on Bussey Hill. One purpose of Dr. Wyman’s travels through Europe during the spring and summer of 1965 was to select from many sources, including botanic gardens, private gardens, nurseries, and the wild, plants worthy of introduc- tion into the United States. During the current fiscal year material of 930 species and varieties of 119 genera was received, recorded, and prop- agated at the Arboretum. Of several thousand individual plants now growing in the greenhouse area, some will be ready for hardiness testing during the next year. These are included among the 167 shipments re- ceived during the year from 11 countries, comprising 1012 species and varieties. In addition, shipments of seeds received numbered 130 from 35 different countries, comprising 364 species and varieties. Many of the for representation in the herbarium of cultivated plants. In response to specific requests, the staff of the greenhouses distributed 109 shipments of 1009 taxa to ten countries. Included in these figures are twelve taxa of Cornus, Ilex, Ligustrum, Magnolia, Malus, Pieris, Potentilla, Syringa, and Viburnum which were offered to codperating nurserymen as plants introduced or selected by the Arnold Arboretum and worthy of greater horticultural use. A total of 389 taxa was propagated for replacements or additions to the collections on the grounds. One hundred and fifty-one taxa were propagated for taxonomic or analytical studies by staff members. Continued experimental work in the greenhouse area concerned methods of plant propagation and the accumulation of data on seed germination. Through the year 112 taxa were processed for such investigations. Work on the registration of cultivars continued, the publications in- cluding a check list for the genus Weigela and one issue of Arnoldia devoted to newly registered names in various genera. Inquiries for infor- mation and for the proper registration forms still far exceed actual A full discussion of the registration procedure is planned for Compilation of a listing of cultivars for the genus Lantana nears com- pletion. Dr. Dudley gave special attention to the material of Viburnum in our 330 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 herbarium of cultivated plants and to that obtained on loan from other institutions, as well as to the living collections. Mr. Green and Mrs. Riidenberg continued their study of Lonicera, annotating many of the specimens in the Arboretum herbarium collections and making additional chromosome counts. The horticultural herbarium received sizable and welcome additions from the collections made by Dr. Dudley on the Burpee Fordhook Farm and at other locations and by the gift of the herbarium of the American Herb Society. A new edition of Dr. Wyman’s book, Trees for American Gardens, was published by Macmillan at the end of December. Changes of various sorts have been made at Jamaica Plain, some on the grounds which will become apparent in better landscaped plant- ings, more convenient paths for visitors, or healthier specimens; others will result in more efficient use of the facilities in the buildings, both for research and maintenance. On Bussey Hill a new retaining wall of natural boulders was constructed so that a level path could be established. Heavy mulching and the laying of a plastic pipe from South Street to the top of the hill, making water available for the first time, will improve the condition of both old and new plantings and enhance the general appearance of the collections. Adjacent to the oak collection on Busse Hill a beginning has been made on the establishment of a collection of Viburnum species. At the base of Hemlock Hill many evergreen rhodo- dendrons have been planted in mass. Work has commenced on the establishment of terraces and walks on the slopes below the bonsai house in the greenhouse area, using rocks obtained from the Case Estates in Weston. When completed this will become the location for the smaller plants of the genetic dwarf conifer study collection. Larger plants will continue to be placed in the area of the chestnut collection. Much effort is being expended on the rehabilitation of the old Centre Street roadbed, obtained from the Department of Public Works by the Department of Parks and Recreation. The city has removed the work house and the Arboretum has contracted for the excavation of the road to a depth of about 3 feet, using this excavated material for filling in a swampy area and a dangerous pond on the South Street tract. Much fill will be needed in the roadbed, but localized plantings will be used in the meantime. In codperation with the Department of Parks and Recreation a six-foot chain link fence is being erected around the new property. The chain link fence bordering the Arborway was painted during the summer by boys in the Job Corps, a project directed by the City of Boston. Also with the codperation of the Department of Parks and Recreation, the gates to the Arboretum were closed to driving on week-ends during the months of May and June. With very few exceptions this practice has 1966} THE DIRECTOR’S REPORT Ge Lo) ~ Asove: Collection of genetic dwarf conifers in a temporary location near the Dana greenhouses. Mulch of gray stone is being used for these plants. eae ELOW: Dry wall of natural boulders bordering the rebuilt beds of Cytzsus on Bussey Hill. 332 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 met with the approval of those commenting, since it permits individuals and families to walk without concern for automobil i special police were hired for the summer months to patrol the grounds during the evening hours. With the codperation of the planning office of the University and the an h ildi Gro nt from cramped and crowded quarters in the asement of the Administration Building. Case Estates: t is pleasant to record again a marked increase in the use of the Case Estates, both by individual visitors and by groups requesting special tours. In both spring and fall, classes were held as afternoon walks around the — and fee bite! an Open House was held on what, unfortunately, was a cold, rainy Sunday. Parking for visitors remains a problem both on W steley Srest and near the barn Display area at the Case Estates, Weston, for authentically named clones of Hosta sent from special collections in Sweden, England, and the United States. i 1966 | THE DIRECTOR’S REPORT 533 One new planting has been established, and additions have been made to others. A display of mulching materials prepared for exhibit at the Spring Flower Show of the Massachusetts Horticultural Society, in 1964, aroused such interest that the plan was developed for use in Weston. Two long beds, 6 by 100 feet, were prepared in the fall and planted in the spring. Each bed is divided into units ten feet long, permitting the use of 20 different kinds of mulching materials. Instead of using the same plant materials in each bed an alphabetical arrangement of peren- nials by their generic names has — ised, omitting only such “exotics” as genera beginning with “Q” and “ Additions have been made to ] collection of Hemerocallis. New taxa of Allium have been acquired by gift and purchase in the hope of establishing a new bed for next year. An experimental nursery planting of 132 trees was set out in the large field along Newton Street, and 89 trees of 37 species and varieties were added as display specimens in the same field. The town of Weston was given 193 trees and shrubs in the fall of 1965 for planting in the town. All had been placed by spring, and a letter of appreciation for the gift was received from the members of the Weston Park and Cemetery Commission. Herbarium: The decision of the Supreme Court granting permission to keep col- lections of the herbarium and library in Cambridge has made feasible a reéxamination of the herbarium building and its facilities. In 1954, when the collections were moved, the capacity of the building for the combined collections of the Arnold Arboretum and the Gray Herbarium was considered adequate for ten years’ growth. It is now twelve years since the move occurred and the increase in the herbarium collections has exceeded what was expected, partly from the study and incorporation of previously unworked collections and partly from expanded and in- creased interest in plant groups and geographical areas not formerly studied by the Arboretum staff. A developing interest in the herbaceous plants of all floras has resulted in such material being studied, mounted, and inserted as a part of the Arboretum herbarium, rather than as formerly, being donated to, or exchanged with, other institutions, or even discarded. An increasing awareness of the importance of the floras of Africa and Australia, and the deficiency of our collections in these areas, have led to active efforts to increase our collections by gift or ex- change with institutions working seriously on these floras. The col- lections so expanded will make available for research and teaching in- teresting representatives of the plant families characteristic of these parts of the world. As a result, the collections are increasing in a normal way for an active herbarium. The number of staff, students, and visitors also increased since 1954 and should continue to do so. This, however, is straining the existing 334 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 working facilities beyond capacity. A small table and simple microscope are inadequate to cope with the newer and more sophisticated taxonomic techniques which require laboratory space and, often, photographic equip- ment. The building has become inadequate both for the staff needs and for housing the collections. In fact, it has become necessary to store certain groups of specimens in cardboard boxes on top of the steel cases. In the horticultural herbarium in Jamaica Plain the situation for growth is more favorable. Case space is adequate for a decade of expansion, and space for more cases and for special laboratory facilities in the Adminis- tration Building is available. During the year 18,647 specimens were mounted and added to the herbarium collections in Jamaica Plain and Cambridge, bringing the total number of sheets to 824,920. Also during the year 14,483 specimens were pee the greater portion by exchange. Special attention was the organization and distribution of plants from Brazil and eres collected about 1940 by Richard Evans Schultes, making available some extremely interesting representatives of that flora. A very important addition to the horticultural herbarium was the gift of the herbarium of the Herb Society of America which is now being appropri- ately marked to indicate its origin, and inserted into the herbarium. Access to this collection will, of course, be available to members of the Herb Society when needed, but the specimens will also be available on loan to monographers for research studies, and to the staff for reference. During the past year staff members and students requested or received for identification 2,496 herbarium sheets representing 51 loans from 36 institutions. The staff filled requests for 120 loans to 65 institutions, including 16,491 specimens. Arrangements were made during the year to collaborate with the Universidad Nacional Autonoma de México on the production of an ecologically oriented Flora of the State of Veracruz, México. The study will be under the direction of Dr. Arturo Gémez Pompa, a former Mercer Research Fellow, now Director of the National Herbarium of México. The staff of the Arnold Arboretum will collaborate in the study of the cultivated plants and by supplying assistance in herbarium, library, and research facilities. Dr. Lorin Nevling is expected to serve as coordinator for the rest of the staff. Mr. Baranov prepared for publication taxonomic studies in the genus Prinsepia of the Rosaceae, and has completed a study of the morphological variation in Sorbaria sorbifolia. It is hoped that additional species and forms of these genera may be procured from the wild sources in Asia for cultivation in the Arboretum. Dr. Brizicky devoted much attention to recent publications on culti- vated plants and native floras from Russia. Since discussions at the 1966 International Horticultural Congress will consider the registration of cultivar names, problems of the code of nomenclature regarding trans- lation and transliteration were the basis of a joint publication with Dr. Howard. He completed his studies of the Sterculiaceae and Goodeniaceae 1966 | THE DIRECTOR’S REPORT 335 for the generic flora of the southeastern United States and has turned to a study of the Malvaceae represented in the area. Dr. Dudley continued his studies of Alyssum and its relatives, in- cluding the evaluation of hardiness trials of many species introduced to the United States for the first time and grown at the Case Estates. He completed a review of Asian species of Viburnum. He also compiled bibliographic information on taxa of Sambucus under cultivation, pre- liminary to the publication of a cultivar registration list for the genus. Dr. Ferguson completed the revision of the Cornaceae, Caprifoliaceae, and Chenopodiaceae for the generic flora of the southeastern United States, Mr. Green was fortunate to be able to establish collaboration with botanists in Australia interested in the floras of Lord Howe and Norfolk Islands, since he has turned his attention to a review of the flora of these islands with special emphasis on the origin and distribution of the plants and the nature of the plant communities involved. He completed a study of the cultivated taxa of Jasminum which was published in Baileya. A review of the species and hybrids in the Lonicera tatarica complex was also completed and has been published. Dr. Hartley’s monographic study of the genus Zanthoxylum in Malesia is being prepared for publication. He is continuing studies of his own col- lections from New Guinea and those of some other expeditions sponsored by the Arboretum or received as exchange. Dr. Howard received several collections of plants representing native and cultivated floras of the West Indies, among which special collections from the Bahamas, Bermuda, Tortola, and small adjacent and poorly known islands, and from Puerto Rico represented valuable additions to the herbarium. The study of the mossy summit forest on Pico del Oeste in the Luquillo Mountains of Puerto Rico is well under way. Cooperative environmental studies done with the National Center for Atmospheric Research have permitted the establishment of accurate weather instru- ments in a small building on the summit and within the elfin forest. Both automatic and manual recording instruments are being used to make various kinds of measurements related to environmental factors. Veg- etational studies include measurement of growth of parts, persistence of leaves, time of flowering and fruiting, observations on pollination methods or pollinators, and on seed and fruit dispersal. Dr. Nevling and Mrs. Riidenberg are collaborating on the cytological study of species within this forest. This complete and detailed biological-environmental study is made possible by a grant from the National Science Foundation and with the codperation of many people and institutions in Puerto Rico and the continental United States. The permission of the Forestry Service of the Commonwealth of Puerto Rico for the use of this land is grate- fully acknowledged. The contribution by Mr. Joseph B. Martinson of facilities for housing the field staff is most important in the development and continuation of this work. Dr. Hu continued her work on the Compositae of China. The first three parts of this study, representing approximately one-third of the 336 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 total, have been published. Dr. Hu is also writing special treatments of the genus Hemerocallis for a Daylily Handbook being prepared by a com- mittee of the American Horticultural Society and the American Hemero- callis Society. Dr. Nevling, as Supervisor of the Herbarium of the Arnold Arboretum and Supervisor of the Gray Herbarium, has directed various projects in- volving the two collections. He has continued his studies of the Thy- meleaceae, devoting special attention to South American genera and species in preparation for field work in Brazil, Venezuela, and Trinidad. He has also worked with Dr. Howard in the field study of the mossy forest noted earlier. Dr. Perry has made much progress in handling the Cape York, Aus- tralia, collections of the Archbold Expedition of 1948. With the collabora- tion of Dr. Hartley these collections have been labelled, at least partly named, and divided into sets for distribution. Dr. Schubert has completed a treatment of the genus Desmodium for the Flora of Panama and is working on treatments of the genus for the Flora of East Tropical Africa and the Flora Zambesiaca, Together with Dr. Lyman Smith of the Smithsonian Institution she is continuing a re- view of the genus Begonia in Colombia, to bring up-to-date a study pub- lished by the same authors in 1946. The first of a series of investigations of the genus Dioescorea was published during the year and further work on the genus is in progress. Dr. Schubert visited Belém, Para, Brazil, in December and spent time there collecting in the Area de Pesquisos Ecologicas do Guama, an ecological study area maintained cooperatively by the Smithsonian Institution, The Instituto Agronédmico do Norte, and the Belém Virus Laboratory of the Rockefeller Foundation. With the aid of representatives of these institutions she was able to obtain speci- mens es considerable interest. . Wood continued his direction of the project concerned with the a. flora of the southeastern United States. He has supervised the preparation of illustrations and edited and contributed to the manuscripts published during the past year. He did field work in Florida during the spring, collecting material for critical studies and for illustration. He continued to serve actively on the Subcommittee for Family Names of the Committee for Spermatophyta, which is now concentrating its efforts on the names of families of living gymnosperms. During the spring Dr. Wood served on a committee of the American Society of Plant Taxono- mists which met at the Smithsonian Institution to consider the feasibility of a Flora of North America. We are pleased to have had many visitors who worked on special groups in our collections. In addition to the Mercer Research Fellows, Dr. Charles C, Tseng and Mr. Pablo Legname, Mr. Lindsay Smith of Brisbane, Australia, visited the Herbarium to study species of Lantana which have become noxious weeds in Australia. Mrs. S. G. M. Carr de- voted her attention to the extensive collections of Eucalyptus. Mr. Ramon Riba, a Guggenheim Fellow, has come from the University of er TO ne ge ee pee all t ft 1966 | THE DIRECTOR’S REPORT 337 Mexico to study tree-ferns. Mr. Lucio Quimbo from the College of Forestry, University of the Philippines, was particularly interested in the large wood collection with herbarium specimen vouchers. Library: The status of the library of the Arnold Arboretum was also affected by the decision of the Supreme Court permitting the retention of books in Cambridge. Since a complete reorganization of the whole library is not now necessary, it has been possible to initiate a rearrangement of the horticultural library in Jamaica Plain to make it more easily usable. The work of cataloguing the material in the new arrangement will, of necessity, continue through the summer, but an important start has already been made this year. The new arrangement will permit greater use of the library in association with the horticultural herbarium and make it easier to use the library in the reference room as an open-stack-library. In the catalogue of the combined libraries of the Arnold Arboretum and the Gray Herbarium in Cambridge the librarian, Mrs. Schwarten, is indicating the ownership of the book on each catalogue card, and for the Arboretum books, specially marking each one housed in Jamaica Plain. This work has been completed for about one-half of the catalogue and the whole task should be finished within this calendar year. During the year 431 bound volumes were added to the library of the Arnold Arboretum, The total count of bound volumes is 53,295. Re- prints numbering 320 were catalogued bringing the total of the reprint collection to 20,594, The annual increment of 3000 cards was added to the Card Index of American Plants published by the Gray Herbarium and 2650 cards were added to the /ndex to American Botanical Literature published by the Torrey Botanical Club. Series 21 to 23 of the Index Nominum Genericorum were received and incorporated into that file. Approximately 900 reference cards were prepared and added to the Rehder index of cultivated plants. Miss Stephanne Sutton continues her work on the history of the Arnold Arboretum, on the biography of Charles Sargent, and on the organization of the historical correspondence of the Arboretum. During the year Miss Sutton made a study trip to England to consult letters of Charles Sargent in the files of the Royal Botanical Gardens, Kew, the British Museum (Nat. Hist.), and at the library of the Royal Horticul- tural Society. The assistance rendered Miss Sutton in this search is gratefully acknowledged, as are the many photocopies of letters to and from Sargent which have been sent from many institutions and individuals in the United States. It is hoped that still other letters representing different periods or other facets of Sargent’s life will be found. The Arboretum received several large collections of reprints and per- sonal files and also collections of books. The work on the reorganization of the reprints of Professor Joseph Faull, described in earlier reports, is now nearly complete, and the collection is in usable condition. The com- 338 JOURNAL OF THE ARNOLD ARBORETUM [VOL. 47 plete files of Harlan P. Kelsey have proven to be extremely valuable for their material pertaining to the Arnold Arboretum, to the introduction of native species into cultivation, to the development of standardized plant names, and to the development of natural areas and national parks. Materials from this gift concerned with city planning have been trans- ferred to the Harvard University Library of the Graduate School of De- sign. The numerous letters are being classified and will be available for research purposes. During the year Mrs. George Hamor presented to the Arboretum many books from the library of her late husband which increase our source material on agriculture and botany in the American tropics, especially in the West Indies. The library also had many visitors during the year. Among the very special projects in which our collection was of great usefulness we note that of the Rachel McMasters Miller Hunt Botanical Library to photo- graph the title pages of books of botanical and horticultural interest. Since much of the material is most conveniently available here, we are glad to allow it to be photographed and reproduced. Comparative Morphology: Professor Bailey has continued his research on the primitive leaf- bearing Cactaceae under an extension of his grant from the National Science Foundation. A paper considering the significance of the reduction of vessels in the Cactaceae is being prepared. Investigations also have been made of the biochemical nature of materials within the tissues of these plants. Professor van der Schijff, Mercer Research Fellow from the University of Pretoria, continued his studies of the peculiar growth characteristics and enlarged subterranean stem development of species of the Leguminosae. Some of the new techniques being developed for work associated with electron microscope studies were modified for application to special prob- lems of fixation and sectioning of this difficult material. Dr. Charles Tseng continued his studies of the pollen and floral mor- phology of the Araliaceae and Umbelliferae. Dr. Howard concentrated his studies of nodal and petiolar anatomy on plants from the mossy forest study area in Puerto Rico, and to variations in the secondary body developed in leaves of more than one year’s dura- tion. Cytology and Genetics: Mrs. Lily Riidenberg developed her studies of the taxa of Lonicera under cultivation in the Arnold Arboretum in conjunction with the systematic investigations of Mr. Green. Chromosome counts have now been obtained for all but a few of the most recent introductions and a publication recording the counts and the documenting herbarium speci- mens is soon to be published. 1966} THE DIRECTOR’S REPORT 339 Dr. Nevling and Mrs. Riidenberg are both participating in the cyto- logical study of the component species of the mossy forest in Puerto Rico. e 84 genera of flowering plants in the area present a variety of diffi- culties in the search for division figures in which the chromosomes can be counted. Accurate chromosome counts have been obtained in certain plants only from buds collected in the early morning hours, while in other taxa the most active stages of microspore division figures seem to occur at different hours or in different sequences. The length of stages in which divisions can be obtained also seems to vary considerably among the species comprising the elfin forest. Education: No formal courses were given by members of the Arboretum staff, but several of them were special lecturers in a course in Plant Geography, or supervised the work of students. Informal field classes were offered at Jamaica Plain under the direction of Dr. Wyman and on the Case Estates in Weston by Dr. Hartley. A course in Plant Propagation was offered throughout the year by Mr. Fordham. The staff took part in the regular series of weekly seminars held in the Herbaria in Cambridge. The Arboretum staff served as host to various visiting groups for con- ducted tours, special seminars, and general discussions or provided space and facilities for meetings. The horticultural judging and identification competition sponsored jointly by the University of Massachusetts, the Massachusetts Department of Agriculture, and the Boston Public Schools was held in the auditorium of the Administration Building. The mid- special open house in Jamaica Plain and an invitation to visit all of the Arboretum facilities was extended to members of the Department of Biology of Harvard. The Arboretum was represented at many national and international meetings and several staff members presented papers. Mr. Fordham at- tended the annual meeting of the International Plant Propagators’ So- ciety, the meetings of the American Society for Horticultural Science and the Horticultural Day program at the University of New Hampshire and spoke on all these programs. Dr. Hartley attended the public symposium “Scientific Aspects of Pest Control,’ arranged by the National Academy of Sciences, National Re- search Council, held in Washington, D.C. Dr. Howard, as lecturer for the American Institute of Biological Sci- ences visited the campus of Ohio University in Athens, Ohio. He spoke about the Spice Trade of Colonial New England at the Gardeners’ Work- shop at Sturbridge Village; to the Connecticut Nurserymen’s Association short course at the University of Connecticut, and on the lecture series of the New York Botanical Garden. He attended the meetings of the 340 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 American Association of Botanic Gardens and Arboretums where he led a discussion of cultivar registration procedures; the annual meeting of the American Horticultural Society, at Callaway Gardens in Georgia; the meeting of the Directors of Systematic Collections at the Bishop Museum in Honolulu, and of the scientific advisory board of the Fairchild Tropical Garden in Miami; and represented the Arboretum at the celebration of the centennial of the Smithsonian Institution. At the request of the Department of Global Medicine of the U.S. Naval Medical School in Bethesda, Maryland, Dr. Howard presented a program on the medical significance of the vegetation of southeast Asia, considering the edible, poisonous, and dangerous plants of the area. The program was later taped in color for television use and as an educational film. Dr. Nevling presented a paper during the symposium on the Biota of the Amazon Basin, on the need for further material and knowledge of tropical American taxa of uncertain systematic position, particularly in relation to activity in the proposed International Biological Programs. Dr. Wood was invited to lecture at Kent State University on his work with carnivorous plants and on the generic flora of the southeastern United States. Dr. Wyman addressed the Scandinavian Horticultural Congress in Aulanka, Finland. He attended the American Horticultural Society an- nual meeting in Georgia, and spoke on the program of the Clara B. Ford Garden Forum in Dearborn, Michigan, the Williamsburg Garden Sympo- sium, the Farm and Home Week of the University of Maine, and at the Western Association of Nursery Men in Kansas City. He also presented a special workshop on “Trees, shrubs and vines for small gardens” for the Brooklyn Botanic Garden. Travel and Exploration: The convenience and speed of air travel and the availability of federal and foundation grants to support field work and attendance at scientific meetings have made it possible in recent years for botanists and horti- culturists to travel widely. Dr. Dudley visited several gardens in New York, New Jersey, and Pennsylvania to collect specimens for the horticultural herbarium. Mr. Fordham continued his field studies of naturally occurring witches’ broom in conifers in several New England states and visited six botanical gardens and arboreta in Ohio. Dr. Howard and Dr. Nevling made two study trips to eastern Puerto Rico during the year. Dr. Nevling attended a sympo- sium in Belém, Para, Brazil, and en route was able to do field work on plants of his speciality near Rio de Janeiro, in Venezuela, and in Trinidad. Dr. Schubert attended a council meeting of the Association for Tropical Biology, also in Belém, and spent a week collecting near there. Miss Sutton made a trip to England to carry on her biographical and _his- torical investigations. Dr. Wood took part in one of the study trips to Puerto Rico and was able to have several weeks in the field in Florida 1966 | THE DIRECTOR’S REPORT 341 where he collected from Tampa to the Florida Keys and northward to Jacksonville. Dr. Wyman completed his che dlc trip during the present fiscal year Members of the staff are most grateful to all those individuals who helped in various ways to make their trips both successful and interesting. Gifts and Grants: Three grants were received by staff members from the National Science Foundation. Dr. Howard was awarded a grant for two years to study the biology and environment of the mossy or elfin forest in Puerto Rico. A one-year grant to Dr. Wood is a terminal grant for the project on the generic flora of the southeastern United States. The National Science Foundation has supported this work generously for the past ten years. Dr. Bailey Sane a grant to continue his investigations of the leaf- bearing Cactacea In addition, a gine, grant from the Association for Tropical Biology supported the work and travel of Dr. Nevling in Brazil. Special anony- mous gifts were used to aid the travel of Dr. W yman in Europe and of Miss Sutton in England. An appeal to the Friends of the inet Arboretum brought renewed contributions from many loyal supporters. This assistance is used en- tirely for work in horticulture and on the grounds of the Arnold Arbo- retum unless otherwise designated. Certain groups of plants and certain of the plantings in the Arboretum have their own special friends who desig- nate the purposes of their gifts. It is a pleasure to carry out such wishes. Every year special gifts of books, herbarium specimens, botanical prints, plants, Kodachrome slides, and photographs are received from many sources. Special acknowledgment is sent directly to these donors for their interest in and generosity to the Arboretum. Publications: The regular publications of the Arnold Arboretum are the Journal of the Arnold Arboretum and Arnoldia. The Journal was edited by Dr. Schubert and the four issues published during the year comprised 430 pages and 25 articles. Publication of the letters from Charles Sprague Sargent to Reginald Somers Cocks, edited by Dr. Joseph Ewan of Tulane University, was completed in this period. A detailed bibliographic study of Grisebach’s Flora of the British West Indian Islands was contributed by the British botanist William T. Stearn, A group of related papers on a new genus of palms by Dr. aaa E. Moore, Jr., and Dr. Natalie Uhl of the Bailey Hortorium and Dr. P. B. Tomlinson of the Fairchild Tropical Garden represent pena - the interesting contributions published during the year. During Dr. Wyman’s absence Arnoldia was edited by Mr. Green. The twelve issues during the year contained a variety of articles on horticul- tural topics ranging from a check list of cultivars in Weigela to new and 342 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 additional pictorial studies of tree barks. One issue devoted to plants which could be used to screen junk yards and dumps has been reprinted twice and, by request, rewritten slightly for other magazines. A new section “Arnoldia Reviews” will appear from time to time to call attention to publications of interest in horticulture, botany, and forestry. A cumulative index to the nine volumes of the Symbolae Antillanae of I. Urban was issued as a special publication. The Index was prepared by Miss Carroll and Miss Sutton, under the direction of Dr. Howard, who prepared a biographical sketch of Dr. Urban A set of 18 postcards was printed to supplement and replace those already issued. In response to special requests cards of the bonsai, fall foliage colors, and winter scenes were included. Two new brochures ‘with new maps of the Arnold Arboretum and the Case Estates were prepared by Miss Jeanne Taylor The bibliography which follows includes 56 articles and books by mem- bers of the staff, published during the past fiscal year. Bibliography of the Published Writings of the Staff and Students July 1, 1965—June 30, 1966 Bartey, Irvinc W. Comparative anatomy of the leaf-bearing Cactaceae, XIV. reliminary observations - the vasculature of the cotyledons. Jour. Arnold Arb. 46: 445-452. 5, . Comparative diet of the leaf-bearing Cactaceae, XV. Some pre- liminary observations on the occurrence of “Protein Bodies.” Jour. Arnold BaRANov, ANDREY I. & SKVORTZOV, B. V. Plantae novae et minus cognitae florae Chinae boreali-orientalis. Quart. Jour. Taiwan Mus. 18: 219-232. 1965. . What is Allium macrostemon from North Manchuria, China? Quart. . Taiwan Mus. 18: 471-475. a tiie Georce K. Cultivated plants and their wild relatives. Taxonomy, geography, cytogenetics, ecology, origin, utilization. [In Russian.| P. M. ZHUKOvsky. 2nd ed. Leningrad, 1964. [Book review] Econ. Bot. 20: era of Tiliaceae and Elaeocarpaceae in the southeastern United States, ae Arnold Arb. 46: 286-307. 1965. : im’ an visio rere cultivar of Syringa. Quart. Newsl. Am. Assoc. Gard. Arb. 64: 22. 1965. e genera a Sterculiaceae in the southeastern United States. Jour. Arnold Arb. 47: 60-74. 1966. (with Fercuson, Ian K.). ecenciararel notes on Dipsacus fullonum and ee sativus. Jour. Arnold Arb. 46: 362-365. 1965. ie) CHARD A.). The fectulition and transliteration of clue names hers some notes on lilacs. Quart. Newsl. Am. Assoc. Gard. Arb. 64: 15-21. 1965. DupLEy, THEODORE “ Alyssum ogy account of]. Jn: Davis, P. H., Ed., Flora of Turkey 1: 362-409. 1965 Aé Ad 1966] THE DIRECTOR’S REPORT 343 ————. The Anatolian collections of Alyssum L. made by Dr. Friederike [sic] Sorger (Linz/Donau, Austria), in 1962, 1963 and 1964. Oesterr. Bot. Zeitschr. 112: 746-761. 1965. . Aurinia [The account of]. Jn: Davis, P. H., Ed., Flora of Turkey 1: 361, Ais 1965. —., mental madworts (Alyssum) and the correct name of the golden- tuft pee Arnoldia 26: 33-48. 1966 & CULLEN, . A as in the Old World Alyssae Hayek. Feddes Repert. 71: 218-228. 196 FerGuSsON, IAN K. i Cornaceae in the southeastern United States. Jour. Arnold Arb. 47: 106-116. 6 The genera of oleae in the southeastern United States. Jour. Arnold Arb. 47: 33-59. 1966. igh on the niacin of Cornus. Jour. Arnold Arb. 47: 100-105. 2 Brizicky, GEorGE K. Nomenclatural notes on oar fullonum and Dipsacus sativus. Tuk. Arnold Arb. 46: 362-365. 1965 ForDHAM, ALFRED J. Abnormal conifers from native pojuibtions in Massachu- setts. Int. Pl. Prop. Soc. Comb. Proc. 14: 115, 116. 1964. . Dwarf white pines from witches’ brooms. Am. Nurs. 123(1): 14, 15, 85. 1966. . White pine witches’ broom seedlings. Int. Pl. Prop. Soc. Comb. Proc. 14: 116-118. 1964. GRAHAM, SHIRLEY A. The genera of Araliaceae in the southeastern United States. Jour. Arnold Arb. 47: 126-136. 1966. GREEN, PETER S. Identification of the species and oe in the Lonicera tatarica complex. Jour. Arnold Arb. 47: 75-88. 196 . The name Jasmine. Arnoldia 25: 71-74. 1965. —. Studies in the genus Jasminum III. The species in cultivation in North America. alles 13: 137-172. 5. HartLey, THomas G. Discovery of the Massachusetts fern in Wisconsin. Rhodora 67: 399-404. 1965. Howarp, Ricuarp A. A check-list of cultivar names in Weigela. Arnoldia 25: 49-69. 1965. [Introduction] A cumulative index to the nine volumes of the Sym- bolae Antillanae seu fundamenta Florae Indiae Orientalis, edited by Ignatius Urban. 1-5, Jamaica Plain, 1965. . The Director’s Report. The Arnold Arboretum Rares the fiscal year ended June 30, 1965. Jour. Arnold Arb. 46: 465-490. . How to Se a cultivar. Quart. Newsl. Am: Asis: Gard. Arb. 64: 13-15. 1965 . Notes on some plants of Puerto Rico. Jour. Arnold Arb. 47: 137-146. 66. The registration of cultivar names, why and how. Rhode Island 66. Nurserymen’s New Ml. 25:1, 4, G6, 7; ; ; —. Sobrevivieron 999. Un analisis de experiencia de supervivencia en el cifico suroesta. 100 pp. 1962. cS bapa by Seccion de Traducciones Doce scnantsat The Woman’s Day Book yr House Plants, by J. Hersey. 7; “1965. New Yor 344 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. 47 & Brizicky, Grorce K. The translation and transliteration of cultivar names with some notes on lilacs. Quart. Newsl. Am. Assoc. Gard. Arb. 64: 15-21. 1965. — ELL, Dutcie A. The Indian Botanic ee lia and the Gardens wg the West Indies. Bull. Bot. Surv. India 7: 1 (wi CHWARTEN, LAzELLA). Abbreviations for pare he ‘publications cited in cultivar registration lists. Quart. Newsl. Am. Assoc. Gard. Arb. 64: 22-38. 1965. Hv, Sxru-Yinc. A Chinese botanist in American gardens. Gard. Jour. N.Y. Bot. Gard. 15: 196-198, 203; 249-251, 255. 1965. . The Compositae of China, Quart. Jour. Taiwan Mus. 18: 1-136. 1965. ——. The Compositae of China (II). Quart. Jour. Taiwan Mus. 18: 233- 333.. £965. . Notes on the flora of China IV. Taiwania 10: 13-62. 1964. NEVLING, Lorin I., Jr. A catalogue of the published works of Robert E. Woodson, Jr. Ann. Missouri Bot. Gard. 52: 229-233. 1965. . IOPB chromosome number reports VI. Taxon 15: 117, 118, 122, 128. 966. ee ees BernicE G. Studies in Dioscorea, I: A collection from British Honduras. Jour. Arnold Arb. 47: 147-159. 1966. pies vita LazELta & Howarp, RicHarp A. Abbreviations for titles of pub- lications cited mn cultivar registration lists. Quart. Newsl. Am. Assoc. Gard. Arb. 64: 22-38. 1965. (with RocERSON, Cxiark T., Rickett, H. W., & BECKER, HERMAN). Index to American ictanieal literature. Bull, Torrey Club 92: 312-328, 416-436, 501-520. 1965; 93: 76-96, 146-171. 1966. SUTTON, cee sails B. The herbarium introduced. Arnoldia 25: 37-40. 1965. ——_ h Carrot, EILEEN), Compilers. A cumulative index to the nine yer of the Symbolae Antillanae seu Hector Florae Indiae Occiden- talis, edited by Ignatius Urban. 272 pp. 8°. Jamaica Plain, 1965. Woop, pee E., Jr. On the identity * Drosera brevifolia. Jour. Arnold . 47: 89-99. 1966. oe DonaLp. The dependable crab apples. Horticulture 44(3): 20, 21, 42. ———. Ground covers. /n: nsec all. U. S. Dep. Agr. Yearbk. 1965. 225-228. Washington, D. C., 196 e hardiest azaleas. peer 26: 17-32. 1966. . How to screen a junkyard. Natl. Gard. eid 35-37. 1966. ———. More plant registrations. Arnoldia 26: 13-16. 1966. ———. More tree trunks. Arnoldia 26: 5-12. . Plants for screening junkyards, gravel pits and dumps. Arnoldia 25: 45-48. 1965. . Privet hedges. Flower Grower 53(4): 32, 58,59. 1966. . Screen the junkyards. Horticulture eae 24-27, 37. 1966. . Snow damage. Arnoldia 26: 1-4. 1966 . A tree for spring. Flower Grower 53( 2): 27. 1966. . Trees for American gardens [Revised]. 502 pp. illus. Macmillan. New York, 1965. I Ricuarp A. Howarp, Director hig. Uj ¥ a A i a i a a a a ali nae 1966 | THE DIRECTOR’S REPORT 345 The Board of Overseers of Harvard College Committee to Visit the Arnold Arboretum AUGUSTIN H. Parker, JR., Chairman, Boston, Massachusetts. OLIVER Wo cott, Vice-Chairman, Hamilton, Massachusetts. Mrs. RALPH BRADLEY, Canton, Massachusetts. GeorGE R. Crark, Philadelphia, Pennsylvania. Witi1amM J. Devine, Commissioner, City of Boston, Department of Parks and Recreation. Henry F. Du Pont, Winterthur, Delaware. Joun M. Foose, Jr., Philadelphia, Pennsylvania. Mirorp R. LAWRENCE, Falmouth, Massachusetts. Haroip E. Moore, Jr., Ithaca, New York Francis W. SARGENT, Boston, Massachusetts. Mrs. ELLERY SepGwick, Beverly, Massachusetts. HERMON DuNLapP SMITH, Chicago, Illinois. Mrs. HoMER N. Sweet, Weston, Massachusetts. GrorRGE TAYLOR, Kew, Surrey, England. Mrs. G. KENNARD WAKEFIELD, Milton, Massachusetts. Mrs. CHarLEs D. WEBSTER, New York, New York. RicHarp P, WuiteE, Washington, D.C. Mercer Research Fellows during at least part of fiscal year 1965-1966 WiiuiaM GENSEL, University of Rhode Island, pune Rhode Island. PaBLo RAUL LECNAME, Instituto Lillo, Tucuman, Argen SANDRA SHANNON, University of Massachusetts, Amber, Massachusetts. Car Es C. TsENG, Windham College, Putney, Verm HERMANUS PHILIPPUS VAN DER SCHIJFF, Chivetsity te Pretoria, South Africa. Francis H. Worre, University of New Hampshire, Dover, New Hamp- shire. 346 JOURNAL OF THE ARNOLD ARBORETUM [voL. 47 Staff of the Arnold Arboretum 1965-1966 § RicHARD AtpEN Howarp, Ph.D., Arnold Professor of Botany, Professor of Dendrology, and Director. IRVING WIDMER Bartey, S.D., Professor of Plant Anatomy, Emeritus. Kart Sax, S.D., Professor of Botany, Emeritus. PRISCILLA JEAN JAMES-ASHBURNER, M.A., Business Secretary. ANpbrEY I. BARANOv, Curatorial Assistant. GrorGE KONSTANTINE Brizicky, R.N.Dr., Botanist, Southeastern Flora Project.* MicHaet ANTHONY Canoso, M.S., Senior Curatorial Assistant.* HENRY DRAPER, Superintendent, Case Estates THEODORE ROBERT Duprey, Ph.D., Assistant Horticultural Taxonomist.** Tan KettH Fercuson, Ph.D., Botanist, Southeastern Flora Project.*7 ALFRED JAMES ForpHaM, Prope ator . PETER SHAW GREEN, B.Sc., Horticultural Taxonomist.+7 WILLIAM Ep Grime, B.A. Curatorial Assistant.* THomAs GorDON Hasrury Ph.D., Assistant Curator. HEMAN ARTHUR iO Menengss Horticulturist. Suiu-Y1nc Hu, Ph.D., anist. MARGARET CATHERINE pty Herbarium Secretary. Lorin Ives NEVLING, JR., Ph.D., Associate Curator and Supervisor of the Herbaria.* Dutcte Aictia PowELL, M.A., Research Fellow. Mary ELizABETH SANDERS, Ph.D., Research Associate.t 77 BERNICE Gipuz SCHUBERT, Ph.D., Associate Curator and Editor. LAZELLA SCHWARTEN, Librarian.* STEPHANNE Barry Sutton, A.B., Research Fellow. RoBERT GEROW WILLIAMS, B.S., Superintendent. CARROLL Emory Woop, Jr., Ph.D., Associate Curator. Donato Wyman, Ph.D., Horticulturist. * Appointed jointly with the Gray Herbarium. 6. +77 Resigned October 1, 1965. ee ¥ 1966 | INDEX 347 INDEX oo. 35 oo sect. aig 165 bies, aad logical Notes on Latin . Fruticulosae, 163 yearn Ml Pinus and, 313 ate di i 166- age ingsionny 16 — griffithii, — 316 —halimus, 16 haitiacoa ad Floram Jordanicae, 160 —-— argutidens, 160 11 Ayenia, 68-70 Ampacus latifolia, 216 Anatomy of gechai age (Palmae), Notes on the Veg ses road : the seca Axis and Flower a New Palm, Aristeyera rhea leerdets ogy an Anatomy of the Leaf- Geisler Cactacea The Develo chosria eh Rhizome, 248; III. Juvenile Phase, 127 (Araliaceae), A New Species fro he 3 tea Islands, Polyscias face ae in the Southeastern United States, The Genera of, 126 — suecica, 10 Aristeyera, A New Genus of Geonomoid Palms, Aristeyera spicata, Morphology and anatomy of the Inflorescence Axis and Flowers of a New Palm, 9 Aristeyera spicata (Palmae), Notes on the Vegetative Anatcm 1020 i 28 spicata, Arrow-wood, 42 sterogyne Aster ae De Atriplex, 160-165 —sect. Ayenia, 69 Bartey, I. W. Comparative Anatomy of the Leaf- bearing Cactaceae, XVI. Th Development of Water- soluble Crystals in Dehydrated Leaves of Pereskiopsis, 213 Battey, I. W. The Significance of the Reduction of Vessels in the Cactaceae, Beach- ari 297 Benthamia, 105 a fragife era Benthassaiee: 105, 110 — florida, 105 Borassus, 9 Hathincery wn: 104, 110 British Honduras, A Collection from, Studies in Dioscorea, I, 14 Brizicky, Grorce K. The nera of Sterculiaceae in the ade United States, 60 Beuixy, GEorR' . The Goodeniaceae in the Southeastern United States, 293 Butomaceae, Cactaceae, seg ont Anatomy of the Cactaceae, — subfam eit 290 = gab tin: Opuntioideae, 290 Calamus, 9 Calyptrogyne, 1, 3, o 6 Calyptronoma, 1, 3, sey sohaagoal in wad Southeastern United Sta he ir capac ae, 3-59 cape 48, 54 348 JOURNAL OF THE ARNOLD ARBORETUM [ VoL. 47 aa er ine tribe “tee Meat 51 — tri in Chamaero 5 Collection fom British Honduras, A, tudie Dioscorea, I, 147 s in Comparative Anatomy pct the Leaf- i he Develop- nek rystals in Dehydrated Leaves of Pereskiopsis, 273 ° = cf oO fu wn i} oa = =r S ta) Conostegia hotteana, 143-144 Cordia bellonis, 138 — wagne rorum, 137-138, 144, 145 Cornaceae the Southeastern United cica, 105 ornus, a on the Nomenclature of, 10 Cornus, hate 05, pet 109, 116 sect. A — subg. Benthamia, 105, 113. — subg. Kraniopsis, suis 110-112, 113 — subg. Macrocarpium, 104 — subg. Mesomora, i 112, 113 5 — Gruppe Benthamidia, 105 Cornus sarees Amblycaryum, 104 ——su t. Bothrocaryum, 1 Se ibectole. 104, 111 mum Amomum, 111 | ° 2 uo) a ° co “5 — macrophylla, 104 —mas, 103, 104 —mascula, 104 CRITCHFIELD, WitLt1AM B. Phenological otes on Latin American Pinus and 1 Crystals in Dehydrated Leaves of Pere- kiopsis dapiverse ig gah 142- —recurvata, 142 inc la ‘105 — floridum, 105 i ea a helleriana, Rediscovery of, ae Beat americana, 265 —helleriana, 262-265 lata, 159 — bartlettii, 159 y a 1966 | oo Santi 148 ulliana, 159 9 — gaumeri, 154-158, 159 — hondurensis, ie ‘154, 159 — pilosiuscula, 1 Peal 7, 158, 159 Dioscoreaceae, 147 Diplotaxis, 167 — sect. Catocarpum, 167 -— harra, —kerakensis, 163, 164, 166-167 Director’s Report, The, 323 ogwood, Dogwood Fam Drosera eer ae the Identity of, 89 90; 9 a: brevitalis, 89, 90, 92-99 —capillaris, 90, 91- 92, 93, 96, 97, 98 — chiapasensis, 90, 97 = — eecaditer 0, —tenella, 92 ne - Elde Esphoria ges ay 170 7, ae Evodin es es atica, 216 he Nyssaceae in the Southeastern United States, 117 Fagara, 173 INDEX 349 Fagara ailanthoides, 188 —ani 93 Ww —laxifoliolata, 177 —nitida, 180 — oblongifolia, 181 — odorata, 1 — ovalifolia, 207 —rhetsoides, 185 — scandens, —serrulata, 216 —torva, 1 —triphylla, 216 — varians, 207 warburgii, 181 Fercuson, I. K. Notes on the Nomen- clature of Cornus, 100 Fer fai) Seal <8 6-68 Floram Rees ‘cugpranie ad, 160 Flcwers of a New , Aristeyera spi- cata, Morphology has ‘Anatomy of the Inflorescence Axis an Genera of Araliaceae in the Southeastern United States, The, 126 350 JOURNAL OF THE ARNOLD ARBORETUM Genera of Caprifoliaceae in the South- 3 enus Linociera sear eas in New Genus of Geonomoi Palms, Aristeyera, New, Geonoma, : 3,26, 27 ae Palms, Aristeyera, A New Genus of, 1 Ginseng Family, 126 Goodenia, 2 Goodenia Pt ily, Goodeniaceae in es ae United States, The, 293 Goodeniaceae, oe he Gen of Atalinceas in ihe PA United GrEEN, P. S. _ Identification of the Species and Hybrids in the Lonicera tatarica Complex, 75 REEN, P. S. The Genus Linociera (Oleaceae) in New Caledonia, 319 GREEN, Peter S., and Liry Ripennenc. af Karyologica Survey of Lonicera, I, Pe Peter §., and Orro T. Sorsric Sciaphila dolichostyla (Triuridaceae), Gum, 121 Hartitey, THomas G. A Revision of the Malesian “ener of Zanthoxylum (oa Rage Wy | su nie, 56 Hon neysuckle ——_: 33 Horse-gen Howarp, seas A. Notes on Some Plants of ee Rico, 137 A Howarp, The Director’s Report, Hybrids nicera tatarica Com- _— Idepiificatton of the Species and, tiie 269 a of the Species and Hybrids n the Lonicera tatarica Complex, 75 fie of Drosera brevifolia, esr the, 89 Inflorescence Axis and Flowers of a New , Aristeyera ~— Mord and Anatonay of the, [voL. 47 Jordanicae, Additamenta ad Floram, 160 venile Phase, Anatomy of the Palm Rhapis excelsa, ITI, 301 ne ear soloed ee of Lonicera, I., A., ene AKI. Additamenta ad Floram oa 160 Latin American Pinus and Abies, Pheno- logical Notes on, 313 Leaf-bearing Cactaceae, Comparative Anatomy of the, XVI. The Develop- ment of Water-soluble Crystals in Dehydrated Leaves of aaaacene 273 Leuc cojum capitulatum, Leycesteria, 3 Linociera {Ghincean’ in New Caledonia, The Genus, 319 Linociera, 319-322 _ brachystachys, a — A sonal Survey of, I, 222 Lonicera — Complex, Identification of the Species and Hybrids in the, 75 Lonicera, 34, a 38, oe ig 222-247 se — sabe. einen ay 230 Py 20s — — subsect. Purpurascentes, 230 — — subsect. Pyrenaicae, 234 — — subsect. Rhodanthae, 235 1966 | INDEX 351 Lonicera subsect. Tataricae, 236 Lonicera leycesterioides, 232 —— subsect. Vesicariae, 232 — flava, 246 — alseuosmoides, 244 —> < hirsuta, : — — “coriacea”, 244 ——.< prolifera, 227 —altaica, 226 ——x aengee cod —— > pallasii, 226 — floribunda, 82, — altmannii, 233 —— > num nner 226 —alpigena, 234 — fragrantissima, 233 1 ——w—nana, 234 ——. standishii, 227 —— ledebourii, 227 — gibbiflora, 227 — X americana, 227, 228, 246 —glauca X hirsuta, 228 —— X sempervirens, 228 — glehnii, 2 — X amoena, 80, 81, 82, 227, 238 — gymnochlamydea, 232 —w— arnoldiana, 80, 238 — ™ heckrottii, 228 — —— alba, — > helvetica, 22 — — — rosea, — henryi —angustifolum phos 228 — — subcoriacea, 245 % — X bella, 80-82, 227, — hildebrandiana, 245 ——w—albida, 82 — hirsuta * prolifera, 227 — — — atrorosea, 82, 240 —— > sempervirens, 227 ——— candida, 82, 24 — hispida, 2 ——— ieee 82 — — bracteat im Ro ee POU , 241 as uinquelocularis, 226 at ruprechtians 83, 227 Be cea 4 x ; bi more’, 82, 241 —implexa, 24 withers —involucrata, 234 — X brownii, 227 is, 234 Seale, "26, 231 — — — serotina, 234 | edulis, 231 — >< italica, 227, 228 7 — —— eraciliflora, 231 — japonica, 245 — canadensis, 230 —koehneana, 243 —captiiolium, 2 — korolkowii, 81, 82, 238 ——-x etrusca, 227, 228 — — aurora, 82, 238 ee oe Rae, 207 — — floribunda, 82, 238 _ periclymenum, 227 —— zabelii, 82 i. Seeearpe 24 —— tatarica, 80, 227 — eheyaantha, ho — ledebourii, —w— longipes, 24 ——— villo maximowiczii, gets roprectina a a — tes microphy ie alepieiys, 14 % 5 mints 82, 227, 241 ——X q esata 227 — — ‘Valencia’, 241 EA a gg 2 — modesta, 2 — dioica, 245 — lushanensis, 2 as Se awn — morrowii, 82-83, 84, 240, 241 —— xX hirsuta, 228 ——— xanthocarpa, 82-83, 240 ——. prolifera, 228 cana. ruprechtiana, 84, 22 — discolor, 235 —— > tatarica, 80 ol — epsomiensis, 228 —— X xylosteoides, 82, 227 —etr pad —X muendeniensis, S34 227, 241 — — — xanthocarpa, Ee eee eon muscaviensis, 84, 227, 241 — ferdinandii, 232 —myrtilloides, 228, 23 —— induta, 232 —myrtillus, 230 352 JOURNAL OF THE ARNOLD ARBORETUM Lonicera myrtillus depressa, 230 —nervosa, 235 ee 235 naar poe ea 228 ——®> ta anes —x sylestcum, 226 uae —x pase erat 221;,259 3 — — — carneo- crete. 85, 239 — — — gilva a es grandiflora, 85 ——W— ochroleuca, 85 ———-rosea, 8 ’ —nummulariifolia > quinquelocularis, 22 — orie entalis, —— longifolia, 235 — X _ pseudochrysantha, 227, 243 —X purpusli, 227, 233 — pyrenaica, 2 — Eesinent ead 228, 244 —_——— bigger, — rupic 230 _ abo 85, Ps 239 —— calvescens, 8 — X salicifolia, 85-86, 227, 240 — X sargentii, 227 _— — vey — sempervire 245 ——x pec ae 28 — — ‘Dreer’s Everblooming’ Rees) 245 — syringantha, 2 — tatarica, 81, 86-88, 236, 238, 239 — nana, geen 236 — — — elegans, 236 [VoL. 47 Lonicera tatarica ers 236 —— xylosteum, 88, 227 —— ‘Arnold Red’, 87, 237 —-——*Hack's Red’, 87. 237 — — ‘Lutea’, —— ‘Morden oe 812237 — ‘Nana Beer once Red’, 87, 237 — tatsiensis, 234 —xt ellmanniana, 228, 245 es, 230 —trichosantha, 2 ——— acutiuscula, 244 — vesicaria, 232 — X vilmorinii, 227, 243 — webbiana, 234 — X xylosteoides, 88, 2. Vy. 228y1209 — xylosteum, 88, 227, , 242 ———com Li Loranthus guadalupensis, 139 — portoricensis, 140 Macrocarpium, 1 Malesian Species a eg (Ruta- ceae), A Revision of t Mapouria nee 140 — ena cope pau, 216 216 RE, Haroitp E., Jr., — A New Genus of es Palms, — EO 1966] Morphology and Anatomy of the In- florescence Axis and Flowers of a New Palm, Aristeyera spicata, 9 Nardostachys, 34 NeEviinc, Lorin I., Jr., and Roy Woop- BURY. Rediscovery of Daphnopsis helleriana, 262 New Caledonia, (Oleaceae) in, 319 Nomenclature at Cornus, Notes on the, The Genus Linociera Notelaea a Pac eaios 319 American Pinus and Abies, Phenological, 313 Notes on Some Plants of Puerto Rico, 137 Notes on the Nomenclature of Cornus, 100 Notes on the Vegetative Anatomy of ae spicata (Palmae), 23 Nothopanax, 27 Nugae sy acum ‘eit 180 Nyssa, 117, — ogeche, fay Nyssaceae in the Southeastern United States, The, 117 Nyssaceae, 117-125 (Oleaceae) in New Caledonia, The Genus Linociera, 319 ea, 104 —instolonea, 104 Pachydiscus, 34 Palm, Aristeyera and Anatomy of and Flowers of a Palm Rhapis deg Anatomy of the, R spicata, Morphology . Inflorescence Axis i, ev ome, 248; _ III. Juvenile “hionage 30 palma, (almae) : Vegetative No Anat > of pone spicata, 23 Palmae — hes Arecoideae, 1 bs cathe — meae, 2 palmito, INDEX $53 Palms, Aristeyera, A New Genus of Pereskia, ais 289 — bleo — aint. 289 — joarei! fei Ran i se evelopment of Wate soluble Crystals in Dehydrated a 0 omparative atomy of the eaf-bearing Cactaceae, XVI, 27 Pereskiopsis, 273, 282, 283, 289 rnin ike on Latin American Abies, 313 Pholidostac hys, 1, 3 27 Picrasma he vanica Pinus bies, Phenotosia Notes on a patula, 3 ge = Chasteuenien 316 —— oaxacana, 317 —radiata, 313 — tenuifolia, 317 —teocote, 317 Piper pinnatum Plants of Puerto ie Notes on Some, 1 7 Polyscias _ verticillata alga aggre A ew Species fro the Solomon Islands, 270 Polyscias, 270-272 —corticata, 270 —pinnata, 270 354 JOURNAL OF THE ARNOLD ARBORETUM Polyscias scutellaria, 270 — subcapitata, 270 e Puerto Rico, Notes on Some Plants of, 137 Quiabentia, 289 Rediscovery of Daphnopsis helleriana, 262 Reinhardtia, 27 Revision of he Malesian a fasny of Zanthoxylum (Rutaceae), A, 171 Rhapis excelsa, Anatomy of the P Palm, II. Rhizome, 248; ITI. Saget oh 301 eel excelsa, shila ieee on of Vessels in vie eke by ea ar the, 2 nhieu me, Anatomy of the Palm Rhapis excelsa, 13 OF = RUDENBERG, L d Perer S. Green = hpi fi et ee of Lonicera, i. a, A Revision of — Malesian Species of Zanthoxylum, 1 Rutaceae, 171 Sabal, Shchacie: 34, - 37, 38—41 — canadensis Sarcocarpaca, 3H Sarlina, 319 SCHUBERT, BERNICE G., Studies corea, I: A Collection from Hondur 4 Sciaphila dolichostyla (Triuridaceae), 266 Sciaphila, — sect. Becton 269 — sect. Oliganthera, 268-269 in Dios- British —andajensis, 268 daichosvi 266-269 — japonica, 268 [voL. 47 Sciaphila neo-caledonica, 267 h 69 — thaidanica, 268 nsis, 268 perma, 27 Sienifcance a the * ange of Vessels oo The, 2 5 4 be Sorsric, Orro T., and Peter S. on ae dolichostyla "(Trluridageae), Aba woodburyi, 138-139, 144, 146 Solomon Islands, Polyscias verticillata A acca A New Species from the, eis Peis States, The Corna- ceae i emi ier ane Un ited een The Genera of Araliaceae in t 126 Southeastern United States, The Genera of Caprifoliaceae in the, 3 pie ter hath United Aron The Genera of Sterculiaceae in the, 6 Southeastern Raise States, The Goode- niaceae in Southeastern United States, The Nyssa- ceae in pipes es — Ea ds in the Lonicera tatarica Complex, Identification of the, 75 Species from the Solomon Island aged verticillata (Araliaceae), A New, 2 Species c rm (Rutaceae), A ion of the Malesian, 171 erculia, terculia Fami ily, 6 STONE N C. Polyscias verticil- lata. moue aay. A New Species from the Solomon Islands, 270 Studies in Dioscorea, I: A Collection i 14 Survey of Lonicera, I., A Karyological, 222 Swida, 1 Symphercarps, 35, 38, 49-51 . Meridionales — subg. Eusymphoricarpos, 49 TT eee a ae 1966 | INDEX 355 Taenianthera, 1, 3 Toddalia, Tomutnson, P. B. Notes on the Vegeta- tive Anatomy of Aristeyera ats (Palmae), 23 Tomutnson, P. B. fue Martin H. Zim- MERMANN. of the Palm Rhapis excelsa, 1 Rhizome, 248; III. hte aaa Tri pero Siaphila dolichostyla, 266 Triuridaceae Tupelo, 1 Tupelo ee 117 HL, NATAL W. Morphology and Anatomy of "the Inflorescence Axis and Flowers of a New Palm, Aristeyera spicata, 9 Uragoga grosourdieana, 140 — parasitica, Waltheria, 72-7 Water-soluble pees in Dehydrated Leaf-bearing — BV Welfia, 1, 2, Weigela, 34, 35. Woop, C:_5.,- Jr. go the Identity of sana brevifolia, Woopsury, Roy, and ogee I. NEVLING, R. Rediscovery of Daphnopsis hel- leriana, 262 Vegetative Anatomy of Aristeyera spicata (Palmae), Notes on the Vessels in the Cactaceae, iy Significance —sect. Pseudotinus, 42-43 Viscoides as, 139 Xanthoxylon, 173" 190 — planispinum, 9 14-215 Zanthoxylum (Rutaceae), A Revision of the Malesian Species of, Zanthoxylum, 171, 173-221 — aca nee 209-211 —accedens, 216 — me 187, 188-190 —w—inerme, 18 —alatum, 21 ——planispinum, 211 —armatum, 2 181 “= ceien sagen 204-205 — cyrtorhachium, — diabolicum, 185 —dimorphophyllum, 209 — — dimin ge 204 — dissitiodes, 180 — div writotonn, 190 — Jaxifoliolatum, iy — leiorhachium — Pa 173, ae 199 216 — minahassae, 197 356 JOURNAL OF THE ARNOLD ARBORETUM ie a er 187 — montanu = mpriacanthum, lias 0-185 ge — ovalifolium, 207-209 7 20 Bes a eg 180 — parviflorum, 199-200 —planispinum, 211, 214 — pluviatile, 201-204 — retroflexum, 195-196 tsa, 19 —robiginosum, 209 — roxburghianum, 217 —rumphianum, 217 ap adi ea 181 77-1 — yunnanense, 177 Zilla spinosa, 170 ——emarginata, 166, 167-170 ZIMMERMANN, MartIN 15 [voL. 47 a -P: ToMLinson. Anatomy of the Palm Rhapis excelsa, oa oo 248; III. Juvenile Phase, PUBLICATIONS FOR SALE AT THE ARNOLD ARBORETUM OF HARVARD UNIVERSITY CAMBRIDGE, MASSACHUSETTS JOURNAL OF THE ARNOLD ARBORETUM. Published quarterly by the eas Arboretum of Harvard eee eaten ae one $10.00 per year. Volumes I-XX, reprinted, and back issues of volum XLVI available from the sun Reprint Corporation, 16 Bast 46th Street, Nex ang N. Yi 10017. ARNOLDIA. A continuation of the Bulletin of Popular Information of the Arnold Arboretum. About 12 numbers per year. Subscription price $2.00. Price of single numbers $0.25 each BIBLIOGRAPHY OF CULTIVATED TREES AND SHRUBS HARDY IN THE COOLER TEMPERATE REGIONS OF THE a ee HEMISPHERE. By Atrrep Renper. xl + 825 pp. 4°. Cam- bridge, Mass., Price $20.00 INDEX paveeeawne The plant names published by C. S. Rafinesque with reductions, and a Say Tg of his methods, objectives, and attainments. By Ermer D, Merri. vi + 296 pp. 4°. Baltimore, Md., 1949. Price $10.00 BOTANICAL EXPLORATION OF THE gsc ppoigge cent ace he 1790-1850. By Susan Detano McKetvey. xli + 1 Large 8 Jamaica Plain, Mass., 1955. ies $25.00 THE GENUS PINUS. By Georce Russert SHaw. 96 pp. 39 pl. 4°. Cam- bridge, Mass., 1914. Photoprint copy, 1958. Price $15.00 THE PINES OF MEXICO. By Gerorce Russert SHaw. 31 PP. ted “ — bridge, Mass., 1909. Photoprint copy, 1961. 15.00 AN ACCOUNT OF THE BOTANIC GARDEN IN THE ces OF ST. VINCENT, FROM ITS FIRST ESTABLISHMENT TO THE PRESENT TIME, By Lanspown Guitpinc. 47 pp., 3 pl., 1 plan. 4°. Glasgow, 1825. Photoprint, 1964. Price $5.00 A CUMULATIVE INDEX TO THE NINE VOLUMES OF THE SYMBOLAE ANTILLANAE SEU FUNDAMENTA FLORAE INDIAE OCCIDENTALIS, EDITED BY I. URBAN. Compiled and prepared by EILEEN Sutron. Introduction by R. A. Howarp. 272 pp. 8°. — Offset. Jamaica Plain, iii, 1965. Price $4.50 Now Available A Cumulative Index to the Nine Volumes of the SYMBOLAE ANTILLANAE seu fundamenta Florae Indiae Occidentalis edited by IGNATIUS URBAN Compiled and prepared by EILEEN CARROLL and STEPHANNE SUTTON Introduction by RICHARD A. HOWARD Unbound, 272 pp., 8° Price $4.50 Published by the Arnold Arboretum of Harvard University Please address inquiries and orders to the Lrsrarian, ARNOLD ARBORETUM, 22 Divinity AveNvuE, Camprince, MAssACHUsETTS 02 138, U.S. A.