JOURNAL

OF THE ARNOLD ARBORETUM HARVARD UNIVERSITY C. E. WOOD, JR.

EDITOR

L. I. NEVLING, JR. LAZELLA SCHWARTEN ASST, EDITOR CIRCULATION

VOLUME XLII

CAMBRIDGE, MASS. 1962

Reprinted with the permission of the Arnold Arboretum of Harvard University

KRAUS REPRINT CORPORATION N or

1968

DATES OF ISSUE No. 1 (pp. 1-108) issued January 15, 1962. No. 2 (pp. 109-221) issued April 12, 1962. No. 3 (pp. 222-350) issued July 10, 1962. No. 4 (pp. 851-468) issued October 15, 1962.

Printed in U.S.A.

TABLE OF CONTENTS

THE GENERA OF RUTACEAE IN THE SOUTHEASTERN UNITED STATES. By George K. Brizicky THE LEAF BASE IN PALMS, ae MorpHoLocy AND MECHANICAL BroLocy. By P. B. Tomlin on BOTANICAL AND OTHER onentene ON REponpA, THE WEST InpiEs. By Richard A. Howard A Taxonomic REVISION oF Popocarpus, XIII. Section Pory- PODIOPSIS IN THE SouTH Paciric. By Netta HE. Gray ............ TAXONOMIC AND NOMENCLATURAL NOTES ON ZANTHOXYLUM AND Gtycosmis (RuTACEAE). By George K. Brizicky ................ THE ae PIFICATION OF DiospyROS EBENUM AND DIOSPYROS EBEN- r. By Richard A. Howard and Tycho Norlinda ............ a ON THE RELATIONSHIPS OF PINUS MERKUSII.” By Nicholas T. Mirov STUDIES IN THE GENUS JASMINUM, IJ. THE Species rrRoM NEw ALEDONIA AND THE Loyatty IsLanps. By P. S. Green. ...... A CyroLocicaL Srupy OF THE GENUS ViBURNUM. By Donald R. Egolf THE GENERA OF SIMAROUBACEAE AND BURSERACEAE IN THE SourTH- EASTERN UnITep States. By George K. Brizicky .........0.0.05. CoMPARATIVE ANATOMY OF THE LEAF-BEARING CACTACEAE, IV. Tue Fusirorm INITIALS OF THE CAMBIUM AND THE ForM AND STRUCTURE OF THEIR Derivatives. By J. W. Bailey and Laht M. Srivastava ON THE ORIGIN OF CARAGANA SINICA. By Raymond J. Moore .... Suruyincuua, A New GENUS OF SCROPHULARIACEAB FROM CHINA By J. Pact PHENOLOGY or Tropical Pines. By Nicholas T. Mirov .............. ON THE Status oF Psiraea (THYMELAEACEAE). By Lorin I. Nevl-

Ulta te JosErPH Horace Fautt, 1870-1961. With portrait. By Anna F. au COMPARATIVE ANATOMY OF THE LEAF-BEARING CACTACEAE, V. THE SeconDARY PuHuoEM. By Lalit M. Srivastava and EW: Bailey VOLCANISM AND VEGETATION IN THE LESSER ANTILLES. By Rich- ard A. Howard Tue GENERA OF PAPAVERACEAE AND FUMARIACEAE Pi THE SOUTH- EASTERN UNItTep States. By Wallace R. Ernst ........0.000... Note oN DapHNOPSIS CRASSIFOLIA fe ON By Lorin I. Nevling, Jr. Two New Asiatic PANDANACEAE. By Benjamin C. Stone ............

EeRNEsT JESSE PauMer, 1875-1962. With portrait. By Clarence BE. Kobuski THE GENERA OF ANACARDIACEAE IN THE SOUTHEASTERN UNITED States. By George K. Brizicky COMPARATIVE ANATOMY OF THE LEAF-BEARING Cacracear, VI. THE XYLEM OF PERESKIA SACHAROSA AND PERESKIA ACULEATA,. By I. W. Bailey SOME GUTTIFERAE OF THE LESSER ANTILLES. By Richard A. Howard DaTEs OF PUBLICATION OF THE JOURNAL LinnakaA. By Robert (’. Foster A MonocrapH OF THE GENUS PLATYDESMA (RUTACEAE). By Benjamin C. Stone THE THYMELAFACEAE IN THE SOUTHEASTERN UNITED States. By Lorin I. Nevling, Jr. THE LEITNERIACEAE IN THE SOUTHEASTERN UNITED States. By Rk. B. Channell and C. E. Wood, Jr. Tue Direcror’s Report BIBLIOGRAPHY OF THE PUBLISHED WRITINGS OF THE STAFF AND STUDENTS, JuLY 1, 1961—JUNE 30, 1962 STAFF OF THE ARNOLD ARBORETUM, 1961-1962 INDEX TO VoLUME XLIII

JOURNAL

OF THE

ARNOLD ARBORETUM

VoL. XLIII JANUARY 1962 NUMBER l

THE GENERA OF RUTACEAE IN THE SOUTHEASTERN UNITED STATES *

Georce K. Brizicky

RUTACEAE Jussieu, Gen. Pl. 296. 1789. (RUE FAMILY)

Armed or unarmed trees or shrubs [sometimes scandent or xeromor- phic], rarely herbs. Leaves alternate or more rarely opposite, simple or compound, usually glandular-punctate at least at the margin, exstipulate; petioles sometimes winged. Flowers bisexual and/or unisexual, the plants monoecious, dioecious, or polygamous, regular [rarely irregular], usually 3—5-merous, the insertion hypogynous, solitary and axillary or in various axillary or terminal, often cymose, inflorescences. Sepals distinct or con- nate, very rarely wanting, often glandular-dotted, usually imbricate in bud. Petals distinct [rarely connate or wanting], often glandular-dotted, imbricate or valvate. Stamens as many as the petals and in 1 series (haplostemonous) or twice as many to more numerous and in 2 series (diplo- or obdiplostemonous), those of the outer series often shorter than those of the inner [or occasionally reduced to staminodes] ; filaments dis- tinct or -- connate, often conspicuously dilated [or rarely appendaged | at base; anthers versatile, introrse, 2[4]-locular at anthesis, often gland- tipped, longitudinally dehiscent. Intrastaminal nectariferous disc ring-, cup-, or cushion-like, rarely wanting. Gynoecium of (1)2-5(-several),

the Arnold Arboretum and the Gray Herbarium of Harvard University which has been made possible through the support of George R. Cooley and the National Science Foundation. 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 40-42 (1959-1961). It should be repeated that the area covered by this work is bounded by and includes North Carolina, Tennessee, Arkansas, and I ouisiana. The descriptions are based primarily on the plants of this area, with any supplementary material in brackets. References which the author has not seen are marked by an asterisk.

The author is indebted to Dr. Carroll E. Wood, Jr., for his criticism and valuable suggestions, and to Mrs. Gordon W. Dillon, for her careful help in the preparation of the manuscript. The illustration has been made by Dorothy H. Marsh under the direction of C. E. Wood, Jr.

? JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIII

sessile or stipitate, distinct or incompletely to completely connate carpels; stigmas simple or lobed; styles basal, + lateral or terminal, distinct, con- nivent or connate; ovaries 1-locular or 2—5(—several) locular, with axile placentae [very rarely 1-locular with parietal placentae|; ovules usually anatropous and epitropous [rarely apotropous|, with 2 integuments and a thick nucellus, 1 or 2—many in each locule of the ovary. Fruit of (1)2— 5(-several) follicles or drupes, or a capsule, berry, drupe, samara, or schizocarp; pericarp often glandular-pitted to -verruculose. Seeds with or without endosperm, sessile or funiculate, 1 or 2-several in a locule; embryo relatively large, straight or curved, with plano-convex, some- times convolute [rarely plicate] cotyledons and a superior radicle. Typr GENUS: Ruta L

A family of about 150 genera and 1600 species, widely distributed in tropical and temperate regions, most abundant in tropical America, South Africa, and Australia, extremely scarce in Europe. Eight genera (five naturalized) in three subfamilies are represented in our area.

The presence of secretory cavities (or at least inner multicellular glands) containing an aromatic volatile oil in stems (in cortex, rarely in phloem), leaves, floral parts, and pericarp of fruits is a characteristic feature of the family, distinguishing Rutaceae from their morphologically very simi- lar allies, especially Simaroubaceae and Meliaceae. Rutaceae are also closely related to Zygophyllaceae, Cneoraceae, and Burseraceae.

The family, in general, seems to be entomophilous, insects being at- tracted by the strong smell of the flowers and/or by usually abundant nectar, sometimes also by showy corollas (e.g., Citrus spp.). Ornithophil- ly has also been presumed for some genera. Cross-pollination seems to be the rule. In some genera self-pollination is prevented by dicliny, or (in monoclinous genera) by proterandry (e.g., Ruta, Ravenia, Barosma) and/or the position of the stigma in regard to the anthers (e.g., Triphasia, Dictamnus). In some genera, however, both cross- and self-pollination seem equally possible and effective (e.g., Choisya, Skimmia, Murraya, Poncirus, Citrus). Cleistogamy has been recorded in the New Zealand alae simplex A. Cunn. Nucellar embryony (apomixis) has been shown in several genera (e.g., Zanthoxylum, Esenbeckia, Ptelea(?), Tri- phasia, Aegle, Murraya, Poncirus, Citrus), but probably is of even wider distribution in the family. Zygotic (gametic) polyembryony seems to be very rare, having been recorded or presumed in only a few cases.

Chromosome counts have been made for about 50 genera and 150 species. On the basis of these counts, nine appears to be a basic number for the family. (Banerji [1954], however, suggested three as a basic number and nine as a result of secondary polyploidy for Citrus grandis (L.) Osbeck.) The family, in general, appears to be + euploid, except the Australian aneuploid tribe Boronieae (which has chromosome numbers based on 7, 8, 9, 11, 13, 17, 19) and apparently some genera from other tribes. Poly- ploidy is widely distributed in Rutaceae and seems to be of importance in the evolutionary development of the family.

1962 | BRIZICKY, GENERA OF RUTACEAE 3

The family is of economic importance, notably for a number of im- portant fruits (Citrus spp., g.v.), timbers, aromatic oils (e.g., Citrus spp., Ruta graveolens L.), various products of medicinal value, and ornamentals. REFERENCES:

CHAKRAVARTBY, R. 5S. Polyembry ony in Murraya Koenigii Spreng. Curr. Sci. Bangalore 3: 361, 362. 1935.* [Nucellar embryony. |

———. Nucellar polyembryony in the Rutaceae. /bid. 5: 202, 203. 1936.* | Murraya exotica L. (= M. paniculata (L.) Jack) and Aegle Marmelos (i) Corral

Desat, S. Cytology of Rutaceae and Simarubaceae. Cytologia 25: 28-35. 1960. [Includes chromosome counts for 8 genera and 10 spp. of Rutaceae; Ptelea trifoliata L., 2n = 42; Zanthoxylum americanum L., 2n = :36.]

ENGLER, A. Rutaceae. Nat. Pflanzenfam. III. 4: 95-201. 15896

—. Rutaceae. Nat. Pflanzenfam. ed. 2. 19a: 187-359, 458, “459. 1931.

———. Studien tiber die Verwandtschaftsverhaltnisse der Rutaceae, Simaru- paces und Burseraceae nebst Beitragen zur Anatomie ae Systematik dieser Familien. Abh. Naturf. Ges. Halle 13: 111-158. pls. 12, 13. 1874.

GALLeT, F. Développement et structure anatomique du tégument Te des Rutacées. Thesis, 66 pp. Univ. Paris, Ecole Supér. Pharm.

Harti, D. Struktur und Herkunft des Endokarps der Rutaceen. Beitr. Biol. Pfi. 34: 35-49. 195

Die eee a nnees des Endokarps der Simaroubaceen, Rutaceen Leguminosen. /bid. 34: 452-455. 1958

pace E. Osservazioni sulla struttura dell’ovulo e sulla cariologia di Calo- dendron capense Thunb. e Pilocarpus pennatifolius Lem. Ann. Bot. Roma 24: 438-448. pls. 19-21. 1954. [Includes brief survey of chromosome numbers in Rutaceae. |

Jourr, B. M., & M. R. Anvya. A contribution to the floral morphology and embryology of Aegle Marmelos Correa. Phytomorphology 7: 10-24. 1957. | Anthers 4- locular at anthesis; nucellar embryony the rule. |

Mauritzon, J. Uber die Embryologie der Familie Rutaceae. Sv. Bot. Tidskr. 29: B1902347. 1935.

Moore, J. A. Floral anatomy and phylogeny in the Rutaceae. New Phytol.

: 318-322. 1936. [Phylogenetic considerations based on Saunders’ data of floral anatomy. |

Necopr, G. Lineamenti sulla cariologia delle Rutaceae e delle ewes. Avo Bot. Forli 13: 93-102. 1937. [Includes Ruta graveolens, 2n Ptelea trifoliata, 2n = 36; Citrus Limon and C, simensis, 2n = 18.

Penzic, O. Studi botanici sugli agrumi e sulle Sh affine. Ann. Agr. Roma 1887: i-vi, 1-590; 58 pls. [atlas (folio) ].

Recorp, S. J., & R. W. Hess. American woods ae as family Rutaceae. Trop. Woods 64: 1-28. 1940.

SaunpeERrS, E. R. On carpel polymorphism. VI. Ann. Bot. 48: 643-692. 1934. [Rutaceae, 643-673. |

Scuuize, H. Beitrage zur Blattanatomie der Rutaceen. Beih. Bot. Centralbl. 12: 55-98. pls. 1, 2. 1902.

SMITH-WHITE, 8. Chromosome numbers in the Boronieae (Rutaceae) and their bearing on the evolutionary development of the tribe in the Australian flora. Austral. Jour. Bot. 2: 287-303. 1954. [Basic numbers 7-19; 69 spp. in 11 genera; aneuploidy and polyploidy. |

4 JOURNAL OF THE ARNOLD ARBORETUM [ VOL, XLII

SWINGLE, W. T. A new taxonomic arrangement of the orange subfamily, Auran- fo dene: Jour. Wash. Acad. Sci. 28: 530-533. 1938.

. The botany of Citrus and its wild relatives of the orange subfamily (family Rutaceae, subfamily Aurantioideae). Chap. IV, pp. 129-474. In: Wesser, H. J., & L. D. BatcHetor, The Citrus Industry. Vol. I. His- tory, Botany, and Breeding. xx + 1028 pp., front. Univ. Calif. Press, Berkeley & Los Angeles. 1943.

TANAKA, T. The taxonomy and nomenclature of Rutaceae-Aurantioideae. Blumea 2: 101-110. 1936.

Titison, A. H., & R. Bamrorp. The floral anatomy of the Aurantioideae. Am. Jour. Bot. 25: 780-793. 1938. [Vascular system does not offer a definite basis for division into the tribes and subtribes; the disc appears to repre- sent a third whorl of vestigial stamens.

Urzan, I. Zur Biologie und Morphologie der Rutaceen. Jahrb. Bot. Gart. Ber- lin 2: 366-403. pl. 13. 1883.

Witson, P. Rutaceae. N. Am. Fl. 25: 173-224. 1911.

KEY TO THE GENERA OF RUTACEAE

neral characters: leaves usually pinnate to 3-1-foliolate, alternate, rarel ee leaflets a least at margin), perianth parts, and fruits glandular. often pellucid-punctat

Plant a perennial herb or subshrub; leaves 2- to 3-pinnately divided; flowers

greenish yellow, bisexual, 4—5-merous with 8-10 stamens; fruits 5-lobed,

WAY SOCCEO CAPSULES. 6 bores oe hi-wd sx PAE RS wie base Peeing 1. Ruta.

. Plants woody, shrubs or trees.

B. Flowers unisexual or uni- and bisexual, small, yellowish or yellowish white, 3-5-merous, sometimes without calyx; stamens as many as petals; Soar apo- or syncarpous, (1)2-5-carpellate; fruits dry; leaves alternate.

>

i Plants dioecious or monoecious, often prickly; leaves pinnate, usually

5 or more foliolate; flowers unisexual, with or without sepals; gynoe-

cium apocarpous, (1)2—S-carpellate; fruits dehiscent, 1-seeded fol-

WGN Gas yoc4 sees wie ae, us Bares one Rha eeee oe ees 2. Zanthoxylum,.

C. Plants polygamous, unarmed; leaves usually 3-foliolate; flowers uni and bisexual; gynoecium syncarpous, 2-loculate; fruit a flat, inde- hiscent, 2-locular, 2-winged samara. .................... 3. Ptelea,

B. Flowers usually bisexdal small to large, usually white, 3- 5- -merous; stamens twice as many as petals, or more numerous; gynoecium syn-

carpous, 2—18-locular, rarely l-carpellate; fruits fleshy drupes or arene leaves alternate or opposite.

D. Stamens twice as many as petals; flowers relatively small; ovary 1—5-

loculate; fruits small, pulp without pulp-vesicles; petioles not winged.

E. Ovary 1-carpellate; fruit a 1-seeded drupe; leaves opposite, 3-

foliolate; flowers usually 4-merous, paniculate; era: aro-

matic, resinous shrubs or trees. .................... 4. Amyris.

E. Ovary 2-5-carpellate; fruit a “ 3-seeded berry; leaves alternate.

F. Flowers (4)5-merous, in short, spikelike axillary panicles; berry

subglobular, sometimes depressed at eke and slightly oblique,

white to pink, 1-3-seeded, edible; unarmed shrubs or trees

with pinnate, usually 1-3(5)-foliolate piers and large leaflets.

gee ehh dete ete ee esau tas 2 NG ane baht State 5. Glycosmis.

1962] BRIZICKY, GENERA OF RUTACEAE 5

F. Flowers usually 3-merous, solitary or in 2’s or 3’s in the leaf axils; berry subglobular, sometimes apiculate, reddish orange to crimson, 1-3-seeded, insipid; shrubs with solitary or paired, axillary spines; leaves 3-foliolate, leaflets cael: Leena cerrare te tcc eeres

eh en ee NERD RRS eyecare d or fo caea 6. Triphasia.

D. Stamens numerous (20-60); flowers relatively large; perianth usually

5-merous; ovary 6-18-loculate; fruits (hesperidia) greenish yellow,

yellow, orange to reddish orange, large, pulp formed by pulp vesicles; usually thorny shrubs or trees, mostly with winged petioles.

G. Flowers on the previous year’s branchlets; stamens distinct; ovary 6—-8-loculate; fruits relatively large, pubescent, pulp with very sour and acrid juice; leaves 3-foliolate, deciduous. ... 7. Poncirus.

_ Flowers on new branchlets; stamens polyadelphous; ovary 8~-18- loculate; fruits large, glabrous, pulp with sweet or sour not acrid juice; leaves 1-foliolate, persistent. ............ 8. Citrus.

7p)

Subfam. RUTOIDEAE Engler 1. Ruta Linnaeus, Sp. Pl. 1: 383. 1753; Gen. Pl. ed. 5. 180. 1754.

Heavy-scented perennial herbs, subshrubs [or shrubs]. Leaves alter- nate, usually glandular-punctate, compound [or simple], odd-pinnate to -bipinnate, with pinnae or pinnules respectively deeply cut (divided) into obovate-cuneate to oblanceolate or oblong segments. Flowers bisexual, 4- or 5-merous, in terminal, panicled cymes with simple or 3-fid bracts. Sepals + connate at base, persistent. Petals yellow or greenish, glandular- punctate, spatulate-cochleariform with incurved, hooded apex, + clawed, denticulate [fimbriate or entire], imbricate in bud. Stamens 8-10, in 2 series, the outer (antipetalous stamens) usually somewhat shorter than the inner whorl (antisepalous stamens) ; filaments filiform, broadened toward

on an intrastaminal, cushion-like, nectariferous disc; stigma small; style central; ovary deeply 4- or 5-lobed, 4- or 5-locular, with [2 to] numerous ovules on axile placentae. Capsule glandular-punctate, 4- or 5-lobed, 4- or S-locular, [few- to] many-seeded, dehiscent loculicidally inward (adaxi- ally) at apex, [or split into indehiscent segments]. Seeds angled, brown, tuberculate; endosperm fleshy; embryo slightly curved, cotyledons some- times 2-lobed. (Including Haplophyllum A. L. Juss.). LECTOTYPE SPECIES: R. graveolens L.; see P. Wilson, N. Am. FI. 25: 212. 1911. (Classical Latin name of the plant [since Cicero in literature], related to and ques- tionably derived from Greek, rAyte, the name of the plant in Nicander; etymology obscure.) — RUE.

A genus of about 60 species, ranging from Macaronesia eastward through the Mediterranean region to central Asia and eastern Siberia. Ruta graveolens L., common rue, 2” = 72, 81, native to the Mediter- ranean region but widely naturalized in temperate parts of the Old World, is’ introduced and more or less naturalized in the eastern United States. There are few reliable records from our area and further data regarding

6 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLII

the distribution of this species in the southeastern United States are very desirable.

The species is proterandrous. The stamens execute peculiar nutation movements during the expansion of the flowers. Pollination agents re- corded are small Hymenoptera, Coleoptera, and Diptera. Polyploidy

tetraploid, and octoploid species are known. The occurrence of a nonaploid chromosomal race of the otherwise octoploid Ruta graveolens has also been recorded. However, no general conclusions can yet be drawn on the basis of the few counts.

The common rue is of ancient culture as an ornamental, spice, and medicinal plant. Because of a volatile oil (“oil of rue”) it was formerly used in medicine, but at present, its use is quite limited because of very unpleasant secondary effects. The species should be regarded as + poison- ous. Contact with fresh leaves of the plant produces a dermatitis in some individuals. Poisoning by oil of rue is characterized by gastroenteritis. The Mediterranean R. chalapensis L. and some other species have also been used locally in much the same way as R. graveolens. The latter and a few other species are cultivated as ornamentals.

REFERENCES:

See also under family references ENGLER (1931. pp. 243-246), GALLeT (1913, pp. 23-27), Necopr (1937, pp. 93-97, 101), UrBan (1883, pp. 372, 373, 401), WILSON (1911, p. 212).

AHLEs, H. E., C. R. Bett, & A. E. RADFORD. Species new to the flora of North or South Carolina. Rhodora 60: 10-32, 1958. [R. graveolens, Orange

ALLEN, N. Beware of garden rue. Horticulture 24: 403. 1946. [Poisoning by t

contact. BeRSILLOoN, G. Les inflorescences de Ruta graveolens L. Revue Gén. Bot. 63: 437-460. 1956.

Branpt, W. Zur Anatomie und Chemie der Ruta graveolens L. Thesis, 33 pp. Berlin. 1915.* [See review in Bot. Jahresb. 43¢1)3 S10, 511, 1913.4 CAPPELLETT!, C. Sterilita di origine micotica nella Ruta patavina L. Ann. Bot.

Roma 18: 145-166. pls. 6,7. 1 ; 18; includes the embryology. ] Outa, T., & T. Mryazaxr. Furoquinolines. XIII. Alkaloids from the pericarps of Kuta graveolens L. (In Japanese.) Jour. Pharm. Soe. Japan 78: 538, 539. 1958. SovEGES, R. Développement de lembryon chez le Ruta graveolens L. Bull. Soc. Bot. Fr. 73: 245-260. 1926

2, Zanthoxylum Linnaeus, Sp. Pl. 1: 270. 1753: Gen. Pl. ed. Oe. LOU, 1754.

Deciduous or evergreen trees or shrubs. often armed with prickles which sometimes become elevated on broad-conical or -pyramidal corky excrescences; bark aromatic. Leaves alternate. odd- or even-pinnate to 1-foliolate; leaflets opposite or alternate. frequently inequilateral, crenu- late or entire, glandular-punctate, at least at the margin; petiole and

1962 | BRIZICKY, GENERA OF RUTACEAE 7

rachis winged or wingless, unarmed or prickly. Plants dioecious, monoeci- ous, [or polygamous]; flowers small, white to greenish yellow, unisexual [and/or bisexual], in axillary short spikes or cymose fascicles or in ter- minal, sometimes corymbiform, panicles. Sepals 3-5[10], distinct or + connate, deciduous or persistent, or apparently wanting. Petals 3-5 [8], imbricate [or valvate| in bud. Stamens 3-5[8], distinct, alternate with the petals, rudimentary (staminodial) or wanting (or sometimes trans- formed into carpels) in ¢ flowers; filaments filiform to subulate; anthers ovate, elliptic to subcircular in outline. Intrastaminal disc small, often pulvinate, or obscure. Gynoecium of (1)2—-5 sessile or stipitate, distinct or partially united [very rarely completely connate] carpels, rudimentary in g flowers; stigmas capitate, distinct or connate; styles sublateral, dis- tinct, connivent or connate toward the summit; ovaries usually 1-carpellate and -locular [very rarely ovary compound, 2—5-carpellate and -locular | with 2 collateral, pendulous ovules in each carpel [or locule]. Fruits 2- valved follicles, distinct or connate at base. stipitate to sessile, firm-walled or fleshy, glandular-punctate, with separating (loose) or adherent endo- carp, 1(2)-seeded. Seeds obovoid to subglobular [or + lenticular], black [blue-black, brown, or dark red], shining, with a crustaceous testa and fleshy endosperm, at maturity often hanging from the carpels on slender funicles: embryo axial, straight or somewhat curved, with a short radicle and flat, almost circular cotyledons. (Including Fagara L., nom. cons., type: F. Pterota L.) Type species: Z. americanum Mill. (‘“Xanthoxy- lum”) (Z. fraxineum Willd.); see F. R. Fosberg, Taxon 8: 103-105. 1959. (Name from Greek, xanthos, yellow, and xylon, wood.) — PRICKLY ASH.

A genus of about 215 species, primarily pantropical, extending with several species into the Temperate Zone of eastern Asia and North America. Two subgenera sometimes regarded as distinct genera are recognized in the present treatment.’

Subgenus ZANTHOXYLUM, with unisexual flowers with petaloid [or sometimes sepaloid], 4—5{10|-merous, “simple” perianth in a single series (presumably petals) and 4-5[8| stamens or [1]3-5 carpels, includes about 15 species, primarily of the Temperate Zone of eastern Asia and North America, but with at least two species in Central America. Zan- thoxylum americanum Mill., common or northern prickly ash, 21 = 68, 136, a shrub or small tree with odd-pinnate leaves, paired pseudostipular prickles (rarely prickleless) and yellow-green flowers in sessile, axillary, umbellate clusters, expanding before leaves, is the only eastern North American species of the subgenus, occurring from Georgia and Alabama northward beyond our area to North Dakota, Minnesota, Ontario, and western Quebec. Zanthoxylum mazatlanum Sandwith is known from

>The occurrence of species apparently transitional in the character of the perianth between Zanthoxylum and Fagara is ample reason to regard both as components of a single genus. See “Taxonomic and Nomenclatural Notes on Zanthoxylum and Glycos- mis (Rutaceae)” in the present issue of this Journal.

8 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIII

Mexico; Z. Williamsii Standl. and Z. ferrugineum Radlk. occur in Hon- duras and Costa Rica respectively.

Subgenus Facara (L.) Triana & Planch. (Fagara L.), with unisexual

[or both uni- and bisexual] flowers with 3-5[—8]-merous perianth in two series (i.e., sepals and petals), 3-5[—7] stamens and/or 3—5 carpels, in- cludes about 200 species, primarily of the tropics of both hemispheres, four in our area. Zanthoxylum Clava-Herculis L., southern prickly ash, Hercules’ club, 2x = ca. 72, armed trees or shrubs with odd-pinnate leaves and usually 5-merous flowers in terminal panicles, occurs on the Coastal Plain from southern peninsular Florida to Texas and southeastern Vir- ginia, northward to southern Arkansas and Oklahoma, The South Ameri- can-West Indian Z. Fagara (L.) Sarg. (Fagara Pterota L.), an ever- green shrub or tree usually armed with pseudostipular prickles, with odd-pinnate leaves having relatively small leaflets and winged petioles and rachises, and 4-merous flowers in short axillary spikes, reaches its north- ern limit in central Florida and in southern and southwestern Texas beyond our area. The West Indian Z. coriaceum A. Rich., a + prickly tree or shrub, usually with even-pinnate, coriaceous leaves and 3-merous flowers in terminal panicles, is known from southern peninsular Florida and the Florida Keys. The West Indian Z. flavum Vahl, yellow-wood, satin- wood, an unarmed tree with odd-pinnate, occasionally 3-1-foliolate leaves and S-merous flowers in terminal panicles, occurs on the lower Florida Keys. Little is known regarding pollination, but bees and various Diptera have been recorded as the most frequent visitors of the flowers of Z. americanum. Nucellar polyembryony has been found in a few species (e.g., Z. ameri- canum, “Z. Bungei Planch.,” Z. alatum Roxb.), but probably is of wider distribution within the genus. The few published chromosome counts (2n = 32, 64, 68, 70, 72, 136) indicate polyploidy and perhaps aneuploidy. No interspecific hybrids have been recorded, A modern monograph is highly desirable.

The dried bark of Zanthoxylum americanum and Z. Clava-Herculis, “toothache bark,” “prickly ash bark,” or “xanthoxylum,” has been ap- plied as a stimulant, tonic, and sialagogue in the United States, and the bark of 2. alatum has been used against fever, dispepsia, diarrhoea, and cholera in India. Zanthoxylum americanum and some eastern Asiatic species are sometimes cultivated as ornamentals. Various other species have been used for timber, spices, and medicines.

REFERENCES: See also under family references Desar (1960, p. 32), ENGLER (1895, pp. 115-119; 1931, pp. 214-224), and Wrtson (1911, pp. 177-199). Barser, C. A. The nature and development of corky excrescences on the stem of Zanthoxylum. Ann. Bot. 6: 155-166. pls. 7, 8. 1892. BocquILton, H. Etude botanique et pharmacologique des Xanthoxylées. Thesis. 125 pp., 4 pls. Univ. Paris, Ecole Supér. Pharm. 1901. Fosperc, F. R. Typification of Zanthoxylum L. Taxon 8: 103-105. 1959,

1962] BRIZICKY, GENERA OF RUTACEAE 9

[Type sp.: Z. fraxineum Willd. (2. americanum Mill.) ; see also Taxon 7: 94-96. 1958. |

Goto, J. Studies on the toxic principle of Xanthoxylum piperitum De Can- dolle. Jap. Jour. Veterin. Sci. 17: 205-215. 1955.*

Huanc, C. C. Preliminary study on Chinese Rutaceae (1). (In Chinese.) Acta Phytotax. Sinica 6: 1-143. pls. 1-36. 1957. [Zanthoxylum, 6-83, pls. 1-19.]

Lonco, B. La poliembrionia nello Xanthoxylum Bungei Planch. senza feconda- zione. Bull. Soc. Bot. Ital. 1908: 113-115. 1908.

Reever, J. R. Xanthoxylum Miller (1768). Taxon 4: 237. 1955. { Proposed for conservation; see also L. H. SHINNERS, Taxon 6: 135-137. 1957, against conservation.

SARGENT, C. S. Xanthoxylum. Sylva N. Am. 1: 65-74. pls. 29-32. 1891; 14: 97-98. 1902.

Torro, F. G. The gum of Fagara xanthoxyloides. Nature 180: $64, 865. 1957.*

WirtH, E. H. Xanthoxylum. Natl. Formulary Bull. 9: 149, 150. 1941.*

Subfam. TODDALIOIDEAE Engler 3. Ptelea Linnaeus, Sp. Pl. 1: 118. 1753; Gen. Pl. ed.'5. 54. 1754.

Unarmed shrubs or small trees, with bitter bark and foliage disagree- ably scented when crushed. Leaves alternate, usually 3-foliolate, rarely 4-S-foliolate, leaflets entire or toothed, glandular-punctate, glabrous to densely soft-hairy beneath. Plants usually polygamous, the flowers bi- sexual and/or unisexual, greenish or yellowish white, aromatic, in termi- nal corymbiform, cymose panicles. Sepals usually 4 or 5, distinct, im- bricate. Petals usually 4 or 5, relatively narrow, surpassing the sepals, imbricate. Stamens usually 4 or 5, alternating with petals, hypogynous, inserted at base of a disc, very short, with imperfect sterile anthers in 9 flowers; filaments subulate, hairy in the lower half; anthers ovate- cordate, introrse. Gynoecium syncarpous, usually 2-carpellate, inserted on a low disc in bisexual and ¢ flowers, very small, imperfect (lacking style and with rudimentary stigmas), raised on a conspicuous truncate- pyramidal to subglobular disc in ¢ flowers; stigma capitate, usually 2-lobed; style relatively short and slender; ovary compressed, usually 2-locular and narrowly 2-winged, with 2 superposed ovules (the lower usually sterile) on axile placenta in each locule. Fruit a flat, subcircular to obovate, glandular-punctate samara with 2 broad, thin, reticulate lat- eral wings completely encircling the indehiscent 2-locular and 1(2)-seeded body. Seeds laterally compressed, semiovate to semilanceolate in outline, acute at apex, rounded at base, dark reddish brown to black, densely papillose and glossy on the surface; seed coat thin, leathery; endosperm fleshy, thin; embryo large, straight, with oblong to ellipsoid cotyledons and short, stoutish superior radicle; germination epigeous. LECTOTYPE species: P. trifoliata L.; see P. Wilson, N. Am. FI. 25: 208. 1911. (Classi- cal Greek name of elm, Ulmus, transferred by Linnaeus to this genus on account of the similar fruit.) — HoP-TREE, SHRUBBY TREFOIL.

A genus of three (according to P. Wilson) or probably more species,

10 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLII

of temperate North America. Ptelea trifoliata L. (including P. micro- carpa Small fide Wilson), common hop-tree, stinking ash, 2n = 36, 42, a shrub or small tree, very variable in regard to size, shape, and pubescence of the leaflets and size and shape of fruits, occurs from northern Florida, throughout our area, and beyond to Texas and northern Mexico, Nebraska, Towa, Michigan, southern Ontario, New York, and Connecticut. Its vari- ant var. mollis Torrey & Gray, regarded by Wilson as P. tomentosa Rai with branchlets, lower surface of the leaflets and inflorescences subtomen- tose, ranges from Georgia to North Carolina and westward to Oklahoma, Arizona, and Mexico. Ptelea serrata Small, an irregularly branched shrub with shallowly serrate leaflets, probably belongs with the preceding. Ptetea Baldwinii Torrey & Gray, a shrub with leaflets smaller and nar- rower than in P. trifoliata, has been recorded only from northeastern Florida in our area but is apparently conspecific with P. angustifolia Benth., widely distributed in the western United States from southern Texas to Colorado, Utah, California, New Mexico, and Arizona, and southward into Mexico.

Vegetative reproduction by root sprouts has been recorded in Ptelea trifoliata. e occasional occurrence of 3- or 4-carpellate ovaries, some- times accompanied by a few additional antipetalous stamens, is note- worthy. Polygamy seems to be a generic character, Although herbarium specimens usually are represented either by male or by female inflores- cences, sometimes both bisexual and male flowers occur in the same in- florescence; more rarely a few female flowers (usually terminal) are found in the male inflorescence. Cross-pollination is the rule because of the predominance of unisexual flowers, but self-pollination seems to be possible in the bisexual flowers. Bees, especially short-tongued species, and a few other Hymenoptera and Diptera have been recorded as polli- nators. The occurrence of a nucellar embryo has been recorded in P. trifoliata (Mauritzon, 1935), but the observations (made on scanty ma- terial) need verification. Only two chromosome counts are reported for

the genus. No hybrids have been recorded, ut, Desai (1960), investi- gating the cytology of P. ¢trifoliata grown in England (2 = 42), ob-

served meiotic irregularities resulting in high sterility and concluded a possible hybrid origin for the plant.

Over 60 species of Ptelea have been described, but these are apparently not well understood at present, and the reduction of all of them to two or three species is open to question. A modern monograph is urgently needed.

Ptelea trifoliata is a generally known ornamental plant. Its bitter fruits have sometimes been used in brewing as a substitute for hops, hence the English name of the genus. The bark is reputed to possess medicinal properties as a weak tonic.

REFERENCES:

See also under family references Desar (1960, pp. 29, 30), ENGLER (1931, pp. 302, 303), Maurirzon (1935, pp. 339, 346), URBAN (1883, pp. 397, 398), and WiLson (1911, pp. 208-210).

1962} BRIZICKY, GENERA OF RUTACEAE 11

Baitey, V. L. Historical review of Pfelea trifoliata in botanical and medical literature. Econ. Bot. 14: 180-188. 1960.

GREENE, E. L. Some Piéelea ss Torreya 5: 99, 100. 1905. [Descrip- tions of 3 new spp., 2 of the se. U. S.]

. The genus Prelea in a western and southwestern United States and Mexico. Contr. U. S. Natl. Herb. 10: 49-78. 1906. [59 spp. recognized, including 55 new. |

Harris, J. A. Teratological fruits of Ptelea. Bull. Torrey Bot. Club 38: 385- 387. pl. 17. 1911.

SARGENT, C.S. Ptelea. Sylva N. Am. 1: 75-78. pls. 33, 34. 1891; 14: 98. 1902.

ScHROEDER, E. M. Germination of fruits of Ptelea species. Contr. Boyce Thompson Inst. 8: 355-359. 1937.

Wixson, P. Notes on Rutaceae — V. Species characters in Pfelea and Taravalia. Bull. Torrey Bot. Club 38: 295-297. 1911

4. Amyris Linnaeus, Syst. Nat. ed. 10. 2: 1000, 1367. 1759.

Usually glabrous shrubs or trees with resinous, fragrant wood. Leaves opposite [subopposite or alternate], odd-pinnate, often 3—-5[rarely 1]- foliolate, with glandular-punctate leaflets and unwinged [or winged] petioles. Flowers small, bisexual [rarely unisexual], usually 4[rarely 3 or 5]|-merous, pediceled, in terminal or axillary panicled cymes. Calyx cuplike, 4-lobed, glandular-dotted, persistent. Petals 4, white, glandular- dotted, imbricate in bud. Stamens 8, in 2 series, inserted at base of a disc or of the ovary; filaments filiform; anthers ovate to oblong, introrse, 2-locular at anthesis. Intrastaminal disc pulvinate, gynophore-like, sup- porting gynoecium, or wanting. Gynoecium 1-carpellate; stigma capitate to discoid-subcapitate; style very short and stout or wanting; ovary 1-locular, with 2 collateral ovules suspended from the top of the locule. Drupes globular, ellipsoidal to obovate, black [or reddish], often glaucous, dotted with glands, aromatic, oily, the endocarp chartaceous, 1-seeded Seed pendulous, with thin, membranaceous testa, lacking endosperm; em- bryo with plano-convex, fleshy, glandular-dotted cotyledons and a short, superior radicle. Lecrorypr species: A. balsamifera L.; see P. Wilson, N. Am. Fl. 25: 216. 1911. (Name apparently derived from Greek, a, with, abounding in, and myron, balsamic juice, resin, with reference to the balsamic properties of the genus.) *— TorcHwoop.

A genus of about 20 species of the West Indies and especially of cis- Amazonian tropical America, extending southward (with one species?) to Peru and northward (with a few species) to Texas and Florida, The South American-West Indian A. balsamifera L., with 3—5-foliolate leaves, leaflets dull underneath, puberulous inflorescences, and puberulous, stipi- tate ovary, occurs in southern peninsular Florida and the Florida Keys. The primarily West Indian A. elemifera L., differing from the preceding especially in the glabrous inflorescences and glabrous, sessile ovaries, is

* More frequently the Greek prefix a has been used in a reverse sense, without, devoid of, not. This has led Little (U.S. Dep. Agr. Handb. 41: 57. 1953) to a differ- ent, hardly probable derivation: “not myrrh,” “not true myrrh.”

12 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIII

also known from southern peninsular Florida and the Keys, and the very similar A. maritima Jacq., with stipitate ovaries, has been recorded from Key West. Wilson and Small included this last species in 4. elemifera, but Urban (1896, 1920) regarded it as distinct.

The genus is, in general, very imperfectly known. The species, especially the West Indian, are closely related, and their delimitation is often weak, sometimes being based on few, questionably specific characters. A modern revision based on field observations, experimental cultures, and cytology is very desirable.

The resinous timber, especially that of Amyris balsamifera and A. elemifera, is of excellent quality but is scarce and of small size. It is used locally for fuel, torches (fragrant when burned), small cabinet work, and to a limited extent as a source of an oleoresin, Mexican elemi, which is used locally in medicine.

REFERENCES: See also under family references ENGLER (1931, pp. 313, 314) and Wrrson (1911, pp. 216-220). SARGENT, C. S. Amyris. Sylva N. Am, 1: 83-86. pl. 36. 1891; 14: 98. 1902. Ursan, I. Additamenta ad cognitionem florae Indiae occidentalis. Particula III. Bot. Jahrb. 21: 514-638. 1896. [Amyris, morphological-taxonomic intro- duction, key to and description of the spp., abundant synonymy, 595-612. ] ———. Amyris L. Symb. Antill. 8: 323. 1920. [Includes synonymy and gen- eral distribution of our spp. |

Subfam. AURANTIOIDEAE Engler 5. Glycosmis Corréa, Ann. Mus. Hist. Nat. Paris 6: 384. 1805,

Unarmed, evergreen shrubs or trees. Leaves alternate [rarely opposite], odd-pinnate, often 1-3(5)-foliolate; leaflets pellucid-punctate, entire [ser- rate or crenulate], + inequilateral at base; petiole articulated with the blade in 1-foliolate leaves. Flowers relatively small, fragrant, bisexual, (4)5-merous, with very short, stout pedicels, in rusty-villosulous, spike- like cymes arranged in axillary panicles. Sepals (4)5, nearly distinct [or + connate], triangular-semicircular, + fleshy, imbricate. Petals (4)5, elongate, somewhat concave, white, imbricate. Stamens (8)10, the anti- petalous shorter than the alternipetalous, filaments flattish [or filiform], dilated upward [or downward] and abruptly narrowed into an acuminate tip; anthers relatively short, ovate-cordate, often apiculate, gland-tipped. Gynoecium 2-—5-carpellate, syncarpous, raised on a cushion-like [or cylin- drical] nectariferous disc; stigma broad, cushion-like (convex-disciform to semiorbicular), persistent; style often indistinct, short and very stout, merging into a subglobular or ellipsoidal 2—5-locular ovary covered with glands; ovules pendulous, one in each locule. Fruit a relatively small berry with thin pulp, + globular, sometimes depressed at apex and -£ oblique, 1—3-seeded. Seeds without endosperm, ellipsoid to subglobular, with a membranaceous testa; embryo with plano-convex cotyledons and a short

1962 | BRIZICKY, GENERA OF RUTACEAE 13

superior radicle. LecroTyPE sPEcIES: G. arborea (Roxb.) DC. (Limonia arborea Roxb.); see P. Wilson, N. Am. Fl. 25: 215. 1911.4 (Name from Greek, glycys, sweet, and osme, scent, odor, referring to the fragrant flow- ers of the genus.

A genus of about 35 species of southeastern Asia, the East Indies, the Philippines, New Guinea, and northeastern Australia. The southeastern Asiatic Glycosmis parviflora (Sims) Little (G. citrifolia (Willd.) Lindl.), with white or pink subglobular berries, widely cultivated in the warm re- gions of both hemispheres, has become naturalized in the hammocks of Key West (Small, Manual, 1933; Everett, 1940). Wilson (1911) and some other authors have included this Specs in “G. pentaphylla (Retz.) DC.” or “G. pentaphylla (Retz.) Corréa.’

The present knowledge of the genus is very imperfect, and both taxonomy and nomenclature are complex. Tanaka has studied the genus critically for many years but has not yet published a monograph.

The genus is of no economic importance.

REFERENCES:

See also under family references ENGLER (1931, pp. 316-318), PeNnzic (1887, pp. 194-209, pls. 19-21), SwIncLe (1943, pp. 153-158), and Wixson (1911, B25)

EVERETT, T. H. Glycosmis citrifolia. Addisonia 21: 29, 30. pl. 687. 1940.

LittLe, E. L., Jr. Notes on nomenclature of trees. Phytologia 2: 457-463. 1946. [Glycosmis parviflora (Sims) Little, comb. nov., 463.

NARAYANASWAMI, V. A revision of the Indo-Malayan species of Glycosmis Correa. Rec. Bot. Surv. India 14(2): 1-72. 1941.

TANAKA, T. A note on Retzius’ Limonia pentaphylla. Bot. Not. 1928: 156-160. 1928. [Records finding a presumed type specimen. |

6. Triphasia Loureiro, Fl. Cochinch. 152. 1790.

Evergreen shrubs with paired, or sometimes solitary, axillary spines. Leaves 3-foliolate [or simple], occasionally 1- or 2-foliolate; leaflets rela- tively small, subsessile, the terminal somewhat larger than the lateral, thick- ish, without evident reticulation, crenulate to crenate, glandular-dotted; petioles short, puberulous, not articulated with the leaf blade. Flowers 1-1.6 cm. long, bisexual, 3[5]-merous, fragrant, solitary or in 2- or 3- flowered cymes in the leaf axils, pedicels short, minutely 2-bracteolate. Calyx cuplike, 3[5]-lobed, persistent. Petals 3[5], linear to lanceolate- oblong, imbricate. Stamens 6[10] in 2 series; filaments slender, broadened toward the base; anthers small, oblong. Disc ringlike to short-cylindric, en- circling the stipelike base of the ovary. Gynoecium 3[5]-carpellate, syn- carpous; stigma + capitate, 3[5]-lobed; style slender, deciduous; ovar ovoid to ellipsoid, narrowed toward the ends, 3[5]-locular, with a solitary ovule in each locule. Berry small, dull reddish orange or crimson, with

4See also “Taxonomic and Nomenclatural Notes on Zanthoxylum and Glycosmis

c (Rutaceae)” in the present issue of this Journal.

14 JOURNAL OF THE ARNOLD ARBORETUM | VOL. XLIII

glandular-dotted exocarp (peel), mucilaginous, pulpy flesh, and 1— —3|-5] seeds. Seeds ellipsoid to subglobular, lacking endosperm, with fleshy or leathery testa; embryo straight, with plano- convex, sometimes unequal cotyledons snd small radicle. (Including Echinocitrus Tanaka.) Typr species: 7’. Aurantiola Lour. ( = T. trifolia (Burm. f.) P. Wils.). (Name from Greek, ¢riphasios, threefold, triple, referring to the usually trifoliolate leaves and trimerous flowers of the type species.) — LIMEBERRY.

A genus of three species, probably native in southeastern Asia, the East Indies, and the Philippines. Triphasia trifolia (Burm. f.) P. Wilson, com- mon limeberry, 2x = 18, 36, apparently indigenous to southeastern Asia and the East Indies, has been recorded as naturalized on the Coastal Plain from Florida to Texas. This species, much cultivated as an ornamental and hedge-plant in the warm regions of both hemispheres, is widely natural- ized in the tropics. Flowers with 4-merous perianth and gynoecium some- times occur in this species.

Cross-pollination by insects seems to be the rule, spontaneous self- polli- nation being prevented by the position of the stigma which considerably overtops the anthers. Nucellar polyembryony has been recorded in Tri- phasia trifolia. A spontaneous autotetraploid form of this species, with leaves thicker and flowers larger than in the diploid, typical form, has been found. Fruits of 7. trifolia are edible.

REFERENCES:

See under family references especially SwincLE (1943, pp. 236- 240); also ENGLER (1931, p. 325), Sac (1935), URBAN (1883, pp. 399, 400, 403), and WILson (1911, p. 22 ALINDADA, R. C., and . G. Gonzales. The graft affinity of Triphasia trifolia

(Burm. f.) P. Wils. with five varieties of citrus. Philip. Agr. 40: 631, 632 Boh] 6s SWINGLE, W. T. New varieties and new combinations in the genera Clausena, Oxanthera, and Triphasia of the orange subfamily Aurantioideae ur, . Acad. Sci. 30: 79-83. 1940. [T. trifolia var. tetraploidea Single:

7. Poncirus Rafinesque, Sylva Tellur. 143. 1838.

Shrubs or much-branched small trees with green twigs armed with stout, axillary thorns often flattened at base; foliage spurs with extremely short internodes developing from dormant buds just above the thorns on the previous year’s branches. Leaves deciduous, palmately 3- foliolate ; leaflets sessile, shallowly crenulate to serrate above the middle; petiole articulated with the blade, narrowly winged. Flowers fairly conspicuous, 3—6 cm. in diameter, Biceaual: subsessile, solitary or in pairs on the preced- ing season’s twigs (just above the thorns). Sepals 4—7, usually 5, distinct, ovate, reece Petals 4—7, usually 5, spatulate to obovate, with claw- like bases, white, soon deciduous. Stamens 20-60; filaments free, un- equal in length, slender, broadened toward the base; anthers ovate to

1962] BRIZICKY, GENERA OF RUTACEAE 15

ovate-oblong in outline, gland tipped. Intrastaminal disc annular to shal- lowly cupular, hairy. Gynoecium 6—8(usually 7)-carpellate, syncarpous; stigma capitate; style short and stout; ovary subglobular, hairy, 6-8 (usually 7)-locular; ovules 4—8 in 2 collateral rows on axillary placentae in each locule. Hesperidium subsessile, globular to pyriform, 3—5 cm. in diameter, dull lemon-colored, fragrant when ripe, finely and densely pubescent, many-seeded; peel (exocarp and mesocarp) soft, 5-10 mm. thick, with numerous oil cavities, rather rough; pulp (forming together with the inner walls of the locules the endocarp of fruits) consisting of elongate, cylindric-conical, slender-stalked vesicles (outgrowths of the inner surface of the tangential walls of the locules) filled with a very acid juice and droplets of acrid oil in the center, and with minute, lateral, irregularly branched appendages (Fic. 1g, h) which presumably secrete a viscous fluid. Seeds lacking endosperm, plump, ovoid, the testa leathery; embryos often several in a seed, differing much in size, with 2 equal or unequal cotyledons and a short radicle; germination hypogeous, the young seedlings at first with bractlike cataphylls, then intermediate forms that soon merge into normal 3-foliolate leaves. Type species: P. trifoliata (L.) Raf. (Citrus trifoliata L.). (From French, poncire, of obscure origin but applied earlier to a variety of citron with large, tuberculate fruits [C. Medica var. tuberosa Risso], perhaps also to other similar varieties.) — TRIFOLIATE ORANGE.

The single species, Poncirus trifoliata, 2n = 18, 36, native to central and nothern China, has become + naturalized in our area on the Coastal Plain from Florida to Georgia and Texas. Poncirus shows only a few insignificant variations and seems to be the most stable species of the group of “citrus fruits” (including also Fortunella, Eremocitrus, Clymenia, Microcitrus, and Citrus) which have been cultivated for long periods by man (Swingle, 1943).

The floral biology of the species does not seem different from that of Citrus. Both cross- and self-pollination seem to be almost equally effec- tive. Nucellar polyembryony, apparently of an induced type, has been proved. A small percentage of autotetraploids sometimes occurs among nucellar seedlings and those of open pollination. Poncirus hybridizes freely with species of Citrus and Fortunella, producing hybrids which usually are female-sterile but which occasionally produce some fertile pollen. Be- cause of competition with nucellar embryos in a seed (and perhaps from other causes) the sexual embryo, when produced, seldom reaches maturity in most of the Poncirus > Citrus hybrids. This leads to the development of maternal-type seedlings, seriously interfering with the normal segrega- tion and recombination necessary in plant breeding (Yarnell, 1942). Vigorous, variable trigeneric hybrids, some very complex, have also been produced. The citrangedin, involving Citrus, Fortunella and Poncirus, is extremely resistant to cold and is notable in that it shows clearly traces of all three genera, but is strikingly different from any species of those genera.

16 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLII

Generic distinctions in the Aurantioideae are weak and are much in need of further critical study. Although genetic evidence shows that Citrus and Poncirus are closely related, the latter has a number of differ-

SHOAL

of pulp vesicle (after Penzig, 1887), x ca. 110; i, seed, X 3; j, k, four of nine m m a single seed (nucellar polyembryony), X 4—note mostly un- equal cotyledons and different sizes and shapes of embryos

1962] BRIZICKY, GENERA OF RUTACEAE 17

ential characters: deciduous 3-foliolate leaves; overwintering flower buds covered with bud scales; densely pubescent fruits; pulp vesicles with minute lateral appendages: pith with transverse plates of thick-walled cells; stomata of the green twigs situated at the bottom of deep, narrow pits; seedlings with spirally arranged cataphylls, which merge gradually into foliage leaves, as in Eremocitrus; and the presence in immature fruits of the glucoside ponciridin, analogous to hesperidin but not found in Citrus. (Cf. Swingle, 1943.) Since the generic problem is very complex, involving a number of other genera, current usage is followed here, and Poncirus and Citrus are maintained as distinct.

Poncirus has been used more or less extensively in many citrus-pro- ducing regions of the world as a rootstock for cultivated citrus fruit trees. The somewhat dwarfing effect on Citrus scions is noteworthy. The species is commonly grown as an ornamental in Asia, southern and central Europe, and North America, and is sometimes used for hedges. The fruits have been used in medicine in China. The juice and peel of fruits sometimes are used after special treatment for flavoring in baking and confectionary.

REFERENCES:

See also under family references ENGLER (1931, pp. 332, 333), PENzIc (1887, pp. 132-149, pls. 11, 13, 14, ‘Aegle sepiaria DC.’), and SwincLe (1943, pp. 366-373).

BenTON, R. J., . Bowman, L. Fraser, and R. G. Kessy. The significance of a ee for citrus rootstock problems. Int. Hort. Congr. Rep. 13: 1235-1240. 1952.*

Cuapot, H. Remarques sur la Se auon des pépins de Poncirus trifoliata (Rafin.). Fruits 10: 465-468. 1955.*

Knorr, L. C. Re- -appraising citrus rootstocks, with particular reference to their susceptibility to virus diseases. 1. Trifoliate orange (Poncirus tri- foliata (L.) Raf.). Citrus Mag. 19(8): 12, 14, 15, 18, 22, 23. 1957.*

MUKHERJEE, S. K., and J. W. Cameron. Tree size and Chromosome number in a trial of tetraploid trifoliate orange as a citrus rootstock. Proc. Am. Soc. Hort. Sci. 72: 267-272. 1958.

SWINGLE, W. T. Poncirus Raf. In C. S. Sarcent, Pl. Wilsonianae 2: 149-151. 1914. [Restoration of the genus. ]

WeatTuers, L. G. The effect of host nutrition on the capac of exocortis in Poncirus trifoliata. (Abs.) Phytopathology 50: 87. 19

Wirtrock, G. L. Poncirus trifoliata. Addisonia 15: 59, 60. a oe 1931.

8. Citrus Linnaeus, Sp. Pl. 2: 782. 1753; Gen. Pl. ed. 5. 341. 1754.

Glandular, aromatic shrubs or trees, usually with solitary, axillary thorns, the older branches often thornless. Leaves alternate, persistent, 1-foliolate; leaflet subcoriaceous, glandular-dotted, entire or toothed; petiole usually winged and clearly jointed with the blade (except in C. Medica). Flowers bisexual, sometimes also unisexual by abortion of the gynoecium, usually relatively large, 2-5 cm. in diameter, often fragrant, solitary or in pairs in the leaf axils or in short, axillary, corymbiform cymes. Calyx shallowly cup-shaped, 4- or 5-lobed. Petals 4-8, usually 5,

18 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIII

linear-oblong, thickish, white, pink, or purplish pink, glandular dotted, imbricate in bud. Stamens 20-60, polyadelphous [or distinct]; filaments linear-lanceolate, subulate upward, white, usually variously connate; an- thers oblong or somewhat sagittate. Disc annular to cushion-like, sup- porting the gynoecium. Gynoecium syncarpous; stigma -- capitate, some- times slightly lobed; style cylindrical, deciduous; ovary ellipsoidal to subglobular, 8— 18 (usually 10—14)-locular, with several, usually 4—8, ovules (arranged in 2 rows) on an axillary placenta in each locule. Hesperidium usually large to very large, ellipsoidal and often mammillate at apex or pyriform to subglobular, sometimes depressed at apex; pericarp differen- tiated into 3 layers: an outer, yellowish-green to orange, leathery exocarp (flavedo) dotted with very numerous oil glands, a middle, thick and spongy, white mesocarp (albedo), and an inner, membranaceous endocarp with juicy, stalked, fusiform, inner outgrowths. (pulp-vesicles) filling the locules (segments of the fruit) and forming the “pulp”; the thin, mem- branaceous partitions (radial walls) of the locules often loosely coherent and easily separated one from another as well as from the spongy, white central column (core) of the fruit. Seeds ellipsoidal to obovoid, plump or flattened, sometimes beaked at apex, usually a few in each locule (seg- ment), the testa leathery; endosperm lacking; embryos 1-several, green- ish to white, with fleshy, plano-convex, often unequal cotyledons and short radicle. Lecrorypre species: C. Medica L.; see P. Wilson, N. Am. Fl. 25: 221. 1911. (Classical Latin name, atiginally used for the wood of Tetraclinis articulata (Vahl.) Mast., tne ‘African sandarac tree, and per- haps other conifers, but transferred to the citron in about the first cen- tury of the Christian Era; etymology of the word obscure.)

A polymorphic genus of an uncertain number of species (16-145), of southern and southeastern Asia and Malaysia. Several species with numer- ous cultivars are widely cultivated and often spontaneous in all warm regions of the world. Five species, Citrus Medica L., citron, 2n = 18; C. Limon (L.) Burm. f. (C. Limonum Risso), 2n = 18, 36; C. aurantii- folia (Christm.) Swingle, lime, 21 = 18, 19-21, 27; C. Aurantium L. (C. vulgaris Risso), sour or Seville orange, 2” = 18: and C. sinensis — Osbeck (C. Aurantium [var.| dulce Hayne), sweet orange, 2n = 36, 45, all presumably natives of southern or southeastern Asia, a ee recorded as more or less naturalized in our area (primarily southern Florida and the Keys). Accurate and recent data, however, are scanty, and per- haps only C. Aurantium and C. Limon (‘rough lemon”) can with some certainty be regarded as extensively naturalized. Seedlings of all the above, as well as of C. paradisi Macf., grapefruit, 27‘= 18, 27, 36, are often found along roadsides, at the edges of woods, and in secondary woods in Florida. Further data as to the extent to which such seedlings persist and reproduce themselves are much needed.

Bisexual flowers usually occur regularly in Citrus sinensis (also in C. aa (L.) Osbeck, pummelo, C. paradisi Macf., grapefruit, and C. reticulata Blanco, tangerine). Both bisexual and unisexual (staminate)

1962 | BRIZICKY, GENERA OF RUTACEAE 19

flowers are common in C. Medica, C. Limon, C. aurantiifolia, C. Auran- tium, and some cultivars of other species (polygamo-monoecious species and varieties). Cross-pollination and self-pollination (including pollina- tion between individual trees of a clone) seem equally effective in the for- mation of embryos and seeds, with some exceptions involving absence or defective development of pollen, or self-incompatibility. Self- or cross- incompatibility has been recorded in some horticultural varieties of dif- ferent species. Thrips, honeybees, bumblebees, and some other insects attracted by the conspicuous corollas, fragrance, and abundant nectar apparently are responsible for natural cross-pollination and for most self- pollination.

Polyembryony is common in Citrus. Gametic (zygotic) polyembryony

gamospermy) seems to be rare, but can arise either by embryonic fission (cleavage polyembryony) or sometimes by the development of two game- tophytes (embryo sacs) in the same ovule. Nucellar (somatic) embryony (apomixis, agamospermy) is widely distributed within the cultivated species, apparently lacking only in C. grandis. (Some cultivars of C. Limon and C. reticulata are also mainly monoembryonic.) The number of embryos per seed (1-18) varies with species and variety, and there is no general consistency within the species. Often only a single seedling or rarely more than two or three seedlings develop from a polyembryonic seed. Although nucellar embryos sometimes occur in a fertilized ovule alongside a zygotic embryo, only pollination (nct fertilization) seems to be necessary for their formation. The occurrence of autonomous nucellar embryony has not been positively demonstrated. Nucellar seedlings show increased vigor, larger leaves and fruits, and a Juxuriant development of spines (“‘rejuvena- tion” or “neophyosis”). Parthenocarpy regularly occurs in some culti- vars of C. sinensis (e.g., navel orange, Satsuma), C. aurantiifolia (e.g., ‘Tahiti lime,’ a triploid), and some others.

The species of Citrus generally are interfertile, and their hybrids are + fertile. Only a few varieties may be cross-sterile, and many inter- and intraspecific hybrids are known. The F, progeny of such hybrids usually shows great variability, but less vigor, than nucellar offspring or hybrids from intergeneric crosses (e.g., with Poncirus and Fortunella). The great variabiltiy of Fy, of inter- and intraspecific hybrids is mainly explainable by parental heterozygosity, which is common. Bud mutants are frequent.

Polyploidy is frequent, triploids (of gametic origin) and tetraploids (somatic autotetraploids) having been recorded in various species, e.g., C. aurantifolia and C. Aurantium. Penta- and hexaploids seem to be ex- tremely rare. Aneuploidy and an unbalanced chromosomal complement, 2n = 28, have been found in a few hybrids.

A mycorrhizal association with an endophytic fungus has been recorded for the genus, but the association seems to be facultative (perhaps + para- sitic), since the formation of root hairs has been observed in young seed- lings grown in artificial cultures (Girton, 1927, Hayward & Long, 1942), as well as under field conditions (Bartholomew & Reed, 1943).

There is little unity of opinion on the generic delimitation of Citrus.

20 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIII

Although Swingle, who is followed here, regarded Poncirus, Fortunella, Eremocitrus, and Microcitrus as distinct genera, some taxonomists include Poncirus in Citrus, and Burkill (1931) united all of these. Much greater differences exist in the concept of the species. Swingle (1943) recognized 16, at least one, C. paradisi Macf., being considered a satellite (or doubt- ful) species, while Tanaka (1952) accepted 145 species. The difficulties in delimiting the species, subspecies, and varieties of Citrus seem to de- pend, for the most part, on the absence of sterility barriers between groups of related forms and their apparent heterozygosity; but the problem is further complicated by nucellar embryony, by rejuvenation by nucellar progeny of + senescent varieties long propagated asexually, and by the spontaneous production of autotetraploids

In addition to morphological and anatomical characters, the chemical composition, especially the presence of certain glucosides in the fruits (e.g., hesperidin and eriodictyol glucoside in C. sinensis, those plus hes- peridin chalcone in C. Limon, hesperidin in C. Medica, aurantiamarin, naringin (?), and hesperidin in C. Aurantium), seems to be of importance in distinguishing the species. The presence or absence, as well as the num- ber, distribution, and character of the acrid oil droplets in the pulp vesicles of Citrus (and relatives) may be of some taxonomic significance.

The diverse uses and application of citrus fruits in the food and bev- erage industries, in essential oil production and the perfume industry, and as sources of vitamin C and “vitamin P,” are well known. The peel (flavedo) of sour and sweet oranges and lemons or the volatile oil ex- tracted from the fresh peel are used in medicine as stimulants, aromatics, and flavoring agents.

REFERENCES:

The vast number of references has been reduced here primarily to those either of general interest or dealing specifically with the southeastern United States. For an extensive list of references see H. J. WEBBER and L. D. BATCHELOR, The Citrus Industry. Vol. I. History, botany, and breeding. 1943; Vol. II. The production of the crop. 1948. (Univ. Calif. Press, Berkeley & Los Angeles). Under family references see especially PENzic (1887, pp. 17-132. pls. 1-10) and SWINGLE (1943, pp. 386-445); also ENGLER (1931, pp. 333-346), TILLSoN & BAMForpD (1938), and WILSON (1911, pp. 221-224).

Aut, S., & Musanis-Up-Din. Morphology of the spines of Citrus. (Abs.) Proc. Pakictan Sci. Conf. 9(3): 23. 1957.*

Barn, J. M. Morphological, anatomical, and physiological changes in the de- veloping fruit of the Valencia orange, Citrus sinensis (L.) Osbeck. Austral. Jour. Bot. 6: 1-24. pls. 1-4. 1958.

BANERJI, I. Morphological and cytological studies on Citrus grandis Osbeck. Phytomorphology 4: 390-396. 1954.

BARTHOLOMEW, E. T., & H. S. Reep. General morphology, histology and physi- ology. Chap. VI, pp. 669- Oe The Citrus Industry. Vol. I. 1943. See WEBBER & BATCHELOR abov

Bove, J., & G. Moret. La silice de tissus de Citrus. Revue Gén. Bot. 64: 34-39, 1957.

Burkitt, IJ. H. An enumeration of the species of Paramignya, Atlantia and

1962 | BRIZICKY, GENERA OF RUTACEAE 21

Citrus, found in Malaya. Gard. Bull. Straits Settl. 5: 212-223. 1931. [ Poncirus, ree Fortunella, and Microcitrus reduced to sections of Citrus, 217-223.

Cuapot, H. Systématique des Citrus en relation avec leur composition chimique. Bull. Soc. Sci. Nat. Phys. Maroc 35: 69-73. 1956.

Ex Azount, M. M., & S. H. CaMEron. Adventitious buds in Citrus. Proc. Am. Soc. Hort. Sci. 38: 363-368. 1941.

Ensicn, M. R. Venation and senescence of polyembryonic Citrus plants. Am. Jour. Bot. 6: 311-329. 1919.

FAUVEL-DECROMBECQUE, J. H. Les variations des Citrus et leurs conséquences botaniques, culturales, promologiques et commerciales. Thesis, 156 pp. Univ. Alger, Fac. Sci. Alger. Philippeville. 1944.

FLORIDA DEPARTMENT OF AGRICULTURE. Citrus Industry of Florida. 256 pp. Tallahassee. 1955.*

Frost, H. B. Genetics and breeding. Chap. IX, pp. 817-913. The Citrus In- dustry. Vol. I. 1943. See WepBer & BATCHELOR, above.

. Seed reproduction: development of gametes and embryos. /bid. Chap. VIII, pp. 767-815. [See also Jour. Hered. 29: 423-432. 1938; Jour. Wash. Acad. Sci. 15: 1-3. 1925; Proc. Natl. Acad. Sci. 11: 535-537. 1925.]

Furr, J. R., & P. C. REECE. Wentincaion of hybrid and nucellar citrus seed- lings by a modification of the rootstock color test. Proc. Am. Soc. Hort. Sci. 48: 141-146. 1946.

. C. Cooper, & P. C. Reece. An investigation of as formation in ‘adult and juvenile citrus trees. Am. Jour. Bot. 34: 1-8. 1947.

Furusato, K., Y. Outs, & K. IsHrpasui. Studies on Se in Citrus, Seiken Zihé 8: 40-48. 1957.*

Girton, R. E. The growth of Citrus seedlings as infeed by environmental factors. Univ. Calif. Publ. Agr. Sci. 5: 83-11

GurcEL, J. T. A., & J. SousiHe. Analysis of ene in Citrus, especially pummelos. (aka Portuguese.) Anais Esc. Super. Agr. “Luis de Oucion? 8: 727-746. 1951.* [English summary. |

Haywarp, H. E., & E. M. Lonc. The anatomy of the seedlings and roots of the velence orange. U. S. Dep. Agr. Tech. Bull. 786: 1-31. 1942.

Hopcson, R. W., & S. H. Caos: Effects of reproduction by nucellar em- Reon on clonal characteristics in Citrus. Jour. Hered. 29: 417-419. 1938.

Hu, C. C. A review of literature on Czfrus nucellar embryony, and some new data on old and young lines of Valencia orange. Mem. Coll. Agr. Natl. Taiwan Univ. 3: 20-33. 1953.

Hume, H. H. Citrus fruits. Rev. ed. 444 pp. Macmillan, New York. 1957.*

Knorr, L. C. The growing of lemons in Florida: historical, varietal, and cul- tural considerations. Proc. Fla. State Hort. Soc. 71: 123-128

,R. F. Suit, & E. P. DucHarmMeE. Handbook of Citrus diseases in Flor- ida. Fla. Agr. Exp. Sta. Bull. 587: 1-157. 1957

Kruc, C. A. Chromosome numbers in the subfamily Aurantioideae with special

reference to the genus Citrus. Bot. Gaz. 104: 602-611. 1943. [Addendum y H. B. Frost, 609, 610. | O. Baccui. Triploid varieties of Citrus. Jour. Hered. 34: 277-283.

Leroy, = F. La polyembryonie chez les Cztrus. Son intérét dans la culture et anielioration: Revue Bot. Appl. Agr. Trop. 27: 483-495. :

LoncLey, A. E. Polycary, Holvepery, and polyploidy in Citrus and Citrus rela- tives. Jour. Wash. Acad. Sci. 15: 347-351. 1925.

7 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIII

Loustav, J., A. OrtuNo, & O. Carpena. La estructura de los cloroplastidios de hojas de Citrus. I. Limonero. Anal, Edafol. Fisiol. Veg. 17: 473-485 1958.*

Mat ack, M. B. Observations on the red color of the blood orange. Pl. Phy- siol. a 729, 730. 1931.

. The juice sac of the orange with some observations on the plastids of Citrus. Jour. Wash. Acad. Sci. 21: 437-440. 1931.

Mrnessy, F. A., & C. A. ScHRoEDER. Pistil development in C7trus flowers. Bot. Gaz. 117: 343-347. 1956.

NAITHANI, S. P., & S. S. RacHuvansnt, Cytogenetical studies in the genus

Pus. Nature 181: 1406, 1407. 1958.

Nauriva., J Self- incompatibility in pumelo (Citrus maxima Merr.). Curr. Sci. Bangalore 21: 347. 1952.

Neusauer, H. F. Beobachtungen an Blaittern von Citrus maxima Merr. Osterr. Bot. Zeitschr. 106: 556-565. 1959.

Reep, H. S., & T. Fremont, Etude physiologique de la cellule 4 my corhizes ie

les racines de Citrus. Revue Cytol. Cytophys. Vég. 4: 327-348. 1935 Factors that influence the formation and development of my ee associations in citrus roots. Phytopathology 25: 645-647. 1935.

Russo, F., & M. Torrisi. Polyploidism in Citrus autopolyploids and allopoly- ploids. (In Italian.) Ann. Sper. Agr. II. 5: 1041-1062. 1951.* [English

summary SHAMEL, A. D. Bud variation and bud selection, Chap. X, pp. 915-952. The Citrus Industry. 1943. See above SuimoMurRA, H. Pharmacognostical studies on the pericarp of Citrus and re- lated genera (1-4). (In Japanese, English summary.) Jour. Jap. Bot. 35: 129-138, a 218, 282-288, 296-303. 1960. SWINGLE, W. T. The batawical name of the lime, Citrus aurantifolia. Jour. Wash. Acad. Sci. 3: 463-465. 1913. . New taxonomic technique in ap eae wild a - major crop plants age by Citrus. (Abs.) Am. Jour. Bot. 35: 1948, aes

BINSON, & E. M. ree New Citrus a. Cire. U.S: Dep. Agr. 181: 119; 10 pls. 1931.

TaNnaKA, T. Species problem in Citrus. A critical study of wild and cultivated units of Citrus, based upon field studies in their native homes. (Revisio Aurantiacearum IX.) 152 pp., pls. 1-3, chart 5. Tokyo

A revision of Assam citrus. (Revisic Auranbacearan XI) Bull. Univ.

Osaka B. 9: 29- oa 1959.

& T. TANINAKA. A revision of Osmocitrus, a ae of the genus Citrus.

(Revisio rene NUT.) Jbid. 10: 1960.

TotKowsky, S. Hesperides. A history of the hie and use of citrus fruits. 371 pp., 113 pls. London Sie

VENNING, F. D. Cortical tracheids: a new vascular element from the orange sub-family oe Aurantioideae). Quart. Jour. Fla. Acad. Sci. 9: O 114. 1946

ns on accessory vascularization in four species of Citrus and their bite application as new taxonomic characters. Jour. Wash. Acad. Sci. 37: 210-217. 1947. Wo tre, H = Some problems of Citrus nomenclature. Proc. Am. Soc. Hort. Sci. Cari ». Region 7: 18-21. YARNELL, 5S. a eaf segregation in Citrus-Poncirus hybrids. Proc. Am. Soc. Hort. Sci. 40: 259-263. 1942.

1962] TOMLINSON, LEAF BASE IN PALMS 23

THE LEAF BASE IN PALMS ITS MORPHOLOGY AND MECHANICAL BIOLOGY

P. B. ToMLINSON

IN PREPARING A SUMMARY ACCOUNT of the palms for the Anatomy of the Monocotyledons (Tomlinson, 1961la) the construction of the leaf base was described in a very cursory fashion, since little material had been available for detailed study. Also, this anatomical survey of palms had been carried out from the standpoint of systematic anatomy, and there seemed to be little information of taxonomic significance in the microscopic structures of the leaf base. More recently, having had access to abundant material from living palms, the problem has been reappraised. It has become evident that, although, from the point of view of systematic anatomy, the initial attitude may have been justified, the construction and behavior of the leaf base in palms is a complicated subject which is of considerable morphological interest. It is a problem which can only be understood by considering mechanical aspects of leaf development in rela- tion to growth of the stem. This subject is virtually unexplored, since the mechanical behavior of the palm leaf has no parallel elsewhere in the plant kingdom, and the construction and growth of palms is itself little understood.

That there is considerable variety in the structure of the leaf base in palms is obvious at once when a large collection of living plants is studied. Some palms have smooth trunks because the leaves are early deciduous, each abscissing cleanly; other palms have a more dishevelled appearance because the leaf base persists on the stem, as fibers, shreds, spines, or solid stumps. The distribution of these visible characters is not random, for the appearance and behavior of the leaf base is very constant for each species. Sometimes a single type may occur throughout an entire genus or larger taxon. In this way the structure of the leaf base in palms affords good diagnostic features. However, in spite of the diagnostic usefulness of these features, they are rarely employed by taxonomists, or, if used, their description indicates a sad lack of understanding of the fundamental con- struction of the palm leaf on the part of the author.. There are some exceptions. The notes given by Wallace (1853), although brief, often in- clude sufficient information to allow the recognition of the type of leaf base which is described. Bailey (1941) considered characters of the stem sur- face to be a basis for the subgeneric division of Acrocomia; these char- acters are a result of differences in the behavior of the leaf base in the two proposed subgenera.

On the other hand, it is not surprising that these features are poorly

24 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIII

understood since many taxonomic descriptions of palms have been made from fragmentary specimens in which the attachment of the leaf base to the stem is not visible. In addition, there is no accurate account of the morphology and anatomy of the leaf base in palms. But this problem is only one of the many aspects of palm structure in which our knowledge is sadly deficient. The reason for these deficiences can best be appreciated by the tropical botanist; living palms must be studied, and it is difficult to obtain material of plants which are scarce or inaccessible or too valuable to horticulture to be used for science. The problems are on a scale larger than that envisaged by most plant morphologists, and special techniques must be devised, for it is difficult to apply orthodox anatomical and mor- phological procedures to buds which may be up to half a meter in diameter.

The present account is intended as a general survey of the problem and fills in some of the larger gaps in our understanding without offering any comprehensive solution. The problem cannot be solved by casual examina- tion of unmolested palms. Buds must be carefully dissected and leaves of all ages examined because structures visible in mature leaves may only be of mechanical significance in early stages of development. Conversely, structures may be visible in the young leaf primordium only to disappear as the leaf reaches maturity. Thus, a ligule is far more commonly de- veloped by palms than an examination of their mature leaves would sug- gest. The ligule is present in the leaf primordium but shrivels, erodes away, or shreds as the leaf approaches maturity.

The only published work on the mechanical features of leaf develop- ment in palms is that of Schoute (1915) who described the morphology, anatomy, and mechanical behavior of the leaf base in Hyphaene. Schoute’s work illustrates how the mechanical demands made on a cylindrical sheathing organ which encloses growing tissues are met by the development of vertical clefts or sutures which appear in precise regions as a result of anatomical modifications. The ultimate configuration of the leaf attach- ment is shown to be very efficient and to utilize available tissues in an economic manner.

However, the leaf base in Hyphaene represents only one method by which the mechanical problems of the leaf base are overcome. Other methods involve either the shredding or partial decomposition of certain tissues of the leaf base, or the complete disappearance of certain parts, or sometimes the abscission of the whole organ. I have recognized a number of distinct biological types and these are described subsequently. Schoute himself was well aware that there was much to be added to his own account, for he wrote, “Il est cependant sir que le corps tout entier des palmiers n’a pas encore été étudié suffisamment sous ce point de vue et que l’examen attentif nous pourra révéler encore beaucoup de détails intéressants.”’

It is emphasized that this present article is only a preliminary essay and cannot hope to elucidate all the details. It deals with the problem entirely in a qualitative way and aims at providing information of use to the taxonomist who may wish to include notes on the diagnostic fea-

1962 | TOMLINSON, LEAF BASE IN PALMS 25

tures of the leaf base in his description of palms. Since problems of growth are involved, future detailed studies must be based on quantitative observations.

As a preliminary aid, an account of the growth processes taking place within the leafy crown of a palm are given in order that the subsequent discussion of mechanical adaptations may be understood. These growth processes have been summarized elsewhere (Tomlinson, 1961a,b), so only a brief outline is presented here.

THE DEVELOPMENT AND MORPHOLOGY OF THE LEAF BASE IN PALMS

Leaves in a palm always originate within the terminal cluster. Most stages of leaf development take place within this leafy crown and are not visible without careful dissection of the bud. As in most monocotyle- dons, each leaf in the palm originates as a minute cowl-shaped primordium at the shoot apex. In its earliest stages, the primordium does not com- pletely encircle the stem, but its two margins soon grow rapidly around the shoot apex so that the base of the primordium is sheathing almost from the first. Eventually the base develops into a closed tube, the future leaf sheath. The distal part of the primordium elongates to form the petiole and rachis, the blade being segmented by peculiar growth processes in the distal part of the primordium (cf. Eames, 1953). Since differentiation of organs and tissues within the leaf occurs in a basipetal direction, the distal parts mature early and complete their growth and ex- pansion while the base is still meristematic. This basipetal order of matura- tion should be clearly recognized, since it is of considerable mechanical significance. In each successive leaf, at the time when the blade is fully differentiated and on the point of unfolding, the leaf base is still actively meristematic. Therefore, in its early stages of unfolding and assimila- tion the blade is borne upon a delicate leaf base in which cell division and cell expansion still continue and which is useless as a rigid sup- porting organ. Support for the open blades of the younger leaves is pro- vided by older, surrounding leaves with rigid, mature sheaths.

Elongation of the parts of the leaf may take place neither at a uni- form rate nor in the sequence in which the parts mature. For example, elongation of the petiole is often very tardy, but rapid once it begins. In some cocoid palms a true petiole is absent since there is no naked portion of the leaf axis between the insertion of the lowest leaflets on the rachis and the mouth of the leaf sheath. In these palms a pseudopetiole is developed, because as the leaf matures the marginal tissues of the distal part of the leaf sheath disappear. This pseudopetiole may be quite long, as in Arikuryroba. In yet other cocoid palms, as in species of Orbignya, the petiole is absent. Other palms with no petiole, or with a very short petiole include species of Copernicia (e.g., C. torreana) and some of the scandent lepidocaryoid palms.

The basal, sheathing part of the leaf in all palms is always tubular

26 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIII

in its early stages of development (Fic. 1). The tube is closed and the leaf insertion completely encircles the stem. In many palms the initial tubular state of the leaf base can be revealed only by dissecting the bud, because, as the leaf matures, decay and dissolution of certain tissues re- duces its size considerably, and at maturity it may have only a narrow insertion.

A ligule is a characteristic feature of a number of palms. It is a tubular extension (rarely the tube is open as in Rhapidophyllum and Zombia) of the distal part of the leaf sheath above the insertion of the petiole. It is most conspicuous and persistent in many scandent palms of the lepidocaryoid alliance in which it may have a specialized func- tion. Thus, in Korthalsia it is commonly occupied by ants. A ligule is also present in many, possibly all, palms of the bactroid and caryotoid groups, although it is visible only in young leaves, disorganizing as the leaf matures. In some arecoid palms, e.g., Ptyvchosperma, and some borassoid palms, e.g., Latania, there are small auriculate outgrowths on each side of the mouth of the sheath. These structures may be homologous with a true ligule.

In order that subsequent descriptions may be clearly understood, it is necessary to define certain terms which are used consistently in later sections (Fics. 1-3). The abaxial surface of a tubular leaf sheath is its outer surface, the adaxial surface being the inner. In accordance with the terminology of Von Mohl (1824), which was also adopted by Schoute (1915), the part of the sheath on the side below the petiolar insertion is referred to as the dorsal side, the opposite side being the ventral side. This corresponds to the same use of these terms in descriptions of the carpel, itself a modified leaf. The tissues on each side of the dorsiventral plane may be described as lateral.

Although the leaf sheath in all palms is fundamentally a closed tube, it is convenient to recognize two main types of leaf base, according to the behavior of the sheath as it ages. In the first group the leaf base may be described as obviously tubular since the base persists and matures as a visible long, closed tube (Fic. 2). Leaf sheaths of this type are always long in proportion to the total length of the leaf. Up to one quarter of the leaf axis may be sheath. This type of leaf base occurs predominantly in members of the arecoid, chamaedoroid, and iriartoid groups and is also characteristic of scandent palms in other groups. In the second category the leaf base is not obviously tubular (Fic. 3). Only in the early stages of development does the sheath have the form of a closed tube. As the leaf base grows certain tissues disorganize and the cylindrical sheath disappears, the leaf sheath at maturity being represented by a broad clasp- ing base which may or may not encircle the stem. Leaf bases of this type are always short and rarely ‘exceed more than one-eighth of the total length of the leaf. It should be emphasized that there is no sharp distinction between these two classes and that intermediate types of leaf base, difficult to categorize, are quite common.

When the leaf base is examined at various stages of development it is

1962 | TOMLINSON, LEAF BASE IN PALMS od.

4 2 ae en —— petiole | distal ——— sheath A <—s

ventral dorsal abaxial (outer)

G J) surface i

adaxial (inner) surface

lateral

Fics. 1-3. Types of a bases in palms. 1, Diagrammatic representation of young leaf primordium. 2, Mature leaf base of obviously tubular type, with representative cross eee on of sheath below. 3, Mature leaf base of not obviously tubular type, with representative cross section of sheath below.

obvious that it can only be described as a uniform cylinder in a very approximate sense. The leaf sheath is always somewhat conical up-

conical outline is a result of the basipetal sequence in maturation. The distal part of the sheath matures early, so its mouth is always narrow, enclosing younger leaves, the tissues of which are not yet bulky. Sub- sequent widening of the mouth of the sheath is to a large extent accom-

28 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIII

plished as a result of continued growth and expansion of younger leaves which are forced through the mouth by their own basal meristematic growth. Whereas the mouth of the sheath is widened mechanically, the base widens through activity of its own meristematic tissues. In ligulate leaves the loss of the ligule is to a large extent caused by internal expan- sion of younger leaves. The ligule, being the most distal part of the leaf sheath, is always much narrower than the rest of the leaf base, and its disappearance from the leaves of larger palms is essential if younger organs which it encloses are to expand without restriction. In fact, the mechanical function of a tubular ligule may be to direct and support the pointed apex of primordia of younger leaves, maintaining them in a vertical position in such a way that they meet no resistance from older leaves.

In the tubular type of leaf base, the walls of the sheath are fairly uni- form except for slight thickening on the dorsal side (Fic. 2), a thicken- ing which is most obvious in the distal region, close to the insertion of the petiole. There may also be slight constrictions in regions in which sutures subsequently appear. In the nontubular type of leaf base the dif- ferences between dorsal and ventral regions are more conspicuous (Fic. 3); the dorsal region is always thick, woody, and persistent, whereas, the ventral region is thin, membranaceous, and often ephemeral.

DEVELOPMENT OF THE LEAF BASE IN RELATION TO THAT OF THE STEM

Growth processes in the leafy crown of a palm are not comparable to those of a woody dicotyledon (Tomlinson, 1961a,b). The apical meristem region is minute and is situated at the base of a shallow depression which terminates the axis of the palm. Leaves originate in the apical meristem proper, and, as they grow, each in turn comes to occupy the center of this depression. Each leaf is, however, displaced from this position to allow the development of younger leaves within, its insertion widening at the same rate as the stem enlarges. Thickening growth of the stem is not induced by the apical meristem itself but by a primary thickening meristem, a tissue which forms the surface of the apical depression. Each annular leaf insertion therefore continually increases in circumference as a result of the thickening of the axis, itself brought about by the primary thickening meristem. The rest of the leaf sheath also increases in cir- cumference at the same rate of enlargement as the leaf insertion. Initially this enlargement of the leaf sheath is effected entirely by cell division and cell expansion, but in the later stages cell division ceases. At the periphery of the apical depression thickening growth is no longer pre- dominant, lengthening growth of the axis occurs, and adjacent leaves become separated by internodal elongation. It is in this predominance of thickening over elongation growth in the early stages of stem develop- ment that palms differ so strikingly from most dicotyledons.

The stem commonly continues to widen as elongation proceeds, but no

1962 | TOMLINSON, LEAF BASE IN PALMS 29

‘longer by activity of the primary thickening meristem. Instead, there is continued growth and expansion of the ground parenchyma. This diffuse secondary thickening (Tomlinson, 1961a, p. 20) may be long continued and obvious, as in Roystonea, the trunk tapering upwards (Fic. 14). Otherwise, secondary growth is either limited or not obvious, since it is restricted to that part of the stem which is still enclosed by persistent leaf bases so that the distal tapering of the stem is not visible without removal of the expanded leaves. However, it is this phase of stem growth which has such a significant effect on the behavior of the leaf base.

It should be emphasized that, although growth of the palm stem can be envisaged as taking place in two distinct phases (an initial phase of thickening growth, followed by elongation growth), there is no marked disjunction between the two. Thickening growth continues well into the phase of elongation growth. In the smaller, canelike palms, and par- ticularly in the scandent palms, elongation growth begins early and may be much exaggerated. It should also be emphasized that a true under- standing of the growth processes at the stem apex in palms can only come from quantitative observations and not merely from a qualitative ap- proach employed in this article. Apart from figures quoted by Schoute (1915) there are no relevant published measurements. It is hoped that these qualitative notes can be amplified by measurements in future work.

The general basipetal sequence of growth and differentiation has al- ready been emphasized. Since growth of the leaf and its associated stem internodes is closely correlated, this basipetal sequence is continuous from leaf to stem. Consequently, growth of the internode, both in length and thickness, continues after the leaf base has ceased to enlarge. The mature leaf base thus is seen as a rigid envelope enclosing growing tissues. Mechanical stresses are set up in the leaf base. Several types of stress are imposed and these are indicated below.

MECHANICAL STRESSES IMPOSED UPON THE LEAF BASE

Four main mechanical stresses are exerted upon the leaf base. These stresses are not of equal significance, and their magnitude differs in dif- ferent palms, largely as a consequence of the great variety in growth habit exhibited by palms.

(a) Stresses due to growth of younger leaves. During the early phases of growth the leaf sheath is subjected to internal stresses arising from the expansion of younger, enclosed leaves. These stresses are not of equal significance throughout a single leaf base because the distal part of the leaf sheath matures before the base. Thus, the still meristematic base is capable of accommodating internal expansion by cell division and cell enlargement. The narrow mouth of the sheath is, however, usually widened passively under the internal stress. This widening may involve splitting or loss of tissues at the mouth of the sheath, changes which are most obvious in the ligule, when present.

(b) Stresses due to expansion of the stem. The growing palm bud is to

30 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIII

some extent in a state of dynamic equilibrium. Much of the stress set up by the growth and enlargement of the young stem internodes is resisted by the rigidity of the encircling, mature leaf bases. That these stresses are considerable is shown when the mature leaves around a palm bud are removed so that there is no restriction on internal expansion: the axis commonly splits vertically. The structural modifications described below are largely a result of changes induced by such internal expansive forces.

(c) Stresses due to the weight of the leaf. Stresses are imposed upon the leaf base by the weight of its own rachis and blade. In view of the great weight of the leaves in the larger palms, these stresses are likely to be enormous. “A single, entire leaf is a load for a man” (Wallace, 1853). Such forces can trap and kill a man. They are not constantly and uniformly distributed because the palm leaf moves in the win he distribution and moments of these forces have been indicated by Schoute (1915) for Hyphaene. Schoute considers that the mode of attachment of the leaf in Hyphaene is of maximum efficiency in catering to these forces. However, Schoute discusses the problem as that of a single leaf on the axis. This condition does not occur in palms, and much of the stability of the leafy crown must be a result of many leaves deriving mutual support as a consequence of their overlapping leaf bases being ar- ranged on a spiral with a high phyllotactic fraction. This probably ac- counts for the scarcity of a distichous leaf arrangement in palms and for the relatively short internodes of the larger palms with persistent leaves. Some of the stresses set up in the dorsal part of the leaf axis by the weight of the blade sometimes cs visible effects. The dorsal tissues may buckle under compression, as in the leaf base in Cocos, or a “geniculum” may be developed, as in many on palms.

(d) Stresses due to the development of the inflorescence. The inflo- rescence in palms is usually axillary, and its enlargement places some strain on the leaf base. This strain may be minimized in several ways. In many of the arecoid palms, for example, the inflorescence does not expand until the leaf has abscissed (Fic. 15). Commonly the inflorescence elongates considerably, and grows nan the mouth of the leaf sheath before it expands. The inflorescence sometimes pierces the leaf sheath, or grows through an existing suture. In general, mechanical stresses imposed by the expansion of the inflorescence are less severe than those imposed during the normal processes of vegetative growth.

Of these four types of mechanical stress, that imposed by the expansion of the stem is undoubtedly the most Sinica for the behavior of the leaf base. Thus, the measurements given by Schoute (1915) for Hy phaene indicate that at least a 30% increase in diameter of the leaf base occurs passively, after its growth has ceased. My own measurements on Sabal indicate that the diameter of the leaf base must at least double as a result of stem expansion. This increase is accommodated entirely by splitting of the leaf sheath. since the sheath widens neither by addition of new tissues nor by stretching of old ones.

Anatomical observations indicate that the leaf sheath has only a limited

1962 | TOMLINSON, LEAF BASE IN PALMS oF

capacity for expansion by passive stretching. Initially the leaf base widens to accommodate internal enlargement by cell division and cell ex- pansion. After cell division ceases, tangential expansion of ground tissue cells continues, the cells often widening two- to threefold. This enlarge- ment involves the cells of the ground parenchyma only, since the tissues of the vascular bundles complete their differentiation early and are in- capable of passive expansion. Tangential expansion of ground tissue cells is most obvious close to the inner surface of the sheath. Here cells which were originally isodiametric may be widened up to tenfold in the tangential direction.

The limit of this passive enlargement is soon reached, however, and subsequent widening of the leaf base takes place entirely by means of certain mechanical contrivances for which the leaf is adapted in various ways. In this way the rigid leaf base is able to persist on a stem which is still widening. The main biological types which I have recognized are described below. There is no clear limit to these categories, and different types intergrade. Some palms may show mechanical features belonging to more than one class.

MORPHOLOGICAL TYPES OF LEAF BASES

The following observations are entirely my own, made on living palms either in cultivation or in their natural habitat. In each of the classes which I have recognized, the bud of at least one example has been dis- sected carefully. A few preliminary remarks should be made concerning observations on leaf base morphology in palms. The larger palms have a long period of growth in which only juvenile leaf characters are seen (Tomlinson, 1960). These juvenile characters may persist longer in the leaf sheath than in other parts of the leaf, and a leaf with the adult char- acters of the blade may still have juvenile features in the sheath. In Mascarena, for example, the leaves at the base of the plant are long persistent and not deciduous like the more distal leaves. In the follow- ing notes juvenile characters are ignored. Observations on palms in cul- tivation can also be misleading, since such palms are commonly leaf pruned at regular intervals and not allowed to show features they would develop if left undisturbed. These abnormal aspects also have been ignored.

The types of leaf base listed below are identified by the name of a genus in which typical features are clearly shown. These types are listed in two groups according to whether the leaf base is obviously tubular or not.

Leaf sheath obviously tubular. Palms in this category can be sub- divided into two main types according to whether the leaf base is de- ciduous or not. However, these two types intergrade.

CaLamus TyPE. Leaf base long-persistent, little modified with age. In palms of this type each leaf base is long, tubular, and wholly or partly covers the slender internode. Stem thickening is slight, and the leaf

oe JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLUI

base compensates for internal stresses entirely by division and expansion of cells. The leaf sheath is subjected to little mechanical strain and shows no mechanical adaptations. Each leaf base dies, its tissues dry and shrivel somewhat, but the leaf is not abscissed. It is only lost in extreme age, by gradual erosion and decay. Where a ligule is present it also may persist, as in most of the scandent palms. Otherwise it may decay, as in Bactris species. The only change brought about in such leaf bases by mechanical means is a widening of the mouth of the sheath early in de- velopment.

This type of leaf base is found in the scandent palms which belong to several unrelated groups: lepidocaryoid palms (several genera), bactroid palms (Desmoncus), chamaedoroid palms (Chamaedorea spp.). It is also found in many of the smaller palms with erect, rigid but canelike stems as in species of Chamaedorea, Geonoma and Pinanga. In these last-men- tioned genera, however, there is a tendency for the leaf to absciss in the manner described below.

VEITCHIA TYPE. Leaf deciduous, developing distinct separation layers. (Fic. 4). In this type the diameter of the stem continues to increase ap- preciably long after the tubular leaf sheath has achieved its maximum diameter through growth. The extent of this belated stem expansion may be considerable, and in Roystonea it may continue throughout most of the life of the palm (Schoute, 1912), although the trunk tapers ob- viously only close to the crownshaft (Fic. 14). The leaf sheath is not resistant to this expansion, but the whole leaf abscisses cleanly as a unit. This results from the development of separation layers in precise ab- scission regions. An annular separation layer at the node frees the leaf from the stem, but since the sheath is a closed tube, the leaf is not yet able to fall away from the stem. Complete release is effected by a vertical split on the ventral side through a narrow, but often conspicuous, separa- tion layer which extends the full length of the leaf sheath (Fics. 1 ae ee In this region the leaf may be slightly constricted; rarely the constriction is so deep as to form a conspicuous ventral groove (Fic. 15). These separation layers are both anatomically specialized and, at maturity, represent weak areas. Under the strains imposed from within, the leaf sheath splits from above downwards through the ventral separation layer (Fic. 12) abscission then being completed by cleavage at the node (Fic. 13). The annular leaf scar is always clean and smooth, indicating that separation is predetermined and not merely a mechanical tearing and also that protective layers are developed before the leaf falls (Fic. 13). Ana- tomical aspects of this abscission require further study.

The ventral line of abscission is always present, although it is more conspicuous in some species than in others. It is probable, however, that its conspicuousness is a good specific diagnostic feature. It is easily seen in Veitchia, for example (Fic. 12), but is not obvious in Roystonea. This ventral separation layer apparently has been overlooked in taxonomic literature. Schoute (1915) certainly missed it in Roystonea, which was

1962 | TOMLINSON, LEAF BASE IN PALMS 33

contrasted with Hyphaene as an example of a palm in which the leaf base was split in a haphazard manner. Admittedly this irregular tearing does occur in Roystonea, but usually only in young palms, and Schoute’s observations may well have been made on young Roystonea plants. In adult specimens cleavage is almost invariably regular.

An additional morphological peculiarity, common in species with the Veitchia type of leaf base, is the presence of a small triangular notch at the node, the apex continuous upward into the ventral separation layer (Fic. 11). This structure must be formed early in leaf development and represents the region in the minute leaf primordium where the encircling halves of the future leaf sheath meet on the ventral side, touching but not fusing completely. This feature is probably a good specific diagnostic character, although it has apparently never been mentioned in taxonomic writing. For example, it is very conspicuous in some species of Veitchia, e.g., V. montgomeryana, but absent from others, e.g., V. joannis. Since the notch marks the exact ventral point of the leaf asecuon and persists as part of the leaf scar, it is a very useful tag by which the original phyllotaxis can be determined on old trunks.

This type of leaf base, exemplified by Veitchia, is common throughout the arecoid, chamaedoroid and iriartoid groups. It is almost universally associated with an infrafoliar inflorescence. It does, however, intergrade with other types. In many small palms with canelike stems abscission is often imprecise and there is a transition to the Calamus type. In many

palms of the arecoid alliance the inflorescence is either interfoliar or not consistently infrafoliar, and abscission layers are not well defined. There is thus a transition to other types of leaf base, and further study would probably reveal them as a new type. Heterospathe is an example of a palm the inflorescence of which is, with fair consistency, infrafoliar, but in which abscission of the leaf is irregular. This makes it difficult to cate- gorize.

Leaf sheath not obviously tubular. This includes a number of distinct biological classes of which the type examples are quite clear cut. Some leaf bases may have characters of more than one type, how- ever, and in many genera it is not easy to categorize the type of leaf sheath.

HypHAENE TYPE. Leaf base cleaving widely on the dorsal side. (Fic. 5.) This type has been described by Schoute (1915), and the present summary of his observations adds little new information. In palms of this type the leaf base is at first shortly tubular. Allowance for the demands of stem expansion is made in two ways. Initially a ventral split (fente ventrale of Schoute) appears and its widening is accompanied by the loss of some ventral tissue. This split permits expansion of younger leaves and also causes the leaf as a whole to fall from a vertical position into a more pendant one. The second adaptation is more complex and accommodates the more extensive forces of stem thickening. These forces are exerted on the thick, woody dorsal tissues of the leaf base, and their

34 JOURNAL OF THE ARNOLD ARBORETUM [| VOL, XLIII

effects are indicated by Schoute’s measurements. They show that the circumference of the leaf base ceases to increase actively after it has reached a diameter of about 20 cm. Subsequent increase, up to a diameter

4 VEITCHIA

5 HYPHAENE

(=> a

lod:

Tics. 4-6. Behavior of leaf base, illustrated diagrammatically, immature to right, mature stage to left, with representative cross sections of bot - :

1962 | TOMLINSON, LEAF BASE IN PALMS 35

of 38.5 cm. is largely due to the development of a dorsal cleft (fente dorsale) which widens at the same rate as the stem thickens and ultimately assumes a broad, rhombohedral shape (cf. Fic. 18). This cleft is no merely the result of mechanical cleavage. It is marked in the immature leaf by a slight constriction, the microscopic examination of which shows certain anatomical peculiarities. The cleft is mechanically weak so that separation occurs easily within it. In Hyphaene the margins of the cleft are clean, but in Sadal and some other palms with this type of leaf base there is some fraying of the margin and fibers frequently remain to con- nect the newly exposed surfaces.

This type of leaf base is usually distinct and easily recognized, but transitional forms exist in some small palms. Thus, in T/vinax, the dorsal cleft is rather indistinct, and the rest of the leaf base shows additional mechanical adaptations.

The overall result of the development of a dorsal cleft is the appearance of a broad diamond-shaped hole in the woody leaf base. Mature stems of palms with this type of leaf base are thus clothed with a crisscross pat- tern of overlapping leaf stumps, each “strand” in the pattern representing one half of a split leaf base. Schoute’s mechanical analysis of Hyphaenc indicates that the clefts do not weaken the leaf attachment since they appear in such a position that they offer mechanical advantages to the leaf insertion. In older leaves the distal part usually breaks away by a fracture of the petiole. Commonly the residual leaf base persists through- out the life of the palm, if undisturbed, as in Hyphaene. In Sabal, the leaf base itself soon falls to leave a fairly clean trunk. In Washingtonia, the whole leaf persists and the stem of an unmolested palm is a striking object, being clothed with a “skirt” of old leaves. In Corypha the extent of lateral movement of the leaf base is revealed by horizontal scratches, scored in the leaf base by the teeth on the margin of the petiole (Fic. 19).

This type of leaf base occurs in larger palms of the borassoid and sabaloid groups. The reason for the consistent correlation between a Hvphaene-type leaf base and a palmate blade is unexplained.

PHOENIX TvPE. Ventral tissues of leaf base eroding without becoming conspicuously fibrous. (Fic. 6.) The leaf base is initially shortly tub- ular, but resists stem expansion verv little. The ventral tissues disor- ganize, decay, and erode, and only the thickened, woody dorsal part of the leaf base persists (Fics. 16, 17). Otherwise, there are no specialized anatomical adaptations in this type. It is possible that expansion in the nodal region ruptures the ventral leaf traces so that the ventral tissues are deprived of food and water and die, subsequent decay being a passive proc- ess. The dead tissues show little tendency to persist as a fibrous material. The annular leaf scar, which is often rather indistinct, is the only indication of the original tubular state of the leaf base. The petiole fractures, and only the dorsal woody stump persists, sometimes throughout the life of the palm. Otherwise it is abscissed or decays gradually. The persistent dorsal part of the leaf base may occasionally show the effects of in-

36 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLII

ternal forces of expansion, in numerous small vertical splits, as in the larger species of Copernicia (Fic. 17). In the larger species of Phoenix, e.g., P. canariensis, in which the internodes are congested, the woody overlapping de of the leaf bases form a complete vesture on the trunk (Fic. 16). This may have some protective value, like the bark of a hardwood tree. In other species of Phoenix, as in P. reclinata, the internodes are longer, and the woody base is less persistent. Copernicia cerifera is an interesting example of the Phoenix type in which the leaf bases are more persistent on the older, lower parts of the stem than on the younger, upper part. Hence, mature specimens of this palm have old woody leaf stumps at the base of the trunk, below the relatively smooth upper trunk

This type is common throughout the palms in many unrelated genera. Its limits are not easy to define and it intergrades imperceptibly with other types, particularly with the two next described. In Phoenzx itself, for example, there is little tendency to form a persistent fibrous mass. In some of the cocoid palms which I have included in the Phoenix type, however, the ventral tissues tend to split into strips before they disappear. One feature which characterizes many cocoid palms is for a part of these strips to persist as blunt spines spaced at regular intervals along the margin of the leaf sheath, as, for exam oe in Arikuryroba, Butia, Elaeis, Jubaea, and Syagrus. The distal part of the petiole beats spines in Phoenie it- self, but these are obviously modified leaflets and not homologous with the spines in the cocoid palms, Another diagnostic feature of the cocoid palms is the absence of a true petiole, the apparent petiole being the persistent thickened dorsal region of the leaf sheath, although this may be quite long. In Afttalea, Orbignya, Scheelea, and Maximiliana there is neither petiole nor pseudopetiole. Only in Cocos itself is a naked petiole present.

Cocos TypE. Ventral tissues of leaf base long persisting as a continuous fibrous network. (Fic. 7.) This class shows a complete intergradation with the previous class. The type example, Cocos, is, however, quite dis- tinct. The ventral tissue decomposes only partly, and its remains persist and serve an important mechanical function. There is some loss of tissue and splitting in the distal part of the leaf sheath. The splits generally separate the thin, ventral tissue from the thickened dorsal tissue, this being necessary for the correct mechanical behavior of the leaf base. The ventral tissues undergo decomposition, the soft, parenchymatous ground tissue decaying and being lost. The fibrous strands and fibrovascular bundles persist and form a material which resembles coarse sacking, not only in appearance but, to some extent, in texture (Fic. 20). This ma- terial owes its texture to the peculiar disposition of the vascular and fibrous bundles. Towards the distal end of the leaf sheath the bundles

one with another, but form two or more systems (cf. Fics. 23, 24). Bundles within each system form parallel hiv st the different systems

1962 | TOMLINSON, LEAF BASE IN PALMS 37 are arranged so that they execute a shallow spiral, the spiral in adjacent

systems running in opposite directions. There are frequent anastomoses and irregularities within this arrangement, but the general analysis de-

MCQOEe@>

8 TRACHYCARPUS

Ai 9 ZOMBIA C) ARI) a

=

C © ® 10 CARYOTA

Fics. 7-10. Behavior of leaf base, illustrated diagrammatically as in Figs. 4-6. 7, Cocos type; 8, Trachycarpus type; 9, Zombia type; 10, Caryota type.

38 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLII

scribes the system quite well. With the decay of epidermal and ground tissues, only this fibrous network persists. Internal expansion pulls the tissue laterally, but the peculiar arrangement of this network allows the angular divergence of bundles in different systems to be increased, with- out affecting the structure or strength of the whole tissue. The tissue thus comes to resemble a coarse cloth, but only in a superficial way. There is no regular interweaving of warp and woof as in a man-made fabric; the palm sacking shows interweaving between fibers in different systems in an irregular manner and only sufficient to hold the adjacent systems of strands tightly together.

Examination of this material shows that it can be stretched sideways, as the mechanical requirements of the palm bud demand, but it is also very elastic. Consequently, while offering no resistance to expansive forces, it functions as an elastic tissue, cushioning and damping the variable strains imposed upon the leaf base by the movement of the blade in the wind. A horizontal widening of this leaf tissue must be compensated by a vertical shortening. This probably accounts for the distal cleaving of the ventral tissues away from the rigid dorsal part of the leaf base, and this cleavage is probably entirely passive. The long, pendulous tongues of ventral tissue, each of which represents the remains of the distal part of the leaf sheath, are conspicuous objects in the leafy crown of a coconut (Fie. 20)

It should be pointed out that the anatomical peculiarities in the dis- tribution of the strengthening tissues in the leaf sheath are not restricted to palms of the Cocos type, but that a similar arrangement is to be found throughout the whole family. This suggests that this construction has the same mechanical significance for all palms, but only in the Cocos type is the arrangement of major and obvious mechanical significance. Further detailed anatomical studies are needed to analyze this construction cor- rectly.

The Cocos type can be recognized in many genera belonging to unre- lated groups, but often only in a modified form. Intergradation with other types is frequent. For example, the texture of the elastic tissue varies considerably. Thus in species of Coccothrinax the fibrous strands are quite woody and the tissue is very rigid (Fic, 24). In Livistona, on the other hand, the fibers are fine and the tissue delicate and almost w oolly. The leaf sheath of Livistona thus resembles the type of leaf base de- scribed as the next class.

In palms of the Cocos type, the useful life of this elastic mechanical device is often short, as in Cocos itself. The ventral tissues are lost and the woody, dorsal stump soon abscisses to leave a fairly clean trunk. Otherwise, the leaf base may be long persistent, as in Paurotis and many species of Coccothrinax.

ie

—

TRAcHYCARPUS TYPE. Leaf-base tissue mostly persisting as fibrous material, (Itc. 8.) This class represents only a modification of the previous type, and all intermediate stages of transition can be seen. In

1962 | TOMLINSON, LEAF BASE IN PALMS Sie)

Trachycarpus all the tissues of the leaf base, except the thickened dorsal region, persist as a fibrous mass which represents the fibrovascular system released by the decay of the ground parenchyma. This fibrous mass offers no resistance to internal expansion. The leaf bases are closely crowded on the trunk and persist, if undisturbed, throughout the life of the palm as a continuous fibrous mat out of which the dorsal stumps protrude (cf. Fic. 25). Trachycarpus itself is of some interest because it will grow at much higher latitudes than any other palm, even as far north as Scotland. One is tempted to correlate this cold tolerance with the shaggy coat, but it should be remembered that Trachycarpus is not a warm-blooded animal. Coccothrinax crinita has a similar shaggy coat but is not par- ticularly cold-resistant.

Isolated examples of this class can be found in several unrelated genera, but it is often difficult to distinguish this type from the Cocos tvpe with which it intergrades. A transition to another type is also shown by Arenga species. For example. in 4d. pinnata (FiGs-25)- ube fibrous remains of the larger vascular bundles protrude prominently from the fibrous mass as long, pliable, but not sharply pointed, spines. as in the next class.

ZomBia typE. Leaf base shredding partly or wholly into protective spines. (Fic, 9.) Examples of this class can be considered a modification of either the Cocos or the Trachycarpus type. The softer tissues of the ventral region decay but the fibrovascular bundles persist, either partly or wholly, as spines. These completely invest the stem and form a very effective armature (Fic. 23). In Zombia, the distal extremities of all those larger vascular bundles which have massive fibrous bundle sheaths end in a sharp point. These extremities become reflexed as the leaf matures and form the prominent stout spines. Anatomically the arrange- ment of the vascular bundles corresponds to the general pattern de- scribed for Cocos. hree systems of vascular bundles exist, but only the bundles of the outer system have massive fibrous tissues. Distally these extend beyond the insertion of the petiole into a long ligule which, unlike the ligule in most palms, is open on the dorsal side CHiGe 222). As the leaf base matures, the spaces between the vascular network are enlarged by internal expansive forces (cf. Fics. 22, 23), while the distal extremities of the future spines are freed by disappearance of the marginal tissues (Fic. 22). The reflexing of the spines is entirely a mechanical process, resulting from the effect of expansion on the peculiar mode of attachment of the ligular spines to the leaf sheath.

In Rhapidophyllum hystrix the mature leaf base has the same biological function as in Zombia, but a few morphological differences are obvious. A ligule is present, open on the dorsal side, the free margins overlapping somewhat. The spines are the larger vascular bundles of the ligule freed by the breakdown of the surrounding tissues. These spines do not be- come reflexed. In mature palms, the spines protrude from a mat of fibrous tissue, which represents the remains of the network of smaller

40 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLII

nonspinous vascular bundles. Arenga is very similar to Rhapidophyllum, although not of the same class because no rigid spines are developed. Trithrinax acanthocoma is a third example of this class, although I have not studied it in detail. However, the spines seem to originate as in Zomobia.

It is hoped in the future to make detailed histological studies of the leaf base in this peculiar biological group.

Caryora TyPE. Leaf base splitting ventrally, split surfaces connected by persistent fibers. (Ftc. 10.) This type is found only in the smaller, caespitose species of Caryota, but seems sufficiently distinct to merit separate description. The leaf base is initially long, tubular and ligulate, the walls of the sheath being much thicker than is usual in tubular sheaths. The ligule shreds into fibers and disappears as the leaf matures. A slight constriction on the ventral side is visible externally as a vertical separation layer. This region is anatomically specialized. As a result of internal expansion, the sheath splits along this line. The two margins of the suture do not separate cleanly, being connected by persistent fibers which are pulled into a horizontal position as the suture widens. These horizontal fibers interweave in a complex and irregular manner along the line of the original suture (Fic. 21).

The leaf base in the Caryota type is quite long and more obviously tubular than in related palms. It is, therefore, intermediate between the tubular and nontubular type of leaf base.

DISTRIBUTION OF LEAF TYPES IN THE MAJOR GROUPS OF PALMS

The preceding account represents a summary of incomplete observa- tions. The main biological classes I have described are quite distinct, but others may exist, and the present ones may be subdivisible. In the succeeding paragraphs the distribution of these types throughout the major tribes of palms is indicated. The taxonomic groups are those employed in Tomlinson (1961a, Table 1), recognized partly on an anatomical basis. Only genera are listed, although it is evident that some genera include more than one leaf-base type. Where a record is based on personal observation, the name is followed by an exclamation mark. Otherwise the records are based on published descriptions and illustrations, and may not be accurate. The list is very provisional and incomplete. Many palm genera are omitted.

Palms in which the stem is more or less subterranean are not considered, since these form a group which has not been studied in detail. Such palms commonly show no mechanical response to the little stem expansion which occurs.

Arecoid palms (a) Inflorescence infrafoliar VEITCHIA TYPE: Acanthophoenix, Actinorhytis!, Archontophoenix!, Arecal,

1962 | TOMLINSON, LEAF BASE IN PALMS 41

Bentinckia!, Brassiophoenix!, Chrysalidocarpus spp.!, Clinostigma, Cyrto- stachys!, Dictyosperma!, Drymophloeus!, Euterpe spp.!, Geonoma spp.!, Gulubia, Hedyscepe, Hydriastele!, Iguanura, Linospadix, ree ees veitchia, Nephrosperma!, Normanbya!, Oncosperma!, Pina spp.! Ptychandra!, Ptychococcus!, Ptychoraphis!, Ptychosperma!, ee Rhopalostylis!, Roscheria spp., Roystonea!, Siphokentia!, Veitchia!,

In the above genera the inflorescences are always infrafoliar and abscission of the leaf takes place unequivocably according to the Veztchia type.

(b) Inflorescence either interfoliar or inconsistently infrafoliar

PHOENIX TYPE: Manicaria.

TRACHYCARPUS TYPE: Oenocarpus spp.

TYPE OF LEAF BASE OBSCURE: The following palms require detailed study: Asterogyne, Balaka, Calyptrocalyx, Calyptronoma, Chrysalidocarpus spp., Dypsis, Geonoma spp., Heterospathe, Howeia, Hyospathe, Jessenia, Lino- spadix, Neodypsis, Neophloga, Oenocarpus spp.. Phloga, Pholidostachys, Pinanga spp., Prestoea, Reinhardtia, Roscheria spp., Stevensonia, Vershaf- feltia, Vonttra.

In these genera the leaf base abscisses fairly early, but often in an irregular manner. The inflorescence may be frequently infrafoliar, although the leaf base differs somewhat from the Veitchia type, as in Heterospathe. In other ex- amples the inflorescence is consistently interfoliar, the leaf base is quite different from that of Veitchia and abscission is long delayed, as in Neodypsis.

Bactroid palms CaLamus TYPE: Bactris spp.!, Desmoncus!, PHOENIX TYPE: Acrocomia!, Aiphanes!, Astrocaryum, Bactris spp., Guilielma!.

Borassoid palms HypHaENe type: Bismarckia, Borassodendron!, Borassus!, Hyphaene!, Latania!, Lodoicea, Medemia It should be noted that this group of palms is the only one with the same type of leaf base in all species.

Caryotoid palms Cocos TyPE: Wallichia spp.! TRACHYCARPUS TYPE: Arenga!, Wallichia spp.! CaRYOTA TYPE: Caryota spp.!

Chamaedoroid palms

CALAMUS TYPE: Chamaedorea spp. !

VEITCHIA TYPE: Chamaedorea spp.!, Mascarena!, Opsiandra!.

In some Chamaedorea species and in Gaussia and Synechanthus, the in- florescence is commonly interfoliar, abscission is imprecise, and these palms are difficult to typify. They require further study

42 JOURNAL OF THE ARNOLD ARBORETUM [ VOL, XLIIL Cocoid palms

PHOENIX TYPE: Arecastrum!, Arikuryroba!, Attalea!, Butia!, Corozo, Elaeis!, Jubaea!, Maximiliana!, Orbignya!l, Rhyticocos, Scheelea!, Syvagrus!. Cocos TYPE: Cocos!.

It should be noted that, although most genera of this alliance have leaf bases of the Phoenix type, those of this type can be subdivided into two more or less distinct groups. Arikuryroba, Butia, Corozo, Elaeis, Jubaea, and Syagrus have a spiny margin to the pseudopetiole, whereas in Arecastrum, Attalea, Maxi- miliana, Orbignya and Scheelea the pseudopetiole is not spinous and is often absent

Iriartoid palms VEITCHIA TYPE: Cafoblastus, Catostigma, Iriartea spp., Socratea, W ettinii- carpus, In /riartea species and Jriartella the inflorescence is commonly interfoliar,

abscission of the leaf base may be imprecise and the palms are difficult to categorize.

Lepidocaryoid palms

CALAMUS TYPE: Avc Ls dass Calamus !, Calospatha, Ceratolobus, Cor- nera, Daemonorop: nospatha!, Korthalsia!, Lepidocaryum spp., Mvyrtale pis !, Pletton omia! “Pleclgeona pus Schizospatha.

PHOENIX TYPE: Coclococcus!, Mauritia spp. (e.g., M. setigera), Metroxylon!, Raphia spp.!

TRACHYCARPUS TYPE: Mauritia spp. (e.g., MW. carana), Raphia spp. (R. tae- digera).

Sabaloid palms

HypHAENE TYPE: Corypha!, Cryosophila!, Nannorhops!, Sabal!, Thrinax spp.!, Washingtonia!l,

PHOENIX TYPE: Copernicia!, Erythea spp.!, Pritchardia spp.!

Cocos TYPE: Coccothrinax spp.!, Colpothrinax, Erythea spp., Licuala spp Livistona!, Paurotis!, Pritchardia spp.!, Rhapis!, Serenoa!.

TRACHYCARPUS TYPE: Chamaerops!, Coccothrinax spp. (e.g., C. crinita!), Trachycarpus!.

ZOMBIA TYPE: Rhapidophyllum!, Trithrinax Zomobia!.

rn

spp. (e.g., acanthocoma!),

—

Genera of uncertain taxonomic position VEITCHIA TYPE: Pse udophoe nix! Cocos Type: Leopoldinia spp. (e.g., L. pulchra), Pelagodoxa.

TRACHYCARPUS TYPE: Leopoldinia spp. (e.g., L. piassaba).

The leaf base in Ceroxylon, Orania, and Sclerosperma needs further study.

1962 | TOMLINSON, LEAF BASE IN PALMS 43

DISCUSSION

In the present era of specialized and sophisticated botany with so much emphasis on experimental procedures, the plea for more research involving simple morphological observation is likely to go unheard. Descriptive botany is regarded as too unscientific and unprofitable a field for startling new discoveries. And yet, in the realm of tropical botany there is enormous scope for research which requires the simplest apparatus: an axe, a knife and a scalpel, a clear eye, and the patience and willingness to settle down and observe elemental things. The present article is based entirely on simple techniques of this caliber. The observations have been made on living palms, mostly in the field or outside the laboratory rather than within it. The palms are a tropical family, their morphology is im- perfectly understood because it differs from most other plants in many peculiar ways. It cannot be understood by regarding palms as aberrant trees. They are a unique but neglected group, and their study demands an outlook unclouded by a knowledge of how hardwood trees behave and preconceived notions as to how palms should behave. In so far as the growth of palms can be said to be known at all, its peculiarities raise problems more easily comprehended by an engineer than a botanist. The palms have discovered many devices by which these mechanical problems can be overcome. If we could ascribe reasoning powers to plants, pas one certainly be granted a high intelligence quotient, for their

“ingenuity” soon becomes apparent to a thoughtful observer. The success and persistence of palms through a long geological history and their size as a taxonomic group and wide present-day distribution bear witness to their success in experimenting with original, and often apparently ill- adapted, growth forms.

One of the most difficult tasks which may confront a morphologist is to recognize the existence of mechanical problems. The purpose of the present article has been to emphasize the question of mechanical be- havior of leaf insertion in palms, just as much as it has been to indicate ways in which the problem has been overcome. Only by further detailed studies using the more elaborate and precise techniques of modern botany can more decisive answers be given.

There are many other aspects of palm morphology, in addition to the leaf base, which require study from a functional aspect. For example, the growth and expansion of the inflorescence enclosed by a leaf sheath needs sympathetic consideration. Some of the answers to these mechanical problems seem easy enough. The leaf may fall before the inflorescence expands. Usually the inflorescence grows through the open mouth of the sheath. Sometimes the inflorescence protrudes through a cavity developed for quite a different purpose, as it protrudes through the dorsal slit in the leaf sheath of Latania and Lodoicea. Commonly the sheath is pierced by the expanding inflorescence. But to what extent this break- through is facilitated by anatomical preadaptation, we do not know. Here is a profitable field for simple observation. The behavior of the spathe

44 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIII

in palms is also a subject deserving careful study. Often the spathes are deciduous and conveniently fall so as not to hinder flowering. I know of no studies on the abscission of the spathe. The large woody spathe in cocoid palms deserves looking at with the eye of an engineer. One of its most common and striking features is the system of deep longitudinal grooves in its outer surface. They are of obvious mechanical significance when the life history of the spathe is considered.

None of these studies, however elementary, can be accomplished swiftly. It takes time to become familiar with the biology of the living palm. Palms in botanic gardens are to a large extent inviolate, and the gift of a palm bud for dissection may come only at long intervals and then, quite literally, as a windfall. However skillful a research botanist may be in planning a budget for the expenditure of large sums on laboratory equipment, he is usually much less adept at acquiring funds for travel and collecting. But a travel fund is an important item of research expenditure for the tropical botanist.

The study of some of these elementary problems in tropical biology may have wider implication than is at first apparent. Thus, the per- sistence or otherwise of the leaf base in palms may be a significant matter to the ecologist. The stems of palms with smooth, self-cleaning trunks are a common habitat for lichens and small bryophytes. Palms with rough, persistent leaf bases provide an important lodging place for many epiphytes — ferns, orchids, bromeliads, figs, and others. The shage fibrous coat of a palm stem is an excellent habitat for small animals — snakes, scorpions, frogs, beetles, ants, spiders, cockroaches, and a host of small insects. In dissecting the buds of palms the botanist should be prepared for constant zoological surprises.

SUMMARY

The problem of the leaf base in palms is introduced in this essay and solutions are offered only in a very general way. The leaf base in all palms is initially tubular, although it does not always mature as a closed tube. It completes tissue differentiation and matures as a rigid sheath long before the stem which it encloses has ceased its thickening growth The rigid sheath, especially in the larger palms, is thus subjected to internal stresses which it cannot accommodate by further active growth and expansion. Instead, this accommodation is effected by various mechanical devices for which the sheath of the palm leaf is often re- markably preadapted. Several distinct biological groups are recognized, although intermediate types are common. The leaf may absciss as a unit due to the development of separation layers in clearly prescribed regions. Otherwise, expansion of the sheath may be permitted through the development of wide clefts. Yet again, tissues may break down and disappear before they can offer resistance to stem expansion. In other types, these same tissues may persist either as an elastic tissue, or as a

1962 | TOMLINSON, LEAF BASE IN PALMS 45

fibrous mass, or sometimes even as protective spines, thus serving a secondary, mreidental, but important, function.

In order to emphaciee the taxonomic value of these morphological characters, which are quite constant in each species, a provisional list of leaf base types throughout the family Palmae is included

It is emphasized that there is much scope for this type of simple ob- servation, not only in palms, but in other groups of tropical plants. These observations are likely to have wide significance.

BIBLIOGRAPHY

BarLey, L. H. Acrocomia—preliminary paper. Gent. Herb. 4: 421-476. 1941.

Eames, A. J. Diese morphology of the palm leaf. Phytomorphology 3: 172-189. 19

Mou, H. von. =. structura palmarum. /m: Martius, Historia naturalis pal- marum 1: I-LII. 1824.

ScHoutTe, J. C. Uber das Dickenwachstum der Palmen. Ann. Jard. Bot. Buiten- zorg II. 11: 1-209. 1912.

. Sur la fissure médiane de la gaine foliaire de quelques palmiers. Jbid. 14: 57-82. 1915.

ToMLInson, P. B. Seedling leaves in palms and their morphological significance. Jour. Nenole Arb. 41: 414-428.

. Anatomy of the monocotyledons, ed. C. R. Metcalfe. Vol. 2. Palmae.

Oxford. 196la.

. Essays on the morphology of palms. VI. The palm stem. Principes 5: 117-124. 1961b.

WaALLaAcE, A. R. Palm trees of the Amazon and their uses. London. 1853.

FAIRCHILD TROPICAL GARDEN, MIAMI, FLORIDA

EXPLANATION OF PLATES PLATE I

cs. 11-13. Leaf bases of the Veitchia type. 11, Ptychandra glauca, leaf insertion, ventral aspect, separating layers and notch distinc , Veitchia johannis, leaf insertion, ventral aspect, leaf sheath splitting from above down- ward along the ventral separation layer, notch not developed. 13, V. johannis, abscissing leaf, ventral aspect, the clean leaf scar indicating preformed pro- tective layers.

PLATE II

Fics. 14-17. Palms with the Veztchia and Phoenix types of leaf base. 14, Roystonea regia (Veitchia type), falling leaf abscissed as a unit — note distinct tapering of stem below crownshaft indicating future extent of internal diffuse secondary growth. 15, Mascarena vershaffeltii (Veitchia type), crownshaft from below, separation region of outermost leaf sheath visible as conspicuous ventral groove. 16, Phoenix canariensis, stem surface with woody armor of

46 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLII

overlapping, gases partly encircling leaf bases. 17, Copernicia curbeloi (Phoenix type), vertical splits in leaf base near insertion reflecting extent of internal ee forces.

PLATE III

Fics. 18-21. Palms with the Hyphaene, Cocos, and Caryota types of leaf base. 18, Latania commersonit elle type), dorsal view of mature leaf base on young nonflowering specimen —=%in flowering specimens inflorescence commonly protrudes through large dorsal cleft. 19, Corypha elata (Hyphaene type), leaf bases showing horizontal scratches scored by marginal teeth of petiole. 20, Cocos nucifera, showing persistent ventral tissues of leaf base visible as tongues of material resembling coarse sacking. 21, Caryota mitis, ventral aspect of mature leaf sheath with peculiar fibrous configuration.

PLATE IV

Fics. 22-25. Palms with leaf bases of the Zombia, Cocos, and Trachycarpus types. 22, Zombia antillarum, immature leaf base, with ligule (open on dorsal side) beginnin ing to separate into individual vascular bundles. 23, Zombia antillarum, stem covered with persistent spinous leaf bases, spines released == note separation of larger vascular bundles (as compared os = 22) indi- cating extent of expansion, 24, ee a esiag acuminata (Cocos type), showing woody and rigid remains of ental tissu 5, Arenga nen (Trachycarpus type), showing long, sa spines ae from ligular tissues and initially standing vertically er

PLATE I

oL. XLII

7, ]

Jour. ARNOLD ARB. \

S

M

, LEAF BASE IN PAL

N

TOMLINSO

Jour. ArnoLp Ars. VoL. XLIII PLaTE IT

Jour. ARNOLD Ars. VoL, XLIII Puate III

&

TOMLINSON, LEAF BASE IN PALMS

Jour. ARNOLD Ars. VoL. XLIII PLATE IV.

ae < ahs 1)

Mae

or A ea. BP ier ys'

TOMLINSON, LEAF BASE IN PALMS

1962 | “HOWARD, OBSERVATIONS ON REDONDA 51

BOTANICAL AND OTHER OBSERVATIONS ON REDONDA, THE WEST INDIES

RicHARD A. HowaArD

THERE ARE MANY SMALL ISLANDS in the West Indies, but none is so intriguing as Redonda to the traveler in the Leeward Islands. Isolated and pinnacle-like, Redonda is a landmark to the sailors who pass it. Since deep water surrounds the rock, one can approach it closely, only to learn that a landing appears to be difficult, if not impossible. From the east, Redonda appears to live up to its descriptive name as a round island, and as the descriptions of Redonda have varied little from the original supplied by Columbus, who named it, one suspects that few of the recent writers have done more than reword an earlier description. At one time, the island of Redonda supported a mining operation and had about 130 peo- ple living on it, but for the last half a century it has been uninhabited. The island, however, is reported to have a “‘king” representing a regency now in the second dynasty. There are only meager botanical records from Redonda, and only one other biologist, an entomologist, has climbed to the crest.

During several weeks of field work on Montserrat in January, 1961,1 I saw Redonda daily in the distance. Finally, the temptation to visit it became overwhelming, and with the assistance of Mr. Kingsley Howes, my wife and I chartered the schooner “Melody” for a day and the trip to Re- donda. Departure before the break of dawn allowed a smooth and rapid passage to the island, where we landed, climbed to the peak, explored the shaft of the phosphate mine, and collected some samples of the vegetation. Descending in midafternoon, we set sail and broke out of the lee of the island with some difficulty, returning to Montserrat before sunset through rougher seas. The diversity of the vegetation, although mostly of weedy species, was unexpected. This alone compensated for the physical difficul- ties of the trip. It is not to be recommended to a tourist.

Redonda lies about 25 miles southeast of Nevis, and 15 miles northwest of Montserrat. It is clearly visible from both of these larger islands. Redonda is recorded as being about 1.5 miles long and 0.3 miles wide, and the highest point is given by Martin-Kaye as 975 feet by aneroid reading (Reports on the geology of the Leeward and British Virgin Islands, p. 77. 1959). The island is obviously volcanic in origin, presumably of the post- Pliocene period, and a part of the volcanic chain of the Leeward Island group including Saba, St. Eustatius, St. Kitts, Nevis, and Montserrat. It is a fragment of a volcano, however, for the western two-thirds have been

‘The financial support of Grant No. G4441 from the National Science Foundation towards a Flora of the Lesser Antilles made this trip possible, and is therefore grate- fully acknowledged.

52 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIII

lost in ocean depths, which reach 600 fathoms a short distance off shore. From the sea, the western face of the island is sheer, broken only by a mining scar at the northern end and one or two flume-like gullies farther south. In places, the vertical face appears to extend the full 975 feet. Various colored rocks and multiple layers of lava flows draw one’s eye. Sea-eroded cliffs 100 feet high surround the remainder of the island, mak- ing the initial assault a formidable feat, impossible except in one place. Clearly some cataclysmic event split Redonda in the past to leave the frag- ment present today. The local story that this split occurred in the 17th century and was observed by a passing sea captain cannot be checked. One can only wonder when the discrepancy developed between the early descriptions of the island and its present condition.

Redonda was discovered and named by Columbus on November 11-12, 1493, on his second voyage. Morrison, in Admiral of the Ocean Sea (p. 410), relates that “from the northern end of Montserrat, Columbus sighted a large island to the northwestward but did not care to beat up to it against a trade wind. He named it Santa Maria la Antigua.”

“Proceeding in a general northwesterly direction the fleet passed a small steep and rounded but inaccessible rock less than a mile long, which Columbus named Santa Maria la Redonda, St. Mary the Round. Redonda retains her name and her importance as a sea mark to this day; but she has never been worth inhabiting.”

Other historians have about the same story. In the Life of Christopher Columbus, by his son Ferdinand (p. 125), it is reported that “from here he proceeded to Santa Maria la Redonda, to which he gave the name because it is so round and smooth that it is impossible to climb its sides without a ladder.” Markham (Columbus, p. 152, 1892) reports Redonda to be “a round islet [that] was seen to the westward, so steep on all its sides that it seemed inaccessible without stairs or ropes thrown from the top.” More recently, Sir Frederick Treves, in The Cradle of the Deep (1908), de- scribed Redonda (p. 196) as “a smooth pale fabric of stone rising out of the sea, like the dome of some immense submarine hall, whose span is a mile. It reaches to the height of 1,000 feet. It is bare as a pebble. .. .” Ober (A Guide to the West Indies, Bermuda and Panama, p. 343. 1920) says of Redonda, ‘It appears scarcely more than a rock pinnacle rising above the sea between Nevis and Montserrat, but it is a mile and a half in length by a mile in breadth, with an altitude of 1,000 feet. The Spaniards called it Redonda, or the Round Island, from its shape.” Thus, the dimen- sions suggested by the chroniclers of Columbus do not vary from those of the present. It is not clear whether Columbus, viewing the island in the late evening, passed to the east and did not notice the steep western face, or whether the island has truly split since the time of discovery.

Although many naturalists passed the island of Redonda in the 17th, 18th, and 19th centuries, apparently not one of them reached it. Only Sir Hans Sloane appears to have noted the island, and even he fails to com- ment on its shape in his published writing. G. R. de Beer (Sir Hans Sloane and the British Museum, 1953) refers to a letter from Sloane which

1962 | HOWARD, OBSERVATIONS ON REDONDA 53

is now in the British Museum. Sloane wrote of his trip north from Bar- in September, 1687, “Between Montserrat and Nieves lies a very sari Island called Redondo or Rotunda, discovered by Columbus in his

AxpovE: Redonda as seen from the southeast. BreLow: Redonda from the northeast. The wave-cut cliffs and the cirques are from one to several hundred feet high.

54 JOURNAL OF THE ARNOLD ARBORETUM [ VOL, XLIII

Second Voyage, who gave it the name Santa Maria Rotonda, from its Figure. It consists of one Rock, very Perpendicular and high, looking like a Pyramide, and if there were nothing but Rock, but I was inform’d by those who have been on it, that there is on its top an Acre or two of very good Ground, that it has a very good Landing-Place, and a Well of very good fresh Water. It has also a great store of Iquanas of black color. Many Boobies, and other Birds that come hither to lay their Eggs at proper Seasons.”

Redonda was not visited by a naturalist until the Smithsonian-Bredin expeditions of 1956 and 1958 stopped at the island. Waldo Schmitt re- lated in the Smithsonian Report (457-8, pl. 6. 1957; 429-30. pls. 9, 10. 1959) of his visit to the shore and of the ascent of Redonda by J. F. Gates Clarke, who climbed the ridge on the first visit, and on the second made the ascent and spent the night collecting insects, while defending himself from the rats which infest the rock. Dr. Clarke gathered a few plant specimens, but some of these were lost, and others so ruined in the descent as to be of little use, and so were discarded. Dr. Clarke’s photographs re- main, to the present, the only published photographs of the island and its summit. Several comparisons between his photographs and my own or between our observations will be made.

Although currently uninhabited, Redonda has not always been so. At some unspecified period prior to 1860, ores containing phosphate were discovered on Redonda, and between 1865 and 1912 this material was mined. Shephard (Am. Jour. Sci. IT. 47: 428. 1869; 48: 96. 1870) de- scribed this mineral as Redondite, a name later shortened to Redonite, and currently considered to be variscite. It is included in a porphyritic or felsparphyric olivine basalt. In 1891, Hitchcock (Bull. Geol. Soc. Am. 2: 7. 1891) reported an analysis to show concentrations of P.O; as high as 42.9%, but a modern analysis of the residue remaining on Redonda yielded only 18% PsQs.

Mr. Fred W. Morse visited Redonda in the company of Prof. Charles H. Hitchcock in the summer of 1890, and reported on the trip (Popular Science Monthly 46: 78-87. 1894). His colorful description of Redonda and the activities thereon is worth repeating. Redonda was reached after 4 sail of three hours from Montserrat, Mr. Morse related. “As we ap- proached the pier, a boat manned by two negroes put off to meet us, with a strongly built man with a pleasant face and brown beard and dressed in white linen sitting in the stern. The man proved to be Captain H —, the superintendent of the mine, who welcomed us to Redonda and transferred us with our baggage to the shore.

“The beach was only a few yards in width, and above us towered the cliffs, over five hundred feet high. Groups of men stood on their brink, looking down at us and appearing like silhouettes against the clear sky. The ascent to the plateau above... was accomplished upon an aerial tramway.

“Two stout, heavy wire cables were stretched up the gorge and firmly anchored at both ends. Upon each cable ran a trolley, from which was

Lert: Western escarpment of Redonda, showing remnants of the cable base and landing. An as- cent must be made in the right-hand (southern) gully. Ricur: View down the gully from the saddle. Plants of Cephalocereus and Melocactus are visible on the right. Notice the narrow coastal shelf and the deep water within a few yards of shore.

[2961

VONOdGHea NO SNOILVAYWASHO ‘CYVYMOH

Sal

56 JOURNAL OF THE ARNOLD ARBORETUM | VOL. XLII

suspended a large iron bucket. To each trolley was attached the end of a light yet strong wire cable, which passed over a set of heavy pulleys at the top of the cliff, thus causing one bucket to ascend as the other de- scended. When passengers or freight were to be raised, the bucket at the top of the cliff was filled with water from a tank, and the lighter load at the bottom was quickly drawn up. The speed was regulated by means of brakes applied to the pulleys.

“The main cables were eight hundred feet long and the load was raised to the height of five hundred and twenty-five feet above the beach. In places the wires ran at a height of sixty feet above the uneven surface of the gorge.

“We were invited to get into the bucket which was at the foot. Captain H — stood on the edge, clinging to the trolley, and we rapidly glided up between the steep walls of the gorge, from whose rocky sides peered round cactus plants like heads of gnomes and several strange shrubs threw down aerial roots as though in a vain effort to reach the thin soil at the bottom. On gaining the landing at the top, we were. . . conducted up the slope a hundred yards to the superintendent’s house. The dwelling and office were really two separate buildings joined together by a wide veranda be- tween them and along their front. . . . the buildings had been framed and ready for putting together, and were small cottages with two rooms and with roofs of corrugated iron. We were met at the house by Mrs. H — and her young daughter, Dorothea, who, with the captain, were the sole white inhabitants of the island. A small black boy called Chalmers showed us to our room, where we prepared for dinner. By this time, the short twilight of the tropics had been succeeded by darkness, and when we returned to the dining room with its bright light we could hardly be- lieve that we were upon an almost inaccessible rock in the Caribbean Sea.

“The next morning, just before daybreak, while yet dark as night in the room, we were awakened by the cries of the sea birds, which made their homes by the hundreds in crevices and niches of the cliffs. Very soon a bell rang in front of the house to awaken the workmen in the huts below

s.”

Later Morse visited the mine. “The path to the mine led us along the eastern slope of the island to the northern face of the main peak, where a wide and deep ravine separated us from the smaller peak. The distance from the house to the mine was about three-fourths of a mile. The path was very steep in places as it ascended towards the summit in order to avoid a deep gorge, and sometimes so narrow that a misstep would give one a bad fall down the slope.”

Morse reported on the birds and lizards around the house, and com- mented that “several sheep and goats, two dogs, some hens, two peacocks, and a white cat comprised the domestic animals of this Crusoe-like home. From time to time the sheep and goats had become wild and had taken to the almost inaccessible parts of the cliffs and gorges, where it was exciting sport to pursue them.

“After lunch, when the sun had begun to descend towards the west,

HOWARD, OBSERVATIONS ON REDONDA

1962 |

SedIe JSIOUL UT punof a19M wWinj0

‘SYIOI ay} Suoure [Sq PUe Dimodagag ‘apIs Ula}sea dy] UO UOTIeAI[a appr ay. Inoqe

adoys 9219 LHOTY “Aqooq Bursar ayy aaoqe udas SI Sapl]s puke] puke wolsola Aq pewoy anbin y apeuutd yor ay} SuULMOYs epUopay Jo avy UIAISAM JO MITA :LETT

‘puvyst ay} Jo jurod ysaysry ay}

‘

58 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIII

Captain H — took us down to the plateau below the house to look at the quarters occupied by the workmen. The buildings consisted of two long sheds with close shutters instead of glass windows, and contained for furniture nothing but a tier of bunks or rather shelves, of rough boards along the walls... . Near the buildings were ovens where the baking was done... . Their fare was simple, consisting of bread and salt beef. The foremen and skilled workmen occupied two smaller houses, but lived in the same manner. Water for drinking was obtained by catching the rain on large inclined surfaces of corrugated iron, and collecting it in reservoirs. Such a reservoir was built at each end of the island for the use of the men, and the superintendent’s dwelling was provided with capacious tanks connected with the roof.

Tempany (West Indian Bull. 15: 22- 25. 1915) visited Redonda in 1915, after a disastrous hurricane on August 7, 1899, destroyed the build- ings Morse had described. Tempany’s descriptions, while less picturesque, give further details on the mining operations and on the living conditions of the workmen. Until the time of Tempany’s visit, phosphate rock was mined in four places on the island. The rebuilt dormitories now had space for 180 men, and the working conditions, including the rations of food and water, indicate that life for the laborers from Montserrat was any- thing but serene. Breakfast and lunch were served at the mine areas. After the cable at the northern end of the island weakened and collapsed, all ore was carried over the treacherous trail in basket loads of 84 pounds each on the heads of the men to the remaining cable near the manager’s house.

Although the last shipment of ore was made in 1912, ore was stockpiled at both ends of the cable until July, 1914. It cannot be definitely estab- lished when the island was evacuated, but it is known that a skeleton crew was maintained on the island until the hurricane of 1929 blew away the remaining buildings, and the company’s lease was finally relinquished in

Martin-Kaye, in describing the resources of the Leeward Islands, sug- gests that “the outlook for Redonda, however, is poor. The reserves are not known and to determine them would be expensive. Reports have it that the material was becoming more difficult to win. . . . There is the possibility that important resources remain, and this is supported to some measure by the length of time for which the defunct company held on to the lease. If they thought that the island held little further reserves the lease would presumably have been abandoned earlier.”

The story of the “King of Redonda” is worth reporting, for it is well known and often repeated by the residents of the nearby islands. The details vary a bit with the teller, and two published versions are also known. Bradley Smith, in his excellent book Escape to the West Indies, gives a straightforward account obtained from Charlesworth Ross, re- cently commissioner on Montserrat. Apparently the story began in 1865, when Matthew Dowdy Shiel claimed the island of Redonda, and a few years later had his son, Matthew Phipps Shiel, later better known as the

1962 | HOWARD, OBSERVATIONS ON REDONDA 59

writer M. P. Shiel, proclaimed King Felipe the First. Smith reports that the British Colonial office eventually “tacitly admitted” his claim, although today the island of Redonda remains under the administration of the gov- ernment of Antigua. Eventually the younger Shiel passed his title to John Gawsworth through an ancient succession ceremony involving the mingling of blood through cuts on the wrist. After Shiel’s death in 1947, Gawsworth became King Juan the First, and has since appointed many prominent figures Dukes of Redonda. A more glamorous version of this tale appeared in a recent issue of a popular magazine (Men Only [London] 75: 58-60. 1960) under the title of ‘King of all the Seagulls.” The coincidence of a disputed regency and a seemingly profitable mining operation makes one wish for more of the details of Shiel’s claim.

THE VEGETATION

A visit to Redonda is becoming increasingly more difficult. Storms of recent years have all but destroyed the loading pier, the remnants of which mark the landing spot. Once on the narrow, boulder-strewn “beach,” the only direction one can move is up. The ascent of Redonda is made along a cleft filled with volcanic gravel, which rises at a 60 degree angle. One climbs using all appendages, and the only resting place is the saddle, or ridge, fully 500 feet above the sea. In the afternoon, this valley assumes all of the characteristics of an oven, and, without a hint of a breeze on this, the leeward side, the rocks become too hot to touch. At the ridge, one finds a small level plateau on which are the foundations of former build- ings and the rusting machinery of the old phosphate mine. A grassy ap- pearing slope descends at an angle of 30 degrees to the east and the south. Broken water catchments, and the foundation of the manager’s house, are on the slope to the north, which is capped with a pinnacle of large boulders. A disintegrating foot path follows a circuitous route of varying altitude to a mine shaft, a short tunnel, at the northern end of the island. In all directions there is a breath-taking view of the northern Caribbean.

Redonda supports a wild herd of nearly one hundred goats, some of the males massive and handsome specimens with beards nearly reaching the ground. Their hooves left prints the width of my hand. Myriads of sea birds nest on the island. During the peak of the laying season, the men from Montserrat visit the island and gather eighty dozen eggs a day for sale on adjacent islands. Rats persist on all parts of Redonda, and sleep is impossible along the shore or in the ruins of the buildings because of the attacks of these animals. Lizards and iguanas are commonly seen, but appear to be more agile than usual in avoiding capture.

The early accounts of the plant life of Redonda are meagre. Morse, who visited the island with Charles Hitchcock, reports of their interest in the minerals, as well as the flora and the fauna. In addition to the “round cactus plants like heads of gnomes” and “several strange shrubs which threw down aerial roots” seen on the west wall of the island, he mentions the “red and yellow blossoms of the cactus’ around the man- ager’s house. He reports that “a few air plants, a species of Tvllandsia,

60 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIII

Apove: The grassy slope of Redonda as seen across the plateau ( (saddle) and site of the former barracks and mining buildings. The ruins of the manager's house are seen toward the upper right. BeLow: A closer view of the rocks and vegetation of the sa me slope. Volcanic “bombs” are infrequent. Opuntia

z

repens, known locally as “su a is abundant. Agave, Croton, Lantana, and an

W edelia species are domina

1962 | HOWARD, OBSERVATIONS ON REDONDA 61

clung to the projections of the rocks and formed almost the sole vegeta- tion at the extreme summit.” He was also impressed that ‘“‘one of the drawbacks to exploring the island was a variety of cactus which the work- men spoke of as ‘suckers.’ It resembles the prickly pear in form and had a yellow blossom. Its joints or sections were thickly covered with thorns or spines, which were from three-fourths of an inch to an inch and a half in length and barbed at the tip. The joints were easily broken off, and clung to anything upon which their spines could catch. The animals about the place were almost always seen with from one to a half a dozen of these ‘suckers’ clinging to them. When a barbed spine became imbedded in the flesh it produced a sore unless removed at once, and it was usually neces- sary to cut it out in order to remove it.”

Although Morse wrote, ‘The remainder of our stay was spent in col- lecting and preparing specimens of the phosphate, and also of the plants and animals,” no records of these collections can be located.

Mr. C. Steffens (Globus 67: 49. 1895) reported that the island appeared treeless, but that near the peak were “Gold- und Silberfarne” and under other plants ‘‘Tillandsien.”

In 1915, Trepany (doc. cit. 23) reported, “There is but little vegetation on the island, the surface being rock-strewn and barren, with no depth of soil. The principal vegetable forms which occur are species of Prickly pear (Opuntia) and Cacti, notably Cereus. It is, however, worthy of note that the silver fern (Gymnogramme trifoliata) and the gold fern (Gymno- gramme chrysophylia) both occur fairly abundantly in places in sheltered crevices in the rocks.”

More recently, Martin-Kaye (loc. cit. 77) states that Redonda possesses “scant vegetation beyond some coarse grass and extensive networks of particularly pugnaceous varieties of prickly pear. . . .’ However, Clarke (Smithson. Rep. 430. 1959) reported, “The inclined plateau forming most of the top of the island is covered by coarse grasses, sedges, a slender nar- row-leafed agave, several cacti in great abundance, lantana and several other scrubby shrubs.” Plates. g and ro associated with Clarke’s report illustrate some of the plants I was able to collect.

Although early observers may have described Redonda to be “as bare as a pebble,” plants exist on the island. Some ornamental or useful plants are remaining following cultivation. and a large number of “weeds” were obviously introduced. The typical strand plants of sandy beach and coral rock of the adjacent islands were missing; however, even the steep, wave- washed cliffs were abundantly populated with plants appearing to be perched in crevices or even on boulders soon to fall to the ocean below. The only level ground in the saddle, which formerly was occupied by min- ing buildings, had a wealth of grasses of wide-spread geographic distribu- tion. The eastern slopes of Redonda were simply piles of boulders, yet on and between the stones were many plants so abundant that the footing was even more treacherous, as the stones were obscured by the growth of Croton flavens, Lantana camara, and Wedelia calycina, the broad- leaved species present.

62 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLII

By far the most memorable plants of Redonda are the “suckers,” Opuntia antillana, Opuntia repens, and Opuntia triacantha. No descrip- tion does justice to the audacity or tenacity of Opuntia repens. Truly its pads are delicately attached and its spines retrorsely barbed the full length. The slightest jar caused a fragmentation of the plant, and the pads penetrated with ease the canvas, leather, or heavy rubber footwear of various members of our party. On the rocky plateau, the plants were abundantly branched, extending in many directions, the segments stretch- ing pad on pad over rocks and shrubs. On the cliff faces, plants of the same species seemed to develop a single stem structure. The plants hung in long festoons from the rocks, and single plants appeared to be as much as twenty feet in length. A slight touch with a stick or a falling rock would send the pads as individuals cascading down the slope on the unsuspecting climber below. If my records of the vegetation are inadequate or in- complete, the blame rests on this species alone.

Clarke reported that the only tree on Redonda was an introduced one, but failed to name it. Casuarina equisetifolia is clearly shown in his paper (pl. ro, fig. 2) near the ruins of the house of the manager of the mining company. In the vicinity are also introduced specimens of Bougainvillea spectabilis, its brilliant red flowers clearly visible from the schooner ap- proaching the island; Citrus aurantiifolia, the lime; and Annona squa- mosa, the sugar apple. Since Clarke’s visit in 1958, seedlings of Ficus citrifolia have been introduced by birds, for a nine-foot sapling is now flourishing from the apex of the central rock in Clarke’s photo of the “very tiptop of Redonda” (pl. 9, fig. 2). Other abundant plants, obviously residual from the period of mining operation, are Ricinus communis and Nicotiana tabacum.

A spire-like cactus (Cephalocereus royvenii) grew in profusion on the steepest slopes of the western escarpment, but ati could the plants be approached with safety in order to make a collectio

Although Clarke reported a cistern on top to ~ in “good condition,” during our visit we found that all of them were dry and so cracked as to appear scarcely capable of holding water. No rainfall records are avail- able, yet fortuitous showers must provide small pools or at least wet places deep in the piles of rock, and many of the scree slopes appeared damp and slippery. In damp, shady places on the eastern slopes, I was surprised to find several plants of Psilotum nudum, Pityrogramma chrysophylla, Pilea microphylla, and Peperomia simplex. Among the rocks, usually in slightly protected areas, were many plants of Tillandsia recurvata, but none was seen at the summit as Morse reported.

The following species represent the most complete account, to the pres- ent, of the vegetation of Redonda. All species cited by number are sup- ported by vouchers deposited in the herbarium of the Arnold Arboretum. Regrettably, large collections could not be made, and several sight records are given for clearly recognized species which either could not be handled in my limited amount of collecting equipment, or just could not be reached over the edges of the precipices.

1962] HOWARD, OBSERVATIONS ON REDONDA 63

Lichenes

The determinations of this group were made by Dr. I. M. Lamb, of the Farlow Herbarium, where the supporting specimens are deposited. The un- numbered collections were some of the more colorful lichen-masses and were not separated in the field from the small rocks on which they grew. Acarospora chrysops (s.n.) Buellia prospersa (s.n.) Calopaca sp. (s.1.) Heppia bolanderi (s.n.) Parmelia sp Ramalina ape pe rata (15227) — abundant in restricted areas of large boulders. Roccella babingtonii (15228) — abundant, pendant from the underside of large

boulders.

Psilotaceae

Psilotum nudum (15220) —a single clump found in a wet crevice under a large

Polypodiaceae Pityrogramma chrysophylla (15231) —the silver form alone was found in a gulley on the east slope, but smaller clumps of both the silver and gold forms were growing in cracks on the walls of the cistern above and to the west of the ruins of the manager’s house

Gramineae Chloris inflata (15251) Digitaria sanguinalis (15249) Eragrostis ciliaris ares 15253) Panicum maximum (15256) Pappophorum pappiferum oe —a clump grass with conspicuous inflores-

Paspalum laxum (15252) —a large clump grass, the largest and most common

Setaria setosa (15250, 15257)

Trichachne insularis (15254)

Trichachne sp. (15234) —a distinct species which, unfortunately has no avail- able name. Specimens are known from a range extending from Mexico to Peru, and this collection must have been introduced as a weed. Dr. Jason Swallen, who supplied this information, will describe and name the species at a later date.

Tricholaena repens —a colorful and easily recognizable grass, but only locally abundant.

Cyperaceae

Cyperus ligularis (15255) —a heavy clump sedge essentially limited to the higher elevations, the edges of the western precipice, and found occasionally on the sheer faces. Only rarely does the plant assume its normal shape, for the clumps are the favorite nesting places of the boobies (Sula spp.). The ae apparently keep the plants trimmed into a cushion shape, and, at the time of our visit, only a few of the lateral shoots bore inflorescences. (See Clarke's illustration loc. cit. pl. 10, fig. 1.)

Cyperus sphacelatus (15230)

64 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLII

Bromeliaceae

Tillandsia recurvata (15221) — certainly not as abundant as Morse and Steffens implied. All plants were found growing on rocks at higher elevations.

Agavaceae

Agave montserratensis — although it was impractical to collect specimens of this large Agave, a complete set of photographs allows this determination. Trelease’s monographic treatment may need revision, but his recognition of ioe) species in the Leeward Island area seems sound. The Redonda speci-

appears to be the same as the plants found on Montserrat. Clarke’s photograph (pl. 10, fig. 2) shows the characteristic habit and the long narrow leaves of this species which it is strange that earlier visitors did not notice.

Agave a several plants of this species were seen near the ruins of the mine headquarters, suggesting that the plant may have been introduced as an ornamental or a possible crop plant.

Casuarinaceae Casuarina ssid ts te (15233) —a single tree of this species was io near the use and is shown in Clarke’s illustration (pl. 10, fig. 2). The specimen 1s cichilate and no seedlings were to be found. The tree is a favorite nesting place for boobies, and all branches were heavily covered with guano.

Piperaceae Peperomia simplex (15222) — Dr. T. G. Yuncker kindly identified this material and indicated that it is commonly known as P. hamiltoniana Miq., which is

properly referred to P. simplex Ham. The present plants were abundant in damp areas under rocks and on the scree in gullies on the eastern side.

Moraceae Ficus ee (15247) — only three plants, all seedlings, were seen on Redonda. he largest grew in a crevice of the top-most boulder of the island, a rock ne en guano. Clarke’s photograph of this same rock, taken in 1958, does not show this plant. The other specimens were in unapproachable loca- tions on the steep western face of the island. Clearly the species has been introduced recently by birds.

Urticaceae

Pilea microphylla (15219) — this species occurred on rocks and the damp scree of the eastern gullies.

Amaranthaceae Amaranthus dubiuns (15245) —a weed near the old ae area. Centrostachya indica —a weed along the old path to the m Tresine angustifolia (15235) —a rampant herb among ees: iv the summit.

Nyctaginaceae Boerhaavia coccinea — on the plateau and on the western scree slope. Bouganvillea aan —a woody vine planted and persisting near the ruins of the manager's hous

1962 | HOWARD, OBSERVATIONS ON REDONDA 65

Portulacaceae Portulaca oleracea — a weed around the ruins and in the gully one ascends. Portulaca halimoides (15218) — this, and the following two species, occur primarily on the scree slopes of the western face of the island. Talinum triangulare (15223) Trianthema portulacastrum

Annonaceae ii squamosa —a small shrub persisting following planting near the man- ger’s house.

Capparidaceae Cleome viscosa (15241) — a most common weed.

Leguminosae Centrosema virginiana Galactia sp. (probably G. stricta) — occurring commonly near the ruins, but all of the plants seen were sterile. Tephrosia cinerea (15237) —an abundant herb found in many locations.

Rutaceae Citrus aurantiifolia (lime) —a shrub persisting after cultivation near the man- ager’s house

Euphorbiaceae Croton flavens (15240) — this species, Wedelia calycina and Lantana camara represent the most common broadleaved plants and dominant shrubs on the island. All of the plants, however, were of lower stature than the species ssumes on other islands. Many of the flat-topped shrubs served as nesting Bee for the birds, and these plants appeared to be trimmed around the

ests. craton lobatus (15244) — common herb in the western gully along which one cends.

Euphorbia hirta (15246) Euphorbia heterophylla Jatropha gossypifolia —a weed around the ruins and in the scree of the western

Phyllanthus amarus — a weed around the ru Ricinus commun Lael iden persisting ae cultivation; common in gullies on the west face of the islan

Malvaceae Sida cordifolia (15239)

Cactaceae meas royenii — columnar cacti common on precipitous western escarp-

eee intortus — found primarily on the sheer western face of the island. Opuntia antillana — a stout SAAN! of frequent occurrence.

Opuntia repens — “the sucker”; extremely abundan

Opuntia triacantha — less frequent than O. repens, and less easily fragmented.

66 JOURNAL OF THE ARNOLD ARBORETUM [VOL, XLII

Plumbaginaceae Plumbago scandens (15248) — abundant on the western escarpment.

Apocynaceae

Catharanthus roseus — occasional in mining area.

Asclepiadaceae Cynanchum parviflorum (15225) — comm mon vine on the plateau. A leafless form of ihis species hung in festoons from many rocks on the western face.

Verbenaceae elisa camara (15236) —an abundant shrubby apes: represented by the ange-yellow color form. All plants were spineles ne involucrata — found primarily at lower tees above the cliffs on the windward side. Stachytarpheta jamaicensis — infrequent. Labiatae Hyptis pectinata (15226, 15242) — frequent in occurrence, but usually browsed

Leonotis nepetifolia

Solanaceae Nicotiana tabacum — occasional plants presumably persisting after cultivation.

Scrophulariaceae Capraria biflora — a weed near the mine ruins.

Acanthaceae Justicia periplocifolia (15238) —a narrow-leaved form usually browsed into abnormal growth forms. Determined by E. C. Leonard.

Rubiaceae Oldenlandia corymbosa — occasional among rocks on the eastern slopes.

Compositae Ageratum deter sleet Emilia coccinea (15233A) Pterocaulon reali (15232) —a few individuals at the northern end of the island

Wedelia calycina — one of the common shrubs on the island.

After this article was set in type, I discov ne a eg > Harold Box entitled “A Note on the Vegetation of Redonda, B. W. dae . 77, 311-313. 1939. r. Box, sailing from Antigua, visited Redonda on rie i ate observations were made from the landing place, since he was unable to ce the cliff. A list of 27 species, including Talinum paniculatum and Lithophila muscoides which I did not encounter, is given as the lithophyte flora of the island.

1962 | GRAY, REVISION OF PODOCARPUS, XIII 67

A TAXONOMIC REVISION OF PODOCARPUS, XIII SECTION POLYPODIOPSIS IN THE SOUTH PACIFIC

Netta E. Gray *

SecTION PoLtypopiopsis was established in 1874 by Bertrand (1) for Podocarpus vitiensis Seemann, then known only from Fiji. In 1903, Pilger (23) included this species and P. minor Parlatore as species of doubtful affinity in sect. NaGetA, chiefly because of their opposite leaves. In 1926, he mentioned Bertrand section and species synonyms briefly but did not change his interpretation. Florin (8) recognized the significant dif- ferences of these species from others in sect. NAGEIA and restored the use of sect. Potypopropsis, to which he added Podocarpus rospigliosi Pilger, a South American species which had been described by then. This section now is firmly established in its use by Orr (22), Wasscher (32), and Buchholz & Gray (4).

The five living species now included in sect. PoLypopiopsis are Podo- carpus vitiensis Seemann, of Fiji, New Guinea, New Ireland, and the Solomon Islands; P. filicifolius sp. nov., newly described from the Moluc- cas; P. comptonii Buchholz and P, minor Parlatore, endemic to New Caledonia; and P. rospigliosii Pilger, of the Andes of Venezuela, Colombia, and Peru. Two fossil species (from Tertiary deposits) referred to this section are P. araucoensis (Berry) Florin, found in Chile, and P. brownei Selling, lately described from Tasmania.

This distribution is striking and becomes very significant in recent studies of regional floras (Smith 28), eaeee = (e.g., Van Steenis 30, Selling 26), conifer geography (e.g., Florin 9, Li 20), and phylogeny (Florin 14). An ancient vast southern eae land mass is empha- sized by Van Steenis (30) in his consideration of the angiosperm genus Nothofagus, which is not only found both in the South Pacific and South America, but which has a subsection of twenty-one species limited to New Caledonia and New Guinea. Some oi the land areas here included are southeastern Asia, Australia, New Guinea, New Caledonia, Fiji, and New Zealand. Since Podocarpus vitiensis was known first from Fiji, the section has tended to be associated mainly with those islands, but with the dis- covery of the closely related P. comptonii in New Caledonia and of P. filicifolius in the Moluccas, the New Guinea area of P. vitiensis becomes

* The author wishes to express her appreciation to Dr. Rudolf Florin, of the Bergianska ie ee Stockholm, Sweden, for critically reading the introduction to this paper, and to Drs. Florin and J. Tengnér, for examinaticn and evaluation of the wood anatomy of two critical specimens of Podocarpus minor. She also thanks Dr.

mith, of the Smithsonian Institution, for criticism of her Latin description of Bodvcurpus filicifolius.

68 JOURNAL OF THE ARNOLD ARBORETUM [ VOL, XLII

of primary importance, with the Fijian specimens then representing an eastern extension of the species. There is a remarkably close similarity to the distribution of Nothofagus species. One would expect it to be possible to distinguish two separate species of Podocarpus from New Guinea and Fiji, but the specimens I have seen from New Guinea seem the same, and only ripe seeds are lacking from New Guinea. The close relationship of Fiji to these land areas, geologically and botanically, is shown especially well in the recent studies of the Fijian flora by Smith (28, 29).

The postulation of a great south land mass, often called Gondwanaland, would connect not only the continental islands of the South Pacific, but also Antarctica, Australia, Asia, South America and Africa.

The leaf anatomy of the species of sect. AFROCARPUS, endemic to Africa, shows the same twisting of the short petiole that is found in sect. PoLypoptopsis, even when the leaves are alternate instead of opposite. The species of both sections have amphistomatic leaves and the leaves lack accessory transfusion tissue; in the leaf epidermis of species of both sections, the subsidiary cells of the stomata show the Florin ring. The South American species, Podocarpus rospigliosii, is very like P. vitiensis, except that its branches are covered only with foliage leaves and we find scale leaves or bracts only on the fertile shoots; the seed usually is some- what larger. Other gymnosperm genera with species in both South America and Australasia are Araucaria, Dacrydium, and Austrocedrus.

Florin (8, 11, 14) has concluded that Podocarpus is a southern genus which probably originated in the early Mesozoic, and we find fossil evi- dence in South America in Podocarpus araucoensis and in Tasmania in P. brownet, Florin’s new fossil genus (11) from Australia belongs to sect. STACHYCARPUS Of Podocarpus which has living species in South America, New Zealand, New Caledonia, and Australia. Buchholz (3) maintained the northern origin of all conifers.

The most significant change in recent gymnosperm systematics has been the separation of the taxads from the podocarps, leaving the latter with the conifers (Florin 10a,b, 12, 13, 14). Pilger (23), in his classical monograph of the Taxaceae, included the podocarps with the taxads. There was increasing indication that these should not be kept so close together, and, in 1926, Pilger arranged the gymnosperms in seven fami- lies with the Taxaceae and Podocarpaceae separated. This treatment was followed by Buchholz (2) in 1946, in spite of the growing realization that the taxads should not be included with other conifer families. Florin (14) compared these two groups morphologically in almost every way in which they could be investigated to uphold the elevation of a class Taxineae. This class has been based chiefly upon the evidence of the development of the female strobilus as found in fossils (Florin 10a,b, 13, 14). Wilde (33) limited her lengthy discussion to comparisons of the male and female strobili of many species of Podocarpus; she included many data, however, which were useful in deriving the interpretations and solutions given by Florin.

1962] GRAY, REVISION OF PODOCARPUS, XIII 69

Only the two sections, Potypopiopsts and Nace, of Podocarpus regularly have opposite leaves; in sect. AFROCARPUS some specimens have all or only occasional branches with leaves opposite or subopposite. But all three of these sections show the peculiar twisting of the leaf bases and stem torsion In both sections Potypopropsis and NacetA, this may result in the leaves being spread in a single plane, as in the compound frond of a fern. This orientation has been described in detail by Florin (8), Orr (22), Gray & Buchholz (17), and Wasscher (32). Orr (22) recog- nized further similarities in sections Potypopiopsis and AFROCARPUS in the leaf anatomy: the leaves are amphistomatic, with more or less hypo- derm, transfusion tissue often extending more than half-way from the midrib to the margin of the leaf, no accessory transfusion tissue, and a single resin canal in all species except P. rospigliosi. The three resin canals below the vascular bundle, with additional ones in the blade of the leaf of P. rospigliosii, is an exception discussed by Gray & Buchholz (17) in relation to reports in the literature (e.g., Bertrand 1, Stiles 31, Mahlert 21) that P. vitiensis has more than a single vascular resin canal. The single resin canal which I have found in all transverse leaf sections of P. vitiensis is in agreement with the findings of Gibbs (16) and Orr (22) who emphasize this fact. Neither have I seen any accessory trans- fusion tissue in the mesophyll of the leaf blades such as is shown by Bertrand (1, Pl. 6, fig. 12). No mention has been made thus far that the notable difference in the leaves in sect. NAGEIA is in the many parallel veins extending the full length of the large leaf blades, in contrast to the single unbranched midrib of the species of sect. PoLYPODIOPSIS.

Orr (22) used only three species in describing the leaf anatomy of sect. Potypopiopsis: Podocarpus vitiensis, P. rospigliosii, and P. minor. The external appearance of the foliage of the two former is quite similar, and to these we may now add P. filicifolius, which has foliage most like that of P. vitiensis. The external appearance of P. minor differs in not showing the pinnate arrangement of the leaves in a single plane except in the foliage of seedlings and on occasional lower branches; the branches are otherwise fully covered with crowded, opposite, decussate, ascending, oval or elliptic leaves. The discovery in New Caledonia of P. comptonii, with transitional foliage, affirms the judgment that P. minor really belongs to this section. Podocarpus comptonii is a large tree, recognized by Compton (5) and others. It generally has the foliage which has been described in such detail, although the reproductive branches lose the pinnate arrange- ment and are covered with decussate, opposite, oval, or elliptic leaves, as seen in P. minor.

The only section of Podocarpus which has both bifacially flattened lanceolate leaves and scale leaves in the mature foliage is PoLyPopiopsis. The two kinds of leaves in sect. DacrycarPus are scale leaves and the needle-like leaves on the pinnate twigs are tetragonal in transverse sec- tion. This remarkable dimorphism shown by the foliage on the main shoots and leafy branches was described by Florin (8) and Wasscher (32) for Podocarpus vitiensis, but it is also evident in the other species.

70 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLII

In his consideration of the external morphology of the coniferous leaves, De Laubenfels (6) described this section as having only Type II (bi- facially flattened) leaves. The scale leaves (De Laubenfels Type IIT) have been found only in connection with mature foliage. This combination is rare among other gymnosperms, in fact, having been found only in seed- lings of a Dacrydium; transitional foliage of Chamaecyparis, Neocalli- tropsis, and a Dacrydium; and mature foliage of Athrotaxis, Taxodium, and a group of Juniperus. Podocarpus vitiensis and P. filictfolius have scale leaves much as described by De Laubenfels, tapering sharply from the point of attachment and closely appressed to the stem in P. vitiensis. They are rarely crowded, except at the beginning of a growth period, but are usually spaced by internodes of about the same length as those between the foliage leaves. They are decurrent, and their arrangement is always decussate except at the one or two nodes immediately preceding distichous foliage leaves. Podocarpus comptonii and P. minor have some scale leaves of this kind but they become obtuse and ovate in shape, divaricate, often elliptic as they approach true leaves, even being abruptly narrowed at the base but not usually becoming more than 4 mm. in length. According to De Laubenfels, only in one group of Dacrydium do scale leaves follow juvenile leaves of his Type II, and, when scale leaves (Type III) are developed, they are never followed by any other leaf type. In sect. Poty- Poptopsis, however, I find that all variations are exhibited in the alterna- tion between scale leaves and foliage leaves. Main shoots may bear only scale leaves; a leafy branch which continues growth may first bear foliage leaves, then scale leaves. Leafy branches may have only one growth period with a dormant terminal bud or they may have at least one other burst of growth, as is evident in the leafy shoots where each growth period shows first an increase in leaf length to about the middle of the growth period, after which it again uniformly decreases; one to several pairs of scale leaves may separate the growth periods. The small- est of the leaves may be very similar to scale leaves and a pair of scale leaves is usually present in ?. comptonii and P. minor at the first node of a leafy branch. Scale leaves have been found only on the special repro- ductive branchlets of P. ros pigliosii.

The leaf anatomy, as seen in transverse sections of the foliage leaves, was described by Orr (22) for Podocarpus vitiensis, P. minor, and P, rospigliosui. The leaves are amphistomatic, and palisade parenchyma may be found on both sides or developed only on the side facing the light. The single vascular bundle is flanked by wings of transfusion tissue which sometimes extend fully half-way to the margin of the leaf. The extent of the transfusion tissue varies too much from one leaf to another for it to be used as a diagnostic character, but it is greatest in P. minor and P. comptonti. There is no organized accessory transfusion tissue, and iso- lated lignified cells with large lumina were detected only very rarely in the mesophyll, with none at all present in P. rospigliosii. I agree with Orr that P. rospigliosii has the greatest number of hypodermal fibers with often a continuous layer at the margin and midrib. In P. vitiensis and

pl

—

1962] GRAY, REVISION OF PODOCARPUS, XIII 71

P. filicifolius the hypodermal fibers are few, small, and often isolated even at the margin and midrib. In P. minor and P. comptonii the hypo- dermal fibers are larger and scattered, or grouped together, with the fewest in P. minor, except for an almost continuous layer at the margin. Vascular fibers are usually large and abundant above the midrib in all the species but are absent or rare below the midrib in P. vitiensis, P. filicifolius, and P. rospigliosii. The leaves of these latter species are thin, usually only between 0.3-0.6 mm. thick, those of P. comptonii are a little thicker (especially on fruiting branches), being from 0.5—0.8 mm., while P. minor has very thick leaves, from 0.6—-1.2 mm.

The female strobili are remarkably similar in sections PoLypopropsIs, NaceIA, and AFrocarpPus, even to the fertile bract adhering after the seed is separated from the axis at maturity. The detailed description of the ovule development of Podocarpus vitiensis by Gibbs (16) indicates a peduncle covered with imbricate scale leaves, succeeded by 6-10 bracts on the strobilus, the terminal one or two fertile. The portion of the pe- duncle covered by bracts might be designated as a woody receptacle, the uppermost 2—4 bracts subtending the ovule sometimes having definitely thickened adhering bases. This was not apparent to Gibbs in P. vitiensts, but that three sterile and one fertile bract unite to form a receptacle has been recognized in P. minor and P. comptonu. A thickened woody re- ceptacle is very apparent on the mature seeds of P. filicifolius, where an additional pair of bracts may become involved.

In her observations on the wood structure of Podocarpus, Kaeiser (19) found that of species in sect. Porypoptopsis usually like that of those in sections AFROCARPUS and NAGEIA.

In sect. Potypopiopsis, Hair & Beuzenberg (18) counted the chromo- somes of P. vitiensis, P. comptonii, and P. minor and found the 2m num- ber to be 20, If one considers ten to be the basic number, this count affirms the suggestion that this section and sect. NAGEIA, in which 27 = 2 in P. blumei, are the oldest groups in the genus.

Key To SPECIES OF SECT. POLYPODIOPSIS

A. All foliage leaves arranged pinnately on the twigs. Bee Mature. seed Crested «ae, <n ten eee were are) at P. rospigliosit.’ B. Mature seed not crested. Ce Mature: seeds. pear shaped.¢ 25 yeeros ek tic Ish vitrensts. €.- Mature seeds, spherical: 2. is.4 oye eee rn ed x: 2. P. filicifolius. A. Some twigs pinnately leaved, but fertile twigs not flattened, and always bearing decussate foliage.

D. Small trees or shrubs; foliage mostly decussate: twigs rarely flattened. ae 4, P. minor,

‘This species treated previously in Jour. Arnold Arb. 29: 118-122. 1948.

72 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLII

1. Podocarpus vitiensis Seemann, Bonplandia 10: 366. 1862, Jour. Bot. 1: 33. 1863, Fl. Vitiensis 266. 1865-73; Van Tieghem, Bull. Soc. Bot. Fr. 38: 169. 1891; Pilger, Pflanzenr. IV. 5(Heft 18): 63. 1903, Nat. Pflanzenfam. ed. 2. 13: 245. 1926; Gibbs, Jour. Linn. Soc. Bot. 39: 182. 1909, Ann. Bot. 26: 533. 1912; Stiles, Ann. Bot. 26: 455. 1912; Dallimore & Jackson, Handb. Conif. 58. 1923, 1931, 85. 1948; Florin, Sv. Vet-akad. Handl. III. 10: 275. 1931; Wasscher, Blumea 3: 425, 1941.

A large forest tree up to 43 m. high with trunk often over 1 m. in diameter (sometimes buttressed in New Guinea) and a crown of spread- ing branches. Young twigs decussate, usually in pairs at distances of 2.5—4 cm., spreading or erect-spreading, dimorphic; main twigs terete or slightly flattened below the ramifications, leafless, alternately bearing axillary decussate lateral leafy twigs and decussate, ovate to orbicular, deciduous bracts nearly 2 mm. long; lateral leafy twigs slender, usually unbranched, with the leaves usually smaller toward each end, decussate but turned in one plane, pinnately arranged, divaricate, the twigs up to 38 cm. long and terminated by small buds which may not develop. Ter- minal buds of main twigs globose or ovate, with decussate, ovate, orbicular, or obovate, obtuse scales, 1.5-2 mm. long. Leaves spread out in one plane by twisting at the base so that the adaxial sides are exposed on one side of the twigs and the abaxial sides exposed on the opposite side; leaf pairs 5-8 mm. apart; leaves sessile, lanceolate, rounded at the base and broadly decurrent, gradually narrowing toward the narrowly obtuse apices, am- phistomatic, 1.5—3 cm. long by 3—5.5 mm. broad; midrib sometimes promi- nent on either surface and with stomata in the epidermis above and below. Leaves in transverse section with scattered hypodermal fibers on both sides, palisade mesophyll on upper, lower, or both sides, depending on the orientation of the leaf; a single median vascular bundle with a single resin canal (never 3) abaxial to the phloem and flanked by wings of trans- fusion tissue; vascular fibers conspicuous above the vascular bundle and no accessory transfusion tissue. Male strobili solitary or clustered 2 or 3, terminal on axillary leafy twigs or bracted pedicels, or on bracted main shoots often branching once or twice so that strobili or fertile branches may arise in bract or leaf axils, strobilus cylindrical, 12-20 mm. long and 2—2.5 mm. in diameter; microsporophylls triangular, with acute or obtuse tips, margins scarious. Female strobili terminal on peduncles in axils of bracts or normal leaves; peduncles 2-8 mm. long, sometimes branched, covered with 6-10 pairs of imbricate scale leaves; strobilus of 6-10 bracts, 1 or 2 fertile; receptacle not differentiated. Seed bluish red when ripe, to 2 cm. long, obliquely attached to receptacle, obovate, narrowed towards the base which may still bear the attached bract when the seed has fallen, apex obtuse; hard-coated inner seed terminated at the micropylar end by a sharp, sometimes recurved point.

DistripuTion: In dense mixed forest, Fiji Islands, at altitudes between 100 and 900 m.; New Guinea, in the western part at altitudes of 1100-1200

1962 | GRAY, REVISION OF PODOCARPUS, XIII (es

m. and in the southeastern part at 1650-2000 m.; and Vanikoru, Solomon Islands.

Fiji Islands. Vir1 Levu: Mba, Nandarivatu, Degener 14483 es ISH, K, us),” 14485 (mo), 14496 (A, Ny), Mead, Sing. Field No. 1974 (x), Parks Bae (BIsH, tuc), Gibbs 674 (BM, K); valley of Singatoka River, Gillespie 3273.2 (BISH), 3712 (BISH, DS, tUC), 3865 (BISH, K, NY, UC, us), 4308.1, 4402.2 (BIsH); Tailevu, e. of Wainimbula River, Swih 7076 (us). VANUA LEVU: Thakaundrove, Yanawei River, Mt. Kasi, Smith 1796 (BISH, GH, NY, tUC, US). WirHovut Locatity: Horne 531 (GH, K), Seemann 576 (Holotype, K; GH), Graff (K). New Guinea. NETHERLANDS New GuINEA: Idenburg River, Ber- nard Camp, Brass 19534 (+a), 12787 (+a), 12787a (A); Cyclops Mts., Ver- steegh BW913 (Laz, +L); Papua: Alola, Carr 14160 (A, +Ny); Lala River, Carr 15666 (+A, BM). Solomon Islands. Santa Cruz Group, Vanikoru, near Lemon River, Walker B.S.J.P.212 (+a). Cultivated. AUSTRALIA: Bot. Gard. Sydney, Boorman, in 1908 (a), Eames, in 1937 (cu

Even though Podocarpus vitiensis has been recognized as a species since 1862, an understanding of its relationships within the genus is quite recent. On the basis of the single vascular bundle in the leaf, stomata on both sides of the leaves, and supposedly two lateral resin canals as well as the one below the vascular bundle, Bertrand (1) put this species in a separate section, Potypopropsis. The confusion brought about by Bertrand and Mahlert (21) in ascribing the three vascular resin canals has been dis- cussed previously (17) and referred to P. rospigliosii from South America which is now included in the same section.

Subsequently, Pilger (23) put Podocarpus vitiensis in his sect. NAGEIA on the basis of the opposite leaves which show the same characteristic turning of the leaf bases so that the lower surface is uppermost on one side of the twig and the upper surface on the other side of the twig when these decussate leaves orient themselves in the pinnate “fronds.” Pilger’s classification was followed by Gibbs (16) and Stiles (31) after Gibbs (15) had briefly considered sect. StacHycarpus. Florin (8) recognized the validity of Bertrand’s sect. Potypopiovsis for P. vitiensis as a sep- parate group and added P. minor. This section is now well established.

1 of these studies, including Gibbs’ (16) description of the female

° The following symbols indicate the location of the specimens cited: Arnold Arbo- retum, Cambridge (a); Bernice P. Bishop Museum, Honolulu (pIsH); Botanic Mu- seum and Herbarium, Brisbane (BRI); British Museum (Natural History), London BM); Jardin Botanique de |’Etat, Bruxelles (Br); Herbarium of the University of California, Berkeley (uc); Wie gand Herbarium, Cornell University, Ithaca (cu) ; Dudley Herbarium, Stanford University, Staniord (ps); Herbarium of the Univer- sity of Georgia, Athens (ca); Gray Herbarium, Cambridge (cH) ; University of Illi-

Rijksherbarium, Leiden (1); Missouri Botanical Garden, Saint Louis (mo); Depart- ment of Forests, ee and New Guinea, Lae (Laz); New York Botanical Garden, New York eM uséum National aonictoie Naturelle, Paris (Pp); United States National Museum, Washington (us); National Herbarium of Melbourne, Melbourne (MEL). A dagger (+) preceding the abbreviation of an herbarium signifies that the details of the ee of this specimen have been examined in transverse section.

74 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIII

strobilus, were on Fijian specimens of Podocarpus vitiensis. In 1944, Wasscher (32) recognized specimens of this species from New Guinea and, for the first time, described the dimorphic foliage which had been aveooked in the Fiji material. It is interesting that the early specimens (Seemann 576, Graff s.n., and Gibbs 674) all have only terminal male cones on leafy twigs, wheueds recent male specimens (Degener 14496, Parks 20653, Smith 7076) definitely show the dimorphic foliage, main twigs with scale leaves, and lateral leafy branches bearing male strobili or fertile branches bearing scale leaves and strobili or axillary branches bearing strobili.

New Guinea specimens may still prove to represent a separate species, but the half-matured ovules which I have seen are still pear-shaped and obliquely attached to the receptacle, as in the Fijian Podocar pus vitiensis.

Both Seemann (25) and Gibbs (16) have waxed ene on the beauties of this species and I must quote from Seemann: “This is one of the finest Coniferae I have ever seen. . . It attains sixty feet in en has a stem nine feet in circumference, and has drooping, extremely grace- ful branches, which would render the species a highly desirable acquisi- tion to our living collections.”

2. Podocarpus filicifolius, sp. nov.

Arbor 15 m. alta ramulis numerosis, divaricatis, spiralibus vel oppositis; alabastris in ramulis principalibus parvis, late ovatis, squamis paucis, late triangularibus decussatis, 1.5—2 mm. longis;: alabastris i in ramulis foliiferis permutatis, ovatis, plerumque constanter dormientibus; foliis dimorphis; squamis in ramulis principalibus, divaricatis, tenuibus; late triangularibus,

mm, longis, acutis vel obtusis; foliis ad ramulis pinnatis terminalibus vel axillaribus, sessilibus, divaricatis, oppositis, 6-24 mm. longis, 3.5—4.5 mm. latis, nceslane tenuibus, planis, apice late acutis, basi fate rotun- datis, decurrentibus, costa non manifesta; strobilis masculis ignotis; stro- bilis femineis solitariis (?), pedunculis 10 mm. longis, squamas 4— 6 binas decussatas vel cicatrices gerentibus; receptaculo parvo lignoso bracteis oppositis 2 late obtusis subtento, 7 mm. longo, 3 mm. lato, bracteis 3 coalescentibus equalibus, apice liberis composito, bractea unica fertili; semine maturo globoso, 1.5—1.8 cm. longo et lato, apice aequaliter rotundo, obtuso.

DISTRIBUTION: Morotai, in the Moluccas. Moluccas. Morotai, A. Kostermans, in 1949 (Holotype, +L).

This tree differs from Podocarpus vitiensis in the spreading scale leaves, thinner foliage leaves, the distinct receptacle 7 mm. long supporting the seeds, and the spherical seeds. The dissected seed shows a smooth, brown outer coat 0.3 mm. thick, a hard, woody, light tan middle layer which is pointed at the micropyle, and a thin papery brown inner layer.

The name refers to the fern-like appearance of the pinnately leaved twigs, reminiscent of the royal fern, Osmunda regalis.

1962 | GRAY, REVISION OF PODOCARPUS, XIII 75

oy pe uiee Reais aie Buchholz, Bull. Mus. Hist. Nat. Paris II. 284. Guillaumin, Acta Horti Gothob. 19: 8. 1952; Che- he pes Mélanésiennes II. 1: 114. 1956.

A tree becoming 10-12 m. high with trunk up to 80 cm. or more in diameter, bearing few branches below but profusely branched above; bark brownish, longitudinally furrowed; young twigs green and showing decur- rent leaf bases; branches erect or ascending, usually opposite while still young; terminal buds appearing naked but protected by special scales which become part of foliage but do not grow into leaves. Foliage di- morphic; scale leaves on main twigs decussate, broadly triangular, de- current, 2 mm. long, becoming obtuse, then ovate-elliptic, divaricate, 3—4 mm. long; 1 pair opposite scale leaves usually at first node of leafy twigs; leaves decussate, emerging in 4 ranks from the twigs and, on vigorous branches, twisted at the base so that the pinnately leaved twigs bear opposite leaves, on one side all with the abaxial side exposed and on the other side the adaxial; on upper reproductive branches of mature trees, leaves held obliquely in 4 vertical ranks; leaves lanceolate to elliptic, obtuse, 7-15 mm. long, 2-5 mm. wide, thick, sessile, with decurrent leaf bases; single midrib not usually distinct, with the surface becoming dry in many longitudinal wrinkles. Usually only foliage leaves on small vigorous plants and seedlings becoming 30 mm. long, 6 mm. wide, very thin, acutish, and opposite on pinnate twigs. Leaves differentially amphistomatic, vary- ing with the orientation; transverse sections showing hypoderm inter- rupted by the stomatal rows, single vascular bundle with single resin canal below phloem, transfusion tissue well developed, no accessory transfusion tissue, vascular sclereids abundant above the bundle with fewer below; palisade mesophyll on either or both sides depending on leaf orientation. Male strobili sessile and terminal on lateral twigs usually bearing regular foliage leaves, subtended by a pair of narrow, short foliage leaves or bracts; strobili ovoid, 5-6 mm. long, 3 mm. wide; microsporophylls broadly triangular, apices acute, margins thin and somewhat erose. Female strobili terminal on lateral twigs which may bear ordinary foliage leaves or spreading scale leaves; each strobilus bearing 3 or 4 pairs of small, opposite, sterile bracts on an axis 10 mm. or more long, with the two upper bracts unequal and forming a nonfleshy receptacle, one of the bracts fertile. Seed obovoid, 25 mm. long, 18-20 mm. diameter, fleshy, red (?) when ripe; kernel with two ridges to a straight or curved beak at the micropyle.

DISTRIBUTION: In forests on mountain slopes of New Caledonia above 1000 m. altitude.

ew Caledonia. Mt. Ignambi, Compton 1524 (+pm), 1587 (+BM); Mt. Mou, Vieillard 1275 (+p), 3064 (+cH), 3264 —) Bitiel 1085 Grint f4 52) 1474, 1578 (1LL), 1684 (Holotype, +1LL), 7697 (1LL), MacDaniels 2323 (cv), Virot in 1938 (a), White 2033 (+a, BRI), a Laubenfels P129 (GA), a 607 (pm); Mt. Humboldt, Schlechter 15331 (+pBr, +BM), Mt. Dzumac,

Laubenfels 153 (GA), Montagne des Sources, De Laubenfels 123 (Ga); Koe,

76 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLII

Balansa 184 (+P); Plain des Lacs, Mons du Mai Forest, Buchholz 1350, 1350a,

LL); Bai des Piroques, White 2120 (+A, 2 sheets); beside River Blanche, Ingle I.66 (+MEL); River Tane & Bourail, Balansa 1385 (x). No specified locality, Buchholz 1539a (+mo).

The twig dimorphism observed and described by Wasscher (32) for Podocarpus vitiensis is likewise very apparent in P. comptonii. The plant bears several kinds of foliage: definitely acute triangular scale leaves on vigorously growing main shoots; ovate or elliptic scale leaves on main shoots and lower parts of twigs; true foliage leaves lanceolate and ar- ranged pinnately; and elliptic foliage leaves arranged on the twigs de- cussately. Variations, such as twigs bearing alternately scale leaves and foliage leaves for three growth periods with no branching, or branches bearing foliage leaves becoming main branches but subsequently bearing only scale leaves of either kind, are numerous.

This species has long been confused with Podocarpus minor which is a small tree along streams and in swamps at lower elevations. Compton (5), who strangely saw only the large trees, remarked on the inappro- priateness of the name “minor.” All the specimens he assigned to P. minor belong here. The specimen White 2120 from the Arnold Arboretum was labeled P. vitiensis, showing that it was recognized as being different from P. minor, Dr. J. T. Buchholz collected both species abundantly and the differences became clear to him. The specimen Compton 1273, listed as an undetermined species, belongs here.

-

Podocarpus minor (Carriére) Parlatore in DC. Prodr. 16(2): 509.

1868; Brongn. & Gris, Bull. Soc. Bot. Fr. 16: 326. 1869; Mahlert, Bot. Centralbl. 24: 281. 1885; Pilger, Pflanzenr. IV. 5(Heft 18): 62. 1903, Nat. Pflanzenfam. ed. 2, 13: 245. 1926; Schlechter, Bot. Jahrb. 38: 16. 1907; Compton, Jour. Linn. Soc. Bot. 45: 425. 1922; Dallimore & Jackson, Handb. Conif. 51. 1923, 1931, 75. 1948; White, Wilson & Guillaumin, Jour. Arnold Arb. 7: 78. 1926; Florin, Sv. Vet-akad. Handl. III. 10: 278. 1931; Guillaumin, Acta Horti Gothob, 19: 8. 1952; Chevalier, Etudes Mélanésiennes II. 1: 114 LOS i;

Nageia minor Carriere, Traité Gén. Conifeéres. ed. 2. 641. 1867. Podocarpus palustris Buchholz, Bull. es Hist. Nat. Paris II. 21: 284. 1949; “hevalier, Etudes Mélanésiennes II. 1: 114. 1957.

A dwarf tree or shrub, 2—3 m. high, trunk 15-30 cm. in diameter with numerous ascending branches; bark rough, brown or dark gray; twigs opposite on young stems or scattered, ridged by decurrent leaf bases; terminal buds normally small with few opposite scales, frequently abnor- mally enlarged due to infestation by insect larvae. Leaves dimorphic; scale leaves on main shoots often deciduous, decussate, triangular-obtuse, keeled, decurrent, 1 mm. long, becoming divaricate: elliptic, thick, to 4 mm. ions: and usually 1 pair of elliptic scale leaves at the first node of leafy branches; foliage leaves crowded on short branchlets, opposite, de-

1962] GRAY, REVISION OF PODOCARPUS, XIII Ue

cussate, ascending and held obliquely in 4 vertical ranks, oval or elliptic, obtuse, sessile, 10-20 mm. long. 3.5—-5.5 mm. wide, thick and longitudinally wrinkled when dry, with broadly decurrent bases; midrib not evident. Leaves differentially amphistomatic with twice as many stomata on the adaxial side, stomata easily seen under low magnification on young leaves as minute white dots. Transverse sections of leaves showing rare hypo- dermal fibers on both sides but abundant at the margins, palisade meso- phyll on both sides, a broad midvein due to extension of transfusion tis- sue half-way to the margin, single resin canal, usually abundant vascular sclereids above and below the vascular bundle. Male strobili terminal on short, bracted peduncles, 1-1.5 mm. long, in clusters of 3-5 or more at ends of short lateral leafy twigs; 5-8 mm. long, 2.5-3 mm. in diameter; microsporophylls broadly hastate with narrowly acute upturned apiculi. Female strobili terminal on lateral twigs, axis 4 mm. long; strobili of 3 or 4 pairs of decussate bracts 1-2 mm. long; immature ovules usually crested, long-pyriform; 1 terminal fertile bract (rarely 2), keeled, 2 mm. long and spread upon the back of the base of the large seed (away from the micropyle) with minute or suppressed apex, the 2 upper bracts fusing into a small receptacle 4 mm. long. Seed crested, obliquely conically pyri- form, 2.8 cm. long, 1.7 cm. wide, becoming maroon-red when ripe; inner woody layer usually obtuse, with a beak formed at the micropyle, usually straight.

DistrIBUTION: Along banks of streams and lakes on Plaine des Lacs and near Prony Bay, New Caledonia.

New Caledonia. Borders of Lac Arnaud, Vieillard 1275 (Holotype of P. minor, +P; BM, ¢{NOUMEA), Vieillard or Deslanche 170 (+P); Plaine des Lacs, River des Tacs Beas 1719 (tILL), 1729 (+1LL), LeRat 607 (+BM), 22 ent Sta- tion, Buchholz 1347 (t1LL), 1348 (1LL), 1421 (Holotype of P. palustris, +IL1; P), flooded lake shore, headwaters of Yate R., De Laubenfels P112 (cA), Rio des Pirogues, White 2261 (srr, tuc, us); Lac en Huit, De Laubenfels P115 (Ga), McKee 3382 (us); river sw. of Grand Lac, Virot 658 (+A), MacDaniels 2544 (t+cu); Prony, Frank 207 Ser. A (+BM, BRI, tF, UC, US); Baie du Sud, Viedllard 1275 (a, BM). Without specific locality: Levormand 9171 (+k), Raoul s.n. (+p), Petit N138 (+p), LeRat 1752 (+P).

At some sites these dwarf trees grow very slowly and become very. old, their bases often growing in standing water and becoming buttressed. At times they may be entirely submerged beneath muddy, silt-laden water, reddish from the ferruginous soil carried down from the surrounding moun- tains during heavy rains. Dr. Buchholz (ms. data) believed there to be a related species, Podocarpus palustris which he described, but the dif- ferences he listed do not fall outside the normal range of variation in P. minor, and the difference in wood density is no more than expected from the slightly different ecological habitats.

LITERATURE CITED

1. BerTRAND, C. E. Anatomie comparée des tiges des feuilles les Gnétacées et les Coniféres. Ann. Sci. Nat. Bot. V. 20: 5-153. 1874

wn fo

a

“I

bo No

BRO bo ios}

JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIII

BucuuHoLz, J. T. Gymnosperms. Taxonomy of Coniferales. Encyclo- paedia Bans 11: 22-34.

: ric and subgeneric distribution of the Coniferales. Bot. Gaz.

110: 30-91. 1948,

E. Gray. A taxonomic revision of Podocarpus I, The sec-

tions of ‘the genus and their eae with special reference to leaf anatomy. Jour. Arnold Arb. 29: 49-63. 1948.

Compton, R. H. A systematic account of the plants collected in New Caledonia and the Isle of Pines. Part II. Gymnosperms. Jour. Linn. Soc. Bot. 45: 421-434. 1922

De LAUBENFELS, D. J. The external morphology of coniferous leaves. Phytomorphology 3: 1-20. 1953.

FLortn, R. Untersuchungen Stammesgeschichte der Coniferales und Cor- daitales. Part I. Morphologie und Epidermisstruktur der Assimilations- organe bei den rezenten Coniferen. Sv. Vet-akad. Handl. III. 10: 1931.

. Die Koniferen des Oberkarbons u. des unteren Perms. Paleonto- ae $5(B1 7): 457-654. 1938-44.

Tertiary fossil conifers of south Chile and ra phyto- _geoeraphical significance. Sv. Vet-akad. Handl. III. 19(2). 19

———. On the relationships of the Taxaceae. Bot. Gaz. 110: A 39. 1948,

sant in Cordaites and conifers. Acta Horti Berg. 15: 285-

388. 1951.

. On two conifers from the Jurassic of southeastern Australia.

Paleobotanist 1: 177-182. 1952.

———. The female reproductive organs of the conifers and taxads. Biol. Reviews 29: 367-389. 1954.

he systematics of the gymnosperms. Pp. 323-403 in A Cen- ae a Progress in the Natural Sciences, 1853-1953. Calif. Acad. Sci.

955

———. Notes on the systematics of the Podocarpaceae. Acta Horti Berg. 17: 404-11. 1958.

Grpps, L. S. A Sage to the montane flora of Fiji. Jour. Linn. Soc. Bot. 39: 130-212. 19

—. On the developmen of the female strobilus in Podocarpus. Ann.

Gray, N. E. & J. T. a CHHOLZ. A taxonomic revision of Podocarpus, III. The American species of Podocarpus: section Polypodiopsis. Jour. Arnold Arb. 29: 118-122. 8.

Harr, J. B. & E. J. Beuzenperc. Chromosomal evolution in the Podo- carpaceae. Nature 181: 1584-1586. 58.

KarIser, M. Microstructure of wood of Podocarpus. Phytomorphology 4: 39-47, 1954.

Lr, H. L. Present distribution and habitats of the conifers and taxads. Evolution 7: 245-261. 1953.

Mauntert, A. Beitrage zur Kenntnis der Anatomie des Laubblatter der Coniferer mit besonderer Beriicksichtigung des Spalt6ffnungs-Apparates. Bot. Centralbl. 24: 278. 1885.

Orr, M. Y. The leaf anatomy of Podocarpus. Trans. Proc. Bot. Soc. Edin- burgh 34: 1-54. 1944.

PiLcer, R. Taxaceae. Pflanzenr. IV. 5(Heft 18): 1-124. 1903.

Podocarpaceae, Nat. Pflanzenfam. ed. 2. 13: 211-249, 1926.

1962 | GRAY, REVISION OF PODOCARPUS, XIII 79

Sib Sy

SEEMANN, B. Flora Vitiensis 266-7. 1865-73. SELLING, O. H. Further studies in Schizaea. Sv. Bot. Tidskr. 41: 431-450. 1947.

———. Some Tertiary plants from Australia. /bid. 44: 551-560. 1950. Smitu, A. C. The vegetation and flora of Fiji. Sci. Monthly 73: 3-15. 1951.

erogam genera with distributions terminating in Fiji. Jour. Arnold eee 36; 273-292. 1955. STEENIS, C. G. G. J. vAN. Results of the Archbold Expeditions. Papuan Nothofagus. Jour. Arnold Arb, 34: 301-374. 1953. Strives, W. The Podocarpaceae. Ann. Bot. 26: 442-514. 1912. WasscHER, J. The genus Podocarpus in the Netherlands Indies. Blumea 4: 359-481. 1941.

. Witz, M. H. A new interpretation of coniferous cones I. Podocarpaceae. 1944.

Ann, Bot. IT. 8: 1-41.

AGNES ScoTT COLLEGE,

DECATUR, GEORGIA

80 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIII

TAXONOMIC AND NOMENCLATURAL NOTES ON ZANTHOXYLUM AND GLYCOSMIS (RUTACEAE) !

GEORGE K. Brizicky

IN THE COURSE of surveying the genera of Rutaceae in the southeastern United States (see Jour. Arnold Arb. 43: 1-22. 1962), the author en- countered several nomenclatural and taxonomic problems in Zanthoxylum L. and Glycosmis Corréa which require further comment. These include the generic limits of Zanthoxylum; the legitimacy of the name Z. cori- aceum A. Rich., in view of the existence of a supposedly earlier homo- nym; and the correct name of the type species of Glycosmis. The prob- lem of Z. coriaceum has led to further bibliographic research on the dates of publication of Achille Richard’s work on the flora of Cuba in Ramon de la Sagra’s Histoire Physique, Politique et Naturelle de VIle de Cuba. These items are dealt with separately below.

THE GENERIC LIMITS OF ZANTHOXYLUM

The nomenclatural confusion concerning Zanthoxylum L. and Fagara L. seems to have been cleared up by the typification of the former by Z. fraxineum Willd. (= Z. americanum Mill.) by Fosberg (1959). (It is notable in this connection that Jussieu (1825, p. 505) and Triana and Planchon (1872, p. 310) indicated Z. fraxineum Willd. as the type of Zanthoxylum L. (“Zanthoxylum Colden—L. J.— Schreb.’”’).) As the matter stands at present, the name Zanthoxylum L. has to be applied either sensu stricto to the genus with one perianth whorl (a simple or haplochlamydeous perianth, according to Engler, or one composed of petals, according to Eichler), or sensu lato to the combined genus, in- cluding Fagara L. (type, F. Pterota L.) with two perianth whorls (a double or diplo- and heterochlamydeous perianth). The recognition of two separate genera or of a single inclusive genus remains a matter of per- sonal evaluation of the evidence. For the students of tropical African, South American, and West Indian floras the generic status of Fagara apparently does not seem questionable, since no species of Zanthoxylum L. sensu stricto have been recorded from these regions. But students of the floras of eastern and southeastern Asia and North and Central America, where species of both these taxa occur, face the problem of the recognition

* Continuing a series of miscellaneous notes and papers on the flora of the south-

presses his sincere gratitude to Dr. Carroll E. Ait. d, Jr., for his Gaitical reading of the manuscript, valuable suggestions, and advi

1962 | BRIZICKY, ZANTHOXYLUM & GLYCOSMIS 81

of Fagara as a segregate genus. Opinions have differed. Thus, Rehder (1945, p. 73) in his study of Asiatic species of Zanthoxylum and Fagara came to the conclusion that “the two genera are close and none of the characters are strong enough for generic separation, so it seems preferable

A different view is represented by Reeder and Cheo (1951, p. 68) who say, “After studying numerous specimens in this complex, it is our feeling that both these genera are worthy of recognition. Although there are no striking vegetative differences, flowering specimens are quite distinct. Accordingly we are accepting both Xanthoxylum and Fagara, an interpretation which is, we believe, in harmony with that of most modern students of the group.”

Such disagreément in regard to the generic status of Fagara is based on differences in views as to the morphological nature of the perianth in Zanthoxylum. The earliest view, that formulated by Linnaeus in his generic descriptions, that the flowers of Zanthoxylum are apetalous while those of Fagara have sepals and petals, was followed by all taxonomists up to 1878 (e.g., Humboldt, Bonpland & Kunth, 1823; De Candolle, 1824; Jus- sieu, 1825; Triana & Planchon, 1872; Engler, 1874). This interpretation of the perianth favored considering Fagara a subgenus of Zanthoxylum, as was done by Triana and Planchon (1872), as well as by Engler (1874). Although maintaining the Linnaean view on the perianth of Zanthoxylum, Jussieu (1825, p. 505) noted, “Sepala interdum plura, sex aut etiam (teste Kunth) novem. Quorum analogia cum petalis confirmatur situ alterno staminibus ovariisque opposito, praetereaque metamorphosi ipsorum in stamina non infrequenti (observante C. Richard). These features of the flowers spoke against the Linnaean interpretation of the perianth leaves of Zanthoxylum as homologous with the sepals, but Jussieu’s note was disregarded or overlooked by his contemporaries.

In 1878, Eichler (p. 323), like Jussieu, taking into consideration the alternate position of the stamens and the opposite position of carpels (in the case of isomery) in respect to the perianth parts (leaves) in Zan- thoxylum americanum (“Z. fraxineum”), concluded that the perianth leaves in this species are homologous with the petals and the flowers are asepalous. Eichler’s view, accepted and followed by most American taxono- mists, also was in favor of the inclusion of Fagara in Zanthoxylum as a subgenus or section of the latter.

Engler (1896, 1931), however, after reconsideration of his previous (Linnaean) view, came to the conclusion that the perianth of Zanthoxylum is simple (haplochlamydeous), not homologous with the sepals, of a primi- tive type which occurs in Rutaceae only in this genus, and that it cannot be derived from the double (diplo- and heterochlamydeous) perianth of Fagara. Consequently, Zanthoxylum and Fagara should be recognized as distinct genera. This view has been adopted by many taxonomists, espe- cially in Europe. Engler’s interpretation of the perianth in Zanthoxylum is objectionable on at least two bases

First, there is no evidence that the simple perianth of Zanthoxylum

82 JOURNAL OF THE ARNOLD ARBORETUM [ VOL, XLUI

really represents a primitive condition which is not the result of sim- plification. On the contrary, Saunders (1934, p. 660, figs. 37-39, & p. 661), analyzing the carpellate flowers of Z. planispinum Sieb. & Zucc. (= Z. alatum Roxb. var. planispinum Rehd. & Wilson), treated its eight- merous perianth as consisting of four sepals and four petals. She also noted that ‘“exsertion of the two median sepals is delayed until after that of petals.” The latter observation, if confirmed by more abundant ma- terial, may also suggest some complexity in ontogenetic development of the perianth expressed in delay, or perhaps discontinuation of develop- ment, of some or all of the sepals (or their homologues). In his considera- tions of phylogeny of the Rutaceae, based on Saunder’s floral anatomical data, Moore (1936, p. 321) stated that his type “A” of vascularization of the floral parts (including Zanthoxylum fraxineum Willd., and appar- ently Z. planispinum Sieb. & Zucc., the species studied in greater detail by Saunders) ‘is more than likely the result of reduction.”

Second, the occurrence in Mexico and Central America (perhaps also in South America) of species of Zanthoxylum which appear to be transitional to Fagara in their perianth structure supports Eichler’s interpretation of the perianth in Zanthoxylum, rather than that of Engler. Thus, the peri- anth of carpellate flowers of Z. ferrugineum Radlk. (Donnell Smith 6468) from Costa Rica is described by Radlkofer (in Smith, 1897, p. 392) as, “nerianthii foliola 9-10, linearia (2-3 mm. longa), alia (plus minus con- spicue exteriora et sepalis respondentia) breviora et angustiora, alia (sub- interiora, reliquiis subalterna) longiora et paullulo latiora, omnia tenuiter membranacea.” The staminate flowers of Hinton et al. 10136 (GH) from Mexico (identified at Kew as Z. ferrugineum Radlk., but perhaps repre- senting a different species closely allied with the latter) possess five to ten, often eight to nine, perianth leaves, all similar in appearance, with one to five -+_ exterior and smaller than the others. The four to five stamens usually are opposite the smaller and alternate with the larger perianth leaves. (One five-merous flower which the author examined cor- responded exactly with the Fagara type with only the difference that the minute sepals were of the same appearance as the petals.) Almost similar conditions were found in staminate flowers of Z. mazatlanum Sandw. (iso- type, Gonzales Ortega 5210, GH), from Mexico, with 4—11-merous perianth and four to six stamens, and Z. Williamsii Standl. (isotype, A. Molina 1078, Gu), from Honduras, with four to eight perianth leaves and usually four stamens. A few carpellate flowers (fruits) of 7. Williamsi exhibited six or seven persistent perianth leaves. It is also notable that Engler (1874, p. 180), having described Zanthoxylum ciliatum as a new species from Venezuela, remarked, ‘Species valde insignis et cum nulla alia Austro-Americana adhuc descripta confundenda. . . . Characteribus suis transitum inter Zanthoxvlum et Fagara efformat.”’ The present author’s conclusion from the above is that the “simple” pcnanth of Zanthoxylum is most likely a secondary condition, derived by reduction from that of the Fagara type by abortion of some or all the sepals. The occurrence of species of Zanthoxvlum which appear in their perianth structure to

1962 | BRIZICKY, ZANTHOXYLUM & GLYCOSMIS 83

be transitional to Fagara not only supports this view, but also is ample reason to regard Fagara as a subgenus of Zanthoxylum.

ZANTHOXYLUM AMERICANUM MILL. VERSUS ZANTHOXYLUM FRAXINEUM WILLD.

Fosberg (1958, 1959), when discussing the problem of typification of Zanthoxylum, introduced an element of confusion into the nomenclature of the type species through the suggested replacement of Z. americanum Mill. (1768) by its later synonym, Z. fraxineum Willd. (1796). He reasoned, “Furthermore, according to Miller’s description, his X. ameri- canum could not be what is now called that as he says of it ‘the flowers grow in loose panicles as on the first sort’ rather than in axillary fascicles which is the case with the modern Z. americanum Mill. Lawrence, when photographing Miller’s types in the British Museum in 1950, was unable to locate a specimen of Xanthoxylum americanum so we have no way of knowing what plant this name actually refers to. . . . As pointed out in Taxon 7(4): 95. 1958, the earliest available name for this is Zanthoxylum fraxineum Willd. (1805).” A number of items argue against Fosberg’s proposition, however.

The general description of Xanthoxylum americanum given by Miller and his mentioning both the natural occurrence of the species in Penn- sylvania and Maryland and its resistance to cold seem to show clearly that Miller’s name referred to the plant of the northeastern United States, 1.e., to the only species of Zanthoxylum which occurs north of Virginia and Arkansas. Furthermore, Aiton (1813, pp. 382, 383) mentioned only three species of Zanthoxylum which were introduced and cultivated by Miller in the Botanic Gardens at Kew. These were “Z. emarginatum Willden.

. Cult. before 1739, by Mr. Philip Miller. Mill. dict. vol. 2 addenda, Lauro affinis 2.0; “Z. Clava Herculis Willden. . . . Cult. 1739, by Mr. Philip Miller”; and “2. fraxineum Willden. . . . Cult. before 1759, by Mr. Philip Miller. Mill. dict. ed. 7. n. 2.” The last quotation indicates that the plant grown at Kew as Z. fraxineum Willd. was that introduced by Miller and described by him in the seventh edition (1759) of his Gardeners Dictionary as “Xanthoxylum no. 2.” Since the latter, in turn, was the same entity as Xanthoxvlum americanum of the eighth edition (1768) of Miller’s work (the corresponding descriptions in both editions are identical), the conspecificity of the latter species with Z. fraxineum Willd. appears to be unquestionable. The circumstance that the type specimen of X. americanum has not been found among Miller’s types but that the species was introduced into the Botanic Gardens at Kew prior to the appearance of the seventh edition of Gardeners Dictionary makes it probable that the description of this species was made from the living plant.

In 1771, Du Roi (pp. 57, 58) mentioned the species as Xanthoxylum americanum Mill., and provided it with a rather detailed Latin diagnosis, including the correct description of the inflorescence and staminate flowers

84 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIII

(which he erroneously took for bisexual). He, however, regarded X. amerit- canum as a variety of Zanthoxylum Clava-Herculis saying, “Cl. Millero species, mihi varietas Zanthoxyli Clavae Herculis L.” In 1772, Du Roi (pp. 511, 512) again treated Miller’s species in the same way, but the description was given in German. Wangenheim (1787, p. 116) treated Xanthoxylum americanum as a species, noting that perhaps it should be considered a variety of Z. Clava-Herculis L. Neither Du Roi nor Wangen- heim had any doubts regarding the entity named and described by Miller.

Willdenow, in describing Zanthoxvlum fraxineum (1796, p. 413), ap- parently was aware that his new species was identical with that of Miller, since he mentioned ‘“‘Zanthoxylum Clava Herculis Du Roi” and ‘“Zan- thoxylum americanum Wangh.” in the synonymy and also noted that “Der Herr von Wangenheim und Miller geben dieser Art ungezahnte Blattchen.” Schkuhr (1803, pp. 467, 468, pls. 323, 3236), although he called the species Z. fraxineum Willd., remarked (p. 466) that this was the species which ‘‘auch Miller, Wangenheim und andere schon unter Zanthoxylum americanum bemerkt haben.”

Since the entity described by Miller as Xanthoxylum americanum is known, neither the occurrence of an error in the description nor the lack of a type specimen makes this binomial illegitimate. Therefore, its re- placement by Z. fraxineum Willd., as suggested by Fosberg, appears un- justifiable in the light of the present Code of Botanical Nomenclature.

ZANTHOXYLUM CORIACEUM A. RICHARD AND ITS DATE OF PUBLICATION

The legitimacy of Zanthoxvlum coriaceum A. Richard in Ramon de la Sagra (Hist. Phys. Polit. Hist. Nat. Cuba. Bot.-Pl. Vasc. [Essai FI. Cuba 1.] 326. pl. 34) is often questioned because of the existence of a supposedly earlier homonym, Z. coriaceum (Desv.) Walpers, Repert. 1: 521. 1842 (“‘coriacea”). The former name refers to the species of south- ern Florida and the West Indies (absent from Jamaica and a few other islands); the latter is a synonym of the Jamaican Z. spinosum (L.) Sw. According to Urban (1894, p. 563) and Kuntze (1898, p. 162) the French edition of the Richard work appeared in 1845 and preceded the Spanish edition which bears the same year of publication. The publication date of the portion of this work that included Z. coriaceum A. Rich. was presumed by Wilson (1911, p. 185) to be 1842 and by Little (1953, p. 437) to be 1842 or 1843. Finally, Van Steenis-Kruseman (1960, p. 741), with refer- ence to data given by Grisebach (1841, 1842, 1847, 1852) in connection with this work, stated, “Up to and including 1840, 28 parts were issued; in 1841 11 others followed. At that time the Ist part of the cellular crypto- gams was out and the phanerogams were published up to the end of the Thalamiflorae (Cand. Syst.). In 1846 parts 1-54 had been published, probably no other parts appeared. The atlas, dated 1845 too, was not finished before 1851 (cf. Wiegmann’s Repert. I.c. 1852, 387).” This statement, however, is not entirely clear and needs further explanation.

1962 | BRIZICKY, ZANTHOXYLUM & GLYCOSMIS 85

The Richard work consists of 42 signatures (printed sheets or Bogen) of 16 pages each, except the last with only 7 printed pages. Similarly, Montagne’s Plantes Cellulaires, of the De la Sagra work, contained 35 such signatures, the last one of 15 pages. Usually several signatures formed an installment (a fascicle or an issue, Lieferung, livraison). It is un- known whether Grisebach’s ‘“Lieferungen” translated as “parts” by Van Steenis-Kruseman should be understood as installments (improbable be- cause of their high number, e.g., 53 by 1846) or referred to the printed signatures (Bogen).

Richard’s work was most likely published irregularly in relatively few, perhaps unequal, installments consisting of at least several signa- tures. Thus, a notice in Hooker’s London Journal of Botany (1: 308- 311), issued in-June or July, 1842, mentioned receipt of ‘“‘several of the first numbers” of the Richard work, accompanied with plates. At approxi- mately the same time, Walpers apparently had at his disposal 192 pages (12 signatures) of this work, including Ranunculaceae through Buttneri- aceae, since he included in the first volume of his Repertorium (1842) new taxa of this group of families (the last species mentioned being Gua- zuma parvifolia). Pages 193-336 (signatures 13-21) became available to Walpers sometime later in 1842 or 1843, since Richard’s new taxa of Tiliaceae (e.g., Belotia) through Rutaceae (e.g., Zanthoxylum coriaceum) appeared in Supplementum 1 which was included in the second volume of the Repertorium (1843-1844). Endlicher, however, apparently had at his disposal both these portions in 1842, since Richard’s new genera, includ- ing Belotia, were mentioned in the Addenda to the 2nd Supplementum to the Genera Plantarum (1842, after March). Grisebach (1842), recording the publication of Richard’s work during 1841, mentioned Simaroubaceae as the last family treated by Richard in that portion. Therefore, one should assume that the part of Richard’s work which appeared in 1841 consisted of pages 1-336 (signatures 1-21), and probably Plates 1-35, and included the families Ranunculaceae to Simarubeae (as well as one page of Ochnaceae with a portion of the generic description of Gomphia).

According to Grisebach’s report of 1847, the part of Richard’s work published during 1846 included the families Ochnaceae to “Portulacceae” up to page 624, Although it is unknown whether the corresponding Plates 36-44(2) also appeared at that time, there is no ground (at least at present) for the belief that the appearance of these plates was delayed. A relatively long interval between publication of this and the first part of the work is also evident from the Richard ‘“Avant-propos” (p. vii) which is dated February 1, 1845. “Les circonstances tout a fait indépen- dantes de notre volonté ont singuli¢rment retardé la publication de la deuxiéme moitié de ce volume. Nous espérons que désormais cet ouvrage marchera avec plus de régularité vers la fin de sa publication.” It is un- known whether this second part of Richard’s work was issued in install- ments or whether it remained undistributed until the completion of the whole volume. One may assume that this part was either rare or not available at all to botanists for at least a few years after its publication.

86 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIII

Thus, Planchon (1846-1847) did not mention Richard’s new species of Gomphia in his review of the genera and species of Ochnaceae; Naudin (1849-1852) did not include either Pachyanthus A. Rich. or Naudinia A. Rich., which was named after him, in the monograph of Melastoma- taceae; andl only one new genus (Belairia, Leguminosae) of the second part of Richard’s work was mentioned in the third edition of Lindley’s The Vegetable Kingdom (1853).

Finally, in Grisebach’s report of 1852 on phytogeographical and taxo- nomic works which were published during the preceding year, we read (p. 375), “Von R. de la Sagra’s Werk tber Cuba wurde der erste Band der von Richard bearbeiteten Gefasspflanzen (s. Jahresb. f. 1846. p. 53) vollendet 14!). Ubersicht der seitdem bearbeiteten Familien: 5 Phytolac- ceen, 3 Cacteen, 4 Umbelliferen, 1 Araliaceae, 1 Hedera, 4 Loranthaceen”’ and (p. 387) “141) R. de la Sagra, Histoire physique, politique et naturelle de Vile de Cuba. Botanique. Plantes vasculaires, par A. Richard. al. tit. Essai d’une flore de Vile de Cuba. T. IT. contenant les Dicotylédones poly- petales. 663 pag. 8. Atlas. Paris 1845. (aber jetzt vollendet).” It is evi- dent that the last parenthetical phrase referred to both the text and the atlas, not to the atlas alone. The circumstance that no new taxa of oy tolaccaceae (Trichostigma A. Rich., p. 627; T. rivinoides, p. 628; Stegnos perma cubense, p. 632), proposed by Richard § in this last part, were mentioned by Moquin (1849) in his monograph of Phytolaccaceae (in- cluding “Addenda et Corrigenda”’), seems to support the last Grisebach statement.

In summary, I assume the following probable dates of publication for Richard’s work:

ean

1841: Part 1 (in a few installments?), pp. 1-336 (signatures 1-21), probably ls. 1-35, Ranunculaceae to seers ae.

1846: Part 2 (in installments?), pp. 337-624 (signatures 22-39), probably pls. 36-44(2), Ochnaceae to Portulacceae.

1851: Part 3, pp. 625-663 (signatures 40-42) and i-vili, probably pl. 44(3), Phytolacceae to Loranthaceae

Since 1841 may be accepted as the publication date for Zanthoxylum coriaceum A. Rich., the priority of this binomial over Zanthoxylum cori- aceum (Desv.) Walpers (1842) is clear and the former is the correct name for the species of southern Florida and the West Indies.

THE TYPE SPECIES OF GLYCOSMIS

Much nomenclatural confusion, with resulting taxonomic discrepancies, surrounds Limonia arborea Roxb. (1798), the type species of Glycosmis Corréa. Tanaka considered this species to be identical with Limonia pentaphylla Retzius (1788) and applied to it the binomial Glycosmis penta- phylla (Retz.) Corréa. Narayanaswami (1941), however, came to the conclusion that Limonia arborea Roxb. and Limonia pentaphylla Retzius are entirely different species, Glycosmis arborea (Roxb.) Corréa and

1962 | BRIZICKY, ZANTHOXYLUM & GLYCOSMIS 87

i pentaphylia (Retz.) Corréa, respectively, A brief history of the genus d the basic species toed is necessary for an understanding of the aa

In 1788, Retzius (p. 24) proposed Limonia pentaphylla based on Koenig’s specimen from the East Indies. The leaves of this species were described as ‘‘Folia plerumque in petiolo quina, alterna, ovata, acuta, integra, magnitudine foliorum Citri Medicae.’ Roxburgh (1798, p. 60, pl. 84) provided the plant which he believed to be Limonia pentaphylla Retzius with a more or less detailed description and drawings, and also described and illustrated a new species, Limonia arborea (p. 60, pl. 85). According to him, the former species possessed relatively small, entire leaflets and staminal filaments conspicuously dilated upwards, while the leaflets of the latter species were larger than in the preceding and toothed, the staminal filaments + filiform. However, Tanaka (1928a, p. 159) pointed out that “Roxburgh unfortunately transposed the figures [of the flowers] in the two drawings given in his above mentioned book.”

In 1805, Corréa founded the genus Glycosmis, basing it on Limonia ar- borea and L. pentaphyilla as they were understood by Roxburgh. Corréa, however, did not make formal transfers of these species into his genus, but only remarked (1805, p. 386), “Le Limonia arborea et le Limonia pentaphylla de Roxburgh, pl. Coromand. vol. 1, fig. 85, 86 [sic] ... m’ont furni le caractére du genre Glycosmis, qui se distingue aisément du reste de la famille.” De Candolle (1824, p. 538) apparently was the first to make the formal transfer. Glvcosmis arborea (Roxb.) DC. was based on Limonia arborea Roxb., and Glycosmis pentaphylla DC. was based on the plant identified and illustrated by Roxburgh as ‘“‘Limonia penta- phylia Retzius” and only questionably on Retzius’ species (‘G. penta- phylla, ... Limonia pentaphylla Retz. obs. 5. p. 24 ? Roxb. cor. 1. t. 84.”’).2 Most taxonomists, however, subsequently used the binomial Gly- cosmis pentaphylla (Retz.) Corréa for this species.

Engler (1896, p. 185) introduced a new combination, Glycosmis cochin- chinensis (Lour.) Pierre ex Engler, based on Loureiro’s Toluifera cochin- chinensis (1790) and including G. pentaphylla (Retz.) Corréa taken in a very broad sense, It is unknown for what reason Engler ascribed the authorship of this new combination to Pierre. As far as the present author has been able to determine, Pierre himself did not make the transfer of Toluifera cochinchinensis into Glycosmis. In 1893 (text to pl. 285), having described a new species, G. montana, from Cochinchina, he men- tioned “le Glycosmis pentaphylla Corr. qui est la méme chose que le Toluzf- era cochinchinensis (Lour. Fl. Coch., p. 262) ou cam ruu des Anna- mites. . .” From a brief description of G. pentaphylia, included in the note, it is clear that Pierre understood this species in a broad sense.

Tanaka, who studied extensively Glycosmis (as well as the other genera

?Further evidence that G. pentaphylla DC. was reais based on the nonexistent basionym Limonia pentaphylla Roxb. is found in the way in which De Candolle ee p. 538) cited oe basionym of his Clausena cee “Limonia pentaphylla

erb. Lamb. non Rox

88 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLII

of Aurantioideae), visited European herbaria in the late 1920’s in a search for generic types. In 1928, he reported finding, in the Lund Herbarium, a Koenig specimen which presumably was the authentic type specimen of Limonia pentaphylla Retzius.* An examination of this specimen showed that the true Limonia pentaphylla Retz. was conspecific with L. arborea Roxb. and different from the plant identified by Roxburgh as L. penta- phylla. Consequently, Tanaka reduced G. arborea (Roxb.) Corréa (= G. arborea (Roxb.) DC.) to the synonymy of G. pentaphylla (Retz.) Corréa. To Roxburgh’s “Limonia pentaphylla Retz.” Tanaka (1928b) applied another name, Glycosmis mauritiana (Lam.) Tanaka, based on Limonia mauritiana Lam. (1792). Narayanaswami (1941), apparently having overlooked Tanaka’s article on the type of Retzius’ species, came to the conclusion, based only on Retzius’ description of Limonia penta- pAylla which mentions the entire leaflets, that Tanaka’s interpretation of L. pentaphylla was entirely incorrect and that L. pentaphylla Retz. and L.. arborea Roxb. are different species.

Narayanaswami was also of the opinion that Corréa should be assigned the authorship of the combination Glycosmis pentaphylla, as well as that of G. arborea. “But when we take into consideration the facts regarding the origin of the genus Glycosmis and the species G. pentaphylla and their acceptance by all botanists up to this day, as having been created by Correa, it logically follows that Correa should be assigned the authorship of G. arborea also, since Limonia arborea Roxb, formed one of the com- ponents of the types of the genus Glycosmis of Correa. There appears to be no rule in the botanical nomenclature, that governs such cases where an author creates a new genus from two species of another genus, and leaves it without making the necessary transfer of the earlier species to the new genus. But when a subsequent worker assigns the two species in their new status to the author of the new genus, does it not become bind- ing on all subsequent botanists to follow this adoption by the first bota- nist, subsequent to the publication of the genus” (Narayanaswami, 1941, p. 25). Narayanaswami was correct insofar as there apparently was no clear rule governing the transfers at that time. However, in 1952 at Paris, the Eighth Botanical Congress formulated a rule concerning validly pub- lished new combinations and illustrated it with very clear examples (Art. 32, Int. Code Bot. Nomencl. 1954). In the light of this rule, Corréa’s mentioning “Le Limonia arborea et le Limonia pentaphylla de Roxburgh”

®JIn 1932, Fischer reported on an examination of the Koenig collection of the Lund Herbarium, which was sent on loan to Kew. Interestingly, Limonia pentaphvlla Retzius was listed neither ran the ee ‘Retzius specimens” examined nor among 33 species mentioned in Retzius’ _Observationes but not found in the Koenig collection. specimen of Citrus decumanus,” quoted by Tanaka (1928a) included in either of the two lists. The presence, however, in the Koenig collection of a number of other specimens which were recognized as authentic Retzius jean by Tanaka (Triphasia trifolia (Burm. f.) P. Wils., Plezospermum oe (W. & A.) Swingle, and — limonia (L.) Swingle ionone acidissima of Retzius]) makes Tanaka’s conclusion in regard to the authenticity of the specimen of Limonia pentaphylla ee very probable.

1962 | BRIZICKY, ZANTHOXYLUM & GLYCOSMIS 89

as the species on which the genus Glycosmis was based does not constitute publication of the new combinations in this genus. Therefore the use of Corréa’s authorship for these combinations is against the rules. De Can- dolle (1824) seems to be “‘the first botanist subsequent to the publication of the genus” who made formal transfer of the above-mentioned species of Limonia into Glycosmis as G. pentaphylla and G. arborea, and his author- ship for these binomials is indisputable.

As was mentioned above, Tanaka (1928a), on the basis of the presumed type specimen of Retzius’ species, stated that Limonia pentaphylla Retzius and L. arborea Roxb. were conspecific and different from the plant de- scribed and illustrated by Roxburgh as Retzius’ species. Then Glycosmis pentaphylla DC., based on Roxburgh’s plant, not on that of Retzius, must be regarded not as a new combination, but as a new name in Glycosmis for the species for which Tanaka later (1928b) created the combination G. mauritiana (Lam.) Tanaka (based on Limonia mauritiana Lamarck, validly published in 1792, not in 1789 as Tanaka believed). Since G. pentaphylla DC. cannot be applied to Retzius’ Limonia pentaphylla, the next available name for the latter species is Glycosmis arborea (Roxb.)

The third species involved in the nomenclatural confusion was Gly- cosmis cochinchinensis (Lour.) Pierre ex Engler (1896, p. 185) which was nomenclaturally based on Toluifera cochinchinensis Loureiro (1890, p. 262). Creation of this combination was of no assistance to our knowl- edge of the entity described by Loureiro as Toluifera cochinchinensis, for the type of the latter is not extant. Merrill has helped to reveal its iden- tity. “Although Engler in taking up Pierre’s transfer [sic! ] of Loureiro’s specific name intended it to replace G. pentaphylla Corr. as a collective species, it is not the same as Limonia pentaphylla Retz. (Obs. 5 1789) = Glycosmis pentaphylla Corr. Loureiro’s species is represented by Clemens 3363, 4448, from thickets at Hue and Tourane [presumed classical localities for most of Loureiro’s species from Cochinchina], and de Pirey’s specimen of cam ruou, Chevalier 41186. Guillaumin’s descrip- tion of Glycosmis cochinchinensis (Lour.) Pierre applies only in small part to Loureiro’s species as he treated it as a collective one, citing 14 synonyms, most of which have to be excluded with the restriction of specific limits to the form actually described by Loureiro. True Glycosmis epee (Retz.) Correa does not occur in Indo-China” (Merrill, 1935a,

21). Merrill commented further, “The Hainan specimens closely ian Chevalier 41186, Anamese com ruou (Loureiro’s cay cam ruu) from Anam, and I believe these to represent Loureiro’s species. This form closely resembles G. citrifolia (Willd.) Lindl., but the leaves are constantly simple. Chun 5722 from Hainan, which represents the same form as the specimens cited above, has been identified by Tanaka as representing Glycosmis citrifolia (Willd.) Lindl. var. obtusa (Miq.) Tanaka” (Mer- rill, 1935b, p. 17). The quoted notes suggest that Toluifera cochin- chinensis Lour. is closely related to Glycosmis parviflora (Sims) Little

90 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLII

(G. citrifolia (Willd.) Lindl.) and perhaps should be included in the

In conclusion, it is to be said that in the light of the current rules of botanical nomenclature and of our present knowledge of botanical bibli- ography, the nomenclature of the two basic species of Glycosmis appears to be the following:

Glycosmis arborea (Roxb.) DC. Prodr. 1: 538. 1824.

Limonia arborea Roxb. Pl. See ie - as 1798.

Limonia pentaphylla Retz. Obs. Bot. 5: 24.

Glycosmis pentaphylla sensu Bea - a other authors, not G. penta-

phylla DC. 1824.

Glycosmis mauritiana (Lam.) Tanaka, Bull. Soc. Bot. Fr. 75: 708.

Limonia mauritiana Lam. Encycl. Méth. Bot. 3: 517. 1792. Limonia pentaphylla Retz. according to Roxb. Pl. Coromand. 1: 60. pl. 84. 98, not Retzius, 1788 Glycosmis pentaphylla DC. Prodr. 1: 538. 1824; incorrectly given as G. Ne de Se Corréa_ by ona: Rec. Bot. Surv. India 14(2): 12. 1941

SUMMARY

With regard to the generic limits of Zanthoxylum L., especially in con- nection with the difference in perianth structure between Zanthoxylum L. sensu stricto and Fagara L., the ‘‘simple” perianth of the Zanthoxylum type is presumed to be a secondary condition derived by reduction from the double perianth of the Fagara type. The occurrence in Mexico and Central America (perhaps also in South America) of species of Zan- thoxylum which appear to be transitional to Fagara in their perianth structure is considered ample reason to regard Fagara as a subgenus o Zanthoxvlum, rather than as a distinct genus.

A number of lines of evidence show that Miller’s binomial Xanthoxylum americanum (1768) referred to the northeastern American species which was later described by Willdenow (1796) as Zanthoxylum fraxineum. Since neither the occurrence of an error in Miller’s description nor the

*Tanaka’s interpretation of Limonia pentaphylla Retzius, based on the presumed type specimen, seems preferable to the Narayanaswami’s concept of the species. This is especially true if ae also takes into consideration that the original description of Limonia pentaphylla Retz. is well applicable to L. arborea Roxb., except for the entire leaflets of the former species and the serrate to crenate ones of the latter. It should be remembered, however, that the serration of leaflets in some specimens of L. arborea is perceptible only with a lens. On the other hand, the Retzius’ species differs from “Limonia pentaphylla Retz.” of Roxburgh in acute leaflets Mes acute” in the latter according to Narayanaswami, 1941, p. 14) and their size, “magnitudine foliorum Citri Medicae,” (“small” according to Narayanaswamil, loc. cit., “from two to three inches long, and about one and a half broad” according to Roxburgh 1798, p. 60). The leaves of Citrus Medica are about 12-20 cm. long,

1962 | BRIZICKY, ZANTHOXYLUM & GLYCOSMIS 91

lack of the type specimen makes Miller’s binomial illegitimate, the re- placement of Zanthoxylum americanum Mill. by Z. fraxineum Willd., re- cently suggested by Fosberg, seems to be unjustifiable in the light of the present code of botanical nomenclature.

In the light of the publication dates of Achille Richard’s work on the flora of Cuba (in Ramon de la Sagra), as reported by Grisebach (1842, 1847, 1852), the priority of Zanthoxylum coriaceum A. Rich. (1841) over its homonym Z, coriaceum (Desv.) Walpers ( 1842) appears indis- putable. Thus, the former binomial is the legitimate and correct name of a well-known species of southern Florida and the West Indies.

A very complex situation in the nomenclature of the type species of Glycosmis Corréa, Limonia arborea Roxb. (1798), is shown, and a brief history of this genus and the basic species is given. Tanaka’s assump- tion of the conspecificity of Limonia pentaphylla Retzius (1788) with L. arborea Roxb., based on the presumed authentic type specimen of the former, seems correct. Since G. pentaphylla DC., based on Limonia penta- phylla sensu Roxburgh (1798), not of Retzius (1788), cannot be applied to Limonia pentaphylla Retz., the next available name for the latter species is G. arborea (Roxb.) DC. The correct name for G. pentaphylla DC. is G. mauritiana (Lam.) Tanaka. The third species involved in the nomen- clatural confusion was G. cochinchinensis (Lour.) Pierre ex Engler, which was based on Toluifera cochinchinensis Loureiro (1790). Merrill’s notes on the latter species, quoted by the present author, suggest a close re- lationship of Loureiro’s species with G. parviflora (Sims) Little (G. citrifolia (Willd.) Lindl.), rather than with G. arborea.

LITERATURE CITED

Arton, W. T. 1813. Hortus Kewensis. ed. 2. 5: 568 pp. London.

CANDOLLE, A. P. DE. 1824. Rutaceae. Jn: Prodr. Syst. Nat. Regni Veg. 1: 709-732.

CoRREA DE SERRA, J. F. 1805. Observations sur la famille des orangers et sur les limites qui la circonscrivent. Ann. Mus. Hist. Nat. Paris 6: 376—

87.

Du Ro, I. P. 1771. Dissertatio inauguralis observationes botanicas sistens. Thesis, 62 pp. Helmstadii.

. 1772. Die Harbkesche wilde Baumzucht theils Nordamerikanischer und anderer fremder, theils einheimischer Baume, Straéucher und strauch- artigen Pflanzen 2: 512 pp. & indices. pls. 4-6. Braunschweig.

Ercuier, A. W. 1878. Bliithendiagramme. 2. pp. ITI-XX & 575 pp. Leipzig.

ENDLICHER, S. 1842. Mantissa botanica sistens generum plantarum supple- mentum secundum. 114 pp. Vindobonae.

ENGLER, A. 1874. Rutaceae. Jn: Martius, C. F. P. Fl. Bras. 12(2): 77-196. pls, 34-37.

. 1896. Rutaceae. Nat. Pflanzenfam. IIT. 4: 95-201.

———. 1931. Rutaceae. Nat. Pflanzenfam. ed. 2. 19a: 187-359.

FiscHer, C. E. C. 1932. The Koenig collection in the Lund Herbarium. Kew Bull. 1932: 49-76, 256.

92 JOURNAL OF THE ARNOLD ARBORETUM [ VOL, XLII

Fosperc, F. R. 1958. Zanthoxylum L., “Xanthoxylum Mill.,” and Thylax Raf. Taxon 7: 94-96.

1959. Typification of Zanthoxylum L. Ibid. 8: 103-105.

GrisEBACH, A. 1842. Bericht tiber die Forschungen in we ie a wahrend des Jahres 1841. Arch. Naturgesch. 8(2): 462.

. Bericht iiber die Leistungen der Sanaa wahrend des Jahres 1846. bid. 13(2): 409-472.

. 1852, Bericht tiber die Leistungen in d. Ape Nae und systema-

tischen Botanik wahrend der Jahres 1851. Jbid. 18(2): 308-429.

Humeotpt, A., A. BoNPLAND, AND C. S. KuntH. 1823, ver Nova genera et species plantarum 6: 535 pp., pls. 513-600,

Jussieu, A. 1825. Mémoire sur le groupe des Rutacées. Mém. Mus. Hist. Nat. Paris 12: 384-542. pls. 14-29.

KunTzE, O. 1898. Revisio generum plantarum 3(2): 1-201.

Lamarck, J. B. A. P. M. ve. 1792. Encyclopédie méthodique. Botanique. 3(2): 361-753.

Littte, E. L., Jr. 1953. Check list of native and naturalized trees of the United States (including Alaska). U. S. Forest Serv. Agr. Handb. 41: 472 pp. Washington, D.C

Loureiro, J. pe. 1790. Flora cochinchinensis: sistens plantas in regno Co- chinchina nascentes 1: 353 pp. Ulyssipone

MerriLL, E. D. 1935a. A commentary on Loureiro’s “Flora cochinchinensis.” Trans. Am. Philos. Soc. II. 24(2): 1-445.

. 1935b. A sixth supplementary list of Hainan plants. Lingnan Sci. Jour. 14: 1-62.

MiItter, P. 1759. The Gardeners Dictionary. ed. 7. London.

. 1768. Ibid. ed. 8

Moore, J. A. 1936. Floral anatomy and phylogeny in the Rutaceae. New Phytol. 35: 318-322

Moaguin—Tanpon, A. C. H. B. 1849. Phytolaccaceae. Jn: DC. Prodr. 13(2): —40

NARAYANASWAMI, V. 1941. A revision of the ae ae an species of Glycos- mis Correa. Rec. Bot. Surv. India 14(2):

NaupIn, C. 1849-1852. Melastomacearum que in museo Parisiensi continentur monographicae descriptionis et secundum epee distributionis tenta- men. Ann. Sci. Nat. IIT. 12-18.

Pierre, L. 1893. Flore foresti¢re de la Cochinchine. Fasc. 18. pls. 273-288. Paris.

PLANcHON, J. E. 1846, 1847. Sur le genre Godoya et ses analogues, avec des observations sur les limites des Ochnacées, et une revue des genres et especes - ce Ce London Jour. Bot. 5: 584-600, 644-656; 6: 1-31.

REEDER, J. p S.-Y. CHEo. 1951. ee on Xanthoxylum & Fagara in China. i face Arb. 32: 67-72. pls. 1,

REHDER, A. 1945. Notes on some euiltivated es and shrubs. Jour. Arnold Arb. 26: 67-78.

Rerzius, A. I. 1788. Observationum botanicarum fasciculus quintus. 32 pp., 3 pls. Lipsiae.

Ricuarp, A. 1841-1851. Botanique.-Plantes vasculaires. (Essai d’une flore

e Vile de Cuba. 1.) Jn: Sagra, R. de la. Histoire physique, politique et naturelle de ile de Cuba. Paris.

1962 | BRIZICKY, ZANTHOXYLUM & GLYCOSMIS 93

RoxpurcH, W. 1798. Plants of the coast of Coromandel. 1(4): 57-68. pls. 76-100. London.

SaunberS, E. R. 1934. On carpel polymorphism. Ann. Bot. IJ. 48: 643-692.

ScHkuUHR, C. 1803. Botanisches Handbuch der mehresten theils in Deutsch- land wild wachsenden, theils auslandischen, in Deutschland unter freiem Himmel ausdauernden Gewachse 3: 564 pp. & Nachtrag und Register, 69 pp., pls. 213-358. Wittenberg.

SmitH, J. D. 1897. Undescribed plants from Guatemala and other Central American Republics. 19. Bot. Gaz. 24: 389-398.

mae eer AN, M. J., VAN. 1960. Dates of publication and bibliographical

Fl. Males. Bull. 15: 730-743.

ee T. 1928a. A note on Retzius’ Limonia pentaphylla, Bot. Not. 1928:

156-160.

. 1928b. Revisio Aurantiacearum I. Bull. Soc. Bot. Fr. 75: 708-715.

TRIANA, J., AND J. E. PLANCHON. 1872. Prodromus vgs es Granatensis. Te mebinthace ye: Rutaceae. Ann. Sci. Nat. Bot. V. 14:

Ursan, I. 1894. Additamenta ad cognitionem florae eee Cae iE. Bot. Jahrb. 19: 562-681.

Watpers, G. G. 1842. Repertorium botanices systematicae 1: 947 pp. Lipsiae.

; 184 3. Ibid. 2: 1029 pp. Lipsiae

WANGENHEIM, F. A. J. 1787. Berag zur teutschen holzgerechten Forstwissen- schaft, die Anpflanzung Nordamerikanischer Holzarten, mit Anwendung auf teutsche Forste. 124 pp., pls. 1-31. Gottingen.

WILLDENow, C. L. 1796. Berlinische Baumzucht, oder Beschreibung, der in

n Garten um Berlin, im Freien ausdauernden Baume und Straucher. 452

pp., pls. 1-5. Berlin.

Witson, P, 1911. Rutaceae. N. Am. FI. 25(3): 173-224.

94 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLII

THE TYPIFICATION OF DIOSPYROS EBENUM AND DIOSPYROS EBENASTER

RicHArp A. Howarp ANp TycHo NorLINDH

IN THE LATE PART of the eighteenth century, botanists were uncertain of the identity of the trees which produced two valuable commercial tim- bers, red sandalwood and true ebony. John Gerhard Koenig is credited with the identification of each during a stay in India and Ceylon. Koenig was born in Polish Latvia in 1728. He moved to Denmark in 1748 and studied with Linnaeus from 1757 to 1759. As a physician, he joined a Danish mission to India in 1767 and died in India in 1785. The present paper concerns the typification of Diospyros ebenum oa source of the true

ebony), the proper citation of the name, and its synonymy.

A paper with the Swedish title, ‘ ‘Diospyros Ebenum fe Akta Ebenholz, beskrifvit af John Gerhard Konig” |Diospyros ebenum or true ebony, described by John Gerhard Koenig|, was published in the first volume of the Lund Physiographiska Salskapets Handlingar, printed in Stockholm in 1781." In addition to a Latin description of Diospyros ebenum, the article contained in a footnote a discussion of the distribution of the plant, its characteristics and its use, the method of formation of the char- acteristic black wood, and, finally, an eulogy of Koenig for supplying the information. The article, in contrast to others in the Handlingar, does not indicate the author. The discussion is written in the third person and we shall show that an original article written by Koenig was translated, edited and published by A. J. Retzius, the founder and the secretary of the Lund Physiographiska Salskapet and its publication. The correct cita- tion of the name given the ebony tree should be Diospyros ebenum Koenig es Retzus,

The original description apppears to be a composite one, including uni- sexual and hermaphrodite flowers and fruit. It was obviously based on a field knowledge of the plant in Ceylon. No specimens were cited, and the selection of a lectotype for the species is now necessary

In the library of the Botanical Museum, in Copenhagen, there are a number of long letters from Koenig in India during the period 1777-1783

"The date of publication of Diospyros ebenum is given in Index Kewensis and by most recent authors as 1776, the date on the title page. The first volume of this eae was published in four parts. It is significant to note that Parts 1 ee 2, pages 1-64 and 65-132 respectively, were published in 1776, but that Part 3, pages 133-220, ee: the description of Diospyros ebenum, was published in 1781, and Part 4, pages 221-318, was issued in 1786. Otto Gertz viene! Fysiogr. Sallsk. i Lun 1772-1940 Historisk Overblick, 16. 1940) reports that “the Secretary of the Society announced on May 2nd, 1781, that the printing of ‘Handlingarna’ Part 1:3 was finished.”

1962 | HOWARD & NORLINDH, DIOSPYROS 95

to Rottboell who was chief of botany in the Chair of Medicine in Copen- hagen until 1797. One letter, written on August 22, 1777, is fifteen pages long. Another pertinent to the present paper was written on February 5, 1778, and is eight folio-size pages in length. Koenig’s letters are hand- written in an old-style German. The structure of the language he used is strange today, as is the spelling of many of the words. We are indebted to Mr. Sigurd Molander, of the Library of the Botanical Museum of Lund, for his assistance in the interpretation and translation of Koenig’s script.

From Tranquebar, India, on February 5, 1778, Koenig wrote to Rott- boell of his recently completed journey to Ceylon and his discovery of the true ebony. On May 10, 1777, the English warship “Seahorse,” under Captain James’ command and with Koenig aboard, sailed from Madras and arrived at Trincomalee, Ceylon, nine days later. Koenig wrote a brief

4 hE Elon: Clay Tgawe

os ben the Parag Llbar aloe eile me : “a Bilin gt foe Gl

alta da we : ee Cathey ges. ee opifal= Ub ns nage ad mith § ine eo tad eomavee fur aS, an ‘ jicd Lee bh is oposce ; i rLeveds gredhes ater eh, nase axafey rel. c oe Ver bigfens, phduls, covet : a Fy f eget, Haren Pt ar es VE pe

2 ‘Pp pee ee eet - » stflert.fujes eb duceti , re aypee’ed ‘Li VRE T, od : ‘ Y i vane L y) — a OTTO. ma oa ah fot, f re aes: ina fo bs Tak je in Yen an me, lip fiellin, ue a, thik Cum m a

tune pare, vem one, oe fle nip oh Forartes ONG, oe. ed prespey evbix'c Bee axe , : pls aa fees cone soar Cadib, vinedfefes. Pract pe ft gl my Lere bee ores, Pad... 2 un Yoo! por jbrcbufeon:, Vlora Oriente

CofE pn ea len hus x a xv yobs mnrille brectcrey:

ms)

TextT-Fic. 1. First page of Koenig’s manuscript, “Descriptio Ebeni Classis Polygama Dioecia.”

96 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLII

description of the Trincomalee Mountains, their geology and vegetation. He stated that their slopes were rugged and densely covered by many kinds of trees, among others very many ebony trees. The season was not well adapted for the collection of flowering material, but Koenig declared that he had been able to make a complete description of ebony which he enclosed and from which the addressee, Rottboell, could see that the ebony was a true Diospyros. The description Koenig sent was titled ‘“De- scriptio Ebeni Classis Polygama Dioecia”’ and is preserved today. The first page of this classic manuscript is reproduced here as TEXT-FIGURE 1. During a six-day stay in the area of Trincomalee, Koenig collected many new herbs and even concluded that he should very much like to live there. Later in his letter Koenig described another landing at Jaffna where the true ebony used to be cut. Here he said there were very tall and well- formed trees, many of which bore fruit. However, he was not able to find a single flower, though he offered rewards of much gold to the finder and many trees were thoroughly searched. In his presence two trees were cut down, and in these only the innermost part had turned black, showing beyond doubt that it was ebony. A forester on the spot told Koenig that holes were cut through the bark into the wood in order to encourage the development of pigmentation. He explained the white streaks found in ebony as incomplete mummification due to premature cutting and con- trasted this with the uniformly colored woods obtained in the French and Mauritian islands. Koenig indicated in his letter that he had collected as many plants as possible and that these had been sent to Copenhagen (‘von welchen allen ich im vOnenten Jahr auch nach Copenhagen zureich- ende Exemplar tiberstandt habe.

Although there is no corespondence to support the conclusion, it seems clear to us that Rottboell received the specimens sent by Koenig as well as the “Descriptio ebeni.” He kept for himself a fruiting specimen and the wood specimen. He sent to Retzius a poorer fruiting specimen, and either the original manuscript, ‘‘Descriptio ebeni,”’ which was later returned to Copenhagen, or a copy thereof. Retzius then edited Koenig’s manu- script, for, when the original and the published documents are compared, we find several changes in Koenig’s Latin constructions but no significant alteration of the diagnosis. The German text of the original manuscript was translated freely into Swedish for publication, and Retzius supplied only the footnote.

The selection of a lectotype, therefore, is a choice between the specimen Rottboell received from Koenig and sent to Retzius and the specimen which Rottboell retained. The specimen in the Retzius herbarium at Lund (PratEe I) bears only two ee “Diospyros ebenum,” in Retzius’ handwriting (PLATE Va), and “e coll. Retzii,” which is probably the writing of Professor Agardh. There are fruits attached to this speci- men but these are not associated with the leaves, which are only at the ends of the branches and probably represent a later period of growth. The leaves are thin in texture, pointed at the apex, and have dried a dark color.

1962 | HOWARD & NORLINDH, DIOSPYROS 97

are the annotations ‘“Diospyros ebenum verum” and “habitat in vastis- simis sylvis prope Jafnapatnam & ad latere montium Trinquemallensium”’ (PLATE Vb). Associated with this sheet is a piece of ebony wood and a section of a stem with both wood and bark.

A second sheet in the Vahl herbarium in Copenhagen bears a specimen in flower and an annotation partly in Koenig’s handwriting (PLATE Vc). A label of more recent date designates this specimen as the “originalek- semplar’”’ and indicates that the specimen was received from Koenig from “Trankebar ca. 1780.” We do not know who supplied this particular label, but that botanist wrongly indicated its origin as “Trankebar.”’ He failed to notice the annotation on the reverse side of the sheet which reads, “habitat in sylvis Zeylanicis copiose” (PLATE Vc). A sample of wood accompanies this flowering specimen. It is important to notice that this specimen has both older, coriaceous, light-colored leaves and younger, thinner textured, dark-colored leaves on the same shoot. It thus indicates the relationship between the types of foliage represented by the Retzius and the Rottboell collections. The more authentic label associated with the specimen in the Rottboell herbarium, the presence of two wood sam- ples, the fruiting condition of the specimen, and the supporting corre- spondence have led us to accept the Rottboell specimen as the lectotype and to disregard the “originaleksemplar” label. A similar specimen is in the herbarium at Lund (PLATE I

There is in the Linnaean herbarium in London a fruiting specimen with similarly shaped heavy leaves (Prat IIIb). We have not been able to trace the origin of the Linnaean specimen in any correspondence from Koenig. The sheet is annotated “Konig 1777” and “Ebenum Verum ex vastis sylvis Zeylonae. Flores non vidi! an Diospyros?” We suspect that Koenig, a former student of Linnaeus, might well have sent a speci- men of his important discovery directly to his former teacher. The chances are that it arrived after the death of the elder Linnaeus (January 10, 1778), for Linnaeus filius described Diospyros ebenum in his Sup- plementum Plantarum Systematis Vegetabilium, page 440, in 1781,” with

2On the basis of the year of issue, Diospyros ebenum L. {. appears to compete with the same name attributed to Koenig. According to the date given in the preface, the Supplementum of Linnaeus was completed by October 1, 1780. Linnaeus states that his determination was confirmed by Thunberg, who worked with him at least in the early months of 1779 (Karsten, Jour. 5S. Afr. Bot. 5: 103. 1939). Stoever (The Life of Sir Charles Linnaeus, 295. 1794) quotes Ehrhart as stating that Linnaeus “sent me the manuscript in the autumn of 1779, to be printed. J perused it, set down my doubts and observations, and sent them to Linnaeus. A correspondence then began between us, which lasted almost the whole of the ensuing winter. After this, I had it copied afresh, and began to get it quite ready for the press; I was how-

98 JOURNAL OF THE ARNOLD ARBORETUM | VOL. XLUI

the comment, “Hoc est verum Lignum Ebenum, cuius originem detexit Konig, confirmavit Thunberg.” Linnaeus filius was obviously unaware of the nearly concurrent publication of the same epithet by Retzius in the Lund Physiographiska Sdlskapets Handlingar. Certainly the date 1777 and the annotation on the Linnaean sheet, in further agreement with Koenig’s letter to Rottboell, suggest that Koenig did send the specimen to Linnaeus after he had first seen ebony but while he was still uncertain of its identity as a species of Diospyros.

After his initial visit in 1777, Koenig returned to Ceylon several times. The majority of his visits were for but a few days. However, in 1781, he made extensive investigations of the flora. It is probable that Koenig continued to collect specimens of ebony, including ample flowering ma- terial, which he sent to Rottboell at Copenhagen. Rottboell was obvi- ously aware of the two publications of the name Diospyros ebenum by Retzius and by Linnaeus, of the incomplete published description, and of the lack of an illustration of this important plant. In 1783, he pub- lished still another description of the plant, this time with an illustration, and changed the name to Diospyros glaberrima (Nye Samling K. Danske Videnskabers Selskabs Skrifter 2: 540. tab. 5. 1783). It is clear that Rottboell was renaming the plants described earlier. He points out that his description is largely compiled from Koenig’s published description and from material Koenig sent him. Rottboell’s own contributions, for which he takes full responsibility for errors, are based on dissections he made of the flowers, thus lending supporting to our conclusions that Koenig sent Rottboell additional material. Fina lly, in the planation of the figures for the ey he supplied, Rottboell notes, ‘“‘Ramus Diospyri glaberrimi sive Eb ’ The upper portion of the illustration (PLaTE IIIa) is clearly a from the Koenig specimen ‘“Diospyros ebenum verum” in the Rottboell herbarium which we have designated as the lectotype (PLatre 11). Rottboell did not annotate the sheet itself, but the outer cover is labeled Diospyros glaberrima in Rottboell’s hand. The lectotype we have selected for Diospyros ebenum Koenig ex Retzius and the holotype of Diospyros glaberrima Rottboell are one and the same specimen. We believe this is what Rottboell intended.

The second specimen of Diospyros ebenum in the Retzius herbarium (Prate IVb) at Lund is comparable to the specimens in the Rottboell (Prate IT), Vahl (Prater [Va) and Linnaean (Piate IIIb) herbaria. The Lund specimen, however, bears a copy of the description published by Linnaeus filius and the correct page reference to this work (PLatE Vd). ever, prevented, by the botanical tour through the electorate of Hanover, with which

S BRITTANIC MAJESTY had expressly charged me. I got it ready at last, in the winter between 1780 and 1781, The work was to be printed at Hanover, under my immediate

of two ducats per sheet, which I sent to Linnaeus after his return from England.” Thus it appears that Diospyros ebenum Koenig has a priority of two or three months over Diospyros ebenum L. {. (See Footnote 1.)

1962 | HOWARD & NORLINDH, DIOSPYROS 99

The label also indicates that the specimen was collected by Koenig in Ceylon and sent to Vahl in 1781 (‘‘misit Praeclar: Demonstr. D: Vahl, anno 1781’). Vahl may have sent this to Retzius at Lund. Vahl was attached to the Botanical Garden of Copenhagen as lecturer in 1779 and was appointed professor of botany there in 1801.

We conclude that Retzius had not seen this flowering specimen, which bears the name Diospyros ebenum L. f. as well as the description and reference, when he published Diospyros ebenum Koenig ex Retzius. We do not know whether it was this specimen from Vahl or whether Retzius saw the specimen which Linnaeus filius must have had at Uppsala, but it must have seemed to him that the heavier and bluntly pointed leaves of the two specimens (PLATES IVb & IIIb) characterized a different plant from the one he had on hand (Prater I) when he published the Koenig manuscript. Therefore, in his Observationum Botanicarum (5: 31. 1789) Retzius published the following:

88. DIOSPYROS Ebenaster foliis ovali-oblongis coriaceis, gemmis glabris.

Diospyros Ebenum L. Supplem. p. 440.

Hebenaster Rumph. Amb. III. p. 13. t. 6.

Habitat in sylvis circa Calcuttam. Konic.

Obs. Manifeste utrasque Diospyri species confundit Nob. @ LINNE Fil. De-

scriptio hujus habetur in Suppl. l.c. quacum conferri meretur Rumphius. $9. DIOSPYROS Ebenum foliis ovato-lanceolatis acuminatis, gemmis hirtis.

Diospyros glaberrima Friis Rottb. in Novis Act. Hafn. II. p. 540. tab. 5.

Diospyros Ebenum Physiogr. Saelsk. Handl. V. I. P. 3. p. 176.

Habitat in Zeylonae sylvis.

Descriptionem ab inventore concinatum vide 1. cit.

Folia circiter tripollicaria, quoad consistentiam tenuia, flexilia, obscure viridia.

A translation of Retzius’ comment under Diospyros ebenaster appears to be the following: “Obviously he (Linnaeus filius) mixes up both species of Diospyros. The description of this (D. ebenaster) is found in Suppl. l.c. with which place Rumphius deserves to be compared.”

It seems clear, therefore, that Retzius was substituting a new name, Diospyros ebenaster, for D. ebenum L. {. and that this species probably was known to him at least by a specimen in his herbarium which he had received from Vahl in 1781. Regrettably, we have not been able to locate any specimens annotated as D. ebenaster by Retzius. Furthermore, it appears that Retzius made two mistakes in the protologue of D. ebenaster. The first is the description of the buds as glabrous. All of the specimens we have seen are more or less pubescent when viewed under the magnifica- tion of the usual hand lens. Mr. George Proctor examined for us the speci- men in the Linnaean herbarium and confirms that the buds on that, too, are slightly pubescent. The second mistake appears to be Retzius’ error in referring this species to a Koenig collection from Calcutta. Certainly the present specimen in the Linnaean herbarium is clearly marked from Ceylon. While Koenig did collect in India, we have not seen any speci- mens of Diospyros ebenum collected by him and reported to be from Calcutta.

100 JOURNAL OF THE ARNOLD ARBORETUM [ VOL, XLIII

It should also be noted that the Rumphius reference does not apply to specimens from Calcutta. Furthermore this name is pre-Linnaean (1743) and technically need not be considered in the synonymy. However, Retzius obviously based his new specific name on this mononomial. Bak- huizen van den Brink in a later publication described the Rumphius species as Diospyros lolin Bakhuizen van den Brink * (Gard. Bull. Straits Settl. 7: 175. 1933 and Bull. Jard. Bot. Buitenzorg III. 15: 152. 1937).

There remains for consideration the difference in leaf shape given by Retzius in the protologues of the two species Diospyros ebenum and D. ebenaster. We have already indicated that the existing Koenig collections cited and illustrated show variation in single specimens comparable to ae described by Retzius for the two species. During 1961, Mr. D. M.

A. Jayaweera, director of the Royal Botanic Garden, Peradeniya, Ceylon, kindly obtained for us a suite of specimens of D. ehenurs series ob- tained from a single tree clearly indicates that the size, shape. and tex- ture of the leaves of the ebony do vary. The apex of the leaf blade may be acuminate, acute, obtuse, or emarginate on one branch. The texture of the blade in successive flushes of growth may be thin and membranaceous and black upon drying or the older leaves of the same branch may be coriaceous and of a lighter color after drying. The young buds on all the specimens examined were densely pubescent, while the buds in the axils of older leaves showed less pubescence.

We have concluded that Diospyros ebenaster is a synonym of D. ebenum and that the correct citation and synonymy should be:

Diospyros ebenum Koenig ex Retzius, Lund Physiogr. Salsk. Handl. P70. 178)

Diospyros ebenum L. f. Suppl. Pl. Syst. Veg. 440. 1781.

Diospyros glaberrima Rottboell, Nye Saml. Kong. Dansk. Vidensk. Selsk. Skr. 2: 540. pl. 5. 1783.

Diospyros ebenaster Retzius. Obs. Bot. 5: 31. 1789.

The lectotype selected for Diospyros ebenum Koenig ex Retzius is the specimen in the Rottboell herbarium (Copenhagen)

Two monographs of the Ebenaceae published in the last century dif- fer in their treatments of this species and its synonyms. Hiern (Trans. Cambr. Phil. Soc. 12(1, 2): 27-300. 1873) recognized Diospyros ebenum and D. ebenaster as distinct species. The treatment published by Bak- huizen is inconsistent and less than clear (Bull. Jard. Bot. Buit. III. 15: 1-515. 1936-41). His first reference to the species was “Diospyros

“Tt should be noted here Diospyros lolin Bakhuizen based on Hebenaster Rumphius

ebenum. By typifying Diospyros ebenaster Retzius with Diospyros ebenum 1 Oe f. t way is clear for the use of Diospyros lolin Bakhuizen.

1962] HOWARD & NORLINDH, DIOSPYROS 101

Ebenum L.” in an observation (loc. cit. 5. 1936). In the Addenda et Corrigenda (loc. cit. 369. 1941) this is corrected to read “Koen. et L. f.” which is incongruous. In the body of the monograph (loc. cit. 216. 1937) he uses “Diospyros Ebenum Koen.” and in the literature cited the refer- ence “Linn. f., Suppl. Syst. Plant. (1781) 440, partim” is given. There is no discussion of this conclusion that Linnaeus filius had a mixed col- lection and no reference to the other part of the Linnaean concept can be found in the monograph. Bakhuizen also established as new two vari- eties for ““D. ebenum Koen.” The first variety “A. var. glaberrima (L.f.) Bakh. — D. Ebenum Koen. typica, D. glaberrima Linn. f.” * if accepted today must be var. ebenum including as it does Koenig’s type. The second variety is “B. var. timoriana (Miq.) Bakh. — ?D. Ebenaster Retz., D. reticulata Willd. var. timoriana A. DC., D. timoriana (A. DC.) Migq.” In the following discussion Bakhuizen states, “D. Ebenaster is a very vaguely described species collected by Koenig in the forests of Calcutta. Hence it is most probable that the plant in question is only a form of D. Ebenum Koen... . It will be better to consider D. Ebenum Koen. and D. Ebenaster Retz. forms of the same species.” The range of D. ebenum var. timoriana, however, is given by Bakhuizen as Timor, Celebes, and the Malayan Peninsula — and Calcutta, if one believes the Retzius reference on which this location is based. We have not examined material of Bakhuizen’s D. ebenum var. timoriana, nor have we seen authentic specimens of D. reticulata Willd. var. timoriana A. DC. We call attention to this problem and leave for others the decision as to whether this is truly a variety of D. ebenum or some other taxon.

The taxon which Hiern called Diospyros ebenaster was reported to occur in the Philippines and the Celebes and to be in such “cultivated places in tropical America, perhaps introduced” as Mexico, Brazil, Cuba and Montserrat. Bakhuizen van der Brink renamed this Diospyros nigra (Gmel.) Perrottet. Many contemporary floras continue to use the name D. ebenaster Retz. for this widely distributed species. The senior author has pointed out in a previous paper (Jour. Arnold Arb. 42: 430-436. 1961) that Bakhuizen was in error in using the epithet D. migra and that two species are represented by Hiern’s D. ebenaster which is Bak- huizen’s D. nigra. One is a native of the Lesser Antilles of the West In- dies, and this species does not appear to be introduced into cultivation out- side of the Western Hemisphere. Its correct name is D. revoluta Vahl, the type of which was collected in Montserrat. The other taxon involved is correctly named D. digyna Jacquin. This plant is a native of Mexico and Central America but has been introduced into many areas of Asia and is cultivated in many botanical gardens.

In addition to the individuals mentioned in the body of this paper, we wish to express our appreciation to Dr. A. Skovsted, of Copenhagen; Dr.

*It should be noted that this reference is in error. Linnaeus filius did not publish “Diospyros glaberrima” but Diospyros ebenum. Bakhuizen either misread the i naean descriptive phrase (q.v.) or intended to credit Rottboell as the author of the basionym.

102 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLII Frances Jarrett, of Kew; and Mr. T. B. Worthington, of Ceylon, for many instances of valuable assistance during the course of this study. ARNOLD ARBORETUM, HARVARD UNIVERSITY AND

NATURHISTORISKA RIKSMUSEET, STOCKHOLM

EXPLANATION OF PLATES

PLATE I Diospyros ebenum in the Retzius herbarium. This specimen presumabl y re- ceived from Rottboell along with Koenig’s original manuscript. Fiare If Lectotype of Diospyros ebenum Koenig ex Retzius in the Rottboell herbarium, Copenhagen. This specimen is also the holotype of D. glaberrima Rottboell.

Prate IIT a, Illustration published by Rottboell for aes hae Rottboell (Nye Saml. Kong. Dansk. Vidensk. Selsk. Skr. 2: pl. 83). Notice by — with Plate II accuracy with which this Hare was ser b, Type specimen of D. ebenum L. f. in the Linnaean herbariu PLATE IV

a, Specimen of Diospyros ebenum in the Vahl herbarium, Copenhagen, anno- tated by an unknown hand as “Originaleksemplar.” This specimen, incorrectly attributed to “Trankebar,” is in flower. ; Flowering specimen in the soot herbarium, Lund. This specimen, bearing the description and reference of D. eben L. f. may well represent Reizius’ concept of D. ebenaster Retzius.

Prats: ¥.

Annotations from specimens of Diospyros ebenum illustrated in Plates ae he and IV. a, Annotation in Retzius’ handwriting from specimen in Plate I. b, Annotations from lectotype in the Rottboell herbarium (PLare II). c, Anno- tation from specimen in the Vahl herbarium (PLATE IVa). d, Annotation from specimen in the Retzius herbarium (PLate IVb),

PuateE I

Jour. Arnotp Ars. VoL, XLIII

von cnpnanne

mg

fee

ect

Howarb & NoRLINDH, DIOSPYROS

Jour. ARNOLD Ars. VoL. XLIII Pirate II Pe et ar se ret ae Ne Ob Re ee eee

=e

Howarpb & NorLINDH, Drospyros

Jour. ARNOLD Ars. VoL. XLIII PriaTE III

n

ie)

24

m

Ou

n

je)

_

A

q

Q

$reg “ as i= ae : ne ee a ee UP 1B) 19 0 21 | A 4

oe * |

[2]

2

a

) Fs 2 a ee =<

i

a

Howarp & Noriinpu, Diospyros

IIITX “1OA ‘day GIONYY ‘Anof

AT aivig

Jour. ARNoLD Ars. VoL. XLIII PLATE V

Seopeyn Mowe’

ox

Tr ~ HB, ROTTBOLL,

Cee De - g

wee oe les ‘Conse tov hag r6foana tip! Oibregs oe ae Lb. Glabhre dy pige con

C.Lny oe ok Mii co ER . ae a0 Cae pte PIN) he PL Ey LES <o

re Via a) oruen OL

ee

oo Le <! |e py wt ae le lb mr Gh yteon..

F oe 2 G oe fIte7e f a-y off a wo farts" ate ta see LOLS 7 ve

Bafa Cer

a ey, oe é | 2

« a 4 *. d

/ Mh non fy Spill rte olin en, | | MM , cif = ase

Lape oe ona WZ | a ie fe 2 “9 Lo eee ae in tage IGE | anne 1Gf,

Howarp & NorLINDH, DiospyRros

108 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLII

NOTE ON THE RELATIONSHIPS OF PINUS MERKUSII

NicHoLas T. Mirov

In nis Genus Pinus (Publication No. 5 of the Arnold Arboretum, 1914), George Russell Shaw with some hesitation placed Pinus merkusii Jungh. & De Vriese in the group Lariciones. Shaw wrote, “Of the habit of this pine I know nothing. . . . In my specimens the pits of the ray- cells of the wood are both large [as in the group Lariciones| and small [as in the group Insignes]. In this particular it may belong in either of two groups of species. Its uniform leaf-hypoderm associates it with this group |Lariciones| or with P. halepensis of the Insignes. I have assumed ee cone to be dehiscent at maturity and have placed it with the Larici- ones, but if further information shows the cone to be serotinous, this 8 should be transferred to the serotinous group [of the Insignes |. '

In February, 1961, I had occasion to observe Pinus merkusii in the field in Thailand, Indonesia, and the Philippines. I noticed that normally P. merkusti cones dehisce on maturity, shed their shiny ochre-yellow seeds, and remain open and attached to the trees. I could not help but admire Mr. Shaw’s wisdom, his sound judgment, and intuition that caused him to place P. merkusii in the group Lariciones. Perhaps it is true that in certain characters, such as the uniform leaf-epidermis, P. merkusii re- sembles some pines of the Insignes group, but I think this pine has much more in common with the species of the group Lariciones.

The geographical distribution of the Insignes and the Lariciones pines is shown below:

PINES OF GROUP PINES OF GROUP REGION INSIGNES LARICIONES Mexico 4 None W. America 4or5 None E. America 6 2 Europe 3 (incl. P. brutia) 3 (incl. P. sylvestris) SE. Asia None 7

It is seen from the above tabulation that the Lariciones pines are pre- dominantly species of southeastern Asia (from Japan to Sumatra). On he other hand, the Insignes pines are not found there. Therefore, the occurrence of Pinus merkusii in southeastern Asia, where most of the Lariciones pines grow, seems to me to be one more point in favor of placing this southernmost pine in the group where Shaw placed it.

The Insignes are a rather heterogeneous group, perhaps the first one of the genus Pinus to be revised in the light of its genetic characteristics. But such a revision would chiefly concern pines of the eastern United States and would not affect pines of southeastern Asia. All pines of that part of the world most likely will remain in Shaw’s group Lariciones.

Casot FOUNDATION, HARVARD UNIVERSITY

JOURNAL

OF THE

ARNOLD ARBORETUM

VoL. XLIII APRIL 1962 NUMBER 2

STUDIES IN THE GENUS JASMINUM, II THE SPECIES FROM NEW CALEDONIA AND THE LOYALTY ISLANDS *

P. S. GREEN

EXAMINATION OF THE EXTENSIVE MATERIAL of the family Oleaceae col- lected on the Franco-Swiss Botanical Expedition to New Caledonia, 1950— 1952, and generously sent on loan from the Botanic Garden and Museum of the University of Ziirich, soon indicated the need for a revision of the species of Jasminum recorded from New Caledonia. Even though Guil- laumin published a key to the species as recently as 1948 in his Flore de la Nouvelle-Calédonie (p. 283), it soon became apparent that some species required reconsideration; and, in addition, two further species, J. lineart- folium and J. velutinum, had been described since 1948. In this revision I have attempted to see as wide a range of material as possible, and although I have not seen all the specimens cited by Guillaumin in his numerous and valuable ‘Contributions to the Flora of New Caledonia,” I have been en- abled to examine the types of all the taxa concerned.

When considering specific delimitation in Jasminum, as in many genera, it is important to bear in mind the phenomenon of heterostyly. This was recorded in Jasminum as early as 1867 (Kuhn, Bot. Zeit. Leipzig 25: 67), but unfortunately it appears to have been overlooked by Guillaumin in his studies of the New Caledonian species. Jasminum artense and J. dzumacense are separated in the key just mentioned by the relative posi- tions of the anthers and style: the former-with the anthers borne in the upper part of the corolla tube above the style and the latter with the reverse condition. Ignoring this as a specific distinction here, it becomes clear there are not two species but only one. The relative positions of the anthers and styles are mentioned in two or three other places in the key as well, and, in 1914, Guillaumin described a variety of J. elatum which he called brevistylis, based as it was upon the short or “thrum” condition of the style.

* Results of the Botanical Expedition to New Caledonia 1950-52 (French-Swiss

Mission). “Studies in the Genus Jasminum, I: Secticn Alternifolia” was published in Notes Bot. Gard. Edinb, 23: 355. 1961

110 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLII

The immediate affinities of most, if not of all, species included in this revision lie with others outside New Caledonia, so it is not possible to arrange them in strict order of their suspected relationships. A rough grouping has been attempted, however, and, where two species are thought to be close enough in affinity, e.g., Jasin artense and J. linearifolium, they have been arranged together.

5 MM.

Calyx types and range of variation in the New Caledonian species Pa si inum. J. didymum, C; J. leratii, F and I to G, rarely almost to D; J. neocaledonicum, G; J. noumeense, H; J. artense, Ct o D; J. linearifolium, cer a oe A an 0, occasionally, almost oo. ey A; J. kriegeri, E awn from Pancher 316; B, Déniker 2270: C, Vieiwlard 913; D, Ddaniker 483; E, Vivillard 2057 +5. Danie 3089; G, Schlechter 15500 Paice 315% I, Daniker 1333)

a

It is perhaps significant that three out of the nine species are only known from a single collection. Furthermore, additional exploration may well indicate what can only be suspected at present; that Jasminum lineari-

1962 | GREEN, STUDIES IN JASMINUM, II ale

folium may not be separable at specific rank from J. artense and that J. promunturianum is equally close to J. elatum. The calyces of these species, as pairs, appear identical, and in Jasminum the shape and dimensions of the calyx and its teeth, although somewhat limited in the possibilities of their variation, often give convenient characters for identification and a strong indication of affinities. Jasminum kriegeri, on the other hand, known only from the one somewhat scrappy specimen collected almost one hundred years ago, has a distinctive calyx. Ficure 1 gives a diagrammatic representation of the calyces and their range of variation in the species under review.

Two alien species widely grown in the tropics throughout the world have been recorded from New Caledonia, and it is suspected that they have established themselves as escapes from cultivation. They are Jas- minum sambac (L.) Ait. (see Guillaumin, Ann. Mus. Col. Marseille II. Dell Ot Bulls Wrus: sist. Nat. Paris:25* 500.1010 Nor, Syst, Paris 35.61. as. 1914; and Fl. Nouv.-Caléd. 284. 1948) and J. multi- florum (Burm. f.) Andr. (as J. pubescens (Retz.) Willd.; see Guillaumin and Virot, Mém. Mus. Hist. Nat. Paris II (B). 4: 48. 1953).

One or two native names are recorded for various species. From the Island of Lifou (Loyalty Islands) the name “queku” is noted for Jasminum elatum on the label of Deplanch 82 (p) and the name “wechu” is reported by Daniker (Viert. Naturf. Ges. Ziirich 78(Beibl. 19): 386. 1933) as being given to J. didvmum, J. elatum, and J. leratii (as J. sim- plicifolium). From the adjacent Island of Maré, Daniker records the name “wekutsch” for J. didymum and J. leratii and Baumann-Bodenheim (14770) “wawekuce” for J. didymum. It would seem that these are gen- eral names for jasmines as a whole. Guillaumin (Candollea 5: 151. 1932) quotes information from Bergeret, who had lived in the Loyalty Islands and knew the language, that on Lifou the name “wexu tremanji’” is used for J. leratii and ‘“‘wexu foe” for J. elatum.

I should like to express my grateful thanks and appreciation to the directors and curators of the cited herbaria for the loan of material or facilities for study. All material cited has been examined, and the re- spective herbaria are indicated by the abbreviations published in Jndex Herbariorum, Ed. 4, 1959. I should particularly like to mention Dr. Alicia Lourteig, of Sas: Dr. H. Hirlimann, of Basle; and D Stauffer, of Ziirich, to whom I am much indebted for their kind foe and Cope ton and Miss Judith Kroll to whom I owe the drawings of calyces.

Rey Or ser Cis

ie ee es trifoliolate; corolla tube up to 12 mm. long and lobes up to 2.5 mm.

lr as Sap a TO alata in Re chi Sete hes a tate gd ela eT 1. J. didymum.

ile ee simple; corolla tube more than 12 mm. long (rarely as short as 8 mm. in J. elatum) and lobes more than 4 n ong.

2. Calyx teeth distinct, more than 1 mm. ice ae 1, F-I).

112 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLUI

3. Teeth of calyx narrowly lanceolate to subulate-filiform, up to 3 mm. long, usually less (Fic. 1, F to G or I); corolla tube 10-22 mm. long.

4. Corolla tube up to 20 mm. long, usually less; leaves usually narrowly ovate but a. from narrowly lanceolate to broadly

ovate, (1—)2.5-5.5(-8) cm. long. ................ 2. J. leratii.

4. Corolla tube 20-22 mm, Re leaves broadly or very broadly ovate, (4-)6-9.5(-11) cm. long. ........ 3. J. neocaledonicum.

3. Teeth of calyx lanceolate, subfoliaceous, 2-5.5 mm. long (Fic. 1, H); corolla tube 24-33 mm. long. ................... 4. J. noumeense.

2. Calyx teeth short or indistinct, up to 1 mm. long (Fic. 1, A-E) 5. Leaf length not more than twice the breadth.

6. Pedicels 5-20 mm. long; leaves usually less than 4 cm. long, rarely up to 5.2 cm., ovate to orbicular, more or less coriaceous, the margin more or less recurved and thickened. .... 5. J. artense.

6. Pedicels 0-2 mm. long; leaves usually more than 4 cm. long narrowly ovate, or if ovate to broadly ovate then always more ah 4 cm. long, more or less chartaceous, the margin not usually

rved or noticeably thickened. ................ 7. J. elatum. of Leaf length more than twice the breadth gth of leaf more than 3 cm.; leaves linear to narrowly ee apex obtuse or acute. 8. Shape of leaf ee not more than 6 mm. broad; calyx

tube 1.25 mm. lon Oy. linearifolium.

8. Shape of leaf ere lanceolate, as 1 mm. broad; calyx UDG: AS TOU. oie ye nn eRe 9. J. brerer:

7. Length of leaf up to 2.5 cm.; leaves oblanceolate or narrowly oblanceolate, apex retuse........... 8. J. promunturianum.

Section TRIFOLIOLATA DC.

Jasminum didymum Forster fil., Prodr. 3. 1786; De Candolle, Prodr. 8: 311. 1844; Drake del Castillo, Ill. Fl. Ins. Maris Pacif. 231. 1892, et Fl. Polyn. Franc. 120. 1893; Jeanneney, Nouvelle-Calédonie Agri- cole 121. 1894; Guillaumin, Ann. Mus. Col. Marseille II. 9: 191. 1911, et Bull. Mus. Hist. Nat. Paris 18: 40, 329. 1912; Guillaumin & Beauvisage, Ann. Soc. Bot. Lyon 38: 102 (‘ Species Montrou- zieranae,’ 28) 1914: Guillaumin, Not. Syst. Paris 3: 61, 65. 1914, et Bull. Mus. Hist. Nat. Paris 25: 291, 652. 1919, et in Sarasin & Roux, Nova Caledonia, Bot. 1: 207. 1921; Moore, Jour. Linn. Soc. Bot. 45: 356. 1921; Guillaumin, Bull. Mus. Nat. Hist. Paris II. 4: 701. 1932; Da niker, Viert. Naturf. Ges. ae 78( Beibl. 19) +365,.1933% Guillaumin, Bull. Mus. Hist. Nat. Paris II. 6: 458. 1934, zbid. 10: 519. 1938, ibid. 14: 455. 1942, et Not. Syst. Paris 11: 55. 1943, et Bull. Mus. Hist. Nat. Paris II. 15: 453. 1943, et Fl. Nouv.-Caléd. 283. 1948, et Mém. Mus. Hist. Nat. Paris II(B). 4: 47. 1953, et Bull. Mus. Hist. Nat. Paris IT. 27: 475. 1955, zbid. 31: 179, 1959, et Mém. Mus. Hist. Nat. Paris II1(B). 8: 161. 1959.

NE Raaerk R. Br. Prodr. 521. 1810; Labillardier, Sert. Austr.-Caled. 27. 1824-25; Endlicher, Wien Mus. Naturg. Ann. 1: 177. 1836; De

1962 | GREEN, STUDIES IN JASMINUM, II 113

Candolle, Prodr. 8: 311. 1844; Montrouzier, Mém. Acad. Lyon 10: 232.

1860; Jeanneney, Nouvelle-Calédonie Agricole, 121. ee Schlechter, Bot.

Jahrb. 39: 231. 1907; Guillaumin, Not. Syst. Paris 11: 1943.

I, ales var. stenophyllum Daniker, Viert. Naturf. on Ziirich 78(Beibl.

19): 1933.

als ee Knobl. ex Guillaumin, Bull. Mus. Hist. Nat. Paris II. 14: 456. 1942, nomen pro syn.

Evergreen climber; stems puberulent, often minutely so, or more rarely glabrous. Leaves opposite, trifoliolate; common petiole 7—17(—22) mm. long, terminal petiolule (2—)4-12 mm. long, the lateral (1—)2-10 mm. long, puberulent, especially in the groove above, or glabrous; lamina thickish or more or less chartaceous, leaflets narrowly ovate, to narrowly lanceolate (or ovate) or more or less elliptic to narrowly elliptic, the ter- minal leaflet (0.8—)2—7(-9) cm. long by (0.4-)0.8-3.5(—7) cm. broad, the lateral (1—)1.5-5(-—7) cm. long by (0.3—)0.6—-2.5(—3.5) cm. broad; margin entire, sometimes slightly thickened, often slightly recurved; apex rounded, obtuse or acute, often with an apiculus or the remains of one; base rounded, subcordate or acute, not attenuate into the petiolule; vena- tion more or less visible as a reticulation: with 4 or 5 (6) primary veins per side. /nfl ces axillary or terminal on side shoots, a contracted ternately rancned panicle, with few to many flowers, nubentiea or glabrous; bracts triangular-lanceolate or subulate, 0.5-2 mm. long; ped- icels 0.5-1.5 (—2) mm. long. Flowers white, fragrant, heterostylous. Calyx glabrous, occasionally slightly puberulent towards the base, the tube 1.5-2 mm. long with 4-6 often obscure teeth 0-2.5 mm. long (Fic. 1, C), usually slightly ciliolate. Corolla hypocrateriform, tube 5—8 (—10.5) mm. long, the lobes 4—7, ovate or broadly ovate, rounded to acute, 1.75-3 mm. long. Stamens 2, the anthers 2-3.5 mm. long, on filaments 0.5—1 mm. long. Ovary 0.75-1 mm. long, with style 2—2.5 mm. long in short- styled flowers or about 5 mm. long in long-styled. Fruit not seen in New Caledonian material.

New Caledonia. Gatope, Vieillard 913 (a, E, GH); montagnes de Gomouen pres Gatope, Vieillard 2935 (BM, K); M’bée, Vieillard 2936 (kK); superieur de Koumac, Balansa 2785 (&, K): im Tale der Koumac, am Weg von Koumac nach Quegona, bei den Kalk Klippen, im Géhodlz am Bache, 23 May 1925, Ddniker 1717 (z); Baie Banaré, Balansa 3189 (BM, E, K); buissons, La Coulée, 28 March 1928, Franc 2313 (A, us); bois de Port-Despointes (Nouméa), +50 m. alt., bois secs des collines littorales, schistes nummulitiques, 21 Nov. 1942, Virot 846 (A), ibid., + 40 m. alt., 6 Dec. 1942, Virot 882 (a); Anse Vata, Brousmiche 577 (A); Anse Vata, low calcareous hill, 23 Apr. 1955, McKee 2428 (HULL, US); Oundjo, 1958, McKee 6512 (HULL); bords de la Dumbéa, 50 m. alt., 6 June 1909, Franc 1377 (Ny); trail from Dumbéa towards Mt. Dzumac, 650 m. alt., mesophilous forest on serpentine soil, 28 Mar. 1951, Hurlimann 1092 (Reale slope south of Paoué valley (Tipindjé), 500 m. alt., meso-xerophilous forest on rocky serpentine soil, 13 Apr. 1951, Hurlimann 1166 (a, upper

forest on calcareous soil, 27 July 1950, Baumann-Bodenheim 5032 (A, ny

114 JOURNAL OF THE ARNOLD ARBORETUM [ VOL, XLII

17 Sept. 1950, Baumann-Bodenheim 6057 (a, z), and 14 Oct. 1950, Guillaumin é& Baumann- Bodenheim 6842 (A, z); Mt. Kouvelée (Dumbéa), forest on ea soil, 9 May 1951, Guillaumin & Baumann-Bodenheim 13127 (a, z);

Kaféaté, 19 May 1951, Guillawmin 13395 (a, z); Mt. Mou, gully forest, RNG 450 m. alt., 1914, Compton 455 (pm); bei Hie shéne, Apr. 1921, Heim 74 (z): Tiebaghikuppen, am Nordostabhang, auf Serpentin, im Gebtisch auf Steinen, oft und verbreitet, 14 Mar. 1924, Daniker 1436 (holotype, var. stenophyllum, z).

F Pines: Le Rat 109 oR near Kuto, on coral soil, 26 May 1951, Baumann-Bodenheim 13454 (a, z); Anse Kanumera, on coral soil, 27 May 1957, Baumann-Bodenheim 13608 (2); a aterie Island, near Isle of Pines, Arau- caria ei Pandanus forest on coral soil, 25 May 1951, Baumann-Bodenheim 13436

ee Tslands. Liou: Mou, oft an den felsigen Abhangen der Ostkiiste, 29 Nov. 1925, Daniker 24635 (z); Natalo, im Buschwalde, 2 Nov. 1925, Déniker 2465a (z); Képénéé, in Wald, 17 Apr. 1912, Sarasin 730 (z). OuvEA: May 1912, Sarasin 890 (z). Mare: Pénélo, on coral soil, 18 July 1951, Bawmann-Boden- heim 14770 (A, 2).

Easily distinguished amongst the jasmines of New Caledonia by its trifoliolate leaves, this species is nevertheless very variable. The vari- ability is most obvious in leaflet size and breadth but also in corolla size, although the corolla tube and lobes are always smaller than in any of the other species. In 1933, Daniker described his variety stenophyllum based upon a small, narrow-leafletted variant (Daniker 1436) but, to judge from the considerable number of specimens I have examined, it falls into the range of the apparently continuous variation exhibited by the species. It is worth noting, too, that Virot 846, with leaflets similar to and perhaps a little narrower than Ddniker 1436, is said, in the field notes on the label with the specimen, to be a “forme jeune’, but I have no knowledge to what extent leaflet shape and size may vary with maturity. The name “Jasminum fitzgeraldii” attached by Knoblauch to the specimen of Daniker 2465 at Zurich was never validly published, and Guillaumin, when drawing attention to the variability of J. didymum (in Bull. Mus. Hist. Nat. Paris II. 14: 455. 1942), remarks that it is only an intermediate between the linear-leafletted expression, as exemplified in the extreme by LeRat 1610 (not seen, also mentioned by Guillaumin in ibid. 6: 458. 1934) and the almost orbicular-leafletted Vieillard 2935

Jasminum didymum lies within a complex distributed over the Aus- tralasian and Malaysian areas; a complex which calls for investigation and treatment as a whole, with geographical distribution as only one of the characters taken into consideration. Jasminum didymum was the first species described in this complex; in addition to J. divaricatum R. Br., described from Australia and included in the synonymy above, the other species are: J. dallachit F. Muell. (Australia), J. degeneri Kobuski (Fiji), J. domatiigerum Lingelsh. (New Guinea), J. gilgianum K. Schum. (New Guinea), J. lineare R. Br. (Australia), J. micranthum R. Br. (Australia), J. parviflorum Decne. (Timor), J. racemosum F. Muell.

1962 | GREEN, STUDIES IN JASMINUM, II FES)

(Australia), J. rupestre Blume (New Guinea), J. smithianum Kobuski (Fiji), and J. triphyllum Merrill (Philippine Is.).

It appears that Jasminum didymum may exhibit a certain amount of androdioecism, for, despite the examination of several flowers, no styles have been found on some specimens (for example, Franc 2313 and Vieillard 913), and the ovary, whilst normal in size, is presumably abortive. The anthers however, were normal and full of pollen. In other specimens both the style and stamens were normal, although exhibiting heterostyly.

Section UNtroLioLtata DC.

2. Jasminum leratii Schlechter, Bot. Jahrb. 40(Beibl. 92): 32. 1908; Guillaumin, Ann, Mus. Col. Marseille IT. 9: 191. 1911, et Not. Syst. Paris 3: 62. 1914, et Bull. Mus. Hist. Nat. Paris II. 6: 458. 1934, ibid. 13: 476. 1941, ibid. 15: 454. 1943, et Fl. Nouv.-Caléd. 284, 1948, et Mém. Mus. Hist. Nat. Paris II(B). 4: 47. 1953, et Bull. Mus: HisteNat,.Pans Il. 272327, 475. 1955.

J. francii Guillaumin, Bull. Mus. Hist. Nat. Paris IT. 5: 323. 1933.

J. neocaledonicum Schlechter var. angustifolium Daniker, Viert. Naturf. Ges. Zurich 78(Beibl. 19): 366. 1933.

J. daenikeri Knoblauch, Repert. Sp. Nov. 41: 150. 1936.

J. paagoumenum Knoblauch, Repert. Sp. Nov. 41: 150. 1936.

J. absimile L. H. Bailey, Gent. Herb. 4: 346. 1940, Hortus Second 397. 1941, et Man. Cult. Pl. ed. 2. 797, 798. 1949: P. Riedal (Calif. Arboretum Foundation), Pl. for Extra-Trop. Regions 335

J. noumeense Schltr. var. microphyllum Guillaumin, Bull. Mus. Nat. Paris II. 1 54. 1943

“J. simplicifolium Forst. f.” sensu Guillaumin, Ann. Mus. Col. Marseille II.

1911, et Not. Syst. Paris 3: 62, 65. 1914, et Bull. Mus. Hist. Nat. Paris 25: 652. 1919, et in Sarasin & Roux, Nova Caledonia, Bot. 1: 207. 1921; Moore, Jour. Linn. Soc. Bot. 45: 356. 1921; Guillaumin, Candollea 5: 151. 1932; Daniker, Viert. Naturf. Ges. Ziirich 78(Beibl. 19): 368. 1933; Guillaumin, Bull. Mus. Hist. Nat. Paris II. 5: 323. 1933; ibid. 6: 458, 1934, ibid. 15: 454. 1943, et Fl. Nouv.-Caléd. 284. 1948

“J. australe Pers.” sensu Montrouzier, Mém. Acad. Lyon 10: 231. 1866; Guillaumin & Beauvisage, Ann. Soc. Bot. Lyon 38: 123 (‘‘Species Mont- rouzleranae,” 49) 1914.

Evergreen diffuse shrub, climber, or liane of variable length; the stem glabrous or minutely puberulent, especially when young. Leaves opposite, simple; petioles 1.5-10(—20) mm. long, glabrous or minutely puberulent, articulated 14 to ¥% way from base; lamina more or less chartaceous, (broad ovate to) ovate to lanceolate (to linear lanceolate), (1—)2.5—5 (8) cm, long by (0.4—)1.5-4(—5) cm. broad; margin entire, not or scarcely thickened, rarely slightly recurved; apex obtuse or acute, sometimes more or less rounded or apiculate; base rounded or obtuse, sometimes more or less truncate or acute or even angustate, often slightly attenuate into the petiole; venation more or less obscure or reticulate or with primary veins

116 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLII

only visible above and below, 3—5(6) per side, the upper 3 or 4 often anastomosing towards the margin. /mflorescences terminal on side shoots or axillary, ternately paniculate with 3—9(—18) flowers per panicle, gla- brous or minutely puberulent; bracts linear-filiform, 1-4 mm. long or occasionally approaching foliaceous and slightly longer; pedicels 2—20 mm. long. Flowers white, feebly to strongly fragrant, heterostylous. Calyx glabrous or minutely puberulent, tube (1.5—)2-2.5(—3) mm. long with 4—6, triangular-lanceolate teeth, (0.5—)2—3.5(—4) mm. long (Fic. 1, ranging from G, through I and F, to rarely almost D), more or less ac- crescent in fruit. Corolla hypocrateriform, tube 11-20 mm. long, lobes 5-7, lanceolate or narrowly lanceolate, acute, 7-12 mm. long. Stamens 2, anthers 3-4.75 mm. long on filaments 0.5-2 mm. long. Ovary 1—1.25 mm. long, style 3-6.5(—9) mm. long in short-styled flowers or 11-16 mm. long in long-styled. Fruit ellipsoid-ovoid, paired (or single by abortion) 6-8 mm. long by 5—6 mm. broad.

ew Caledonia. Magenta, Jan. 1903, Le Rat 172 (isotype, Pp); Noumeéa, Franc 1544 (syntype J. francii, p), Franc 1544A (syntype J. francii Pp; isosyn- type, A); Port de France (Nouméa), 1855-60, Vieillard 911 (A, G, K); Port Despointes, 50 m., bois secs des ee littorales, schistes nummulitiques, 6 Dec. 1942, Virot 876, 881 (A), ibid., 13 Dec. 1942, Virot 922, 927, 929 (a); Prony, coteaux niles: roches ales aise 20 Fe b. 1914, Gi 1906 (A, 2); Koné, oft im niederen Gebusch der Kiistenhiigel, 11 Feb. 1925, Daniker 770a, 3088 (z); an der Miindung des Flusses von Koné, zerstreut im Gebische der sumpfigen Flussdeltas, 10 Feb. 1925, Daniker 1171 (holotype J. neocaledonicum var. angustifolium & J. daenikeri, z); bei Paagoumen, im Grasbestand auf Serpentingebiisch, 22 March 1925, Ddaniker 1518 (holotype J. paagoumenum, z); Ile de la Table, bei Koumac, in Acaciengebtisch, 5 March 1925

1333, 1335 (z); auf der Insel Neba, im Sumpfwaldchen bei dem Teiche, 8 April 1925, Ddniker 1599 (z); am Anse longue bei Carticaté, St. Vincent, im Ufergehélz, 12 Dec. 1924, Daniker 770 (z); Anse Vata, low calcareous hill, remnant of dry forest with trees close together, 23 Jan. 1955, McKee 1973 (A, HULL, US), 475 and 2342 (HULL); slope of Ouen Toro, 22 March 1955, McKee 2251, (A, US); Ouen Toro, promenade Pierre Vernier, 10 m. alt., climax littoral a Acacia spirorbis, nummulitique, terraine schisto-calcaire, 30 Jan. 1941, Virot 442 (a); N.W. of Mt. Natégou, beneath the road to Yaté, shrubby forest on rocky serpentine ridge, 270 m. alt., 29 Jan. 1951, Hirlimann 769 (A, z); “Bois du Sud” (towards the Yaté valle y, Spermolepis forest on serpentine, 29 March 1951, Guillaumin & Baumann-Bodenheim 11663 (A, z); south of Mt. Poindas (Tip SS mesophilous forest on serpentine ridge, 520 m. alt., 21 April 1951, Hiurlimann 1290 (a, z); Mt. Koniambo, on serpentine soil, 21 Dec. 1950, Guillaumin & Baumann-Bodenheim 9514 (a, z); Mt. Kaféaté, on serpentine soil, 22 Dec. 1950, Guillaumin & Baumann-Bodenheim 9665 (A, z), ibid., on cal- careous soil, 10 April 1951, Guillaumin & Baumann-Bodenheim 12043 (A, Z); Moro Island, on coral soil, 5 & 6 March 1951, Guillaumin & Baumann-Boden- heim 11106, 11135, 11161 (a, z); Col de Vulcain, xerophilous forest on serpentine soil, 900 m. alt., 12 Nov. 1950, Baumann-Bodenheim 8285 (A, 2); Gatope, 1861-67, Vieillard 2938 (a, G, GH, P) and 2935 (holotype J. noumeense var. microphyllum, vp); Voh, March 1921, Heim 49 (z). IsLE or PINEs:

1962] GREEN, STUDIES IN JASMINUM, II 117

Germain 13 (A); Kuto peninsula, on calcareous soil, 28 May 1951, Baumann- Bodenheim 13642 (a, z)

Loyalty Islands. Lirou: bei Képénéhé, im Buschwalde, 31 July 1925, Daniker 1886 (z), ibid., 22 Oct. 1925, Daniker 2269 (z), 19 April 1912, Sarasin 185 (z); in der Umegebung von Natalo, oft im Gebiisch, 2 Nov. 1925, Denver 2388 (z); zwischen Natalo und Tingeting, in der Savanne zerstreut, 2 Nov. 1925, Déaniker 2388a (z); Kiatheng, in der Savanne, 21 Nov. 1925, Daniker 3089 (z). Ouvea: May 1912, Sarasin 885 (z); Fayaoué, in den Gebiischen, 4 Aug. 1925, Déniker 1909 (z). Maré: uberall auf dem Plateau von Maré, 29 Nov. 1911, Sarasin 424 (z); Eneni, oft in der inneren Ebene von Maré, besonders lings der Wege, 29 Dec. 1925, Daniker 2626 (z); Tadine, forest on coral soil, 12 July 1951, Baumann-Bodenheim 14529 (a, z); Rawa, mesophilous forest on coral soil, 17 July 1951, Baumann- Bodenheim 14721 (A, z) without locality, Jan. 1908, Franc. 1021 (ny ie

tie

a rol ° TE oe Kee be ce Et

it

ese

i) o

laced ios re)

AONANDSYS

5 LEAF : AVERAGE LENGTH IN CM.

Fic. 2. Frequency histograms of leaf eee in Jasminum sei (L}, volubile (V), and J. simplicifolium (S). LH1 and NH indicate the posi of rf specimens examined from Lord Howe eee and the New Hebrides eee ae.

118

Cult

1917, cae enone

(BH.

Fic. A fact

4

o. ext

FREQUENCY °

BH )

JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIII

(A);

San Dieg

ivated. United States of oe California: Santa Barbara, 27 July g

19, Woodcock 630 (holotype of J.

; Golden Gate Park. San Francisco, 23 March 1927, Bailey 9623 -R, Menzie’s place, Mill Valley, 24 July 1927, Walter 65 (A).

Greatest leaf breadth factor: measurement B divided by measurement ae

LEAF °

i

9 32 35 .38 41 44 47 80 GREATEST BREADTH FACTOR

Fic. 4. Frequency histograms for greatest leaf breadth factors in Jasminum leratii (L), J. volubile (v), and J. simplicifolium (s). LHI and Nu indicate the position of the specimens examined from Lord Howe iad and the New Heb- rides respectively.

1962 | GREEN, STUDIES IN JASMINUM, II 119

A close relationship exists between this species, Jasminum sim plicifolium Forst. f. and J. volubile Jacq. (J. gracile Andr.), and, after considerable study of material of all three, I have come to the conclusion that, for the present at least, they should be treated as separate species. Without doubt all three lie very close together and it is not without significance that J. leratti was placed as a doubtful synonym under J. simplicifolium by Guillaumin in 1921 (in Sarasin & Roux, Nova Caled. Bot. 1: 207. 1921) and later by Daniker (Viert. Naturf. Ges. Ziirich 78(Beibl. 19): 368. 1933), although the two were maintained as distinct in Guillaumin’s rela- tively recent Flore de la Nouvelle-Calédonie 284. 1948. Bailey (Gent. Herb. 4: 342. 1940) separates J. volubile (as J. gracile) and J. simplici- folium and inférs that the former is confined to Australia, an inference supported by this investigation. For this study a careful analysis has been made of leaf length, distance from the base to the greatest breadth of the leaf, length of calyx teeth, length of pedicel, and length of corolla tube; and each has been considered in relation to geographical distribu- tion. The results may be expressed as follows: (i) LEAF LENGTH. The extra-Novo Caledonian J. simplicifolium (from Fiji & Tonga) has longer leaves than either J. leratii (from New Caledonia) or J. volubile (from Australia), although, using the average leaf length of a specimen, a slight overlap occurs around 5.5-6 cm. (Fic. 2). (ii) GREATEST LEAF BREADTH FACTOR. Bailey (doc. cit.) indicated the possible value of the position of greatest breadth of the leaf in distinguishing between these species. In this analysis a “greatest leaf breadth factor” was calculated by measuring the total length of a leaf, and dividing by the distance up the midrib to the position of greatest breadth (see Fic. 3; B divided by A). Using this method it was found that J. simplicifolium had the lowest factor, less than 0.3 whilst J. leratii and J. volubile both had factors ranging from 0.3 to 0.45 (Fic. 4). That is to say, J. simplicifolium has leaves which have their greatest breadth in the lower third and the other species between the lower third and middle of the leaf. (iii) LencrH OF CALYX TEETH. Jasminum volubile is more or less separable from J. leratii and J. simplici- folium on this character. In the first species, the teeth range from prac- tically nothing to no more than an average of 0.75 mm. in length, usually less, whilst in the latter two species they vary from an average of just under 0.75 mm. to 2 mm. in J. simplicifolium and to nearly 4 mm. in J. leratii, which is very variable in length of calyx teeth (Fic. 5). (iv) LENGTH OF PEDICEL. There is no clear cut separation of the taxa on this character but only a very slight tendency for J. volubile to have shorter pedicels than the other two species. (v) LENGTH OF COROLLA TUBE. Jasminum volubile is distinguishable by its shorter corolla tube which, in the material examined, ranges from about 8 to 11 mm., whereas in J. lerata and J. simplicifolium (which together seem indistinguishable on this character) it ranges from 11-20 mm. (Fic. 6).

These details may be summarized by saying that Jasminum leratii has leaves which are shorter and slightly more elliptical than those of J.

JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIII

FREQUENCY

t 5 | U

] | T : 2 3 4 CALYX TEETH : AVERAGE LENGTH IN MM. . Frequency histograms for length of calyx teeth in Jasminum leratu (L), J. volubile (v), and J. simplicifolium (s). Lui and NH indicate the position of the specimens from Lord Howe Island and the New Hebrides respectively.

5

FREQUENCY a pit WN

1.0 COROLLA TUBE : AVERAGE LENGTH IN CM.

Fic. 6. Frequency histograms for length of corolla tube in Jasminum (L), J. volubile (v), and J. simplicifolium (s), Lut indicates the position i i d.

leratit of specimens examined from Lord Howe Islan

the

1962] GREEN, STUDIES IN JASMINUM, II 121

simplicifolium and generally has a larger corolla tube and calyx teeth than J. volubile.

Whilst not strictly observations on New Caledonian material, it is interesting to make two notes on allied material. The only specimen seen from the adjacent New Hebrides (Kajewski 404, a, NY) is Jasminum lerati. It was originally identified as J. simplicifolium (Guillaumin, Jour. Arn. Arb. 13: 17. 1932), but, although it has no corollas on it to examine and measure (being in fruit), in leaf dimensions and shape, and in calyx type, it falls into the range of J. leratii. The two specimens examined from Lord Howe Island in this investigation (McCornish 4 [A] and “Phytologic Museum of Victoria, Government Botanist Melbourne” [E]) have leaves and calyx teeth which in length lie on the edge or outside the ranges of the Australian specimens of J. volubile. Yet in greatest leaf breadth factor and length of corolla tube they agree with the rest of the Australian material.

At the same time as Jasminum leratii, J. volubile, and J. simplicifolium were being investigated, the opportunity was taken to examine the cultivated species J. absimile L. H. Bailey, and a separate note about this will appear in Baileya. It is sufficient to point out here that J. absimile is conspecific with J. leratii, although in the former there are usually more flowers per inflorescence and the leaves are a little larger on the average. But both these differences can be accounted for by favorable conditions of growth in cultivation.

Jasminum leratii is very variable in leaf shape and size, and it is sus- pected that this may be a reflection of the habitat conditions under which the plants are growing. Some of the largest leaves (up to 7 cm.) are exhibited by a specimen of Ddniker 1909; yet under this same number is a shoot with quite typical leaves half their size. Ddniker 1171 exhibits leaves which are consistently small and narrow. Because of this it was used as the type of both J. daenikeri and J. neocaledonicum var. angustifolium, but some of the leaves on Le Rat 172, the type of J. leratii, are an exact match, as are those of Guillaumin & Baumann-Bodenheim 11663, in which they range from broad-ovate to linear, although con- siderably larger than in Le Rat 172. Similarly, a small-leaved scrap (Vieil- lard 2935) was the basis of the published name J. noumeense var. miicro- phyllum (although the type sheet at Paris bears only the name J. nco- caledonicum var. microphyllum in Guillaumin’s hand). The leaves are shorter than those of Ddaniker 1171 but they can be matched exactly by the smallest leaves of several other specimens and in floral characters this specimen is identical with the rest of J. leratit.

Dr. Hirlimann has remarked in correspondence that it is uncommon for endemic species in New Caledonia to occur on calcareous as well as on serpentine soils, but using dried specimens I have been unable to discern any consistent differences between individuals from the two habitats. It is possible that there are differences which are masked in the herbarium by the limitations of the material and the wide variability of the species,

122 JOURNAL OF THE ARNOLD ARBORETUM [ VOL, XLIII

and an examination of populations in the field with this point in mind might well prove profitable.

Although Guillaumin described Jasminum francii in 1933, he placed it in synonymy the following year (see Bull. Mus. Hist. Nat. Paris Il. 6: 458. 1934) after he had seen authentic material of Schlechter’s species.

When describing Jasminum kriegert, Guillaumin cited two numbers, one of which was Vieillard 2938; however, all the specimens of this number which | have seen, including one from Paris, are clearly J. leratii. This number is presumably a mixed gathering, for Guillaumin (in Bull. Mus. Hist. Nat. Paris II. 15: 453, 454. 1943) referred Vieillard 2938 to both J. kriegeri and J. leratii, adding “pro parte” to the citation under the latter.

3. Jasminum neocaledonicum Schlechter, Bot. Jahrb. 39: 231. 1906; Guillaumin, Ann. Mus. Col. Marseille II. 9: 191. 1911, Not. Syst. Paris 3: 65. 1914, et Bull. Mus. Hist. Nat. Paris 25: 652. 1919, ibid. II. 2: 169. 1930; Daniker, Viert. Naturf. Ges. Ziirich 78(Beibl. 19): 366. 1933; Gullioumin, Bull. Mus. Hist. Nat. Paris II. 13: 127, 1941, ibid. 20: 371. 1948, et Fl. Nouv.-Caléd. 284. 1948, et Bull. Mus. Hist. Nat. Paris IT. 27: 475. 1955, ibid. 30: 397. 1958.

J. pulchrefoliatum Guillaumin, Not. Syst. Paris 3: 62, 65. 1914, et Bull. Mus. Hist. Nat. Paris 25: 500. 1919, et in Sarasin & Roux, Nova Caledonia, Bot. 1: 207. 1921, et Bull. Mus. Hist. Nat. Paris 27: 560. 1921, zbid. II. 4: 701. 1932; Daniker, Viert. Naturf. Ges. Ziirich 78(Beibl. 19): 367. 1933; Guillaumin, Bull. Mus. Hist. Nat. Paris IIT. 10: 520. 1938, et Fl. Nouv.- Caléd. 284. 1948

Evergreen liane; stem glabrous, even when young. Leaves opposite, simple; petioles 6-35 mm. long, glabrous, articulated 4%—'™% from the base, occasionally on the largest leaves an additional articulation toward t:e top; lamina broadly ovate to very broadly ovate, (4—)6—9.5(-11) cm. long by (3- a —8.5(—10) cm. broad; margin entire, not noticeably thickened or recurved; apex acute or acuminate, more rarely obtuse or even retuse: base pounced. truncate or subcordate, only slightly attenuate into the petiole; venation more or less evident, especially above, primary nerves (4)5-7 per side. Juflorescences axillary and terminal on side shoots, ternately paniculate, (5—)7—50 flowered, glabrous; bracts more or less linear subulate, 2-7 mm. long; pedicels 3-10 mm. long. Flowers, white, fragrant, heterostylous. Calyx glabrous, tube 2-3 mm. long with 4-6 subulate-lanceolate teeth, 1-2.5 mm. long, often slightly recurved (Fic. 1G). Corolla hypocrateriform, tube 20-22 mm. long, lobes 5—7, lanceolate, acute, about 13 mm. long. Stamens 2, anthers 4.75-5 mm. long on fila- ments 0.25—0.5 mm. long. Ovary 1.25-1.5 mm. long, style about 10 mm. long in short-styled flowers and 20 mm. long in long-styled. Fruits (one loose fruit only seen, Bawmann-Bodenheim 5581), ellipsoid, 22 mm. long by 9 mm. broad.

1962 | GREEN, STUDIES IN JASMINUM, II 123

New Caledonia. Bei Oubatche, in den waldern der Berge, c. 900 m. alt., 27 ee 1902, Schlechter 15586 rel ea B [destroyed, not ea: isotypes, BM,

, K, Z); entre Bourail et Houailou, montagnes, c. 400 m. alt., 6 Feb. 1912, coe 525 (syntype J. ee iii: P [photograph seen]; isosyntype, z); Bourail, Balansa 1299 (Pe); Prony, buissons du littorale, Jan. 1915, Franc 1935 (P); road to Montagnes des Sources, 650 m. alt., 4 March 1955, McKee 2198 (us); south of Mt. Kongouaouri, mesophilous forest on serpentine, 300 m. alt., 10 Sept. 1950, Baumann-Bodenheim 5874, 5881 (A, z); Mé Amméri, 700 m alt., on serpentine soil, 28 Nov. 1950, Guillaumin & Baumann-Bodenheim 8883 (z); Mé Aoui, hygrophilous forest on schistaceous soil, 500 m. alt., 7 & 8 Feb. 1951, Guillaumin & Baumann-Bodenheim 10301, 10440 (a, z).

A handsome species with large leaves and usually numerous flowers to the inflorescence which must be very becutiful as a living plant. Vegeta- tively it is sometimes very similar to large-leaved examples of Jasminum elatum, but the two may easily be distinguished by their calyces and, to a limited extent, by the position of the petiolar articulation (see under J. elatum). To judge from some of the specimens examined, there is also a tendency for specimens of J. meocaledonicum to blacken on drying.

In his protologue for this species Schlechter cites his specimen no. 15585 but this must be a misprint, for 75586 is the number on all the Schlechter specimens examined, and, in the same publication (Bot. Jahrb. 39: 224. 1906), he cites no. 15585 as the type of another of his species: Leucopogon septentrionalis.

4. Jasminum noumeense Schlechter, Bot. Jahrb. 40(Beibl. 92): 33 1908; Guillaumin, Ann. Mus. Col. Marseille IT. 9: 191, 1911, et Not. Syst. Paris 3: 65. 1914, et Bull. Mus. Hist. Nat. Paris IT. 6: 458. 1934, ibid. 10: 519. 1938, ibid. 15: 454. 1943, et Not. Syst. Paris 11: 55. 1943 (sphalm. J. noumeanum), et Fl. Nouv.-Caléd. 284. 1948, et Mém. Mus. Hist. Nat. Paris II(B). 4: 48. 1953, et Bull. Mus. Hist. Nat. Paris IJ. 27: 475. 1955.

J. bouquetii Jeanneney, Nouv.-Caléd. eae ie oa nom. nud.; Guil- laumin, Not. Syst. Paris 11: 55. 1943,

ne Hiaeeniae Guillaumin, Bull. Mus. Hist. "Nat Paris TL 2: 169. 1930; zbid. 4: 697.

Evergreen liane of variable length; stem puberulent or occasionally glab- rous. Leaves opposite, simple; petioles 3-17 mm. long, puberulent or occasionally glabrous, articulated near the base; lamina more or less chartaceous, ovate to very broadly ovate, (2—)3-5(-—6) cm. long by (1.5—) 2-3.5(—5) cm. broad; margin entire, not thickened, occasionally very slightly recurved; apex rounded to obtuse, or retuse, occasionally acute, often with an apiculus; base truncate or occasionally rounded or sub- cordate, only slightly attenuate into the petiole; venation with primary nerves usually visible, sometimes more or less obscure, 3—5 per side. J/n- florescences terminal or axillary on side shoots, ternately subcorymbose- paniculate, with up to about 20 flowers per panicle, usually pubescent,

&

124 JOURNAL OF THE ARNOLD ARBORETUM [ VOL, XLIII

especially towards the base of the inflorescence, occasionally glabrous; bracts more or less foliaceous, especially the basal ones, and up to 2 cm. long, occasionally the upper ones ranging to subulate, 1-2 mm. long; pedicels 3-12 mm. long. Flowers white, fragrant, sometimes strongly so (McKee 2728) or inodorous (Ddniker 1846), heterostylous. Calyx pubes- cent, occasionally glabrous, the tube 2—3(—4) mm. long, with 4 or 5 sub- foliaceous teeth 2—5.5 mm. long, often wavy margined (Fic. 1,H), some- what accrescent in fruit. Corolla hypocrateriform; tube 24-33 mm. long, lobes 5, lanceolate or broadly lanceolate, acute or very acute, 8.5—15 mm. long. Stamens 2; anthers (3—)4—5.5(—6) mm. long, on filaments 0.2-2 mm. long. Ovary 1-1.25 mm. long; style 10-11 mm. long in short-styled flowers or 20-30 mm. long in long-styled. Fruits (Balansa 2232, Virot 842) bi- lobed, united for about 24 length of each half, 8-9 mm. long by 7-8 mm. broad, occasionally single sided with a lateral projection 2-3 mm. in diameter representing the abortive loculus.

New Caledonia. Magenta (near Nouméa), June 1903, Le Rat 574 (isosyntype, p) also, May 1903, Le Rat 588 (isosyntype, Pp); bois de Port Despointes (Nou- méa), +50 m. alt., bois secs des collines littorale, schistes nummulitiques, 21

(a); Kiistenhiigel der Halbinsel Nouméa, in den Buschwaldchen am Abhang, 24 July 1925, Daniker 1846 (z); Nouméa, Aug. 1926, Franc 2232 (holotype J. magentae, P; isotypes, A, NY, US); Nouméa, foot of Ouen Toro, just above beach, 6 July 1955, McKee 2728 (A, HULL); Anse Vata, remnant of dry forest on low calcareous hill, 27 March 1955, McKee 2298 (a, HULL, us). Without locality, Franc 845 (A, BM, E, K, NY, P, US); Vieillard 2936 (BM); Pancher 315 (A) and Pancher sn. (labeled J. bouquetii, a).

One of the most distinctive features of this species is the large, almost subfoliose, calyx lobes (Ftc. 1,H). The corolla tube, too, is longer than in any other New Caledonian jasmine. Vegetatively it can be very similar to Jasminum artense; for vegetative characters which can be used to dis- tinguish between them see under that species.

It was recognized as a distinct species as early as 1894 when, without description, Jeanneney used the name Jasminum bouquetii. In 1930, Guil- laumin published his J. magentae with a short description in French, and two years later (in Bull. Mus. Hist. Nat. Paris II. 4: 697. 1932) he amplhi- fied the description and published it in Latin. Later (7bid. 6: 458. 1934), after having seen syntype material of Schlechter’s species published in 1908, he placed it in synonymy under J. noumeense, where it undoubtedly belongs.

5. Jasminum artense Montrouzier, Mém. Acad. Lyon 10: 231. 1860; Guillaumin. Ann, Mus. Col. Marseille IT. 9: 191. 1911, et Bull. Mus. Hist. Nat. Paris 18: 468. 1912, et Not. Syst. Paris 3: 61, 65. 1914; Guillaumin & Beauvisage, Ann. Soc. Bot. Lyon 38: 102 (“Species Montrouzieranae,’ 28) 1914; Dianiker, Viert. Naturf. Ges. Zurich 78

1962 | GREEN, STUDIES IN JASMINUM, II 25

(Beibl. 19): 365. 1933; Guillaumin, Fl. Nouv.-Caléd. 284. 1948, et Mém. Mus. Hist. Nat. Paris II(B). 8: 161. 1959

J. dzumacense Guillaumin, Not. Syst. Paris 3: 63, 65. 1914; Daniker, Viert. Naturf. se Zurich 78(Beibl. 19): 366. 1933; Guillaumin, Fl. Nouv.-Caléd. 284. 1948, et Mém. Mus. Hist. Nat. Paris II(B). 4: 47. 1953, ibid. 8: 161. 1959.

J. velutinum Guillaumin, Mém. Mus. Hist. Nat. Paris II(B). 4: 48. 1953, non Kobuski (1941); Virot, Vég. Canaque, 175. 1956.

Evergreen liane; stem glabrous or puberulent, rarely somewhat velutin- ous, especially when young. Leaves opposite, simple; petioles 3-20 mm long, glabrous or puberulent, rarely velutinous, articulated 4—'% from the base, sometimes on larger leaves showing an additional articulation toward the top; lamina more or less coriaceous, glabrous or rarely puberulent, ovate to orbicular, rarely narrowly ovate, (1.2—)2—4.5(—5.5) cm. long by (1—)1.5-4(—5.2) cm. broad; margin entire, slightly thickened, usually re- curved; apex rounded, obtuse or even retuse to acute, often with a small apiculus; base rounded or subcordate, rarely obtuse; venation with primary veins visible, sometimes obscure above, (2)3 or 4 per side, occasionally reticulate. Inflorescences axillary and terminal on side shoots, ternately paniculate, with 5—7(—15) flowers per panicle, glabrous or puberulent, rarely velutinous; bracts subulate, 1-1.5 (—3) mm. long; pedicels 5—20 mm. long. Flowers white, fragrant, heterostylous. Calyx glabrous or puberulent, rarely velutinous; tube 1.5—2.5 mm. long with 4 or 5 shallowly triangular teeth 0.1-0.7 mm. long (Fic. 1, C to D). Corolla hypocrateri- form; tube 10-20 mm. long with 5—8 lanceolate or narrowly lanceolate, acute (to rounded) lobes 8-13 mm. long. Stamens 2; anthers 2.75—3.5 (—5) mm. long on filaments 0.5—1 mm. long. Ovary 1-1.25 mm. long; style about 7 mm. long in short-styled flowers or 9-17 mm. long in long-styled. Fruit (Balansa 3629) ovoid-ellipsoid, paired (or single by abortion) 10-11 mm. long by 7-8 mm. broad.

New Caledonia. Valleé du Diahot (Dothio), collines eruptive, Jan. 1872, Balansa 3629 (syntype J. dzumacense, P; isosyntypes, BM, E, K); . Dzumac, Jan. 1909, Le Rat 165 (syntype, J. amen Pp); bords des creeks desecheés, base des versants ouest du Mt. Kaala, + 20 m. alt., maquis serpentineux, 10 Nov. 1943, Virot 1302 (holotype J. velutinum, Pp); im Tale des vom Mt. Hum- boldt kommenden Zuflusses der Kalouéhola und an den Abhangen des Tontou- tatales, im Serpentingebiisch, 18 Nov. 1924, Daniker 483 (z) and 10 Nov. 1924, Daniker 588 (z); am Mt. Bacon am dusseren Abhang des siidlichen Serpentin- massivs gegen die Tontoutaebene, im Buschwalde, 15 Nov. 1924, Daniker 700 (z); Prony, 1890, Bougier (Ly); Ile Art, in montanis et silvis, Sept. & March, Montrouzier 143 (holotype, P); slope west of “Baie des Pirogues,”’ on serpen- tine soil, 40 m. alt., 16 Dec. 1950. Hiirlimann 355 (a, z), ibid., meso-xeroph- ilous shrubby forest on serpentine soil with rocks, 170 m. alt., 17 Dec. 1950, Hiirlimann 381 (A, z); range of Mt. Podchoumié (Dumbéa), on rocky serpentine ridge, 180 m. alt., 23 July 1951, Hiirlmann 1625 (a, z); Col de Plum (Mt. Dore), on serpentine soil, 200 m. alt., 22 Aug. 1950, Baumann-Bodenheim 5577, 5593A (A, 2), ibid., 8 Nov. 1950, Guillaumin & Baumann-Bodenheim 7879, 7933

126 JOURNAL OF THE ARNOLD ARBORETUM [ VOL, XLIII

z); Col de Vulcain, serpentine bush, 900 m. alt., 11 Nov., 1950, Bauwmann- es ee 8178 (A, 2). IsLe or Pines: south of Pic Nga, mesophilous forest on serpentine, 29 May 1951, Baumann-Bodenheim 13699 (a, z), and creek on the southwest slope of Pic Nga, mesophilous forest, 30 May 1951, Baumann-Boden- heim 13786, 13789, 13816 (A, 2)

Jasminum artense is a distinct species which may perhaps be confused vegetatively with J. noumeense, but the leaves of the latter tend to be thinner in texture and the articulation of the petiole appears never to be more than 2 mm. from the base, even on the longer petioles. In J. artense the articulation is usually at least 2 mm. from the base, except in the case of very short petioles, where however, it is found at least between one- third and one-half of the way from the base of the petiole. When material bears an inflorescence there is no difficulty in distinguishing the two species, for the calyx lobes are quite distinct: in J. artense short and broadly triangular, in J. noumeense long and almost subfoliaceous.

Jasminum dzumacense was separated from J. artense on the relative lengths and positions of the style and stamens, but I cannot separate them as species in view of the frequency of heterostyly in the genus. In 1953, Guillaumin also distinguished and published the name of J. velutinum, but examination of the type (Virot 1302, not 1303 as cited in the protologue) shows that in everything but its velutinous indumentum the plant matches J. artense. Furthermore, J. artense has been found to be very variable in the amount of its pubescence, and Virot 1302 is only the extreme expression of this character. Similar treatments of hairy variants as species have oc- curred in other species of Jasminum, e.g. J. angulare, J. humile, J. sub- humile, and J. floridum (see Verdoorn, Bothalia 6: 560. 1956, and Green, Notes Bot, Gard. Edinb. 23: 355. 1961), and in each case an examination of more extensive collections than had been available earlier has shown that separation at specific rank cannot be justified.

In addition to the collections cited above, there is a specimen without sign of flowers or fruit (south slope of Ouen Toro, near Nouméa, on cal- careous soil, 30 Oct. 1950, Guillaumin 7414, |A, z|) which appears very similar to Jasminum artense. However it bears orbicular, retuse leaves which are especially small, varying from 0.8 cm. to 2 cm. in diameter, and was gathered from a plant growing on calcareous soil. All the specimens of J. artense of which there is mention of the soil type in their field notes are from serpentine soil and it would be interesting to see more material of this plant, especially flowering material.

6. Jasminum linearifolium Guillaumin, Mém. Mus. Hist. Nat. Paris II(B). 4: 47. 1953; Virot, Vég. Canaque, 175s 1956.

Evergreen liane of variable length or diffuse shrub; stems glabrous, slender. Leaves opposite, simple; petioles (2—)3—5 mm. long, glabrous, passing imperceptibly into the lamina, articulated at approximately the middle; lamina linear or very narrowly lanceolate, (3-)3.5-5(—7) cm. long by (0.1—)0.2—-0.4(—0.6) cm. broad; margin entire, slightly thickened;

1962 | GREEN, STUDIES IN JASMINUM, II Lea apex acute; base narrowly attenuate into the petiole; venation obscure above and below with 3 or 4 primary veins per side. /nflorescences terminal on side shoots, ternately branched with about 5 flowers per inflorescence, glabrous; bracts subulate, 0.5-2 mm. long; pedicels 5-10 mm. long. Flow- ers white. Calyx glabrous: tube 1-1.25 mm. long with 4 or 5 shallowly triangular teeth, 0.1-0.2 mm. long (Fic. 1, C). Corolla hypocrateriform: tube 16 mm. long with 4 narrowly lanceolate, acute or obtuse lobes 10 mm. long. Stamens 2; anthers 3.75 mm. long on filaments 2.5 mm. long. Ovary globose (fide Guillaumin); style 5 mm. long (fide Guillaumin) in a short- styled flower. Fruit unknown.

New Caledonia. Sommet du Dome de la Tiébaghi. +600 m. alt., maquis ser- pentineux, 27 Oct. 1943, Virot 1275 (holotype. P).

The specimen upon which this species is based is unique amongst all the material examined in this investigation. However, one gathering of Jasminum lerati (Guillaumin & Baumann-Bodenheim 11663) bears some shoots and leaves which approach (and in their extreme match) in slender- ness and narrowness the stem and leaves of J. dinearifolium. But a glance at the calyx shows that this species is not immediately related to J. leratii. Rather is the relationship with J. artense and, in fact, except for the leaves, the two species are identical, and further collections may perhaps produce intermediates. Yet the difference between linear leaves and ovate or or- bicular ones is so great that, for the present at least, it is felt best to main- tain this species, based though it is, upon a single few-flowered specimen with only one corolla.

Jasminum elatum Pancher ex Guillaumin, Not. Syst. Paris 3: 63, 65. 1914, et Bull. Mus. Hist. Nat. Paris IJ. 1: 217. 1929 et Candollea §: 151. 1932; Daniker, Viert. Naturf. Ges. on 78(Beibl. 19): 366. 1933; Guillaumin, Fl. Nouv.-Caléd. 284.

J. elatum var. brevistylis Guillaumin, Not. Syst. Paris 3: 64, 1914, et in Sara- sin & Roux, Nova Caledonia, Bot. 1: 207. 1921, et Bull. Mus. Hist. Nat. Paris IT. 15: 453. 1943.

Evergreen climbing shrub; stems glabrous or puberulent, often minutely so and especially when young. Leaves opposite, simple; petioles 3-22 mm. long, glabrous or puberulent, articulated near the base; lamina more or less chartaceous, narrowly ovate to broadly ovate, or even very broadly so, (2.5—)4-8(-10) cm. long by (1.2—)2—7(-—7.5) cm. broad, margin entire, not usually recurved or noticeably thickened; apex obtuse or acute, or more rarely slightly retuse, often with an apiculus; base truncate, rounded, obtuse or cuneate, often more or less attenuate into the petiole and rarely almost subcordate; venation usually with primary nerves only visible above and below, 4-7 per side. /zflorescences axillary or terminal on side shoots, dense, ternately subcorymbose, many flowered, glabrous to densely and minutely puberulent; bracts subulate, narrowly oblanceolate or elliptic,

128 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIII

the lower ones subfoliaceous, 1-6 mm. long; pedicels short, 0-2 mm. long. Flowers white, fragrant, heterostylous. Calyx glabrous and ciliolate to densely and minutely puberulent; tube 1.5—2.5 mm. long with 4-6 stout, blunt, teeth 0.1-1 mm. long, which in the extreme are aan flattened laterally (Fic. 1, A & B, occasionally almost to C). Corolla hypocrateri- form; tube 8-21 mm. long; lobes 4—6, broadly ovate or ovate, rounded or acute, 4—-7.5 mm. long. Stamens 2; anthers (2.5—)3-4 mm. long on fila- ments 0.2-0.5 mm. long. Ovary 0.75-1.25 mm. long; style 3-6.5 mm. long in short-styled flowers or 12-22 mm. long in long-styled. Fruit (Balansa 1298) ovoid, paired (or single by abortion), 9-10 mm. long by 7-8 mm. broad.

New Caledonia. Environs de Nouméa, bosquets, Sept. 1868, Balansa 578 (syn- type, P; isosyntype, BM, NY); Port de France [=Nouméa], Vieillard 907 (syn- type var. brevistylis, P; isosyntypes, A, G); Kanala, Vieillard 908 (syntype var. brevistylis, ep); Mt. Canala, 650 m., Nov. 1911, Sarasin 363 (z); au Said de Canala, foréts, vers 1,000 m. alt., Nov. 1869, Balansa 1702 (syntype var. brevistylis, Pp); prés de Koe, bords de la Dumbéa, 14 Dec. 1869, Balansa 1298 syntype, P; isosyntype, K); bei La Foa an der Briicke iiber das Fliisschen La Foa, im sneer 27 Sept. 1924, Ddniker 131 (z); Gatope, Vieillard 2929 (BM, kK); northwest slope of Mt. Mou, forest, 15 Oct. 1950, Guillaumin & Baumann- Bodbuhem 6958 (A, z); Mé anita, on serpentine soil, 700 m. alt., 28 & 30 Nov. 1950, Guillaumin & Baumann- Bodenheim 8770, 9066 (A, 2); above the “Ermitage” valley (Mt. Koghi), hygrophilous forest on serpentine, 400 m. alt., 4 July 1951, Baumann-Bodenheim 14479 (a, z); without locality, Pancher 316 (=Vieillard 2930) (syntype var. brevistylis, P; isosyntype, BM),

massifs sur les sols argilo-schisteux, March, Pancher 313 es Vieillard 2941) (lectotype, P

Loyalty islands: Lirou: Képéneéhé, im Buschwald in Kustennahe, 22 Oct. 1925, Daniker 2270 (z); Thosip, 10 Nov. 1925, Daniker 2270a (z); without locality Deplanche 82 (syntype, P).

Jasminum elatum is distinguished amongst the simple-leaved New Cale- donian species by its sessile or almost sessile flowers. The apparent ‘“‘flat- ness” of the inflorescence branches and the size of the upper bracts are also distinctive. In length of calyx lobes it is similar to J. artense, but the lobes of the two species are nevertheless different: in J. artense they are flat, broadly triangular and acute, or at least with an angular apex, but in J. elatum they are blunt and thick, or even somewhat flattened laterally in those with the largest teeth. In addition, although the corolla tube in these two species is the same length, the lobes in J. edatum are shorter than those in J. artense. Large-leaved vegetative specimens of J. elatum may be confused with J. neocaledonicum, but in the former the petiolar articula- tion is usually found more or less towards the base and in the latter well up the petiole. Some purely vegetative specimens with large leaves can be difficult to identify and‘a good example is Guillaumin & Baumann- Bodenheim 8732 (Mé Amméri, 700 m. alt., on Serpentine soil, 28 Nov. 1950, z) which is perhaps best left as “cf. elat

With his original description Guillaumin ee five collections; it has

1962 | GREEN, STUDIES IN JASMINUM, II 129

been decided to select as lectotype the specimen of Pancher 313 at Paris: it is the only Pancher specimen amongst the syntypes, and on the label with the Paris specimen he has written ‘Jasminum elatum — de la hau- teur de la tige.”’

8. Jasminum promunturianum Daniker, Viert. Naturf. Ges. Zurich 78(Beibl. 19): 367. 1933; Guillaumin, Bull. Mus. Hist. Nat. Paris II. 14: 456. 1942, ibid. 15: 454. 1943, ibid. 20: 371. 1948, et Fl. Nouv.- Caléd. 284, 1948.

Evergreen shrub; stem puberulent. Leaves opposite, simple; petioles 1-2 mm. long, puberulent, articulated at the top; lamina thickish, oblanceolate or narrowly oblanceolate, (1—)1.5-2.5 cm. long by (0.2—)0.4—0.6(—0.7) cm. broad; margin entire, strongly recurved; apex retuse; base very nar- rowly cuneate, attenuate to the petiole articulation; venation more or less obscure, especially above, with 2 or 3 primary nerves per side. /nflores- cences terminal on side shoots, single flowered; peduncle 1—1.5 mm. long, glabrous. Flowers white, ? heterostylous. Calyx glabrous; tube 2 = mm. long with 5 or 6 stout, blunt, ciliolate teeth, 0.6-1.1 mm. long (Fic. 1, A). Corolla hypocrateriform: tube 20 mm. long: lobes 5, broadly lanceolate, acute, 7 mm. long. Stamens 2; anthers 4.25 mm. fore on filaments 0.25 mm. long. Ovary 1 mm. long; style 7 mm. long in a short-styled flower. Fruit unknown.

w Caledonia. Insel Art der Belepgruppe, an den Kiiste, niedrigen Gebtischen nee und exponierter Nordoststeilhang, 8 May 1925, Déniker 1681 (holotype, z).

Vegetatively this species, with its small oblanceolate leaves, is very dis- tinct. However, the calyx is extremely similar to that of Jasminum elatum, which itself is quite distinct from the other New Caledonian species in this respect. The field notes which accompany the one and only specimen state that it was collected on a rocky and exposed steep northeast slope on the coast, and it seems possible that J. promunturianum is really only a habitat form of J. elatum. Field observations and additional material may prove that the two ought to be united, but, since as yet I have seen no inter- mediate material, it seems best to maintain them for the present as separate species.

9, Jasminum kriegeri Guillaumin, Bull. Mus. Hist. Nat. Paris II. 15: 453. 1943, et Fl. Nouv.-Caléd. 284. 1948.

Evergreen; stem glabrous. Leaves opposite, simple; petioles 6-12 mm. long, glabrous, articulated 14-2 from the base; lamina thickish, narrowly lanceolate, 3. ae 5.5 cm. long by 0.6—1.2 cm. broad; margin entire, slightly thickened; apex obtuse or acute, subapiculate; base attenuate- obtuse, somewhat decurrent onto the petiole: venation clear and raised above and below, 2 (or 3) primary veins each side from near the base, the upper two

iE) JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIII

prominent, running the length of the lamina and anastomosing, joined by 3 or 4 (5) slightly less prominent, short veins per side from the midrib, occasionally 1 or 2 of these being prominent. Juflorescences terminal on side shoots, ternately corymbose-paniculate, 5—12 flowers per inflorescence, glabrous; bracts subulate, 1-2 mm. long; pedicels 8-12 mm. long. Calyx glabrous, tube 4.5—5 mm. long (at least in dried material with about 20 longitudinal raised lines), with 5 acute or obtuse teeth, shallowly triangu- lar, 0.2-0.5 mm. long (Fic. 1, FE). Corolia hy pocrateriform:; tube (10—)2 mm. long with 5 or 6 ia. acute, or obtuse lobes 11-13 mm. long. Stamens 2; anthers 4.5 mm, long, subsessile. Ovary 1.25 mm. long; style 18 mm. long in the long-styled flower examined. Fruit unknown.

New Caledonia. Gatope. leg. M. Krieger, Vieillard 2937 (lectotype, P).

Jasminum kriegeri, although based on a scrappy specimen collected in y an infantry lieutenant named Krieger, was not described until

1943, and, despite the many collections that have been made in New Caledonia during the last forty years, the original, rather poor gathering is still the only example of this species in the herbaria I have examined.

When describing Jasminum kriegeri, Guillaumin cited two specimens, Vieillard 2937 and 2938, but all the material of the latter that I have seen, including one from Paris, is undoubtedly J. leratii. The specimen of the former number at Paris, with the name written in Guillaumin’s hand and collected by Krieger, is the only one which the original description fits, and it is therefore selected as the lectotype.

The calyx is perhaps the most noticeably distinct feature of this species. In none of the other species is the calyx tube normally longer than 3 mm, at the most, although rarely in Jasminum noumeense it reaches 4 mm. but with large subfoliaceous lobes on top. Usually in the New eee jas- mines it is only 1.5—2.5 mm. long, whereas in J. kriegeri it is 4.5—5 mm. in length and contrasts with the very small comparatively obscure ey The calyces on the dried specimen which constitutes the type are furrowed throughout their length by shallow grooves which presumably correspond to the spaces between the veins. They are never so noticeable in the other species, and it is suspected that they appear only on drying.

The other part of this specimen which is noticeably distinct is the leaves. In shape they are very similar to a narrow-leaved expression of Jasminum leratii, but the venation is more or less unique amongst the New Caledonian species. In several species the main lateral and basal nerve runs the length of the lamina near the margin and joins up with the other lateral nerves given off by the midrib, but in this species two prominent nerves run the length of the lamina and the nerves joining them are fewer in number and more obscure. Nevertheless, this difference is only one of degree, and within the genus a whole range can be found from a regularly pinnately nerved leaf to the extreme as exhibited, for example. by J. crassi- folium Blume where the only two prominent lateral nerves run the length of the leaf, one on each side of the midrib.

1962]

GREEN, STUDIES IN JASMINUM, II

131

None of the intact flowers on the type specimen are quite mature, and it is feared that the measurements of the corolla given in the description above may be too short, for, judging from specimens of other species, it is believed that the corolla tube, for example, lengthens right up to the

time of anthesis. point.)

(Observations on living plants would perhaps settle this

INDEX TO EXSICCATAE

The list is arranged alphabetically by the last name of the collector. Numbers in parentheses refer to the corresponding species in the text.

Bailey 9623 (2 Balansa 578, 1298 (7); 1299 (3); 1702 (7); 2785, 3189 (1); 3629 (5) Baumann- Bode nheim 5032 5593A (5); 5874, 5881 (3); 6057 rae 8285 (2); 13436, 13454, 13608 (1); 13642 (2); 13699, 13786, 13789, 13816 (53); 14479 (7); 14529, 14721 2 14770 1

Gi) mae 131 ae 483, 588, 700 (5); 770, 1335 (2); 1436 (1); 1681 (8); 1717 (1);

1846 (4); 1886, 1909, 2269 (2); 2270 B270G Ce pees 2388a (2); 2465, 2465a (1); 2626, 3088, 3089 (2)

Deplanche 82

Eastwood, s.n. (2

Franc 845 (4); 1021 (2); 1377 (1);

> 1906 (2); 1935 (3); (1)

Guillaumin 7414 (under 5); 13395 (1)

Guillaumin & Baumann-Bodenheim 6842 Gre 6958 CI) 7879, 7933 (5); 8732 (under 7); 8770 (7); 8883 (3); 9066

(7) ; 9514, 9665 (2); 10301, 10440 (3); 11106, 11135, 11161, 11663, 12403 (2); 13127- i)

Heim 49 (2); 74 (1)

Hiirlimann 355, 381 (5); 769 (2); 1092, 1166 CLV S 1200, Oe 150711) = 1625 (3

Le Rat 709 (1); 165 (5); 172 (2); 574 588 (4)

McKee 475, 1973 (2); 2198 (3); 2251 (2) = 2208 (Ay 2342) 2428" (1); 2728 (4); 6512 (1)

Montrouzier 143 (5)

Pancher 373 (7); 315 (4); 316 (7);

(

4)

Sarasin 185 (2); 363 (7); oe Q)525 (37307 Gl) 85a (2) | 2890 eCL)

Schlechter 15586 (3)

Vieillard 907, 908 (7); 911 (2); 913 (1);

2929, 2930 (7); 2935 (1); 2935 (2); 2936 (1); 2936 (4); 2937 (9); 2938 Cyee2047 (7)

Virot 442 (2); 842 (4); 846 (1); 865 (Gee 870; 88). (2); 8821). 922, 92 O20 mn (Oneweire (6): 1302 “(5))

Walter 65 (2)

Woodcock 630 (2)

~

132 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLII

A CYTOLOGICAL STUDY OF THE GENUS VIBURNUM !

DoNALD R. EGoLr

THE GENUS VIBURNUM includes some 250 species from diverse habitats in Asia, Europe, North America, and South America. Though viburnums are varied in form, the ideal ornamental that combines fragrance, colored flowers, small stature, evergreen or brilliantly colored foliage, and luxuriant fruit does not exist. The present cytological study was initiated to estab- lish a basis for genetical research that might yield interspecific hybrids combining in one plant the ornamental characteristics of several species. Such a study of chromosomes is obligatory for the plant breeder to initiate and pursue intelligently an interspecific hybridization program. Many Viburnum species and varieties involve complexes that have not been adequately covered by any comprehensive taxonomic treatment. The data obtained from such related disciplines as cytology and genetics when combined with taxonomy may help to resolve the species complexes and clarify the classification of the genus

To the extent that this publication is a portion of more extensive cyto- genetical and cytotaxonomical studies in progress, only that portion of the research concerned with cytology is reported. Although the literature has been frequently consulted to verify the identification and relationships of the taxa studied, this paper is not intended as a taxonomic study. The author has followed the taxonomic nomenclature of Rehder’s Manual of Cultivated Trees and Shrubs (30) and Bibliography of Cultivated Trees and Shrubs (31).

REVIEW OF LITERATURE

All the chromosome counts reported by various authors are incorporated under the respective species in Tapie I. The earliest cytological study of Viburnum was that of Sax and Kribs (32), who reported that eleven species had a gametic chromosome number of nine (7 = 9). The Asiatic

*This study includes a portion of the research completed for the Ph.D. thesis in the Departments of Plant Breeding and Floriculture, Cornell University, Ithaca, N. Y., and later work in the United Kingdom and at the U. S. National Arboretum, Washing- ton, D.C.

Th e author is indebted to all who have co-operated in supplying seeds, cuttings, or

for a Fulbright Scholarship in the United Kingdom, 1956-58. The many courtesies extended in the United Kingdom by the Royal Botanic Garden, Edinburgh; the British Museum (Natural History), London; the Royal Botanic Gardens, Kew; the Royal Horticultural Society’s Gardens, Wisley; the John Innes Institute, Bayford- bury; and the University of London are gratefully acknowledged.

1962] EGOLF, CYTOLOGY OF VIBURNUM 133

and American species studied included six of the nine taxonomic sections of Rehder (30). Sax and Kribs stated that the chromosomes are large, have an affinity for chromosomal stains, and consequently provide favor- able material for study. A second basic chromosome number of eight (n = 8) was reported for V. fragrans by Simonet and Miedzyrzecki (35), who also published counts for seven additional species as n = 9. gametic and somatic number of V. tinus was determined by Feng (15) to be n = 18 and 2m = 36. Sugiura (38) reported 2” = 20 for V. awabuki (syn. V. odoratissimum). In 1946, Poucques (27) listed the gametic chromosome counts for five species, four of which were previously un- published; in a later publication (28) he listed two additional species.

Janaki Ammal determined the chromosome number 2m = 16 for Vi- burnum fragrans and V. grandiflorum, and for V. x bodnantense, a hybrid produced from a cross between these species. The chromosomes of the two species paired normally in the hybrid, and the pollen fertility was as high as one hundred per cent (36). The extensive cytological study of Janaki Ammal (18) included thirty-seven determinations, of which twenty-one were reported for the first time. Her survey reported somatic chromosome numbers for species in cultivation at the Royal Horti- cultural Society’s Garden, Wisley, England; the Royal Botanic Gardens, Kew, England; and the Jardin des Plantes, Paris, France. Seventeen of these counts are at variance with the somatic chromosome numbers of the present study; while three are at variance with previous reports. These differences are considered in the discussion. She has interpreted the 2n = 18 of the hybrid V. & juddii (V. carlesii, 2n = 20, X V bitchiuense, 2n = 16) as a synthesis of a plant with » = 9, and in addi- tion, proposed that V. carlest (2n = 20) arose as a backcross between a chance triploid (2m = 24) and the normal diploid (2m = 16) of V. bitchinuense. This was considered to be an example of the possible manner of evolution of Viburnum species in nature, which finally resulted in a large number of species with the basic number = 9. A chromosome count of V. lobophyllum, 2n = 20, determined by Enoch for a plant grown at Exbury, Southampton, England, was included in the publication of Janaki Ammal (18).

The cytological study by Thomas (40) included twenty-nine of the plants cultivated in the Arnold Arboretum. A few of these counts were obtained from root tips, but most were made from pollen mother cells; however, only partial designation is given as to which counts were gametic. Thomas concurs with the author that the origin of species with m = 9 as postulated by Janaki Ammal (18) is questionable. He states that it is more likely that species with basic chromosome numbers of eight and ten originated from species with a base number of nine by the loss or gain of a chromosome. His study indicates that translocations occur rather frequently, as evidenced by bridge formations observed in several of the species. He noted a relatively high percentage of aberrations in Viburnum cassinoides, V. carlesii, V. dentatum, V. plicatum, Ve. xX rhy- tidophylloides, V. sieboldii, V. trilobum {. compactum, and V. veitchit.

134 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIII

MATERIALS AND METHODS

The seventy-seven species, sixty-one varieties and forms, thirteen hy- brids, and two unidentified accessions of Viburnum investigated are a sented with their sources, in Taste TI. Plant material was obtained < seed, cuttings, or plants from native habitats, botanic gardens, cho and estate gardens throughout the world, exclusive of Central ay South America. In all cases an effort was made to secure representatives o each taxon from three or more sources to provide a check on identification and chromosome counts of each taxon. Since commercial nurseries often propagate horticultural forms of Viburnum by grafting, all plant material or cytological study has been propagated from cuttings or from seed collected when possible in the native habitat of the species. Seed from native habitats have been used chiefly because seed from botanical col-

in the collection. The plants for cytological study were maintained during the summer months in frames or in a lath house. The remainder of the year the plants were grown under long-day conditions of twenty to twenty-two hours of light in a 70° F. greenhouse. Under these conditions it was possible to keep the plants actively vegetative and to avoid any dormant period.

In so far as feasible, plants propagated from seed and cuttings, and representatives of all sources, will be maintained at the Cornell Planta- tions and the U. S. National Arboretum for further study. Herbarium specimens, which are identified with the code accession numbers, were prepared for each collection that provided sufficient material. These are maintained as part of the permanent record and will be supplemented with flowering and fruiting material when the plants mature. Original descriptions and many of the type specimens have been studied to verify identifications. Photographs of type specimens and photostats of pertinent taxonomic literature were prepared to provide a basis for cytotaxonomic research. The identification of many of the plants previously studied (14 has been checked since they flowered and fruited. In those cases in which a positive determination was made, the alteration has been entered on Tas_LE I. However, the documentation numbers have not been altered and will be the same as in the previous list (14).

The root-tip smear technique was used exclusively in this eae 7. liminary use of McClintock’s permanent aceto-carmine (19) a La Cour’s (20) acetic-orcein stain techniques revealed that the ee gave best results. In a portion of the early work the root tips were pretreated in aqueous paradichlorobenzene (22), fixed in Baldwin’s modified Carnoy’s (2), hydrolized in a solution of equal parts of 95% alcohol and con- centrated hydrochloric acid, and smeared in acetic-orcein on the slide. After trial of numerous schedules and variations of these procedures a modification of La Cour’s (20) technique was employed. Three- to five- millimeter-long root tips were pretreated in aqueous paradichlorobenzene for one to two hours. The root tips were placed in a watch glass contain-

1962] EGOLF, CYTOLOGY OF VIBURNUM 135

ing one part 1.0 N. hydrochloric acid to nine parts 0.5% acetic-orcein stain. The watch glass with root tips was passed two or three times over the flame of an alcohol lamp to heat the mixture, but great care was taken to keep the solution under the boiling point. After the heated watch glass had cooled for several minutes, a root tip was placed in a drop of 0.5% acetic-orcein on a slide, smeared, and the cover slip applied.

The Feulgen technique (10) has been used in recent studies with ex- cellent results. Root tips were collected and pretreated with 0.1% col- chicine for two hours and fixed in La Cour’s 2BD general fixative for twelve hours. The root tips were thoroughly washed with distilled water before immersion in a peroxide-oxalate bleach consisting of equal parts of ammonium oxalate in distilled water and hydrogen peroxide, and placed in direct sunlight or under a spot light for five minutes. After the root tips were washed again, they were hydrolized in 1.0 N. hydrochloric acid

t 60° C. for twelve minutes, stained in leuco-basic fuchsin for 2 hours, and smeared. The edges of the cover slip were sealed with a mixture of gum mastic and paraffin in equal parts.

Slides were observed immediately or stored in a 40° F. refrigerator. After critical examination of the temporary smears was completed, camera lucida drawings made, and photomicrographs taken, selected slides were made permanent. The method of Conger and Fairchild (6) accomplishes the separation of the cover slip from the slide by freezing on dry ice. More recently, compressed carbon dioxide has been utilized for freezing slides to separate the cover slip from the slide. Immediately before thaw- ing, the separated frozen slide and cover slip are placed in 95% alcohol which contains 5—-10% acetic acid. After two or three minutes they are placed in absolute alcohol for another few minutes before mounting in diaphane. The permanent preparations made by this method are almost always equal in excellence and clarity to temporary slides and are superior for photomicrographs. Permanent slides of virtually all accessions here reported have been prepared.

Critical examination of the preparations was made with a binocular pean equipped with 98 & fluorite objectives, N.A. 1.30, and 12.5 oculars. A minimum of ten countable cells was located before ascertaining the ee of chromosomes in the somatic complement. Drawings were made with a camera lucida at table level, using 15 % oculars, giving the drawing a magnification of approximately 2400 x. In addition to the drawings, photomicrographs were taken on 35 mm. microfilm at a mag- nification of approximately 1500 x

RESULTS

The somatic chromosome counts of the 153 taxa of Viburnum included in this investigation are presented in TasiEe I. The species are arranged alphabetically within the taxonomic sections, as designated by Rehder (31). Under each taxon the chromosome counts published by other authors precede those from this study. The general geographic distribution and

136 JOURNAL OF THE ARNOLD ARBORETUM [VOL, XLIII

ey «s

“~

Fics. 1-6. Photomicrographs of chromosomes of Viburnum to show varia- tions in chromosome complements, approximately & 1100. 1, Viburnum erubes- cens, 2n = 32; 2, V. lobophyllum, 2n = 18; 3, V. sieboldii, 2n = 32; 4, V.

carlesii, 2n = 18; 5, V. bracteatum, 2n = 72; 6, V. scabrellum, 2n = 72.

138

JOURNAL OF THE ARNOLD ARBORETUM

[ VOL. XLIII

the source of material are given for each accession. A series of photomicro- graphs (Ficures 1-15) illustrates variations in the chromosome comple- ment.

In order to make the table more concise, the sources of material have been abbreviated as follows:

AA

Arnold ma Jamaica Plain 30, Mas

Arboretum des rance

Arturo Anealogs Bologna, Italy

Barres, Loiret,

Armstrong Nursery Ontanio. Calif.

Botanisches Museum, Berlin- Dahlem, Germany.

as Botanic Garden, Brook- Y.

lyn

Birr Cae Birr, County Kings, Trela

Collected near Bedford, Pa.

Bodnant Garden, Tal y-Cain, Denbighshire, N. Wal

Borde ill, Haywards Hea, Sussex, En

U_S. Botanic cen Washing- on, D.

Collected ak Brighton, Sussex, Englanc

Boyce eae Arboretum, Yonkers Y.

Botanic Gz aven of the University of Copenhaget Copenhagen, Denmark.

eee Castle, St. Austell, Cornwall, England.

Col lected near Chattanooga, Tenn,

R. Chenault, Orleans, France.

W. B. Clarke & Co., San Jose 3, Calif.

Crathes Castle, Crathes, Kincard- shire, Scotland.

Carnell oo Ithaca, N. Y.

Dominion Arbor Ottawa, O

ntario, ee Collected near South Downs, Engla “ T. Dawson, es tucket, Mass. National ene c Gardens, Glas- nevin, Dublin, Treland Royal oe Ps rden, Edin-

iques, Geneva, ae land.

GA

Gardens of the Blue Ridge, Ash- ord, N. C.

Goteborgs Sear Tradgard, Goteborg, Swed

The Great Park, Windsor, Berk- shire, England.

N. G. Hadden, West Porlock,

Somerset, Ruckint Hattori Botan ical Laboratory, Nichinan, Miyasaki, Japan.

Headford Court, Meath, Ireland.

Henry Foundation for Botanical Research, Gladwyne

Hermann A. Hesse, Weener, Ger-

Kells, County

Hong Kong Dent. of Agriculture,

Highland i. Rochester, N. Y. Collingwood Ingram, Cranbrook, Kent, Englan George Jackman & Son, Woking, d.

Innes Institute, Bayford- . Hertfordshire, England. Jackson and Perkins Co., New-

5 Ns ¥

Botanie Gardens, Kew, \ and. Botanic Gardens of Indonesia oe Raya), Bogor, Indo-

nesi

Teen Kohankie and Son, Paines- ville, Ohio.

iGnesvi lle Nurseries, Md.

Kornik Gardens, Kornik, Poland.

Collected — Lake Cayuga, Ithaca, N.

V. Lemoine & Fils, aoa France.

Kingsville,

Linn agen Iunsevies Center Point, Los ‘State & County

a, Calif. Lu Shan ii Garden, Kiu- kiang, Chin

EGOLF, CYTOLOGY OF VIBURNUM 139

Grier peaeueae Wimborne, Dor- set, Englan

W. ies ae Chapel Hill, N.C.

McLean Bog, McLea

i ie Philadelphia

Morton eee Lisle, III.

Myddelton House, Enfield, Mid- dlesex, England.

Nanking Botanic Garden, Nan- king, Kiangsu, China.

R. C. Notcutt, Ltd., Woodbridge, Suffolk, England

Nymans Gardens, Haywards Heath, Sussex, England.

New York Botanical Garden,

enV. Onarga Nursery Co, Onarga, Ill. Muséum d’Histoire

nce. University,

Pa. .S. Plant Tatreduction Garden, ee Dale Princeton Nurseries, Princeton, N

Unknown Source, Warsaw, Po- nd

and. Regel & Kesselring, Rome, Italy.

Ringwood Forest, Ithaca, New

York.

Royal Horticultural Society's Eng-

ere Wisley, Surrey, and,

‘ Ruys, Holland

Ltd., Dedemsvaart,

Siebenthaler Co., Dayton 5, Ohio.

ollected South Hill, Ithaca, N. Y. L. Spath, Berlin, Germany. Bergius Botanic Garden, Stock-

holm, Sweden.

Swyncombe House, Oxford, Eng- land.

Taiwan Forest eee rae Taipei, Taiwan,

Taiwan Pineapple ae Taiwan, China

Tokyo Uni eae eae Gar- dens, Tckvo

Univ. of California SBemhical

Garden, Berkeley, Calif.

Taipei,

Univ. of Washington Arboretum, Seattle, Wash.

Vaughan Nursery, Chicago 6, Illinois.

Arnold Vogt, Erlenbach-Zurich, Switzerland.

ae Place, Ardingly, Sus- Sex and.

’

Wayside Teen Mentor, Ohio.

Wildlife Research Laboratory, ee Sar c L. Williams, Exeter, N. H.

Willows ood Far Gia ctone:

womans Garden Centers, Inc., Framingham, Mass.

Wyoming Nurseries, Cincinnati ike

140 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIII

TABLE I. Chromosome Numbers of Viburnum

SOURCE GENERAL AuTHOoR- DocuMEN- OF DISTRIBU- SPECIES nm ITy* TATION MATERIAL ° Sect. I. Thyrsosma (Raf.) Rehd. V. X bodnantense Aberc. (V. fragrans < grandiflorum) 16 9 76E CL cult = 9 874E UW539-S0(HL) ¢ ds - 16 9 C02E RH as 7 16 9 1007E K660—48 (BG) Re ne 16 9 1102E HL . ‘Dawn’ 16 6* — —- cult a 16 ge 1357E BG “ ‘Deben’ 16 g* 1937E NN cult. V. erubescens Wall. 48 6 — — Himal. 32 9 972E AA7602-A ‘ * 2 9 1149E K '

7 32 9* 170K EX is var. gracilipes Rehd. 32 9 100K HP2158 Himal. e 32 9 502E RU e " 32 9 1092K HL et ““ 32 g* 1594E DU183P er

V. foetens Decne. 16 6 — = Himal as 16 9 441E RH e 16 9 1061E K as M 16 9 1091E HL a . 16 Q* 1571E HA n V. fragrans Bge. 16 2 — — n. China i. 16 6 HI “6 16 9 25E CU 7s 16 9 27E PI-82380 a as 16 9 174E AA11588(K) iE “ 16 9 597E MT1007-39 s 16 9 1074E RH a “ 16 9* 1154E HA «“ var. album Kriiss, 32 6 — — n. China fas var. candidissimum) 16 8 AAS55-50-B AAS55-50-B n. China 16 9 250E 35-38 uf 7 16 9 598E MT235-38(VN) « 7 16 9 896E RH fs " 16 g* 1419E MY .

*1, Sax and Kribs (1930); 2, aa and eae (1932); 3, Feng (1934) ; Sipiura (1936); 5, Poucques (1946) ; , tbid. (1949); 6, Janaki Renae (1953); 6*, ibid. (1950) ; ; ea (1953); 8, Thomas nae 9, Egolf (1956) ; 9*, ibid., reported here for the past

he abe listed in this column are the accession numbers and voucher herbarium specimen numbers for the plants studied. Specimens of others are given as “AA” (Arnold oo and “K” (Kew). . mber or designation ‘after the code letters identifies specific plants. The abbre- aaa ye Bt ” signifies that the accession was obtained as seed. e of material reference in parentheses refers ie the original source from which the pare: or plants were procured.

4,

1962] EGOLF, CYTOLOGY OF VIBURNUM 141

TABLE I (Continued)

SOURCE GENERAL AUTHOR- DocUMEN- OF DISTRIBU- SPECIES TATION MATERIAL TION var. nanum Boom 16 9 431E UW198-41 cult. sf 16 9 504E “ te 16 9 600E MT685-50(CL) = ‘Roseum’ 16 9 438E RH cult. a 16 9* 1422E JG cs V. grandiflorum Wall. 16 6x — — Himal. ae 16 9 1022E K S as 16 9 1082E RH “ 16 9 1093E HL oi « 16 gx 1426E GP ‘i “ 16 ox 1427E G 2 “ 16 gx 1428E E3025 (Cooper) “ z 16 9* 1576E IN “ V. henryi Hemsl. 48 6 — — c. China ef 32 9 984E Cc ry ce 32 9 1039E K 5 “ 32 9 1069E DU a te 32 9 1097E HL * ce 32 g* 1168E E i st 32 9* 1175E ss ae 32 g* 1435E BH1398 sf es 32 g* 1441E DU10-39 (Henry) ss V. X hillieri Stern (V. henryi & erubescens) ‘Winton’ 32 gk 1442E HL cult. V. odoratissimum Ker-Gawl. (as j (0) 4 — — Malaysia : 40 6 —_ — i. ce 32 9 68E CU (Wash.) " fs 32 9 119E FN ‘ “ 40 9 293K UW g MY 32 9 371E LO < “ 40 9 392E NY (BO) ‘: a 32 9 427E ‘ sf 32 9 919E HL : te 32 9 983E © g ‘ 32 9 1011E K i V. photinioides Fashiro 32 9 691E KB Malaysia V. sieboldii Miq. 16 6 2454K. RH Japan e 32 8 AA616-6-B AA616-6-B ms ss 32 9 62E E ne : 32 9 105E HP2161 _ 16 9 270E SN > i 32 9 656E MT159-38(KH) " . 32 9 718E R2143 % ss 32 9 839E AA616-6-B “ - 32 9 903E H ne - 32 9 1120E P “ f. reticulatum Rehd. 32 9 657E MT281-51(AB) Japan t 32 9 878E ‘ ~ 32 9* 1056E K986-36(LE) as

142 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIII TaBLe I (Continued) SOURCE GENERAL AuTHOoR- DoCUMEN- OF DIstTRIBU- SPECIES 2n TATI MATERIAL TION V. suspensum Lindl. (as V. san- dankwa) 18 5 — -- Malaysia 16 6 — —- _ " 16 9 289E UC 16 9 389K NY a : 16 9 2 KB uf Sect. II. Lantana Spach. V. bitchinense Makino 16 6 -- w. Japan - 18 9 4k PI-82381 e . 18 9 7OE sd. PI-82381 i sf 18 9 307E KR,s “ 18 9 522E VZ, sd. - 18 9 788E AA (Wilson) _ 18 9 980E MR3 » 7 18 9 998E < = . 18 9 1078E RH ef 18 9 1143E CU ia 18 9* — 1354E . . 18 g* 1355E E430-29 a V. buddleifolium Wright 18 2 = c. China “ 20 6 a “ aa 18 9 7E PI-111380 - 18 9 8E CU 18 9 422E BB 18 9 423E KN e . 18 9 710E MR “ a 18 9 808k AA7533 (Veitch) 7 a 18 9 1049E K « . 18 g* 1328E jl “ 18 g* 1360E EX burejaeticum Rel. & Herd 18 g 278E GB, sd n. China 18 9 428E O ae " 18 9 434K UW 160-46 . a 18 9 585E MT546-32(AA,sd) “ 7 18 9 586K MT475-40(ST, sd.) : s 18 9 628E MT1144-40 ns 18 9 675E KH 7 18 9 772E AA4942(RE) V. & burkwoodii ony & Skip. (Vv. cuntese S< utile) 18 8 AA815-41-B AA815-41-B cult 18 9 1E te ey 18 9 518E sd. “ * 18 9 587E Me 295— 36(SN) - 18 9 1006K K1923HK 18 9* — 1363E i) - ‘Park Farm Hybric 18 9 923K HL cull Vv. &X ee Han Burk. ex. Pike (V. carlesti >< macro- ce phalunr) 8 AA618-53-A AA-618-53-A cult

1962 | EGOLF, CYTOLOGY OF VIBURNUM 143

TABLE I (Continued)

SOURCE GENERAL AvuTHOR- DocUMEN- OF DISTRIBU- SPECIES ITY TATIO} MATERIAL TION a 18 9 77E ONE, s 18 9 433E UW st + 18 9 689E WG es i‘ 18 9 1023E K V. carlesii Hemsl. 18 D5 ae — Korea - 20 6 = a e 18 8 AA17981-A AA17981-A e 18 9 112E HP2193 es iv 18 9 421E BB : e 18 9 521E WZ, sd. ss “ 18 9 536E JP, sd. i. i 18 9 676E KH e a 18 9 785E AA17981 (Gibbs) s om 18 9 1009E K972-34(Trickett) “ os 18 9 1144E CU " 2 20, 22 9 1145E HN, sd.( Korea) n u 20, 22 9 1146E TA, sd. : if 20, 22 9 1147E HM, sd. (s.China) “ 18 g* 1369E DU s ss 18 gx 1596E E529-38 " V. < carlotta Hort. (V. x burk- woodii * carlesii) 18 9 78E CL cult. V.& chenaultii Chenault 18 9 406E RU cult oS 18 9 409E MT741-50 s 18 9 519E WZ, sd. 2 “ 18 9 592E MT218-51(SN) ef oh 18 9 1057E K227-48( AB) - V. cotinifolium D. Don 18 5 — -- Himal. : 18 9 819E AA1236—-52(K) ef ae 18 9 871E UW17493 L&S (Bhutan, sd.) 4 S 18 g 1037E K56-81 Himal. x 18 9 1040E K1067-83 es & 18 9 1067E DU ee a 18 9 1086E L ‘ V. glomeratum Maxim. 18 9 817E AA, W180-5(K, sd.) China a 18 9 992E MT406-44 i te 18 9 1052E K sf V. < juddii Rehd. (V. carlesi << bitchiuense) 18 — — cult. ¢ 18 9 81E CL * oe 18 o* 9O7E MR “ 7 18 9 606E MT451-48(AA) % be 18 9 756E AA1107-27 (Gibbs) ue 18 g* 823E AA813-49 es 18 g* 824E AA284-44 . a 18 9 Q73E P : “e 18 g* 1312E RH V.lantana L. 18 1 — AA Eur., w.As.

144 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLII TABLE I (Continued) SOURCE GENERAL AuUTHOR- DoCUMEN- OF DISTRIBU- SPECIES I TAT MaArtERIAL TION 7 18 5*,6 — — ae . 18 9 31E U ‘ - 18 9 239E MT1010-39 6 - 18 9 301E KR, sd. ff | . 18 se) 385E Y as - 18 9 607E MT437-37 (PI-107644) a ” 18 9 608E MT640-34 nt " 18 9 722E BT78-38 - ' 18 9 996E K 40-33 = 18 9 1041E K ae . 18 9 1142E P Ki : 18 g* 1325E DO « a 18 9* 1456E R es . 18 g* 1467E E(B193-36) s ‘Aurea Marginata’ 18 9 609E MT838-37(B) cult. ‘Floribundum’ 18 9 1138E Pp cult. ‘Lanceolatum’ 18 9 612E MT616-39 cult 7 18 9 940E P . ‘Lees’ 18 9 825E AA(K) cult. ‘Macrophyllum’ 18 9 72E MT cult var. rugosum Lange 18 8 AA907-27-A_ AA907-27-A cult. > 18 9 9E HP es is 18 9 456E MT352-46 a “ 18 9 615E MT213-41(KH) i 7 27 9 679E KH a ° 18 9* 775E AA907-27(DA, sd.) “ f. variegatum (West.) Rehd. 18 9 618E MT1053-41 cult. = 18 g* 1470E E " V. macrocephalum Fort. f. sterile Dipp. 18 9 82E CL cult. “t 18 9 424E KN ut 18 9 713E MR47-121 et ms 18 9 975E P . at 18 9 1015E Kk ut sf 18 g* 1597E E tf 7 micro phyllum (Oerst. ) Hemsl. 18 9 624E MT642-36 Mexico .m ongolcun: (Pall.) Rehd. 16 6 — ~- e. As. 18 9 277E GB, sd. - ee 18 9 673E MTSOI1-S3 ° V. & rhytidocarpum Lemn. (V. buddleifolium rhytidophyl- lum) 18 9 418E R cult 18 9 643E 03-53 “ « 18 9 806E AA412-36(LE) ee V. rhytidophylloides Suring. (V. rhytidophyllum *K lantana) 18 8 AA711-36-A AA711-36-A cult. af 18 9 102E HP “ af 18 9 426E KN(WW) “

1962] EGOLF, CYTOLOGY OF VIBURNUM 145 TABLE I (Continued) SOURCE GENERAL AutTHOR- DocuUMEN- OF DISTRIBU- SPECIES TATION MATERIAL TION oe 18 9 449E Y e 3 18 9 671E MT265-37 s - 18 9 754E AA711-36(LE) - a 18 9 826E AA1481-52(WW) te ; 18 9 1038E K387-29 Bs V. rhytidophyllum Hemsl 18 2 —_ — c. & w. China oe 18 6 2451K RH - S 18 9 51E CU oT a 18 9 52E PI-58813 oy ee 18 9 91E MR * “ 18 9 304E KR, sd. : = 18 9 1048E K 201-07 (Veitch) se Hi 18 9* 1203E E401-37 ie e 18 g* 1205E E493-36 i eS 18 g* 1259E RH “ e 18 9* 1278E HI( Wilson) SS s 18 9* 1505E E493-36(CR) es f. roseum (Gard. Chron.) Rehd. 18 9 295E UW(CL) cult te 18 9 800E AA310-41(HL) ‘ f. Fe doen Boom 18 9 822E AA2 cult. V. EAA EED Maxi 18 9 128E AA15570 nw. China 18 9 3907E NY(HP) We 18 9 465E MT1051-401 “ ss 18 9 652E coca 35(AA) iy H 18 9 726E ts 18 9 1014E KUAASG? 30) “ wu 18 9 1090E o V. stellulatum Hemsl. 18 9 506E a Himal Zs 18 9 1017E K131-38(DU) - V. utile Hemsl. 18 6 — — c. China s 18 9 214E DA, sd. s sf 18 9 425E KN : ss 18 9 507E DU oe i 18 9 694E PI-111380 i a 18 9 1008E K Wy 18 Q* 1528E BH oS i 18 Q* 1529E WA ef 18 g* 1530E E a V. veitchii Wright 18 8 AA7198 AA7198 c. China : 18 9 113E HP2177d sf a 18 9 261E MT1101-36 sf 18 9 685E KH - i 18 9 727E NY67480 “ s 18 9 753E AA7198( Veitch) " 18 9 1005E K101-13 (Veitch) Me

18 9* 1560E DU1288W

146 JOURNAL OF THE ARNOLD ARBORETUM

TaBLe I (Continued)

[ VOL. XLII

SOURCE

GENERAL AvutTHOoR- DocUMEN- OF DISTRIBU- SPECIES 2n TATION MATERIAL Sect. IIT. Pseudotinus Clarke V. furcatum Bl. ex Maxim. 18 6 — Japan “a 18 9 740E AA17988 (Wilson) ‘ 7 8 9 1099E HL a lantanoides Michx. (as V. alnifolium) 18 1 —— AA e.N. Am. 18 6 — ‘ . 18 9 497E RF os ™ 18 g* 1876E GA ae V. svi podiale Graebn. 18 9 451E NY(LS) China V. urceolatum S. & Z. 18 8 \A876-51 A A876-51 Japan i 18 9* 1645E PI-227284 « Sect. IV. Pseudopulus (Dipp.) Rehd. V. plicatum Thunb 16 8 AAI8016-1 AA18016-1 e. AS. 18 ) 92E MR681 m : 18 9 103E HP 2165 = 18 9 157E AA18016-1 ‘ 16 ) S577E PN 16 9 S78E PN ‘ 16 9 779E AA933- aa _ 16 9 10021 K629 f. eabren (Nakai) Rehd. 16 9 532E TI, a shad Japan ea, urt 16 9 436E RH cult, 16 9 843E AA134-53 a f. lanceolatum Rehd. 16 8 AA6122-1 AA6122-1 cult. = 16 9 658E MT1204-41 ~ 16 763 FE AA6122-1(Sargent) {. — (Veitch) Ret a is ymentosum var. oe 18 6 6K RH cult. 7 18 8 AA870-51-A AA870-51—-A ‘s 16 9 SOOKE RS 16 9 762E AA19355(K) a . 16 9 1010E K : 16 9 1079E RH - ‘Roseum’ 16 9 580E PN cult f. roseum (Doney) Rehd. 16 8 AA856-34 AA856-34 cult : 16 9 737E rf ee 16 9 742E AA856—34( BB) ‘Rowallane’ 16 9 437E RH cult. ‘St. Keverne’ 9 S505E RU cult f, tomentosum Thunb.) Rehd. (as V. tomentosum) 18 1 = NA e. As. - 18 6 2452K RH eS ed 18 9 43K CU ne a 16 9 95E MR1218 . “ 16 9 110E HP2195 .

1962 | EGOLF, CYTOLOGY OF VIBURNUM 147 TABLE I (Continued) SOURCE GENERAL AuTHOR- DocuMEN- OF DISTRIBU- SPECIES TATION MATERIAL TION i 18 9 484E MT176-37, 1204-41 “ i. 16 9 688E KH : 16 9 846E CU ‘. i 16 9 1001E K257-33(NN) : ss 18 9 1084E RH . Sect. V. Lentago DC.

V. cassinoides L. 18 8 AA17997 AA17997 e. N. Am. ie 18 9 12E C ay a 18 9 106E HP2790 mf a 18 9 S75E ML, sd. 8 18 9 590E MT, 570-43 a i 18 9 677E KH ee - 18 9 1000E K "i var. nanum Kriiss. 18 9 1088E HL cult

; jackii Rehd. (V. len-

tago xX prunifolium) 18 9 257E MT1037-40(AA) cult.

‘ 18 9 965E AA17992-1-B - V.lentago L. 18 1 oe e.N. Am (v3 18 6 __ ae “e a 18 9 166E AA18021-A ey = 18 9 535E CU . sf 18 9 574E 50 -

f. sphaerocarpum (Fern.) Rehd. 18 9 619E MT892-35 e.N. Am, is 18 9 068E AA11316 nt V. nudum L. 18 6 — — e.N. Am. ye: 18 9 570E MC ies = 18 9 681E KH “ ms 18 9 988E MT as ie 18 9 1059E K961-31(Gibbs) ss 18 9 1151E G V. obovatum Walt. 18 9* 982E C se. U.S es 18 9* 1802E HF(Fla.) fs V. prunifolium L. 18 1 — AA e.N. Am. ss 18 6 — — a a 18 9 46E CU re 18 9 47E E ss if 18 9 156E AA1805-2 of «“ 18 9 223E DA, s “ A 18 9 960E CU - wf 18 9 1020E K . V. rufidulum Raf, 18 9 517E WZ se.U.S ss 18 9 569E C af sf 18 9 645E MT1012-39 se a 18 9 1150E E <

148 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIII TaBLe I (Continued) SOURCE GENERAL AuTHOR- DOCUMEN- OF DISTRIBU- SPECIES 2n TATION MATERIAL TION Sect. VI. Tinus Maxim. V. atrocyaneum Clarke 18 9 918E HL Himal. Re 18 9 1043E K566—-48(HL) of " 18 g* 1173E nt wt 18 Q* 1267E DU9198K.W. ms V. calvum Rehd. 18 9 922E HL w. China . 18 9 1054E K(E23-48) s a 18 g* 1285E E16-52(HL) . V. cinnamomifolium Rehd. 18 6 — — w. China a 18 9 85E CL af 18 9 884E UW163-41(AR) as es 18 ce) 895E RH . “ 18 gx 1270E DU963W “ “ 18 Q* 1276E HC _ " 18 Q* 1372E BH a 18 Q* 1374E WA i V. davidii Franch. 18 6 — _ w. China " 18 9 86E ” . 18 9 181E E2485, sd es “ 18 9 415E RU a 18 9 877E vo a 18 9 885E UW . ss 18 9 900E RH 18 9 1029E © 18 9* 1390E Ec63W a Af 18 9* 1391E A te ‘Femina’ 18 9 503E cult. te 18 9* 1315E DU(MA) « V. harryanum Rehd. 18 5 —_ — w. China me 18 9 Q17E HL v “ 18 g* 1165E EX fe 18 g* 1194E E56—42 i i 18 g* 1275E BC x ss 18 g* 1314E DU es V. propinquum Hemsl. 18 5 — = w. China Zs 18 9 559E K “ 18 9 875E VO " - 18 9 886E UW79-50 a ‘Lanceolatum’ 18 9 924E HL cult. V. rigidum Vent. 18 9 556E K Canary Is. “ 18 9 858E c « “ 18 o* 1174E HI & V. tinus L. 36 2; 3; 5,6 — —_— Eur _ 36 9 118E FN “ 36 9 267E AN, sd . 36 9 291E UW 7 i 36 9 314E C ne 36 9 1035E K ot

1962 | EGOLF, CYTOLOGY OF VIBURNUM 149 TABLE I (Continued) SOURCE GENERAL AuTHOR- DocuMEN- OF DIstTRIBU- SPECIES n ITY TATION MATERIAL TION

ie 36 9 1064E K a

a 36 9* 1180E E -

i 36 gx 1260E RH sf ‘French White’ 36 9 915E L cult. var. hirtulum Ait. 36 9 1028E K22-11-47 Eur

ce 36 9 1100E HL n var. lucidum Ait. 72 9 288E UC Eur

2 72 9 292E UW ce

i 72 9 343E DU, sd i

ct ees 9 997E K pt

e 72 Q* 1198E E

‘Variegatum’ 72 9 999E K cult ‘Purpureum’ 36 9 1046E K1936 cult. Var. variegatum 36 9 1096E HL cult

36 9 1181E E s

Sect. VII. Megalotinus Maxim. V. coriaceum Bl. 18 9 879E vo As.

3 18 9 1026E KS0-51(E) tf

i 18 9 1101E HL 6

ok 18 Q* 1167E EX te

" 18 9* 1274E C i

V. sempervirens Koch. 18 9 276E HN, sd. Malaysia Sect. VIII. Odontotinus Rehd. V. acerifolium L. 18 1 — AA e.N. Am. ss 18 6 — = %

rf 18 9 1E CU s

a 18 9 2E SO s

as 18 9 357E DU 5

be 18 9 581E MT202-38(WM) ss

Hy 18 9 799E AA19181 “

y 18 9 1066E K258-53 (HP) fe

cs 18 9 1144E LA oe

V. betulifolium Batal. 18 6 — — c..& w. China is 18 9 297E UW(CL) eS

7 18 9 447E NY xt

tt 18 9 674E KH sé

* 18 9 757E AA550-26

(Rock 13476) ot

nS 18 9 1003 E K61-08 at

i 18 Q* 1187E E346F a

c 18 g* 1247E RH uf

rs 18 g* 1338E NS ms

18 9* 1340E WA 7

oe 18 Q* 1342E HL “

oH 18 9* 1347E DU238A-W fs ‘Aurantiacum’ 18 9* 283E cult.

150 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIII

TABLE I (Continued)

SOURCE GENERAL AuTHOR- DOCUMEN- OF DISTRIBU- SPECIES 2n ITY TATION MATERIAL TION V. bracteatum Rehd. 72 9 455E MTS60-37 se. U.S. 72 9 552E K is V.dasyanthum Rehd. 18 9 294E UW c. China ms 18 9 736E s os 18 9 1055E K910-39( LE) o 7 18 9 1386E E£910-39( LE) " V. dentatum L. 54 6 — _ e.N. Am. a 36 8 AA17985 AAI7985 My : 36 9 16E CU e af 36 9 107E HP2212 = a 36 9 142E AA17985-B a ei 36 9 391E NY (HP) “ 7 36 9 572E SO tf - 36 9 593K MT63S-34(0ON) “ ¢ 36 9 890E kH ue a 36 9 1036E K(JG) ae 7 36 9 1148E RE of var. deamit (Rehd.) Fern. 72 9 144E AA100-38-A C. 72 9 636E MT1006-39 (Deam) ‘ var. pubescens Ait. 36 8 AA18008 AA18008 e.N. Am. et 36 8g AA18009 AA18009 ” - 72 9 49 CU a ia 36 9 179E AA18009 ee " 36 9 639E MT355-46 (AA2106-4A) ‘“ “ 36 9 743E AA18008( DS) ef - 72 9 7T77E AAS5O70-1 7 a 72 9 974E HP a : 72 9 1019E K1938HK . V. dilatatum Thunb. 18 8 AA229-46-B AA229-46-B ce. As. 18 9 19E CU oe a 18 9 23E PI-76383 a “ 18 2) 30 PI-C3R29 «“ ~ 18 9 93E MR4 et a 18 9 525E TG, sd. (Japan) «“ . 18 9 526E TG, sd. (Japan) e " 18 9 S64E HB, sd. (Japan) ‘ / 18 9 784E AA7665—-D(CH) “ 18 9 832E AA137-52 (Japan, sd.) “ 18 9 835E AA138—52 i 18 9 888E UW248-49(E, sd.) " 18 9 1021E K i f{. hispidum Nakai 18 8 AA17486-1-A AA17486-1-A e. Asia ’ 18 9 146E AA17986-1-A te 7 18 9 386E NY(TG) a ie 18 9 731E N¥2311-36(TG, sd.) “ 18 9 S18E AA647-53

(Japan, sd.)

1962 | EGOLF, CYTOLOGY OF VIBURNUM Lod TABLE I (Continued) SOURCE SNER/ AvuTHOR- DOCUMEN- OF DISTRIBU- SPECIES 2n ITY TATION MATERIAL TION sf 18 9 828E AAS567-53 (Japan, sd.) ef : 18 9 831E AAS569-5 (Japan, sd.) s se 18 9 833E AA139-52 (Japan, sd.) 18 9 870E UW91-46(NY,sd.) “ f. pilosum (Thunb.) Nakai 18 9 446E NY(TG) e. As. s 18 9* 534E TG, sd.( Japan) si 18 9* 562E HB, sd.( Japan) se “ 18 9 563E es te f. xanthocarpum Rehd. 18 8 A A10140 AA10140 cult. a 18 9 40E A10140 a : 18 9 373E y . 18 9 S94E MT457-45(KH) = os 18 9* 1397E E(IC) _ V. ellipticum Hook. 18 9 430E LO w. USS. s 18 9 1060E K316-32(EN) "s V.erosum Thunb. 18 9 24E PI-4276P e. As. < 18 9 89E MR2015 ef " 18 9 515E HB, sd. s 18 9 595E se SS(AAI1S06 18 9 734E NY179 oe 18 9 810E ee aries g 18 9 881E UW168-S50(NY, sd.) “ ot 18 9 1081E s 18 g* 1398E E15163 Yii ee var. punctatum Franch. 18 9 511E TC(Japan, sd.) e. As. as 18 9 512E = 18 9 527E a sd.(Mt. Tawa) var. taquetii Rehd. 18 9* 1401E ~orea V. flavescens W.W.Sm 18 9* 202E eo 86, sd. China * 18 9* 882E UW274-49(E, sd.) 18 9* 1407E E w V. foetidum Wall. 16 6 — — Himal. x 18 9 440E RH e 2 18 9 1087E HL M var. rectangulatunt (Graebn.) Rehd. 16 6 — — w. China yi 18 Q* 1158E HA _ e 18 9* 1163E EX = V. hanceanum Maxim. 72 9 477E MT564-39 China os 72 9 821E AA1507-51(HL) ae a 72 9 1033E K124-26(AA) “ i 72 9 1132E P = V. hirtulum Rehd. 18 9 883E UW264-49(LS, sd.) China cs 18 g* 1784E wy i V. hupehense Rehd. 18 1 — AA c. China 18 2 — — “ (as V. hirtulum) 18 8 AA708-37-B AA708-37-B _

152 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLII TaBLe I (Continued) SOURCE GENERAL AUTHOR- DOCUMEN- OF DISTRIBU- SPECIES 2n TATION MATERIAL TION . 18 9 302E KR, sd. , 18 9 355E DU, sd. " : 18 9 463E MT, 390-39 e 18 9 678E . 18 9 T11E MR601A( Wilson) a" - 18 9 738E BB " us 18 9 765E AA18020 (Wilson 601) “ 18 9 769E AA708-37-B (LS, sd.) ‘ . 18 9 978E MR601A-5 ‘ 18 9 995E (Wilson 601) . oe 9 1080E RH be V. ichangense (Hemsl.) Rehd. 18 9 OE CL c. & w. China ‘ 18 9 695E PI-114819 ee 4 18 9 838E AA(KN) «“ ss 18 9 1072E RH et a 18 9 1085E HL i = 18 Q* 1162E EX 7 ot 18 g* 1319E BC V. japonicum (Thunb.) Spreng. 18 9 560E K apan oe 18 9 566E HB, sd. (Japan) - ae 18 9 876E VO re “ 18 9 887E UW631-50(HL) “ ns 18 9 1047E K167-37(DA) “ . 18 gx 1159E ld V. lobophyllum Graebn. 18 1 —_ AA c. & w. China T9 20 6 ae en (<9 . 22 7 EX ~ M 18 9 108E HP 2152 at re 18 9 131E AA19494-C a = 18 9 351E Du, sd. = 27 9 620E MT1303-35 (NB, sd.) “ - 18 9 730E NY67491(LS) a ee 18 9 CO7E —! ne 18 g* 1004E K88-08 (AA238w) se 7 18 9 1128E P i 7 18 9* 1472E ts V. molle Michx. 36 9 391E NY (HP) c.US ee 36 9 625E MT1044—40 * ee 36 9 803E A A18294-A (Palmer) i " 36 9 1016E 104-35(DU) f. leiophyllum Rehd. 18 8 AA4643-1-A AA4643-1-A c.US " 36 9 804E AA4643 (Bush) _ - 36 9 985K MT 1097-36 - V. ovatifolium Rehd. 18 1 -—- AA w. China 18 8 AA20078A AA200784 as

1962 | EGOLF, CYTOLOGY OF VIBURNUM 153 TABLE I (Continued) SOURCE GENERAL AUTHOR- DOCUMEN- OF DISTRIBU- SPECIES 2n TATION MATERIAL TION hy 18 9 290E UW (MT, sd.) i _ 18 9 764E AA20078 (Wilson 590) . sf 18 9 1025E K1008-34(NN) t 18 9 1076E H ty V. parvifolium Hayata 18 9 889K RH Formosa V. phlebotrichum S. & Z. 36 9 531E TG, sd. (Japan) an is 36 9 723E BT37-39 be ‘“ 36 9 869E UW38-—49(HL) ef or 36 9 1095E HL V. rafinesquianum Roem & Schult. 36 9 365E K e.N. Am SS 36 9 573E N10) : ‘6 36 9 76E LA t var. affine (Schneid.) House 20 8 AA4622-2-B AA4622-2-B e.N. Am. ie 36 9 71E : 36 9 768E AA17972 (Bush) tt “ 36 fe) 10S51E 78-36 oy V. recognitum Fern. 36 9 707E MR54-94 e.N. Am V. scabrellum Chapm. 72 9 C6E MR se. U.S oy 72 9 396E NY (BT) cs oe 72 9 567E MC us os 72 9 579E PN(Va.) n uf 72 9 714E MR2200 “ 72 9 745E AA11549-B (Harbison) i V. setigerum Hance (as V. theiferum Rehd.) 18 2 — c. & w. China 2 36 8 AA20189 AA20189 2 i 36 9 55E CU s ss 18,36 9 57E PI-104128 « 36 9 88E MR11( Wilson) is v 36 9 654E en an es R) s ey 36 9 655E MT52 a 36 9 709E MR218A (Wilson e - 36 9 744E AA20 aes 236) cs {. aurantiacum Rehd 36 8 AA812-32 AA812-32 c. & w. China a 36 9 SSE PI-023027 ae ie 36 9 60E PI-TN-R8 ef a 36 9 94E MR12 eee 236)---% fe 36 9 815E AA19 ana 236) sf V. wilsonit Rhed. 18 9 1053E K262-23(HL) w. China be 18 9 1068E DU a a 18 g* 1562E E262-33(HL) s V. wrighti Miq. 16 6 2450K RH Japan ~ 18 9 4E CL = 18 9 116E HP2202 +f i 18 9 286E Cc

154 JOURNAL OF THE ARNOLD ARBORETUM [| VOL. XLUI TABLE I (Continued) SOURCE GENERAL AuTHOR- DOCUMEN- OF DISTRIBU- SPECIES 2n TATION MATERIAL . 18 9 S16E HB, sd.(Japan) a ne 18 9 667K MT165-37 e “ 18 9 686K ‘ se " 18 781E AA18015 (Sargent) at 18 9 912 RH “ var. eglandulosum (Miq.) Nakai 9 565E HB, sd.( Japan) Japan var. hessei (Koehne) Rehd. 18 9 402K NY Japan 18 9 1031E K - Sect. IX. Opulus DC. V. edule (Michx.) Raf. 18 9 873E UW 1053-50 (White R.) n.N. Am. - 18 g* 1098E HL _ V. kansuensis Batal. 18 9 1083E RH w. China ~ 18 9 1094E HL “ nt 18 g* 1453E GP 13248 L&S : V. opulus L. 18 1 = AA Eur., n. Afr. “ 18 —— — “ (as V. sargentii) 20 6 2459K RH s 18 9 39E “U 18 9 101E K2491, sd. is “ 18 0) 324K PO 7 e 18 9 344E DU _ = 18 9 375E NY ae - 18 9 400E NY (Austria) ag 18 9 459E MT1189-35 zs 7 18 9 S45E < is 18 9 672E MT49-52(AA) « “ 18 9 687K KH a 18 9 741E AA * 18 9 786E AA20736(E, sd.) a " 18 9 1013E AA562-30 = ag 18 9 1030E K53-46(RH) es ff 18 9 1032E K94-29(G) he 7 18 9 1063F K106-63 . ‘Aureum’ 18 9 629E MT1046-40(SP) cult “ 18 9 842K AA997-52 ™ 18 9 892E RH en . 18 9 1012E K “i ‘Compactum’ 18 9 417E RU cult “ 18 9* 1485E JG a var, nmanum (David) Zabel. 18 9 524E CU cult " 18 9 630E MT118-53( HL) 3 ae 18 9 981E C ns ‘Notcutt’ 18 9 670K MT211-51 cult - 18 9 747E AA814-38( NN) °

1962 | EGOLF, CYTOLOGY OF VIBURNUM 155 TABLE I (Continued) SOURCE GENERAL AuTHOoR- DOCUMEN- OF DISTRIBU- SPECIES 2n ITY TATION MatTERIAL TION f. roseum (L.) Heg. (as V. OC var. sterile) 18 6 2463K RH cult. 18 ) 41E CU ve e 18 9 246E MT156-38 - Z 18 9 632E MT1098-36( AA) A oo 18 9 790E AA26-47-A Z i 18 9 1107E a i 18 9 1118E P “ f, variegatum (West) Zabel 18 9 pee DA cult ee 9 33E T528-51(DA) ce {. xanthocarpum (Endl.) Rehd. 18 8 AAn258 28-A AA1298-28-A cult. iz9 18 9 O4E P2 ae oe 18 9 oe U * fe 18 9 634E MT1099-36(HP) . 18 9) 791E AA1298-28-A ms V. orientale Pall. 18 9 175E AA677-33 w. As Me 18 g 296K UW(MT, sd.) “ # 18 9 837E AA934—52 _ V.sargentii Koehne 18 1 — A! ne. As st 18 9 74E PI-81798 if o 18 9 300E KR mn es 18 9 337E PO a i 18 9 404E RU - . 18 9 794E AA18012-B(Korea) “ “ 18 9 827E AA571— ee sd.) i i 18 9 829E AA saan, sd.) = - 18 ) 830E AA saan, sd.) ee 18 9 834E AA es sd.) a * 18 9 COlE RH ss f. calvescens (Rehd.) Rehd. 18 8 AA467-26 AA467-26 ne. As. i 18 Q* 4C4E AA467-26 es : 18 9 797E ee sd.) st 18 9 798E eae 1QAGEHe at a 18 9 1042E K234-38 Marsh o f. flavum Rehd 18 9 117E HP2159 cult. 18 g* 130E AA21419 se ‘ 18 9 647E MT2195-22(HP) es in 18 9 795E AA11037(HP) . ‘Puberulum’ 18 9 649E MT975-38(GB) cult. V. trilobum Marsh 18 1 — AA n.N. Am 18 6 — — 4: 18 9 327E PO a is 18 9 659E MT1228-38 af oe 18 9 816E AA15673 i i: 18 9 1058E K205-48(KR) oe

156 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIII

TaBLeE I (Continued)

SOURCE GENERAL AuTHOR- DoCUMEN- OF DISTRIBU- SPECIES 2n ITY TATION MATERIAL TION ‘Andrews’ 18 9 248E MT1055-—40 cult. 18 9 793E AA292-—42(LI) es af 18 9* 1663E WLA66-104 ee ‘Compactum’ 18 8 AA871-51 AA871-51 cult. - 18 9 664E MT216-53(KH) _ " 18 9 684E KH ne ‘Hans’ 18 9 661E MTS801-40 cult. “ 18 9 789E AA293-42(LI) it ‘Wentworth’ 18 9 663E MT1057-40 cult. a 18 9 783E AA294-42(LI) « Unidentified species 18 9 693E PI-22978 uf 18 9 700E PI-111382

CONTROVERSIAL CHROMOSOME COUNTS

Poucques (27, 28) reported 2” = 18 in Viburnum sandankwa, a syno- nym of V. suspensum, while a somatic chromosome number of 16 is re- ported here. However, since several different evergreen forms have been referred to V. sandankwa, it is possible that Poucques’ plant was not the same as V. suspensum. Simonet and Miedzyrzecki (35) reported a chromo- some number 2” = 18 for V. setigerum (V. theiferum). In the present study a single plant of this species was found with a chromosome number 2n = 18; whereas other determinations, including that of Thomas (40), revealed a chromosome number of 2” = 36.

The somatic chromosome counts reported in this study for thirteen species and four varieties differ from those reported by Janaki Ammal (18). In order to check Janaki Ammal’s counts, and, if possible, to resolve the differences between our studies, an attempt was made to secure all the species grown at the Royal Horticultural Society’s Garden, Wisley; at the Royal Botanic Gardens, Kew; and in the Jardin des Plantes, Paris. It is assumed that plants from these sources are similar to the material studied by her, but there is no assurance that the same plants were sampled. Since she published no record of the particular plants involved, it is uncertain whether she studied the same species in all three gardens or in only one. In any case, it seems logical to assume that some of the plants included in this study were the same as some of those she studied.

As with most cytological investigations, many of the studies must be made with cultivated plants that may be variants of the native species. The plant material used or the technique used could result in differences of chromosome counts. In a personal letter from Janaki Ammal it was stated that she had used the lacmoid leaf-bud technique (10), whereas

1962] EGOLF, CYTOLOGY OF VIBURNUM Lo?

in the present study root-tip smears were used. A critical comparison of her cytological study with the present one is impossible, for no slides, drawings, or photomicrographs and only a limited number of herbarium specimens were available. For clarity of discussion the differences between her and my counts are grouped into two classes: 1) species differing by a few chromosomes per complement and, 2) those differing in the number of sets of chromosomes in each complement.

In the first category, Viburnum bitchiuense, V. foetidum var. foetidum, V. foetidum var. rectangulatum, V. mongolicum, and V. wrightii were determined to have 2” = 18 chromosomes, two more than reported by Janaki Ammal (18). Among the plants studied were V. bitchiuense, from Kew, and V. bitchiuense, V. foetidum vars. foetidum and rectangulatum, and V. wright from the R.H.S. Garden, Wisley. In the collection at Wisley is a plant identified as V. foetidum var. rectangulatum which is probably the plant studied by Janaki Ammal. Plant 1084E of this study, which was received from the R.H.S. Garden as V. foetidum var. rectangu- latum and which has horizontally spreading branches, oblong-ovate leaves, and an inflorescence with fertile flowers surrounded by sterile marginal flowers, is correctly identified as V. plicatum {. tomentosum. Upon visiting the R.H.S. Garden the author further verified the identification of this individual plant.

The present study has revealed that Viburnum carlesi is composed of a complex 2” = 18, 20, and 22 chromosome forms. All the plants studied that were obtained from cultivation, including a plant from Kew, had 2n = 18 chromosomes, which agrees with the reports by Poucques (28) and by Simonet and Miedzyrzecki (35). Janaki Ammal (18) reported this species have 2m = 20 chromosomes.

The present study agrees with the previously reported counts of 2n = 18 in Viburnum buddletfolium (35) and in V. lobophyllum (32), but Janaki Ammal (18) reported 2” = 20 and 22 (the count of Enoch) for V. lobophyllum. Included in my study was a plant of V. lobophyllum from Kew.

In this study both Viburnum plicatum {. plicatum and f{. tomcntosum were determined to have forms with 2n = 16 and 2m = 18. Janaki Ammal (18) reported both these and f. mariesii to have 2n = 18. The plants of f. mariesii and f. tomentosum from Kew and of f. mariesii from the R.H.S. Garden used in this study have chromosome complements of 2n = 16. The herbarium specimens deposited at Kew by Janaki Ammal clarify the discrepancies, and this documentation has been added to Tasre I. Speci- men 2456K of V. plicatum {. mariesii (V. tomentosum mariesu) has an annotation note “2m = 16?” which definitely indicates that her published count was questionable. The specimen of V. plicatum (V. tomentosum sterile), 2463K, collected at the R.H.S. Garden, is identified by me as V. opulus £. roseum. Likewise, specimen 2459K of V. sargentii, collected at the R.H.S. Garden, is V. opulus. The plants of V. odoratissimum from Kew were found to be of the 2m = 32 variation. Janaki Ammal reports 2n = 40 for this species. This last number, however, was found elsewhere

158 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLII

in the V. odoratissimum complex in this study and has likewise been reported by Sugiura (38).

The number of genomes reported by Janaki Ammal (18) in the comple- ments of four species and one variety differ from those found in this study. Viburnum erubescens, V. henryi, and V. dentatum, including in the present study specimens of all three from Kew and of V. dentatum from R.H.S., were found to be tetraploid, whereas Janaki Ammal re- ported them all to be hexaploid. Janaki Ammal’s theory that the species evolved after spontaneous doubling of an unstable triploid obviously is not supported by this new evidence. Viburnum sieboldii, reported by Janaki Ammal (18) to be a diploid, was determined in this study to have both diploid and tetraploid forms, the tetraploid occurring much more frequently. The plant from the R.H.S. Garden studied was the tetraploid form. Janaki Ammal (18) reported V. fragrans ‘Album’ to be a tetraploid, but plants of this variety from R.H.S. Garden, and from other sources utilized in this study, were determined to be diploids. Thomas (40) also reports this variety to be diploid.

Among the 536 counts listed by the author (14) were a number of plants procured from the Arnold Arboretum. Thirteen of the plants studied by Thomas (40) are identical with those from which the author acquired material, while another six plants have the same Arnold Arbo- retum accession number but are not necessarily the same individual plant. The author lists the counts of an additional sixty-six plants from the Arnold Arboretum. Two of Thomas’ counts, those for Viburneum rafi- nesquianum var. affine (2n = 20) and V. molle f. leiophyllum (2n = 18), differ significantly from the present work. Viburnum rafinesquianum var. affine was determined to be 2m = 36 for all plants studied. A plant of V. molle {, leiophyllum bearing the Arnold Arboretum accession number 4643 was determined to be 2” = 36, while Thomas reports 2n = 20 for plant 4643—-1—A, the latter having been propagated vegetatively from one of the original lot. Quite possibly, either a mixed lot or a mistake in labeling may be involved.

VARIATION IN CHROMOSOME NUMBERS WITHIN SPECIES AND VARIETIES

Differences in chromosome number were found within six species and five varieties. These differences can be placed in two classes for dis- cussion: 1) those species that differ by a few chromosomes and, 2) those species that differ by a number of genomes.

The first category includes Viburnum plicatum and V. carlesii. Morpho- logically indistinguishable plants of V. plicatum, with 2n = 16 and 2n — 18, have been found. Comparison of the chromosome complements of these two forms reveals that the form with 2” = 18 has an extra pair of metacentric chromosomes. This species will be discussed later in more detail. All plants of V. carlesii from cultivation have 2” chromosome complements of 18. However, among plants produced from seeds col-

1962 | EGOLF, CYTOLOGY OF VIBURNUM 159

lected in Korea, the native habitat of the species, chromosome comple- ments of 2” = 18, 20, and 22 have been found.

Six species that differed in number of genomes were studied. One plant of Viburnum lobophyllum (2n = 27) was determined to be a triploid, with three cytologically identical genomes. All other specimens of this species studied were diploid, with 21 = 18. Since the triploid plant appears iden- tical with the diploid, it does not seem likely that this is a hybrid between the diploid V. lobophyllum and one of the tetraploid species. It is possi- ble that this triploid could have resulted from the fertilization of an un- reduced functional gamete by a normal gamete. Likewise, it could have originated as a cross between diploid and tetraploid plants of V. lobophyl- lum, although this does not seem probable since no tetraploid V. lobophyl- lum is known.

One species, Viburnum odoratissimum, is represented by tetraploid and pentaploid forms. These two forms, which have distinct vegetative dif- ferences, comprise a taxonomic complex that is given further consideration later. The pentaploid form, with 27 = 40, has five similar genomes, each of which is porphologically, identical with the genomes of the tetraploid. It is unknown whether the four pentaploids and seven tetraploids are representative of the variation that occurs in native populations.

Individual diploid plants of Viburnum sieboldii and V. setigerum were discovered in species that otherwise are tetraploid. The diploid plant of V. setigerum was isolated from a group of plants grown from seed obtained from the U. S. Plant Introduction Garden, Glenn Dale, Maryland. This plant probably resulted from parthenogenesis wherein an unreduced gamete developed without syngamy. The diploid plant of V. steboldii was secured from a commercial nursery and it is not known whether the plant was propagated asexually or from seed. The diploid plants of these species are still immature, making it impossible to compare them critically with tetra- ploid plants. Although at this time they do not appear morphologically distinct from the tetraploid, they are obviously somewhat weaker, and growth has been slower. This may be partially or entirely due to environ- mental conditions, however.

Only six of the sixty-one varieties observed in this study had chromosome complements with numbers different from the species (i.e., the typical varieties). Viburnum tinus var. lucidum, V. dentatum var. deamtt, and V. dentatum var. pubescens, all 2n = 72, had double the number of gen- omes of the typical variety of the species. One of the five collections of V. lantana var. rugosum was found to be a triploid, 2x = 27. Though there are three cytologically identical genomes in this particular plant, it does not appear to be morphologically distinct from the other collections. It is conceivable that this plant developed from a chance unreduced gamete that was fertilized by a normal gamete.

In Viburnum plicatum {. tomentosum and V. plicatum f. mariesit, as well as in V. plicatum f. plicatum, occur both 2m = 16 and 18 chromo- some forms which cannot be distinguished by vegetative characteristics. All other taxa of this species which have been examined (V. plicatum f.

160 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLII

glabrum, V. plicatum ‘Lanarth,’ V. plicatum f. lanceolatum, V. plicatum f. mariesit, V. plicatum ‘Roseum,’ V. plicatum ‘Rowallane,’ and V. plicatum ‘St. Keverne’) have chromosome complements of 2” = 16. The forms with 2m = 18 differ from those with 2n = 16 by an additional pair of metacentric chromosomes. Since the 2m = 16 plants produced the most abundant fruit, it is to be expected that these forms should have a higher rate of survival. The evidence indicates that most of the minor variations in Viburnum that have been given varietal rank are the result of genic or intrachromosomal changes, rather than the result of changes in chromo- some numbers.

CYTOTAXONOMIC COMPLEXES

Differences in the chromosome numbers of different collections of V1- burnum odoratissimum, V. carlesti, and V. dentatum suggest that each of these is a species complex.

VIBURNUM ODORATISSIMUM. Two distinct forms are evident in V. odora- tissimum, the plants of which were secured from nine different sources. The one has smooth-barked branches; thin, coriaceous, elliptic-ovate leaves; and indistinct axillary buds. The other has stout, lenticular branches; thick, coriaceous, elliptic-obovate leaves; and prominent axillary buds. The illustrations of V. odoratissimum by Dippel (13) and of V. awabuki by Nakai (23) portray two types which are respectively similar to the two forms observed in the present study. Nakai (23) recognized V. odoratis- simum, V. liukinense, V. awabuki, and V. awabuki var. serratum in this complex, but other taxonomists generally have accepted only one species, V. odoratissimum. Because all the plants used in this study have not yet produced flowers, it is impossible to make a positive identification of these variants. Chromosome counts of 2m = 32 (tetraploid) and 2” = 40 (pen- taploid) have been observed among both morphological forms of this species complex. Two of the three plants with 27 = 40 are from the same original source and are of the variant with coriaceous leaves and stout, lenticular branches. Sugiura’s (38) report of 27 = 40 in V. awabuki also indicates that his V. awabuki may be different from V. odoratissimum. Do these represent species or are they variants of one species? Although the cultivated material studied probably does not differ from representa- tives of the native populations of this species complex, additional material from known populations and further study will be necessary to determine the relationships within the species.

VIBURNUM CARLESII. In this study all plants of V. carlesii received from cultivation have chromosome counts of 2m = 18. However, seeds sup- posedly collected from native populations in China and Korea produced an array of plants with 2m = 18, 20, or 22 chromosomes. It is difficult, if not impossible, to explain this variation if this is a true species. Of course, there is no assurance that this seed was from isolated plants and not from

1962 | EGOLF, CYTOLOGY OF VIBURNUM 161

plants growing near other species with which there could be cross-pollina- tion.

As previously noted, Janaki Ammal (18) proposed that Viburnum car- lesii arose as a backcross between a chance triploid (2m = 24) and the normal diploid (2m = 16) of V. bitchiuense. In view of the present study, her explanation appears inadequate, for V. bitchiuense has 2n = 18 chro- mosomes, rather than the 16 chromosomes required by her proposal. If triploid plants exist in nature, this variation in the chromosome complement of seedlings might be the result of self-pollination or cross-pollination be- tween the triploid and diploid forms, which would produce progeny with additional chromosomes. It is doubtful, even if the triploid does exist, that fruit would normally result from self-pollination.

Most taxonomists consider Viburnum bitchiuense and V. carlesii to be closely related. However, few have gone to the extreme of Nakai (24), who not only reduced V. bitchiuense to a variety of V. carlesii, but also put both of these in a new genus Solenolantana. It is doubtful if this complex requires the latter action. Pollinations by the author of V. bitchiuense (2n = 18) X V. carlesii (2n = 18) produced no seed, while the reciprocal cross between these species produced, from ninety-nine flowers pollinated, forty-three seed which have yielded forty-one plants. A meiotic chromo- some study of these plants should indicate more clearly the natural rela- tionships of these species. Until a sporocyte study is made of authentic materials collected from native populations, this remains an unsolved cytotaxonomic complex.

VIBURNUM DENTATUM. Viburnum dentatum var. dentatum and V. denta- tum var. pubescens, represented by 2n = 36 and 2n = 72, form the third cytotaxonomic complex. According to Rehder (31), V. pubescens is a synonym of V. dentatum var. pubescens. In the V. dentatum—pubescens complex specific delimitation has been based almost entirely on the pres- ence and distribution of pubescence on petioles, leaf surfaces, and inflores- cence branches or combinations of these. Blake (4), who examined So- lander’s manuscript of Hortus Kewensis, an early treatment of this com- plex, and who also studied native material, recognized V. pubescens, V. pubescens var. canbyi, and V. pubescens var. longifolium. Rehder (29) de- scribed two new varieties from Indiana, V. pubescens var. deamti and V. pubescens var. indianense, which have only minor differences from each other.

Svenson (39) commented on the variations of this group, which he separated on the basis of leaf shape and pubescence into Viburnum denta- tum, V. dentatum var. lucidum, V. dentatum var. pubescens and V. pubes- cens var. semitomentosum. He also concluded that V. pubescens var. deamii probably is a separable variation.

Fernald (16) was in disagreement with Svenson’s reduction of the glabrous-twigged form of Viburnum dentatum to varietal rank as V. denta- tum var. lucidum, and elevated this glabrous variation to the rank of species with the specific name V. recognitum. Viburnum recognitum is dis-

162 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIII

tinguished from V. dentatum by its glabrous branchlets and cyme, glabrous or glabrate foliage, and flowers produced ten days to three weeks earlier. Fernald further stated that V. dentatum and its variations and V. recogni- tum “are both hopelessly variable in leaf outline and toothing of leaves, each of them with blades varying from lance-ovate to ovate-oblong to or- bicular, with veins prominent beneath or obscure, with length from 2.5 to 10 cm. and breadth from 2 to 8 cm.” Viburnum recognitum, V. recognitum var. alabamense, V. crenatum, V. dentatum, V. carolinianum, V. carolinta- num var. deamii, and V. pubescens are entities recognized by McAtee (21) in this species complex.

The present study follows Rehder’s treatment which reduces this com- plex to Viburnum dentatum, V. dentatum var. pubescens, and V. dentatum var. deami. The material studied here was received under practically every one of the different names applied by past authors to these varia- tions. Many of these collections were from cultivated plants, since mate- rials from native populations were not available for all species or varieties. Therefore, it must be realized that the material studied is not necessarily an adequate sampling of variability of the native population, but it does give indications which may establish a basis for further research. The plants are being maintained for further study and identification.

Two chromosomal forms, 27 = 36 and 2m = 72, were located among plants of Viburnum dentatum var. pubescens. All plants of V. dentatum var. deamti had 2n = 72. Are these plants of another variety or polyploid forms of the same species? Are these variations within a species the result of natural hybridization that has been followed by segregation and selec- tion? The variations between V. dentatum var. dentatum, var. pubescens, and var. deamii have been described taxonomically but are only of minor magnitude. Likewise, V. recognitum is only slightly different from V. dentatum, though it has been elevated to specific rank. Are these cases in which speciation is resulting from natural hybridization but in which di- vergence of types is not yet great enough for complete delimitation? Is the glabrous-branched V. recognitum a plant of a segregating hybrid population or does it represent another species?

It has been suggested that Viburnum dentatum may have crossed with V. rafinesquianum, another related species, to produce by introgressive hybridization a different ecological population. Normally, however, these species are isolated by season of bloom, V. dentatum flowering approxi- mately ten days later than V. rafinesquianum, and by habitat, V. dentatum being located on moist soil and V. rafinesquianum on dry upland. It is possible that a few late flowers of V. rafinesquianum may be shedding viable pollen when the first V. dentatum flowers open, and the distance between plants would not prevent cross-pollination. The offspring produced by such crosses could, over a period of time, produce a population of a type differing from the original species of the locale. In the particular area of South Hill, near Ithaca, N.Y., where plants were used for controlled pol- linations in this study, there is no evidence of natural hybridization, but in similar situations elsewhere the densely pubescent V. rafinesquianum may

1962 | EGOLF, CYTOLOGY OF VIBURNUM 163

have crossed with the ay glabrous V. dentatum to produce a population with additional variation

A further pees arises from the report by Hes Ammal (18) that Viburnum dentatum has a chromosome number of 2” = 54. In the present study, 2x = 36 has been counted in all cases. It is ‘gical to think that her plant was a hybrid, unless the count was incorrectly determined, since all of her work was done on cultivated plants growing in close proxim- ity and hence subject to crossing. Since it is the custom of many botanic gardens to raise plants from seed, not realizing the seed may be from cross-pollination and not true for the species type, it is possible that the plants studied by her may have originated in this way.

One approach to the problem is experimental; that is, to reproduce a similar plant by controlled hybridization. All the possible pollination combinations between Viburnum dentatum var. dentatum, V. dentatum var. pubescens, and V. rafinesquianum were made in the present study. From 386 flowers of V. dentatum pollinated by V. rafinesquianum were produced 190 seeds from which 123 plants have been grown. From 130 flowers pol- linated in the cross V. dentatum var. dentatum * var. pubescens, 49 seed and 15 plants were obtained. The only other combination to yield seed was V. dentatum var. pubescens * V. rafinesquianum which, from 329 pollinations, produced five seed that yielded two plants. The other com- binations failed to produce seed, but this cannot be attributed necessarily to sterility or incompatibility, for climatic conditions and technique may have been variable factors. That seeds and plants were procured from crosses between these species and varieties indicates a relationship within the complex. As these plants attain flowering size, a study of the meiotic chromosome configurations of the sporocytes should reveal the inter- relationships more definitely. It will be desirable to repeat those crosses that produced no seed and to attempt additional crosses with other related species from section ODONTOTINUS.

Because polyploids are found in this complex, a study of the native populations will be required to resolve the problem. Such experimental studies should not be concentrated within a few isolated populations but should cover the distribution range of this species complex so that dif- ferences and relationships between populations, as well as within popula- tions, can be determined. With the union of the evidence from cytology, genetics, and taxonomy the intricate relationships of this complex should eventually be clarified further.

The section ODONTOTINUS of Rehder’s classification includes other species that occur in the same geographical areas as members of the Viburnum dentatum—pubescens complex. Other taxa related to this complex and war- ranting study include V. molle, V. scabrellum, V. bracteatum, and V. rafines- gquianum. The Chinese V. Aanceanum appears to be allied closely to this complex. When the plants studied have produced fruit and the identifica- tion has been checked, a more valid interpretation of V. kanceanum may result.

This cytotaxonomic complex in eastern North America and the two com-

164 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIII

plexes in Asia may be representative of those that exist in other species groups of this genus. The author is inclined to believe that a similar com- plex may exist with V. cassinoides, V. nudum, and V. lentago of eastern North America.

PHYLOGENETIC RELATIONSHIPS

The phylogenetic relationships within a genus can be accurately formu- lated only when the number, morphology, and behavior of chromosomes are correlated with anatomy, morphology, and taxonomy of the species. A tentative scheme for the species of Viburnum has been constructed utiliz- ing all the available data (Fic. 16).

The genus Viburnum includes polyploid series with the basic numbers of eight and nine. Polyploidy can take place effectively only in one direction; the diploid must nearly always be the parent of the polyploid (11, 33). Stebbins (37) emphasizes the point that diploid members must be older than the polyploids, although they are not necessarily more primitive in the sense that they are less specialized in structure. Since higher polyploids (tetraploids, hexaploids, etc.) usually cannot revert to the diploid without abnormalities in the reproductive cycle, it is likely that one of the lower gametic numbers, i.e., 8 or 9, is primitive.

Wilkinson (44) presented additional evidence of the natural relation- ships of selected species in the genus. She reported on fourteen species, representing all but two of the sections of the genus, which were placed in five groups on the basis of their internal morphology and the vascular anatomy of their flowers. Viburnum sieboldii (n = 8) is the most primitive of those studied, with only two characteristics that might be considered ad- vanced: reduction of peripheral bundles to five and the reduction of the sepal supply to a single unbranched trace. In no other species included in her study were so many primitive characteristics present. A group of relatively primitive species includes V. carlesti, V. lantana, and V. denta- tum, all n = 9. Another group of less primitive species would include V. lantanoides (n = 9), V. plicatum {. tomentosum (n = 9), and V. lentago (x = 9). The group of more advanced nine-chromosome species includes V. rhytidophyllum, V. nudum, V. cassinoides, V. dilatatum, and V. trilobum. Viburnum opulus (n = 9) is considered the most advanced. Her work supports the theory that the species with the basic number of eight are primitive, while those with a basic number of nine are more advanced.

From a study of the stem anatomy, De Vos (12) concurred that V. opu- lus, V. lentago, and V. cassinoides are the most advanced and that V. pli- catum {. tomentosum, V. lantanoides, and V. sieboldii are the most primi- tive.

To date, this is the extent of comparative morphological and anatomical studies of Viburnum species. Viburnum sieboldii (n = 8) is on these grounds considered to be the most primitive species of the genus. All other species with a basic number of eight have many characteristics in common with V. sieboldii, so that it can be assumed that this group Is more primitive

1962 | EGOLF, CYTOLOGY OF VIBURNUM 165

*— SPECIES WITH MORE come HAN ONE CHROMOSOME NUMB

V. bracteatum tu

2n = 40

* V. adoratissimum

srubescens

" <<<<<< * ox N

=. a 2

*

2n = 27 * V. lobophyllum

x V. lantana In =18 V. acerifolium * V.lantona V. atrocyaneum V. lantanoides V. betulifolium V. lentago V. bitchivense * V. lobophyll V. buddleifolium V. macrocephalum V. burejaeticum V. microphyllum V. x burkwoodii V. mongolicum Vic V. nudum Vix “corleephalom V. obovatum *V. carlesii V. opulus V. x carlotta V. orientale V. cassinoides V. ovatifolium chenaultii V. parvifolium V. cinnamomifolium * V. plicatu coria V. propinquu V. cotinifolium V. prunifoliu yonthum V. x rhytidocarpum vidii V. x rhytidophylloides V. dilatatum V. chytidophyllum pad . edul V. rigidum Q@n=18 V. ellipticum V. rufidulu V. eros V. saorgent V. fHlavescens V. schensianum foetidu V. sempervirens V. furcatum * V. setigerum . glomeratum V. stellulatum =16 V. harryanum Vv. sympodial Von bodaenteare V. hirtuly V. trilobum hupehense V. urceolatum V. foetens , ' V. ichangense V. utile V. fragrans V. x jockii V. veitchii V. grandiflorum V. joponicum V. wilsonii * V. plicatum V. x juddii Vv. ig iv sieboldt V. kansvense Unidentified V. suspensum

Fic. 16. Diagram alan on chromosome complements to show interrelation- ships of species of Viburn

166 JOURNAL OF THE ARNOLD ARBORETUM [ VOL, XLIII

than those species with a basic number of nine. But the possibility re- mains that another species within the m = 8 group may be more primitive than V. sieboldii, and likewise that V. opulus need not necessarily be the most advanced species of the genus. Only a small sample of species and of the variation that exists in this widely distributed genus has been evaluated critically. Opinion concerning the relative position of species in regard to primitiveness may shift when additional evidence is available, but the basic cytological relationships of the various chromosome groups appear to be well established.

On the basis of floral anatomy Wilkinson (43, 45) postulates that there are in the Caprifoliaceae two lines of development. Through a form re- sembling the prototype postulated for Viburnum one line leads to Viburnum and Sambucus; the other stems through Leycesteria and branches sepa- rately to the Loniceraeae and Linnaeeae. The evidence at hand is in- adequate to hazard an interpretation or conclusion as to what was the prototype for the family and from whence Viburnum arose, but the present evidence favors the » = 8 forms as the more primitive.

ALTERATION OF BASIC CHROMOSOME NUMBER

Evolutionary changes within a genus may be due to polyploidy, to the addition or subtraction of one or a few chromosomes of a complement, to gross structural rearrangements of the chromosomes, to submicroscopic changes, probably involving the chemistry of the chromosomal material, or to any combination of these. It can be assumed that all these changes have probably functioned in speciation in Viburnum. However, the genus has not been studied sufficiently to ascertain the evolutionary significance of each. For this reason, this discussion of phylogenetic relationships will be centered primarily around the evolutionary significance of the basic chromosome numbers of the genus.

Navashin (25) realized that changes in the basic number must involve loss or gain of the existing kinetochore, since kinetochores or kinetochore modifications cannot arise de novo. In this light, the increase or decrease in chromosome number attributed to “fragmentation” and “fusion” (7) could occur only when it involved a gain or loss of the kinetochore.

Darlington (8) presented a procedure favoring the loss or gain of a chromosome by means of an equal translocation between two different nonhomologous chromosomes with subterminal kinetochores. An inter- change involving the long arm of one and the short arm of the other would produce one long metacentric chromosome and one very short chromo- some or fragment. It has been pointed out that the consequences of un- equal translocation depend on whether the regions about the kinetochore are genetically active or inert. An inert centric fragment may be elimi- nated, with a consequent reduction in chromosome number. If the fragment chromosome is genetically active it may persist as a univalent and be passed at meiotic metaphase to the same pole as the other interchange chromosome. This will yield, in addition to normal gametes, gametes

1962] EGOLF, CYTOLOGY OF VIBURNUM 167

with both or none of the interchange chromosomes. The union of gametes with additional chromosomes may yield trisomic or eventually tetrasomic plants which evolve into species with permanently increased basic numbers. Morphological differences may arise by virtue of either dosage effect or by divergent gene mutation in the duplicated chromosomes. Further cyto- logical divergence may arise from reciprocal translocation between one of the new extra chromosomes and another chromosome of the comple- ment (5, 37, 1).

Tobgy (41) confirmed Darlington’s postulate with the demonstration that a Ea: reciprocal translocation between two chromosomes of Crefis neglecta gave rise to one chromosome of Crepis fuliginosa. Of the two chromosomes resulting from this translocation, the one with a genetically inactive region adjacent to the kinetochore was lost. Likewise, Sherman (34) obtained evidence that the origin of the Crepis kotschyana (n = 4) complement involved reciprocal translocation in the reduction from five to four pairs of chromosomes

Chromosome number can also be increased or decreased by aberrations such as the translocations observed by Thomas (40) in the meiotic cycle.

Asynapsis, desynapsis, nondisjunction, and chromosome lagging may be responsible for the production either of gametes with a single extra

mosome, several extra chromosomes, or with the entire unreduced eeicntee or of other gametes with chromosomal deficiencies. The union of such gametes may result in individuals deficient in or with addi- tional chromosomes.

Chromosome numbers may be increased by supernumeraries or by mis- division of the kinetochore. White (42) considers the formation of super- numeraries, fragments produced by deletion or translocation, and frag- mentation of the kinetochore to be probably the chief method whereby chromosome numbers have become increased in the course of evolution in animals. The fragment lacking the kinetochore region is lost in sub- sequent divisions unless it is translocated to another chromosome. Thus, fragmentation, in association with translocation, provides a mechanism or chromosome number increase. According to Darlington and Mather (11), misdivision of the kinetochore is the only single change that can affect both the number and structure of the chromosomes in a single stroke. The kinetochore, rather than dividing lengthwise, divides cross- wise, resulting in two telocentric chromosomes which at a later division may produce two pairs of unlike isochromosomes (9). Races of Campan- ula persicifolia (11) have been found in which two telocentric chromo- somes occur instead of a single chromosome and thus add an additional chromosome to the haploid number

The backcrossing of a triploid, produced by a cross between a tetra- ploid and diploid plant, to a diploid has experimentally produced a great variety of segregant types, from among which have been recovered a small proportion of fertile types. Examples of such results have been reported in Triticum by O’Mara (26) and in Gossypium by Beasley and Brown (3) and have been summarized for Nicotiana by Goodspeed (17). This

168 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIII

system of hybridization has artificially changed the basic chromosome number but has not been established as occurring in the natural evolution of species.

These examples individually, or in combination, illustrate numerous means whereby the basic chromosome number could increase or decrease as the taxonomic group evolved. Likewise, the progressive increase or decrease in basic number could be followed or accompanied by amphi- diploidy to produce a complicated aneuploid series such as occurs in Carex, Iris, Sedum, Viola, and other genera. Stebbins’ (37) summary of the types of aneuploid series in higher plants shows that the number of descending basic series is far greater than either the ascending basic or interchange amphidiploid series.

It is possible that in Viburnum the change in the basic chromosome number from eight to nine was adequate to keep the two types isolated and to allow each to evolve independently. As the types were exposed to new and changing environments the selection pressure further subdivided the major groups into minor groups with differential adaptability. It is conceivable that under certain circumstances the polyploid was favored over the diploid, or in others vice versa, thus increasing the variability. During this interval adaptive mutations could arise in certain subgroups. As the minor groups became further subdivided and isolated for survival in specific environmental niches, the variability within the genus expanded until today many Viburnum species have great morphological divergence. Although the phylogenetic relationships suggested by the chromosome numbers of Viburnum species provide a basic framework from which the specific differences evolved, the pathway remains obscure.

CHROMOSOMES AND THE TAXONOMIC SECTIONS

The nine taxonomic sections of the genus, based on morphological characters, as recognized by Rehder (30) may be correlated with the pattern of cytological relationships. Section THyrsosmMa, which includes Viburnum sieboldii, is composed entirely of species with a basic chromo- some number of eight. This number occurs in only one other section, PsEUDOPULUS, in which, however, V. plicatum and V. plicatum {. tomento- sum have forms with both » = 8 and m = 9. The question arises as to whether V. plicatum and V. plicatum {. tomentosum, both with cytological forms that are morphologically indistinguishable, represent the connecting link in the evolution of the genus between the basic numbers of eight and nine. Wilkinson (43) concluded from the study of floral anatomy and morphology that V. plicatum f{. tomentosum was relatively primitive, but not as primitive as V. carlesti,n = 9, of sect. LANTANA, and V. dentatum, of sect. Opontotinus. This does not support the proposition that V. plicatum forms an evolutionary bridge between the groups of species with m = 8 and » = 9. It is probable that » = 9 may have evolved more than once and in various places. However, with additional study this relationship may be clarified.

1962 | EGOLF, CYTOLOGY OF VIBURNUM 169

Only diploids are found in sects. PsEUDOPULUS, LENTAGO, MEGALOTINUS, Oputus, and LaNnTANA, with the exception of V. carlestt (2n = 18, 20, 22) and V. lantana var. rugosum (2n = 27

Tetraploids and higher polyploids are found in sects. THyRsosMa, TINUS, and OpontToTINus. The scheme showing the relationship between chromo- some complements is presented in Fic. 16, which shows the evolutionary trend to be from 2m = 16 to 2m = 32, and from 16 to 18 to 36 to 72. In the evolution of the genus the diploids probably have had the highest adaptive value and today are represented by the largest number of species. At the present time it appears impossible to separate the diploid species into taxonomic sections on the basis of chromosome morphology. The distinct gross morphological differences used by the taxonomist to divide the genus into sections appear to be the result of genic rather than structural chromosomal changes. When karyotype analysis has been com- pleted for these species, differences of arm length, secondary constrictions, kinetochore position, satellites, and size of chromosomes may reveal natural relationships between species and sections.

A study of the chromosome complements of polyploid species reveals that the genome is duplicated. These species, preceded by an asterisk in Fic. 16, provide additional evolutionary information. For example, in sect. ODONTOTINUS the tetraploid V. setigerum has the same genomes duplicated that occur in the diploid. In the same section, V. dentatum var. pubescens and V. dentatum var. deamit are represented by octoploid forms with eight genomes duplicating the four genomes of the tetraploid. Viburnum carlesii (sect. LANTANA) and V. plicatum, V. plicatum {. mariesit, and V. plicatum f{. tomentosum (sect. PSsuDopULUS) are the only aneu- ploid species yet found in Viburnum. The forms with 2n = 20 and 2n = 22 can be considered to have developed from the 2m = 18 form which is the most common. The plants with 20 and 22 chromosomes have, respectively, one and two pairs of chromosomes not found in the 2” = 18 form, but at present the origin and relationship of these additional chromo- somes to the usual 18 in V. carlesii is unknown.

The geographical distribution of polyploids is much more restricted than that of diploids. Viburnum species from all the major centers of distribution, except Central and South America, are well represented in this study. Since the origin of most of the varieties, whether natural or by man’s selection, is uncertain, in many cases they cannot be assigned to a specific geographic area and are omitted in the following discussion. The greatest number of species studied is in the 18-chromosome group, and these are distributed over a wide geographic area. The diploid species include fifty from Asia, four from Europe, and twelve from North America, while the polyploids include eight species from Asia Minor, one from Europe, and six from North America.

From the foregoing it is obvious that the distribution of polyploids in Viburnum provides little evidence for one distinct center of origin of the genus. It is probable that polyploidy has evolved several to many times and in various places: eastern Asia, the Himalayan and Mediter-

170 JOURNAL OF THE ARNOLD ARBORETUM [VOL, XLIII

ranean regions and eastern North America. In eastern Asia occur all the species with the basic number of eight, along with a large number of diploids and a few polyploid forms with the basic number of nine. The Mediterranean species are few in number and belong to the group with a basic number of nine. The eastern North American species include the greatest number of octoploids, possibly indicating that this geographical niche has been more favorable for their evolution and establishment.

Only for certain sections does a relationship exist between the geographi- cal distribution and the taxonomic sections of the genus. Sections THyr- SOSMA and MEGALOTINUS are entirely of Asiatic distribution. Viburnum plicatum and its varieties, composing sect. PSEUDOPULUS, are native only to China and Japan. All species of sect. LENTAGO are limited to eastern North America. Representative species of sects. Tinus and LANTANA are distributed both in Asia and in Europe. Sections Oponrotinus and PSEUDOTINUS are represented both by North American and Asiatic species. The species of the sect. OpuLtus occur both in Europe and in North America.

At present, the cytological evidence suggests that Rehder’s sectional classification of Viburnum corresponds favorably with the natural rela- tionships. It is hoped that as this study is continued and expanded a more accurate evaluation of the classification can be achieved.

PROPOSED RESEARCH

Portions of the preceding discussion are based principally on inference which indicates where the problems lie and suggests methods of approach. Definite conclusions cannot be drawn until much additional research is completed. Therefore, the present study is basic both for plant breeding and for cytological studies to be continued in the genus Viburnum.

To the present, it has been impossible to secure the species native to Mexico and Central and South America, but by expedition or other- wise, it is hoped that these may become available for future study. Within these areas are many species of diverse form which, when secured from higher elevations, should prove hardy and noteworthy ornamentals in this and other latitudes. These species not only may provide additional genetical variability for interspecific hybridization, but also are repre- sentatives of one of the centers of diversity in the evolution of the genus.

Cytological studies, in addition to providing a basis for plant breeding, have provided a useful tool for taxonomists in classifying certain plant groups, and there is every indication that such information can likewise be useful in studies of Viburnum. It is anticipated that a sporocyte study and karyotype analysis, associated with genetical and taxonomic studies, will aid materially in revealing natural relationships which can be utilized in the classification of the genus Viburnum.

—

U.S. NationaL ARBORETUM, WASHINGTON 25, D.C.

1962 | EGOLF, CYTOLOGY OF VIBURNUM 171

BIBLIOGRAPHY

_

Bazscock, E. B. Cytogenetics and speciation in Crepis. ae pee in Genetics, pp. 161-206. Academic Press, Inc., New York,

BaLpwin, J. T., Jr. Kalanchoe: the genus and its Fete Am. Jour. Bot. 25: 572-579. 193

BEASLEY, J. O., & M. S. Brown. ‘The production of plants having an extra pair of chromosomes from species hybrids of cotton. Rec. Genet. Soc. Am. 12: 43. 1943

Bake, S. F. On the names of some species of Viburnum. Rhodora 20: 11- 15.

CLAUSEN, JENS. Stages in the evolution of plant species. 206 pp. Cornell Univ. Press, Ithaca, N.Y., 1951.

Concer, A. D. & L. M. Farrcuiip. A quick freeze method for making smear slides permanent. Stain Tech. 28: 281-283. 1953.

Dartincton, D. C. Chromosomes and plant breeding. 112 pp. Macmillan and Co., London, Z

. Recent ee in cytology. Ed. 2. 671 pp. J. and A. Churchill,

bo

Ww

al

On

oy

“I

8. Ltd., London,

9, ———. oe and the genetics of the centromere. Jour. Genet. 37: 341-364. 1939.

10: & L. F. La Cour. The handling of chromosomes. Ed. 2. 180 pp. George Allen and Unwin, Ltd., London, 1947.

ila & K. THER. The elements of genetics. 446 pp. Macmillan Co.,

New York, 1949.

12. DE Vos, F. The stem anatomy of some species of the Caprifoliaceae with reference to phylogeny and identification of the species. 81 pp. Ph.D thesis, Cornell University, 1951.

13. Dippet, L. Handbuch der Laubholzkunde Beschreibung der in Deutsch- land heimischen und im Freien kultivierten Baiime und Straucher 1: 172-200. 1889.

14. Ecotr, D. R. Cytological and interspecific okies ane in the genus Viburnum. 131 pp. Ph.D. thesis, Cornell University,

15. Fenc, Y. A. Observations sur la présence de centrosomes s U'asters présidant 4 la caryocinése dans un genre de Caprifoliacées: Lonicera. Le Botaniste 26: 3-20.

16. Fernatp, M. L. Another oilitier of additions to the flora of Virginia. Rhodora 43: 635-657. 1941

17. GoopspEeD, T. H. The genus Nicotiana. 536 pp. Chronica Botanica Co.,

altham, Mass., 1954.

18. Janaki AmMAL, E. K. Chromosomes and the species problem in the genus Viburnum. Curr. Sci. Bangalore 22: 4-6. 1953.

19. JoHANSEN, D. A. Plant eng 533 pp. McGraw-Hill Book Co., New York and London,

20. La Cour, L. Acetic-orcein a new a fixative for chromosomes. Stain Tech. 16: 169-174. 1941.

21. McAtTer, W. L. A review of the nearctic Viburnum. 125 pp. Chapel Hill, North Carolina, 1956.

22. Mever, J. R. Prefixing with paradichlorobenzene to facilitate chromo- some study. Stain Tech. 20: 121-125. 1945.

Naxal, T. Tentamen systematis Caprifoliacearum japonicarum (System-

wn oO

bo Ww

JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLII

atic treatment of sg i guages of Japan), Tokyo Imp. Univ. Jour. Coll. Sei. 42> ‘Art. 2; 139 pp. 1.

Classes, sate familiae, subfamiliae, tribus, genera nova quae attinent ad plantas Koreanas. Jour. Jap. Bot. 24: 8- 4: 1949,

. NAvASHIN, M. The dislocation hypothesis of evolution of chromosomal

numbers. Zeits. Ind. Abst. Ver. 63: 224-231. 1932.

O'Mara, J. G. Cytogenetic studies on Tyriticale. I. A method for de- termining the effects of individual Secale chromosomes in Triticum. Genetics 25: 401-408. 1940.

Poucques, M. L. pe. Etude caryologique de quelques Viburnum. Compt. Rend. Soc. Biol. Paris 141: 183-185.

———. Recherches caryologiques sur les Rubiales, Revue Gén. Bot. 56: 5-27, 74-138, 172-188. 1949.

REHDER, A. New species, varieties and combinations from the herbarium and the collections of the Arnold Arboretum. Jour. Arnold Arb. 5: 49-59. 1924.

—. Manual of oo trees and shrubs. Ed. 2. 996 pp. Macmillan

Co., New York,

A oc of cultivated trees and shrubs. 825 pp. Arnold

Arboretum, Harvard University, 1949.

Sax, K. & D. A. Kriss. Chromosomes and phylogeny in Caprifoliaceae. coe Arnold Arb. 11: 147-153. 1930,

SHARP, L. W. Introduction to cytology. Ed. 3. 567 pp. McGraw-Hill Book Co., New York and London, 1934.

SHERMAN, M. Karyotype evolution: a cytogenetic study of seven species and six interspecific hybrids of Crepis. Univ. of Calif. Publ. Bot. 18: 369-408. 1946.

SIMONET, M., & C. Miepzyrzecxi. Etude caryologique de quelques espéces arborescentes ou sarmenteuses d’ornement. Compt. Rend. Soc. Biol. Paris 111: 969-973, 1932.

STEARN, W. T. X Viburnum bodnantense, Caprifoliaceae. Bot. Mag. 167: t. 113. 1950.

STEBBINS, G. L., Jr. Variation and evolution in plants. 643 pp. Columbia Univ. Press, New Yor

Suciura, T. Studies on the chromosome numbers in higher plants with special reference to cytokinesis. I. Cytologia 7: 544-595. 1936.

Svenson, H. K. Plants of the ra United States. I. Viburnum dentatum. Rhodora 42: 1-6.

. Tuomas, J. L. The cytology of some cultivated species of Viburnum. 1961.

Jour. Acold Arb. 42: 157-164.

. Topcy, H. A. A cytological study of Crepis fuliginosa, Crepis neglecta

and their F, hybrid and its bearing on the mechanism of phylogenetic reduction in chromosome number. Jour. ees 45: 67-111. 1943.

. Wuire, M. J. D. Animal cytology and evolution. 454 pp. Cambridge

Univ. Press, Cambridge, 1954

. Wivxinson, A. M. The floral anatomy and morphology of some species

of Cornus and the Caprifoliaceae. 148 pp. Ph.D. thesis, Cornell

University, 1945.

: | anatomy and morphology of some species of the genus

Viburnum of the Caprifoliaceae. Am. Jour. Bot. 35: 455-465. 1948. Floral anatomy and morphology of Triostewm and of the Capri-

foliacéae in general. /bid. 36: 481-489. 1949.

1962] BRIZICKY, SIMAROUBACEAE AND BURSERACEAE 173

THE GENERA OF SIMAROUBACEAE AND BURSERACEAE IN THE SOUTHEASTERN UNITED STATES *

GEORGE K. BRIZICKY

SIMAROUBACEAE A. P. de Candolle, Ann. Mus. Hist. Nat. Paris 17: 422. 1811, “Simarubeae.” (QuassIA FAMILY)

Trees and shrubs [rarely subshrubs], often with bitter bark. Leaves usually alternate, never glandular-punctate, pinnately compound or simple [rarely rudimentary], with predominantly entire margins, exstipulate [or eae Flowers usually small to minute, greenish or variously colored,

ogynous, regular, bisexual or Griseauel by abortion, or both, with biseriate, [3]4— —6[8]-, usually 5-merous perianth, in terminal and/or axillary many- or few-flowered cymose panicles or racemes, rarely solitary. Sepals distinct or united, imbricate or valvate. Petals distinct, imbricate or valvate, rarely wanting, Stamens distinct, as many or twice as many as the petals [rarely more numerous], usually inserted at base of an intra- staminal disc, rudimentary or wanting in @ flowers; filaments usually slender, sometimes appendaged at base on the ventral (adaxial) side; anthers usually versatile, 2- or rarely 4-locular at anthesis, introrse, fone tudinally dehiscent. Intrastaminal nectariferous disc usually present, an- nular, cupular, cushion-like [to columnar], mostly lobed or crenate, rarely obscure or wanting. Gynoecium usually inserted on or encircled at base by the disc, sessile [or rarely raised on a gynophore], 2-6[8]-carpellate, apocarpous to syncarpous, rudimentary or wanting in ¢ flowers; stigmas distinct or united; styles basal, lateral or apical, distinct or partially to completely united; ovaries | -carpellate and -locular (gynoecium apocar- pous) or 2—3[4]- -carpellate and -locular with axile placentae (gynoecium syncarpous); ovules anatropous or rarely orthotropous to campylotropous, usually epitropous, very rarely apotropous, 2- or rarely 1- integumented, with a thick nucellus, 1 or 2 (collateral or superposed) [very rarely more

1 Prepared for a generic flora of the southeastern United States, a joint project of the Arnold Arboretum and the Gray Herbarium of Harvard University which has been made possible through the support of Ge eorge R. Cooley and the National Science Foundation. This treatment follows the pattern established in the first paper in the series (Jour. Arnold Arb. 39: 296-346. 1958) and continued through the seventeen subsequent papers in volumes 40-43 (1959-1962). It should be repeated that the area covered by this work is bounded by and includes North Carolina, Ten- nessee, Arkansas, and Louisiana. The descriptions are based primarily on the plants of this area, with any supplementary material in brackets. References which the

The author is indebted to Dr. Carroll E. Wood, Jr., for his criticism and valuable suggestions, and to Mrs. Gordon W. Dillon, for her careful help in the preparation of the manuscript.

174 JOURNAL OF THE ARNOLD ARBORETUM [ VOL, XLII

numerous| in a locule, pendulous from the top or ascendant from the base of the latter. Fruit apocarpous, of 2—6|8] one-carpellate and -locular drupes, [berries] or samaras, or a syncarpous, 2—3|4|-carpellate and -locular (occasionally 1-locular by abortion) berry, samaroid capsule [drupe or schizocarp| with 1-seeded locules. Seeds with membranaceous or leathery testa; endosperm scanty or wanting; embryo rather large, straight or rarely curved, with narrow, mostly fleshy, plano-convex or flat cotyledons and usually a very short radicle. (Including Surianaceae.) Type GENUS: Simarouba Aublet.

A pantropical family of about 30 genera and 200 species, at least two species in temperate eastern Asia. Four of twelve genera indigenous to the New World occur in subtropical Florida. One species of Adanthus is also naturalized in our area.

The rather heterogeneous family includes six sharply delimited sub- families. Evidence from wood anatomy supports the segregation of at east some as distinct families, but, since data from gross morphology appear to be inconclusive and those from floral anatomy, embryology, and cytogenetics of the family are fragmentary, it seems preferable to retain the family in the larger sense.

Simaroubaceae are closely related to Rutaceae and Burseraceae, differ- ing from the former mainly in the absence of multicellular glands (secre- tory cavities) with aromatic oils in leaves, axes, and floral parts, and from the latter in the absence of schizogenous resin ducts in the bark.

Wood and/or bark of some species (e.g., Picrasma excelsa (Sw.) Planch., “Jamaica Quassia”’; Quassia amara L., “Surinam Quassia”) yield a bitter principle employed as tonics and vermifuges, as an insecticide, and some- times as a substitute for hops in brewing. Some genera produce timber of local importance; a few species are ornamentals.

REFERENCES:

Boas, B. Beitrage zur Anatomie und Systematik der Simarubaceen. Thesis, 58 pp. Dresden. 1912. {Also in Beih. Bot. Centralbl. 29(1): 303-356. 1913.]

Cronourist, A. Studies in Simaroubaceae—IV. Résumé of the American genera. Brittonia 5: 128-147. 1944. [Suriana, Recchia, Alvaradoa. Hola- cantha, Picrasma, Picrolemma, and Quassia are treated in detail. |

DEsat, S. Cy tology of Rutaceae ane Simarubaceae. Cytologia 25: we 35. 1960. | Ailanthus altissima, 2n = 80, Quassia amara, 2n = 36. |

ENGLER, A. Simarubaceae. . Paanzentani. III. 4: 202-230. 1897; ibid. ed. 2. 19a: 359-405. 1931.

Harti, D. Die Ubereinstimmungen des Endokarps der Simarubaceen, Rutaceen, und Leguminosen. Beitr. Biol. Pil. 34: 453-455. 1958

Japin, F. Contribution a l'étude des Simarubacées. in. a Nat. Bot. VIII. 13: 201-304. pl. 1. 1901.

——-—. Essai de classification des Simarubacées basée sur les caracteres anato- miques. Compt. Rend. Assoc. Fr. Avanc. Sci. 30(2): 477-483. 1902.* [See review in Bull. Soc. Bot. Fr. 49: 223, 224. 1902. ]

Narr, N. C., & T. S. Josepu. Floral morphology and embryology of Samadera indica. Bot. Gaz. 119: 104-115. 1957.

1962 | BRIZICKY, SIMAROUBACEAE AND BURSERACEAE 175

& R. K. JosH1. Floral morphology of some members of the Simarou- baceae. Bot. Gaz. 120: 88-99. 1958. [Five spp. of Adanthus, including A, altissima ; Picrasma quassioides (D. Don) Benn.. and Brucea sumatrana Ox as amarissima (Lour.) Desv. ex Gomes). |

& N. P. SukUMARAN. Floral morphology and embryology of Brucea amarissima. Bot. Gaz. 121: 175— ee NARAYANA, L. L. Embryology of two Siniaroubeccae Curr. Sci. Bangalore 26:

323, 324. 1957.* [Includes ee excelsa. | & M. SAYEEDUDDIN. Floral anatomy of Simarubaceae — I. Jour. Indian Bot. Soc. 37: 517-522. 1958. gee amara L. and Ailanthus excelsa.| Recorp, S. J.. & R. W. Hess. Timbers of the New World. 640 pp. ey 59. New Haven. 1943. [Simaroubaceae, re 514; Gane S2o22.4 SARGENT. C. S. Manual of the trees of North ie (exclusive of Mexico). ed. 2.897 pp. 1922. [Simaroubaceae. 641- SAUNDERS, E. R. Floral morphology. Vol. 2. New York. 1940. [Simaroubaceae. 195-197. SMALL, J. K. Simaroubaceae. N. Am. Fl. 25: 227-239. 1911. [Suriana placed in a Separate family. | WEBBER, I. E. Sy stematic anatomy of the woods of the Simarubaceae. Am. Jour. Bot. 23: 577-587. 1936. WEST. Ee. & L. E. ARNOLD. The native trees of Florida. 212 pp. Gainesville. 1946. | Simaroubaceae, 101-104. | Wicer. J. Embrvological studies on the families Buxaceae. Meliaceae. Simaru- baceae and Burseraceae. Thesis. Lund. 1935.* [See Mauritzon’s criticism. 935: 490-502. 1935; Wigers reply. Bot. Not. 1936: 585-539. 1936. |

KEY TO THE GENERA OF SIMAROUBACEAE

General characters: trees or shrubs with alternate. predominantly pinnate leaves; flowers small, uni- or eee or both; perianth st usually 5-merous ;

stamens 5 or 10, distinct; intrastaminal aie usually present; gynoecium apo- or synearpous, 2-5- parpelluter ovules 1 or 2 in each eee

A. Leaves simple. small (not over 5 cm. long) and narrow: flowers bisexual. in short. terminal, few-flowered. corvmb-like panicles. rarely solitary; stamens 10. the 5 antipetalous usually ag or rudimentary: intrastaminal disc wanting; carpels distinct; fruit of r fewer achene-like drupelets: strand Dlaniwor subtropical Plordac ss exces ee eety Fee eee ace te 1. Suriana,.

A. Leaves pinnately compound; flowers usually unisexual (occasionally also bi- sexual) in large, many-flowered, complex terminal panicles or racemes; in- trastaminal disc distinct; carpels connate. sometimes only by the styles. rudimentary or wanting in 4 flowers; fruits various, never achene-like drupelets B. Stamens 10 in 6 flowers. much reduced or wanting in @ flowers; carpels

4—6, usually 5; styles lateral. connate; ovaries distinct, connivent; fruits

aggregate.

C. Leaves usually even-pinnate. persistent; leaflets alternate. + leathery. entire; staminal filaments appendaged at base; fruit of 5 or fewer olive-shaped drupes about 2 cm. long: hammocks of subtropical Florida. 2. Simarouba.

176 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIII

C. Leaves odd-pinnate, deciduous; leaflets usually opposite, thin, with a few glanduliferous teeth or lobules near the base; staminal filaments without appendages; fruit of 5 or fewer oblong samaras 3-5 cm. long, each with a flattened seed at the middle of the membranaceous wing; northern Florida and northward. .................. 3. Ailanthus.

B. Stamens 5 and alternate with the sepals in 6 flowers, reduced or wanting in 2 flowers; carpels 2 or 3; styles terminal, connate or distinct; ovary compound; ieives odd- epinnates hammocks of subtropical Florida.

D. Leaflets 5-9, 5-12 cm. long; ¢ and @ flowers with linear-lanceolate petals, in terminal panicles with slender racemose or spikelike branches; stigma 2-lobed; style one, very short and stout; fruit a 1- or 2- joe EOE: 4 cyan waewae eee epider ene ee sea e ook 4. Picramnia.

D. Leaflets 19-51, 1-3 cm. long; 4 flowers with linear-filamentous ae the ? flowers without petals, in terminal and axillary racemes; stigmas 3; styles 3, distinct, subulate, short; fruit a flat, papery, 1- seeded sama- roid manele: be kde whe 45 odor hateueees cee pans 5. Alvaradoa.

Subfam. SURIANOIDEAE Engler 1. Suriana Linnaeus, Sp. Pl. 1: 284. 1753; Gen. Pl. ed. 5. 137. 1754.

Shrubs or small trees 1-8 m. high with most parts more or less densely pubescent with simple and glandular hairs. Leaves sessile, simple, entire, narrow (seldom over 4 cm. long), thickish, with indistinct lateral veins and centric mesophyll. Flowers relatively small, bisexual, obdiplostemonous, with 2-bracteolate pedicels, in terminal, few-flowered, corymb-like cymose panicles (sometimes reduced to a solitary flower). Sepals 5, narrow (about 6-10 mm. long), connate at base, imbricate, persistent. Petals 5, yellow, broad, clawed, about as long as the sepals, imbricate. Stamens 10, distinct, the antipetalous usually antherless or rudimentary; filaments subulate, hairy below; anthers subcircular in outline, emarginate at both ends, in- trorse, 2-locular at anthesis; pollen grains medium sized, suboblate, 3-col- pate, finely striate. Intrastaminal disc apparently wanting. Gynoecium apocarpous, 5-carpellate; stigmas distinct, capitellate; styles distinct, fili- form, nearly basal; ovaries 1-locular, sessile; ovules 2 (collateral), ascend- ing from the base of the ovary locule, 1-integumented, orthotropous, be- coming campylotropous after fertilization. Fruits 1-seeded achene-like drupelets (not over 5 mm. long) with thin flesh and crustaceous endocarp (stone), obovoid-subspherical, pubescent. Seeds broadly obovoid, slightly flattened laterally, endospermless, with a thin membranaceous testa; em- bryo horseshoe shaped, with flat, oblong cotyledons and an elongate radicle descendent toward the micropyle which is side by side and in contact with the hilum. Type species: S. maritima L. (Named in honor of Joseph Donat Surian, a French physician and botanist, “Dioscorides Americanus futurus” |[Plumier, Nov. Pl. Am. Gen. 37. 1703], companion and _ col- laborator of Plumier.) — BAY CEDAR.

A monotypic genus represented by Suriana maritima, widely but sporadi- cally distributed on the seashores of the New and Old World tropics, but

1962] BRIZICKY, SIMAROUBACEAE AND BURSERACEAE Liz

apparently absent from the Pacific coasts of the Americas, the islands of the Central Pacific, and the Atlantic coast of Africa. It occurs in the coastal sand dunes and hammocks of the Florida Keys and of peninsular Florida northward to Brevard and Pinellas counties, and beyond our area in Bermuda, the West Indies, and from Yucatan south to Brazil. Mill- spaugh (Publ. Field Mus. Bot. 2: 241. 1907) believed the fruits (stones) to be carried on the feet of sea birds, but according to Guppy they “could readily be carried in the crevices of floating logs, or in the cavities of float- ing pumice, such as is stranded on the beaches of tropical regions all over the world. But it is on their great floating powers, which fit them for dis- persal by currents, that we must mainly rely.” Buoyancy of the stones is produced by the unfilled space in the fruit locule (Guppy, Schimper). Record & Hess (1943, p. 521), Small (Man. 761. 1933), Wilson (1911), and a few others thought the genus to represent a distinct, monotypic family Surianaceae; but, if a separate family is recognized, it should also include the closely related genera Cadellia F. Muell., Guilfoylia F. Muell., and perhaps also Recchia Sessé & Moc. ex DC. (Rigiostachys Planch.). The branches exude some kind of ‘“‘manna.”

REFERENCES:

See also under family references, Boas, ENGLER (1897, pp. 208, 209; 1931, pp. ae 368), JADIN, REcorp & Hess, WEBBER (pp. 577-579, 586, 587), and Wi

IGE

ee H. B. Plants, seeds, and currents in the West vee and Azores. 531 pp., 3 maps. bendon: 1917. [Fruit dispersal, 239-24

Rav, M. A. An embryological study of Suriana maritima a Proc. Indian Acad. Sci. B. 11: 100-106.

cases A. F. W. Die Indo- IMalayische Strandflora. (Botanische Mittheilungen

us den Tropen. 3.) 204 pp., pls. 1-7. Jena. 1891. [Floating power of

fe 163, 165. ]

SOLEREDER, H. Ueber die systematische Stellung der Gattung Rigiostachys, zugleich ein Beitrag zur nadheren Kenntnis der Simarubeae-Surianoideae. Jn:

kanische Gattung Rigiostachys. Verh. Bot. Ver. Brandenburg 47: 41-61. 1905. [Includes morphology of ovules, fruits, and seeds. | WILSON, P. Surianaceae. N. Am. Fl. 25: 225. 1911.

Subfam. SIMAROUBOIDEAE Tribe SIMAROUBEAE

2. Simarouba Aublet, Hist. Pl. Guiane Fr. 2: 859. 1775.7 Trees or large shrubs with bitter bark and wood. Leaves even-pinnate [or odd-pinnate]|, persistent; leaflets 3-21, mostly 8-16, alternate, rarely opposite, entire, -_ coriaceous, shortly petiolulate, Plants polygamo-dioe- cious or dioecious Flowers small, unisexual or uni- and bisexual, obdiplo- ? Aublet’s generic name should be conserved against its earlier homonym, Simaruba

Boehmer, 1760 [nom. rejic. vs. Bursera Jacquin ex Linnaeus, 1762, nom. cons.]. See A. A. Bullock, Taxon 8: 199. 1959.

178 JOURNAL OF THE ARNOLD ARBORETUM [ VOL, XLIII

stemonous, in large terminal and axillary complex panicles. Sepals 4—6, usually 5, rather small, connate at least at base, imbricate. Petals 4—6, usually 5, yellowish to yellow [or whitish], much longer than the sepals, imbricate. Stamens 8-12, usually 10, as long as the petals, reduced, rudi- mentary or absent in @ flowers; filaments subulate, broadened toward base, with an adaxial, liguliform, pubescent appendage at base; anthers oblong in outline, slightly emarginate, versatile, 2-locular at anthesis; pol- len grains small, subprolate, 3-colpate. Intrastaminal disc + cushion- like, depressed in 6 flowers. Gynoecium (4)5(6)-carpellate, sessile on the disc, partially syncarpous, rudimentary or wanting in é flowers; stig- mas 4—6, usually 5, long, slender, and divergent in ¢ flowers, short, lobe- like in bisexual flowers; styles connate into a short column: ovaries dis- tinct, 1-locular, cohering |or perhaps sometimes weakly connate] by their ventral sutures and thus resembling a compound deeply lobed ovary; ovules solitary in each carpel, pendulous from near the top of the inner angle of the locule, anatropous and epitropous, 2-integumented. Drupes 5 or fewer from each flower, ellipsoid to obovoid, slightly laterally compressed, usually 2-ridged, with thin flesh and a crustaceous stone. Seeds endosperm- less; testa thin, membranaceous; embryo straight, with fleshy, plano-convex cotyledons and a very short, superior radicle macy included between the cotyledons (retracted). Typr species: S. amara Aublet. (Name derived from the Carib Indian name of the type species in French Guiana.)

A tropical American genus of about six species. Simarouba glauca DC., paradise tree, occurs in coastal hammocks on the Florida Keys and in southern peninsular Florida (Dade, Broward, and Palm Beach counties). It is widely distributed in the West Indies and in Central America from Costa Rica to southern Mexico. The ellipsoidal fruits, about 2 cm. long, are scarlet, changing to dark purple or black when ripe. A white-fruited form has been reported from El Salvador

The timber of some species, especially. the South American Simarouba amara, is of commercial importance. The bitter bark of roots of S. amara and S. glauca is said to be efficient against diarrhea and post-dysenteric disorders. Seeds of S. glauca yield about 60 per cent of edible oil and a crystalline glycoside, glaucorubin, which reportedly has amoebicidal prop- erties and is now being introduced in most of the major tropical areas under the trade name of ‘‘Glaumeba.”’

The genus is a representative of a very natural group, Simarouboideae— Simaroubeae, and is closely related to the tropical South American Simaba Aubl. and Quassia L. (monotypic, according to Cronquist), as well as to the tropical African Odyendea (Pierre) Engl.

REFERENCES:

See also under family references, ENGLER (1897, pp. 211, 213; 1931, pp. 372- 347), Recorp & Hess (pp. 513, re SARGENT (pp. 643, 644), SMALL (pp. 227— 229), and West & Rete (p ).

ArMourR, R. P. spmpeteete on Simarouba glauca DC. in El Salvador. Econ.

1962 | BRIZICKY, SIMAROUBACEAE AND BURSERACEAE 179

Bot. 13: 41-66. 1959. [Commercial importance and cultivation as a vege- table oil crop. |

Cronguist, A. Studies in the Simaroubaceae — II. The genus Simarouba. Bull, Torrey Bot. Club 71: 226-234. 1944.

Kryn, J. M. Simarouba, paradise-tree, marupa, Simarouba spp., family: Simarou- baceae. U. S. Forest Serv. Forest Prod. Lab. Rep. R. 1956. 10 pp. 1953.*

KUKACHKA, B. F. Marupa: Simarouba amara Aubl., Simaroubaceae. U. S. Forest Serv. Forest Prod. Lab. Rep. 1856. 4 pp. 1960.*

SARGENT, C. S. Simaruba. Silva N. Am. 1: 89-92. pls. 38, 39. 1891.

SoLA, F. pE. Notes on the aceituno tree (Simaruba glauca DC.) and its adapta- tion as a vegetable oil crop. Ceiba 4: 351-358. 1956.

Tribe PIcRASMEAE Engler

3. Ailanthus Desfontaines, Mém. Acad. Sci. Paris 1786: 265. 1788, nom, cons

Trees, sometimes strong smelling. Leaves odd-pinnate, large, deciduous; leaflets opposite or alternate, thin, with a few glanduliferous, blunt teeth or lobules near base [or entire or rarely coarsely toothed to lobed through- out], petiolulate. Plants polygamo-dioecious. Flowers small, uni- and bisexual, obdiplostemonous, of unpleasant odor, pediceled, in large, terminal panicles. Sepals 5 (6), connate in the lower third or higher, imbricate. Petals 5 (6), longer than the sepals, greenish or yellowish, induplicate- valvate in aestivation. Stamens inserted at base of an intrastaminal disc, 10 (12) in 6 flowers, sometimes fewer in bisexual, much reduced and sterile or absent in @ flowers; filaments subulate, without appendages; an- thers oblong or oblong-ovate in outline, versatile, 2-locular at anthesis, subintrorsely dehiscent; pollen grains small to medium sized, subspheroidal, 3-colpate, reticulate. Intrastaminal disc annular, thick, usually deeply 5(6)- or 10(12)-lobed, or crenate, of receptacular origin. Gynoecium 5(6)-carpellate [rarely 3-carpellate], partially syncarpous [or apocar- pous|, sessile within the disc, rudimentary or absent in ¢ flowers; stigmas distinct, tongue-shaped and divergent or + capitellate; styles lateral, fili- form, usually connate [or distinct]; ovaries 1-carpellate and ocular, much compressed laterally, distinct but + cohering by their sutures and resembling a compound, deeply lobed ovary; ovules anatropous, 2-integu- mented, solitary in each carpel, hanging from below the insertion point of the style. Fruit of (1)2—5(6) distinct, oblong to oblong-elliptic samaras, each with a flattened seed at the middle of the thin, veiny, adaxially emargi- nate wing. Seeds lenticular, with thin, membranaceous testa and sparse, fleshy endosperm; embryo with flat, obovate to orbicular cotyledons and a short, superior radicle. Type species: A. glandulosa Desf. (= A. altis- sima (Mill.) Swingle). (Name derived from the Moluccan name for A. integrifolia Lam., ailanto, Sea in allusion to the height of the trees.) — TREE-OF-HEAVE

A primarily tropical genus of about 15 species, distributed in eastern

180 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLII

and southern Asia, the Philippines, Malaysia, Melanesia (east to the Solo- mon Islands), and northeastern Australia, extending into the Temperate Zone to northeastern China and Korea (to about 40° N. Lat.). The north- ernmost species, Ailanthus altissima (Mill.) Swingle, Chinese tree- of- heaven, stinkweed, 2” = 80, introduced into North America in 1784 as an ornamental tree, has become naturalized in native woodlands of the south- eastern United States as far south as northern Florida. Northward and westward beyond our area it seems to be primarily a ‘‘weed” tree of cities, but the exact extent of naturalization needs to be more carefully recorded. An irregularly branched, rapidly growing tree which reproduces by seeds, stump sprouts, and root suckers, it becomes a weed difficult to eradicate. However, it has sometimes been used for fastening sterile, sliding declivities and for aforestation of bare hill and mountain slopes (Alps, Caucasus) and of grasslands (Asia Minor). The leaves contain a substance toxic to seed- lings of many species of gymnosperms and angiosperms. “It is possible that the toxic substance is washed from the | fallen] leaves by the rain and influences the composition of plant communities” (Mergen).

Small flies and beetles have been recorded as pollinators in Adanthus altissima, but bees must participate since it is regarded as an undesirable honey plant. Wind pollination is also possible (cf. Wodehouse). Floral anatomy has been studied in this and a few other species, embryology only in the Indian A. excelsa Roxb., 2n = 62. Apparently only two chromosome counts have been recorded.

Wood of Ailanthus altissima and the other species appears to be of relatively little value, being used mostly for fuel, occasionally for cabinet work, musical instruments, etc. It also seems to be fit for paper pulp. Leaves and bark of A. altissima, and bark of a few other species are used locally in Asia as anthelmintics or antidysenterics.

The genus occupies a somewhat isolated position in the tribe Picrasmeae of Simarouboideae, representing a subtribe of its own.

REFERENCES:

The large number of references. has been reduced here primarily to those either of general interest or dealing > eormaeay| with the southeastern United States. Under family references see ENGLER (1931, pp. 390-393), Desat (p. 39), Narr & JosHI, NARAYANA, NARAYANA & SAYEEDUDDIN, and SMALL (p. 234). ANbERSON, E. The tree of heaven Ailanthus altissima 1. A blessing and a curse. Missouri Bot. Gard. Bull. 49: 105-107. 1,

AnpREAE, E. Uber abnorme Wurzelanschwellungen bei Adlanthus glandulosa. Thesis, 34 pp., 3 pls. Erlangen. 1894.

CampREDON, J. Etudes des propriétés physiques et mécaniques de quelques bois exotiques. III. Le bois d’Ailante (Ailanthus dees Desf.). Ann. Ec. Natl. Eaux Foréts Nancy 5: 211-217. 1934.*

Davies, P. A. Leaf arrangements in Ailanthus altissima. Am. Jour. Bot. 24: 401-407. 1937. [See also ibid. 26: 67-74. 9

; aie glands in Ailanthus altissima. Trans. Ky. Acad. Sci. 11: 12-16.

: on glands on Ailanthus altissima, Ibid. 12: 31-33. 1945.*

1962 | BRIZICKY, SIMAROUBACEAE AND BURSERACEAE 181

& E. Bennett. Abnormal branching in Ailanthus. Jour. Hered. 20: 348, 349. 1929,

& E. W. TuHeiss. Factors affecting the method of branching in Ailan- thus altissima. Bull. Torrey Bot. Club 64: 229-233. 1937. [See also DAvIEs, Bull. Torrey Bot. Club 63: 139-146. pl. 5. 1936.]

Inyick, J. S., & E. F. Brouse. The Ailanthus tree in Pennsylvania. Penn. Dep Forests Waters Bull. 38. 29 pp. 1926. [Includes biology and prospects ie the species. |

Kriz, V., M. CHLEBEK, & M. PEKar. Ailanthus [altissima] from the view- point of breeding. (In Czech.) Lesn. Prace 36: 116-118. 1957.*

MerenpI, A. Ailanthus: cellulose plant. (In Italian.) Gior. Agr. Roma 66: 137. 1956.* [See also A. pet Lunco, Terra Sole 122: 345-349. 1952.*]|

MERGEN, F. A toxic principle in the leaves of Ailanthus. Bot. Gaz. 121: 32-36. 1959. [A. altissima. |

Miter, R. Zur Anatomie der Ailanthus-Rinden. Pharm. Praxis 7: 261-263. 1908.* [For review see Bot. Jahresb. 36(1): 473, 474. 1908.]

Petaj, V. Die extrafloralen Nektarien auf den Bittern as Ailanthus glandu- losa. (In Croatian.) Rad Jugosl. Akad. Znan. Umjet. 215: 59-81. 1916.* [For review see Bot. Centralbl. 137: 385. 1918. ]

SayA, I. Ferite su fusto di Adanthus glandulosa Desf. e reazioni di gemme dormienti. Nuovo Gior. Bot. Ital. II. 64: 680-682. 1957. [Wounds on the stem of A. altissima and reactions of dormant buds. |

SWINGLE, W. T. The early European history and the botanical name of the

e of heaven, Ailanthus altissima. Jour. Wash. Acad. Sci. 6: 490-498. 1916.

TiEGHEM, P. van. Ailante et Pongéle. Ann. Sci. Nat. Bot. IX. 4: 272-280. 1906. [Segregation of Pongelion gi from Ailanthus. |

WopeHousE, R. P. Hayfever plants. xx + 245 pp. Chronica Botanica. Wal- tham, Mass. 1945. [Azlanthus en 115, 116.]

Subfam. PICRAMNIOIDEAE Engler

4. Picramnia Swartz, Prodr. Veg. Ind. Occ. 2, 27. 1788, nom. cons.

Trees or shrubs, usually with slender, curving branches, the bark and wood often very bitter. Leaves odd-pinnate, persistent; leaflets [3]5—9[21], opposite or alternate, entire, chartaceous to + coriaceous, petioluled. Plants dioecious [or occasionally polygamous]. Flowers minute, unisexual [occasionally also bisexual]. Inflorescences paniculate, terminal [some- times opposite the leaves], with slender, raceme- or spikelike branches. Sepals 5 [3 or 4], connate from %4 to % of their length, imbricate. Petals 5 [3 or 4, or wanting|, narrowly linear-lanceolate [or lanceolate], as long [or twice as long] as the sepals, imbricate. Stamens as many as and oppo- site the petals, inserted below and between the lobes of a low intrastaminal disc, reduced to staminodia in @ flowers; filaments subulate; anthers basi- fixed, almost globular, with thick connective, introrse, 4-locular at anthesis. Gynoecium 2(3)-carpellate, syncarpous, rudimentary in 6 flowers; stigma deeply 2-lobed, the lobes thick, divergent; style very short and stout, usu- ally inconspicuous; ovary sessile, 2(3)-locular; ovules anatropous, epitro- pous, 2-integumented, pendulous from near the top of the carpels, 2 in

182 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIII

each locule. Fruit a rather juicy, subglobular, ellipsoidal or obovoid berry, (3)2- or (by abortion) 1-locular, with 1-seeded locules. Seed ovoid, often plano-convex, filling the cavity of the locule, the testa membranaceous, ad- nate to the undifferentiated embryo; endosperm wanting. (Pseudobrasilium Adans., 1763; Tariri Aubl., 1775; nom. rejic.) Type species: P. antidesma Swartz. (Name derived from Greek, pikros, bitter, and thamnos, shrub, in allusion to the bitterness of the vegetative parts.) — BITTERBUSH.

A genus of about 40 species of tropical American distribution. Picramnia pentandra Sw., a shrub or small, slender tree, of the West Indies, Colombia, and Venezuela, occurs in our area in the coastal hammocks of southeastern peninsular Florida (Dade County) and on the Florida Keys. Berries of this species are olive-shaped, 9-15 mm. long, red, turning black when fully ripe; the seeds are light brown and lustrous. Leaves, bark, and roots of this and other species have been used locally in the tropics as febrifuges.

The genus, which represents a subfamily of its own, is in need of a re- vision.

REFERENCES: Under the family references see ENGLER (1931, pp. 402, 403), Recorp & Hess, SARGENT, SMALL, and West & ARNOLD (p. 101). RaADLKOFER, L. Ueber die Gliederung der Familie der Sapindaceen. Sitz-ber. Akad. Wiss. Miinchen IT. 20: 105-370. 1890. | Picramnia, relationship with Alvaradoa, 139-143. |

Subfam. ALVARADOIDEAE Engler 5. Alvaradoa Liebmann, Vid. Medd. Nat. For. Kjgbenh. 1853: 100.

Shrubs or trees up to 15 m. high, with slender, terete, pubescent | or glabrous| branchlets and bitter bark. Leaves odd-pinnate, many-foliolate, crowded at the end of branches, apparently persistent; leaflets alternate, small (not over 5 cm. long), thin and firm [or leathery], entire, petiolu- late. Plants dioecious. Flowers minute, unisexual, in slender, many- flowered, axillary or terminal racemes. Sepals 5, usually distinct in @ flowers and variously connate from near the base to the half of their length in 4 flowers, valvate. Petals 5, linear-filamentous, present [or absent| in ¢ flowers, wanting in @ flowers. Stamens 5, alternate with the sepals, inserted below and between the lobes of the disc, wanting in 9 flowers; filaments filiform, hairy in the lower part; anthers basifixed, ob- long in outline, with a conspicuous, almost orbicular, swollen connective and introrse anther-halves, 4-locular at anthesis; pollen grains small, prolate- subspherical, 3-colpate. Intrastaminal disc thickish, deeply 5-lobed in é flowers, thin and scarcely lobed in @ flowers. Gynoecium 3-carpellate (but only 1 carpel fertile), syncarpous, sessile on the disc, wanting in é flowers; stigmas small, simple; styles 3, distinct, sibulate, short, re- curved; ovary densely villous, flattened, obtusely triangular in cross sec-

1962 | BRIZICKY, SIMAROUBACEAE AND BURSERACEAE 183

tion, imperfectly 2- or 3-locular on account of 2 incomplete partitions (apparently deeply intruded parietal placentae) demarcating a sole fertile locule in the obtuse angle of the ovary; ovules 2 in the fertile locule, basal, ascendent, anatropous, apotropous, 2-integumented. Fruit a compressed, 2|3]-winged, + papery samaroid capsule [or leathery samara], crowned by remnants of the styles, appearing 3- or 1-locular, with 1 seed in the lower half. Seeds + terete, narrowly ellipsoidal (the shape of rice grains) [or rather compressed and broadly elliptical in outline]; testa mem- branaceous; endosperm wanting; embryo straight, with plano-convex [or flat], fleshy cotyledons and a short, inferior radicle. Typr SPECIES: A. amorphoides Liebm. (Named in commemoration of Pedro de Alvarado, one of the chief aides of Hernando Cortez in the conquest of Mexico.)

A genus of about five species, the range disjunct, including southern Florida, the West Indies, Mexico, and Central America, Bolivia, and Argentina. Alvaradoa amorphoides occurs in a few hammocks in southern peninsular Florida (Dade County) and the Florida Keys and in the West Indies, Mexico, and Central America.

Although the genus was originally placed with the Sapindaceae, Radl- kofer showed Alvaradoa to be simaroubaceous with the closest relation- ship to Picramnia. Engler, however, placed the two in separate unigeneric subfamilies.

REFERENCES:

See also under family references, Cronqutst (pp. 132-137), ENGLER (1931, p. 404), Recorp & Hess (p. 510), SARGENT (pp. 644, 645), SMALL, and WEsT & ARNOLD (p. 104).

RApDLKOFER, L. Ueber die Gliederung der Familie der Sapindaceen. Sitz-ber. Akad. Wiss. Munchen II. 20: 105-370. 1890. [Alvaradoa, morphology of the flowers and fruits, relationship with Picramnia, 139-143.]|

BURSERACEAE Kunth, Ann. Sci. Nat. 2: 346. 1824. (TorcHwoop FaMILy)

Trees or shrubs, the inner bark with resin ducts. Leaves alternate, usually once pinnate, deciduous [or persistent], usually exstipulate. Flowers small, hypogynous, regular, apopetalous, usually unisexual by abortion, 3—5-merous, in axillary |or terminal} cymose panicles. Plants mostly dioecious. Stamens 6—10, usually distinct, inserted below an intra- staminal [rarely extrastaminal] nectariferous disc; anthers versatile, in- trorse, longitudinally dehiscent, sterile in @ flowers. Gynoecium 3[2-5]- carpellate,. syncarpous, rudimentary or wanting in 4 flowers; ovary 3|2—5]|-locular, with axile placentae; ovules anatropous, epitropous, 2 in each locule. Fruit usually drupaceous, with + dry [or fleshy] exo- and

without endosperm, solitary in each locule; embryo straight [or curved],

184 JOURNAL OF THE ARNOLD ARBORETUM [ VOL, XLIII

cotyledons contortuplicate [or flat], aes + deeply lobed, radicle superior. TyPrE GENUS: Bursera Jacq. ex

A pantropical family of 15 or 16 genera and about 600 species. Six genera (including Bursera and Protium) with about 200 species occur in the New World. The family is subdivided into the tribes Protieae Engl., Bursereae [H. J. Lam] (Boswellieae Engl.), and Canarieae Engl. The subdivisions, although based exclusively upon the structure of the fruits, seem to be rather natural ones.

Burseraceae are closely allied with Rutaceae, Simaroubaceae, and Meli- aceae. A close relationship to Anacardiaceae, proposed by some taxono- mists (e.g., Radlkofer), is supported by evidence from anatomy and palynology.

Resinous substances obtained from some species are of economic sig- nificance. The most important of these are myrrh, from Commiphora molmol Engl. and C. abyssinica (Berg) Engl., and olibanum or frankin- cense, from species of Boswellia, especially B. Carteri Birdw. Myrrh is used in perfumery, and both are used in medicine and as incense.

REFERENCES: See also RADLKOFER under Alvaradoa (Simaroubaceae).

ENGLER, A. Burseraceae et Anacardiaceae. 71: CANDOLLE, A. & C. pe. Monogr. Phaner. 4: 1-573. pls. 1-15. 1883. [Burseraceae, 1-169. pls. 1-3.]

Burseraceae. Nat. Pflanzenfam. ed. 2. 19a: 405-456. 1931.

GUILLAUMIN, A. Répartition géographique et biologie des Burséracées. Revue Gén. Bot. 20: 321-327. pls. 11-14. 1908

. Les produits utiles des Burséracées. 73 pp. Paris. 1910.

. Recherches sur la structure et le développement des Burséracées. Ann. Sci. Nat. Bot. IX. 10: 1-302. 1909.

Heimscu, C., Jr. Comparative anatomy of the secondary xylem in the ‘“Gru- inales”’ and “Terebinthales’, of Wettstein with reference to taxonomic grouping. Lilloa 8: 83-198. pls. 1-17. 1942. |[Burseraceae, 122-124.]

Japtn, F. Recherches sur la ees et les affinités des Térébinthacées. Ann. Sci. Nat. Bot. VII. 19: 1-51. 1894. |Terebinthaceae divided into tribes Anacardieae and Bursereae. |

Lam, H. J. Beitrage zur Morphologie der Burseraceae insbesondere der Cana- rieae. Ann. Jard. Bot. Buitenzorg 42: 97-226. pls. 9-16. 1932. | Pinnate leaves of Burseraceae and related families are considered phylogenetically reduced shoots; but cf. Stnta, 1938.

. The Burseraceae of the Malay Archipelago and peninsula, with anno- tations concerning extra-Malayan species, especially of Dacryodes, San- tiria, and Canarium. Pace Jard. Bot. oo IIT. 12: 281-561. 1932. [The “General Part,’ 281-317, contains many important data on mor-

ology, dispersal, andl phylogeny of vm family.

ee P. W., C. Katkman, & H. J. Lam. Burseraceae. Jn: C. G. G. J.

N STEENIS, Fl, Males. ds 5: 209-296. 1955. [Includes notes of general nepaien on ecology, dispersal, distribution, wood anatomy, morphology, and taxonomy. |

1962 | BRIZICKY, SIMAROUBACEAE AND BURSERACEAE 185

MarcuHanp, L. Recherches sur l’organisation des Burséracées. 56 pp., ls. 1-6. Paris. 1868. pane L. L. Microsporogenesis and female gametophyte in Boswellia ser- a Roxb. Curr. Sci. Bangalore 28: 77, 78. 1958.* ———. Studies in Burseraceae. I. Jour. Indian Bot. Soc. 39: 204— -209. 1960. [ Boswellia serrata and Garuga pinnata Roxb., floral anatom s in Burseraceae. II. Jbid. 402 ~409. 1960. [Bursera serrata Colcne (= Protium serratum (Wall. ex Colebr.) Engl.), floral anatomy and Ee eee Garuga pinnata, embryology. REcorp, S. J., . W. Hess. Timbers of the New World. 640 pp., pls. 1-59. New oe ee { Burseraceae, 105-110. ] Rose, J. N. Burseraceae. N. Am. Fl. 25: 241-261. 1911. oe C. S. Manual of the trees of North America (exclusive of Mexico). ed. 2. 897 pp. 1922. [Burseraceae, 645-648. | SHUKLA, - D. Studies in the family Burseraceae—I. Floral anatomy of Balsamodendron mukul Hook. Agra Univ. Jour. Res. Sci. 4: 567-573. L9552% . Gametophyte in Balsamodendron mukul. Curr. Sci. Bangalore 23: 333. 1954.*

SintA, H. R. Zur Phylogenie der Fiederblatter der Burseraceen und verwandter Familien. Ann. Jard. Bot. Buitenzorg 48: 69-100. pls. 13, 14. 1938 [Against Lam’s theory of the caulomic nature of pinnate leaves of Bur- seraceae ae their allies. |

WeBBER, I. E. Systematic anatomy of the woods of the “Burseraceae.” Lilloa 6: 441-465. pls. 1-4. 1941.

WIGER, J. Embryological studies on the families Buxaceae, Meliaceae, Sima- rubaceae and Burseraceae. Thesis. Lund. 1935.*

1. Bursera Jacquin ex Linnaeus, Sp. Pl. ed. 2. 1: 471. 1762; Gen. Pl. ed. 6. 440. 1764, nom. cons.

Trees [or shrubs]. Leaves odd-pinnate [sometimes bipinnate], 3—9[- many |-foliolate [rarely 1-foliolate], usually crowded at the end of branch- lets, deciduous; leaflets opposite, chartaceous to subcoriaceous [or cori- aceous|, entire [or toothed], manifestly petiolulate [to sessile]. Flowers very small, unisexual [and/or bisexual], in axillary, raceme-like panicles, appearing prior to or with [or after] the leaves. Plants dioecious [or polygamous? J. Sepals 3-5, minute, connate at least at base, imbricate in bud. Petals 3-5, whitish to creamy, much longer than the sepals, spreading and recurved, induplicate-valvate in bud. Stamens 6-10, non- functional in @ flowers; filaments subulate; anthers oblong in outline, dorsifixed near the base, shorter and without pollen in @ flowers; pollen grains medium-sized, 3-colpate, reticulate-striate. Intrastaminal disc an- nular, 6-10-lobed, orange or red. Stigma capitate, 3-lobed; style short, ovary sessile, ovoid, 3-carpellate and -locular, with 2 collateral. pendu- lous ovules in each locule, rudimentary in 4 Aigwers. Drupes eubelobular or obliquely ellipsoid, renee triangular, with resinous, fleshy, leathery exo- and mesocarp detaching in 3 [or 2] valves when the fruit matures; stones (bony endocarp) covered with a thin, membranaceous, light-pink coat (probably the innermost layer of mesocarp remaining attached to the

186 JOURNAL OF THE ARNOLD ARBORETUM [ VOL, XLIII

endocarp), usually solitary, attached to the persistent axis of the ovary, ovoid-trihedral, essentially 1-locular, but bearing 2 minute, sterile locules or rudimentary stones in the upper half on the adaxial (ventral) side, usually 1-seeded. Seeds without endosperm; testa membranaceous; em- brvo straight Jor curved] with foliaceous, contortuplicate cotyledons and a short superior radicle. Germination epigeous, cotyledons 3-fid, the first leaf 3-foliolate. (Elaphrium Jacq., 1760; Simaruba Boehmer, 1760; nom. rejic.) Typr species: B. gummifera L. (= B. Simaruba (L.) Sarg.). (Named in honor of Joachim Burser, 1583-1639, a German physician and botanist. )

A genus of about 100 species of tropical America. Bursera Simaruba (Elaphrium Simaruba (L.) Rose), gumbo-limbo or West Indian birch, of the West Indies, Mexico, Central America, and northern South America, occurs in coastal hammocks on the Florida Keys and in peninsular Florida about as far north as Brevard and Pinellas counties. Easily recognizable by its lustrous, smooth, copper-colored bark which exfoliates in thin, papery lavers as in some species of Betula, in winter the tree (to 20 m. ) is conspicuously leafless, in contrast with its evergreen associates. The species is a common “fence-post” tree in tropical America, for pieces of the trunk or branches set in the ground quickly develop roots and grow into trees. The Mexican B. fagaroides (HBK.) Engl. and B. microphylla Gray occur in southern Arizona, the latter in southeastern California as we

Insect pollination has been presumed for the genus, but no data are available. Bursera Simaruba apparently is dioecious; records of polygamy are in need of verification. Staminate flowers in this species are 5-merous (or more rarely 4-merous). while the carpellate appear to be almost in- variably 3-merous and only very rarely 4-merous.

The species of Bursera yield a fragrant glutinous resin which is locally applied in domestic medicine. The resin of B. Simaruba is also used as a substitute for glue and as cement for mending broken china and glass.

The genus seems to be most closely related to the paleotropic Boswellia Roxb. ex Colebr

—

i

REFERENCES: See also under family references. ENGLER (1931, pp. 423-429). Recorp & Hess. Rose, and SARGENT. Buttock. A. A. Contributions to the flora of tropical America: XNNVII. Notes the Mexican species of the genus Bursera. Kew Bull. 1936: 346-387 ee {Includes kev. economic notes. and references to economic uses and anatomy. See also Kew Bull. 1937: 447-457. 1937, and 1938: 163-168 1938. for further notes. including the identification of Hinton’s Mexican collections. | SarGENT, C. S. Bursera. Silva N. Am. 1: 95-98. pls. 41, 42. 1891. West, 2, 1. 2. ARNOLD. The native trees of Florida, 212 pp. Gainesville. 1946. [B. Simaruba, 103.]

1962 | BAILEY & SRIVASTAVA, CACTACEAE, IV 187

COMPARATIVE ANATOMY OF THE LEAF-BEARING CACTACEAE, IV THE FUSIFORM INITIALS OF THE CAMBIUM AND THE FORM AND STRUCTURE OF THEIR DERIVATIVES

—

I. W. BartLey ! AND LAtitT M. SRIVASTAVA ”

Ir HAS BEEN DEMONSTRATED that in both dicotyledons and mono- cotyledons vessels originated by modification of long tracheids having scalariform pitting (Bailey & Tupper, 1918; Cheadle, 1942, 1943). In the case of the dicotyledons, highly advanced stages in the phylogenetic special- ization of vessels occur in plants having short vessel members with simple perforation plates throughout both the primary and secondary xylem (Bailey, 1944). In such plants scalariform perforation plates are elimi- nated. During the evolutionary specialization of vessels there commonly tend to be concomitant changes in the ground mass of imperforate tracheary cells which, by elimination of the borders of their pits, become libriform fibers, which, in turn, may at times retain their living contents, become septate, and function in the storage of starch (Bailey, 1936).

In dicotyledons, the differentiation of sieve tubes in the secondary phloem has appeared to afford a phylogenetic parallel to the development of vessels in the secondary xylem (Esau, 1953, p. 275; Esau, Cheadle & Gifford, 1953). In both tissues, axial translocation seems to be facilita- ted by modification in the more or less extensively overlapping ends of adjacent cells in axial seriations: in the xylem by loss of pit membranes to form perforations and in the phloem by formation of sieve plates having larger connecting strands than those in the sieve areas of the lateral walls. Some investigators have suggested that there is a direct correlation between the degree of evolutionary spcialization of sieve plates in the end walls and decrease in conspicuousness of the sieve areas in the lateral walls (Cheadle & Whitford, 1941; Cheadle, 1948; Cheadle & Uhl, 1948; Esau & Cheadle, 1959). Zahur (1959), however, did not find such a correlation.

It should be strongly emphasized in this connection that huge volumes of anatomical data regarding the secondary xylem of dicotyledonous families have accumulated during the last half-century. Wood, in general, is adequately preserved by simple drying, and industrial pressures have stimulated the assembling and study of large collections of wood samples by institutions in various parts of the world. No comparable information is available at present regarding the secondary phloem of dicotyledons.

' This investigation was financed in part by a grant from the National Science Foundation. I am indebted to the pee Philosophical Society for the loan of a Wild microscope.

? Mercer Fellow of the Arnold Arboretum.

188 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIII

Further, as explained by Cheadle (1956), this tissue is difficult to study because of its physiological and structural peculiarities. Even the most extensive reconnaissance to date (Zahur, 1959) deals with only a rela- tively few representatives of the 85 families sampled.

Furthermore, as one of us (Bailey, 1957) has emphasized: ‘There are certain details of the trends of specialization in the xylem which need to be more clearly and generally understood in shifting from a consideration of the dicotyledons as-a-whole to investigations of the taxonomy of individual taxa of restricted size. In revealing salient trends of evolu- tionary specialization by analyses of data obtained from the dicotyledons as-a-whole, variations due to obtaining specimens from different parts of the plant, from plants of different growth rates, from genetically different taxa, etc., tend to neutralize one another. In addition, various localized, divergent trends of specialization do not obscure or confuse the major trends of evolution in the dicotyledons as-a-whole. However, when one becomes concerned with taxa of decreasing size, viz. families, sub- families, tribes, genera and species, such variations and deviations become increasingly significant.”

For example, data from the dicotyledons as-a-whole clearly reveal evolutionary changes from scalariform to simple perforation plates, and from scalariform to alternate multiseriate intervascular pitting. In any randomly selected minor taxon, one of these trends may be retarded or accelerated in relation to the other. Therefore, with present inadequate information regarding the secondary phloem in most orders, families, and genera, it is not possible to determine with certainty what some of the more important trends of phylogenetic specialization in the dicotyledons as-a-whole may actually be. This is particularly the case in those paren- chymatous cells that are physiologically and ontogenetically related to sieve elements.

During the phylogenetic specialization in the secondary tissue of dicoty- ledons, there is a progressive shortening of fusiform cambial initials which not infrequently culminates in storied forms of cambia, involving longitudinal rather than pseudotransverse anticlinal divisions and the elimination of intrusive elongation following such divisions (Bailey, 1923).

In general, the phylogenetic shortening of fusiform cambial initials is most closely paralleled in the secondary xylem and phloem by shortening of fusiform parenchymatous derivatives and by changes in the length of parenchyma strands (for ordinarily fusiform parenchymatous cells and the mother cells of parenchyma strands do not elongate during tissue differentiation). As a concomitant of shortening fusiform cambial initials, the parenchyma strands commonly tend to be composed of fewer and generally of shorter cells.

Statistical data obtained from the dicotyledons as-a-whole indicate that shortening of the fusiform initials likewise is closely reflected in the length of vessel members, there being only slight elongation at times during the maturation of primitive, long, slender vessel members, and a slight con- traction at times during the differentiation of short, very broad, highly

1962 | BAILEY & SRIVASTAVA, CACTACEAE, IV 189

-specialized ones. On the contrary, fiber-tracheids and libriform fibers become longer than the fusiform cambial initials from which they are derived owing to more or less extensive intrusive elongation during tissue differentiation.

In the secondary phloem of such vesselless gymnosperms as the Pinaceae, Taxodiaceae, and Cupressaceae, statistical averages indicate that sieve cells simulate fusiform cambial initials in length, there being no appre- ciable elongation, as occurs at times in tracheids, during their maturation from cambial initials. In the phylogenetically specialized dicotyledons having very short fusiform cambial initials, particularly those having storied cambia, the sieve-tube members in statistical averages closely simulate the fusiform initials in length. On the contrary. in some dicoty- ledons having less highly specialized cambia, the parallelism in length may be modified by more or less numerous divisions of mother cells prior to the formation of sieve-tube members, companion cells, and some paren- eat cells (cf. Esau & Cheadle, 1955; Cheadle & Esau, 1958; Zahur,

1959). In such plants, the sieve-tube members tend statistically to be considerably shorter than fusiform cambial initials.

Statements in the literature regarding detailed structure of the xylem and phloem of leaf-bearing cacti are casual and fragmentary and are based largely upon Pereskia aculeata Mill. The occurrence of septate libriform fibers and porous vessels in the wood of this species is recorded by Schenck (1893) and Solereder (1899). If the most primitive living cacti occur among the leaf-bearing ones, as is generally assumed to be the case, it is essential to obtain comprehensive information regarding the levels of phylogenetic specialization that they have attained, particularly for future use in understanding trends of increasing structural specializa- tion that occur in the Opuntieae and Cereeae.

In all of the putative species of Pereskia, Pereskiopsis, and Quiabentia that we have studied (mentioned in previous papers of this series | Bailey. 1960, 1961a, 1961b]) the fusiform initials of the cambium have attained a high level of evolutionary modification. ‘They are palaaaaiy short, commonly ranging in length from only 150 to 400 microns. Althou Bh they are not consistently in perfect storied or stratified eee’ they frequently exhibit a tendency to become storied, at least in some parts of a mature plant. Where they approach a perfect storied arrangement, they have abruptly tapered ends and a hexagonal form as seen in tangential longitudinal sections of the cambium (Fic. 1) and the cells of one stratum do not extensively overlap those of higher and lower levels. On the con- trary, where stratification is imperfect, the cells have more gradually tapered ends, and there is more overlapping of the cells of different levels (Fic. 4). It should be noted in this connection that where stratification occurs in the leaf-bearing cacti it tends to differ from that which occurs in plants of other dicotyledonous families in exhibiting a conspicuous tendency for the strata of fusiform initials to have a diagonal, rather than a transverse, orientation as seen in tangential longitudinal sections of the cambium.

190 JOURNAL OF THE ARNOLD ARBORETUM [ VOL, XLII

In dicotyledons having very short, especially storied, cambial initials, where adequately preserved material of the cambium is not available, the length and the arrangement of fusiform initials mav be detected both in certain of their derivatives in the xylem and in recently formed phloem,

The distribution of wood parenchyma in Pereskia, Pereskiopsis, and Quiabentia is of a specialized paratracheal rather than a_ primitive apotracheal form, but more or less extensive arcs of thick-walled lignified or thin-walled unlignified parenchyma occur at times and may prove to be associated in some manner with successive zones of growth in the enlargement of stems and roots. The fusiform parenchymatous cells and parenchyma strands are short, the latter being composed usually of two or three cells only. In tangential longitudinal sections of the second- ary xylem, the fusiform parenchyma and parenchyma strands (where not deformed and displaced by excessive enlargement of vessels) closely simu- late fusiform cambial initials in overall form. Furthermore, where paren- chyma is sufficiently abundant, particularly in areas of zonal distribution. the arrangement of fusiform cambial initials, whether storied or non- storied, is clearly revealed by the arrangement of the parenchymatous cells,

The vessels in stems and roots of the leaf-bearing cacti are of a highly advanced evolutionary form having simple perforation plates throughout the primary (Fic. 8) and secondary xylem (Fic. 5). In the secondary xylem, the vessels vary markedly in diameter from 20 microns to as much as 200 microns in certain cases. The vessels occur independently and in clusters of varying size and form. The smaller vessels, as seen in trans- verse sections, are angular, whereas the larger ones are more nearly cir- cular or oval, except where they are modified by compression in clusters. Such variations in the vessels occur, not only in different species and dif- ferent parts of a single plant, but also in closely oe areas of the sec- ondary xylem. The members of the smallest vessels (i.e., those which do not expand appreciably in tangential diameter during maturation) re- semble fusiform cambial initials in overall form when viewed in tangential longitudinal sections of the xylem. In such sections, the vessel members may be confused with vascular tracheids, but the perforations in their more or less abruptly tapered ends are clearly visible in radial longitudinal sec- tions. Where sufficiently abundant, and particularly in association with parenchyma, such small vessel members reveal a storied (Fic. 2) or non- storied arrangement of fusiform initials in the cambium.

In the secondary xylem of the leaf-bearing cacti, vessel members of the smallest diameter tend to have diagonally oriented perforation plates in their terminal radial walls. Larger vessel members, although as short or shorter than fusiform cambial initials, vary more or less markedly in form owing to excessive lateral enlargement during their maturation. The perforation plates, consisting of a circular or oval opening surrounded by a clear zone of unpitted wall, may be transversely (Fic. 5) or more or less diagonally oriented. Furthermore, where vessels are closely associated

1962 | BAILEY & SRIVASTAVA, CACTACEAE, IV 191

in larger clusters and where vessels deviate from a longitudinal course to pass diagonally through the broad multiseriate rays (Fics. 6, 7), the vessel members may at times have perforation plates in their sides rather than in their more or less inclined ends.

The bordered pitting in the sides of adjoining vessels tends to be of the evolutionarily advanced alternate-multiseriate form (Fics. 9, 11), but areas of scalariform pitting (Fics. 12, 13) and of opposite-multiseriate pit- ting are of not infrequent occurrence. The bordered pits of multiseriate ar- rangements vary in size and considerably in form from circular te oval and to angular where the pits are closely crowded. (Compare Fics. 9, 11, 13, 15, 16.) The pit apertures likewise vary from small, nearly circular (Fic. 9), to oval (Fic. 15), and to transversely slitlike (Frc. 151. Furthermore, the borders of the pits exhibit at times reduction in conspicuousness in rela- tion to the size of the apertures. (Compare Fics. 11, 15, 16.) In areas of scalariform pitting, unconformity in the pitting of the walls of adjacent vessels frequently occurs (Fics. 12, 13), i.e.. a transversely extensive pit in one wall being related to smaller pits in the adjoining vessel wall.

The pits of vessels in contact with wood parenchyma and rays exhibit a wide range of variability. At times, they closely resemble those of inter- vessel contacts in size, form, and arrangement. (Compare Fics. 9, 10; 13, 14; 16, 18.) However, in the case of the leaf-bearing cacti, there commonly is a conspicuous tendency for vessel pits in contact with parenchyma to have enlarged apertures and to reduce or eliminate their bordering area (Fic. 20). In other words, the pits in the walls of vessels in contact with parenchyma become simple pits in conformity with the unbordered pits of adjacent parenchymatous cells. Where vessels are completely jacketed by wood parenchyma and have large, transversely elongated unbordered pits, the vessel members, when isolated by maceration, resemble vessel members of those parts of the primary xylem which have reticulate forms of wall thickening (Fic. 22).

The libriform fibers of the leaf-bearing cacti, unlike the vessel mem- bers, fusiform wood parenchyma, and parerichyma strands, elongate more or less extensively by intrusive elongation during tissue differentiation. Therefore, they (Fics. 24-27) become longer than the fusiform cambial initials (Fic. 23) from which they are derived. However, the amount of elongation varies greatly, not only in different parts of a plant, but also in closely adjacent parts of the secondary xylem. In some cases, the elonga- tion at one or both ends of the cell is only a few microns, whereas, in other cases, the elongation at both ends of the cell is extensive. Thus, libriform fibers may be only slightly longer than the fusiform cambial initials from which they are derived, i.e., 150-400 microns (Fic. 24) or they may attain maximum lengths of as much as 1000-1300 microns (Fics. 26, 27). Some libriform fibers are relatively slender and taper gradually and uniformly toward their ends (Fic. 26); others are broader in their central part and taper abruptly at one or both ends into long narrow tips (Fics. 25, 27). In both instances the small oval or slitlike simple pits tend to be concen- trated in the central part of the fibers which corresponds roughly in length

192 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLII

to that of the fusiform cambial initials. (Compare Fic. 23 with Fics. 24— 27.)

The slitlike pits usually are oriented more or less diagonally to the long axis of the cell but may at times be oriented parallel to its axis. This in- dicates that the orientation of cellulosic microfibrils in the central or Se layer of the secondary wall may, at least occasionally, deviate from helical to Jongitudinal.

he fibers in any particular part of the secondary xylem may be pre- vailingly septate (usually a single septum to each cell, Fic. 26), nonseptate, or in varying mixtures of septate and nonseptate. Both the septate and the nonseptate (Fics. 17, 19, 21) fibers retain their protoplasts and nuclei and are capable, at least in certain parts of the plant or at certain seasons of the year, of storing abundant starch.

In the members of Pereskia, Pereskiopsis, and Ouiabentia which we in- vestigated, the phloem derivatives, viz., sieve elements and parenchyma, as seen in tangential longitudinal sections in close proximity to the cambium, commonly reflect the form and distribution of the fusiform cambial initials (Fics. 3,4). The fusiform parenchyma and parenchyma strands, however, are more reliable indicators of the overall form of fusiform initials than are the sieve elements. This is because the sieve-tube members commonly dif- ferentiate in one or more parts of a cell complex that results after longitudi- nal and sometimes transverse divisions in a single phloem mother cell. Therefore, although the length of sieve-tube members in statistical averages generally corresponds to that of the fusiform cambial initials, their overall form frequently does not. The phloem-parenchyma strands are usually composed of two or three cells arranged in a vertical file and appear to originate by divisions in a single phloem mother cell.

The sieve-tube members present features of advanced structural speciali- zation. The sieve plates commonly occur in the more or less transverse (Fic. 28) or slightly oblique end walls (Frcs. 29, 30) but sometimes may be present in the lateral walls (Fic. 29). In certain cases, sieve plates may not be present at both ends of a sieve-tube element but may occur only at one end and in one of the lateral walls (Fic. 29). Usually the pores and the callose cylinders are distributed rather evenly throughout the sieve plate and the latter may be interpreted as a simple sieve plate. Occasional variations in sieve-plate structure occur, however. In some sieve elements with somewhat sloping end walls the pores and the callose cylinders may be distributed in two or more distinct groups in the sieve plate (Fics. 30, 31). There is no strict uniformity with regard to the distribution of hese two forms of sieve plates. They occur in different parts of the same plant and even at the two ends of a single sieve-tube member. These variations in sieve-plate structure are related in a general way to the orientation of the end walls in the sieve elements which, in turn, seems to be related to the character and orientation of end walls in the fusiform cambial initials (Fie, 23)

In addition to sieve plates which usually occur at or near their end walls, sieve-tube members have numerous small sieve areas that are scattered

1962 | BAILEY & SRIVASTAVA, CACTACEAE, IV 193

throughout both radial and tangential lateral walls (Fic. 32). The pores and their callose cylinders in these sieve areas are much smaller than those in the sieve plates. (Compare Fics. 31, 32.) The size of these lateral sieve areas, as determined by the number of callose cylinders stained with lac- moid in one area, varies considerably. In the same sieve-tube member, some lateral sieve areas may be very small, composed of one or two strands and their callose cylinders only, whereas others may be composed of sev- eral strands and their callose cylinders.

Generally, one or more companion cells occur in association with a sieve-tube member (Fics. 28, 29). In addition, one or more members of a cell complex, in which sieve-tube elements and companion cells arise, may differentiate as parenchymatous elements. The exact relationships between the different cells that originate after divisions in a single phloic mother cell and the sieve-area connections between them require detailed ontogenetic study.

During the transition from “functional” to “nonfunctional” phloem * the parenchymatous cells of both the axially oriented part and the rays retain their living contents, capacity for division, and commonly undergo more or less extensive changes in size and form. The most striking changes of taxonomic significance are those which occur in the nonfunctional phloem of Pereskia, these being absent in Pereskiopsis and Quiabentia. As one of us has shown (Bailey, 1961a) three distinct categories of pereskias can be segregated upon the basis of differences in the formation of sclereids in the nonfunctional part of the secondary phloem. In two of these cate- gories of species, sclereid formation involves extraordinary enlargements of parenchymatous cells in both diameter and length. In the third cate- gory, the changes in size and form are more nearly comparable to those that occur in the formation of ordinary sclereids

Although crystals are commonly present in the ray parenchyma of functional phloem in all three genera, they rarely, if ever, occur in axially oriented parenchyma. However, they do occur at times in axial parenchyma of nonfunctional phloem (Bailey, 1961b). Variations in the presence and form of these crystals may prove to be of some taxonomic significance when more adequate and extensive collections are available for detailed investigation.

DISCUSSION

It is evident from our reconnaissance of the secondary vascular tissues of Pereskia, Pereskiopsis, and Quiabentia that these tissues have attained highly advanced levels of evolutionary specialization. This is shown, for example, by the dimensions and form of the cambial fusiform initials and their derivatives, by the perforation plates and the lateral pitting in ves- sels, by the storage of starch in both septate and nonseptate libriform fibers, and by the distribution patterns of vessels and wood parenchyma.

° Following Esau (1953, p. 299) “nonfunctional” phloem refers to that part of the phloem in which sieve elements and companion cells have ceased to function

194 JOURNAL OF THE ARNOLD ARBORETUM [ VOL, XLIII

Such evidence at least suggests that the early leaf-bearing representatives of the Cactaceae (those of nearly typical dicotyledonous woody arborescent and shrubby form) had attained high levels of internal anatomical speciali- zation before increasing succulence and other morphological changes led to the differentiation of the Opuntieae and Cereeae

The ranges of anatomical variability in different parts of a single plant and in the same clone when grown under different environmental in- fluences are very extensive. Potential diagnostic criteria, such as differ- ences in the size, form, and distribution of the constituent cells of xylem, in the form and orientation of perforation plates; in pitting between vessels and between vessels and parenchyma; in the presence or absence of septa in libriform fibers, etc. (which have been utilized so commonly in the differentiation of taxa in other dicotyledonous families) appear, in gen- eral, in the leaf-bearing Cactaceae to be of a quantitative rather than a qualitative character. Likewise, in phloem, the orientation of the sieve plates near the ends of sieve-tube members may vary from more or less oblique to perfectly transverse, and the number of pore groups (viz., sieve areas) on the sieve plate may range from one to three or more. The difference between the size of pores and connecting strands in the lateral sieve areas versus the sieve plates is very great in all species of the three leaf-bearing genera. Therefore, in the case of these cacti, large volumes of material must be studied statistically in order to attain results of valid taxonomic significance.

In the past, unjustifiable phylogenetic and taxonomic inferences have resulted from comparisons between the frequency of diagonal and trans- verse orientations of terminal perforation plates in closely related taxa without taking into consideration various factors involved in the statis- tical differences. Important influences in this connection are the length of fusiform cambial initials, the amount of transverse enlargement of ves- sel members during tissue differentiation, the deformation when vessels occur in crowded clusters, and the aberrations produced in xylem of dis- torted or burly grain. In straight-grained xylem of the leaf-bearing cacti, the broader vessel members tend, on an average, to have transversely or nearly transversely oriented terminal perforation plates, whereas the narrowest vessels commonly have diagonally oriented ones. Where the grain of the xylem is distorted, as so frequently happens in stems and roots of the Jeaf-bearing cacti (owing in large measure to the frequent diagonal dissection of broad rays by conversion of ray initials to fusiform ones), steeply diagonal and even laterally placed perforation plates are of not infrequent occurrence. Similar aberrations occur in the orientation of sieve plates in the phloem. From the point of view of salient trends of phylogenetic specialization in the dicotyledons as-a-whole, the occurrence of simple perforation plates throughout both the primary and secondary xylem is of greater evolutionary significance than are deviations from a transverse orientation of perforation plates in the vessels of certain parts of the stems and roots of the leaf-bearing cacti.

The occurrence of septate libriform fibers in various dicotyledonous

1962 | BAILEY & SRIVASTAVA, CACTACEAE, IV 195

families has been based largely upon surveys made of dried wood samples. It should be recognized, in this connection, that a great many of the samples consist of heart-wood. Where specimens of undecayed sapwood are available, they are sectioned after boiling in water and frequently after softening treatment in hydrofluoric acid. Such sections do not con- tain starch, but the assumption is made, probably correctly, that the septate fibers in the sapwood retained their living protoplasts at functional maturity and were capable, at least for a time, of storing starch. The com- mon occurrence of starch in nonseptate libriform fibers of the leaf-bearing cacti raises the question whether the storage of starch in nonseptate fibers of sapwood is not of much commoner occurrence in dicotyledons than has been realized. From physiological as well as phylogenetic points of view, the whole problem of retention of living protoplasts and storage of starch in nonseptate libriform fibers, whether confined to newly formed sapwood, whether capable of seasonal depletion and renewal as in the case of starch in ray tissue, etc., needs to be thoroughly investigated, not only in the Cactaceae, but also in other families of the dicotyledons.

In this paper, we have omitted discussion of ray initials in the cambium and of their derivatives in the xylem and phloem. The rays in the leaf- bearing Cactaceae are obviously of a highly specialized form, being pre- vailingly multiseriate, uniseriate rays having been eliminated. It seems best to discuss rays in subsequent papers of this series, dealing in greater detail with the structure of various putative species of Pereskia, Pereskiop- sis and QOuiabentia.

LITERATURE CITED

Bartey. I. W. 1923. The cambium and its derivative tissues, IV. The increase in girth of the cambium. Am. Jour. Bot. 10: 499-509.

—_—. 1936, The problem of differentiating and classifying tracheids, fiber tracheids and libriform fibers. Tropical Woods 45: 18-253.

_ 1944. The development of vessels in angiosperms and its significance in

morphological research. Am. Jour. Bot. 31: 421-428.

__. 1957. The potentialities and limitations of wood anatomy in the study

of the phylogeny and classification of angiosperms. Jour. Arnold Arb. 38:

243-254.

————. 1960. Comparative anatomy of the leaf-bearing Cactaceae, I. Foliar vasculature of Pereskia, Pereskiopsis and Quiabentia. Ibid. 41: 341-349. - 1961a. IL. Structure and distribution of sclerenchyma in the phloem of Pereskia, Periskiopsis and Quiabentia, Ibid, 42: 144-150. - 1961b. IIL. Form and distribution of crystals in Pereskia, Pereskiopsis and Quiabentia. Ibid. 334-340.

___& W. W. Tupper. 1918. Size variations in tracheary cells. I. A com- parison between the secondary xylems of vascular cryptogams. gymno- sperms and angiosperms. Proc. Am. Acad. Arts Sci. 54: 149-204.

Cueap_e. V. I. 1942. The occurrence and types of vessels in the various organs of the plant in the Monocotyledoneae. Am. Jour. Bot. 29: 441-450.

_ 1943. The origin and certain trends of specialization of the vessel in Monocotyledoneae. Ibid. 30: 11-17

196 JOURNAL OF THE ARNOLD ARBORETUM [ VOL, XLII

———. 1948. Observations on the phloem in the Monocotyledoneae, II. Ad- ditional data on the occurrence and rier ee in structure of the sieve tubes in the metaphloem. /bid. 35: 131; . 1956. Research on xylem and ee in fifty years. Jbid. 43: 719-731 & K. Esau. 1958. Secondary phloem of Calycanthaceae. Univ. Calif. Publ. Bot. 29: 397-510 N. W. URL. 1948. “The relation of metaphloem to types of vascular bundles in the Monocotyledoneae. Am. Jour. Bot. 35: —583.

& N. WHITFORD. 1941. Observations on the ae in the Mono- cotyledoneae, I. The occurrence and phylogenetic specialization in structure of the sieve tubes in the metaphloem. Am. Jour. Bot. 28: 623- soe

Esau, K. Plant Anatomy. John Wiley & Sons, Inc., New CHEADLE. 1955. Significance of cell cighioue in differentiating eae phloem. Acta Bot. Neerl. 4: 346-357. & 1959. Size of pores and ha contents in sieve elements of dicotyledons. Prog. nine Acad. ae Ac —162. & E. FFORD, JR. 1953. ‘Compaitive ape and possible trends of cea of the cee m. Jour. 40: 9-19 SCHENCK, H. 1893. ay zur Biologie meu ne ve Lianen. ‘Vol 2 7s 6 148. Gustav Fischer. a se a 1899, aise Anatomie der Dicotyledonen. F. Enke. Stuttg ZAHUR, M. . 1959. Comparative study of secondary phloem of 423 spec of woody oo belonging to 85 families. Mem. N. Y. State College Agr. 358:

EXPLANATION OF PLATES PLATE I

Fics. 1-4. TANGENTIAL LONGITUDINAL SECTIONS OF CAMBIUM. XYLEM AND PHLOEM, X 110. 1, Pereskia cubensis Britt. & Rose [Atkins Gard.], storied fusiform eta of the cambium. 2, Pereskiopsis blakeana Ortega [Kimnach & Moran 82]|, diagonally storied peseel members and wood- -parenchyma strands. 3, Pereskia autumnalis (Eichlam) Rose [Moore 8210], storied sieve-tube members; terminal sieve plates are black. 4, Pereskia aff. sacharosa Griseb. [ Cardenas |, im- perfectly storied cambial Hea initials (left of ray) and phloem derivatives (right of ray).

PLATE II

Fics. 5-8. OCCURRENCE AND ORIENTATION OF SIMPLE PERFORATION PLATES

IN VESSEL MEMBERS. 5, Pereskia grandifolia Haw. | Castellanos], transverse sec- tion of the secondary xylem showing transversely oriented perforation plates, x 260. 6, Pereskia colombiana Britt. & Rose | Romero |, radial longitudinal section of the secondary xylem showing laterally placed perforation plate, Pereskia nicoyana Web. [ Rodriguez 662], tangential longitudinal section of the secondary xylem showing divergence of vessel ee through multiseriate ray; vessel members at right have perforation plates, as seen in sectional view at (X), in their lateral radial walls, X 180. 8, be aguosa (Web.) Britt. & Rose [Dressler], radial longitudinal section of the primary xylem showing perforation plates in helically thickened vessel members, X 260.

1962 | BAILEY & SRIVASTAVA, CACTACEAE, IV 197

PLATE III

Fics. 9-15. PIrTING BETWEEN VESSELS AND BETWEEN VESSELS AND PAREN- CHYMA. 9, Pereskia grandifolia | Castellanos|, crowded alternate-multiseriate pitting in walls of contact between vessel members, X 410. 10, Pereskia colom- biana [Romero], pitting between vessel and ray parenchyma, X 410. 11, Peres- kia en [ Castellanos |, variations in form of bordered pits and their aper- tures, X 1130. 12, Pereskia colombiana | Romero |, unconformity i in the pitting of vessels; many Oi the transversely elongated pits in the wall of one vessel are in contact with rows of shorter pits in the wall of the adjoining vessel, x 960. 13, The same, scalariform and transitional pitting in lateral walls of adjoining vessels, X 410. 14, The same, scalariform pitting between vessel and parenchyma, & 410. 15, Quiabentia aff. chacoensis Backbg. [Tucuman], bor- dered pits with large apertures in the lateral walls of adjoining vessels, * 410.

PLATE IV

Fics. 16-22. LONGITUDINAL SECTIONAL VIEWS OF VESSELS, PARENCHYMA AND NONSEPTATE LIBRIFORM FIBERS. 16, Pereskiopsis blakeana [| Kimnach & Moran 82], reduction in the borders and increase in size of apertures in intervascular pitting, X 410. 17, Pereskia aculeata Mill. { Atkins Gard.], iodine stained starch in a nonseptate ineionn fiber, X 750. 18, Pereskia bleo DC. [Atkins Gard.], reduction in borders and increase in size of apertures in pitting ne vessel and ray parenchyma, X 410. 19, Pereskiopsis chapistle Britt. & Rose [Boke B-3], nuclei (black) and unstained starch in nonseptate libriform fibers, 1000. 20, Pereskiopsis blakeana | Kimnach & Moran 82], pitting between ves- sels and parenchyma, * 410. 21, Quiabentia aff. chacoensis |Tucuman], nucleus with nucleolus in nonseptate libriform fiber, X 1000. 22, Quiabentia pereziensis Backbg. [Cardenas], large unbordered pits between vessel and parenchyma, X

PLATE V

Fics. 23-32. DIAGRAMMATIC ILLUSTRATIONS OF CAMBIAL FUSIFORM INITIALS, LIBRIFORM FIBERS AND SIEVE-TUBE MEMBERS. 23, Pereskia sacharosa [Tucuman], fusiform cambial initials, drawn from tangential longitudinal section of cambium. 24, Pereskia conzattu Britt. & Rose [Dressler], libriform fiber drawn from maceration. 25, Pereskia aculeata [Atkins Gard.], libriform fiber drawn from maceration. 26, 27, Pereskia conzatti [Dressler], septate and nonseptate libri- form fibers drawn from maceration. 28, Pereskia sacharosa [Tucuman], sieve- tube member and companion cells, ein from tangential eon Bee tion of phloem. 29, Pereskiopsis aff. chapistle [Boke B-3], sieve-tube member and companion cell, drawn from tangential longitudinal section of Set 30, The same, sieve-tube member drawn from radial longitudinal section. 31, The same, part of a sieve plate, drawn from radial longitudinal section of phloem. 32, The same, lateral sieve areas, drawn from radial longitudinal section, companion cell on right. Figures 23-30, * 175; figs. 31, 32, & 875.

PLATE I

Jour. ARNOLD Ars. VoL. XLIII

=

‘ASTAVA, LEAF-BEARING CACTACEAE, IV

f

BAILEY & SRIV

PLATE II

Jour. ARNOLD Ars. VoL. XLIII

IV

BEARING CACTACEAE,

BAILEY & SRIVASTAVA, LEAF

Jour. ARNOLD Arps. Vor. XLIII PiaTE III

ae

— = i —

BaILey & Srivastava, LEAF-BEARING CACTACEAE, IV

Jour. Arnoip Ars. VoL. XLITI PLATE IV

BAILEY & SRIVASTAVA, LEAF-BEARING CACTACEAE, IV

Jour. ARNOLD Ars. VoL. XLIII PLATE Y¥

rae 26

BarLey & Srivastava, LEAF-BEARING CACTACEAE, IV

1962] MOORE, CARAGANA SINICA 203

ON THE ORIGIN OF CARAGANA SINICA?

RAYMOND J. Moore

THE SHRUB LONG KNOWN as Caragana chamlagu Lam. probably was introduced to Europe by Father Pierre d’Incarville, who sent seed of many plants collected near Peking, China, to the Jardin des Plantes, Paris, in the period 1740-1756 (Bretschneider, 1898). Loudon (1844) stated that the species was introduced to Great Britain in 1773. Appar- ently it has remained since the eighteenth century one oi the less com- monly cultivated shrubs. Rehder (1941) has pointed out that the over- looked epithet Robinia sinica of Buc’hoz has priority over that of Lamarck and, to avoid confusion, the name Caragana sinica (Buc’hoz) Rehder will be used exclusively hereafter, although Komarov, Pojarkova, ef al. have treated this entity under the name C. chamlagu. Chiefly on the basis of flower and calyx size, Pojarkova (1945) has divided C. sinica, sensu lato, into C. sinica and C. ussuriensis.

Caragana sinica, sensu lato, occurs widely in eastern China and far- eastern Siberia. It is a shrub 1-2 m. tall, apparently common in dry, rocky, or other well-drained sites. The leaves consist of 2 pairs of leaflets, usually pinnately arranged, but often so close together as to appear almost palmate; the leaflets are obovate, 10-35 mm. long, 5-15 mm. broad, rather coriaceous, glaucous above, the base cuneate, the apex retuse, mucronu- late. The leaf rachis thickens and develops into a spine up to 25 mm. long which persists after the leaflets drop. The persistent rachises and the spiny stipules (5 mm.) give the bush its conspicuous spiny character. The flowers are large, 20-30 mm. long; the calyx 9-14 mm. long, 5-6 mm. broad, with lobes 2-3 mm. long; the pedicel attachment asymmetrical; and the calyx almost gibbous. The flowers are borne one or two per node on stalks 10-20 mm. long, articulated in the middle between pedicel and peduncle. The corolla is pale yellow, usually with a rose or bronze tinge which deepens with age.

It is generally agreed that Caragana frutex and related species comprise the only species-group to which C. sinica is clearly related. Komarov (1908) placed C. sinica in his series FRUTESCENTES. He recognized its unique character and regarded it (oc. cit., p. 370) as the closest claimant to the position of generic prototype. He based this view on its leaf char- acter, noting that at an early age the leaf is palmate but later becomes pinnate. This transition he held to be a phylogenetic recapitulation, ap- parently regarding palmate leaves as ancestral to pinnate. Although not clearly stated, it appears that Komarov did not refer to the form of early seedling leaves but to the varying forms seen on adult plants. He believed

1 Contribution No. 128 from the Plant Research Institute, Canada Department of Agriculture, Ottawa.

204 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIII

(p. 371) that different environmental conditions then had resulted in the evolution from C. sinica of both palmate-leaved species (FRUTESCEN TES) and pinnate-leaved species (ALTAGANAE) and subsequently from these, all the other series. Caragana sinica thus assumes special importance as the foundation of Komaroy’s phylogenetic scheme. Doubtless as a re- sult of this opinion, he designated east Asia (range of C. rosea, C. fruti- cosa) as the primary center of species formation.

To show specific relationships more exactly, Pojarkova (1945) divided the series FRUTESCENTES Kom. into three more homogeneous but related series: CHAMLAGU, FRUTESCENTES emend., and GRANDIFLORAE. The series CHAMLAGU Pojark. consists of C. sinica and C. ussuriensis (Regel) Pojark. The latter taxon had been described by Regel as C. frutex var. USSUTIENSIS, and, as such, was placed in the synonymy of C. sinica by Komarov (1908).

Pojarkova separates Caragana ussuriensis and C. sinica chiefly on floral characters: flower length, 23-25 mm. vs. 28-30 mm.: and calyx size, to 9 mm. long X 5 mm. broad vs. 12-14 mm. long & 6 mm. broad. The leaflets of C. stnica are said to have a broader, obovate blade and broader base, giving a more rounded outline, whereas those of C. ussuriensis are narrow with a narrowly cuneate base. Only C. ussuriensis occurs in the U.S.S.R. (region of the river Ussuri just north of Manchuria) from whence it extends southward into Manchuria and northern China. It is reported that C. sinica has a more southerly distribution in China: Hopeh (Chili) to Yunnan. The impression is conveyed by Pojarkova that the range of C. ussuriensis is more northerly than that of C. sinica and that, although they overlap in northern China, C. sinica alone occurs in southern China. The two taxa are illustrated in Fics. 1 and 2.

The illustration accompanying Buc’hoz’s (1779) description of Robinia sinica is apparently drawn life-size and shows the larger-flowered plant. Mature open flowers are 30 mm. long, the calyx being 11-12 mm. long and 6 mm. broad, measurements which fall within the limits indicated by Pojarkova for C. sinica.

MATERIALS AND METHODS

In the course of a survey of chromosome numbers in Caragana, a special attempt has been made to obtain living material of C. sinica. Seeds have been received under this name from twelve botanic gardens (eleven Euro- pean) but the resulting seedlings proved to be not the desired species but typical plants of C. frutex, C. arborescens, C. pygmaea, or C. aurantiaca. Some of these species are not closely related to C. sinica; all are readily distinguished from it and it is considered that the error is due solely to incorrect identification of the seed parent. In two cases, leaflets in the seed packet indicated that the source of the seed was C. arborescens and C. pygmaea, as were the resulting seedlings.

Living plants of Caragana sinica have been obtained from only one source: F. J. Grootendorst & Sons, Boskoop, Holland. The species has been propagated asexually in this nursery. It was received there at some

1962 | MOORE, CARAGANA SINICA

L, Baikal

CHINA -*..-°

one 3 ee $0 oe. ai : @o0 e eh are soe eo?

U.S,S.R-

MANCHURIA

=

JAPAN

Y

ussuriensis sinica

intermediate

Fics. 1-4. pp eae ae details and distribution of oh iaee simica. 1, ae

and flower of C. sin 2. Lea

lato,

C. ussuriensis. plement of C. sinica “(eugene triploid).

3, Somatic chromosome c 4, Dcebition of C. sinica, sensu

206 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLII

time before 1928, but the source is unknown (H. J. Grootendorst, personal communication, 1961).

Cuttings of the rare species Caragana rosea were received from the Arnold Arboretum, Harvard University, where there is a single shrub grown from seed collected by William Purdom (seed lot 9a), in 1909, in Weichang, Chili (Hopeh) Province, China. This collection was determined as C. rosea by Rehder (1926).

Herbarium material of Caragana sinica from the Arnold Arboretum (A) and the Gray Herbarium (cu) of Harvard University and from the United States National Museum (us), as well as that in the Herbarium, Canada Department of Agriculture (pao) has been examined. Particular attention has been paid to the characters which separate C. sinica and C. ussuriensis and to the condition and size of the pollen. Pollen was mounted in 45% acetic acid and stained with dilute cotton blue. Grains with a protoplast which completely filled the cell and stained blue were counted as viable. The grains are thin walled and smooth; the diameter stated is the measurement to the outer surfaces.

MORPHOLOGY, POLLEN CONDITION, AND NATURAL DISTRIBUTION

Little variation was found in the size of the flower on any single her- barium specimen, and it is believed that these size-characters are reason- ably constant for any individual. A similar opinion has been formed from observation of numerous living plants of various species of Caragana over a number of years, and it is believed that the size-characters used by Pojarkova reflect genetic differences and are reliable for systematic purposes.

Following the floral and leaf characters designated by Pojarkova, 34 of 38 herbarium specimens were referred either to Caragana ussuriensis or to C, sinica. The determination of the remaining four, all wild Chinese collections, was less certain. Two had the flower size of C. sinica with a smaller calyx, and two had the smaller flower of C. ussuriensis with a calyx similar in size and appearance to that of C. sinica.

The localities of the Asiatic collections are plotted on the map of eastern Asia (Fic. 4) and are listed below. A few collections were omitted, either because of uncertainty about the location or because only a gen- eral area was indicated. Symbols were placed in the center of a province when data or knowledge were insufficient to place them more accurately.

It is concluded, in agreement with Pojarkova, that the small-flowered plant occurs from northern China to Siberia and that the large-flowered plant is dominant in southern China. The small-flowered plant seems to extend as far south as approximately 27° N, well into the range of the large-flowered entity, and a few intermediates were found in the zone of central China in which both entities occur.

Pollen of 34 herbarium specimens, 27 of these wild Asiatic collections, was examined, and in none was the pollen 100% normal. The samples

1962] MOORE, CARAGANA SINICA 207

consisted of mixtures of normal grains with full, densely staining proto- plasts; grains full but with a thinner, lightly staining content; grains only partially filled with a thin protoplast; grains empty and collapsed; and empty micro-grains. Difficulty was experienced in classifying some grains since the first two classes intergrade. Only grains with a full, dense protoplast were classed as normal in the following counts. The micro-grains probably were formed from lagging chromosomes excluded from the tetrad nuclei and indicate a high degree of meiotic irregularity.

The maximum percentage of normal pollen found among the herbarium specimens was 80, the minimum, 3; 12 lay in the range 0-39%, 14 had 40-50%; 8 had over 50%. The size of the normal grains ranged from 18 to 30 » but this range was not found in any single sample. The modal range was 20-40 » (13); in 6 samples the lower limit was 18-20 p», and the upper limit was lowered correspondingly. The range in 10 samples was from 20 p to 25-27 yp, and in 6 the smallest grains were 23 yp, the largest 27-30

Pollen measurements (unpublished) made on other Caragana species do not show a correlation between the diploid and tetraploid chromosome number and pollen size. The samples are almost invariably 100% normal. Pollen of seven diploid species falls in the range 20-27 »; two additional diploids measured 23-29 » and 28-30 p». Pollen of one tetraploid species (C. frutex) measured 20-27 yp, of another (C. spinosa) 24-34 pw. No correlation between pollen size, probable chromosome number, or mor- phological features is believed to exist in C. sinica, and the pollen size variability is attributed to gene differences between the various popula- tions. The significant feature lies in the consistent absence of wholly good pollen and the similarity of the condition of all samples. No dif- ference was found in the range of size or of fertility between specimens classed as C. sinica, C. ussuriensis, or intermediate. Pollen samples of the latter group fall in the range of 40 to 60% normal and thus were no more “hybrid” than those of the more typical groups. No certain correlation was found between the degree of pollen fertility and geographical loca- tion, although there may be a tendency for plants with the higher per- centages (70% plus) to occur on the extremes of the range (Northern China and Yunnan).

Caragana sinica (Buc’hoz) Rehder, Jour. Arnold Arb. 22: 576. 1941. Robinia sinica Buc’hoz, Pl. Nouv. Decouv. 24. ¢. 22. 1799. Flowers 27-30 mm. long, calyx 11 mm. or more, leaflets broadly obovate.

China. CHEKIANG: without location, Barchet 146 (us). FUKIEN: Diongloh Hsien, Chung 1239, Apr. 5, 1923 (A). Hope (Chili): Hsi Yu Ssu, Liu 2286, June 8, 1929 (A). Hunan: Tschangscha, Handel-Mazzetti 638, Apr. 14, 1918 (a). Hupeu: W. Hupeh, Wilson s.n., Apr. 1900 (us); W. Hupeh, Wilson 2203, May 1907 (cH); without location, Henry 5378 (cH), 3812 (us). Kansu(?): Fengwangschan, Forbes 113, Apr. 22, 1877 (A). Kriancsu: without location, Tsu 436, Apr. 26, 1920 (a): Yun Dai Shan, Nanking, Tso 79, Apr. 17, 1926

208 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLII

(a): without location, Faber s.2. (A). KwetcHow: Kweiyang, Teng 90079, Apr. 15, 1936 (A). YUNNAN: East of Tengyueh, Forrest 19343, _ Apr. 1921 (a, uS); Yunnanfu bei Puchli, Schneider 214, Mar. 7, 1914 (a); zwischen Ssiao und Schin lung, Schneider 4046, Mar. 9, 1914 (a, a - without ee Bonatt or Maire 7286 (us); Yunnanfu, Smith 1582, 8/3, 1922 (aA)

Cultivated. Japan. Hondo: Musashi, Teizo a ee y 2, 1904 (A); without location or collector, May 13, 1910 (us 1311798); Nagasaki, Maximowicz s.1., 1863 (cH). FRANCE: Paris, Gay, s.n., 1822 (GH). Bavaria, Scherzer s.1., Apr. 1, 1906 (pao). Unrrep States: Washington, D.C., Apr. 24, 1886 (us 137850); Washington, D.C., 1915 (us 786433); Chico, Calif., Apr. 25, 1922, S.P.I. 22981 (originally collected in Soochow, Kiangsu, China) (A). CANADA: Royal Botanic Gardens, Hamilton, Ont., Rhodes & Florian 3587, June 11, 1952 (pA0); Dominion Arboretum and Botanic Garden, Ottawa, Ont. Woore, 53- 249-1, May 27, 1960 and Moore, 60-231-45, May 29, 1961 (pao); Experi- mental Station, Morden, Man., Rhodes & Vitens 4583, Sept. 10, 1953 (DAO).

Caragana ussuriensis (Regel) Pojark. Flora U.S.S.R. 11: 395. 1945.

ses frutescens 2 ussuriensis Regel, Mem. Acad. Sci. St. Petersb. ser (4): 44. 1861.

Flowers 23-26 mm., calyx to 9 mm., leaflets narrowly obovate with cuneate base.

China. Hopen (Chili): Western a eee 3 21361, May 7, 1929 (A); Nankow, Peiping, Chiao 21253, May 11, se Hunan: Yi Chang District, Tsang 23430, Mar. 91-29, 1934 ay. ee URIA: Ussurl super, Maximowicz s.n., 1860 (us). NortH CHINA: without a. Bunge s.u. (GH).

Cultivated. JAPAN: without location, Zuccarini s.n., 1842 (GH).

Specimens intermediate between Caragana sinica and Caragana ussuri- ensts.

China: Hupen: W. Hupeh, Wilson 2203, May 1907 (us); Patung, W7lson s.m., Apr. 1907 (A). Kranosr: Kipkiang, Bullock 151, Apr. 20, 1892 (us). TFu- KIEN: Diongloh Hsien, Chung 1239, Apr. 5, 1923 (A)

CULTIVATED CARAGANA SINICA

The five plants purchased from the Grootendorst nursery are identical in appearance, doubtlessly having been propagated asexually from a single plant. Flowers are 30 mm. long, with the calyx 11-12 mm. long and 6 mm. broad. The leaflets are broadly obovate, to 20 mm. long, 9 mm. broad, coriaceous, and glaucous. On the basis of flower and calyx size and leaflet shape, the plants are to be classed as the larger-flowered species C. sinica sensu Pojarkova. In all characters they are sidictineuichable from collections of other cultivated and wild Asiatic plants.

Cytotocy. The chromosome number was determined from leaf squashes to be 2n = 24 (Fic. 3). The basic number of the genus is 8: hence this number is triploid. Meiosis was studied in one plant growing outdoors at Ottawa. The configurations at Metaphase I could be fully analyzed in

1962 | MOORE, CARAGANA SINICA 209

only eight cells but these seem to be representative. The minimum pair- ing seen was 16 univalents and 4 bivalents; 8 bivalents and 8 trivalents were seen in two cells. The trivalent was the largest association found, and three trivalents was the maximum number found in a single cell. The average pairing for eight cells was 8! + 6.71 4+ 0.86UI

Approximately 40% of mature pollen grains are normal in appearance. Grains judged to be normal have a full, dense protoplast and measure 20-24 » in outer diameter. Pollen was spread on a mixture of 2% agar plus 5% sucrose, and germination of at least 50% of these full grains was observed. This agar medium has been found to give good germination of pollen of many species of Caragana.

FLowER Biotocy. At Ottawa, the shrubs flower from late May to mid-June, meiosis occurring in buds 5-6 mm. long in the period May 10-17. Over the past ten years it has been observed that varying weather conditions cause little variation in the onset of flowering in Caragana species, at most, two to three days.

Small aborted pistils were noted, but probably not over 10% of the flowers are defective. No other abnormality was observed. The style elongates markedly and extends approximately 2 mm. beyond the keel at anthesis. This feature favors cross-pollination and probably makes insect action necessary even for self-pollination. The plants studied were located approximately 100 feet from other Caragana bushes (C. arbo- rescens, C. frutex, C, aurantiaca) and separated by various other trees and shrubs. It is highly probable that the 5 plants of C. sinica were not pollinated from other species. Most flowers dropped without ovary en- largement; some showed slight enlargement but dropped in two to three days. A smaller number, estimated at 8-10%, developed conspicuously enlarged ovaries (25-28 mm.) which turned green and for a week ap- peared to be forming seed. However, all turned brown and fell. Some of the shrubs have been observed for three years and no seed has been formed. Plants from the same source grown at the Experimental Farm, Morden, Manitoba also are sterile (personal communication). Pollen of C. sinica was used to pollinate shrubs of C. arborescens, C. microphylla and C. frutex but no seed resulted. Two of 38 flowers of C. frutex volli- nated by C. sinica showed slight ovary enlargement before ¢:°:.j:ng.

OvuLE Histotocy. Ovaries of various sizes, pre- and post-flowering, were fixed for histological study of the ovules. Several different types of ovule development were observed, but in all cases all ovules within a single ovary were similar.

Ovaries which lacked ovules or contained abnormal and partially de- veloped ovules were found. Undeveloped ovules consisted of a small nucellus partially enclosed by a tissue two or three cells thick which presumably represented the outer, or perhaps both integuments. This integument did not enclose the micropylar end of the nucellus. No sporo- genous tissue was differentiated and all tissues appeared unhealthy. Such

210 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLII

ovules were found in mature open flowers, a stage when mature embryo sacs should have been present.

Structurally normal ovules were found in ovaries of some mature flowers. These ovules, approximately 525 » long and 375 » broad con- sisted of two integuments and nucellus, all apparently normal. Cell for- mations suggestive of aborted embryo sacs were observed in the nucellus. A linear formation of three compact cells, the micropylar one of which was the largest, was probably the product of meiosis. The innermost of the three was the most healthy and normally probably would be the func- tional megaspore. A large two-nucleate cell surrounded by disorganized tissue seemed to be an aborted early embryo sac. A four-nucleate cell (75 w long, 15-20 » broad) also was apparently an immature embryo sac. The most advanced structure seen was interpreted as a disintegrating mature embryo sac. This structure was 125 » long, 20 » broad and con- tained five recognizable nuclei and remnants of at least two antipodal cells. In all the above examples both sporophytic and gametophytic tis- sues were obviously in an unhealthy condition and incapable of continued development. Many ovules were so disorganized that the stage of embryo sac development could not be determined.

The large ovaries mentioned above (to 28 mm.) contained ovules that were enlarged by some 50% beyond the size of the most normal mature ovules. However, none of these contained embryos. Counts of the cell layers and measurements of representative cells indicated that ovule enlargement was due solely to cell enlargement in both the integuments and nucellus. By the time such ovaries drop, breakdown of the nucellus is advanced. A patch or narrow zone of empty collapsed cells in the area in which an embryo would normally be found is first seen. Dissolu- tion of cell contents and collapse of the walls progresses through the middle of the nucellus from the micropylar to the antipodal end. It was usually observed that the two-celled peripheral layer of the nucellus at its micropylar end remained longest in apparently healthy condition. A zone of the inner integument adjoining the micropylar end of the nucellus appeared more active than the remainder of the integument endodermis. These observations together suggest that the integument cells at this point were absorbing the contents of the nucellus. In the final stage of ovule collapse only remnants of walls remained of the nucellus. No evidence of apomictic seed formation was observed and there is no reason to be- lieve that these plants would ever set seed through either sexual or apomic- tic processes.

DISCUSSION

The occurrence of a major amount of aborted pollen in the living cul- tivated (Grootendorst) plant and in all known collections of Caragana sinica from the wild suggests that the species is of hybrid origin. Whether the species in nature is triploid, like the cultivated plant studied, cannot be decided with the available evidence.

1962 | MOORE, CARAGANA SINICA vans

Morphology of the species alone suggests a hybrid origin. That the leaf, which is pinnate with two pairs of leaflets, frequently appears to be palmate due to the lack of elongation of the rachis suggested to Komarov that the species was ancestral to both the pinnate and the regularly palmate series. An alternative explanation of this phenomenon is that C. sinica is a hybrid between a pinnate- and a palmate-leaved spe- cies. The variable-leaf type occurs otherwise only in the series SpINOSAE and DAsyPHYLLAE, groups of central Asia morphologically very unlike C. sinica. It seems obvious that these do not bear on the present problem.

All authors agree that Caragana sinica is most closely related to C. rosea Turcz. The affinity appears in the number, shape and texture of the leaf- lets, the persistent spiny leaf rachis, the large flower (more than 2 cm.) which is pale yellow with a rosy tinge, and the large calyx which is longer than broad. The rose flower color is not known in other species of eastern China, Caragana rosea has a range in eastern Asia (Manchuria, Hopeh, Honan, Kansu, Chekiang [Rehder, 1926]) very like that of C. sinica.

If the hybrid nature of Caragana sinica is accepted, C. rosea must be proposed as one parent. A pinnate-leaved species of eastern China is re- quired as the other parent, but the exact species cannot be named with an equal degree of assurance. The large flowers of C. sinica, larger even than those of C. rosea, point to another large-flowered species as the second parent. Such is to be found in C. microphylla Lam. (flowers 25 mm., calyx 9-12 mm.) a species of suitable leaf type and range as well.

It is therefore postulated that Caragana sinica, sensu lato (including C. sinica (Buc’hoz) Rehd. and C. ussuriensis (Regel) Pojark.) is a hybrid between C. rosea Turcz. and a pinnate-leaved species, probably C. mi- crophylla Lam. In the opinion of the author it would be preferable to recog- nize the entities simica and ussuriensis at infraspecific level under C. sinica (Buc’hoz) Rehd. It seems obvious that they have shared a common origin and are separated by characteristics of a minor order, quantitative rather than qualitative. The two populations have achieved geographic separa- tion, and taxonomic recognition is justified.

A variation within Caragana rosea in flower and calyx size and leaflet shape, parallel to that used by Pojarkova to split C. sinica was noted in thirteen specimens (US) examined. These could be divided into large- flowered (flower 25-27 mm., calyx 9-11 mm.) and small-flowered (flower 20-24 mm., calyx 6-8 mm.) plants. The more rounded leaflet shape was not invariably associated with the greater flower size, nor was there an evident geographical correlation. The range in pollen size in C. rosea was 20-27 ww. The variation within a single plant was not more than 4 np. However, no correlation between larger pollen size and flower size was found. The existence in C. rosea of variation of the same type as that seen in C. sinica is an additional indication of their close affinity.

The chromosome number of the single available accession of Caranga rosea was found to be 2x = 16. This diploid number is surprising, since it might be expected that this species would be tetraploid like the closely related C. frutex. The latter was reported to be tetraploid (2n = 32) by

Zig JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIII

Tschechow (1930, as C. frutescens), and only this number has been found in nine accessions in the present work. Caragana rosea appears to be mor- phologically a more advanced species than C. frutex. The persistent thickened petiole seen in C. rosea is apparently developed from the de- ciduous petiole of C. frutex. Throughout the genus there is an evolu- tionary trend toward transformation of the deciduous petiole, first, to a persistent but essentially unthickened petiole and, finally, to a persistent and much thickened organ. The latter is the stout spine seen in many species. This trend is found in both the pinnate- and the palmate-leaved series.

The lower chromosome number of Caragana rosea can be reconciled with the view that the species is derived from C. frutex by postulating that diploid populations of C. frutex do or did exist. It is, indeed, not improbable that the series FRUTESCENTES arose at the diploid level from a pinnate-leaved ancestor. All pinnate-leaved species yet examined are diploid (Moore, 1958 and unpublished). The series FRUTESCENTES Kom. em. Pojark. contains six species which extend from the Black Sea to central Mongolia. Three other Chinese species not treated in the Flora of the U.S.S.R. but which should doubtless be referred to the series, ex- tend the range of the series across northern China and Manchuria to the Pacific. Caragana frutex has by far the largest range of any single species (Black Sea to northwest Mongolia). Four species of relatively narrow distribution are found in Central Asia, south of Lake Balkhash. It seems possible that ancestral diploid ‘‘frutex” has spread from central Asia east- wards, developing in the most eastern part of the series range into C. rosea. The morphologically more primitive “frutex” stock may have become autotetraploid in central Asia, and, enjoying an advantage, accom- plished the present wide distribution, particularly to the west and north from Central Asia. It seems probable that Caragana frutex in cultivation has come from the European part of the range, the western extreme, and thus is tetraploid. Diploid populations may still exist in Central Asia. Unfortunately the chromosome numbers of the more limited species of Central Asia are unknown. The range of C. rosea is second only to that of C. frutex, suggesting considerable age for the former species.

The triploid number of the Grootendorst plant of Caragana sinica sug- gests a hybrid origin between a diploid and a tetraploid species. Two serious difficulties at once arise. The tetraploid number is not known in either postulated parent species. A triploid hybrid, if at all like the Grootendorst plant, will be seed sterile, and it seems impossible that the extensive range of C. sinica was accomplished by any means other than by seed dispersal. Moreover, it is known that the species in China does set seed.

These difficulties may be relieved by suggesting that Caragana sinica in eastern Asia is diploid, a hybrid between diploid C. rosea and C. mi- crophylla. A reduced, but still appreciable, seed fertility will then be pos- sible. The triploid condition of the Grootendorst plant may have arisen in cultivation. The plants are morphologically indistinguishable from wild

1962 | MOORE, CARAGANA SINICA 213

collections, and it is therefore improbable that a cross with a different tetraploid species has occurred in cultivation. The triploid condition may, owever, have developed from the fertilization of an unreduced egg. In- deed, the different climactic conditions of Europe acting on a somewhat unstable hybrid genome may have caused the formation of an unreduced gamete. By random chromosome segregation, the additional haploid set could add an equal number of chromosomes from each parent to the diploid hybrid complement, converting some former bivalents into triva- lents. The presence of equal numbers from each parental species might maintain the gene balance to such a degree that the triploid appears iden- tical with the diploid. The absence of a noticeable difference in pollen grain size between diploid and triploid may be disregarded since no cor- relation between chromosome number and pollen size has been observed in the genus, and it is believed that pollen size depends solely on the genes governing this character.

It may be pointed out in conclusion that the widespread occurrence of pollen abortion in Caragana sinica is explained better by the hybridity hypothesis than by the prototype hypothesis of Komarov. It seems improbable that an ancestral stock would have retained a condition of abortive pollen, presumably due to meiotic irregularities, which had been eliminated from its descendants.

SUMMARY

It has been possible to obtain Caragana sinica (Buc’hoz) Rehd. from only one source in cultivation. The plant is triploid (2n = 24), pollen is 40% normal, no seed is set. All herbarium specimens, of wild and cultivated collections, have partially aborted pollen. It is postulated that C. sinica is a hybrid between C. rosea Turcz. and probably C. microphylla Lam. The hybrid in nature is probably diploid; the triploid cultivated clone may have arisen in Europe by the production of an unreduced gamete. The chromosome number of C. rosea Turcz. is 2n = 16

The author wishes to express appreciation to the curators of the herbaria from which material was borrowed and particularly to members of the staff of the Arnold Arboretum and Gray Herbarium with whom the author has corresponded. Dr. J. L. Thomas, formerly of the Arnold Arboretum, was most co-operative in making observations and procuring specimens and cuttings of the Caragana rosea in the Arboretum. At the author’s request, Mr. H. J. Grootendorst looked into the source of the Caragana stnica sold by their nursery. Dr. T. Koyama, University of Tokyo, has informed the author that collections of C. sinica from Japan are undoubt- edly introductions.

214 JOURNAL OF THE ARNOLD ARBORETUM [ VOL, XLIII

BIBLIOGRAPHY

BRETSCHNEIDER, E. 1898. History of European botanical discoveries in China, Vol. 1. London.

KOMAROV, yi L. 1908. Generis Caraganae monographia. Acta Horti Petrop. 29(2): 177-388.

Loupon, J. C. 1844. Arboretum et fruticetum Britannicum. Ed. 2. Vol. 2. London.

Moore, R. J. 1958. The status of Caragana Boisti, Baileya 6: 188-193.

Poyarkova, A. I. 1945. Caragana. In: Flora of the U.S.S.R. 11: 327-368. Moscow and Leningrad.

Renper, A. 1926. Enumeration of the ligneous plants of northern China III. Tour Arnold Arb. 7: 164-65.

. 1941. New species, varieties and combinations from the collections of the Arnold Arboretum. Jour. Arnold Arb. 22: 569-579. [Caragana sinica (Buc’hoz), comb. nov., pp. 576, 577.]

TscHECHOW, WL. 1930. arvelocicel systematische Untersuchung des Tribus Galegeae, Fam. Leguminosae (Vorliufige Mitteilung). Planta 9: 673-680.

1962] PACLT, NEW GENUS OF SCROPHULARIACEAE 215

SHIUYINGHUA, A NEW GENUS OF SCROPHULARIACEAE FROM CHINA

J. PACLT

IN HER MONOGRAPH of the genus Paulownia Sieb. & Zucc. (Scrophu- lariaceae), Hu (1959, p. 47) excluded P. silvestrii Pampanini & Bonati (1911) from that genus and transferred it to Catalpa Scop. (Bignoniaceae). I have had the opportunity of examining a photograph of the type speci- men, P. C. Silvestri 3286 (PLATE I), which shows characters of a tree clearly different from Catalpa. The characters which do not permit one to classify this plant as a member of the genus Catalpa may be summarized as follows:

1. The conspicuously dimorphic shape of leaves which are broadly ovate (cordate) in the axial position and elliptically lanceolate in the abaxial position on flowering branches.

2. The flowering branches which bear both leaves and almost laterally situated (axillary) loose cymes of flowers, as in Paulownia fortunei (Seem.) Hemsl. and substantially all other species of Paulownia. In Catalpa, the inflorescences are formed terminally and correspond to racemes or true panicles (thyrses).

3. The calyx which is five lobed and patelliform, as in Paulownia but not Catalpa in which the calyx is gamosepalous, splitting into two strongly convex lobes at anthesis. Also, the flower buds are generally oblong in Paulownia silvestrii, instead of showing the typically subglobular shape of the bud of Catalpa.

Although the fruit of Paulownia silvestrii is unknown, the original description of this species by Pampanini and Bonati leaves, in my opinion, no serious doubts about its correct position in the system. This place is to be found in the tribe PAULOWNIEAE Pennell of the family Scrophu- lariaceae.

However, the general shape and consistence of the flower buds of Pau- lownia silvestriti alone separate this species sufficiently from all known members of Paulownia. This character also seems to be Hu’s basic reason for excluding P. silvestrit from Paulownia. In addition, her analysis of a flower bud of P. silvestrii revealed another difference concerning the structure of the young stigma (Hu 1959, p. 47). Some apparently less important differences between P. silvestrii and the other species of Pau- lownia may be found in the general appearance of the flowering branches and in the size of flowers. Accordingly, Paulownia silvestrii is best inter- preted as the type of a distinct new genus which I have the pleasure of naming in honor of Dr. Shiu-ying Hu, of the Arnold Arboretum, to whom botany is indebted for her contributions to the knowledge of Chinese plants.

Jour. ARNOLD Ars. VoL. XLIII PLATE I

j

HoLotyPE OF PAULOWNIA SILVESTRII (Silvestri 3286)

1962 | PACLT, NEW GENUS OF SCROPHULARIACEAE ral

Shiuyinghua, gen. nov.

Plantae lignosae foliis simplicibus oppositis vel etiam ternatim verticilla- tis. Inflorescentia axillaria cymosa. Alabastra oblonga. Calyx lobis 5 membranaceis vel + crassis. Corolla gamopetala quinquefida tubo inflato, lobis subaequilongis. Stamina didynama basi tubi inserta. Stigma (in alabastro visum) bilamellatum. Fructus ignotus. — Hab.: Asia temperata.

PUS GENERIS: Shiuyinghua silvestrii (Pamp. et Bonati), comb. nov. Pau- oe suvestri Pampanini et Bonati in Pampanini, Nuov. Giorn. Bot. Ital. II. 18: 177. 1911. Catalpa silvestrii (Pampanini et Bonati) S. Y. Hu, Quart. Jour. Taiwan Mus. 12: 47. 1959

TERRA TypICA: China, provincia Hupeh, praeter ripam fluminis VYang-tze Kiang (et praecipue fluvii Han Kiang), alt. 700 m., 20-30 Juni 1907. P. C. Silvestri 3286 in hb. Fr (et fragmentum, a).

The new genus Shiuyinghua becomes now the second known genus of the tribe PAULOWNIEAE. However, another genus, Wighktia Wallich, is sometimes considered to belong here also. Wightia is likewise a wood genus and is represented in the Himalayan, Burmese, Chinese (Yunnan), Vietnamese (Tonkin), and Malayan floras. The relatively limited knowl- edge of the morphology of Shiuyinghua does not make it possible to elaborate for the time being more than the following key to the practical identification of the three genera.

Stamens surpassing the top of corolla; calyx entire, tight-fitting. .... Wightia. Stamens not surpassing the top of corolla; calyx with 5 + outstanding lobes. Flower buds oblong and slender, with stigma ee bilamellate at this stage of immaturity; corolla not longer than 3 cm. ........ Shiuyinghua. Flower buds broadly ovate and robust, with a appearing punctiform at this stage of immaturity; corolla at least 3 cm. long, mostly much |e Nec mm sien A Orn ICR An Ae A a ev _ Paulownia.

My warmest thanks are due to Professor Richard A. Howard, director of the Arnold Arboretum, for his valuable help in sending me a photo- graph of the type specimen for study, as well as to Dr. Carroll E. Wood, Jr., editor of this journal, for his very kind criticism.

Poe INSTITUTE,

K ACADEMY OF SCIENCES, on CZECHOSLOVAKIA

218 JOURNAL OF THE ARNOLD ARBORETUM [ VOL, XLIII

PHENOLOGY OF TROPICAL PINES? -

NicHo.as T. Mirov

IN THE TEMPERATE ZONE pines shed their pollen during the season of the year vaguely designated as spring. At that time pollination occurs, and the ovulate strobili begin to develop slowly and, in warmer parts o the zone, continue to develop throughout the ensuing winter.* Ovules are fertilized the next spring, the cones develop throughout the second grow- ing season, and the seeds ripen in the fall. Time elapsed between pollina- tion and ripening of the seed thus is equal to about 15 or 16 months and embraces two summers and one winter. The farther south, the earlier pine pollen flies. Of course, many local environmental and genetic factors determine the time of pollen shedding. Pinus radiata D. Don sheds pol- len in March both in its naturalvenvironment on the Pacific Coast and in low elevations of the Sierra Nevada. It is interesting that when Pinus radiata is planted in the southern hemisphere, for example, in New Zealand where it is known as Pinus insignis Douglas, it sheds pollen when it is spring there, i.e., during August or September.

It should be noted that photoperiod sensu stricto does not affect flower- ing of pines; they are photoperiodically neutral. That is, when a northern (long day) pine is moved to a more southern (shorter day) location, its flowering pattern is not changed. When a tropical (short day) pine is cultivated in a more northern nee? day) latitude, it continues to flower as freely as in its southern home.*

The closer to the Equator, the more distorted is flowering in pines. Even in the southern parts of the United States, pines, for instance Pinus elliottii Engelm. in southern Texas or in northern Florida, shed their pol- len sometimes as early as the end of January. When you go farther south to the highlands of Mexico, early ‘flowering’ of pines becomes a wide- spread phenomenon, and its relation to the four seasons of the year be- comes really distorted.

I had occasion to observe Pinus oocarpa Schiede at the southernmost limits of pine distribution in Nicaragua. It was on a south slope of the mountains at an elevation of 4000 feet above sea level. It was the middle of February; the trees had just completed blooming (probably at the end of January), and numerous female strobili were still pink and tender, just having passed their ‘receptive stage.” But the trees also possessed many full-sized cones, still green in color but already containing ripe seeds.

* Regarding aecargt of the term “tropical pines” see my paper on “Some taxonomic problems of tropical pines,” Proceedings, 13th Congress of the International Union of Forest een Organizations, Vienna, 1961. (In press.)

* Gifford, Ernest M. Jr., and N. T. Mirov. Initiation and ontogeny of the ovulate strobilus in ponderosa pine. Forest Sci. 6: 19-25. 60.

®Mirov, N. T. Photoperiod and flowering of pines. Forest Sci. 2: 328-332. 1956.

1962 | MIROV, PHENOLOGY OF TROPICAL PINES 219

Squirrels were busy cutting the cones and eating the fresh seeds. There it was evident that the timetable of events leading to the production of seeds was somewhat distorted.

Springtime in Nicaragua is not an upsurge of life as in the North; tropical pines never cease to grow. We cannot make the statement that in the mountains of Nicaragua it takes two growing seasons, or two calendar summers, for seeds to mature. The ovulate strobili continue to develop, apparently without much winter slowing, for there is no win- ter; and it takes them only a little over one year to mature. That is why a Honduran botanist told me once that in his country it takes only one year for Pinus oocarpa to produce seed.

In Indonesia, late in February and early in March of 1961, I observed even more distortion in the flowering of pines. The pine there was Pinus merkusit De Vriese, moved from the mountains of northern Sumatra (about 3° N. Lat.) to the mountains and lowlands of Java (about 6° S. Lat.). I am not familiar with the flowering habits of this pine in Sumatra, but I suspect, judging from its performance in the mountains of northern Thailand, that it sheds pollen in January.

In the mountains of Java (elev. 4900 ft.) near Bandung, Pinus mer- kusit sheds its pollen twice a year: in January-February, and in July— August. Better seeds are obtained from the latter pollination,

A 20-year-old pine plantation at sea level was visited February 25, 1961. The forest ranger procured phenological records taken for ecvera! years. These records showed that Pinus merkusii pollen had been pro- duced and dispersed intermittently all year round; the ovulate strobili emerging from the buds, as well as the mature cones, were also recorded throughout the year. But Pinus kasya Royle, a pine of Burma and Indo- china, growing naturally at elevations higher than those of P. merkusii, neither produced pollen nor developed ovulate strobili in the plantation. Apparently high air humidity is detrimental to the normal seed produc- tion of this pine in the humid and hot lowlands of Java.

These cursory observations suggest that a more comprehensive study of the phenology of tropical pines would be interesting and profitable. Both Central America and Indonesia are well suited to such a study. In Central America there are several institutions in Honduras, Nicaragua and Guate- mala where such work could be done. In Sumatra, where the southern- most of all pines, Pinus merkusti grows naturally (about 2° south of the Equator) travel is at present hazardous; but in Java there are many easily accessible pine plantations where phenological records have been diligently kept. There a phenological project could be conducted either in the world renowned Herbarium Bogoriense, in Bogor, where the Forest Research Institute is also located, or in the Division of Biology of the Institute of Technology at Bandung

CABoT FOUNDATION, HaArvarD UNIVERSITY

220 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIII

ON THE STATUS OF PSILAEA (THYMELAEACEAE)

Lorin I. NEVLING, JR.

THE GENUS PsILAFA was described, with a single species P. dalber- gioides, by Miquel in 1861! (FI. Ind. Bat. Suppl. 355). The full com- bined generic-specific description was based upon a single Teysmann col- lection from “Sumatra occid., ad littus prope Siboga.” Nine years later, Kurz (Jour. As. Soc. Beng. 39(2): 83) placed Psilaea dalbergioides as a synonym of Linostoma pauciflorum Griffith, and, consequently, Psiaea was treated as a generic synonym of Linostoma. In 1922, Psilaea was used by Hans Hallier (Med. Rijksherb. 44: 28) to typify his section Psiaea of Linostoma. The typification of the section, based on Miquel’s taxon, is indicated by parenthetical credit i.e., “Sect. Psilaea (Mig.) m.” Un- der section Psilaea, Hallier described two new species, L. leucodipterum and L. longiflorum (both synonymous with L. pauciflorum

Following Kurz’s publication, in which he expressed the belief that Linostoma pauciflorum and Psilaea dalbergioides were conspecific, it appears as though subsequent authors accepted his determination with- out re-examining the ee specimen. In my recent revision of Linostoma (Jour. Arnold Arb. 42: 295-320. 1961) I placed P. dalber- gioides and the names _ based ae it in a category of undetermined status. This procedure was followed because the Teysmann type appar- ently never had been re-examined taxonomically in the one hundred years since Kurz’s publication; Kurz recognized only a single genus (Linostoma) where today we recognize two (Linostoma and Enkleia); recent revision of the generic limits of the two genera has produced some shifting of species from one genus to the other; and, finally, the genus Enkleta is known from several collections from Sumatra, but Linostoma is not (ex- cluding the oe type), although it is found on the neighboring Malay Peninsula

To resolve my “hesitancy to accept the validity of Kurz’s determination of the Teysmann type required first locating and then examining the specimen. This proved more difficult than anticipated. Many of Miquel’s types (particularly Teysmann specimens) are deposited currently at Utrecht. The specimen was not located at Utrecht or at any of the other major American or European herbaria. However, from Kurz’s brief remarks on Psilaea it is evident that he had seen authentic Teysmann specimens during his curatorship (1864-1878) at the Royal Botanic Gar- den, Calcutta (now Indian Botanic Garden); and in addition, Teysmann Sumatran material is deposited at Calcutta (see M. J. van Steenis-Kruse- man, Flora Malesiana 1: 525. 1950). After considerable effort, it was established, through Dr. H. Santapau, S.J., that the Teysmann specimen

' For date of issue see Steenis, C. G. G. J. van, Flora Malesiana 4: ccii. 1954.

1962 | NEVLING, STATUS OF PSILAEA 221

was extant, and a black and white photograph of it has been sent to me through the courtesy of Mr. Basu, the specimen itself being judged too fragile to be loaned. It is unquestionably the holotype (Teysmann s.n.) of Psilaea dalbergioides, and, further, is referable to Linostoma pauciflorum Griffith. Thus, Kurz’s identification is confirmed and the faith with which subsequent authors have accepted his determination is justified. A photo- graph of the holotype has been deposited in the herbarium of the Arnold Arboretum.

The full synonymy of Linostoma and of L. pauciflorum is emended to read:

Linostoma Wall. ex Endl. Gen. 331. 1837 (Type: L. decandrum (Roxb.) Wall. ex Endl.).

Linostoma Wall. Cat. n. 4203. 1831, sine descript.

Linostoma sect. Eulinostoma Meissn. in Mart. Fl. Bras. 5(1): 72. 1855 (Type: L. decandrum (Roxb.) Wall. ex Endl.).

Psilaea Miquel, Fl. Ind. Bat. Suppl. 355. 1861 (Type: P. dalbergioides Miq.).

Linostoma subg. Nectandra Kurz, Jour. As. Soc. Bengal 39(2): 83. 1870 (Type: L. decandrum (Roxb.) Wall. ex Endl.).

Linostoma sect. Psilaea Hallier f. Med. Rijksherb. 44: 28. 1922 (Type: Psilaea dalbergioides Miq.).

Linostoma pauciflorum Griffith, Calcutta Jour. Nat. Hist. 4: 234. 1844 (TypeE: Griffith 4376).

Psilaea dalbergioides Miquel, Fl. Ind. Bat. Suppl. 355. 1861 (Type: Teys- MN SN. Linostoma leucodipterum Hallier f. Med. Rijksherb. 44: 28. 1922 (Type:

Hallier f. B.2 Linostoma ee eae Hallier f. ibid. 29 (Type: Haviland 1759).

JosEpH Horace FAULL

JOURNAL

OF THE

ARNOLD ARBORETUM

VoL. XLII JuLy 1962 NUMBER 3

JOSEPH HORACE FAULL, 1870-1961

ANNA F. FAULL With portrait *

JosEpH Horace Fautt was born in L’Anse, Michigan, on May 3, 1870, the eldest son of James and Catherine (Bennetts) Faull. He died June 30, 1961, at his home in Cambridge, Massachusetts, Professor of Forest Pathol- ogy, Emeritus, at the Arnold Arboretum, Harvard University. His studies of plant pathogens had led him from the Arctic Circle to the Tropics in North America, filled his classrooms with students, and won for him world- wide recognition as an authority on forest diseases and in particular the rust fungi.

His early life and training can be summarized quickly. Born in the pine forests of Michigan, he grew up as a minister’s son on the nothern shore of Lake Ontario. Here he often drove long distances through the farm and forest lands with his father talking about Charles Darwin whose theories on evolution James Faull was inclined to accept. He began to teach before finishing high school, interspersing study with teaching until he had earned three Normal School certificates and the first degree in Arts from the Uni- versity of Toronto, where he had enrolled at Victoria College with the class of 1898. From 1898 to 1900, he taught mathematics in Belleville, On- tario, at Albert College, where he had completed his preparation for the University some years earlier.

He married Annie Bell Sargent, of Bellwood, Pennsylvania, late in 1903. They had met during the summer in Cold Spring Harbor, Long Island, New York, where he replaced Albert F. Blakeslee as assistant to Dr. Dun- can S. Johnson, of Johns Hopkins University in the summer botany course. Miss Sargent, as a candidate for the doctorate, was studying life-histories of spiders. She had obtained her A.B. degree from the University of Penn- sylvania, as one of the first small group of young women ever to be admitted to the regular courses of study at that institution. Together until her death in 1953, she shared with him an unabated interest in natural history, as well as a great love for students and all children. Of their own three, Cath-

* This portrait was taken by J. Horace Faull, Jr., in 1949 at the Pathological Labo- ratory of the Arnold Arboretum, Jamaica Plain, Boston, Massachusetts.

224 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIII

erine Sargent died at the age of sixteen, an invalid following an attack of influenza in early infancy; Anna Forward followed her father into botany; while J. Horace, Jr., has made a name for himself in chemistry.

Dr. Faull’s career in botany began in 1900 when he returned to the Uni- versity of Toronto to study with Edward Charles Jeffrey, Lecturer in Biol- ogy, whose sectioning of lignites (coal), along with the application of Dar- win’s theory of evolution to research on the anatomy of woody plants, was already attracting attention. Mr. Faull’s problem was the vascular struc- ture of the Osmunda ferns. His first paper was published by the Botanical Gazette in 1901. More important, Dr. Jeffrey told him about William Gilson Farlow and Roland Thaxter at Harvard, the two American pioneers in the new field of mycology.

Mr. Faull went to Harvard in 1901-1903 as an Austin Fellow to begin under Professor Thaxter the studies on the cytology of the Ascomycetes which were to be the major subject of his published research until 1913. His papers on the formation of the ascus and its spores are still standard references in this field. In 1904, he received his doctorate from Harvard.

In 1903, Dr. Faull had already accepted a lectureship in botany at the University of Toronto, the same lectureship (under a changed name) left vacant by Dr. Jeffrey’s appointment to Harvard. He remained with his wife and family in Toronto until 1928, with sabbatical leave in 1909-1910 for a brief visit to Harvard and a year’s study abroad, mostly with Robert Hartig at the Forst Botanische Institut of the University of Munich, Ger- many. Botanically he found the trip abroad somewhat disappointing, al- though he profited in prestige, in an increased facility with the German language, and in a deepened interest in forest pathology. He had been made an associate professor at Toronto in 1907. In 1918, this was changed to a full professorship while he continued as head of the department which he had created. Before he left, his department taught students in the gen- eral (liberal) arts, applied science, household science, pharmacy, and forestry; his classrooms were crowded; his graduate students and assistant staff were increasing in numbers, and a new building for botany alone was being proposed in the Provincial Parliament.

Summers at the University, with an early closing date in April or May, were long and, for Professor Faull, varied. He continued his research with publications on the Ascomycetes and other fungi, a natural history of the Toronto region and a paper on Charles Darwin. He built a large collection of lantern slides, photographs, and preserved material for teaching. He attended courses in forestry at the College of Forestry, Syracuse Univer- sity, New York. He taught in the summer schools at Cornell University, Ithaca, New York (1915-17), and at Harvard (1925). He also taught his own students in Toronto. He collected in the Adirondacks with G. F. Atkinson, in the Allegheny Mountains of Pennsylvania with his wife (1913-1917), in the country a few miles or more from his Toronto home, and in the forest lands of Ontario and Quebec (Timagami, Algonquin Park, the Rideau Lakes, Georgian Bay, the Laurentians). Wherever he went he collected fungi and pathological specimens. These impressive

1962] JOSEPH HORACE FAULL, 1870-1961 225

collections of teaching, research, and reference material are one of his legacies to botany.

Apart from the University and at the request of the Canadian Govern- ment, Dr. Faull established a department of forest pathology with a field laboratory in the coniferous forests on Bear Island in the Lake Timagami Forest Reservation of Ontario. Commencing in 1918, much of his summer time was spent in the development of this laboratory, in long-range experi- ments there, and with one of his students, Wade Watson, in the compilation of a check list of the Timagami flora. At the same time his services as a consultant were in increasing demand by lumber companies and others with forest or tree problems. These requests, although remunerative, were to him often annoying interruptions to his major interests at the University and in the field laboratory. When he left, the Government laboratory and department, as well as a shade-tree laboratory, were as firmly established as the department at the University, while his students were beginning to appear as responsible men and women in botany and forestry in the uni- versities, schools, and government laboratories and departments through- out the Dominion. He had indeed earned the citation of “Father of Cana- dian Botany” given him in 1959 at Montreal in the opening address at the IX International Botanical Congress.

In 1928, he came to Harvard University as Professor of Forest Pathol- ogy to take part in the expanding program of research envisioned by Charles Sprague Sargent and Oakes Ames for the Arnold Arboretum. For Professor Faull this meant freedom from the increasingly heavy teaching and administrative load at Toronto and a chance to devote most of his time to writing, research, and the advancement of forest pathology. He brought with him his collections of research and reference material, his graduate students and his reputation as a consultant. The Arboretum provided a laboratory and greenhouse built to his specifications on the adjacent Bussey Institution grounds; the Farlow Herbarium gave workin space there; and, later, after their completion, the Biological Laboratories provided additional laboratory rooms,

Professor Faull’s conception of the new appointment and of forest path- ology can best be stated in his own words taken from letters to Edward C. Jeffrey and to Oakes Ames at this time. He wrote: “My mind is set on advancing Forest Pathology — completing and writing up accumulated researches of which there is a good stock in hand, taking up others in mind, organizing the position at the Arboretum on a broad and sound basis and perhaps working towards a treatise of the Hartig type — these are possibilities within reasonable expectation of attainment.”’ Again, “The field is a broad one for it involves the study of the etiology of arboreal dis- eases and of the principles of their control. As for etiology only a begin- ning has been made in America; there exists a host of diseases the causes of which are yet unstudied and unknown, and in several instances that come to mind wrongly ascribed. Intelligent control is possible only when causes are known, but even then there are principles to be worked out. This is particularly true of our untamed and abused American forests

226 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIII

where the solution seems to lie along prophylactic lines frequently based on a knowledge, mostly not yet acquired but acquirable, of the relation of the age of the hosts to susceptibility, of the rate of progress of the patho- gens and the relative amount of destruction caused. There is likewise a phase in connection with the utilization of immense quantities of waste diseased living timber. All these aspects constitute a vast field within which researches may be undertaken and if you subscribe to these limits your project will be as broadly based as the one in taxonomy which has been so long and so successfully developed.”

Scarcely pausing to settle into the new laboratory, Dr. Faull began at once the pregram that he had laid out for himself: the completion of earlier research, the study of “immediate pathological problems presented by trees in the Arboretum” and “of any or all diseases of all kinds of trees wherever found and whether in plantations or in the forests.” The latter was to lead him far afield from one end of the continent to the other, while his days in Cambridge were filled with writing, research, students, and visitors from all over the world. One can trace his collecting and travel in his accessions books,

The first summer found him in Timagami in August checking on the experiments he had last seen in June just before coming to Harvard; re- turning to old locations at Proulx, Quebec; and extending his range into the Gaspé. A month later he was in Maine examining nursery spruce and white pine for the Brown Company. His first paper appeared in the Jour- nal of the Arnold Arboretum in 1929, ““A Fungous Disease of Conifers Re- lated to Snow Cover.” It was based on the work at Proulx, Quebec, begun some years earlier, and on observations made at the Brown nurseries.

From 1929 to 1933, he continued this program of revisiting old locations and experiments, adding new ones, advising and co-operating in experi- ments with the lumber companies and others. He also gave the undergrad- uate half-course in pathology agreed upon. In 1929, he made collections at the Arnold Arboretum; at the Kelsey nurseries, in Boxford; at Oqussoc, in Maine; and again at Proulx, in Quebec, and at Lake Timagami. He extended his trips into Nova Scotia at the request of the Provincial forester where he notes finding “Milesina polypodophila (Bell) Faull on Abies and Polypodium side by side.” He repeated the June trips in the fall and again in 1930. He visited the West and the Pacific Coast in 1931, collecting in the Yellowstone National Park, in Wyoming; at the North Priest River Forest Station, in Washington; in Oregon; in California; in Idaho; and in Illinois. He concentrated on the New England mountains in 1932: Ver- mont, the Presidential Range, the Berkshires. In 1932-1934, the Arbore- tum published the first of his monographs on rusts, that on Milesia.

From 1933 to 1940, Dr. Faull included a winter trip to the tropics in his regular program for the year. In December and January, 1933-34, he visited the mountains of Jamaica, Panama, and the Canal Zone. He re- turned to Jamaica in February and March, 1935, and again in 1935-36, after a three week stop in Cuba. He wrote in the records of the Harvard Botanical Garden at “Soledad,” near Cienfuegos, Cuba: “Arrived Dec,

1962 | JOSEPH HORACE FAULL, 1870-1961 rs

17, 1935. Departed Jan. 5, 1936. Activities: 1) Made a pathological re- connaissance of the Arboretum and of the forest tree plantations at an near Soledad. 2) Examined critically a disease of Marabu [Dichrostachys cinerea] between Soledad and San Blas, a disease of Ficus in Soledad and several heart rot diseases of trees; assembled some relevant material. 3) Collected Pucciniastreae in the Trinidad Mts.” The following winter he went to Guatemala where, in addition to rusts, he collected material of an Abies growing at 10,000 feet which Rehder later described as A. guatema- lensis. In November, 1937, he was in Mexico with Professor Maximino Martinez collecting in the remote areas of the states of Michoacan, Mexico, Morelos, Hidalgo, Veracruz, and Chiapas, as well as in the Federal District. He returned to Mexico in 1938 to visit Oaxaca and in 1939 to collect again in Chiapas and Hidalgo, as well as in the northern states of Chihuahua and Durango. In the meantime, his summer trips continued: the Gaspé, Maine, New Brunswick, Long Island, and Chicago, in 1933; Quebec, Mt. Washing- ton and Maine, in 1934; Maine and Metis, Quebec, in 1936; Vermont, Maine, and Long Island, in 1937; Vermont and Connecticut, in 1938; New Hampshire, in 1939; Massachusetts every year.

In 1938, the Arboretum published the second of his monographs on rusts, that on Uredinopsis. Actually, this was the third of his long papers on the rust fungi dealing with the taxonomy, morphology, physiology, host relations, etc., which he considered not only knowledge of general botanical interest but basic for control of plant diseases specifically or in principle. In 1926, two years before his Harvard appointment, he had presented a long paper on the Puccineastreae at the International Congress of Plant Sciences, Ithaca, New York. While not themselves “treatises of the Hartig type” contemplated in his letter to Dr. Jeffrey, they undoubtedly were working towards it. One more monograph was intended but never written, although he had collected the material and begun the necessary studies. Like all of his work, these papers are the scholarly presentations of careful study and experiment in field and laboratory by a man with a capacity for detail combined with breadth of vision and a keen insight.

Simultaneously with the field work, Dr. Faull organized the laboratory at the Arboretum. By September, 1928, the small building was ready with its own library “of several thousand pamphlets,” periodicals, and reference books; the important collections of diseased plants (about 1000 speci- mens); and the usual apparatus for work with cultures and microscope. Addition of the experimental greenhouse in 1929 or 1930 completed the small laboratory, while more space for students was provided as needed at the nearby Bussey Institution, at the Farlow Herbarium in Cambridge, and, later, in the new Biological Laboratories there. On July 1, 1929, Dr. Faull reported on a reconnaissance of the living collections of the Arbore- tum, in addition to five major projects under active investigation by him- self, his assistant (G. D. Darker, from Toronto), and the first of his Har- vard graduate students: (1) rusts of fir and spruce, (2) Phacidium blight and snow cover, (3) lilac disease, (4) needle-cast diseases of conifers, (5) browning of white-pine transplants. The first two were a continuation of

228 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLII

work begun at Toronto, in the Timagami forests, and at Proulx, Quebec. The last two came from field work with lumber companies in Maine or from general collecting. The lilac was an Arboretum problem.

By 1931, the pathological laboratory envisioned by Oakes Ames and Dr. Faull and established in fulfilment of “the expressed wish of the late Di- rector, Charles Sprague Sargent, and as part of Dr. Sargent’s conception of the Arboretum as an institution for the study of woody plants in all fun- damental aspects” was in full swing with the functions of the laboratory defined “as comprising interest in the Arboretum’s living collections, ex- tension services, instruction and research.” With the propagating houses next door and the living collections of the Arboretum itself within hand’s reach it was ideally located both for his own researches and for friendly co-operation with the rest of the Arboretum group. Contact with the under- graduate College was maintained through the half-course in plant patho- gens, while graduate students, postgraduates, and scholars continued to come in twos and threes and sevens. The United States and foreign govern- ments from Canada to Jamaica, estate owners, forest interests, and others called with increasing frequency for extension services. This, in turn, of- fered not only new problems but often supplementary financial support for their study and for the postgraduate work which Dr. Faull encouraged his students to do both in his own laboratory and abroad.

For the next ten years the laboratory continued to operate under Dr. Faull’s direction with the same efficiency and devotion. The problems brought in by field work, collecting trips, extension services, and the Ar- boretum itself were numerous and specific. Many of them were quickly answered. But many involved much broader botanical issues: “the host of unknown and unstudied” causes of plant disease; ‘the principles of control yet to be worked out”; the relationships of host, parasite, and en- vironment; morphology; anatomy; physiology; and even some aspects of genetics — in short, the entire range of botanical science. At least twenty- two such studies were carried through to the publication of significant con- clusions, a broadly based beginning in the “vast field” of unstudied ar- boreal pathology for which the laboratory had been organized. Of most popular interest is the Dutch elm disease, which Dr. Christina Buisman, of Holland, found for the first time in the United States at the end of a year’s work (1929-30) on American elm-diseases. Early hope of eradication faded when some years after its elimination from Ohio it reappeared in New Jersey; thereafter, the Arboretum effort was directed at control. Equally or more important both botanically and practically were other studies: a disease of Fagus, lilac blight (a graft incompatibility leading to work on viruses), taxonomy of Ganoderma (initially a question of classi- fication), morphology of rust spermogonia as taxonomic criteria, my- corrhiza (morphology and physiology involving mineral nutrition of pine), wilt diseases of elm (other than Dutch), etc. The record of the laboratory can be traced in the annual reports and papers published in the Arbore- tum’s Journal or as one of its Contributions, a series inaugurated by two papers from the pathology laboratory: Dr. Darker’s Ph.D. thesis on

1962] JOSEPH HORACE FAULL, 1870-1961 229

needle-cast fungi and Professor Faull’s monograph on “Milesia.” That the laboratory was not continued and his position remained unfilled after his retirement was a disappointment to him.

Professor Faull retired in 1940, a world figure in forest pathology and the world’s authority on rust fungi associated with ferns. His students have gone on to make names for themselves in the botanical world and to serve in responsible teaching, administrative, and research positions both in the United States and Canada. The extension services have been taken over mostly by government agencies, but the broad and comprehensive kind of research begun at the Arboretum seems not to have been continued in this country.

For several years, Dr. Faull himself remained in his laboratory, assisting with Arboretum affairs, helping in the fight against Dutch elm disease, serving as an associate editor of the Journal, organizing his collections, ad- vising at the Farlow Herbarium, and welcoming the scientists who came to see him. The visits of two of these gave him particular satisfaction. One was that of Professor Maximino Martinez, of Mexico, who had ar- ranged for his earlier collecting in the Mexican forests and had accompanied him on some of the trips. The other was Dr. Krishnadas Bagchee, from the Forest Service in India, who shared Dr. Faull’s love of the forests and his interest in fungi. Dr. Bagchee brought with him a collection of Indian rusts to my father’s delight. In 1956, the Journal of the Arnold Arboretum published his last paper. Shortly thereafter, Dr. Faull gave up his last small room in the Cambridge laboratories.

The proper disposal of his large collections of pathological material occu- pied much of his time in these later years. Some, of course, had been left at the University in Toronto. But the thousand specimens brought with him to Harvard had grown through exchange, communication, and his own collecting trips and field work and that of his students and associates to nearly fourteen thousand. The small room where the earlier collections had at first been kept had become inadequate long before his retirement. In 1939, he had placed a thousand duplicates of wood destroying fungi in the Farlow Herbarium and three hundred in the Bureau of Plant Industry in Washington (later transferred to the National Herbarium). In 1940, he noted in his report the deposition of his large collection of polypores at the Farlow Herbarium. Finally before his death, he placed a collection of fern rusts (type specimens and material documenting his publications) in the U.S. National Herbarium, where he hoped it would be both adequately cared for and available for reference and future investigations. The large remainder of the collections with duplicates has been placed in temporary storage at the Farlow Herbarium and the Arnold Arboretum awaiting more permanent quarters. A few are still at his home in Cambridge.

As his health failed, Dr. Faull spent more and more of his time with his family, partly in Texas, but most of it at their home in Cambridge. Here he enjoyed his garden, his neighbors, a little carpentry, a great deal of reading, a little cribbage, and a voluminous correspondence with friends and students dating back to those whom he had taught as little children in

230 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIII

the Canadian rural schools. As a student at Toronto, he had been the gold medalist in his graduating class of 1898. Now he reread the entire works of Dickens and Shakespeare, along with his other books. As a young man, too, he had played ice hockey on the Victoria College team. Now he listened to the sports broadcasts on the radio. He never lost his interest in the schools or in the education of women or his sympathy for the under- privileged. At his death, a volume of Dickens was near his chair by the radio along with a copy of the Cambridge School Report, a box of fern rusts near his desk.

As a teacher and a scientist, Joseph Horace Faull enriched the two coun- tries in which he lived. His students continue to do so.

Professor Faull had been a member of the American Academy of Arts and Sciences, the American Association for the Advancement of Science, the American Phytopathological Society, the American Society of Natural- ists, the Mycological Society of America, the National Agricultural Chemi- cals Association, the New England Botanical Club, the Royal Canadian Institute, the Royal Society of Canada, the Society of American Foresters, the Society of the Sigma Xi, and the Sociedad Botanica de México.

72 FRESH Ponp LANE, CAMBRIDGE, MASSACHUSETTS

BIBLIOGRAPHY *

— 1901 — The anatomy of the Osmundaceae. Bot. Gaz. 32: 381-420. pl. 14-17,

— 1905 — Development of ascus and spore formation in Ascomycetes. Proc. Boston Soc. Nat. Hist. 32: 77-113. pl. 7-11, Diseases of timber. Canad. Forestry Jour. 1: 105-108, 3 pl.

— 1906 —

Further studies on ascus. (Report of paper read at the annual meeting of the Central Botanists held at Ann Arbor, Michigan, December 28, 29, 1905. Abstracted by B. M. Davis.) Science 23: 134.

A preliminary note on ascus and spore formation in the Laboulbeniaceae. Sci- ence 25: 157, 153,

— 1907 — Bunt, or the stinking smut of wheat. Canada Dep. Agr. Seed Branch Bull. $3. 13 pp.

— 1908 — Arceuthobium pusillum Peck. Ottawa Nat, 21: 175. Notes on Rondeau Park. Ontario Nat. Sci. Bull. 4: 99-103.

* Compiled by Lazella Schwarten; with supplementary references added by the author from Professor Faull’s own files and from those of the Farlow Herbarium.

1962] JOSEPH HORACE FAULL, 1870-1961 231

10002 The influence of Darwin on botanical science. Ontario Nat. Sci. Bull. 5: 31-37. Stele of Osmunda cinnamomea. Trans. Roy. Canad. Inst. 8: 515-534. pl. 4-6.

— 1911 — The cytology of the Laboulbeniales. Ann. Bot. 25: 649-654.

— 1912 — The cytology of Laboulbenia chaetophora and L. gyrinidarum, Ann. Bot. 26: 325-355.

— 1913 — The natural history of the Toronto region, Ontario, Canada. 1-419. Toronto.

— 1914 — Bark disease of the chestnut in British Columbia. (With G. H. Graham.) Forestry Quart. 12: 201-203.

— 1916 — Chondromyces thaxteri, a new myxobacterium. Bot. Gaz. 62: 226-232. Peed, 0. Fomes officinalis (Vill.) —a timber destroying fungus. Forestry Quart. 14: 737-739 Fomes officinalis (Vill.), a timber destroying fungus. Trans. Roy. Canad. Inst. Toronto 11: 185-209. pl. 18-25.

— 1918 — The menace to our white pine. Canad. Forestry Jour. 14: 1685-1687. The fight to save our white pine. /bid. 1743-1747.

— 1919 — Forest pathology. Rep. Minister Lands, Forests & Mines. Prov. Ontario 1919: 11 25.

Pineapple fungus or enfant de pin or wabadou. Mycologia 11: 267-272.

— 1920 — Forest pathology. Rep. Minister Lands, Forests & Mines. Prov. Ontario 1920: 224-235.

— 1921 — Plant pathology; its status and its outlook. (Presidential address.) Trans. Roy. Soc. Canada III. 14(sect. v): 1-16.

— 192 Records for four years on the needle blight of Pinus strobus. Phytopathology 1] .

Some problems of forest pathology in Ontario. Needle blight of white pine. Jour. Forestry 20: 67-70. — 1923 — Balsam rusts. Rep. Minister Lands & Forests. Prov. Ontario 1923: 253, 254.

252 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLII

— 1924 — The aecial stage of Hyalopsora psi (Peck) P. Magnus. (With G. D. Darker.) Phytopathology 14: 3 Stereum sanguinolentum as the cause of ‘“‘sapin rouge” or red heart rot of bal (With Irene Mounce.) /bid. 349. The treatment of decayed wood in and outside the mill. Pulp & Paper Mag. 3pp. Feb. 28.

— 1928 — Living cells in heart-wood of trees. Science 67: 296, 297.

— 1929 — A a disease of conifers related to the snow cover. Jour. Arnold Arb. 10:

The ina biology and ae of the Pucciniastreae. Proc. Int. Congr. Pl. Sci. 1926. 2: 1735-17

— 1930 —

Blister rust in Nova Scotia. Blister Rust News 14: 63.

The health of the forests. Illus. Canad. Forest & Outdoors 26: 146-149. (March 1930.

Notes on forest diseases in Nova Scotia. Jour. Arnold Arb. 11: 55-58.

Some general remarks regarding forest pathology in relation to forestry and notes on forest diseases in Nova Scotia. Rep. Dep. Lands & Forests Nova Scotia 1930; 33-40.

The spread and the control - the Phacidium blight in spruce plantations. Jour. Arnold Arb. 11: 136-14

— 1931 — Milesina rusts on Aspidium braunii Spenner. Jour. Arnold Arb. 12: 218,719;

— 1932 — “Hubert, E. E., An outline of forest pathology.” (Review.) Phytopathology 22: 393-395, Taxonomy and a ar distribution of the genus Milesia. Contr. Arnold Arb, 2: 1-138. pl. 1 231984 <3

Arthur Bliss Seymour (1859-1933). Proc. Am. Acad. Sci. 69: 543, 544.

The biology of Milesian rusts. Jour. Arnold Arb. 15: 50-86. pl. 84-86,

Blister rust of Pinus longifolia Roxb. Ibid. 154-157.

A remarkable spruce rust, Peridermium parksianum n. sp. Ibid. 86, 87.

Winter hardiness of trees and shrubs growing in the Arnold Achavetuni, (With

. G. Jack, W. H. Judd, and L. V. Schmitt.) Arnold Arb. Bull. Pop. Inf.

IV, 2: 29-47, 53-60.

Wehmyer’s “The genus oo Nitschke and its segregates.” (Review.) Jour. Arnold Arb. 15: 157—

— 1935 — Can we eradicate the Dutch elm disease? Address before the Annual Meeting

of the Massachusetts Forest and Park Association, January 31, 1935. 4 pp., illustr, Mass. Forest and Park Assoc., Boston

1962] JOSEPH HORACE FAULL, 1870-1961 233

— 1936 — Pathological studies on beech at the Arnold Arboretum. Proc. Natl. Shade Tree Conf. 12: 21-29 Two spruce-infecting rusts, Chrysomyxa piperiana and Chrysomyxa chiogenis. Jour. Arnold Arb. 17: 109-114 The viewpoint of the Arnold Arboretum on the Dutch elm disease. Arnold rb. Bull. Pop. Inf. IV. 4: 15-20. 1 pl

— 1937 — Chrysomyxa empetri—a spruce-infecting rust. Jour. Arnold Arb. 18: 141-148. pl. 202, 203.

New England i in autumn array. Yankee 3(10): 26, 27.

— 1938 — The biology of rusts of the genus Uredinopsis. Jour. Arnold Arb. 19: 402— 436. The Dutch elm disease situation in the United ee at the close of 1938. rnold Arb. Bull. Pop. Inf. IV. 6: 75-78. pl. 7 Pucciniasirem on Epilobium and Abies. Jour. Rerold Arb. 19: 163-173. my and geographical wee of the genus Uredinopsis. Contr. Ar- ane Arb. 11: 1-120. pl. 1 — 1939 —

A review and extension of our knowledge of Calyptospora goeppertiana Kuehn. Jour. Arnold Arb. 20: 104-113.

— 1942 —

Report of Northeastern Committee on Dutch elm disease. (With C. C. Hamil- ton, D. S. Welch, H. H. York, and J. S. Boyce.) 4 pp. Mass. Forest and Park Assoc., Boston.

— 1947 — Tropical fern hosts of rust fungi. Jour. Arnold Arb. 28: 309-319.

— 1956 — A rust on Woodwardia fimbriata. Jour. Arnold Arb. 37: 314-316.

234 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIII

COMPARATIVE ANATOMY OF THE LEAF-BEARING CACTACEAE, V

THE SECONDARY PHLOEM Lauit M. Srivastava! ano I. W. BarLey 2

IN AN EARLIER PAPER (Bailey and Srivastava, 1962), it was indicated that the vascular cambium, tracheary tissue, and sieve elements in the leaf- bearing Cactaceae show structural features that suggest an advanced level of evolutionary specialization. The arrangement of fusiform initials in the cambium and their derivatives in xylem and phloem, and the structure of xylem tissue were described in detail in this paper. In contrast, the account of phloem was cursory and restricted mainly to the form of sieve elements and their sieve plates and sieve areas. In order to understand the structure of phloem tissue, particularly the relationship between sieve elements and parenchyma cells associated with them, an ontogenetic study of the tissue was considered essential. This study forms the subject of the present paper.

Three species of leaf-bearing Cactaceae, Pereskia sacharosa Griseb, | Tu- cuman ], Pereskiopsis aff. chapistle Britt. & Rose [Boke B-3], and Quiaben- tia aff. chacoensis Backbg. [Tucuman] were selected for detailed onto- genetic work. The material was killed in FAA. Small pieces from the stems of these species were embedded in paraffin and sectioned on a rotary micro- tome. Some additional material was embedded in celloidin and sectioned on a sliding microtome. Serial transverse, radial, and tangential sections were obtained and, later, stained by a combination of tannic se iron chloride, and lacmoid as described by Cheadle, Gifford & Esau (19

In order to determine the origin of phloem elements, several te (that is, radial files of derivatives of single cambial initials) were studied in de- tail for each of the three species. Tiers were drawn from serial cross sec- tions with the use of a camera lucida attachment. The cells in each tier were reconstructed from these drawings. The entire height of a tier was examined in order to determine the exact relationship between sieve ele- ments and the parenchymatous cells associated with them. The results obtained from a study of cross sections were later substantiated by a study of tiers from serial tangential sections. Radial sections were only of lim- ited use in the present study, partly because of the difficulty of obtaining good radial sections and partly because of the irregular planes of divisions in the phloic initials and the small size of some of the phloem derivatives.

As is well known, sieve elements and the associated companion cells

* Mercer Fellow of the Arnold Arboretum.

* This investigation was financed in part by a grant from the National Science Foundation. I am indebted to the American Philosophical eres for the loan of a Wild microscope.

1962] SRIVASTAVA & BAILEY, CACTACEAE, V 235

usually function for a limited period only. They die in old phloem and, in most cases, are eventually crushed by the expanding parenchyma cells and the pressures resulting from new growth (cf. Esau, 1953, pp. 299, 300). As a result, the arrangement of phloem derivatives, typical of young phloem, is distorted and the study of tiers cannot be pursued with accuracy in very old phloem. The tiers drawn in the present study, therefore, represent the functional phloem and only as much of nonfunctional phloem * as had not yet been distorted.

The terms used here are common in literature on phloem (cf. Esau, 1950, 1953; Esau & Cheadle, 1955; Cheadle & Esau, 1958), but some of them are explained again for clarity. A phloic initial is the daughter cell formed towards the phloem after a periclinal division in the cambial in- itial. The phloic initials either directly, that is, without any further divisions, or after a few divisions produce the various cell types in the phloem. For instance, a phloem-parenchyma strand is formed after one or more horizontal divisions in the phloic initial. The term precursor refers to a cell that would either differentiate directly as a definitive phloem element or in which further divisions would occur; in any case, it denotes an undifferentiated cell in which further changes are going to occur. Thus a phloic initial may behave as a precursor of a fusiform phloem-parenchyma cell, or one of the daughter cells after a division in the phloic initial may behave as a precursor of the sieve-tube mother cell and a parenchyma cell. The term sieve-tube mother cell refers to a cell in which one or more divisions occur and within the confines of which a sieve element and its companion cells are formed. Occasionally a sieve- tube mother cell may differentiate directly as a sieve element and no companion cells may be formed. The terms sieve element and sieve-tube member are used interchangeably. Various kinds of parenchyma cells occur in phloem. In the present paper, the terms used to describe them have an ontogenetic implication. Thus, a fusiform phloem-parenchyma cell is derived directly from a phloic initial; and a phloem-parenchyma strand is formed after one or more horizonta! divisions in the phloic initial. Some other parenchyma cells and companion cells are formed in association with a sieve element after divisions in a single phloic initial. For convenience of description, these cells ontogenetically related to the sieve elements are often referred to as parenchymatous cells or elements.

ORIGIN OF PHLOEM ELEMENTS IN PERESKIA Analysis of Tiers A cross section through the phloem of Pereskia sacharosa is shown in Fic. 32, in which the cell types present in the tissue and their general arrangement are seen. Another cross section is drawn in Fic. 1, but the individual tiers are separated tangentially for a better illustration of

° Nonfunctional, in contrast to functional, phloem refers to that part of phloem in which sieve elements are no longer conducting (cf. Esau, 1953, p. 299

236 JOURNAL OF THE ARNOLD ARBORETUM [ VOL, XLIII

Fic, 1. Pereskia sacharosa, transection of cambial zone (below) and phloem, < 495. Tiers separated from one another to show radial extent of each; sieve elements unmarked; companion cells stippled; parenchyma cells with nuclei; ray cells on extreme right without nuclei. Tiers A-E analyzed in detail in Fics. 2-6 and in text.

their radial extent. A study of similar drawings from serial cross sections reveals the axial extent of each tier and the cell types that compose it. Tiers A, B, C, D, and E in Fic. 1 are analyzed in detail. The cross sec- tion drawn in Fic. 1 did not pass through the middle (in terms of axial extent) of all the five tiers selected for study. The cross sectional views at different levels in the axial extent of these tiers are illustrated in Fics. 2-6, a—e; the levels at which these cross sections were taken are indicated along the margins of Fics. 2—6, f. In Fics. 2-6, a-e, sieve elements are un- marked except for sieve plates which, when present, are indicated by hatched areas; companion cells are stippled; and parenchyma cells onto- genetically associated with sieve elements are shown with nuclei. Fusi-

1962 | SRIVASTAVA & BAILEY, CACTACEAE, V 237

form phloem-parenchyma cells and phloem-parenchyma strands are shown with a circle drawn with a heavy pen. Cells internal to the cell complex in which sieve element 1 occurs are not considered — they are assumed to be undifferentiated cells. Some of them may have differentiated as fusiform phloem-parenchyma cells or phloem-parenchyma strands, but we were not certain about this. They are shown without nuclei. FIGURES 2-6, f, represent schematically the axial extent of the sieve elements and the parenchymatous elements ontogenetically associated with them in tiers A-E; the sieve elements are drawn with numbered solid lines, the companion cells with dotted lines, and parenchyma cells with broken lines. Lines representing companion cells and parenchyma cells are placed arbitrarily on the left and right, respectively, of the lines representing sieve elements. If a strand of companion or parenchyma cells is formed in association with a sieve element, it is represented by a single line; but the individual cells in the strand are demarcated by short oblique lines intersecting the vertical line that represents the strand. All cells that are ontogenetically related and derived from a single phloic initial are included within pairs of horizontal lines drawn at the upper and lower limits of the complex of cells. The numbers 0-225 or —300 at left in each drawing represent the length in microns. The purpose of this schematic representation is twofold. First, it shows at a glance how many and what kinds of cells are ontogenetically related with one another. Second, the combined length of the cells in a complex of cells, barring the overlap of cells, reflects the length of the phloic initial (and, hence, that of the fusiform cambial initial). The lengths of different phloic initials in a tier represent approximately the axial extent, or the height, of the tier. Fusiform phloem-parenchyma cells and phloem-parenchyma strands are not considered in Fics. 2-6, f.

Tier A: Seven sieve elements, indicated by arabic numerals, are pres- ent in tier A (Fic. 2, c, f). Sieve element 1 is associated with two com- panion cells and one parenchyma cell, sieve element 2 with one com- panion and one parenchyma cell, sieve element 3 with three companion and two parenchyma cells, sieve element 4 with two companion and two parenchyma cells, sieve element 5 with one companion and one paren- chyma cell, and sieve elements 6 and 7 with one companion cell each. Two fusiform phloem-parenchyma cells occur in the tier and were de- rived directly from their phloic initials.

Each of the sieve elements 1-5 with their associated parenchymatous cells and the sieve elements 6 and 7 with their companion cells originated by divisions within the confines of a single phloic initial. The origin of these elements is considered in detail.

The phloic initial within the confines of which sieve element 1 is present divided obliquely and longitudinally. Of the two daughter cells formed, the one towards cambium was the precursor of a parenchyma cell, the other away from the cambium was the mother cell for sieve element 1 and its companion cells. In this mother cell a longitudinal,

238 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLII

é' es ee ee a es es nt oom ba ae: oe ee | ee ee b ca eee Gee | ee “| {1 "| a yy : c 1 ae ee ee oe ee | no | real ; d at ae r an i | >a e iT | bho pe | | | = \" I oO —L f

Fic. 2. Pereskia sacharosa, Tier A of Fic. 1: a-e, cross sections of tier at levels indicated in schematic representation of certain members of tier (f) by arrowheads opposite a—a, b—b, etc., respectively; a-e, X 495.

For details of drawings in Fic. 2 and in Fics. 3-6, 10-14, 21-25, see text, p. 280;

more or less radial division separated a narrow cell from a larger cell. The narrow cell divided horizontally and formed a strand of two cells that differentiated as companion cells; the larger cell matured as sieve ele- ment

The sequence of divisions in the phloic initials responsible for sieve ele- ments 2 and 5 and their parenchymatous cells was similar to that in the

1962] SRIVASTAVA & BAILEY, CACTACEAE, V 239

phloic initial for sieve element 1—the first oblique longitudinal wall separated the precursor of a parenchyma cell from the sieve-tube mother cell, and a subsequent division in the sieve-tube mother cell produced a small and a large cell; the small cell differentiated as a companion cell, the large cell as a sieve-tube member.

The order of divisions in the phloic initials within the confines of which sieve elements 3 and 4 occur is more complicated. The cell com- plex with sieve element 3 is analyzed as follows. A more or less tangential division extending through most of the length of the phloic initial re- sulted in two daughter cells, one away from the cambium was the pre- cursor of a parenchyma cell (Fic. 2, a—e), the other towards the cambium was the precursor of a second parenchyma cell and the mother cell of sieve element 3. An oblique longitudinal division in the lower half of this latter precursor formed two daughter cells: one was the precursor of the second parenchyma cell (Fic. 2, d-e), the other, a larger cell, was the sieve-tube mother cell for element 3. A longitudinal division along the right radial and outer tangential walls of the sieve-tube mother cell produced a narrow precursor that divided horizontally to form a‘ strand of two companion cells (Fic. 2, a-e); another longitudinal division along the inner tangential wall, but confined to the upper half of the sieve- tube mother cell, produced a third companion cell (Fic. 2, a-c). The larger cell left after these divisions in the sieve-tube mother cell differ- entiated as sieve element 3. The following sequence of divisions is visualized for the origin of sieve element 4. Two successive oblique longitudinal divisions in the phloic initial, one along left radial and the other along right radial wall, separated two precursors of parenchyma cells from the sieve-tube mother cell. A more or less radial longitudinal di- vision in the sieve-tube mother cell produced a narrow cell, which di- vided horizontally and formed two companion cells (Fics. 2, a—-d), and a larger cell that differentiated as sieve element 4.

Sieve elements 6 and 7 are ontogenetically related. A more or less radial longitudinal division in the phloic initial separated two sieve-tube mother cells, each of which divided subsequently and formed a sieve-tube element and a companion cell.

Tier B: Twelve sieve elements (indicated by arabic numerals) with their associated parenchymatous cells and one fusiform phloem-paren- chyma cell occur in tier B (Fic. 3, c, f). The divisions in phloic initials responsible for sieve elements 3, 7, and 12 are least complicated and, hence, are analyzed first. A tangential longitudinal wall laid down in the phloic initial of sieve element 3 separated two daughter cells, one (to- wards cambium) was the precursor of a parenchyma strand, the other (away from cambium) was the sieve-tube mother cell. A radio-longi- tudinal wall divided the sieve-tube mother cell into a narrow precursor,

visions in the phloic initials of sieve elements 7 and 12 is similar except

240 JOURNAL OF THE ARNOLD ARBORETUM [ VOL, XLIII

that the precursor of companion cells did not divide horizontally. In the cell complex including sieve element 12, the precursor of the paren- chyma cell also did not divide horizontally to form a strand of paren- chyma cells (Fic. 3, f).

2257 | 2 3 4 8 6 7 89 10 I 2 ze 7 = a | Io: : ee a ae oe |

o4 oat Pfr port tbe

Bea Ee) a de ee

Oe fees eed ed dL

: : a is 1: a

d eV a) a Sign lS) oa ps

“| at GV ey ah

e | a aT bof) i} sft be ees | es eens) aes

re) et mes ee 4 —_J

f

Fic. 3. Pereskia sacharosa, TIER B of Fic. 1: a-e, cross sections of tier at levels indicated in schematic representation (f), as in Fic. 2; a-e, & 495. Oblique line connecting sieve elements 8 and 9 in diagram (f) shows that the two originated after division in a single mother cell. For details of drawings, see text

1962] SRIVASTAVA & BAILEY, CACTACEAE, V 241

Sieve elements 1 and 2; 4, 5, and 6; 8 and 9; and 10 and 11 are onto- genetically related and were formed with their associated parenchymatous cells after divisions in single phloic initials. One parenchyma cell occurs in common association each with sieve elements 1 and 2, 8 and 9, and 10 and 11. Sieve elements 4, 5, and 6 do not seem to have any parenchyma cell. The sequence of divisions in the phloic initials in all these in- stances is more complicated than in the case of sieve elements 3, 7, and 12. It appears that in the phloic initial responsible for the cell complex in which sieve elements 1 and 2 occur, an oblique anticlinal wall was laid down. Of the two resulting daughter cells, one towards the cambium, the precursor of a parenchyma cell and the mother cell of sieve element 1, divided obliquely, but the dividing wall did not extend to the cell tips (Fic. 3, f). The longer cell formed after this division was the precursor that matured as the parenchyma cell; the shorter was the sieve-tube mother cell for element 1. In this sieve-tube mother cell an oblique wall was laid down that separated a narrow precursor (left and above), which divided horizontally and formed two companion cells, from a larger cell that matured as sieve element 1. The cell away from the cambium, formed after the first anticlinal division in the phloic initial, was the mother cell for sieve element 2 and its companion cell. In this mother cell a more or less radial division separated the precursor of a companion cell from a precursor that matured as sieve element 2. In the interpretation of the origin of this cell complex, it has been assumed that the parenchyma cell was more closely related to sieve element 1 than to sieve element 2. It is possible, however, that the first division in the phloic initial of this cell complex was not the one that separated the sieve-tube mother cell of element 2 from the precursor of the mother cell of element 1 and the parenchyma cell, but it was one that separated the parenchyma cell from a precursor in which later divisions formed sieve elements 1 and 2. In the latter interpretation the parenchyma cell would be equally related to the two sieve elements.

In the analysis of the cell complex in which sieve elements 4, 5, and 6 occur, it appears that a more or less tangential longitudinal division of the phloic initial resulted in two daughter cells. The daughter cell to- wards the cambium was the sieve-tube mother cell of element 4; a radial longitudinal division in this mother cell followed by two transverse di- visions in the narrower of the two cells resulted in three companion cells and the sieve-tube member 4. In the daughter cell away from the cambium an oblique division extending through the upper half of the cell set off the sieve-tube mother cells of elements 5 and 6. The mother cell of element 6 was shorter than that of element 5 (cf. Fic. 3, a-f). A longitudinal division in the mother cell of element 5 produced a long, narrow companion cell and the sieve-tube member 5. Two successive divisions occurred in the mother cell of element 6 and formed two pre- cursors of companion cells and a precursor that matured as sieve element 6. In one of the precursors of companion cells a horizontal division re- sulted in two companion cells; the other precursor matured directly as

242 JOURNAL OF THE ARNOLD ARBORETUM [ VOL, XLII

a companion cell. As a result, sieve element 6 is associated with three companion cells.

In the phloic initial that formed the cell complex including sieve ele- ments 8 and 9, an oblique radio-longitudinal division formed two daughter cells. One was the precursor of a parenchyma cell; the other divided obliquely near its lower end and produced two mother cells, one each for sieve elements 8 (Fic. 3, a-d) and 9 (Fic. 3, d-e). An oblique di- vision in the mother cell of element 8 formed a narrow cell that differ- entiated as a companion cell and a larger cell that matured as the sieve element 8. The precursor of sieve element 9 differentiated as a sieve-tube member without any divisions. The sequence of divisions in the phloic initial within whose confines elements 10 and 11 appear is similar to that in the phloic initial of elements 1 and 2,

Tier C: Nine sieve elements with their associated parenchymatous cells and one fusiform phloem-parenchyma cell occur in tier C (Fic. 4, c, f). The origin of the cell complexes including sieve elements 5 and 9 is simple and followed the pattern described for sieve elements 1, 2, and 5$ in tier A, and 3, 7, and 12 in tier B. A strand of four companion cells was formed in association with sieve element 5S.

Sieve elements 1-4 and 6-8 with their associated parenchymatous cells were derived from single phloic initials. A new feature is seen in the origin of sieve elements 1-4. A more or less transverse division in the phloic initial formed two daughter cells. An oblique, more or less tan- gential division in the upper daughter cell formed two precursors, the one away from the cambium matured without any further divisions as sieve element 2, the other towards the cambium divided radially and formed a parenchyma cell and sieve element 1 (Fic. 4, a~-b). No companion cells were formed in association with either sieve element 1 or 2. In the lower daughter cell (formed after the first, transverse division in the phloic initial) an oblique longitudinal division separated the mother cell of sieve element 4 from the precursor of the mother cell of sieve element 3 and a parenchyma cell (Fic. 4, c-e). An oblique division in the mother cell of element 4 separated a narrow cell that differentiated as a com- panion cell (lower right corner) and a larger cell that matured as the sieve element 4. In the precursor of the mother cell of sieve element 3 and the parenchyma cell, a tangential division produced two daughter cells: one towards the cambium was the parenchyma cell, the other away from the cambium was the sieve-tube mother cell of element 3. In this mother cell a tangential longitudinal division separated a precursor of two companion cells from a precursor that differentiated as the sieve element 3 (Fic. 4, c-e). In the origin of sieve elements 6—8, two successive oblique divisions, one somewhat anticlinal and separating the phloic initial into an upper and a lower cell (cf. Fic. 4, a-e), the other some- what tangential and oblique and separating the lower cell into an outer and an inner daughter cell (Fic. 4, d-e), resulted in three mother cells, one each for elements 6, 7, and 8. The mother cell of element 6 was the

1962] SRIVASTAVA & BAILEY, CACTACEAE, V 243

TTL EPR

a b c d Oc. mca: OC oz | 23 4 § 6 7 8 9 210 2 =] Sa ae ee | ; | | aoe | a | qm | Pa | a on os | a a aL ‘ | |! a e Pg day int d ah <I! | : Si alert a ; : if E

f

Fic. 4. Pereskia sacharosa, T1ER C of Fic. 1: a-e, cross sections of tier at levels indicated in schematic representation (f), as in Fic. 2; a-e, X 495. Details as given in text.

longest and extended from the top end of the phloic initial to well below its middle, that of element 7 was in the lower half of the phloic initial but did not extend to its lowermost tip, and that of element 8 extended through lower one-third of the phloic initial (Fic. 4, f). Two successive longitudinal divisions along the outer tangential wall of the mother cell of element 6 produced two precursors of companion cells and a precursor that matured as sieve element 6. In one of the precursors of companion

244 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIII

cells a horizontal division yielded two cells. As a result, three com- panion cells were formed in association with sieve element 6. In the mother cell of element 7 a longitudinal division along the outer tangential wall produced a narrow precursor of two companion cells and a large cell that matured as sieve element 7. In the mother cell of element 8 two longitudinal divisions, one along the radial wall on left and the other along the outer tangential wall, produced two precursors of com- panion cells and a precursor that matured as sieve element 8. The pre- cursor of companion cells along the outer tangential wall divided hori- zontally, Eventually three companion cells were formed.

That the sieve elements 1—4 originated by divisions in a single phloic initial is shown clearly in Fic. 4, f, where the combined lengths of the parenchyma cell associated with sieve element 1 and the sieve element 3 (which are the longest derivatives of the upper and lower daughter cells formed after the first horizontal division in the phloic initial) approxi- mately equal the length of other phloic initials in the tier. Likewise, the combined lengths of elements 6, 7, and 8, barring the overlap, is the same as that of the phloic initials responsible for the complexes of cells in- cluding elements 5 and 9.

Tier D: Six sieve elements with their associated parenchymatous cells and one fusiform phloem-parenchyma cell occur in tier D (Fic. 5, c, f). Sieve elements 3 and 6 and their associated companion and parenchyma cells originated in the usual manner described for sieve elements 1, 2, and 5 in tier A; 3, 7, and 12 in tier B; and 5, and 9 in tier C, and are not considered. Sieve elements 1 and 2, and 4 and 5 are ontogenetically re- lated. In the phloic initial within the confines of which sieve elements 1 and 2 occur, a longitudinal wall was laid down. This wall has a some- what radial orientation in the upper half of the phloic initial (Fic. 5, a—c), but in the lower half it is somewhat tangential (Fic. 5, d) and finally oblique (Fic. 5, e). (It would appear that the resultant daughter cells were placed somewhat radially in relation to one another near their upper end and somewhat tangentially near their lower end. Such ar- rangements were commonly seen in tangential sections, cf. Fics. 7-9.) Of the two cells thus formed, one (below and right) was the mother cell of element 2. A longitudinal division along the outer tangential wall of this mother cell separated a narrow cell from a larger cell; the former divided horizontally thrice and formed a strand of four companion cells, the latter matured as the sieve element 2. In the other cell (above and left) was laid down a wall which had a tangential orientation in the upper end of the cell (Fic. 5, a-c) and a more or less radial orientation in the lower end (Fic. 5, d-e) and which separated the mother cell of sieve element 1 from a precursor of a parenchyma cell. A longitudinal division along the outer tangential wall of the sieve-tube mother cell of element 1 produced a short, narrow cell, which divided horizontally and formed two companion cells, and a larger cell that differentiated as sieve

1962] SRIVASTAVA & BAILEY, CACTACEAE, V 245

q b c d e | 2 3 4 5 6 ale es ee | * 150 Bln il Sfr 2 a a ae Sole: Se ~ eae an b 2 ee “| ‘tt aT > ¢ ae 2 ah alee lt be : TI d ae |! ae lf bd S : é : | e : 2 | | I De pi Ream | NG meni | re) Lae —l_.

5. Pereskia sacharosa, TIER D of Fic. 1: a-e, cross sections of tier at levels indicated in schematic representation (f), as in Fic. 2; a-e, X 495. De- tails as given in text.

element 1. Sieve elements 4 and 5 and their associated parenchymatous cells originated in a similar manner.

Trier E: In tier E there are eight sieve elements with their associated companion and parenchyma cells (Fic. 6, c, f). Sieve elements 1, 4, 5, and 8 originated in the usual manner —a tangential or a radial longi- tudinal division in the phloic initial separated the precursor of a paren-

246 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIII

TI cw - {I a | TB ge a ly ly Gf OM - 2K 2 pa oa ‘| tt yh 2 | : b br tt] | a il ee i eit at | :|! ili ‘ : i \ : : c TI ee | a be SJE sf s i 3 eS e at :| I | a ee bd ai ai. | Ae ie 4 : “TI Z| I | | 2 a be |] a ail cat ; i 0 hi i: rene

Fic. 6. Pereskia sacharosa, TrER E of Fic. 1: a-e, cross sections of tier at levels indicated in schematic representation (f), as in Fic. 2; a-e, X 495. Details as given in text.

chyma cell (or strand) from the sieve-tube mother cell. In the sieve-tube mother cell itself a longitudinal division separated the precursor of one or more companion cells from the precursor of the sieve-tube member. Sieve elements 2 and 3, and 6 and 7 are ontogenetically related and originated in the manner common to that of sieve elements 1 and 2, and 10 and 11 in tier B; and 1 and 2, and 4 and S in tier D.

1962 | SRIVASTAVA & BAILEY, CACTACEAE, V 247

Discussion

The five tiers studied from cross sections reveal some patterns about the method of origin of sieve elements and related parenchymatous cells in Pereskia sacharosa. In the following, this information is summarized and confirmed from a study of tangential sections.

In the five tiers examined, a total of 32 phloic initials was studied. Cells internal to the cell complex in which sieve element 1 occurred are excluded from this discussion. As mentioned earlier, they may have been fusiform phloem-parenchyma cells or phloem-parenchyma strands, but we were not certain that these cells had differentiated. Of the 32 phloic initials, 5 had matured directly as fusiform phloem-parenchyma cells: the remaining 27 phloic initials had divided in various planes and had formed cell complexes in which one or more members had differentiated as sieve elements. Most commonly, the first division in the phloic initial is longitudinal and radial, oblique, or tangential. The resultant daughter cells may be unequal in size, but one is the precursor of a parenchyma cell, the other the precursor in which the sieve element and its companion cells arise (sieve-tube mother cell). This method of the origin of paren- chyma cells, sieve elements and companion cells was encountered in 14 of the 27 silat initials studied (cf. Fics. 2—6,

Less commonly, the first division in the Snleie: initial, which again is longitudinal, but may vary from radial, oblique, to tangential, separates two daughter cells, one of which behaves like the phloic initial, the other like the sieve-tube mother cell mentioned in the previous paragraph. As a result, two sieve elements, at least one parenchyma cell (or strand), and some companion cells are formed within the confines of a single phloic initial (cf. Fics. 2-6, f). As mentioned during the analysis of the cell complex including sieve elements 1 and 2 in tier B, the sequence of the first two divisions is not very clear in such instances.

Occasionally, two successive oblique divisions in the phloic initial sep- arate three cells. The dividing walls may be oriented in different planes but usually one is somewhat radial, the other somewhat tangential. Also, one or both walls may not extend to the tips of the phloic initial, thus producing cells shorter than the initial. (Compare the lengths of sieve elements 3 and 4 and their associated parenchyma cells in Fic. 2, f; sieve elements 4, 5, and 6 in Fic. 3, f; sieve elements 6, 7, and 8 in Fic. 4, f.) Of the three cells that are formed, all three may behave as sieve-tube mother cells or one may behave as a sieve-tube mother cell and the other two as precursors of parenchyma cells, or strands. As a result, three sieve elements with their companion cells (sieve elements 4—6, Fic. 3, f; sieve elements 6-8, Fic. 4, f), or one sieve element with its companion cells and two parenchyma cells (sieve elements 3 and 4, Fic. 2, f) are formed within the confines of a single initial.

Three other patterns of divisions in phloic initials were recorded. A longitudinal division in the phloic initial may separate two cells. Each daughter cell may then behave as a sieve-tube mother cell (sieve elements

248 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLII

6 and 7, Fic. 2, f), or one may behave as a precursor of a parenchyma cell (or strand) and the other as the precursor of two sieve-tube mother cells (sieve elements 8 and 9, Fic. 3, f). Finally, a phloic initial may be di- vided almost horizontally into two daughter cells. The upper and lower daughter cells may then behave as more or less “independent” phloic initials producing sieve elements and parenchymatous cells in manners described above (sieve elements 1-4, Fic. 4,

In order to confirm some of the conclusions drawn from a study of cross sections, several tiers were also studied by use of tangential sections. As is clear from Fics. 2-6, a—e, the planes of divisions in the phloic initials and their daughter cells are often irregular, with the result that, even in almost perfect tangential sections, the outlines of cells appear at different levels of focus in the same section and even in different sections. An- other difficulty with tangential sections is that the tangential walls of derivatives (and, hence, the tangential limits of phloic initials) are not easy to determine. Because of these complications serial sections must be studied.

Three representative tiers are drawn in Fics. 7-9. Although partly corrected, these figures show the outlines of cells at different levels of focus as truly as was necessary for a clear understanding of the planes of divisions. Only a few derivatives in each tier are shown; the last cell in each set of drawings represents the fusiform cambial initial. The sieve cells are left unmarked, except for the sieve plates which are hatched; companion cells are stippled; parenchyma cells ontogenetically related to sieve elements are shown with nuclei; and parenchyma cells ontogenetically unrelated to sieve elements are shown with a circle drawn with a heavy pen.

A phloem-parenchyma strand occurs in Fic. 7, a. It was derived by a single transverse division in the phloic initial. The derivatives shown in Fic. 7, b-c, d; and at a—b, and c-d in Fics. 8 and 9, are ontogenetically related and originated after divisions within the confines of a single phloic initial. One parenchyma cell, a sieve element and a strand of two com- panion cells occur in the cell complexes shown in Fic. 7, d; Fic. 8, c-d; and Fic. 9, a—b, c-d. In the cell complex shown in Fic. 7, b-c, three sieve elements and their companion cells were formed; whereas in the cell complex in Fic. 8, a-b, two parenchyma cells were formed in associa- tion with the sieve element. The planes of divisions leading to the forma- tion of cell complexes shown in Fic. 7, d, Fics. 8 and 9, c-d, are easily understood. Others are somewhat more complicated but can be explained on the basis of our information obtained from the study of cross sections.

The wall separating the sieve element from the strand of two compan- ion cells in Fic. 9, a-b, appears radial at the lower end, tangential at the upper end. This is shown by the fact that these elements are placed radially to each other at their lower end but appear in different planes of focus near their upper end. Similar orientations of derivatives were en- countered in the study of cross sections (cf. cell complex including sieve elements 1 and 2 in Fic. 5, a-e). It would appear from these illustra-

d e

8 9

Fics. 7-9. Pereskia sacharosa: a-e, in each figure, successive phloem derivatives from outer to inner phloem (a-d) and fusi- form cambial initial of each tier (e), in tangential view, X 325. Derivatives at (a), (b-c), and (d) in Fic. 7, and those at (a—b) and (c-d) in Fics. 8 and 9 are ontogenetically related. Broken lines indicate cell outline seen more clearly at another fo- cal level. Details of figures and of Fics. 15-20, 26-31 given in text, p. 248.

A ‘AVAOVLOVO ‘AAD TIVE ® VAVLSVAIYS [Z961

6b~

250 JOURNAL OF THE ARNOLD ARBORETUM [ VOL, XLIII

tions that, during cytokinesis in phloic initials and their daughter cells, the cell plates may, in their upper and lower extensions, intersect different walls of the mother cell. Possibly, also, the cells formed after divisions within the confines of a phloic initial expand differentially in mutual adjustment with the related cells derived from the same phloic initial and with the cells in neighboring tiers, so that the shape of the mature cells and the orientations of the walls that separate them are different in final stages from those in early stages of ontogeny. These factors may account for the peculiar configurations of cells seen in Fic. 7,

and 9, a—b, and for the lateral intrusion of cells in one tier into neighbor- ing tiers (Fic. 1).

ORIGIN OF PHLOEM ELEMENTS IN PERESKIOPSIS AND QUIABENTIA

Tiers in Pereskiopsis aff. chapistle and Quiabentia aff. chacoensis were studied from serial cross sections in the same manner as in Pereskia sacharosa. Five tiers from each of these species are analyzed in detail in Fics. 10-14, a—d, and Fics. 21-25, a-d. The cross sectional views of the tiers and the schematic representation of sieve elements and _ related parenchymatous cells in them are drawn in the same manner as in Pereskia sacharosa. However, the cross sections are drawn only at three levels in the axial extent of each tier, because the difference in the struc- ture of the tier from one level to another was not very marked. These figures are mostly self-explanatory and do not need much comment. In the following, therefore, only the important points concerning the oc- currence of fusiform phloem-parenchyma cells and phloem-parenchyma strands, and the number and planes of divisions in phloic initials (and their daughter cells) preceding the formation of sieve elements and re- lated parenchymatous cells are mentioned.

Altogether, 18 phloic initials in Pereskiopsis aff. chapistle and 35 in Ouiabentia aff. chacoensis were studied in the five tiers examined for each species. (Cells internal to the cell complex including sieve element 1 in all tiers, except that in Fic. 24, are omitted from this consideration. In

considered as fusiform phloem-parenchyma cells, because the cells in neighboring tiers at that distance from the cambium appeared mature.) Two of the 18 phloic initials in Pereskiopsis aff. chapistle had given rise to one fusiform phloem-parenchyma cell (Fic. 13, a-c) and one phloem- parenchyma strand (Fic. 14, a-c), and 15 of the 35 phloic initials in Ouiabentia aff. chacoensis had matured as fusiform phloem-parenchyma cells (Frcs. 21-25, a-c). The remaining 16 phloic initials in Pereskiopsis aff. chapistle and 20 in Ouiabentia aff. chacoensis had formed cell com- plexes with at least one cell in each case differentiating as a sieve element.

As in Pereskia sacharosa, the most common mode of origin of sieve elements and related parenchymatous cells is one in which the phloic initial divides longitudinally and radially, obliquely, or tangentially. Of

—

1962 | SRIVASTAVA & BAILEY, CACTACEAE, \V 251

fe) d Ke)

300 l 2 3 4 Aue? ald mane al 5 eee ti Tk alt ota fee ac aaa |.

i al ot b 4 lh ly :{/ De aT Ne | Seek an elie. a

| ml ase ae

a ee ge ip 2

aa

6) d II Fics. 10, 11. Pereskiopsis aff. chapistle: a-c, cross sectional views of two tiers at levels shown in diagrams (d), schematic representations of some mem- bers of the two tiers, as in Fic. 2; a-c, * 390. See explanation in text.

252 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIII

the two resultant daughter cells, one cell behaves as a precursor of a parenchyma cell, or strand, the other cell as a sieve-tube mother cell. Such an origin of sieve elements and related parenchymatous cells was recorded in 8 of the 16 phloic initials in Pereskiopsis aff. chapistle (Fics. 10-14, d) and 15 of the 20 phloic initials in Quiabentia aff. chacoensis (Fics. 21-25, d).

~~ ] |e

b < bb c 4 4 A 0 d l2 I 2 3 2257 el mM a |! be | af | 1] J a} all b 4 ah ti pb af ar “| I “| ats TY c 4 i 2 ‘|i be =e an o! at d

Re) Fics. 12, 13. Pereskiopsis aff. chapistle: a—c, cross sectional views of two tiers at levels shown in diagrams (d), schematic representations of some members of the two tiers, as in Frc. 2; a-c, & 390. For details, see text.

1962] SRIVASTAVA & BAILEY, CACTACEAE, V 253

an i 2 3 qytoqo7 a 4 | ‘ | pa | -[I | le tel lt |! 1 b 4 a | pb 1 at et ae | TI ate an Eta | c 4 oll ma: PA, gel ! oe il SB. ja d fs) O a b Cc d e f h I5 2

Fics. 14, 15. Pereskiopsis aff. chapistle. Fic. 14, a-c, cross sectional views of a tier at levels indicated in diagram (d), a schematic representation of some members of tier, X 390. For details, see text. Fic, 15, successive phloem deriv- atives (a-g) and the fusiform cambial initial (h) of a tier in tan ential view,

220. Derivatives at (a), (b-c), (d-e), (f), and (g) are ontogenetically re- lated. Symbols explained in text.

254 JOURNAL OF THE ARNOLD ARBORETUM | VOL. XLIII

iss Coe sayy tee ee ee ee

16 9 Fics. 16-20. Pereskiopsis att. chapistle: representative sieve elements and despciated sacar adege cells, drawn from tangential sections, 220. De rivatives in each figure originated after divisions in a single phloic initial; some derivatives separat ted in Fics. 17-20 for better i sieve elements

in Frc. 18 with a sieve plate on their common lateral wall. Broken lines represent outlines of cells seen more clearly at a different focal level; other symbols ex- plained in text.

Two other patterns of divisions in phloic initials recorded in Ouiabentia aff. chacoensis were similar to those seen in Pereskia sacharosa. In one method two successive, oblique longitudinal divisions in a phloic initial resulted in three daughter cells, one of which matured as a sieve element, the other two as parenchyma cells (sieve elements 3 and 4, Fic. 22, a-d). Such an origin of sieve elements and parenchyma cells seems to be a variation of the same method as was recorded for sieve elements 3 and 4 in Fic. 2, f, except that in this instance the sieve-tube mother cell differ- entiated directly as the sieve element without forming any companion cells. Sometimes the phloic initial divided obliquely and the resulting daughter cells behaved as sieve-tube mother cells of two elements (sieve elements 2 and 3, Fic. 23, a-d; cf. also sieve elements 6 and 7, Fic. 2, f).

In some other patterns recorded in Pereskiopsis aff. chapistle and Quia- bentia aff. chacoensis the number of divisions in phloic initials preceding the differentiation of sieve elements and parenchymatous cells was much Sometimes the sieve-tube mother cell, formed after the first longi-

less.

i 4 O ama d 1G. 21. Quiabentia aff. chacoensis: a-c, cross sectional views of a tier at levels indicated along the margins of diagram (d), a schematic representation of some members of the tier, as in Fic. 2; a-c, X 390. For details, see text.

tudinal division in the phloic initial, matured directly into a sieve element. A parenchyma cell (or strand) was formed in association with the sieve element but no companion cells were formed (sieve element 2, Fic. 10, d). In a few instances, the phloic initial behaved as the sieve-tube mother cell, the first division setting off a precursor of one or more companion cells and a precursor that matured as a sieve element. No parenchyma cells were formed in these cell combinations (sieve elements 1 and 3, Fic. 10, d; sieve element 2, Fics. 11-14, d; sieve element 4, Fic. 21, d; and sieve element 5, Fic. 22, d). In still other cases, the phloic initial did not divide at all and matured directly into a sieve element (sieve element 3, Fic. 12

?

256 JOURNAL OF THE ARNOLD ARBORETUM [ VOL, XLIII

Pb

De

L [

J |

Fic. 22. Quiabentia aff. chacoensis: a-c, cross sectional views of a tier at levels indicated along the margins of diagram (d), a schematic representation, as in Itc. 2; sieve element 6 and associated companion cell partly crushed; a-c, X 390. For details, see text.

Several tiers in Pereskiopsis aff. chapistle and Quiabentia aff. cha- coensis were studied from serial tangential sections. The results obtained from this study generally confirmed the observations made from cross sections (cf. Fics. 15, 26). Some cell complexes including sieve elements and ontogenetically related parenchymatous cells selected from different tiers of these two species are shown in Fics. 16-20 and 27-31. These figures emphasize the variations in number and planes of divisions in different phloic initials, and also give an indication of the number of sieve elements that may be formed within the confines of a single phloic initial.

Fic. 23. Quiabentia aff. chacoensis: a-c, cross sectional views of a tier at levels indicated along the margins of diagram (d), a schematic representation of some members of tier, as in Fic. 2; a-c, X 390. For details, see text

t may appear from our analysis of tiers that in Pereskiopsis aff. chapistle and Ouiabentia aff. chacoensis, as compared to Pereskia sacha- rosa, relatively few divisions occur in the phloic initials and relatively few parenchymatous cells are formed in association with the sieve elements. It would also seem that the planes of divisions are lesss irregular and confus- ing in the first two species than in the last. It must be stated, however, that 2 or 3 sieve elements and related parenchymatous cells are frequently formed from a single phloic initial in Pereskiopsis aff. chapistle and Quia- bentia aff. chacoensis (Fics. 17-20, 30), and sieve elements with no onto- genetically related parenchymatous cells may be formed in Pereskia sacha- rosa, particularly in the neighborhood of rays (Fic. 35). Also, the planes of divisions in the phloic initials and their daughter cells in Pereskiopsis aff. chapistle and Ouiabentia aff. chacoensis (Fics. 17, 18, 20, 30) are often as irregular and confusing as in Pereskia sacharosa. If quantitative dif- ferences exist between these species, only a wider sampling of tiers and material would reveal them and make the results statistically significant.

258 JOURNAL OF THE ARNOLD ARBORETUM [ VOL, XLIII

oC

nm nN on Ss fe, Sa ee ee ey ees een ee Sees ees i

TSO Oe TE “Ty eee Fe re]

ce)

Pic. 24. Quiabentia aff. chacoensis: a-c, cross sectional views of a tier at levels indicated along margins of diagram (d), a schematic representation of some members of tier, as in Fic. 2; a-c, * 390. Three cells internal to cell com- plex including sieve element 1 considered to be fusiform phloem-parenchyma cells. Details of drawings in text.

1962 | SRIVASTAVA & BAILEY, CACTACEAE, V 259

|

|

|

|

|

|

|

|

|

: abl

: |

|

: : : sie

| ral : oh O =i ea a

1G. 25. Quiabentia aff. chacoensis: a-c, cross sectional views of a tier at levels indicated along margins of diagram (d), a schematic representation of some members of tier, as in Fic. 2; a-c, X 390. Details of drawings in text.

d

260 JOURNAL OF THE ARNOLD ARBORETUM [ VOL, XLII

@

f g h

Vic, 26, Quiabentia aff, chacoensis: successive phloem derivatives (ah) and the fusiform cambial initial (i) of a tier in tangential view, < 220. Derivatives at (a), (b-c), (d-e), (f-g), and (h) derived from single phloic initials. Symbols explained in text.

1962] SRIVASTAVA & BAILEY, CACTACEAE, V 261

2? 28

Fics. 27-31. Quiabentia aff. chacoensis: representative sieve elements and associated parenchymatous cells in tangential view, x 200. Derivatives in each figure originated after divisions in a single phloic initial. In Fics. 27 and 28, parenchyma cell is seen both above and below sieve element with which it is associated. Sieve elements in Fic. 30, b, are connected by a lateral sieve plate. Symbols explained in text.

262 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIII

SIEVE ELEMENTS

Sieve elements are considered to be the principal conduits through which carbohydrates and other food materials are transported within a plant. Their structural peculiarities are important, therefore, not only to physiologists concerned with problems of translocation, but also to morphologists interested in the evolutionary aspects of the phloem tissue as a whole. Esau, Cheadle, and Gifford (1953) suggested some possible trends of specialization in the phloem. In recent years, some new information, particularly about the length of sieve elements in relation to that of the phloic initials from which they are derived and about the association of sieve elements with parenchymatous cells, has been added through de- tailed ontogenetic work (Esau & Cheadle, 1955; Cheadle & Esau, 1958: Evert, 1960). In the following paragraphs some structural features of the sieve elements in leaf-bearing Cactaceae, their contents, wall struc- ture, sieve areas, and length, are emphasized for a better understanding of their physiologic and evolutionary specialization.

As is typical of this tvpe of cell, the sieve elements in the leaf-bearing Cactaceae are enucleate in their functionally mature state. The fixa- tive and the staining procedures used by us were not conducive to a critical cytological study. However, numerous, doughnut-: shaped bodies, probably carbohydrate granules, were seen in the sieve elements. These granules aggregated in large numbers near the sieve plates and_ sieve areas (Fics. 39, 41). In addition, the material described in the litera- ture as “slime” was often present in copious quantities. It was usually deposited on one side of a sieve plate in the form of a plug and extended like a strand through most of the cell lumen (Figs. ody 61), Jn 1onpi- tudinal sections, this feature was often useful in determining the extent of a particular sieve element. It was rather scanty in Pereskiopsis aff, chapistle (Fic. 39).

Nacreous walls, reported in the sieve elements of various dicotyledons by Esau and Cheadle (1958), were uniformly lacking in the material studied by us (cf. Figs. 32, 38, 40). In permanent slides the walls of sieve elements appeared rather thin, but they were birefringent under polarized light

The common walls between two sieve elements bear numerous sieve areas, some of which are more specialized than others and are termed sieve plates (cf. Cheadle & Whitford, 1941; Esau, 1950). Sieve areas occur on radial as well as tangential walls (Fics. 44, 48): and their size, as determined by the number of callose cylinders stained by lacmoid, varies considerably (Fic. 51). The difference in size of connecting strands and pores in sieve areas that occur in the sieve plates on the one hand and those that occur elsewhere on the walls is very marked (Fics. 50, 51 The sieve plates are generally of the simple type with a single sieve area, but sieve plates with two or more sieve areas are present also (Fic. 50 As a rule, every sieve element has two sieve plates (cf. Fics. 7-9, 15— 20, 26-31); but some sieve elements may have as many as 3 or 4 sieve plates

eu

a

1962] SRIVASTAVA & BAILEY, CACTACEAE, V 263

(Fics. 18, b; 30, b) others may have only one (Fics. 7, b, c; 18, a; 20, a). The position of sieve plates in the sieve elements varies likewise. Com- monly, the sieve plates occur at or near the ends of the sieve elements, but sieve plates on lateral walls are not unusual (Fics. 7, b; 8, b; 18, a, b; 30,b). The final shape, size, and orientation of the mature sieve elements is often different from that of the phloic initials from which they are de- rived. Some sieve elements may be rather grotesque in shape (Fics. 7, b; 8, b; 17, b); and the long axis of some may be placed slantingly in rela- tion to the vertical axis of the phloic initial and, hence, that of the stem (Fics. 9, b; 20, a; 28; 29, b; 41). Such slantingly placed sieve elements are often connected through sieve plates with sieve elements in laterally adjacent (rather than vertically adjacent) tiers (Fic. +1). Often, two sieve elements derived from the same phloic initial jorm parts of two different sieve tubes, one with a vertical, the other with a diagonal orientation in relation to the long axis of the stem. Laterally placed sieve plates are encountered rather frequently in areas where diagonally placed sieve elements abut on the lateral walls of adjacent sieve elements.

The length of a sieve element depends not only on the length of the phloic initial from which it arises, but also on the number and planes of divisions in the phloic initial and its daughter cells and on the number and planes of divisions in the sieve-tube mother cell.t The length of sieve elements as measured from transections (Fics. 2—6, f; 10-14, d; 21-25, d) is subject to an error which must be mentioned. When the sieve plates occur near but not at the end of the sieve element (e.g., Fics. 7, d; 17, a; 30, b), the tapering end of the cell may be missed from calculations. The following statements about the lengths of sieve elements and the phloic initials from which they are derived are, therefore, liable to some error.

Since the phloic initials in a tier are usually derived by symmetric periclinal divisions in a fusiform cambial initial, one normally expects them to be of approximately equal length (e.g., Fics. 4, f; 10, d). The fluctuations in the lengths of phloic initials in several tiers (e.g.. Fics. 2,f: 6, f; 25, d) are, therefore, somewhat unusual. A possible explanation for the different lengths of the phloic initials in a tier would be that the individual phloic initials elongated intrusively in varying degrees. But this explanation is contrary to the commonly held belief that the phloic initials, in contrast to fusiform cambial initials, do not elongate intru- sively. It appears that in the leaf-bearing @ictaceae the periclinal di- visions in the fusiform cambial initials are often not strictly longitudinal and the dividing walls fall short of one or both tips of the mother initial. The daughter cells formed after such an asymmetric periclinal division may be equal or unequal in length, but they are vertically dis- placed (at one or both ends) in relation to one another. If a number of such asymmetric periclinal divisions occur in the daughter cell that be-

‘ This reasoning is based on the assumption that the sieve clements do not undergo intrusive elongation during their differentiation from the phloic initials or their daughter cells. Several studies on the phloem of gymnosperms and dicotyledons support this assumption (Bannan, 1950; Cheadle & Esau, 1958; Evert, 1960; Srivastava, 1962).

264 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLII

haves as the fusiform cambial initial after each division and if such di- visions are interpolated among others that are symmetric, the net result as seen in mature tissue would be comparable to that shown in Fics. 2,f; 6,f; 15; 25.d; and 26. In its radial extent, the tier would show fluctuations in height that would be independent of the intrusive growth of the fusiform cambial initial. The intrusive growth of the fusiform cambial initial, if it occurred after a periclinal division, would further com- plicate the pattern.

In instances where the phloic initial matures directly as a sieve element and no parenchymatous cells are formed (Fic. 12,d, sieve element 3); the length of the sieve element is approximately the same as that of the phloic initial. Also, if the phloic initial behaves as a sieve-tube mother cell, that is, one or more precursors of companion cells are formed as sister cells of the precursor that matures as the sieve element but no parenchyma cells are formed (Fics. 10,d, sieve elements 1 an sieve element 2; 21,d, sieve element 4; 22,d, sieve element 5), the length of the sieve element closely reflects the length of the phloic initial. In other instances, the length of the sieve element bears no predictable rela- tionship to that of the phloic initial —it may be the same (cf. Fics.

, C; 15, d-e; 26, f-g) or it may be much less (cf. Fics. 7, b; 8, b, c; 9, b: 15, b-c, g; 26, a).

In concluding this section on sieve elements, a few remarks must be made about the relative lengths of the sieve-tube members and the paren- chymatous elements ontogenetically associated with them. The length of parenchyma cells (or strands) may be equal to, or more than, or less than that of the sieve element with which they are associated (cf. Fics. 2—6,f: 10-14,d; 21-25,d). The divisions that separate the precursor of the companion cell (or strand) from that of the sieve element are usually longitudinal with reference to the long axis of the sieve-tube mother cell, but the dividing walls are laid down in such a manner that the precursor of the companion cell (or strand) may be almost as long as that of the sieve element or it may be shorter (cf. Fics. 7-9; 15-20; 26-31). Oc- casionally, the dividing wall is laid obliquely and short of the cell tips at either end; as a result, the precursor of the companion cell (or cells) at one end projects a little beyond the end of the precursor that matures as the sieve element (e.g., Fics. 7,c; 9,a; 19,a; 30, b; 33). Exceptionally, the precursor of the companion cell (or cells) may be a little longer than the precursor of the sieve element (e.g., Fics. 5, f, cell complex in- cluding sieve element 2; 9,c).

PARENCHYMA CELLS

Although the parenchyma cells in the phloem of leaf-bearing Cactaceae differ in their ontogeny, we were not able to discern any significant cytological differences between them with the fixing and staining pro- cedures that we used. The fusiform phloem-parenchyma cells and the individual cells in phloem-parenchyma strands, which in other plants

1962] SRIVASTAVA & BAILEY, CACTACEAE, V 265

commonly store starch, oil, tannins, etc. (cf. Esau, 1953, p. 284), did not show any such inclusions in the functional phloem. The parenchymatous cells ontogenetically related to the sieve elements also did not show any of these inclusions. It is commonly held that companion cells have a denser cytoplasm and, in general, stain more deeply than other paren- chyma cells in phloem (cf. Esau, 1953, p. 283). We utilized this feature in identifying companion cells and were partially successful. But it was not always possible to use this criterion because the deeply staining appearance of a companion-cell protoplast depends partly on the size of the companion cells and partly on the plane of the section with reference to the cell lumen. For instance, if a longitudinal section passes close to a lateral wall rather than through the middle of the cell, the protoplast may appear lightly stained. Because of these features it was difficult for us to be certain about the identity of individual parenchymatous cells in isolated cross and tangential sections. With the use of serial sections one can normally distinguish between fusiform phloem-parenchyma cells and phloem-parenchyma strands, on the one hand, and the parenchymatous cells ontogenetically related to the sieve elements, on the other. It is also possible in most cases to distinguish between companion cells and paren- chyma cells ontogenetically associated with the sieve elements, because the length of the companion cells (or strand of companion cells) usually does not exceed that of the sieve element, whereas that of the parenchyma cells (or strands) frequently may. Still, in several cell configurations (e.g., those including sieve elements 3 and 4, Fic. 22, a—d) it was almost impossible to tell whether one was dealing with companion cells or paren- chyma cells or both. (In this instance, the cells were interpreted as parenchyma cells.)

Sieve elements are usually connected with their associated companion cells through one-sided sieve areas. Such sieve areas have densely stain- ing connecting strands and callose cylinders only in the wall belonging to the sieve element; in the companion cell wail there is only a primary pit field. Cheadle and Esau (1958) and Evert (1960) reported such one- sided sieve-area connections between sieve elements and the parenchyma cells ontogenetically associated with them also. In the material of leaf- bearing Cactaceae that we have examined, sieve-area connections were common between sieve elements and companion cells and between sieve elements and ontogenetically associated parenchyma cells (Figs. 45, 46, 49). Although no counts were made, our general impression is that the number of sieve-area connections between sieve elements and companion cells is usually higher than that between sieve elements and ontogenetically associated parenchyma cells. Cheadle and Esau (1958) made a similar observation. One notable feature about the sieve-area connections between sieve elements and ontogenetically related parenchyma cells was that the connections occurred in those portions of the common wall that bulged somewhat in the lumen of the parenchyma cell (Fics. 46, 52). In contrast, the common wall between the sieve elements and their companion cells, although covered with numerous sieve areas, was usually straight (Fic.

266 JOURNAL OF THE ARNOLD ARBORETUM [ VOL, XLHI

45). Sieve-area connections usually occur between sieve elements and parenchymatous cells derived from the same phloic initials; but occasion- ally sieve elements derived from one phloic initial may be connected to a parenchymatous cell that forms part of a cell complex derived from a different phloic initial. Such connections were recorded especially in those parts of a parenchymatous cell that projected beyond the end of the sieve element to which it was ontogenetically related.

Sieve-area connections were also observed between sieve elements and some fusiform phloem-parenchyma cells (Fic. 47). The sieve areas oc- curred in those portions of the common wall that bulged into the lumen of the parenchyma cell. The fusiform phloem-parenchyma cells that were so connected with sieve elements occurred in the same tier as the sieve element or in neighboring tiers, but their number was relatively very small and rela- tively few sieve-area connections occurred on the common wall. Fusiform i eee cells in such intimate contact with the sieve elements were more common in Pereskiopsis aff. chapistle and OQuiabentia aff. cha- coensis than in Pereskia sacharosa.

In the nonfunctional phloem, sieve elements and companion a lose their living contents, collapse, and sooner or later are crushed (Fics. 36,

37). It is our impression that the parenchyma cells ontogenetically a to the sieve elements survive for a longer time than the related sieve ele- ments and companion cells, but eventually they also lose their protoplast, collapse, and are crushed. The fusiform phloem-parenchyma cells an the individual cells of phloem-parenchyma strands seem to survive and sometimes store starch and even accumulate druses of calcium oxalate. However, in the absence of any definite cytological differences between the parenchyma cells ontogenetically related to sieve elements and those that are not so related, and because of a general distortion of the tissue, it is difficult to determine in nonfunctional phloem which cells or strands may or may not have been ontogenetically related to the sieve elements. Consequently, we are not sure whether all fusiform phloem-parenchyma cells and phloem-parenchyma strands continue to live in the nonfunctional phloem; we are also not sure whether @// parenchyma cells and strands ontogenetically related to sieve elements die and collapse in the nonfunc- tional phloem.

mes

DISCUSSION AND CONCLUSIONS

Our studies on the phloem of leaf-bearing Cactaceae were restricted in scope, partly because of the difficulty of obtaining and sectioning the material of cacti and partly because of the fixative used to kill it. Only those species were selected for a detailed study which could be sectioned in a satisfactory manner, and only those aspects of cellular detail were followed which were not directly related to seasonal changes and the elucidation of which did not require elaborate cytological techniques. For these reasons we did not concern ourselves with such aspects of

1962 | SRIVASTAVA & BAILEY, CACTACEAE, V 267

phloem study as the development and maturation of sieve elements, the cytological inclusions in companion cells and parenchyma cells onto- genetically related and unrelated to sieve elements, and the longevity of sieve elements. We also did not study the changes from functional to nonfunctional phloem in any detail, and our investigation of ray tissue was only cursory.

The secondary phloem of the three species that we investigated is rela- tively simple. Sieve elements and parenchyma cells of various types were the only constituents in the axial tissue. Oil cells, mucilage cells, fibers, etc., were absent from the functional phloem of these species. In the nonfunctional phloem several parenchyma cells accumulated druses of calcium oxalate and, in Pereskia sacharosa, some were modied as sclereids (Bailey, 1961a, b). Intercellular spaces were commonly seen in functional and nonmineliondl phloem (Tics. 32, 38, 40). The rays are typically high and multiseriate and include, especially near the ray margins, some upright cells that may be almost as long as some of the fusiform derivatives. High and broad rays are often dissected by a conversion of ray initials into fusiform initials. In the early stages of such conversion the sieve elements that are formed are often of the same size as other ray cells and frequently lack companion cells (Fic. 35). Fusiform initials are converted into ray initials also, with accompanying changes in the nature of their derivatives.

Ontogenetic studies of phloem yield valuable data about the origin and interrelationships of different phloem elements. Our analysis of tiers in the phloem of leaf-bearing Cactaceae revealed several possible ways in which sieve elements and parenchymatous cells may arise from phloic initials. Most of these methods of origin have been discussed in detail elsewhere in this paper. In the following some of the main conclusions are summarize

The three representatives of the leaf-bearing Cactaceae studied, Pereskia sacharosa, Pereskiopsis aff. chapistle, and Quiabentia aff. chacoensis, do not seem to differ basically in the method of origin of phloem elements. Fusiform phloem-parenchyma cells and phivem-parenchyma strands arise either directly from or after a few horizontal divisions in the phloic ini- tials. In contrast, the differentiation of sieve elements and parenchyma- tous cells ontogenetically related to them is usually preceded by several divisions in the phloic initials and their daughter cells. Evidence of such divisions was seen not only in mature tissue (cf. Frcs. 33, 34), but also in immature tissue close to the cambium (cf. Fics. 42, 43). Commonly, a parenchyma cell (or strand) is formed in association with a sieve ele- ment and its companion cells; but sometimes the companion cell, or the parenchyma cell, or both are lacking, and sometimes several parenchyma cells, sieve elements, and companion cells are formed within the confines of a single phloic initial. Apparently, the phloic initials and their daughter cells show varying degrees of mitotic activity before cells are differenti- ated. Such variations among different phloic initials may be encountered, not only among different tiers of the same species, but also within the

268 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIII

same tier. Also, one tier may be composed predominantly of fusiform phloem-parenchyma cells and phloem-parenchyma strands and may have very few sieve elements (Fic. 24, a—c), whereas a neighboring tier may have numerous sieve elements and ontogenetically related parenchymatous cells and relatively few fusiform phloem-parenchyma cells and phloem- parenchyma strands (Fic. 25, a-c). We do not know what factor or fac- tors may be responsible for these variations in the behavior of phloic initials in the same tier and in neighboring tiers.

Although the final number of parenchyma cells, sieve elements, and companion cells arising within the confines of a phloic initial varies con- siderably, the sequence of divisions seems to be such that the precursors of parenchyma cells arise as sister cells of precursors that behave as sieve- tube mother cells, and the precursors of companion cells arise as sister cells of the precursor that matures as a sieve element. Two observations support this assumption. First, in cross sections, the wall separating the parenchyma cell from the complex of sieve element and companion cells seems to be the first wall laid down in the phloic initial (or its daughter cell) and the wall separating the sieve element from the companion cell seems to be laid down in one of the daughter cells formed after the first division (Fic. 40). Second, the length of the associated parenchyma cells may frequently exceed that of the sieve element, but that of the companion cells rarely does (Fic. 41). If our interpretation of the sequence of divi- sions within the confines of a phloic initial is correct, the parenchyma cells in their ontogenetic relationship would be somewhat more distantly re- lated to the sieve elements than the companion cells. But this interpreta- tion is applicable only if parenchyma cells, as well as companion cells, are formed in association with a sieve element.

It is significant that the sieve elements in the leaf-bearing Cactaceae have sieve-area connections, not only with ontogenetically related, but also with ontogenetically ee parenchyma cells. These connections are numerous between sieve elements and companion cells, less numerous between sieve elements and ontogenetically related parenchyma cells, and are only infrequently observed between sieve elements and the parenchyma cells that are derived from independent phloic initials. Apparently, the varying degrees in which the parenchyma cells are ontogenetically related to sieve elements are paralleled by the degree of their physiological affinity to sieve elements: those with the closest ontogenetic association are also those that have the most intimate physiological association. For future research, it will be of extreme interest to determine the cytological features of these parenchyma cells and to see in what important ways they differ from one another

From an evolutionary viewpoint, the parenchyma cells physiologically associated with the sieve elements present a complicated picture. The albuminous cells of gymnosperms, although physiologically comparable to companion cells of angiosperms, generally have no ontogenetic association with the sieve elements (cf. Strasburger, 1891; Srivastava, 1962). Com-

1962] SRIVASTAVA & BAILEY, CACTACEAE, V 269

panion cells are lacking in Austrobaileya scandens, a primitive dicotyle- don (Bailey & Swamy, 1949); and it is possible that some parenchyma cells derived from independent phloic initials serve as companion cells in this plant. Holdheide (1951) reported some parenchyma cells in the secondary phloem of dicotyledons that he compared to albuminous cells of gymnosperms, thus implying that these cells originated from independ- ent phloic initials. Other reports include varying degrees of specialization among companion cells themselves (Resch, 1954). In view of these findings and the ontogenetic studies carried out by Cheadle and Esau (1958), Evert (1960), and in the present work, it may seem that the evolution of the parenchyma cells physiologically associated with the sieve elements has progressed in the direction of a closer ontogenetic asso- ciation with the sieve elements. But the presence of sieve-area connections between sieve elements and parenchyma cells ontogenetically unrelated to them in a family as highly specialized as Cactaceae presents serious im- pediments in the acceptance of this idea. It is clear that more and detailed ontogenetic and cytological studies of the phloem of vascular plants need to be carried out before any valid generalizations about the evolution of parenchyma cells physiologically associated with the sieve elements can be made.

In the preceding paper of this series (Bailey & Srivastava, 1962), based upon the study of numerous putative species of Pereskia, Peres- kiopsis and Quiabentia, we demonstrated that the fusiform initials of the cambium and their derivatives in phloem have attained a high level of phylogenetic specialization. This is shown by their short length and their tendency to occur in stratified arrangements as seen in tangential longi- tudinal sections. The well-developed structure of the sieve plates in the sieve elements of the leaf-bearing Cactaceae and the marked difference in the size of pores and connecting strands in the sieve plates, on the one hand, and the lateral sieve areas, on the other (Fics. 50, 51), also suggest an advanced degree of evolutionary specialization.

In this paper we have been concerned primarily with the phenomenon of cell divisions in phloic initials and their daughter cells and the inter- relationships between sieve elements and different kinds of parenchyma cells. It is evident from our studies that the ontogenetic changes in the differentiation of sieve elements, companion cells and parenchyma cells are diversified and variable. They obviously are qualitatively similar in the three genera that we examined. Although there are some evidences of possible quantitative differences in these forms, it is clear that a great deal of material must be critically studied before valid taxonomic conclu- sions can be drawn. This is in marked contrast to the structure of non- functional phloem, where, as one of us (Bailey, 1961a) has shown, Pereskia differs from Pereskiopsis and Quiabentia by the presence of sclereids, and where three groups of species of Pereskia may be differenti- ated upon the basis of form and distribution of such sclerenchymatous elements.

270 JOURNAL OF THE ARNOLD ARBORETUM [VOL LTT

LITERATURE CITED BaILey, I. W. Comparative anatomy of the leaf-bearing Cactaceae, I. Struc- ture and distribution of sclerenchyma in the phloem of Pereskia, Pereskiop- sis and Quiabentia. Jour. Arnold Arb, 42: 144-156. 1961a

Comparative anatomy of the leaf-bearing Cactaceae, HI. Form oe distribution of crystals in Pereskia, Pereskiopsis and Oia Ibid. 3 346. 1961b.

& L. M. Srivastava. Comparative anatomy of the leaf-bearing Cacta- ceae, IV. The fusiform initials of the cambium and the form and structure of their derivatives. /bid. 43: 187-202. 1962.

B.G wamy. The morphology and relationships of Austrobaileya.

ibid. 30: 211-226. 49,

BANNAN, M. W. The frequency : anticlinal divisions in fusiform cambial cells of ae ee Amer. Jour. Bot. 37: 511-519. 1950.

CHEADLE, V. I. & K. Esau. cae phloem of Calycanthaceae. Univ. Calif. Publ Bot. - 397-510. 58.

. WuitForp. Observations on the phloem in the Monocotyle- doneae. - The occurrence and phylogenetic raga in structure of the sieve tubes in the metaphloem. Amer. Jour. Bot. 28: 623-627. 1941.

, E. M. Girrorp, Jr., & K. Esau. A ee ‘combination for phloem and

contiguous tissues. Stain Tech. 28: 49-53. 1953,

Esau, K. a ee and structure of the phloem tissue. II. Bot. Rev. 16: 67-114. 1950.

ee Anatomy. John Wiley and Sons, N.Y., 1953. V. I. CHEADLE. Significance of cell divisions in differentiating sec-

ondary phloem. Acta Bot. Neerl. 4: 346-357.

Wall ghickenine in sieve elements. Proc, Natl. cicad, Sc. 445 546-553. 1958.

—______—— & E. M. Girrorp, Jr. Comparative structure and possible trends of specialization of the phloem. Amer. Jour. Bot. 40: 9-19.

Evert, R. F. Phloem structure in Pyrus communis L. and its seasonal changes. Univ. Calif. Publ. Bot. 32: 127-194.

HotpHEIDE, W. Anatomie mitteleuropdischer Gehdlzrinden (mit mikrophoto- graphischem Atlas). Jv: H. Freund’s Handbuch der Mikroskopie in der a Vol. 5, Part 1, p. 193-367. Umschau Verlag, Frankfurt-am-Main,

a . Beitrage zur Cytologie des Phloems. Entwicklungsgeschichte der Siebrdhrenglieder und Geleitzellen bei Vicia faba L. Planta 44: 75-98. 954.

one ae L. M. Secondary phloem in the Abietineae. Ph.D. Thesis (unpubl.), Univ. California, Davis.

Pen E. Ueber den Bau und die Verrichtungen der Leitungsbahnen in den Pflanzen. Histologische Beitrage, Vol. 3. Gustav Fischer, Jena 1891.

1962] SRIVASTAVA & BAILEY, CACTACEAE, V DHA

EXPLANATION OF PLATES

The following symbols have been used consistently in all the plates: C, cam- bial zone; CS, companion cell or strand; FP, functional phloem; Gr, carbo- hydrate granules; IS, intercellular space; NP, nonfunctional phloem; P, paren- chyma cell or strand ontogenetically related to sieve elements; Par, fusiform phloem-parenchyma cell or phloem-parenchyma strand; R, ray; S, sieve ele- ment; SA, sieve areas; Sl, slime; SP, sieve plate.

PLATE I

32-34. Transverse and tangential views of phloem and cambium of Poe sacharosa. 32, Transverse section, X 455. 33, Tangential section of phloem showing a strand of parenchyma cells (P) and companion cells (CS) in ontogenetic association with a sieve element (S), * 220. A strand of companion cells (indicated by arrow) projects below the end of the sieve element with which it is associated. Fusiform phloem-parenchyma cells (Par) give an idea about the length and size of phloic initials. 34, Tangential section of the cam- bium, X 220

PLATE II

Fics. 35-37. Tangential and radial views of phloem of Pereskia sacharosa.

, Tangential section through functional phloem. 36, Radial section of func-

Ce (right) and nonfunctional eae phloem. 37, Tangential section of non- functional phloem. All figures & 45

PLATE III

Fics. 38-40. Transverse and ee views of phloem. 38, Pereskiopsis aff. chapistle, transverse section of phloem and cambium. 39, The same, saa tial section of phloem. 40, ee “ail chacoensis, transverse section of phloem and cambium. Cell complexes including a Daren chviia oe 1), sieve element (S) and companion cell (CS) are shown. All figures & 45

PLATE IV

Fics. 41-43. Tangential sections of phloem and cambium of Quwiabentia aff. chacoensis. 41, section through mature phloem. The sieve element on lower eft is slantingly placed in relation to the parenchyma cell which occurs in association with it. 42, section through immature phloem showing divisions in phloic initials marked with oy 43, section through cambium. Fics. 42 and 43 are photographs of successive sections in a tangential series; the cambial initials that produced the phloic na in Fic. 42 marked with (x) are similarly marked in Fic, 43, All figures & 2

PLATE V Fics. 44-47. Tangential sections of phloem showing sieve areas between sieve related parenchyma cells, and sieve elements and ontogenetically unrelated

parenchyma cells. 44, Pereskia sacharosa, * 910. 45-47, Quiabentia aff. chacoensis, X 910.

272 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIII

PLATE VI

Fics. 48-52. Sieve plates, lateral sieve areas and sieve-area connections be- tween sieve elements and related parenchymatous cells. 48, Pereskia sacharosa, transverse section showing a sieve area (indicated by arrow) on the tangential wall between two sieve elements, X 1140. These sieve elements formed part of a cell complex that had originated from a single phloic initial. 49, Quiabentia aff. chacoensis, transverse section showing one-sided sieve-area connections (in- dicated by arrows) between a sieve element a the companion and parenchyma cells that are associated with - < 910. 50, Pereskiopsis aff. chapistle, sieve plate in radial view, * 1140. Dien all. chacoensts, lateral sieve areas between two sieve elements in renee view, X 1140. 52, The same, tangen- tial section showing ongesiged sieve-area connections between a sieve element and related parenchyma cell, he sieve areas occur in those portions of the common wall that ae into the lumen of the parenchyma cell.

—

Jour. ARNOLD Arg. VoL. XLIII

PLATE I

V

BEARING CACTACEAE,

LEAF-

)

SRIVASTAVA & BAILEY

Jour.

ARNOLD Ars. VoL. XLIII

PLaTE II

SRIVASTAVA & BAILEY, LEAF-BEARING CACTACEAE, V

'

Ga Gy

SRIVASTAVA & BarLey, LEAF-BEARING CACTACEAE, V

IIITX ‘10A ‘aay dTonay ‘anof

III SLV1d

Jour. ARNOLD Arp. VoL, XLIII

PLATE IV

SrIvASTAVA & BAILEY, LEAF-BEARING CACTACEAE, V

Jour. ARNOLD Ars. VoL. XLIII PLATE V

Jour. ARNoLpD Ars. VoL. XLIII PLaTE VI

SRIVASTAVA & BAILEY, LEAF-BEARING CACTACEAE, V

1962] HOWARD, VOLCANISM AND VEGETATION 279

VOLCANISM AND VEGETATION IN THE LESSER ANTILLES

KicHarp A. Howarp

THE MAJORITY OF THE ISLANDS of the Lesser Antilles in the Caribbean archipelago are volcanic in origin. Two of the islands have active vol- canoes which currently are dormant. However, Mt. Pelée on Martinique erupted in 1902 and 1930, and the Soufriére on St. Vincent erupted in 1902. Nine of the islands from Grenada, in the south, to St. Kitts, in the north, have active fumaroles or soufriéres, indicating residual volcanic activity. Of the few volcanic islands without historic volcanic activity, Saba, Redonda, Union, and others show the classic forms of their vol- canic origin and prehistoric volcanic activity (11, 13, 48, 60).

The eruptions of the twentieth century on Martinique and St. Vincent have been well observed, studied, and recorded. An extensive descriptive literature is available for the volcanic and seismic activities of the area. The literature regarding the geological phenomenon of fumaroles in the Lesser Antilles is less complete, although a specialized interest in utilizing the thermal power of one, the Qualibou soufriére in St. Lucia, has been recorded recently (8, 49, 50).

In nearly all of the geologic studies, as well as in many floristic, phyto- geographical and ecological papers on the area, some mention is made of the effects of volcanoes and soufriéres on the vegetation. These range from Perret’s mention of the attempted use of the sensitive plant, Mimosa pudica, to record earthquakes, to the papers by Stehlé and Beard con- sidering the progressive changes in the regrowth of vegetation on devas- tated volcanic slopes. The present paper, a survey of the nature of vol- canic activity and its effects on the vegetation in the Lesser Antilles, is based on observations made in 1950 during a field trip which began in Trinidad, extended to the northern and western islands, and ended in Jamaica. Data and specimens were collected in and around many sites of past and present volcanic activity. More recent visits to the Lesser An- tilles, particularly to Montserrat in 1961, have allowed comparative ob- servations after an elapsed time, as well as the gathering of specific in- formation. For comparison with the West Indies, I have had the oppor- tunity of studying the effects of the recent eruptions of Kilauea and Kilauea-iki and of the various fumaroles on Mauna Loa, in Hawaii.

These trips, observations, and collections form part of a continuing study of the vegetation of the Lesser Antilles. I acknowledge with gratitude the support of the American Philosophical Society during the 1950 trip and that of the National Science Foundation for grants which have aided more recent work. Many people in the Lesser Antilles have aided my work in

280 JOURNAL OF THE ARNOLD ARBORETUM [ VOL, XLII

various ways. I mention in particular the assistance of Hugh McConnie,

f St. Vincent; James Ross, of Grenada; Frank Delisle, now of Antigua; Kingsley Howes, of Montserrat; Harold Simmonds, of St. Lucia; John Knowlton, formerly of Dominica; and the late Malcolm Smith, of St Kitts.

HISTORY OF VOLCANISM IN THE LESSER ANTILLES Perret (46) has recorded the volcanological history of the Lesser An- tilles. A chronology of the important eruptions is the following:

1692 — earthquakes followed by eruption on Mt. Misery, St. Kitts 1694 — earthquakes followed by eruption on the Soufriere, Guadeloupe. 1718 — first recorded eruption of the Soufriére, St. Vincent. 1765 — earthquakes and gas emission on ae 1766 — eruption of the Qualibou soufriére, St. Luc 1798 — earthquakes and eruption of the Soufriere, ‘eunsebupe 1812 — great eruption of the Soufriére, St. Vincent. 1838 — eruption of the Soufri¢re, Guadeloupe. 1851 — earthquake and lateral outbreak of Mt. Pelée, Martinique. 1880 —ash eruption on Dominica. 1898 — start of three years of earthquakes ~ es emission on Montserrat. 1902 — most destructive eruptions of Mt. Pelée, Martinique, and the Sou- friére, St. Vincent 1929 - mnt of Mt. Pe lee, Martinique. 1934 —start of four years of earthquakes and gas emission on Montserrat.

The occurrence of earthquakes and actual eruptions is usually well re- corded. Nevertheless, some of the earlier dates cited above have been questioned. Anderson (2), for example, cites both the picturesque report of the 1718 eruption of the Soufriére on St. Vincent, as given by Defoe (37), and the subsequent questions raised by the report. To the present day, the 1718 eruption is a questionable one. However, the geologists apparently have overlooked the work by Rev. Mr. Smith (64) who, while on Nevis, “heard six or seven dull bounces of noise resembling those of Cannon at a great distance pretty quickly following each other at the exact time of this Explosion: as the Sky was quite clear in the eye of the Wind, and as none of my acquaintance there took the same notice of the thing, I durst not venture to insist much upon hearing those dull bounces till I had seen Mr. Boyd.” Mr. Boyd was previously identified as the captain of a merchant ship en route from St. Kitts to Barbados who noted that the sky grew dark and a horrible noise “far surpassing the loudest thunder” and a “falling likewise instantaneously so thick a Shower of Ashes, that the Sloop’s Deck was covered two or three inches

ar)

deep with them. ... They in fright enough turned back homewards [and]... it was soon after found out, That a large Mountain in the Island of Saint Vincent . . . abounding in Veins of Sulphur and Brim-

stone blew up at once, viz. Woods, Rocks &c. all together, which must be allowed to cause a most dreadful Explosion.” By contrast, the 19Q2 erup-

HOWARD, VOLCANISM AND VEGETATION

1962] CF ao H @ SABA st. EUSTATIUS ST. KITTS @ NEVIS bys ANTIQUA » REDONDA j MONTSERRAT 7 GUADELOUPE @ MARIE GALANTE DOMINICA Mi PEEE MARTINIQUE ST. LUGIA , BARBADOS SOUFRIERE Z| ST VINCENT o “THE GRENADINES G GRENADA

Diagonal lines indicate islands and the Soutriére (St. Vincent) Solid black areas

VOLCANIC AREAS OF THE LESSER ANTILLES. with active fumaroles. Mt. Pelée Seen are currently dormant but have erupted in the present century.

are volcanic in origin. Unmarked areas are nonvolcanic islands.

282 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIII

tions of Mt. Pelée and the St. Vincent Soufriére were observed by parties of scientists from eight different countries, and, today, modern seismo- logical stations record the minute signs of activity of these two dormant volcanoes.

The occurrence and duration of fumarole activity have been less accu- rately reported. Nugent’s account of 1810 (42) is regarded as the original scientific observations for soufriéres on Montserrat and other islands. Earlier accounts of hot springs do appear, with the emphasis generally placed on their curative medical properties or the ability to coagulate the protein of eggs. Pére Labat (35), a French missionary to the Antilles between 1694 and 1705, recorded in six volumes his recollections of the Lesser Antilles. In 1696, he visited the Ance de Goyave in Guadeloupe and described the boiling fountains (fontaines boiillantes) in the bay, as well as in a neighboring swamp. He reported the water within a few feet of the coast to be warm enough to cook an egg held in his handkerchief. He attempted to determine the source of the heat and reported that, although the surface of the beach sand was without heat, when he dug to the depth of a foot he encountered sand and water too hot for his hand. This layer, he reported, smoked ‘‘comme on voit fumer la terre qui couvre le bois dont on fait le charbon.” The fumes of sulphur were nearly unbearable. In a neighboring swamp he found a lake approximately 45 feet in diameter which boiled at the edges and also more strongly but less frequently in the center. The quiet periods, Labat reported, were of a duration suffi- cient to allow one to say both a Pater and an Ave. Odors of sulphur were strong here, too, and a sulphurous taste was present in the water. Even today, previously unknown areas of prehistoric soufriére activity are dis- covered in remote places on most of the Lesser Antilles, and new out- breaks of fumarole activity have been recorded in the last decade on Nevis and Montserrat. In general, the most active areas are well known, and the locations of the principal soufriéres are to be indicated in summary for the various islands.

TYPES OF VOLCANIC ACTIVITY

Two basic types of volcanic activity occur in the Lesser Antilles. The most spectacular is the eruptive form so well recorded for Martinique and St. Vincent. Lava eruptions on these and other islands apparently occurred only in prehistoric times. The historic eruptions have been characterized by the forming of muées ardentes, plus ash fall and mud flows. Hill (20) has described Mt. Pelée as an ash pile, a description which applies equally well to the Soufriére on St. Vincent.

Less spectacular, but of longer continuous duration in activity, is the fumarole. The fumarole is generally defined as a hole or vent (in or near a volcano) from which fumes are emitted. Subterranean as well as surface noise is present at most fumaroles. The gases produced may be saturated, ejecting with them large quantities of water, or they may be dry and either hot or cold. The dry, hot gases seem to condense atmospheric mois-

1962 | HOWARD, VOLCANISM AND VEGETATION 283

ture some distance from the orifice, producing steam or clouds of vapor, thereby giving fumarolic areas an eerie appearance. With one exception, the gases and liquids produced by fumaroles in the Lesser Antilles are acid in reaction. This acidity is responsible for the chemical alteration, the coloration, and the physical decomposition of the soil and rocks char- acteristic of the areas around fumaroles.

A cluster of fumaroles is called a “soufriére,” although in the Lesser Antilles this term is loosely applied to solitary as well as clustered vents, to mountains, and even to towns near sulphur-producing vents. A fuma- role producing an odor of sulphur compounds is occasionally called a “solfatara” in this area. Technically, fumaroles are also classified accord- ing to their location as crateral fumaroles, when located in a volcanic crater, and noncrateral, when the point is lateral to a volcanic mass. Noncrateral fumaroles may be primary or secondary, depending on their association with the magmatic heat source. Representatives of all these types are to be found in the area. When the fumarole is depressed, sur- face waters may collect, and the dissolution of rock forms a mud pond or lake. Such bodies are also produced by the damming of valleys below fumarole areas by land slides, as in the case of the Boiling Lake in Domi- nica, Clear-water hot springs commonly found in alkaline areas elsewhere in the world have not been found in the Lesser Antilles.

VOLCANIC ERUPTIONS AND THE VEGETATION

The two active volcanoes on Martinique and St. Vincent are unique in producing only ash accumulations and a characteristic eruptive form termed nuées ardentes during the eruptions observed in modern times. The few lava flows or blocks recorded from the Lesser Antilles are of great age.

The recent and observed volcanic activity in the Lesser Antilles began with earthquakes and noise, the emission of gas and steam, and the ejection of ash And boulders. Some of the steam was derived from the vaporization of a crater lake from the caldera, at or near the summit. In the eruptions of the Soufriére in St. Vincent in both 1812 and 1902, either an eruptive force or a break in the crater wall produced a cascade of boiling water which descended the slopes to the sea. While this boiling torrent followed the established valley patterns on the mountain slope, its heat, volume, and speed killed and removed the vegetation. During its descent, it scoured the river valleys clear of accumulations of ash which thickened the consistency of the water and led to the reports of a lava flow. Anderson (2), who studied the 1812 deposits, concluded that it was a mud flow of accumulated ash and not a true lava flow.

The most destructive of the eruptive features of West Indian volcanoes have been the nuées ardentes. (See Figs. 1, 2.) The French seismologist Lacroix (36) proposed this term for the eruptions of Mt. Pelée and the Soufriére, and it has now become a standard term in all languages, al- though it is occasionally translated into English as “incandescent ash.”

284 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIII

Fic. 1. Nuée ardente descending slopes of Mt. Pelée, April 13, 1931. Photo ei from Perret (44). Fic. 2. ee of nuée ardente on Mt. Pelee, 1930. Photo reproduced from Perret (46). Fic. 3. Ash accumulation on Wallibu fields, St. Vincent, 1902. Photo reproduced ae Anderson and Flett (4).

Smith (63) has described the phenomenon as a “strange black cloud, which, laden with hot dust, swept with terrific velocity down the moun- tainside, burying the country in hot sand, suffocating and burning all living creatures in its path, and devouring the rich vegetation of the hill with one burning blast.” Perret (46) explains its formation as follows: “Tn the acidic type [of volcano| its conduit normally closed because of the relative infusibility of the lava... the lava will have become sur- charged with gases rising slowly fron below, with possible assimilation from meteoric precipitation from above . . . the accumulated gas charge will, with eventual rupture of whatever restraint may have been imposed, go into paroxysmal explosive manifestation in the early stages of an erup- tion. A good instance of this is seen in the nuée ardente. . . . The viscous liquid has, through its own high gas content now coming out of solution

1962 | HOWARD, VOLCANISM AND VEGETATION 285

in distributed gas vesicles, become autoexplosive as a whole and thus capable, on release, of lifting itself . . . clear out of its pocket, in a stupendous en masse expansion. It will be evident that this will convert what had been a liquid into an infinitely subdivided mass; in a word, largely into a cloud of vapor and solidifying particles. Its great weight will precipitate this upon the mountain flanks in a down rushing ava- lanche of very heavy materials at high temperature, emitting gas from every pore amid rising clouds of ash, and constituting what has been termed nuée ardente. This phenomenon was not described prior to 1902, the date of the destruction of St. Pierre, Martinique. . . . The velocity of the onrushing avalanche is phenomenal, and this fact is not unrelated to the continuing evolution of gas. Vapor films between all sclid particles —from which particle the gas continues to be emitted during the ‘life’ of the nuée — <ffectively prevent all solid contacts, leaving the moving mass quite frictionless and capable of flowing upon the slightest inclines.

. The original temperature (certainly on the order of more than 1,000 degrees C.) tends to be maintained by this insulating atmosphere.”’ Per- ret also indicates that in a nuée ardente there is little if any free oxygen. As a result, no combustion occurs during its passage and materials are therefore carbonized and not burned.

The effects of a nuée ardente are localized but lethal to the life of the narrow belt over which it progresses. Nuées ardentes have been photo- graphed from close range and have even been entered by volcanologists. Lava flows in Hawaii may leave little islands of vegetation (kipukas) un- disturbed but. surrounded with incandescent molten rock. Thus, too, in the case of a nude ardente, areas adjacent to the descending cloud may be unaffected by the passage of the ‘incandescent avalanche.” A study of the slopes of the Soufriére in St. Vincent today will show forests of massive trees cleft cleanly by the passage of a muée ardente. On one particular slope there are 50-80-foot trees with trunks 2 to 3 feet in diameter and within a few yards the ash and scoria of the nuée ardente still barely colonized after 60 years. The uée ardente carbonizes the vegetation in its path, while the residue of scoria left behind (in some cases 80 feet thick) retains its heat and kills the Te eae parts .of the plants and the seeds and fruits on the ground. (See F

Quite in contrast. is the effect of the free fall of ash ic may precede or follow the nuge ardente. The volume of free ash is tremendous and the distance it may be carried by the wind difficult to believe. Anderson and Fleet (3, 4) cite reports of a fall of ash on ships 830 miles east-southeast of Barbados, following the eruption of the St. Vincent Soufriére. Ash fell on Barbados, 100 miles away, at the rate of 3.94 tons per acre. Nearly all of St. Vincent was dusted or submerged in free ash from the eruption.

The term “volcanic ash” carries the popular connotation of a light- weight material. Perret states that this is not so, particularly in the ash derived from a nuége ardente. The crushing effect of the masses of ash is clearly documented, and the flattening effect on the vegetation implied. In the case of free-fall ash, the weight of individual particles may be less,

286 JOURNAL OF THE ARNOLD ARBORETUM [ VOL, XLIII

but the cumulative weight equally great on any given area. Fields of arrowroot and other crops in St. Vincent buried under inches or feet of the cold free-fall ash did not recover. The accumulation of ash on leaves and branches caused defoliation and breakage of plants throughout the island. Although the size of the ash particles may have been small, the abrasive effect was noted in the puncture of leaves of crop plants and the removal of bark. A more modern report following the eruption of Paricutin in Mexico appropriately termed this “plutonic sandpaper” (47). The effects of the free ash on the vegetation is limited to abrasion, crush- ing, and smothering. The ash is generally chemically inert and without heat. The plants affected may recover or sprout from the base. Under- ground portions of the plants are not killed, and seeds germinate and may push up through the covering of ash.

The days which followed the eruptions of Pelée and the Soufriére have been reported as stormy. Heavy rain showers occurred. Perhaps the ac- tivity of the volcanoes caused the formation of the heavy rain clouds; but in any case, violent rain storms swept the area, and exceptionally heavy rain descended on the ash-covered slopes. Russell (52) noted that “in- stead of being a protection to the surface on which it rests, the fresh debris is in many instances of assistance in its more rapid erosion. On steep slopes, and even when the surface is nearly level, the rills formed during the numerous tropical showers quickly cut through the loose sur- face material, and aided by the angular particles in suspension, corrode the soil or rocks beneath. The rains, as it seems, are heavier than usual, owing to two causes: first, the great amount of water contributed to the atmosphere as steam, and second, the vast amount of dust blown into the air, each particle of which serves as a center for condensation.’ River- beds filled with ash and scoria were swept clean, scoured to new depths and often altered from their previous courses. As the ash deposits were removed, the lower soil areas, too, were eroded in sheets of mud which were deposited along the coast in fans recorded as 70 to 80 feet wide and seven miles in length. Subsequent earthquakes caused landslides in shaken and loosened terrain. North of Chateaubelair in St. Vincent, Russell re- ported “strips of nearly flat alluvial land, adjacent to the sea, have dis- appeared leaving fresh bluffs of loose debris some 30 or 40 feet high.”

The revegetation of areas affected by volcanic activity has been the basis for classic studies in the case of Krakatoa. For the West Indies, Perret (44, 46) and Anderson (2) made initial studies and Stehlé (66) and Beard (6, 7) some recent summaries. A more detailed study is still required, with consideration given to the specific nature of the disturbance in each area. Perret noted the occurrence of ‘several mosses and fungi” within two years of the 1930 eruption of Mt. Pelée in an area affected by a nuée ardente. Judging from his photographs, the “fungi” are lichens of the genus Stereocaulon. Perret also noted the invasion of the “golden back fern” and a plant with “tiny blue blossoms, said to be invariably yellow in other localities.” As both gold and silver forms of Pityro- gramma calomelanos and P. chrysophyila occur in the Lesser Antilles,

1962 | HOWARD, VOLCANISM AND VEGETATION 287

the scientific determination of plants cited in the geologists’ reports are not to be trusted. Perret also commented that while the golden form of the fern occurred at lower elevations, only the silver form was found at higher level on the disturbed areas. Although the plant with blue blos- soms can not be identified, Pitcairnia spicata is represented by the yellow- flowered var. sulphurea near craters and by the red-flowered typical form where volcanic fumes are not present.

Both Anderson (2) and Sands (53) visited the St. Vincent Soufriere in 1907, five years after the last eruption, and noted the species which invaded the area. Although Beard cites Anderson’s paper, he quotes only from Sands’ less critical report in which 154 species of plants were col- lected from low-altitude areas badly devastated by the eruption. Ander- son, who was present during the eruptions, differentiates between areas affected by the nuée ardente and those affected by an ash fall. He states, “The incandescent avalanche swept down the Wallibu valley and spread out over the old fan or plateau at its mouth. it then turned south round the lower end of the Richmond Ridge and destroyed the Richmond Works and all the vegetation near them. The ash still remains to a depth of two to six feet in different parts, and the old roots are completely buried and thoroughly destroyed, but the avalanche was confined to the bottom of the valley and none of its effects are visible on either side. . . . The sur- face of the ash near the Works has not consolidated, but is rapidly break- ing up under the influence of plant roots, and humus is being formed. . . .

e chief new plants are Castor Oil (Ricinus communis), which grows in luxuriant masses along and around the ruins of the Works, and a plant, Cattle Tongue (Pluchea odorata) which has already formed flourishing bushes taller than a man. Besides these, Indigo (Jndigofera anil), Sensi- tive Plant (Mimosa pudica), Guinea Grass (Panicum maximum), Eupa- torium odoratum, and two grasses were also noted.” Where the surface of the ash was consolidated, no vegetation can spring up; where it was broken by water or the hooves of animals the first invaders proved to be silver fern, grasses, and Pluchea. For areas covered with free fall ash, Anderson also related, ‘‘On the south slope leading up to the Morne Garu range the return has been considerable. The surface is still studded with the charred and bleached skeletons of trees, which appear to have been killed universally with the single exception of a small lateral valley north of the lower part of the Richmond Ridge, where a few palm trees in a sheltered position have recovered. The shrubs and herbaceous vegetation, which were all burnt level with the ground, are gradually returning, in many cases from the old roots, since the removal of the thin covering of ashes by the rain.

“The top of the Wallibu plateau was entirely devastated. The trees remain only as bleached trunks except a few which have recovered in sheltered positions at the ends and south edge of the plateau. The dead trunks show that the ash was never more than a few feet thick at the most, and the whole is now covered with a luxuriant growth chiefly from the old roots.

jan

288 JOURNAL OF THE ARNOLD ARBORETUM [ VOL, XLIII

“At about 800 feet Tree Ferns (Cyathea arborea) become very abundant and large sheets of them are common, and generally it may be said that vegetation is luxuriant up to a height of about 1,000 feet, abundant up to 1,500, and very sparse above that height, with only a few grasses and silver ferns; higher up nothing but mosses and lichens are found.

“At the lower lip of the crater and just inside it mosses and lichens only are found. The mosses have been identified at Kew as: Pogonatum tortile, P. Beauv. and Philonotis tenella, Jacq. and the lichen as Stereo- ee sp.’

Beard (6) presented a study of the altitude zonation of vegetation types in undisturbed areas and, in comparison, the types he found on the slope of the Soufriere in 1940, approximately 40 years after the erup- tion. Beard found at 2,400 feet altitude that the “plants are knee high only, still composed of silver fern with now very stunted Charianthus and Cyathea, Pitcairnia, Lobelia, ferns and grasses.

“At 820 m. Charianthus and silver fern begin to thin out, and Cyathea is only 60 cm. high. Towards 900 m. the vegetation is simulating ‘paramo’ and is distinctly alpine in appearance. The ground is entirely and thickly covered with lichens, thallose, foliose and crustose. Higher plants are scattered and tufted and include the Lobelia and Pitcairnia, several grasses, a Lycopodium and a few ferns. Cyathea is now at most 30 cm. high, and the general vegetal covering is ankle deep only.

“At the highest levels one finds a tundra only. The stony ground is thickly covered with lichens and, in between, a few clumps of Lodelia and Pitcairnia, The lichens vary greatly in colour, being predominantly grey or orange, and having a striking appearance at a little distance.”

It is interesting that the inside walls of the crater are covered with larger plants, right down to the water line nearly 1,000 feet below. Here Bac- charis cotinifolia, Gleichenia, Lycopodium, ferns, grasses, and a few Cyathea occur in greater stature than at comparable levels on the outer slope of the crater.

Stehleé (66) described the revegetation of the slopes of Mt. Pelée and also determined that the altitudinal zonation existing before the eruption was re-established afterwards. On both Mt. Pelée and the Soufriére it was apparent that the nuée ardente completely removed all plants and their propagulss, The ash fall, however, may have killed the es of plants

ove the ground but allowed subterranean root and stem ystems, as well as seeds, to survive and grow. In both cases, the iene was more rapid at lower elevations where a weedy flora was the first type to be established. At all but the highest elevations, lateral invasion from the areas of vegetation unaffected by the nuée ardente was apparent. At the highest elevations where no unaffected areas were present and the soil was isolated both vertically and horizontally from seed sources which would tolerate those environmental conditions, the reinvasion and _ re- establishment was slow, and often an unexpected composition was cre- ated. Lichens and mosses were the first and most successful invaders. Wind-blown spores of ferns were introduced, and the plants created pure

1962] HOWARD, VOLCANISM AND VEGETATION 289

stands of many individuals. Although spores of Lycopodium were eventu- ally introduced, only scattered plants were found. The seeds of Pitcazrnia, Cladium, and various grasses could have been distributed by wind to these new areas. Stehlé suggests that the fruits of Charianthus, Lisianthus, Lobelia, Tibouchina, Ilex, Weinmannia, and Clusia are distributed by birds or bats, and these plants are slow to become established in the devastated areas of high elevation. At lower altitudes birds, cattle, and man have played a role in starting new plants. Stehlé, for example, no- ticed the invasion of sugar cane on the slopes of Martinique and suggested that a wind-blown seed allowed this plant to be established on the scoria. Perhaps a node dropped by a wandering cane-chewing sightseer might be a more logical interpretation. The seeds of Acacia and Mimosa observed by Sands, Perret, and others were spread by cattle wandering over the newer ash accumulations

It is of some interest. to note that in all cases the growth of invading species appeared to be more rapid inside the crater of the volcano than at comparable levels on the outside. This growth can probably be asso- ciated with the greater protection afforded within the crater walls, as well as the more abundant moisture available on the less porous soil surface.

FUMAROLES AND THEIR EFFECTS

The occurrence of fumaroles, either singly or clustered as soufriéres, can be reported for all the major islands of the Lesser Antilles from Grenada, in the south, to St. Kitts, in the north. The fumaroles are found at many elevations — submarine near Nevis, coastal on Montserrat, at low eleva- tion on Grenada, and at the summit of the Soufriére (4800 feet) on Guadeloupe. The fumaroles are present in many vegetational types, in- cluding the thorn scrub of Grenada, old field vegetation on Dominica, mesophytic forests on Montserrat, and the volcanic facies on Guadeloupe. Active fumaroles can be determined from some distance by the production of heat, steam, and fumes laden with sulphur compounds. Near by, they show a characteristic destruction or absence of adjacent vegetation, if the fumes contain sulphur compounds, or an equally characteristic altera- tion of the rock texture and color. Extinct or dormant fumaroles may be recognized in open areas where mosses, algae, or perhaps a few higher plants cover the surface or where the softened rock in a restricted area reveals the former fumarole and its activity.

The activity of the fumarole may vary considerably in intensity over a long period of time or even in relation to the annual rainy or dry seasons. Some recurrence of fumarolic activity has been noted and associated with active volcanism on adjacent islands or with earthquakes of considerable geographic latitude. Within the Lesser Antilles are dormant fumaroles, those with long records of continuous activity, new ones, and those sporadic in their action.

Nearly all fumaroles are characterized by the emission of heat. Stehlé

290 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLII

(67) gives temperatures of the crateral fumaroles of Mt. Pelée as high as 700° C. Robson and Willmore (50) have published the most recent studies of heat emission from fumaroles in the Lesser Antilles. They have suggested as a base point the temperature of 24° C. for meteoric water or ordinary cold springs and for the air temperature. Steam temperatures from Ie Qualibou soufriére in St. Lucia have been measured as high as 185° C., while water outilow temperatures from ponds in the same soufriére varied from 63° t On Montserrat, Robson and Will- more recorded the flow of water from Gages Upper Soufriére at tempera- tures of 34-86° C. from three sources and at Galway’s Soufriére from

© to 97° C. from two sources. Crater lakes, which may not show vio- lent actions typical of fumaroles may show the effects of their proximity to sites of volcanism. The mean lake temperatures for the crater lake in the Soufriére of St. Vincent was 27.1° C., indicating at least a 4° rise in temperature due to the heating effect of submerged fumaroles. There is only one record of plant life in the boiling lakes of the Lesser Antilles, but this has never been verified in more recent observations. No animal life has been reported. Crater lakes, by contrast, have an algal or higher plant flora, and the crater lake on St. Vincent’s Soufriére has recently been stocked with fish. The heat of water flowing from fumaroles or boil- ing lakes is soon dissipated. Blue-green and green algae, diatoms, and even sedges have been collected from outflows with temperatures of 56° to 85° C. within five feet of vents on several islands. Unfortunately, no identifications have ever been received from specialists who might name these collections of plants properly.

The direct emission of steam could have a noticeable destructive effect on the surrounding vegetation. Few plants could stand blasts of tempera- tures of 100° to 185° C. However, the direct blast of steam heat prob- ably is of minor aoe or none at all in comparison with soil heat, toxic gases, and cumulative acidity. Although MacGregor (39) gives in tabular form records of heat from soufriéres on Montserrat, he also points out that the maximum temperature recorded depends on how far the thermometer can be pushed down the orifice against the gas pressure. Temperatures are higher below than at the surface, and such heat as is emitted is dissipated rapidly in the atmosphere.

For many of the West Indian fumaroles there appear to be few if any toxic substances included in the air being forced out at the present time. Around the fumaroles of Wotten Waven, in Dominica, the vegetation, in- stead of being adversely affected or killed, appears to grow more luxuri- antly in the local area of higher heat and humidity. However, in many fumaroles there must be toxic substances, for the fumes have a definite effect on the vegetation. MacGregor and other authors have stated that the fumes kill the local vegetation. Even the names of the fumaroles and their areas suggest the toxicity. The largest soufriére on St. Lucia is known as ‘“Qualibou,” a name meaning “place of death,” and some soufriéres in Dominica are located in the “Valley of Desolation,” It is clear that the fumes are often noxious and toxic to men and to plants. The death of

1962 | HOWARD, VOLCANISM AND VEGETATION 291

two visitors to the Boiling Lake of Dominica has been substantiated. The geologist Perret was overcome by fumes while residing in his field station within Gages Lower Soufriére, and workmen on an earlier project in the same area were overcome. Although several geologists have attempted to collect and analyze the gases, the results are inconclusive. Romer (51) has reported the fumes from fumaroles on Mt. Pelée to contain only car- bon dioxide, sulphur dioxide, and traces of hydrogen chloride; however, some fumaroles on Montserrat failed to give positive tests for carbon

xide. Obviously the composition of the gases may vary with the nature a the activity of the fumaroles. Most fumaroles of the Lesser Antilles produce hydrogen sulphide, but all characteristically reveal an absence of fluorine, relatively few chlorides, and negligible amounts of boron. Perret described the gas of Gages Lower Soufriére and concluded it was ‘“‘a gas having a traitorously pleasant odor, but extremely irritating to the eyes and respiratory passages. . It was highly unstable and decomposed into hydrogen sulphide and siphur. It may have been one of the per- sulfides of hydrogen.” Willmore (71) stated, “These vents emit a mix- ture of steam and sulphur gases which, during quiet periods, usually have a temperature a little below that of boiling water. During periods of abnormal activity the temperature of the exhalations often rises, and the predominant sulphur constituent may change from hydrogen sulfide to sulphur dioxide. — The tendency for the temperature to become stabilized near boiling point is readily explained on the assumption that the sou- friére is supplied by hot juvenile gases which bubble through meteoric water on their way to the surface. The change in gaseous constitution during active periods is presumed to occur because the original magmatic gases undergo reactions on their way to the surface, so that the enc products depend on the temperature and velocity of the flow.” The a sence of fluorides and the relatively small amounts of chlorides and boron support the conclusion that meteoric and not juvenile water is involved in most soufriéres of the Lesser Antilles, at least most of the time.

The destruction of the vegetation around fumaroles, therefore, is ap- parently not caused by the chemical elements of the fumes or the water but by their nature as compounds. The various sulphur compounds in the presence of water can produce a series of acids. Litmus papers and “pHydrion” papers respond quickly, giving acid reactions. [Perret re- ported that during the seismic crisis on Montserrat egg shells dropped in the fumarolic waters give rapid effervescence. Another mildly acid con- stituent of the fumarole fumes is hydrogen sulphide. The presence of this compound is evident, not only in the odor, but in the tarnishing of all silver brought into the vicinity. The town of Soufriére in St. Lucia is unfortunately in the lee of the Qualibou soufriére, and one is continu- ously aware of the derivation of the name of the town. Silverware tar- nishes in the town of Plymouth in Montserrat which is downwind from the Gages soufriéres. Perret depended on the reaction of lead-acetate paper to sulphur to record the intensity of such emission from the Gages

—

292 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLII

soufriéres, and during the same seismic crises even the lead paint black- ened on ships in the roadstead.

The acids derived from the various sulphur compounds also affect the soil and rocks in the vicinity of the fumaroles. MacGregor (38, 39) con- cluded that ‘mineral changes brought about in the country rock render it locally pale to pure white in color.” The pale color of soufriére areas renders them conspicuous and easily recognizable from a considerable dis- tance. Accompanying the color change in the rock is a loss of structural hardness. Rocks in fumarole areas may retain their original shape and texture but become soft and nearly putty-like in consistency. The novice learns quickly of the treacherous footing in areas of fumaroles, where even the largest stones are not to be trusted.

When both hydrogen sulphide and sulphur dioxide are present, a chemi- cal reaction may take place oxidizing the former and reducing the latter to produce water and crystalline sulphur. The deposits of sulphur crystals around the vents of Lesser Antillean fumaroles are colorful. Local needs for sulphur are easily met and at times small quantities of sulphur have been exported from the area. The commercial extraction of sulphur as a by-product of the use of steam for thermal power has been proposed

(8).

EFFECTS OF FUMAROLES ON VEGETATION

During a visit to many of the fumaroles in the Lesser Antilles, it be- came apparent that there were areas around most fumaroles where no vegetation survived; that leeward from the vents certain plants were affected or killed by the emissions of the vent, while to the windward the same species approached or persisted much more closely; that certain species tolerate the effects of the fumaroles longer than others; that cer- tain species invaded an affected area more readily than others; that some physiological change took place in the species which did grow success- fully near the fumaroles in comparison with the same species elsewhere; and that the effects of the fumaroles on the vegetation were greater at certain periods than others. A review of the literature reveals that the earliest descriptions of the Boiling Lake on Dominica do not vary from observations made recently, and that Sapper’s sketch (57) of the limit of devastation of vegetation at Galway’s Soufriére on Montserrat differs little from a sketch map made last year. As long as the fumarole remains ac- tive and reasonably constant the effects on the vegetation are noticeable to a distance limited primarily by the terrain. Furthermore, the chemi- cally altered rock and soil near the fumarole remain an inhospitable area for plants for decades, and revegetation is an extremely slow process.

owever, new vents do open and I have had the opportunity of record- ing a sequence in the change of the vegetation near a new fumarole over a ten-year period on the island of Montserrat. In 1950, I visited both Gages Lower Soufriére, studied by Perret in 1934-37, and Gages Upper Soufriére, which was reactivated by an earthquake in 1937 when Gages

1962] HOWARD, VOLCANISM AND VEGETATION 293

Lower Soufriére was dormant. Gages Upper Soufriére was active, but the vegetation around it appeared to have been stabilized in relation to the vents. Photographs were taken at both soufriéres, notes were made, and specimens were collected approaching and adjacent to the devastated area. I returned to the area in 1961 expecting to determine the identity and the role of the plants which had successfully invaded during the past decade. At Gages Lower Soufriére I was not disappointed. Several four-foot plants of Clusia alba, as well as other weeds, were growing in the center of the ruins of Perret’s field station. Young plants of Andropogon, Philodendron giganteum, and Pitcairnia were seen, and a coating of mosses colored many of the rocks.

The situation at Gages Upper Soufriére was another matter. New vents had opened in a thicket downstream from the previously affected area, affording an opportunity of determining the initial effects of a fumarole on previously undisturbed vegetation. The exact date of the inception of this vent is not known. Martin-Kaye (40) observed the area affected by Gages Upper Soufriére in 1954. A marker tree, a large buttressed rooted specimen of Sloanea massonii from which I had col- lected a specimen in 1950 was now dead, and only the stump remained some 50 feet inside the devastated area (Fig. 13). A vent had opened in the last decade directly under that tree, the base of which was now discolored white and yellow. The remaining stump was hot from the steam and wet and rotted, so that an 18-inch machete penetrated easily to its hilt, while “pHydrion” paper pressed against the sodden wood showed a pH of 1.5. The activity of the Gages Upper Soufriére had not only been extended downstream but up the hillside nearly to the ridge (Fig. 12). It apparently will be only a short time before the fumarole activity breaks through the crest of the ridge and effects a new drainage system.

In the newly affected area readings of temperature and the pH were taken on occasions between 8 A.M. and 5 P.M. over a period of two weeks with no significant variation in the figures. The vent of greatest interest (Fig. 18) was in a thicket composed of species of Blechnum, Charianthus, Clusia, Cyathea, Ficus, Ilex, Miconia, Palicourea, Philodendron, and Symplocos. No noticeable hole or pore had developed, and the steam appeared to be issuing from an area of dry soil about four inches in diameter. Outside of this small focal point some organic material was present but this was extremely dry. The diameter of the area of dry duff was approximately four feet. Outside of this area the debris on the soil surface was noticeably warm and moist, and dead and dying plants were obvious and readily identifiable. The effects of this new vent ex- tended for a distance of only 20 feet. The temperature reading in the vent area was 94-96° C., a measurement obtained by inserting to a depth of three inches a laboratory thermometer calibrated to 250° C. In the area of dry and barren soil around the vent the temperature was 84° C. at a depth of three inches. The most tolerant plants in the area of wet, warm soil nearest the vent were Blechnum, Clusia, and Philodendron.

294 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLII

Beneath these plants the soil temperature was 54° C. Progressing out- ward on several radii, it was determined that species of Ficus and Lyco- Hey grew in soils at 48° C., while Miconia and Palicourea withstood sevond this point the soil temperature dropped to 30° C., and no ice was evident. The dead or nearly dead plants in soil temperatures of 40° C. and above were Charianthus, Cyathea, Ilex, Miconia, Palicourea, and Symplocos. About 100 yards away from this vent in the undisturbed forest, soil and air temperatures varied between 24° and 26° C. at dif- ferent times of the day. It is apparent that one previously unrecognized factor in the death of plants around new fumaroles is the heating effect of the soil. Although the fumes from this small vent did discolor lead acetate paper they gave no reading on “pHydrion” paper. Litmus paper gave an acid reaction very slowly. The same papers used on the moist soil zone where plants had been killed gave readings of pH 1-1.5. The vegetable material present in the dry zone crumbled easily in my fingers, That of the moist zone, by contrast, was slippery or almost mucilaginous. The structure of the duff had been altered, and I judged from the ease with which my feet slipped that a heavy shower of rain would cause severe erosion. The areas unaffected by heat or low pH were more fibrous to the touch and more resistant under foot. Readings taken in the vege- table material in the unaffected forest showed pH 4—4.2. Hardy, Rod- rigues, and Nanton (19) report that the majority of agricultural soils on Montserrat range from a pH of 6.1 to 7.4. Their two sets of figures for forest sites indicate pH readings of 4.3—4.8 (for lower montane rain forest soil on Frenchman's Hill) to a pH of 5.2—6.6 (for soils in a palm brake on Chance’s Mountain). The initial killing effect of fumaroles may well be, first, soil temperatures beyond the range of tolerance of many species and, secondly, high acidity.

General observations were made on the vegetation within a few yards of the older area of devastation around Gages Upper Soutfriére. Although the heat of the steam emissions could be felt in this area, it was not oppressive, and temperature readings were only a degree or two above those recorded deeper in the forest. The odor of hydrogen sulfide was noticeable, as was a higher humidity. Wet litmus paper did not indicate an acid reaction during a 30-minute exposure. Nevertheless, plants were dying and a slowly cumulative toxic effect of the fumes is the only ex- planation. The genera most susceptible appeared to be Cyathea, Fr eziera, Hirtella, Licania, Palicourea, Peperomia, Philodendron (except P. gigan- teum), Psychotria, Symplocos and Weinmannia. Species of Gonzalagunia, Hex, and Sloanea appeared more tolerant and those of Clusia and Tern- stroemia, as well as Philodendron giganteum, most resistant.

The oldest areas of Gages Upper Soufriére showed scattered plants of Clusia alba, Cyperus ligularis, Philodendron giganteum, and Pitcairnia spicata, It is not clear whether these plants were residual from the former vegetation or more recent invaders. They grew both as individuals and in groups. In numbers they were more abundant on the fumarole affected soil than in comparable areas of the undisturbed forests. Vents of two

1962 | HOWARD, VOLCANISM AND VEGETATION 295

types were present in the areas where these plants grew. One type issued only fumes while the other had steam so laden with moisture that it con- densed around the orifice and ran off, forming a small stream. The air issuing from the dry vents was recorded as high as 93° C., but the soil temperature immediately around the vent was 37° C. Incrustations of crystalline sulphur a such orifices. The moist steam vents, however, gave readings of 37° to 70° C. taken in a comparable manner, and the “pHydrion” paper gave a reading of pH 1.5. Barren soil temperatures around these vents were 30° C. Temperatures recorded from around the root systems of Philodendron and Pitcairnia were 32° C. and the

was 3. The highest water temperature recorded in any of the small pools within Gages Upper Soufriere was 94

The Qualibou soufriere in St. Lucia alowed an example of vegetation stabilized to the long-term effects of fumes from eee This sou- friére is located in a valley with a high hill on the windward side so that fumes from the fumaroles tend to rise along the hill face. The vegeta- tion of this hillside consisted of stunted low plants of Blechnum, Clusia, Cyperus, Gleichenia, Lycopodium, and Pityrogramma. Just over the crest, in an area where the fumes were dissipated by the prevailing winds, the native forest flourished in an unaffected manner. The species which tol- erated the fumes on the hillside were those with heavy cuticles, for the wax-rigid nature of the foliage was conspicuous in contrast to that of the same species growing in an area where they were unaffected by the fumes. The waxy coating on the leaves apparently made this vegetation more Susceptible to burning, for a large fire scar dominated the face of th n a few places a clear line of demarcation was evident between the plants which burned and those which only wilted down. This de- marcation appeared to be associated with the plants exposed to the fumes and those protected from it.

An example of the periodicity of activity of fumaroles and the killing effect on the vegetation can be found on the slopes of the Soufriére on Guadeloupe. Stehlé (65) has described the units of vegetation which

ave been established at successive altitudes to the summit. His “Clusi- etum,” ‘“‘Lobelietum,” ‘“Pitcairnietum” represent, in part, tolerance of the component species to temperature, wind, and cloud cover but also to the pH of the ash terrain and to the effects of the fumaroles at the summit. During periods of little fumarole activity, the shrubs and trees may de- velop normally. When the fumaroles become active, the acid fumes will kill many of the branches and occasionally entire plants. Generally these plants recover by sprouting from the lower branches or the base, but the upper branches remain dead and leafless.

A unique fumarole area, perhaps more properly termed a hot springs, is found in Chambourg section of St. David’s parish in northern Grenada. The twelve small bubbling hot springs apparently are influenced by the old volcanic mass below the crater lake, Lake Antoine. The soils appear to be coral limestone and a sandstone which includes volcanic ash. Dur- ing certain seasons the waters of the springs overflow and several sizeable

296 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLII

cones have been created. An alkaline litmus reaction was obtained from the waters collected in overflow pools. This, to my knowledge, is the only alkaline fumarole in the Lesser Antilles. No plants from the surrounding scrub of Acacia, Acrostichum, Avicennia, Chrysobalanus, and Laguncularia have encroached on the rock areas built up by the hot spring.

The Grenada hot-spring area is the only one in the Lesser Antilles ven- erated by the local people. One vent had an animal sacrifice placed there by religious cultists. Several springs had fresh or wilted bunches of flowers in the orifices. Candles and coatings of wax were present in the largest vent, along with a painting of the Sacred Heart. The late Father Bryan Proudman, who was my host in this area attributes the tokens to “‘shandoo”’ ceremonies.

CATALOGUE OF THE FUMAROLE AREAS OF THE LESSER ANTILLES

Grenada

THe GraNnp Etanc. This crater lake is in the caldera of an old vol- cano. Beard (7) has described and illustrated this area and its vegeta- tion. The lake temperature indicates the presence of some heat source. Sulphur odors are noticeable at times, and particles of sulphur occasion- ally appear on the surface of the lake. No effect of the heat or acidity can be noticed on the surrounding vegetation.

LAveRA Hot Sprincs. (Figs. 4, 5.) About twelve small vents or pot holes near Lake Antoine produce an occasional weak odor of hydrogen sulfide. These are the only alkaline reacting springs that I have encoun- tered. The vegetation has not encroached on the sheets of hard rock formed by the overflow from these springs.

St. Vincent

THE SOUFRIERE. Interest in the eruptive volcano on St. Vincent centers on the accuracy of the reports of its eruption in 1902. Prior to that time, eruptions have been recorded for 1718 and 1812. Reports of the former eruption have been considered by Anderson and Flett (4). They point out that while Huggins (32) discredits such reports, Defoe (37) only in part exaggerates in his story, which may indeed have been based on fact. The letters of Rev. Smith (64) may be considered corroborative evidence that the Soufriére did erupt in 1718.

Mr. James Anderson apparently was the first white man to ascend the volcano. In an account published in 1785 (1) he described the summit. “We found very little obstruction in our way up until we got to the place where I returned, and there, for about a quarter of a mile, we had con- siderable difficulty to clear our way through grass and ferns. After we came within a quarter of a mile from the top, we found ourselves in an- other climate all at once, the air very cold, and the vegetable productions changed; here was nothing but barrenness over the whole summit of the

1962] HOWARD, VOLCANISM AND VEGETATION 297

mountain. On the confines of the grassy region and the barren I found some beautiful plants. Moss grows here in such plenty that I frequently sunk up to my knees in it. This is the only place in the West Indies that produced any moss that I have seen. About noon we gained the top of the peak . . . when in an instant, we were surprised with one of the grandest and most awful scenes I had ever beheld.” For the crater An- derson reported, “In the centre of the bottom is a burning mountain, of about a mile in circumference, of a conic form, but quite level. . From the external appearance of this mountain, I imagine it has only Perce to burn lately, as on several parts of it I saw small shrubs and grass, which looked as if they had been scorched and burnt. ... On ae opposite sides of the burning mountain, . . . are two lakes of water.’

An anonymous account published in the Evening News of July 30th, 1812, is reprinted by Anderson and Flett (4). This account described an ascent of the Soufriére on April 26th, only a few hours before its eruption at noon on the 27th. On viewing the crater, the anonymous report stated, “Exactly in the centre of this capacious bowl rose a conical hill, about 250 or 300 feet in height, and about 200 in diameter, richly covered and variegated with shrubs, brushwood, and vines about half-way up, and for the remainder powdered over with virgin sulphur to the top. . . . The precipitous sides of this magnificent ampitheatre were fringed with various evergreens and aromatic shrubs, flowers, and many alpine plants. On the north and south sides of the base of the cone were two pieces of water, one pure and tasteless, the other strongly impregnated with sulphur and alum.”’ A comparison of the two reports indicates a secession of the ac- tivity of the cone and a re-establishment of the vegetation. The eruption changed all this, for as Shephard (62) stated, ‘all the former beauty of the Soufriére was, of course destroyed; the conical mount disappeared, and an extensive lake of yellow-coloured water, whose agitated waves perpetually threw up vast quantities of black sand, supplied its place.” Hooper’s account in 1886 (24) suggests that a single lake remained in the crater. Anderson and Flett (4) reported that the crater “contained three small lakes of water, greenish and turbid, that in the southeast cor- ner was throwing up jets of mud and steam with a hissing noise.” In the years which followed, an accumulation of rain water apparently raised the surface of the three lakes and made them confluent. Beard (6) found only one lake in the crater during his visit in 1940, and such is the situa- tion today. The southeast corner of the lake, however, is often discol- ored with suspended material and flakes of sulphur, indicating that all is not quiet in the fumarole area.

Beard has described the progress of plant succession on the Soufriére in a paper published in 1945. A new study, now that an additional two decades have passed, appears to be in order. I visited the summit of the Soufriére in 1950 and found sufficient differences from Beard’s description to suggest that the initial invasion of plants had been successful, that a favorable environment had been created, and that rapid changes were then taking place. At that time, the floral composition of the summit was

298 JOURNAL OF THE ARNOLD ARBORETUM [ VOL, XLIII

closer to that described by Hooper prior to the 1902 eruption, with the sole exception of an abundance of fleshy liverworts unmentioned by either Hooper or Beard.

St. Lucia

THE QUALIBOU SOUFRIERE. (Figs. 20, 21.) This appears to be the only active soufriere on the island of St. Lucia and is situated about a mile and a half southeast of the town of Soufriére. Hooper visited the area in 1SS6 and reported briefly that “the exhalation . . . has reduced the vegetation on both sides within reach to a carpet of brake fern.” In 1903, Sapper (56) and, in 1904, Hovey (26) described the area as between two and three acres in extent, but neither made reference to the vege- tation. In 1949, Beard (7) noted that a specialized vegetation was found around the Sulphur Springs (the Qualibou vents). “Clumps of Pitcairnia are as usual found nearest to the vents with some Cyperus ligularis. Far- ther out, these are joined by various species of ferns, principally Blech- num serrulatum dominated by large clumped bushes of Clusia plukenetii and C. alba. The whole hillside above the Springs is carpeted with Blech- num ferns with an occasional Clusia.”

The vegetation around the Qualibou soufriére appears to be com- pletely stabilized. Algae are found in dense mats in the drainage of the hot springs. Cyperus and Eleocharis plants are common in shallow warm pools of seepage areas. None of the larger plants seen in the area showed any damage from the constant emissions of the fumaroles.

Recent interest has centered in the possible utilization of the thermal heat for generating electricity. The reports of Bodvarsson (8) and of Robson and Willmore (49) are pertinent to this problem.

Martinique

Mr. PeLveée. The destructive eruption of Mt. Pelée in 1902 eliminated the town of Saint Pierre, killing all but one person. The latter stages of this eruption, as well as that which occurred in 1930, are well described by numerous authors (2, 3, 29, 36, 39, 44, 52). Bibliographies of signifi- cant publications are given by Hovey ( (28) and Russell (52). Few specific data are available on the regrowth of the vegetation, and a modern survey seems desirable. Stehlé visited the area and published in 1938 a general consideration of the regrowth (66). He recognized conditions comparable to those of Guadeloupe, excepting the sulphur fumes, and suggested that the vegetational zones were a ‘Clusietum martinicense,” a ‘*Lobelietum martinicense,’ and a “Pitcairnietum martinicense.” Stehlé attributed to larger animals the introduction of Mimosa pigra and Acacia tortuosa to the areas of recent lavas. Birds, he felt, accounted for the introduction of Ficus, Melia, Trema, Tetrazvgia discolor, and Miconia striata. At the lowest elevations on new lavas, Stehlé found common weedy species of Crotalaria, Cassia, Emilia, Fuphorbia, Bidens, Ageratum, Indigofera, Pluchea, etc. Slightly higher and probably introduced by wind were plants

1962 | HOWARD, VOLCANISM AND VEGETATION 299

of Ochroma pyramidale, Ceiba antillana, and Tecoma stans. Near the top of Mt. Pelée, Stehlé reports the occurrence of two species of Des- modium and single species each of Stenotaphrum, Erigeron, Sauvagesia, Mikania, Arthrostemma, Hypoxis, Cuphea, Rubus, Phenax, and Cyperus. The majority of these are not found at higher altitudes in undisturbed vegetation on Martinique or nearby islands. No information is available on the length of their persistence. Stehlé indicated that the most power- ful colonizers were Pitcairnia bracteata, Guzmania plumieri, Tibouchina chamaecistus, Baccharis cotinifolia, Pityrogramma calomelanos, and two species of Lpidendrum.

Thermal springs are recorded from the vicinities of the towns of Frégate and La Redoute, along the coast near Marigot, and on the siepes of Mt. Pelée above Macouba. I have not located descriptive literature regarding their activity as fumaroles or the vegetation around them.

Dominica

The earliest record of fumaroles on Dominica appears in Atwood’s ac- count of the island pend in 1791 (5). His vivid, but questionably accurate description stated, ‘“‘These sulphurous mountains are certainly among the most wonderful phaenomena of nature, and command our astonishment and admiration. To see vast tracts of land on fire, whose smoke, like clouds, stretches far around: brimstone in flames, like streams of water issuing from the sides of precipices; in the vallies large holes full of bituminous matter, boiling and bubbling like a caldron; the earth trembling under the tread, and bursting out with loud explosions, are objects truly terrefic to the beholder; who, on the spot, are struck with awe and admirations, on viewing such dreadful works of the Almighty, who causes them to exist, for purposes only known by him.” Edwards (15) also reported about Dominica that ‘several of the mountains contain ee volcanoes, which frequently discharge vast quan- tities of burning sulpht ‘rom these mountains also issue springs of ot water. ...° There are no reports of eruptions on Dominica in his- toric Himes! and these two early descriptions can not be definitely located geographically. It is possible that these both refer to the imposing Val- ley of Desolation and its famous Boiling Lake, and represent its earliest record. The discovery of the Boiling Lake is generally credited to Dr. H. A. Nicholls around 1880. Nicholls’ route was soon followed by Ober (43), Johow (33), and others, all of whom comment in one way or an- other on the nature and composition of the vegetation in the Valley of Desolation and around the Boiling Lake.

Hovey (30) and Earle (14) were the first to consider the geology of the area. Later in the twentieth century Domin (13a) and Hodge (21, 22) studied the vegetation of Dominica, and both wrote extensively on this subject. Hodge has said of the Valley of Desolation, ‘““The combined sulphurous fumes of all the vents and fumaroles have had a blighting effect far up the slopes, wherever they have come against the vegetation.

=

300 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIII

The end result has been the formation of what is known locally as the Valley of Desolation.’ [See Fig. 6.| Along the ridges above lies Elfin woodland but in the valley only patches of it remain and in these Clusia venosa and an /lex seem best able to tolerate the noxious gases. In the proximity of the fumaroles vegetation is absent or sparse and except for the colorful algae to be seen in the warm water of the streams, is appar- ently unable either to tolerate the noxious fumes or to root in the mineral- impregnated soil. The only pioneer species observed close to the vents were the bromeliad Pitcairnia spicata var. sulphurea, a grass, Ischaemum latifolium, and occasional plants of Dicranopteris bifida.” Hovey (30), Sapper (56), and Robson and Willmore (51) give data on the water temperatures of the Boiling Lake.

Earle (14) reported additional soufriéres in Dominica at Ravine D’Or (Layou Valley), Morne au Diable (Portsmouth), Laudat and Soufriére villages, and Wotten Waven. He states, “All of these are insignificant besides the Grande Soufriére, though the Wotten Waven springs are quite numerous and besides a few violent fumaroles also display small mud volcanoes.’

At Wotten Waven I was able to visit the fumarole area which is located in an old field now devoted largely to pasture. The fumaroles at the time of my visit were noisy and emitting steam, but almost without sulphurous odors. The high humidity and heat appeared to encourage plant growth, and the shrubs of Clusia alba were more luxuriant than specimens we had passed a few yards away (Fig. 7.)

The fumarole area behind Soufriére village was represented by an enor- mous scar on the hillside where landslides had recently exposed fresh sur- faces and obscured some of the vents. The fumaroles were not noticeably active, and although the odors of sulphur dominated the air, there was no evident destruction of the vegetation. The area was well stabilized, but few plants other than species of Blechnum, Dicranopteris, and Cyperus had invaded the areas.

Guadeloupe

THERMAL SPRINGS. The island of Guadeloupe is reported to have ther- mal springs along the coast at the towns of Boiuillante, Monchy, and Pigeon; somewhat inland near La Lise and Dolé; on the slopes of the Soufriére near Baines Jaunes; and at the summit of the Soufriére. All of these are on Basse Terre and probably are associated with the volcanic massif of the Soufriére. I have not visited the thermal springs at low altitude and have been unable to find any descriptive literature regarding them or the vegetation surrounding them.

The oldest reference to hot springs and fumaroles is found in Pére Labat’s writings. His description of the thermal springs of “Ance de Goyave”’ presumably refer to the area currently called Boiillante and not to Goyave on the west coast. Labat also climbed the Soufriére and de- scribed the stunted vegetation near the summit. A trip from St. Claude

1962] HOWARD, VOLCANISM AND VEGETATION 301

to Baines Jaunes and along the old carriage trails past the warm and cold baths to the summit of the Soufriére er that few changes have occurred in the intervening 260 years (Figs. 8, 9).

Stehlé (65, 67, 68) has published several ‘admirable treatments of the vegetation of the slope of the Soufriére. He indicates that the effects of high altitude and of the sulphur vapors cause a limitation in height of plants but seem to encourage the development of many lateral branches. He also states, “Les gaz sulfureux émis par la Soufriére sont néfastes a la végétation environnante, qu’ils détruisent peu a peu.” Stehlé concludes that the species which exist in this environment are particularly interest- ing and rare. Many are thought to be endemics, but as more is known of the vegetation of the high mountain peaks from St. Kitts to St. Vincent it becomes increasingly difficult to separate those species which are re- stricted to high elevations on the ash of volcanic ejecta and those which are there because of their tolerance to fumarole emissions. However, the red-flowered form of the bromeliad Pitcairnia spicata occurs at high alti- tudes on St. Kitts, Guadeloupe, Dominica, Martinique, and St. Vincent, while the yellow-flowered form, P. spicata var. sulphurea, has not been found on St. Kitts and occurs on the other islands in areas influenced by the crateral fumaroles. The endemic species of these high mountains are not the most distinct of species, and recent collections from other islands have indicated that their distribution is wider in the Lesser An- tilles than has been previously reported.

Montserrat

GALWAY’s SOUFRIERE. (Fig. 15.) One of the earliest accounts of the sou- friéres on Montserrat was made by Dr. Nicholas Nugent, an honorary mem- ber of the Geological Society of London who visited Montserrat in 1810. His point of view in visiting Galway’s Soufriére was entirely geological, but his description applies to the present as well. “A rugged horse-path was traced along the brink of the ravine, which we followed amidst the most beautiful and romantic scenery. At the head of this ravine is a small amphitheatre formed by lofty surrounding mountains, and here is situated what is termed ‘The Sulphur.’ Though the scene was extremely grand and well worthy of observation, yet I confess I could not help feeling a good deal disappointed, as there was nothing like a crater to be seen, or anything else that could lead me to suppose the place had any connexion with a volcano. On the north, east and west sides were lofty mountains wooded to the tops, composed apparently of the same kind of porphyry we had noticed all along the way. On the south, the same kind of rock of no great height, quite bare of vegetation, and in a very peculiar state of decomposition. And on the south-eastern side, our path, and the outlet into the ravine. The whole area thus included, might be three or four hundred yards in length and half that distance in breadth.”

Karl Sapper visited Montserrat in 1903 (57) and visited both Gages and Galway’s soufriéres. He published a sketch map of Galway’s Sou-

302 JOURNAL OF THE ARNOLD ARBORETUM [| VOL, XLIII

friere indicating the principal fumaroles where he recorded temperatures of 34.2-93.2° C. and commented on the production of hydrogen sulfide y each. Sapper estimated the area of destruction of vegetation as between three and four hectares. MacGregor (39) reproduced Sapper’s sketch- map and concluded that the site of fumarole activity had changed by the time of his visit in 1936. Perret also visited Galway’s Soufriére during the volcanic crisis in Montserrat in 1934. His description of the area is meager, for he concluded that, ‘“Galways Soufriére was a more ordinary solfatara having less chemical activity and water of less acidity — alto- gether, a vent less capable of serving as an indicator of increasing vol- canic activity than the more accessible Gages Soufriére.” He did note that the water from Gages Soufriére had an acid content ten times as great as that from Galway’s. My tests in 1961 with “pHydrion” paper gave a reading of pH 5 for the waters in Galway’s fumaroles and 1.5 for those in Gages.

Sapper commented only on the dead and dying vegetation in the vicinity of Galway’s Soufriére, and his outline map of the area of activity and destruction has recently been questioned by Martin-Kaye (40), who felt from seeing MacGregor’s reproduction of Sapper’s map that the main stream was in reality the southeast stream as drawn by Martin- Kaye (40). I find it difficult to agree with this conclusion and suggest that Sapper and MacGregor are correct. Certainly the locations of the principal vent are similar on the two maps in relation to the area of destroyed vegetation.

Martin-Kaye has reproduced temperature records at Galway’s mae friere taken by the Montserrat Department of Agriculture from 1936 t 1952. My own reading taken in January, 1961, showed a vent pe ture of 96° C. (200° F.) and pool temperatures 83° and 77° C. (180° and 170° F.), respectively, for the pools where, I was told, the tempera- tures were generally recorded.

The vegetation at Galway’s appears to be completely stabilized. The prevailing winds blowing down the mountain gorges across the soufriére and towards the sea apparently have carried the toxic fumes for years at definite elevations. Plants have developed in the lee of fans, piles of scree, and even behind boulders. To the windward, the vegetation is unaffected, and not at all stunted. The principal plants nearest the soufriere were Clusia, Freziera and Ternstroemia. Within the soufriére proper the ex- isting and often flourishing plants belong to the genera Andropogon, Clusia, Cyperus, Philodendron, Pitcairnia, Pityrogramma, and Ternstroemia.

GacES LOWER SOUFRIERE. (Fig. 14.) Nugent does not mention Gages soufrieres and Shafer’s (61) description of his visit to Montserrat speaks so generally of the three soufriéres that one can not place his actual visit. Gages Lower Soufriére is said to have originated in 1896 following a cloudburst and to have been moderately active during an earthquake series

98. Perret’s visit to and study of Gages Lower Soufriére during its period of renewed activity beginning in 1933 is well recorded. Gages Lower Soufriere consists of two parallel valleys, one old and inactive,

—

1962 | HOWARD, VOLCANISM AND VEGETATION 303

the other a “new rift” which opened around 1930. The one valley re- mained active through 1935, but by the 6th of May, 1937, the gas phase had subsided and the water from the gorge was nearly chemically inert. I visited this area in 1950 and again in 1961. Little change was observed in the first visit from the condition shown in Perret’s photographs. In 1961, only minor fume emissions were noticeable with faint odors of hydrogen sulfide with sufficient heat and steam to cloud one’s glasses. Fume temperatures at the most active vents were 94-96° C. The area of devastation, which Perret judged to be 600 feet long and 350 feet wide, had not increased, but by 1961 an occasional patch of green from pro- thallia or young gametophy tes of mosses was evident, with an occasional plant of Andropogon, Capraria biflora, Clusia, or Cyperus present in the valley. On the limonitized zone at the head of the valley grasses and sedges were becoming more abundant, apparently developing from seed sources above the devastated area.

Gaces Upper SourrikreE. (Figs, 10-13, 18, 19.) Accurate dating is not available for the age of Gages Upper Soufriére. Sapper (57) visited it in 1903, and Perret indicates that it was dormant prior to the earth- quake of November 11, 1935, when its activity increased enormously. Perret does not go into detail on his observations on Gages Upper Sou- friere, although he delayed his departure from Montserrat to study this new outbreak. He does indicate that the temperatures were 99° C, i the steam vent and 96° and 98° C. in the pools. His illustrations, and par- ticularly Fig. 31 reproduced here, are noteworthy for the remnants of the vegetation shown. Clearly, the area of Gages Upper Soufriére had been relatively small prior to the 1935 activity, for the density of the vegetation indicates a recent kill. Perret noted that there was a threat for Gages Upper Soufriére to “extend its domain by reason of gas emission and rain erosion.” Martin-Kaye’s observations and map for his visit in October, 1954, led to the conclusion that ‘“‘much of this area has comparatively re- cently opened up as many rotted tree stumps are still to be seen more or less in the position of growth.” My own comparative observations al- ready mentioned were made in 1950 and again in 1961 and lead only to the conclusion that Gages Upper Soufriére is not only as active as it has been in the past but is continuing to increase to the southwest and may well cross the ridge into the next valley. The area devastated at present is about eight acres at an altitude between 1300 and 1400 ft. along the stream bed, but extending up the hill to around 1700 ft.

Tar River SourrizRe. MacGregor (39) has illustrated in his Figure 37 the appearance of the “New Cow Hill soufriére” (Tar River) shortly after its inception in 1936. The fumaroles apparently appeared in the center of a banana plantation. When Martin-Kaye and Willmore visited the area the fumarole was relatively quiet, and only about an acre of vegetation was affected by the fumes. In 1961 steam and fumes were not apparent from a convenient observation point on the slopes leading to English’s Mountain

LANG’ s Soursrtre. (Figs. 16, 17.) With the expert guidance of Kings-

304 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIII

ley Howes we visited the recently discovered Lang’s Soufriére on the north flank of the Soufriére Hills above Paradise Estate. The soufriére itself is essentially dormant, giving only occasional indication of hydrogen sulfide and some degree of heat. It is located at the head of a deep ravine, and landslides not of recent time have covered the actual fumaroles. The decomposition of the rock in the area, plus the occurrence of one large dead specimen of Sloanea and the abundance of Philodendron giganteum attested to the past activity of this fumarole.

OTHER SOUFRIERES. Through the Soufriére Hills are other indications of dead or dormant fumarole areas. Several were seen in Gages Gut en route to Gages Upper Soufriére. Neither fumes nor heat is emitted from these areas at the present. The older ones were green with mosses and lichens, and the younger ones still showed the color of the altered rocks. These fumaroles were apparently never of major importance, for the sur- rounding vegetation was neither affected nor characterized by the genera usually found in active areas. Mulcair Soufriére was known by reputa- tion to the local people, but, as Martin-Kaye discovered, it seems to be inaccessible, for no local individual was capable of guiding me to it.

Hot WATER Ponp. This coastal pond is thought to be heated by sub- terranean springs, possibly the drainage from the Gages soufriéres. It did not support any unusual plant life nor was the surrounding vegetation affected in any way.

St. Kitts

Martin-Kaye (40) supplies the only references I have encountered to fumarole activity on the island of St. Kitts. My visits have been at times when the cloud cover made any ascent of Mt. Misery of doubtful value. Martin-Kaye reports however, that the crater of Mt. Misery is between one-half and three-fourths of a mile in diameter at the rim, 500 to 60 feet deep, and possesses a small shallow lake in the southwest part of the floor, which “except in the vicinity of two small areas of fumarole activity, is otherwise covered with vegetation.” Martin-Kaye also reports that “now defunct fumarolic activity has caused rock alteration at Rainbow Cliff, Frigate Bay where variously colored ochres are associated with yellowish, brownish and grey clays, kaolin, some gypsum and probably opal powders and alunite as are associated with the active soufriéres in Montserrat, Nevis, Mt. Misery itself and elsewhere in the Lesser Antilles.’’ Regrettably no data on the surrounding vegetation are available.

Nevis

The Rev. Mr. Smith (64) in a series of letters presents some of the earliest records of the natural history of Nevis. In his second letter he refers to a hot river and to baths “supposed to flow from the sulphur ground which is not above three quarters of a mile higher in the country.” He also notes that “In my parish of St. John in the Island of Nevis, there is

1962] HOWARD, VOLCANISM AND VEGETATION 305

a considerable spot of sulphurous ground on the south side, at the upper end of a deep rupture in the earth vulgarly called Sulphur Gut, which is so excessive hot . . . as to make us immediately feel it through our Shoe Rev. Smith refers to another at a “place two miles and _ half south-ward of Charles Town” and reports that ‘a new hot spring was in 1718, discovered in Windward Parish, upon clearing of a wood . . . just above Camp-ground.”

Bryan Edwards (15) said for Nevis, “That the island was produced by some volcanic explosion, in ages long past, there can be no doubt; for there is a hollow, or crater, near the summit, still visible, which con- tains a hot spring strongly impregnated with sulphur; and sulphur is frequently found in substances, in the neighboring gullies and cavities of the earth.” In his excellent treatment of the geology of the Leeward Islands, Martin-Kaye has reported two active soufriéres on Nevis at Farm’s Estate and Cades Bay, the latter first noticed in 1953. In addi- tion there is some gas seepage noticeable in the Belmont area and thermal pools at Bath Springs near Charlestown and in the Camp Springs area. The thermal pools apparently are those described by the Rev. Smith, in- dicating an activity of nearly 250 years.

The Farm’s Estate Soufriére is both old and relatively quiet. Hydrogen sulphide odors and wisps of steam are noticeable. Willmore recorded tem- peratures of 42-44° C. in small fissures and temperatures of 96° C. in a vent. Apparently the vegetation has been affected by the activity of the fumaroles but has now become stabilized.

The Cades Bay Soufriére, according to Martin-Kaye, was first noted in 1953 when the “vegetation began to die and hydrogen sulfide was evident.” In six months an area twenty by thirty feet containing sunken boiling steam holes had developed and odors of sulphur dioxide were evi- dent. Shortly thereafter several areas became confluent by the loss of vegetation and a barren area of about 1.5 acres was produced containing small pools with water registering a temperature of 118° F. Martin-Kaye reports that in 1958 the vents “seem to have mainly concentrated in one small bubbling pool a few feet across’ and that the ‘‘vegetation had re- established itself over the western part.’’ Regrettably no specific informa- tion on the nature of the vegetation of the area is available.

SUMMATION AND COMPARISONS

Volcanism, either eruptive or fumarolic, is destructive to vegetation, as ample examples show in the Lesser Antilles. Nuées ardentes, character- istic of recent eruptions, are totally destructive of plant life, and ash falls less so. The revegetation of ash-covered areas is related to the type of destruction but also to the altitude of the area considered and its prox- imity to sources of new plants. At lower elevations or adjacent to un- disturbed areas invasion by weedy species is rapid. Plants covered only with free-fall ash appear to sprout rapidly from the basal portions and quickly supply a covering vegetation. At highest elevations where a

306 JOURNAL OF THE ARNOLD ARBORETUM [ VOL, XLII

specialized vegetation occurs the re-establishment of vegetation is much slower. Even when the proper seeds are introduced from adjacent moun- tain tops the environment is inhospitable, and decades pass before the original introductions have formed an even moderately complete cover.

Yew areas of fumarolic activity appear to act first through a heating effect of the soil to which the native species are selectively tolerant. The emission of sulphur compounds creates an acid environment in the soil and air-borne acidic fumes. Natives species appear to be killed off in direct relationship to their ability to stand increasing acidity. The air- borne acid particles appear to be cumulative in their action on the sur- rounding vegetation. Ferns with leaves of delicate textures and broad- leaved plants without heavy cuticles are most sensitive to these fumes. Plants with coriaceous leaves or heavy cuticles are most resistant. At the highest elevations on soils composed of volcanic ash it is difficult to de- termine the role of fumarole emissions in the establishment and _per- sistence of the species present.

In the Lesser Antilles the species which appear to be most tolerant of fumarolic gas emissions and soil acidity are the following: Andropogon pertusus, Blechnum indicum, Cladium restioides, Clusia alba, Clusia mangle, Clusia plukenettu, Cyperus ligularis, Cyperus planifolius, Dicra- nopteris bifida, Guzmania plumieri, Lycopodium cernuum, Lycopodium meridionale, Lycopodium tortum, Philodendron giganteum, Pitcairnia an- gustifolia, Pitcairnia albucifolia, Pitcairnia spicata var. sulphurea, Pityro- ramma calomelanos, Pityrogramma chrysophylla, and Ternstroemia pe- duncularis.

A comparison of the plant species from other geographic areas which are tolerant of fumarolic environments or which are invaders of volcani- cally disturbed areas is interesting yet difficult due to the differences in composition of the floras. Much has been written of the thermal springs of the United States, particularly the area of Yellowstone National Park where emphasis has been given to heat tolerant algae and to the animal life of the hot springs. Similar studies are available for the hot-spring areas of Iceland. In both cases, the families, genera, and species of the plants can not be compared with those of the Lesser Antilles. The ex- tensive fumarole areas of Indonesia and of Japan have received some study. Mizushima (41) studied the plant community of the fumarole areas of the Idzu Islands, midway between Tokyo and the Bonin Islands. He found that the plants tolerant of fumarole conditions were: Chamae- syce hirta, Chrysanthemum pacificum, Cyperus polystachyos, Fimbri- stylis dichotoma {. floribunda, Imperata cylindrica var. koenigii, Linder- nia crustacea, Lycopodium cernum, Lygodium japonicum, Ophioglossum petiolatum, Paspalum orbiculare, and Pouzolzia zeylanica (Urticaceae). He noted that plants of Lindernia crustacea are of interest as the indi- viduals gain in stature as they approach the fumaroles. Hedyotis, Pas- palum and Pouzolzia were limited in their distribution on the island to the areas of the fumaroles. Cyperus, Fimbristylis, Imperata, Lycopodium, and Lygodium were indicated as abundant near the fumaroles.

1962] HOWARD, VOLCANISM AND VEGETATION 307

The locations of the numerous areas of volcanism in Indonesia are effectively shown in a map published by Ter Braake (9). Braun-Blanquet (10) indicates that Pteris incisa thrives on the soft, sulphurous soils ad- jacent to springs that are rich in alum. Polypodium vulcanicum has been reported to survive on rocks which are occasionally flooded by waters with temperatures as high as 75° C, According to Holtermann (23), the dominant shrubs in such areas are Agapetes vulgaris and Rhododendron retusum.,

Braun-Blanquet suggests that the plants of such fumarole areas are mesophytic in nature. Schimper (59) however, indicates that the flora of the solfatara areas is composed of xerophilous species. Zollinger (72) and later Junghuhn- (34) were the first to observe in Java that the vegeta- tion close to fumaroles was composed of alpine species, even when the fumarole was 1000 to 1500 meters below the natural alpine region. Schim- per described the fumarole area as follows: ‘Where these pools are col- lected, usually in large numbers and of different sizes, the soil is a wet white clay, which is said by Junghuhn to arise by the action of sulphuric acid on trachyte; it is usually covered by a yellowish efflorescence of sulphur. The ground is frequently so hot that to remain standing on it is impossible. From all the crevices and pools there escape hot vapors of suffocating odour, sometimes of sulphuretted hydrogen, at other times of sulphurous acid. The water has an acid taste and sets the teeth on edge.” He noted further that “the bushes of the solfataras are much lower in stature than the surrounding forest and quite sharply marked off from it. None of the small trees and shrubs that form the underwood in the high-forest appear among them; of forest herbs only a few species repre- sented by detached individuals occur; plants that pecan elsewhere in the open situations of the same region are entirely abse

In the vicinity of solfataras Schimper indicates hee are es: numer- ous plants of Ficus diversifolia, Rhododendron javanicum, and Vaccinum varingiacfolium. In solfataras of lower regions some other species, notably Gaultheria leucocarpa, Pree aes javanensis, Myrsine avenis, Rhododen- dron retusum, and Rhododendron tubiflorum, occur, and, in addition,

“isolated ferns with leathery leaves and lycopods are always present in such spots.” Plants of the Melastomataceae and Myrsinaceae, although abundant in the Lesser Antilles, do not tolerate fumarolic environments. The Ericaceae are poorly represented in the Lesser Antilles, and the genera do not occur in areas where fumaroles are active.

Junghuhn noted that in solfatara areas the shrubs are swept clear of all epiphytic vegetation, including mosses and lichens, This is also true throughout the Lesser Antilles. A new fumarole seems to be most toxic to the epiphytic moss, liverwort, and lichen flora of trunks and branches of the adjacent vegetation. In dormant fumarole areas, however, mosses are among the first invaders.

The steam vents on Mauna Loa, Hawaii, contain minimal amounts of sulphur and do not seem to be influential in determining the vegetational components around them. Ferns, particularly Nephrolepis, are abundant,

308 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIII

while Dodonea, Styphelia, Vaccinium, and many rubiaceous genera grow more vigorously in the humid environment of the vent than at a distance.

The revegetation of surfaces disturbed by volcanic eruptions has been the subject of a classic study in the case of Krakatoa (69). Fosberg (16) has outlined the nature of the invasion and revegetation of lava flows and ash-cinder areas of Hawaii. Only preliminary studies are available for the Paricutin eruptions in Mexico. Calvert and Calvert (12) discuss the effects of volcanic activity and subsequent fumarolic activity for Volcan Irazu, Costa Rica. However, the altitude and the species involved in all these areas do not permit comparisons with similar areas of the Lesser Antilles.

BIBLIOGRAPHY

1. ANpeRson, J. An account of Morne Garou, a mountain in the island of ot, Vincent, with a description of the volcano on its summit. Phil. Trans. Roy. Soc. London 75: 16.

2. ANDERSON, T. Report on the eruptions of the Soufriére in St. Vincent in 1902, and on a visit to Montagne Pelée in Martinique. Part II. Phil. Trans. Roy. Soc. London A. 208: 275-304. 1908.

3. ANDERSON, T. & FLeETT, J. S. Preliminary report on the recent eruption of the Soufriére in St. Vincent, and of a visit to es Pelée, in Mar- tinique. Proc. Roy. Soc. London 70: 423-445.

4, : Report on the eruptions of the Soufriére, in - Vincent in 1902,

and on a visit to Montagne Pelée, in Martinique. Part I. Phil. Trans. Roy. Soc. London A. 200: 353-553. 1903. 5. Atwoop, T. The history of the island of Dominica. London

6. BEARD, 7 S. The progress of plant succession on the i ae hes St. Vin- cent. Jour. Ecol. 33: 1-9. 5,

i . The natural vegetation of the Windward and Leeward Islands. Oxford, 1949.

8. Bopvarsson, G. Production of thermal energy on the Island of St. Lucia. United Nations, Technical Assistance Administration. New York, 1951.

9. Braake, A. L. TER. Volcanology in the Netherlands Indies. /n: Honic

& VERDOORN, Science and scientists in the Netherlands Indies. New York, 1945.

10. BrauN-BLANQUET, J. Plant Sociology. McGraw Hill, 1932.

11. Burtrertin, J. La constitution geologique et la structure des Antilles. Gap, France, 1956.

12. CALVERT, A. S., & CaAtvert, P. P. A year of Costa Rican natural history. Macmillan, 1917.

13. Davis, W. M. The Lesser Antilles. Am. Geogr. Soc. Publ. 2. 1926.

13a. Domin, K. The Pteridophyta of the island of Dominica. Mem. Roy. Czech. Soc. Sci. Div. Nat. Hist. Math. ser. 2, 2: 1-259.

14. Earte, K. W. Geological notes on the island of Dominica, BW. Geol. Mag. 65: 169-187. 1928.

15. Epwarps, B. History, ee and commercial, of the British Colonies in the West Indies. Vol. 1.

16. FosBerc, F. R. ae excursion III, Tenth Pacific Science Congress. Honolulu, 1961.

1962 | HOWARD, VOLCANISM AND VEGETATION 309

17.

Frotow, S. Les ee de la Montagne Pelée. Revue Martinique 1: 39-54. 1944. [Not se

GovAULT, J. orn ee sur la Soufriére de la Guadeloupe pour servir a l’étude des appareils éruptifs Peléens. Pointe-a-Pitre, Guadeloupe. 1942. [ Not seen. |

Harpy, F. & Ropricues, G., with Nanton, W. R. E. The agricultural soils of Montserrat. Trinidad, 1949.

Hirt, R. T. Pelé and the evolution of the Windward Archipelago. Bull. Geol. Soc. Am. 16: 243-288. 1905

. Hopce, W. H. The vegetation of Dominica. Geogr. Review 33: 349-375.

1943. . Flora of Dominica, B.W.I. Lloydia 17: 1-238. 1954

. HOLTERMANN, C. Der Einfluss des Klimas auf den Bau der Pflanzengewege.

Leipzig. 1907: Hooper, E. D. M. Report upon the forests of St. Vincent. London, 1886. . Report upon the forests of St. Lucia. Madras, 1887.

Hovey, E. O. Some erosion phenomena observed on the islands of Saint

Vincent and Martinique in 1902 and 1903. Bull. Geol. Soc. Am. 15: 560-1. 1904 _ Grinde Soufriére of Guadeloupe. /bid. 15: 561-2. 1904. . Bibliography of literature of the West Indian eruptions published in the United States. /bid. 15: 562-566. 1904. . The present condition of Mount Pelé. /bid. 16: 566-9. 1905. . Boiling lake of Dominica. /bid. 16: 570-571. 1905. —. Soufriére of Saint Lucia. /bid. 18: 569-570. 1905. Huccins, P. F. Report on the Soufriére, Saint Vincent, B.W.I. St. Vincent, 1902.

. JoHow, F. Vegetationsbilder aus West-Indien und Venezuela. Kosmos

Stuttgart 2: 270-285. 1884.

. JUNGHUHN, F. Java, deszelfs gedaante, bekleeding en invendige structuur.

German ed. (ed. Hasskarl) 1: 453.

. LaspaTt, PERE. Nouveau voyage aux Isles ‘de Amerique. Vol. 2. 1772. . Lacrorx, M. A. L’Eruption de la Martinique. 44 pp. Privately printed,

Paris, 1903.

. Lee, W. Daniel Defoe: his life and recently discovered miscellaneous

writing. Vol. 2. 1869.

MacGrecor, A. G. Royal Society Expedition to Montserrat, B.W.I.: Pre- liminary report on the geology of Montserrat. Proc. Roy. Soc. London B. 121: 232-252. 1936.

. The volcanic history and petrology of Montserrat, with observa- tions on Mt. Pelé in Martinique. Phil. Trans. Roy. Soc. London B. 229: 1-90. 1938.

Martin-Kave, P. H. A. Reports on the geology of the Leeward and British Virgin Islands. 1-117. St. Lucia, 1959.

. Mizusui1ma, M. A preliminary report on the florula of Isl. Aogashima,

Idzu Islands. Misc. Rep. Res. Inst. Nat. Resources 38: 102-126. 1955.

. Nucent, N. An account of “The Sulphur” or “Soufriere’” on the island

of Montserrat. Trans. Geol. Soc. London 1: 185-190.

: See F. A. Camps in the Caribbees. Boston, 1880. . PERR F. A. The eruption of Mt. Pelée, 1929-1932. Carnegie Inst.

ae Publ. 458. 1935.

on Ww

nn

JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIII

. The volcano-seismic crisis at Montserrat, 1933-1937. Carn. Inst. Wash. Publ. 512. 1 Volcanological observations. Carnegie Inst. Wash. Publ. 549.

1950.

Poucn, F. H. Paricutin comes.of age. Nat. History 53: 342-348. 1944.

REED, F. R.C. The geology of the British Empire. Ed. 2. London, 1949.

Rogpson, G. R. & Wit_more, P. L. Geological and eps eee on the Qualibou Soufrieére of St. Lucia. 1-6. Trinidad,

. Some heat measurements in West Indian ee “Bull. Vol- can. ser. 2. 17: 13-39.

Romer, M. [Title unknown.| Bull. Serv. Meterol. et de ?Observ. Geophys. 1953. | Not seen. |

Russet, I. C. Volcanic En uOns on ec and St. Vincent. Ann. Rep. Smithson. Inst. 331-34 903.

Sanps, W. M.. An account e the return Fe vegetation and the revival of agriculture, in the area devastated by the Soufriere of St. Vincent in 1902-03. West Indian Bull. 12: 22-33. 1912.

Sapper, K. Besuch den Insel Grenada. Centralbl. Min. Geol. Palaeont. 1903: 182-186. 1903.

Bericht iiber einen Besuch von St. Vincent. /bid. 248-251. 1903.

Zur Kenntniss der Insel S. Lucia in Westindien. Jbid, 273-278.

—

1903.

. Ein Besuch der Insel Montserrat. /bid. 279-283. 1903.

Scuwabe, G. H. Beitrage zur kenntnis Islandischer Thermalbiotope. Archiv Hydrobiol, puDpL, Ser, 6: 161-352... TO3D.

Scuimper, A. F. W. Plant-geography, upon a physiological basis. English

Pacing by W. R. Fisuer. Revised and edited by P. Groom & I. B.

our. Oxford. 1903.

ees C. Historical geology of the Antillean-Caribbean Region. New York, 1935

SHAFER, Wi Report on a visit to the Island of Montserrat. Jour. N.Y. Bot. Gard. 8: 81-88. 1907.

SHEPHARD, Mr. Historical account of the Island of St. Vincent. London, 1831

SmitH, LONGFIELD. Volcanic eruptions in the West Indies. West Indian Bull. 3: 271-293. 1902

SmirH, Rev. Mr. W. A natural history of Nevis. Cambridge, 1745.

STEHLE, H. Flore de la Guadeloupe et Dépendances. Essai d’écologie et de eae botanique. Vol. 1. Basse Terre, Guadeloupe, 1935.

ae associations végétales de la Martinique. Bull. Agr.

19

Martniga 6: etre

; rest” aie of she) Carib bean Islands. Carib. For. 6(Suppl.): 273- — ecrc s types forestiers des Iles Caraibes. /bid. 7: 337-709. 1946.

VAN pone M. N. Catalogues of the active volcanoes of Indonesia.

(Part I in Catalogue of the active volcanoes of the world.) Naples, 1951 — I. Herbaceous angiosperms of the Lesser Antilles. Puerto Rico,

195

Wittmore, P. L. The earthquake series in St. Kitts-Nevis, 1950-51. Nature 69: 770-772. 1952.

ZOLLINGER, H. Ueber Pflanzenphysiognomie im Allgemeinen und diejenige der Insel Java insbesondere. Zurich, 1

1962] HOWARD, VOLCANISM AND VEGETATION Sl

EXPLANATION OF PLATES PLATE I

Fic. 4. Cone of fumarole at Lavera rae Springs, Grenada. Fic. 5. Path of mineral deposit from cone in Fig. 4, nada. Fic. 6. Valley of Desolation, Dominica. Fic. 7. Steam vent at Tee en Dominica. Fic. 8. eClaseran guadelupense” on windward slopes of La Soufriére, Guadeloupe. Dicranopteris bifida and Cladium restioides are seen in foreground, Clusia mangle in middle, and /lex species on distant slope. Fic. 9. The same on leeward slopes, showing killing by periodic emissions of sulfur fumes. Dead plants are Clusia, Ilex, and Freziera; living plants are Philodendron giganteum.

PLATE II

Fic. 10. Gages Upper Soufriére, Montserrat, after its reactivation on No- vember 11, 1935. Photo reproduced from Perret (45). Fic. 11. Gages Upper Soufriére, ‘Montserrat, as it appeared in June. 1950. Fic. 12. Gages Upper Soufriére, Montserrat, as it appeared in January, 1961. Fic. 13. Dead and acid-rotted stump of Sloanea dentata at Gages Upper Soufriére in January, 1961. A fumarole had opened under this tree within the last decade, for the tree was alive in 1950. Fic. 14. Gages Lower Soufriére, Montserrat, in January, 1961. Vegetation has descended on the limonitized area from above. Fic. 15. Main steam vent at Galway’s Soufriere, Montserrat, in January, ge Effect of the prevailing winds from the mountain is shown by the path of the steam. Effects of the fumes are visible on the hillsides by the reduction of vegetation.

PLATE III

Fic. 16. Lang’s Soufriére, Montserrat, as seen from Paradise Ridge. Point of arrow locates the dead Sloanea trunk shown in Fic. 17. Rear of the arrow marks dense growth of Philodendron giganteum. Fic. 17. Point of emission of fumes from Lang’s Soufriére, Montserrat. A dead trunk of Sloanea is on the left. Recent rock fall shows the depth of the effects of the acidity on the rock face. Fic. 18. A new area of fumarole activity at Gages Upper Soufriére, Montserrat, January, 1961. Plants of Clusia alba are in the foreground. Cyathea and other components of adjacent woodland are shown in the background, as yet, not affected by fumes of the new fumarole. Destructive effects to the pres- ent are due to heat and soil acidity. Fic. 19. Gages Upper Soufriére, Montserrat, January, 1961. The upper valley of this soufriére showing the proximity sat the undisturbed woodland on the windward side of the fume area.

Qualibou Soufriére, St. Lucia. Boiling lakes and steam vents with Saas rock are in the foreground. The profile of the hill in the background ee the absence of shrubs or trees. Fic. 21. Qualibou Soufriere, St. Lucia. A fac view of the hill shown in Fig. 20. The hillside vegetation is of Blechnum a Lycopodium where affected by the fumes of the soufriére. The edge of the affected area is shown on the right,

Jour. Arnotp Ars. VoL. XLII PLATE I

Howarp, VOLCANISM IN THE LESSER ANTILLES

Jour. ARNOLD Ars, VoL. XLIII PriatTE II

eh

Howarpb, VOLCANISM IN THE LESSER ANTILLES

Jour. ARNOLD Ars, VoL, XLIII PiatTE III]

Howarb, VOLCANISM IN THE LESSER ANTILLES

1962 | ERNST, PAPAVERACEAE & FUMARIACEAE 315

THE GENERA OF PAPAVERACEAE AND FUMARIACEAE IN THE SOUTHEASTERN UNITED STATES !

WaLLace R, ERNST

PAPAVERACEAE A. L. de Jussieu, Gen. Pl. 235. 1789, nom. cons. (Poppy FAMILY)

Annual or perennial, sometimes rhizomatous herbs |shrubs, or small trees|, often glaucous and pubescent with multicellular [or unicellular | hairs or glabrous, usually with acrid, reddish-orange to yellow |or white | sap. Leaves rosulate and/or cauline, usually alternate, reduced upward, exstipulate, petiolate or sessile, pinnately (seld om palmately) [or tern- ately | veined, lobed, or dissected |or entire]. Inflorescences terminal and determinate, 1-many -flowered, usually cymose [or rarely paniculi- form|; buds erect [or nodding], subtended by a leaf or bract. Flowers erect, bisexual, open, essentially regular and hypogynous [seldom perig- ynous ]. Perianth cyclic, imbricate, usually 3-seriate, 2- or 3-merous. Sepals 2 or 3 [seldom synsepa lous}, usually with dissimilar left-right- oriented margins [seldom reversed], tightly inclosing the buds until anthesis, then caducous. Petals usually 2-seriate and twice as many as sepals (sometimes polypetalous) |or seldom apetalous], opening diurnally,

"Prepared for a generic flora of the southeastern United States, a joint project of the Gray Herbarium and the Arnold Arboretum made possible through the support of George R. Cooley and the National Science Foundation and under the direction of Reed C. Rollins an Carroll E. Wood, Jr. This et follows the style established in the first paper the series, Jour. Arnold Arb. 39: 296-346. 1958 (and continued through volume a The area covered in this, as in eeyaiae treatments, is bounded by and includes North Carolina, Tennessee, Arkansas, and Louisiana. The deseriptions apply primarily to the plants of this area with any supplementary information in brackets. References which have not been verified by the eee a marked by an asterisk. The il Ee were prepared by Mrs. Dorothy H. Marsh, the manuscript by Mrs. Gordon W.

This ee ie is ine in part, on a survey of the floral morphology of Fuma- riaceae and a doctoral dissertation on the floral morphology of Papaveraceae. The subfamilies described here, discussed in detail in the above dissertation, represent a re-evaluation of the generic affinities. Altogether, these four subfamilies constitute the subfamily Papaveroideae, as used by Fedde, their names and included genera more or less corresponding to of his five tribes. In the terminology of Fedde, the Romneyeae Benth. & Hook. (including Rone eva Harv. and Arctomecon Torr. Frém.) are placed with the Papavereae, and Glauciunt Mill. has been removed from the ee ereae and placed with the Chel ee Reichen

e Russian literature citations were translated by Dr. ie K. Brizicky; the elo- Bene cae descriptions were prepared by Dr. R. C. Foster; Dr. G. W. Gillett spe Mr. R. C. Bean provided material for cytological study; rs. H. Francia Hom mersand verified some distribution records in the herbarium of the University of North Carolina; and Dr. C, E. Wood, Jr., has given much expert advice.

worn

316 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLII

sometimes wrinkled, usually fugacious; the outer series larger and alter- nate with the sepals. Stamens usually many, indefinite, varying in num- ber with the size of the flower, usually shorter toward the petals; the filaments sometimes expanded lor toothed| and often constricted below the anther: anthers basifixed, 2-locular at anthesis, usually extrorse, dehiscent by longitudinal slits; pollen often 3- colpate, sometimes poly- porate |polycolpate or rugate]. Gynoecium syncarpous, of 2 to many carpels; stigmas laterally confluent |or discrete|, as many as and alternate with the placentae, the margins of adjacent stigmas often conspicuous over the placentae [see Eschscholzia|; style definite to absent; ovary 1-locular [locule rarely occluded; or rarely multilocular; or rarely the carpels conduplicate about a central cavity]; ovules usually numerous [seldom 1], obovoid, amphitropous, 2- integumented, the micropyle be- neath the funiculus; placentation parietal, sometimes intruded into the locule. Fruit dehiscent by 2 to many valves alternate with the persistent placentae [or fruit nonvalvate and dehiscent through the placentae|; de- hiscence usually incomplete, acropetal or basipetal. Seeds [1 to] many, obovoid to -_ reniform, free [or rarely fruit fragmenting into 1-seeded joints|, variously sculptured or smooth, sometimes arillate and attractive to ants; endosperm present, often oily; embryo small, straight or some- what curved. Seedling cotyledons linear | very rarely bifid] or expanded ace petiolate. (Excluding Fumariaceae DC. and Hypecoum L.) TYPE Nus: Papaver L.

About 25 genera and at least 200 species, widely distributed in the Northern Hemisphere from sea level to ca. 5800 m. (19,000 t.)* Papaver, local in both South Africa and Australia, extends northward into the Arctic; Bocconia and Argemone occur in the New World tropics. Of the 13 genera indigenous to the United States, only three occur in our area.

he genera may be divided conveniently into four subfamilies. The Chelidonioideae Ernst, the most heterogeneous of the subfamilies, tend- ing to have ~aulecelluiay uniseriate hairs, 2-merous perianths, 3-colpate or polyporate pollen, 2-valved gynoecia (3-S-valved in Stylophorum diphyllum) with well-developed median carpellary traces, and arillate seeds, are represented in our area by Sanguinaria and Stylophorum, and by Chelidonium majus, a naturalized weed. Glaucium flavum Crantz (2n

* Chelidonioideae, subfam. nov. Papaveraceae pilis mollibus, bog multicellu- laribus, terminaliter uniseriatis, septis perpendicularibus (in onia Y in Maes ramosis, fastigatis et imbricatis) ; perianthio dimero Gn Eomecone synse- palo; in Saucers polypetalo; in Macleaya et a apetalo) ; gynoecio biplacen- tato et bivalvato (valvis 3-5 in Stylophoro diphyllo); vena carpellare mediana bene explicata; valvis acropetaliter vel basipetaliter as lait sed dehiscentia plerumque laterale et subcompleta; seminibus saepe arillatis (arillus deest in Glaucio, Dicranostig- mate, et Macleaya microcarpa) ; polline tricolpato vel polyporato. Herbae perennes, saepe rhizomatosae vel in Bocconia ogee in hemisphaerio septentrionale. Typus GENERICUS SUBFAMILIAE: ae ie ee L. Genera AltA: Bocconia L., Coreanomecon Nakai, Dicranostigma Hook. & Thom age Hance, Glauctum Mill, Hylomecon Maxim., Macleaya R. Br., Sanguinaria L, Stylophorum Nutt

Ww —. a

1962 | ERNST, PAPAVERACEAE & FUMARIACEAE 317

= 12, 24) and Macleaya cordata (Willd.) R. Br. (2n = 20) are reported as garden escapes or weeds to the north of our area. The dehiscence of the valves is acropetal in Sanguinaria and Chelidonium but basipetal in Stylophorum, Glaucium Mill., and Macleaya R. Br.

The Papaveroideae,* see with us by Argemone and at leas three introduced species of Papaver, tend to have multicellular-multi- seriate hairs, 2- or 3-merous perianths, 3-colpate |or several-aperturate | pollen, and usually more than 3-valved gynoecia; the median carpellary traces tend to be absent and are replaced, in Papaver and some other genera, by pse eee ae traces originating from the placental bundles; dehiscence of the valves is basipetal, and the seeds usually lack arils.

e Seeder Ernst,? occasionally represented in our area by Eschscholzia californica, are glabrous or have unicelluiar hairs, usually 2-merous perianths, polycolpate pollen, and 2-valved, conspicuously ribbed gynoecia with well-developed median carpellary traces: the dehiscence of the valves is acropetal and violent; the seeds usually lack arils.

latystemonoideae Ernst,’ vernal herbs of the western United States, have multiseriate hairs, 3-merous perianths, 3-colpate pollen, 3- or several-carpellate gynoecia with median carpellary traces, and discrete stigmas; the fruits open by splitting through the placentae without the formation of valves, and the seeds lack arils. he family is rich in alkaloids, some of them important narcotics, with both medicinal and poisonous properties claimed. The alkaloid protopine, originally thought common and unique to both Papaveraceae and Fu- mariaceae, has been identified in the Berberidaceae; berberine, having

e same empirical formula, is reported for Papaveraceae and six other families (see Willaman & Schubert). The acrid sap, usually colored and 4 averoideae, subfam. nov. Papaveraceae pilis multicellulari-multiseriatis ; perianthio di- vel trimero (aliquando polypetalo) ; gynoecio triplacentato vel placentis valvisque numerosioribus; vena carpellare mediana plerumque deest, saepe vena pseudo- mediana de fasce placentale substituta; fructu basipetaliter oe nte; in Arctomecone

humile seminibus arillatis; polline tricolpato vel pluvi-aperturato. Herbae annua vel perennes (raro sublignoste), praecipue in hemispherio Cee TYPUS GENERI- CUS SUBFAMILIAE: Papav . GENERA ALIA: Arctomecon Torr. & Frém., Argemone

L., Canbya Parry, Meconepss Vig. (Cathcartia Hook. f.), Roemeria Medic., Romneya Harvey, Sivlomecon G. Tayl.

‘Eschscholzioideae, subfam. nov. VPsapaveraceae plerumque glabrae, vel pilis Trea et papillatis; perianthio trimero vel plerumque dimero (in Eschscholzia synsepalo); gynoecio biplacentato et bivalvato, longitudinaliter conspicue nervato; vena carpellare mediana bene explicata; valvis acropetaliter ees dehiscenti- us; in Dendromecone seminibus arillatis; polline polycolpato. erbae annuae vel perennes breve victi, vel lignosae, Civitatium Consociatorum bee et Mexiconis septentrionalis orientalisque. YPUS GENERICUS SUBFAMILIAE: Fschscholzia Cham. GENERA ALIA: D

Platyste ape foliis angustis et integris, saepe verticillatis; perianthio trimero; gynoecio tri- pluri-placentato, vena carpellare mediana rikawern ndo tarde explicata; fructu semi- a carpico, carpellis partim vel in toto divisis per placentis sine !ormatione valvaru seminibus exarillatis; polline tricolpato. Herbae vernales Californiae, etiam erin et Washingtonis. TyPUS GENERICUS SUBFAMILIAE: Platystemon Benth. GENERA ALIA: Hesperomecon Greene et Meconella Nutt. ex Torr. & Gray

318 JOURNAL OF THE ARNOLD ARBORETUM [voE Stir

darkening after exposure, apparently contained in articulated tubes or nonarticulated sacs, is an important characteristic of the family.

The interpretation of the gynoecial structure has been hotly contested by morphologists, some concluding, for example, that the 2-valved fruits are the equivalent of four carpels arranged in two series (see Bancroft), while Arber favored the classical 2-carpel interpretation. The latter view, both simple and appealing, is applicable with equal success to all mem- bers of the Papaveraceae and the Fumariaceae. The 2-valved fruits in Papaveraceae seem less specialized in some respects than the multivalved ruits,

It seems generally agreed that the Papaveraceae are to be placed near the Ranales, on the one hand, and near Fumariaceae, Cruciferae, and Capparaceae, on the other. Papaveraceae and Fumariaceae have 2-seriate corollas of twice as many petals as sepals, while the corollas of Cruciferae and Capparaceae are + uniseriate and the petals as many as, and alternate with, the sepals. The insertion of the perianth segments is fundamentally similar in Fumariaceae and Papaveraceae, but the patterns of the traces to the perianths are + different. The sepals of Papaveraceae inclose the buds until anthesis, while those of Fumariaceae being mere scales, sometimes without traces, do not

Chromosome aamibers are reported for many species (2n = 12, 14, 16, 18; 20, 22). 29, 20, Cay, 32,84, 20, 38, 40,42, 44, 50;-Caz 58,. 70,00, 34,

1. 112). Gametic numbers of 6, 7, or multiples thereof are + common, while 8-11, 16, and 19 are uncommon and usually from rather exceptional taxa in the family. Species small in habit and usually annual tend to have low chromosome numbers, while those large in habit, among them the woody members of the family, tend to have higher numbers.

Economically the family is important as the source of showy garden plants; edible poppy seed, poppy-seed oil, and opium and its derived alkaloids are obtained from Papaver somniferum.

=

REFERENCES:

ALEXANDER, I. Entwicklungsstudien an Bliiten von Cruciferen und Papaveraceen. Planta 41: 125-144. 1952.

ARBER, A. Studies in flower ae IV. On the gynaeceum of Papaver and related genera. Ann. Bot. II. 2: 649-663. 1938. | Basic structure of gyno- ecilum, with detailed Keren of the inverted portions of the placental

bundles. |

Bancrorr, H, A review of researches concerning floral morphology. Bot. Rev. bE 77-99, 1935, [Synopsis of controversy over carpel polymorphism, 80- 84. |

BerNuaArpr, J. Uber den Charakter ae die eg der Papaveraceen und Fumariaceen. Linnaea 8: 401-477. 1833. [See also ibid. 12: 651-668. 1838. |

Bie. J.-D., & G. Bocguer. se campylotrope : les Rhoeadales. Actes Soc. Helvét. Nat. Argovie 1960:

BeRSILLON, G. Recherches sur ae ee iéracées; contribution a l’étude du développement des dicotylédones herbacées. han Sci. Nat. Bot. XI. 16: 225-447. 1955. [Primarily Papaveraceae; a general coverage of phyllo-

>

1962] ERNST, PAPAVERACEAE & FUMARIACEAE 319

taxy, histology, differentiation of conductive tissue, and floral morphology ; extensive bibliography.

Buxatscu, F. Zur Analyse der Bakterienhemmstoffe aus der Wurzel vom Schoéllkraut und ahnlichen Pflanzen. Arch. Mikrobiol. 24: 281-296. 1956. [Includes Chelidonium majus and Sanguinaria canadensis. |

Cuist1aKkov, I. D. Organogensis in the family Papaveraceae Hook. Moscow. 1871.* [Fide A. L. TAKHTAJAN, pp. 21, 22. Problems in evolutionary plant morphology. AIBS English translation by O. H. GANKIN, ed. G. L. STEB- bins. 139 pp. Washington. 1959. |

Crete, P, Papavéracées. Le développement de l’embroyn chez le Dicranostigma franchetianum (Prain) Fedde. Compt. Rend. Acad. Sop. Paris 245: 720= fide. 1997;

Dickson, J. Studies in floral anatomy. II. The floral anatomy of Glaucium flavum with references to other members of the Papaveraceae. Jour. Linn. Soc. 175-224. 1935. [A polycarpellate interpretation of the gynoecium. |

Eames, A. J. The vascular anatomy of the flower with refutation of the theory of carpel polymorphism. Am. Jour. Bot. 18: 147-188. 1931. | Directed at the works of E. R. Saunders. |

Morphology of the angiosperms. xili + 513 pp. McGraw-Hill, New

York. 1961. [Reaffirms polycarpellate interpretation of the gynoecia, 198,

243: reports embryos undifferentiated when seeds shed in several genera

___— & C. L. Wutson. Crucifer carpels. Am. Jour. Bot. 18: 638-656. 1930. [Spirited reassertion of polycarpellate explanation of the cruciferous gyno- ecium, extended to closely allied families; gynoecium of Corydalis aurea shown as though 3-locular, 654; see earlier paper (1928) for application to Fumariaceae and Capparaceae, therefore indirectly to Papaveraceae. |

Ernst, W. R. Chromosome numbers of some western Papaveraceae. Contr. Dudley Herb. 5: 109-115. 1958. Chromosome numbers of some Papaver- aceae. Ibid. 137-139. 1959.

A comparative morphology of the Papaveraceae. vii + 213 pp. 202 figs. Ph.D. Dissertation, Stanford Univ., Calif. 1962. | Excludes Fumaria- ceae; deals primarily with floral morphology at the generic level, recognizes 4 subfamilial groups. |

Feppe, F. Die geographische Verbreitung der Papaveraceae. Bot. Jahrb. 36(Beibl. 81): 28-43. 1905. [Includes Fumariaceae. |

———. Papaveraceae-Hypecoideae et Papaveraceae-Papaveroideae. Pflanzen- reich IV. 104(Heft 40): 1-430, 199° [The basic work; family treated as composed of subfamilies Hypecoideae, Papaveroideae, and Fumarioideae, the last omitted. |

Papaveraceae. Nat. Pflanzenfam. ed. 2. 17b: 5-145. 1936. {Important synopsis covering for the first time subfam. Fumarioideae. |

Gates, B. N. Carunculate seed dissemination by ants. Rhodora 45: 438-445. 1943. [Includes Chelidonium majus and Dicentra spectabilis. |

Gray, A. Ordo Papaveraceae. Gen. Pl. U. S. 1: 108-116. pls. 47-49, 1848. [Argemone, Stylophorum, Sanguinaria. |

Harms, H. Reihe Rhoeadales. Nat. Pflanzenfam. ed. 2. 17b: 1-4. 1936.

Harvey-Grpson, R. J., & M. Brapitey. Contributions towards a knowledge of the anatomy of the lower dicotyledons. I. The anatomy of the stem of the Papaveraceae. Trans. Roy. Soc. Edinb. 51: 589-608. pls. 1-3. 191

320 JOURNAL OF THE ARNOLD ARBORETUM [ VOL, XLIII

HeGNAUER, R. Die Gliederung der Rhoeadales sensu Wettstein im Lichte der Inhaltstoffe. Planta Med. 9(1): 37-46. 1961.*

HoneyMan, J. M. On the occurrence of cyanogenetic glycosides in the order Rhoeadales. Taxon 5: 33, 34. 1956.

Hurcuinson, J. Bocconia and Macleaya. Kew Bull. 1920; 275-282. 1920.

—. Contributions towards a phylogenetic classification of flowering plants: V. The genera of Papaveraceae. /did. 1925: 161-168. 1925. [A radical classification with keys to tribes and genera; suggests a phylogeny orig- inating from forms with few stamens.

Lecer, I.. Recherches sur l’appareil végétatif des Papavéracées Juss. Mém Soc. Linn. Normandie 18: 195-623. 1894-1895. [| Anatomical study, meladiag Fumariaceae, of vegetative organs and laticiferous system; an unlikely phylogeny presumes Fumaria basic to Fumariaceae, these giving rise to 2 ultimately confluent phylads of Papaveraceae, 5 7]

LIGNIER, O. Notes anatomiques sur l’ovaire de quelques Papavéracées. Bull. Soc. Bot. Fr. 58: 279-284, 337-344, 429-435. 1911

ManesHwakr!, P., & K. K Intraovarian selina in Eschscholzia cali- fornica Chan, Argemone mexicana L., and A. ochroleuca Sweet. Nature 191; 304. 1961.*

MarkKcrafF, F. ci ee phylogenetische Studien an Bluten von Cruciferen und Papaveraceen. Proc. 7th Int. Bot. Congr. 1950: 352-354. 1953. eae phylogenetische Studien an Bliiten der Rhoeadales. J/bid. 857,

ee g, Untersuchungen Uber den Blutenbau der Papaveraceen. Vet-akad.

Handl. 50: 1-168. pls. 1-28. 1912. |General morphological approach to most genera of Papaveraceae and Fumariaceae showing many floral anom- alies; separates Pteridophylloideae from Hypecoideae and Fumarioideae. |

NorDHAGEN, R. Remarks on some new or little known myrmecochorous plants from North America and East Asia. Bull. Res. Council Israel D 201. 1959. fees ee Chelidonium, etc.; discusses termi- nology of seed-ant relationship; thinks stem anatomy degenerate.

Norris, T. Torus aan and nectary ee as phylogenetic criteria in Seni Am. Jour. Bot. 28: 101-113. 1941. [Includes Fumariaceae. |

Payer, J. Traité d’organogénie comparée de la fleur. 2 vols. Masson, Paris, Lot Excellent illustrations of developmental stages, including Fumaria- ceae. |

Popov, M. G. On the relationships and the history of Papaver and Roemeria. (In Russian.) Bot. Zhur. 42: 1389-1397. 1957. | Published posthumously ; deals little with Papaver and Roemeria at the generic level; a general phylogenetic consideration of the Papaveraceae and Fumariaceae, strongly influenced by weg of color and hue of the petals and modern dis- tribution of the s

Roper, I. Ae ctont atte ond fluoreszenzoptische Untersuchungen an Same Papaveraceen. Osterr. Bot. Zeitschr. 104: 370-381. 1958. | Shows ane

of seed coats, including some Fumariaceae. |

SAGDULLAJEVA, A. L, Pollen morphology of Papaveraceae. (In Russian; very brief paeee sy hae Probl. Bot. 4: 11-50. 1959. [Reports on 84 spp. in 28 genera, including Fumariaceae; 10 types of pollen are described; well oe a: i of Dendromecon shown is atypical.

SAUNDERS, E. R. Floral morphology. Vol. 1. vii + 132 pp. W. Heffer & Sons, Ltd., Cambridge. 1937. | Papaveraceae and Fumariaceae, 45-56; her

1962 | ERNST, PAPAVERACEAE & FUMARIACEAE 321

complete literature references, 23-25; see particularly New Phytol. 29: 44-55. 1930; advocates the polycarpellate interpretation of the gynoecium. |

SHaw, C. H. Note on the sexual generation and the development of the seed- coats in certain of the Papaveraceae. Bull. Torrey Bot. Club 31: 429-433. pl. 15. 1904. [Includes Sanguinaria, Chelidonium, and Eschscholzia; re- ports division of microsporocytes of Sanguinaria in early spring. |

SIMS, . Bocconia cordata. Bot. Mag. 44: pl. 1905. 1817. [= Macleaya cor- dat

SLAVIK, "y, & L. StavixovaA. Alkaloids of the Papaveraceae. V. Chem. Listy 48: 1557-1559. 1954; VI. Ibid. 49: 106-110. 1955; VII. Lbid. 1546-1549; IX. Ibid. 51: 1923-1926. 1957. (In Czech); VIII. Collect. Czech. Chem. Commun. 22: 279-285. 1957; XIV. Ibid. 25: 756-760. 1960; XVI-AVII, XV. bid. 1663— 1675, 1698, 1699. (In German.) [None seen; order of

Sources, R. Embryogénie des Papavéracées. Développement de l’embryon chez le Roemeria violacea Medic. (R. hybrida DC.). Compt. Rend. Acad. Sci. Paris 226: 979-981. 1948.

WittaMaN, J. J., & B. G. Scuusert. Alkaloid-bearing plants and their con- tained alkaloids. U. S. Dep. Agr. Tech. Bull. 1234: 1-287. 1961. [Gives botanical names and reported alkaloids, cross-referenced with empirical formulas and plant index; literature citations through 1959, from Chem. Abstr. through 1957. |

Key TO THE GENERA OF PAPAVERACEAE

General characters: plants herbaceous, sometimes rhizomatous, with acrid, usually colored sap; leaves rosulate or usually alternate (rarely opposite) ; in- florescence terminal and cymose (rarely paniculiform|; flowers solitary or clustered, subtended by a leaf or bract, bisexual, hypogynous (or rarely perig- ynous), essentially regular; perianth cyclic, 2- or 3-merous; sepals inclosing the buds until anthesis, then caducous; petals twice as many as sepals and 2- seriate (occasionally polypetalous) frarely apetalous|; stamens many; gynoe- cium 1-locular with 2 or more parietal placentae; fruit dehiscent by 2 or more valves A. Inflorescence ees or cymose, flowers solitary or several and clustered;

petals 4-6 or m

B. Gynoecium ae 2 placentae and valves

C. Leaf blades palmately veined and lobed, glabrous; flowers solitary on scapes; dehiscence - valves acropetal; seeds arillate; low perennial fromthick: rhizOme:. 45 co.ee ee ee ee tet eae 1. Sanguinaria.

C. Leaf blades Reet veined, lobed or dissect D. Flowers hypogynous, aposepalous: leaves esac broadly pin-

nately dissected. E. Inflorescence subumbellate, flowers subtended by small, nar- row bracts; dehiscence of valves acropetal; seeds arillate. Pe ET ee en tates cy Ce gt eee ae 3. Chelidonium. E. Inflorescence an expanding cyme, flowers subtended by bracts; locule occluded, dehiscence of valves basipetal; seeds nonarillate; fruit linear, becoming 10 cm. or more long; ne liage BEVig: tt 2 Rests tees aon A Mee mc Ree neni een Ash St a | Glaucium., | D. Flowers perigynous, synsepalous; leaves pinnately finely eee

oz2 JOURNAL OF THE ARNOLD ARBORETUM [ VOL, XLIII

glabrous (or hairs pagan, fruits longitudinally ribbed, de- NISGERCE CIO DOIAls. @ 4 arcade kw ee Ro egw Ad Verma 4. Eschscholzia.

B. Gynoecium with 3 or more placentae and valves; dehiscence basipetal. F. Flowers on + weak pedicels subtended by small bracts and 2 or 3 + opposite leaves on a naked shoot; hairs soft, multicellular-uniseri- ate; fruits nodding, dehiscence + complete: seeds dae Ae sae OVENS coined a Sine 4d eee we Wee ore deb ee We ye . Stylophorum. “, Flowers on ‘rigid pedicels subtended by leaves or hone bracts on leafy shoots; hairs multicellular- oO fruits erect, dehiscence

incomplete; seeds nonarillate; ann

G. Flowers on short pedicels’ eee undulate, style short; leaves Tes A OI cess cd Sem, had Gon Bini SusseG Sana wa cee : Arsemone, G. a on long pedicels; stigmas appearing radiate and embedded a cartilaginous disc, style absent; leaves not prickly. 6. Papaver. A. ipa paniculiform, flowers numerous, apetalous; fruits 2-valved, flattened, nodding, dehiscence basipetal; tall herbaceous perennials wit th broad, long: DeLwIte LEAVES, 6 oc unica wekey ad ee ekwwe ee xu eas | Macleaya. |

— 4

Subfam. CHELIDONIOIDEAE Ernst 1, Sanguinaria Linnaeus, Sp. Pl. 1: 505. 1753; Gen. Pl. ed. 5. 223. 1754.

Low, glabrous, perennial herb with red-orange sap, from a thick rhizome. Leaves few, with broad, basically reniform-sagittate, palmately veined and variously lobed Bladas, long-petiolate, subtended by 2-6 ovate to spatulate bracts. Flowers solitary on naked scapes (or very rarely 2 or 3, each subtended by a bract); scape at first enfolded by a leaf. Sepals 2, thin, + notched apically. Corolla often appearing quadrate, 2 (or more )-seriate, with 8 (rarely 6-12), narrow, white or pale-pink, sometimes irregularly inserted petals. Stamens many; anthers oblong; pollen poly- porate with 9-16 apertures. Stigma 2-lobed; style + definite; placentae 2, Fruit broadly fusiform, 2-valved; dehiscence + acropetal, usually in- complete apically. Seeds few to many, black or reddish-orange with a conspicuous aril along the upper surface. Seedling cotyledons hypogeal, oblong, short-petiolate. Typr species: S. canadensis L. (Name from Latin, sanguinarius, bleeding, from the color of the sap.) — BLooproor, PUCCOON

A monotypic genus of eastern North America. Sanguinaria canadensis,

= 18,° with autumnal meiosis of the pollen mother cells, is among the earliest vernal herbs of deciduous woodlands, often occurring on limestone or calcareous soils, from Texas to northern Florida north to Nova Scotia and Manitoba. The leaf shape is variable; the seeds are reported to be attractive to ants.

The closest affinity is with the monotypic Eomecon Hance, of eastern Asia, which shares with Sanguinaria a preference for the forest habitat,

* Original determination by the author from squash preparations of microsporocytes from plants cultivated in the Beal-Garfield Botanic Garden, Michigan State University, East Lansing, Michigan (Gillett 1264, collected 1 November 1961); nine pairs of chromosomes were observed in several cells

1962] ERNST, PAPAVERACEAE & FUMARIACEAE o23

palmately veined, long-petiolate leaves with broad blades, white petals, 2-valved, fusiform fruits, arillate seeds, and polyporate pollen; but the leaves of Eomecon are cordate and sparsely pubescent, the calyx is syn- sepalous, and the inflorescence is a branched scape with each of the several flowers pedicellate and subtended by a small bract.

REFERENCES: Under family references see BUKATSCH, GRAY, NoRDHAGEN, and SLAVIK & SLAVIKovA (XVII BAKER, R. O. The en development, gross structure, ecology and ee ical distribution of Sanguinaria canadensis. Nat.-Study Rev. 15: 62-73 1919,*

Bastin, E. S. Some further observations on the structure of Samguinaria canadensis. Am. Jour. Pharm. 67: 4-9. 1895.

Curtis, W. Sanguinaria canadensis. Bot. Mag. 5: pl. 162. 1791.

Gates, B. N. The dissemination by ants of the seeds of bloodroot, Sanguinaria canadensis. Rhodora 44: 13-15. 1942. [Seeds red-orange. |

erg S. Some notes on the pollination of flowers. Bull, Wis. Nat. Hist.

_ II. 4: 12-21. 1906. [Flowers proterogynous. |

eo G. A. Alkaloids from Sanguinaria canadensis and their influence on growth of Phymatotricum omnivorum, Pl. Physiol. 14: 377-380. 1939.

GREENE, E. L. ee regarding Sanguinaria. ee 5: 306-308. 1905. [Treats 6 spp., 4.n

Harris, J. A. A mire ane study of the morphology of ue fruit of the bloodroot, Sanguinaria canadensis. Biometrika 7: 305-351.

ion in the inflorescence of Sanguinaria. Biol. eared 30: 629-633. 1910.

ae J. W. Juvenile and adult forms of bloodroot. Pl. World 6: 106- 108. 19 iE lustrates variation in leaves

Hom, T. Sa to the knowledge ae the germination of some North American plants. Mem. Torrey Bot. Club 2: 57-108. 1891. [Cotyledons of Sanguinaria hypogea al. |

Medicinal plants of North America. 18. Samguinaria canadensis L. Merck’s Rep. 17: 209-212. 1908.

Jounson, R. H. Aberrant societies of Sanguinaria and Trillium, Torreya 9: 5, 6. 1909. [Petals number 6-12, primarily 8.]

MEEHAN, Sanguinaria canadensis. American blood-root. Meehans’ Monthly J oes 900.

eae J. A. Notes on the seedlings of bloodroot. Am. Midl. Nat. 1: 199-203. 1910. | Extensive observations on morphology and history. | Spencer, W. P. Variation in petal number in the blood-root, See canadensis. Am. Nat. 78: 85- a 1944.

SURFACE, F. M. Contribution to the life history of Sanguinaria canadensis. Ohio Nat. 6: 379-385. 1905. |Eight-celled embryo sac; autumnal division of microsporocytes. |

2. Stylophorum Nuttall, Gen. N. Am. Pl. 2: 7. 1818.

Low, pubescent, perennial herbs with yellow to orange sap, developing a stout rhizome; hairs multicellular, terminally uniseriate. Leaves ros- ulate, petiolate, deeply 5-7-pinnatifid, pale beneath. Inflorescence um-

324 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLII

belliform, few flowered, bracteate, and subtended by 2 or 3 + opposite leaves terminating an elongated naked shoot. Sepals 2, attenuate, api- cally involute and reflexed, the upper left margin lobed. Petals 4, 2- seriate, yellow. Stamens many; filaments narrow, often constricted below the oblong anthers; pollen usually with several + obscure apertures. Stigma truncate with 3 or 4 shallow lobes; style conspicuous; placentae 3 or 4 [or placentae 2, stigma deeply 2-lobed, and style short]. Fruits ellipsoid, usually 4-valved |or + linear and 2-valved], pubescent and nodding; dehiscence basipetal and + complete. Seeds few to many, with a conspicuous aril along the upper surface. Lectotype sprcies: S. diphyllum (Michx.) Nutt.; see Britton & Brown, Illus. Fl. No. U. S. ed. 2. 2: 140. 1913. (Name from Greek, stylos, style, and phoros, bearing, “from the distinct and conspicuous style which distinguishes this plant from all the rest of the Papaveraceae,”’ a distinction no longer true.

carpellate ovaries, in cross section, diagramma tic, x 4 of, mature fruit 7 g, four-valved fruit after dehiscence, one valve removed, x 4%; h, arillate seed, the micropyle below, < 7

A genus of two or three species of eastern North America and eastern Asia. Stylophorum diphyllum, celandine poppy, wood poppy, mock poppy, or yellow poppy, 2” = 20,‘ occurs in rich woods and on bluffs, sometimes on calcareous soil, from northern Arkansas and Tennessee to Pennsylvania and Wisconsin. The degree of dissection of the leaves is variable. Our species, described by Michaux as a species of Chelidonium and reunited with Chelidonium by Prain (1895), is remarkably similar in the vegetative

* Original determinations by the author from squash preparations of microsporocytes from a plant cultivated at Stanford University, California (grown from seeds from the Royal Botanical Garden, Uppsala, Sweden), and from a plant cultivated at Waverley, Massachusetts (£rust 800 [GH]); twenty chromosomes were observed in the former and ten pairs were observed in several cells of the latter

1962] ERNST, PAPAVERACEAE & FUMARIACEAE 325

stage to the Old World C. majus. Stylophorum, with characteristic com- pact inflorescences usually subtended by two + opposite leaves on a naked shoot, pubescent fruits with basipetal dehiscence, and pollen with several + obscure apertures, contrasts markedly with Chelidonium majus, in which the subumbellate inflorescences arise from branched, leafy stems, the dehiscence of the glabrous fruits is + acropetal, and the pollen is 3-colpate. The closest affinities of S. diphyllum, characterized by ellip- soidal, usually 4-valved fruits and + rounded divisions of the leaves, is with S. lasiocarpum (Oliver) Fedde (described by Oliver as a species of Chelidonium) which has elongated, almost linear, 2-valved fruits, and + pointed divisions of the leaves.

REFERENCES: Under family references see Gray and NORDHAGEN , under Chelidonium see PRAIN. Boynton, K. R. Stylophorum diphyllum. Addisonia 3: 31, 32. pl. 96. 1918. Hooker, W. J. Stylophorum diphyllum. Bot. Mag. 81: pl. 4867. 1855. SCHLOTTERBECK, J. O., & C. R. Ecker. The development and structure of the seed of Stylophorum diphyllum. Proc. Am. Pharm. Assoc. 50: 401-404.

3. Chelidonium Linnaeus, Sp. Pl. 1: 505. 1753; Gen. Pl. ed. 5. 224.

Low, branched, pubescent, biennial or perennial herb with yellow sap, developing a stout rhizome or taproot; hairs multicellular, terminally uniseriate. Leaves rosulate and cauline, + petiolate below, deeply 5-7 (or more)-pinnatifid, with rounded |or laciniate| divisions. Inflorescences subumbellate, with several flowers, bracteate, terminal and axillary, aris- ing from branched, leafy shoots. Sepals 2, somewhat hooded, the upper

Fic. 2. Chelidonium. a—h, C. majus: a, flowering and fruiting shoot, X %; b, opening flower bud, showing emerging stigma, X 3; c, flower, X 3; d, stamen,

10: e, gynoecium from flower bud just before anthesis, X 10; f, two-valved fruit after dehiscence, X 2; g, portion of persistent placenta showing attachment of seeds, X 4; h, arillate seed, X 20.

326 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIII

left margin lobed. Corolla appearing cruciform; petals 4, 2-seriate, yeilow. Stamens several to many; anthers elliptical; filaments narrow, usually constricted below the anther; pollen 3-colpate. Stigma 2-lobed; stvle short or indefinite; placentae 2. Fruit narrowly oblong or linear, some- times curved, 2-valved; dehiscence + acropetal, usually incomplete api- cally. Seeds few to many, with a conspicuous aril on the upper surface. Seedling cotyledons petiolate. Lectotype species: C. majus L: see Britton & Brown, Illus. Fl. No. U. S. ed. 2. 2: 141. 1913; typified by removal of other Linnaean species to other genera. (Name from Greek, chelidon, swallow, because, according to Aristotle and other scholars, the mother-swallows bathed the eyes and thereby strengthened the sight of their young with the saffron-colored sap.) — CELANDINE, SWALLOWWORT, ROCK POPPY.

An Old World genus, probably monotypic. Chelidonium majus, 2n = ee usually of damp, shady areas, roadsides, rock walls, and waste places, widely naturalized in eastern North America, occurs in northwestern North Carolina and probably elsewhere in our area. Forms with large flowers or with variously dissected or laciniate leaves are occasionally seen in collec- tions from both Europe and Asia.

Linnaeus (1753) included two species of Glaucium and one of Roemeria Medic. in his Chelidonium. Later, Prain united Stylophorum, Hylomecon Maxim., and Dicranostigma Hook. & Thoms. with Chelidonium: but Glaucium, its generic distinction from Dicranostigma debatable, was not included. The vegetative stages of Chelidonium majus and Stylophorum diphyllum (q. v.) are similar.

REFERENCES:

Under a references see BuKATSCH, NoRDHAGEN, and SLAvIK & SLAVI- KOVA (XVII BANDELIN, : J.. & W. Matesn. Alkaloids of ipa majus, L., leaves

and stems. I. Jour. Am. Pharm. Assoc. Sci . 45: 1956.*

BoRATYNSKA, W. Die Variabilitat des rn aes im Schéllleraut (Cheli- donium maius L.) unter dem Einfluss des radioaktiven Isotopes 60Co. (In Polish; German summary.) Poznan. Inst. Przem, Zielarskiego. Bull. 6(34): 255- 262. 60.

Crete, P. A propos de la strophiole du Chelidonium majus L. Bull. Soc. Bot. Fr. 91: 88. 1944, [Complains that his work (bid. 84: 196-199. 1937) re- garding the epidermal origin of the strophiole was overlooked by Szemes. |

KaczMarEK, F., & B. MALex. Zur Papier- chromatographie der Alkaloide von sided majus L. (English summary.) Planta Med. 7(2): 171-173.

59.

ea A. La chélidoine double et la chélidoine laciniée en Belgique. Bull. Jard. Bot. Bruxelles 25: 409, 410. 1955.

Prain, D. A revision of the genus Chelidonium. Bull. Herb. Boiss, 3: 570-587. 1895. [Combines Chelidonium, Stylophorum, a and Dicrano-

stigma: oo intimately related to this complex, is not considered. | SZEMES, G. ur Entwicklung des Elaisoms von Oaioaaa. majus. Wiener Bot. ae 92: 215-219, 1943

1962] ERNST, PAPAVERACEAE & FUMARIACEAE 327

Ternpau, M. Die aus Chelidonium majus L. isolierten Alkaloid. (Abstr.). Pharm. Zentralh. Deutsch. 97(3): 147. 1958.*

Subfam. ESCHSCHOLZIOIDEAE Ernst 4. Eschscholzia Chamisso in Nees, Horae Phys. Berol. 73. pl. 15. 1820.

Low, annual or perennial herbs with thin yellow to orange sap, sometimes developing a stout taproot; usually glabrous (or hairs 1-celled). Leaves rosulate and/or cauline, alternate or subopposite, petiolate, + ternately finely dissected [or rarely entire]. Flowers and fruits erect on long pedi- cels; early flowers usually terminating short shoots, later flowers often terminal and axillary on long shoots; receptacles perigynous. Calyx syn- sepalous, usually attenuate. Petals usually 4, 2-seriate, yellow or orange [sometimes pinkish or white|. Stamens usually many, often adhering to base of petals; filaments thick, sometimes with a dark spot; anthers usually narrowly oblong; pollen polycolpate with 5-7 [10] slitlike |sometimes confluent] apertures. Stigmas 2 to several (the secondary stigmas over the placentae); style absent; placentae 2. Fruit columnar, constricted be- low the apex, conspicuously ribbed, 2-valved; dehiscence acropetal, in- complete, usually violent. Seeds many, reticulate [papillate or variously pitted]. Seedling cotyledons linear, bifurcate [but usually entire]. TYPE species: Eschscholzia californica Chamisso. (Named for J. F. Eschscholtz,® 1793-1831, zoologist of Dorpat, friend and companion of Chamisso on the Kotzebue expedition which visited California in 1816 on the Rurick.) — CALIFORNIA POPPY.

At least 14 species of northwestern Mexico and the western United States, concentrated in California. Perhaps half of the 123 species listed by Fedde are referable to Eschscholzia californica, 2n = 12, cultivated for its attractive flowers, adventive in many parts of the world, and col- lected in our area at least in North and South Carolina; it is distinguished by the apparently unique bifurcated cotyledons and the fleshy receptacle rim. In nature, as in cultivation, the species is exceedingly variable.

The 2-valved, deeply ribbed fruits of Eschscholzia, sometimes described as polycarpellate, are most easily interpreted as the equivalent of two carpels. The longest stigmas are centered over the median carpellary traces; secondary stigmas, perhaps a peculiarity of reduction in the vas- culature, often develop over the placentae. The combination of synsepaly, perigyny, and stigmatic structure suggest that Eschscholzia is relatively

specialized. The genus is allied most closely with the monotypic Hunnemannia Sweet, 22 = 56, of eastern Mexico, in which the flowers are somewhat

perigynous but aposepalous. Synsepaly is also characteristic of Homecon,

°-The name of the man is usually given thus, but the generic name was rendered Eschscholzia in the original description (misspelt Escholzia on the plate) and in the later reprint (Linnaea 1: 553-555. 1826), seemingly confirming the intended orthogra- phy.

328 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIIT

of eastern China, but the affinities of the latter are with Sanguinaria and other Chelidonioideae.

Chromosome numbers of 27 = 12, 14, 22, 24, 34, and 36 have been reported

REFERENCES: Under family references see MAHESHWARI & KANTA.

Hooker, W. J. Eschscholzia californica Cham. Bot. Mag. 56: pl. 2887. 1829. [Includes history; see also /bid. 63: pl. 3495. 1863 (as E. crocea).]

Jepson, W. L. Eschscholtzia. Fl. Calif. 1: 564-575. 1922. [The most complete account of variation in E. californica, with observations on transplants. |

Lewis, H., & F . Went. Plant growth under controlled conditions. IV. Response of finns annuals to photoperiod and temperature. Am. Jour. Bot. 32: 1-12. 1945. | Brief account of modification of vegetative characters in E. californica and E. lobbii. |

Licnier, O. Eschscholtziées. Explication anatomique de la fleur. Bull. Soc, Bot. Fr. 62: 298-315. 1915.

SACHAR, R. C., & H. Y. M. Ram. The hea of Eschscholzia californica Cham, Phy tomorphology 8: 114-124. 1958. | Polygonum-type embryo sac. |

Saito, K. On a haploid plant eae from the induced autotetraploid plant of common California poppy. (In Japanese; English summary.) Jap. Jour. Breed. 7(3): 152-156. 1958.*

Sources, R. Embryogénie des Papavéracées. Développement de Vembryon chez PEschscholzia californica Cham. Compt. Rend. Acad. Sci. Paris 229: 485— 487. 1949

Subfam. PAPAVEROIDEAE 5. Argemone Linnaeus, Sp. Pl. 1: 508. 1753; Gen. Pl. ed. 5. 225. 1754.

Stout, prickly, annual [or biennial] herbs [rarely shrubs] with yellow [or white to reddish-orange| sap; prickles and hairs multicellular-multiseri- ate, harsh, smooth and polished [or rough and/or branched]. Leaves rosulate and cauline, pinnately veined and lobed, sessile, clasping, or some- what decurrent, sometimes narrowed into a petiole- like base. Inflorescence an expanding cyme, few to several flowered and bracteate. Sepals 2 or 3, + hooded, usually attenuated into a sharp prickle. Corolla 2-seriate, oath twice as many petals as sepals (occasionally polypetalous), yellow or litte [bronze or lavender]. Stamens numerous; filaments narrow; anthers nar- rowly oblong, apically revolute after anthesis; pollen 3-colpate. Stigmatic region deeply undulate, the lobes 3 — several (margins of adjacent stigmas forming conspicuous lobes over the placentae); style short or obscure; placentae as many as stigmatic lobes. Fruit usually more than 3- valved; dehiscence basipetal and incomplete. Seeds numerous, reticulate. ee ling cotyledons linear-lanceolate. LECTOTYPE SPECIES: A. mexicana L.: Britton & Brown, Illus. Fl. No. U. S. ed. 2. 2: 138. 1913. (Name ne an herb mentioned by Pliny; perhaps “from the Greek, argema, the cataract of the eye for which the juice of a poppy-like plant of the same name was a supposed remedy” [Ownbey].) — PrickLy poppy.

1962] ERNST, PAPAVERACEAE & FUMARIACEAE 329

About 28 species (Ownbey) lacking clear subgeneric groupings, often of disturbed habitats in the New World; one presumed endemic in the Hawaiian Islands. Argemone mexicana, devil’s fig, 2n = 28, with yellow petals and leaves usually blotched with white along the main veins, occasional (mostly in the Coastal Plain) in our area, is a cosmopolitan weed of warmer climates, and may be native to the West Indies, Central America, and Florida (Ownbey). Forma /eiocarpa (Greene) G. B. Ownbey, of southwestern peninsular Florida, where perhaps introduced from Key West (Ownbey), is prickly only on the margins of the leaves. It is in- frequently collected, and the chromosome number is unreported. <Arge- mone albiflora Hornem. (A. alba of authors), with white petals, occasion- ally polypetalous, is reported from Florida to Texas and discontinuously northward to Illinois and Connecticut. Subspecies albiflora, 2n = 28, with oblong to oblong-elliptic capsules and uniform, evenly spaced spines, pri- marily of the Coastal Plain, is perhaps native only to northern Florida and introduced northward; nian texana G. B. Ownbey, 2” = 28, with nar- rowly elliptical capsules and unequal spines and prickles, is native only to eastern Texas and is introduced in Arkansas. The two subspecies are com- pletely interfertile or nearly so (Ownbey).

The degree of prickliness, often inconsistent and difficult to describe, is apparently of diagnostic importance. Interspecific hybrids, possible be- tween many species, can, if the original parents were of the same polyploid level, be carried beyond the first generation, but show a decline in vigor and fertility. Sterility barriers probably have not played the primary role in speciation (Ownbey). Chromosome numbers of 2n = 28, 56, 84, and ca. 112 have been reported.

The characteristic hooded sepals terminated by a harsh prickle are rarely found outside of Argemone. As in Arctomecon Torr. & Frém., Papaver, Roemeria, Stylomecon G. Tayl., some species of Mecano pss Vig., and usually in Romneya Harv., the median carpellary traces are lack- ing, the fruits are at least three-valved, and the continuous margins of adjacent stigmas become conspicuous over the piacentae.

REFERENCES: U family references see GRAY, MAHESHWARI & Kanta, and SLAvIK &

SLAVIKOVA (VII, XIV).

Bort, H.-G., & H. FLentje. Alkaloide aus Argemone platyceras. Naturwissen- schaften 47: 323. 1960

Bose, P. K. Cytological studies of Argemone mexicana Linn. Jour. Indian Bot. Soc. 16: 197-208. pls. 12, 13. 1937. [Reports normal 8-nucleate embryo sac, ovules anatropous with 2 integuments, mature pollen uninucleate; suggests polyembryony in Glaucium. |

CuHopra, R. N., & K. S. Rar. Response of ovules of Argemone mexicana L. to colchicine treatment in vivo. Phytomorphology 8: 107-113. 1958. [Some seeds with 2 embryos, the smaller possibly haploid. |

Crete, P. A propos de l’embryologie de l’Argemone mexicana L. Phytomor- phology 6: 144-148. :

Curtis, W. Argemone mexicana. Bot. Mag. 7: pl. 243. 1793.

330 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIII

GiraAL, F., & A. SoreLo. Alcaloides de Argemone ochroleuca. Ciencia 19: 67, 68. 1959

Hakim, S. A. E. Extraction and detection of poppy alkaloids. Jour. Physiol. 138: 8P. 1957. | Argemone. |

———. Poppy alkaloids and glaucoma. J/bid. 40P, 41P. | Argemone. |

eee G. S., & W. G. Rose. Mexican prickly poppy seed oil. Oil Soap 20(3): 33- 35. 1943.*

con. S. kK. Natural hybrid between Argemone mexicana and A. ochroleuca. Curr. Sci. Bangalore 29: 282. 1960.*

Ownbsry, G. B. Monograph of the genus Argemone for North America and the West Indies. Mem. Torrey Bot. Club 21: 1-59. 1958. | Illustrations, maps, and new chromsome numbers. |

. The genus Argemone in South America and Hawaii. Brittonia 13: 91- 109, 1961.

Pratn, D. An account of the genus Argemone, Jour. Bot. London 33: 129-135, 176-178, 207-209, 307-312, 325-333, 363-371. 1895. [Monograph; the first section a history. Footnote, p. 134: “. . . organically the placentas [sic! = stigmas] are in Papaver, as in every other Papaveraceous genus, alternate with the placentas.”

SACHAR, R. C. The embryogeny of Argmone mexicana L.—a criticism. Phyto- morphology 6: 148-151. 1956. |The important paper and reference guide

a controversy with Crété. | Sims, J. Argemone albiflora Hornem. Bot. Mag. 49: pl. 2342. 1822.

6. Papaver Linnaeus, Sp. Pl. 1: 506. 1753; Gen. Pl. ed. 5. 224. 1754.

Annual [biennial or perennial] herbs, usually with wae sap and pu- bescent with multicellular-multiseriate [rarely prickly] hairs. Leaves rosulate and/or cauline, pinnately veined and lobed or dissected, usually without a distinct petiole. Buds nodding |or erect]; flowers on long peduncles subtended by a leaf or bract [or rarely sessile between a pair of foliaceous bracts on a scape, or in a cymose spike]. Sepals 2 or 3, somewhat hooded. Corolla 2-seriate, with twice as many petals as sepals (occasionally polypetalous) white to purple or scarlet [or yellow], often with a dark spot toward the base. Stamens many; filaments filiform, often constricted below the elliptical or oblong anthers; pollen often 3- colpate |or seldom with several apertures]. Stigmatic furrows, usually more than 3, appearing radiate, embedded in a flattened or + conical, cartilaginous, disclike structure [or disc rarely absent]; placentae as many as stigmas. Fruit globose to clavate, truncate, usually more than 3-valved; dehiscence basipetal and incomplete. Seeds numerous, some- what reniform or obovoid-falcate, usually reticulate. Seedling cotyledons + linear. Lectotype species: P. somniferum L.: see Britton & Brown, Illus. FI. No. U.S. ed. 2. 2: 136. 1913. (Name oe ae poppy, ap- parently derived from papa, pap, thick milk.) — Po

At least 40 species of the Northern Hemisphere, including in the New World the Arctic, the Rocky Mountains, and western California; also reported locally in South Africa and Australia. Representatives of three

1962] ERNST, PAPAVERACEAE & FUMARIACEAE 331

of nine sections recognized by Fedde, adventive from Europe or garden escapes, are to be expected in our area. These include Papaver Rhoeas L. (§ OrtHorHorapEes Fedde), 2” = 14, or its cultivated derivative, “Shirley Poppy,” with deeply dissected, pubescent leaves and smooth fruits about as wide as long; P. dubium L., 2n = 42, resembling P. Rhoeas, with fruits about twice as long as wide; and er comm ierumn (S$ PAPAVER, § Mecones Bernh.), opium poppy, 2” = 22, with smooth, globose fruits and usually glabrous, broad, clasping, shallowly dissected or incised Sate Papaver Argemone L. (§ ARGEMONORHOEADES Fedde), 2n = 42, clavate, sparsely pubescent fruits, has been collected immediately to oe north of our area.

Several weedy species of Papaver often associated with the cultivation of cereal crops may inhabit the same or nearby habitats in Britain and have common insect visitors, but interspecific hybrids are rarely found. The relatively self- incompatible diploid, P. RAoezs, and the relatively self- compatible hexaploid, P. dubium, can be hy ‘pridized experimentally in either direction, the hybrids showing anomalous variation and reduced fertility (McNaughton & Harper); bees sometimes discriminate between these species.

The absence of a style and the more than three stigmatic furrows (i.e. the margins of adjacent deltoid stigmas) usually embedded in a disc characterize the genus. In most species of Papaver, as in those of Roemeria, in Meconopsis nirics (L.) Vig., and in the monotypic Séylomecon, the median carpellary traces are absent and replaced by pseudomedian traces originating from the placental bundles; the vascular bundles of the pedicels of some species are + scattered.

Chromosome numbers of 2” = 12, 14, 15, 21, 22, 28, 40, 42, 44. 56, 70, and 84 have been reported.

REFERENCES:

Under family references see ARBER and Popov. Many references from the large pee have been omitted here.

AcHESON, R. M., J. L. Harper, & I. H. McNaveHTon. Distribution of antho- cyanin oe in poppies. Nature 178: 1783, 1284.

Anprews, A. C. The opium poppy as a food and spice in the necical period. Act ish 2674152—15 5. 1952,

BocpasHevskata, O. V. Biosynthesis of morphine in poppy (Papaver somnif- erum) and its movement in the plant. (In Russian.) Fiziol. Rast. 5(2): 185-187. 1958.*

Bort, H.-G., & H. Eumxe. Alkaloide von Papaver paeoniflorum. Naturwis- senschaften 45: 315. 1958.

Cie1j, G. Beet vellows in poppy. Euphytica 10: 225-228. 1961. [Beet virus, transmitted by aphids, also infecting P. sommniferum, controllable by s temic eta

CoMMISSION INTERNATIONALE DES INDUSTRIES AGRICOLES. Bibliographie sur la morphologie et l’histologie de certaines variétés de pavots (Papaver). Commiss. Int. Ind. Agr. Bibliogr. B/2570. 6 pp. 1954.*

Danert, S. Zur systematik von Papaver somniferum L,. Kulturpflanze 6: 61-88. 1958. [Describes 52 Varietéten divided into 4 Convarietaten.

332 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLUI

Faperce, A. C. Genetics of the Scapiflora section of Papaver. III. Interspecific hybrids and genetic homology. Jour. Genet. 46: 125-149. pl. 3.

FAIRBAIRN, J. W., & L. D. Kapoor. 5 7 ase vessels of Papaver somnif-

Sri. Planta Med. 8(1): 49-61.

Harve, J. L., & I. H. McNaucuron. cs inheritance of dormancy in inter- and intraspecific hybrids of Papaver. Heredity 15: 315-320. 1960.

HasirscHKA, G. Bildung von Chromosomenbindeln nach Art der Speichel- driisenchromosomen, i igs es Ruhekernchromosomen und andere Struk- ear aipeneanchieiian n den endopolyploiden Riesenkernen der Antipoden von Papaver Rhoeas. nee 8: 87-113. 1956.

Hiuis, kK. L., & C. N. RopwEL_. The recombination of some varietal characters in the opium poppy. Austral. Jour. Agr. Res. 1: 118-131. 1950. | Quality and quantity of opium—morphine content: some plants self-compatible. |

Hrisui, N. J. Cytogenetical studies on Papaver somniferum L. and Papaver setigerum DC. and their hybrid. Genetica 31: 1-130.

Iv’ina, G. M. Embryological investigations of Papaver somuniterum L. (In Rusern English summary.) Bull. Mosk. Obshch. Isp. Pri. Biol. II. 66(4): 13-25.

Kant, K. Take -ovarian pollination in Papaver Rhoeas L. Nature 188: 683, 634. 1960.

Karp, M. L. On the occasional absence of opium in the opium poppy. Compt. Rend. Acad. Sci. URSS II. 52: 537-540. 1946.

KawatTanlI, T., & H. ASAHINA. External characters and alkaloids of the arti- ficial interspecific F, hybrid between Papaver orientale L. (2) and P., somniferum L. (6). Jap. Jour. Genet. 34: 353-362. :

KLEINSCHMIDT, G., & K. Motues. Zur Physiologie und Biosynthese der Alka- loide von Papaver somniferum, Zeitschr. Naturf. 14b(1): 52-56. 1959.*

KNABEN, G. On the evolution of the Radicatum-group of the Scapiflora Papavers as studied 7 70 and 56 chromosome species. B. Experimental studies. Op. Bot. 3(3): 1-98. pls. 1-6. 1959.

Lisovsku, G. r Anomalies in the structure of the perianth in Papaver somnif- erum L. (In Russian.) Bot. Zhur. 42: 1282-1285. 1957.

Love, A. Cytotaxonomical remarks on the Icelandic Papaver. Nytt. Mag. Bot.

-18.

Tee I. H. Internal breeding barriers in Papaver. Scott. Pl. Breed. Sta. Rep. 1960: 76-84. 196?. [Chromosome numbers, some interspecific hybrids, and pairing patterns of chromosomes in some hybrids

. L. Harper. The comparative biology of closely related species living in the same area. I. External breeding-barriers between Papaver species. New Phytol. 59: 15-26. 1960; II. Aberrant aaa and virus- like syndrome in hybrids between Papaver Rhoeas L. and P. dubium L. Ibid. 27-41; III. The nature of ee Page sympatric populations of Papaver dubium and P. Lecoqit. Ibid. ts

MauesHwarl, N., & M. Lav. In vitro sa of excised ovules of Papaver somiferim L. Phytomorphology 11: 307-314. 1961.

MIKA, Studies on the growth and development and morphine content of opium ee Bot. Gaz. 116: 323-339. 1955.

Nanpa, K. K. Some observations on the emergence, growth and flowering of branches i in Papaver Rhoeas L. Phyton Buenos Aires 16(1): 27-43. 1961.*

SARKANY, S., & E. Percs. Histogenetical observations in the stem tip of ned somniferum L. Acta Biol. Acad. Sci. Hungar. 7(2/3): 183-201.

1962] ERNST, PAPAVERACEAE & FUMARIACEAE 333

ScHIJFSMA, L., M. Hoespercen, & F. E. Nrjpam. A study of the colour and other characters of the seed of some varieties of oil seed poppy. Euphytica 9: 127-140. ; Sosa-BourpouiL, C. Sur le chimisme des étamines de Pavot. Différences entre lignées fertiles et lignées stériles. Compt. Rend. Acad. Sci. Paris 223: 43- 45, 1946

STERMITZ, F. Re & H. Rapoport. The biosynthesis of opium alkaloids. Alka- loid interconversions in Papaver somniferum and P. orientale. Jour. Am. Chem. Soc. 83: 4045-4050. 1961

True, R. H., & W. W. SrockBerceER. Physiological observations on alkaloids, latex and oxidases in Papaver somniferum, Am. Jour. Bot. 3: 1-11. 1916.

Vototov, E. N. The effect of polyploidy upon the tapetum in poppy. Compt. Rend. Acad. Sci. URSS II. 53: 561-564. 1946.

WiLpeMAN, E. be. La variabilité du nombre de stigmates chez Papaver Rhoeas L. Bull. Acad. Sci. Bruxelles V. 31: 459-470. 1946.

ZaiTsEvA, A. A. On the latex of the opium poppy. (In Russian.) Agrobiologiia 3: 122-126. 1951.*

FUMARIACEAE A. P. de Candolle, Reg. Veg. Syst. Nat. 2: 105. 1821, nom. cons.

(FuMIToRY FAMILY)

Annual, biennial, or perennial, sometimes scandent herbs from taproots, bulblets, or rhizomes [or tubers]; plants usually glaucous, rarely pu- bescent with unicellular [or multicellular] hairs. Leaves rosulate and/or cauline, usually alternate, reduced upward, exstipulate, petiolate, or sessile, pinnately or ternately veined, lobed, or dissected. Inflorescences terminal and/or axillary, determinate or indeterminate, cymose or race- mose, scapose or branched, few [rarely 1| to many flowered; buds erect or nodding, bracteate. Flowers erect to nodding, bisexual, hypogynous, usually + closed by the apical connation of the inner petals (sometimes cleistogamous), bilaterally symmetrical or irregular; perianth cyclic, 3- seriate, 2-merous. Sepals 2, bractlike, often + peltate, with + similar margins, not inclosing the buds at anthesis [seldom persistent]. Petals 4, in 2 series, imbricate (very rarely sympetalous) ; outer petals alternate with the sepals, 1 or both [rarely neither] + saccate basally; inner petals opposite the sepals, characteristically connate over the stigmas |or seldom free]. Stamens 6, diadelphous, the phalanges alternate with the sepals, sometimes spurred basally and/or nectariferous; anthers di- morphic, 2- and 1-locular at anthesis, applied to the stigmatic region [or stamens 4, free, anthers 2-locular|]; pollen 3-colpate or several-rugate [rarely 2-colpate or polyporate]. Gynoecium syncarpous, of 2 carpels; style narrow, persistent |or articulated and deciduous], usually flattened apically in the nonplacental plane into a 2- to several-lobed structure bearing the stigmatic surface [or seldom attenuated into 2 simple stig- mas]; ovary 1-locular [rarely occluded]; placentae parietal, ovules few to many, 2-integumented, usually campylotropous, lenticular and + reniform [rarely anatropous and obovoid|. Fruit 2-valved [seldom in-

334 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIII

flated|, the inner wall of valves often transversely ribbed; dehiscence usually incomplete, acropetal or basipetal, the placentae persistent [or fruit indehiscent, or very rarely dimorphic]. Seeds 1—many, free [or fruit rarely fragmenting into 1-seeded joints], + flattened, reniform or orbicular [rarely obovoid], often lustrous, arillate, and attractive to ants; endosperm present; embryo small, straight or somewhat curved. Seedling cotyledons 2 [1], often petiolate (seldom + linear-lanceolate). (Papa- veraceae subfam. Fumarioideae Endl. Gen. Pl. 858. 1839, ““Fumariaceae”’ ; including Hypecoum L. and Pteridophyllum Sieb, & Zucc.) TYPE GENUS: Fumaria L.

About 19 genera and perhaps 425 species, primarily of the Northern Hemisphere, concentrated in Asia where the species are poorly known, oc- curring also in Africa. The family has not been monographed in detail.

Of the three subfamilies recognized on the basis of floral structure, the monotypic Pteridophylloideae Murbeck (restricted to Japan), and the monogeneric Hypecoideae Prantl & Kiindig (Eurasian), are distinguished by free petals and four free stamens. The genera of subfam. Fumarioideae, characterized by diadelphous stamens applied to the stigmatic mechanism and the apical connation of the inner petals, fall into two tribes. The Corydaleae Reichenb., of worldwide distribution in the Northern Hemis- phere, represented in aie area by indigenous species of Dicentra, Adlumia, and Corydalis have persistent, nonarticulated styles; the more specialized Fumarieae, of limited Old World distribution, represented in our area by at least one species of Fumaria, have articulated, deciduous styles. The character of the styles seems more important than either the dehis- cence or indehiscence of the fruits or the number of seeds in the fruits, the criteria by which the two tribes are usually separated. Some of the Old World species now referred to Corydalis have deciduous styles, suggesting that their taxonomic status should be reconsidered.

The Fumariaceae, with some medicinal and poisonous properties claimed, appear to be as rich in alkaloids as the Papaveraceae, sharing about ten per cent of their approximately 140 known alkaloids (Willaman & Schu- bert). Of their common alkaloids, eight are reported only for these two families, two (with identical empirical formulas) are found in six other families, and three (with identical empirical formulas) are found in four other families. The sap of the Fumariaceae, somewhat acrid, watery, and apparently colorless, is contained in nonarticulated sacs; some authors consider the family devoid of laticifers.

The Fumariaceae and Papaveraceae share bracteate inflorescences, two- seriate corollas of twice as many petals as sepals, and similar two-valved fruits, the latter most easily understood as the equivalent of two carpels; however, the patterns of the vascular traces to the perianths and the nature of the sepals, stamens, and seeds usually are different. Hutchinson has suggested that the Fumariaceae, nearly as closely allied with some Berberidaceae as Papaveraceae, probably have not arisen directly from the

1962] ERNST, PAPAVERACEAE & FUMARIACEAE SE)

ancestors of the present Papaveraceae. The least specialized Fumariaceae do not closely resemble Berberidaceae, however.

Chromosome numbers reported for Fumariaceae include 2n = 12, 14, 16, 18, 20, 24, 28, 32, 48, 56, 64, and ca. 150; somatic numbers of 18 and ca. 150 are exceptional. Gametic numbers of 8 and multiples thereof are common, while 6 and 7, or multiples thereof, are rare or have not been verified in recent years (cf. Papaveraceae); only one report of 2m = 28 is verified (Ryberg).

REFERENCES:

Under Papaveraceae see BANCROFT, BERNHARDI, BERSIER & BoQueET, BERSIL- LON, Eames (1961), Eames & WiLson, FEeppe (1905, 1909, 1936), LEGER, MurBECK, PAYER, RODER, SAGDULLAJEVA, SAUNDERS, and WILLAMAN & ScHU- BERT.

ARBER, A. Studies in floral morphology. III. On the Fumarioideae, with special

reference to the androecium. New Phytol. 30: 317-354. 1931. [Analysis of fundamental structure of the fumariaceous flower; includes Corydalis, Dicentra, and Fumaria.| IV. On the Hypecoideae, with special reference to the androecium. /bid. 31: 145-173. 1932. [Establishes morphological rela- tionship of Hypecoum to the Fumarioideae. |

Ernst, W. R. On the family status of the Fumariaceae. (Abstr.) Am. Jour. Bot. 48: 546. 1961.

Gray, A. Ordo Fumariaceae. Gen. Pl. U. S. 1: 117-124. pls. 50-52. 1848. [Dicentra, Adlumia, Corydalis. |

HvuTcHINson, J. e genera of the Fumariaceae and their distribution. Kew Bulle 1921 Obs. Loo:

KLOIMWIEDER, R. Beitrage zur Kenntnis der Schlauchzellen der Fumariaceen, speziell der aes Dicentra s.). Akad. Wiss. Wien Sitz-ber. 138: 517- 55.0)

LIGNIER, O. a fleur des Crucifeéres comparée a celle des Fumariées. Assoc. Fr. Avanc. Sci. Compt. Rend. 25(1): 403-406. 1897.

MartTEL, E, ane sur le diagramme floral des Cruciféres et des Fumariacées. Jour. Bot. Morot 12: 29-31. 1898

RysBerc, M. A morphological study of the Fumariaceae and the taxonomic sig- nificance of the characters examined. Acta Horti Berg. 19: 121-248. pls. I-12. 1960. [Indispensable survey of the family; includes chromosome numbers and extensive bibliography. |

SLAVIK, J., & L. Sztavixova. Alkaloide der Mohngewdchse (Papaveraceae). XVIII. Alkaloide aus Hypecoum procumbens a und H. eee Hook. f. et Thoms. Collect. Czech. Chem. Commun. 26: 1472-147 =

SouEcEs, R. L’embryon chez le Corydalis cheilanthifoli Hemsl. Les Fumaria cées et la classification embryogénique. Ann. Sci. Nat. Bot. 17 1946. [Comparisons with Medicago, Fumaria, and Bip couny

KEY TO THE GENERA OF FUMARIACEAE

General characters: plants herbaceous, with watery sap, sometimes rhizomatous or with bulblets (rarely scandent) ; leaves rosulate or alternate, glabrous. In- florescence terminal or axillary, bracteate, racemose or cymose; flowers bi- sexual, hypogynous, bilaterally symmetrical or irregular; perianth cyclic, 2- merous; sepals not inclosing the buds at anthesis; petals 4, two-seriate (very

336 JOURNAL OF THE ARNOLD ARBORETUM [ VOL, XLIII

rarely sympetalous), 1 or both outer petals + saccate basally, the inner petals

connate apically; stamens 6, diadelphous, 1 or both median stamens basally

spurred and/or nectariferous; anthers applied to the lobed and flattened stig- matic mechanism; gynoecium 1-locular with 2 parietal placentae ; fruit 2-valved

[rarely ndehiscent|,

A. Flowers bilaterally symmetrical, nodding, usually bracteolate and bracteate; petals usually + free (or sympetalous); outer petals identical; inner petals symmetrical.

B. Plants acaulescent or rhizomatous; petals + free basally; stamen fila- ments usually + partially free basally; valves of fruit em basi- petally, only the median vein persistent apically. .

Plants caulescent and scandent; petals united through half ae length,

spongy and persistent; stamen filaments united basally and adnate to

corolla; valves of fruit dehiscent acropetally, free at the base. 2. Adlumia.

A. Flowers irregular, variously ascending or descending, bracteate; petals + free at base (or adnate to the stamens); outer petals dissimilar, the upper petal conspicuously saccate, the lower sometimes gibbous; inner petals ages but asymmetrical.

. Style persistent; fruit elongate, several seeded, dehiscent + bi are

<3

BnTAScOY DIGMMIALS:. d.2 6.6 cod Kaede andwit yaaa aay ee weeaee 3. Corydalis. C. sa articulated and dehiscent; fruit globose, hard, 1-seeded, indehiscent ; AP vente MRNAS on Sica ca we Rede pe hee eet a eames 4, Fumaria.

Tribe CoryDALEAE Reichenb.

1. Dicentra Bernhardi, Linnaea 8: 457, 468. 1833. nom. cons.

Perennial [sometimes scandent| herbs from rhizomes, bulblets [tubers or taproots], glabrous [or very rarely with a few multicellular-multiseriate hairs] and occasionally glaucous. Leaves few to many, rosulate [or cauline and alternate], pinnately or ternately dissected |petioles very rarely sub- terranean and absorptive]. Inflorescence terminal, simple or branched {sometimes axillary or leaf-opposed], determinate [or rarely indeter- minate]|, cymose or racemose [very rarely with solitary, 1-flowered scapes |. Flowers usually nodding [or seldom erect], bilaterally symmetrical, usually bracteolate as well as bracteate. Sepals ovate to lanceolate, non- peltate. Outer petals identical, + saccate basally and reflexed apically; inner petals symmetrical and identical, + spatulate, externally crested and apically connate. Stamens in 2 similar phalanges with filaments fused or usually partly free basally; median stamens spurred basally and/or nec- tariferous; pollen 3-colpate or 12-rugate. Style narrow, persistent, flattened apically into a lobed stigmatic structure. Fruit nodding [or seldom erect], 2-valved, elongate, dehiscence basipetal, incomplete basally, the median caepellaey vein persistent apically [or dehiscence acropetal, free at the base and incomplete apically|. Seeds few to many, reniform-suborbicular, flattened, usually lustrous and + arillate. (Bikukulla Adans. |= Bicu- culla of authors], nom. rejic.) Type species: D, Cucullaria (L.) Bernh. (Name from Greek, dis, twice, and kentron, a point.) — BLEEDING HEART.

1962] ERNST, PAPAVERACEAE & FUMARIACEAE 337

About 19 species and three subgenera distributed in North America and eastern Asia. Subgenus DactyLicapNnos (Wall.) K. R. Stern, is scandent and confined to northern India and adjacent China; CHrysocaPpNos Engelm. occurs in California and eastern Asia; DicENTRA, with one species in eastern Asia and several in North America, is represented in our area by three species. Dicentra Cucullaria, Dutchman’s breeches, 2n = 32, occurring from South Carolina to Arkansas, north to Minnesota and Nova Scotia, and discontinuously in the Pacific Northwest, is distinguished by white to pink, asymmetrical bulblets and basally fused stamen filaments with the nectariferous spur directed diagonally; D. canadensis (Goldie) Walp., squirrel-corn or turkey-corn, 2” = ca. 64, flowering somewhat earlier and sympatric almost throughout the eastern distribution of the above, has yellow, pea-shaped bulblets and filaments partially free basally with the short nectariferous spur oriented vertically. These two, seem- ingly most closely related to each other, have similar scapose, racemose inflorescences terminating blindly in a linear lobe, an aborted bud, or a flower. A strictly racemose inflorescence, also occurs in D. spectabilis (L.) Lem., 2n = 16, of Asia. Dicentra eximia (Ker) Torr., bleeding heart, 27 = 16, with cymose, usually branched inflorescences, strongly allied with D. formosa (Haw.) Walp., 2n = ?24, of the Pacific States, is distributed from eastern Tennessee and western North Carolina and northward to New Jersey and Pennsylvania.

The flowers of Dicentra, as in most Fumariaceae, are somewhat tubular and closed by the apical connation of the inner petals over the anthers and stigmas. Although abundant pollen is shed directly on the stigmas and adheres to them, the germinating pollen tubes barely penetrate the stigmatic tissue in a number of species (Stern, 1961). In some, ovaries abort if protected from the visits of insects; bees sometimes gain direct access to the nectaries by biting through the base of the corolla.

REFERENCES:

Under family references see GRAY; under Papaveraceae see GATES.

Camp, W. H. Dicentra Cucullaria. Addisonia 20: 21, 22. pl. 651. 1937. Hooker, W. J. Dielytra canadensis. Bot. Mag. 57: pl. 3031. 1830. [Dicentra

MEEHAN, T. Dicentra Cucullaria. Meehans’ Monthly 6: 41, 42. 1896. [Brief

history. |

Roginson, C. Flowers and insects. I. Bot. Gaz. 14: 120-126. 1889. [Insect visitors of D, Cucullaria and biting of holes through corolla. ]

Rypperc, P. A. Bicuculla eximia. Addisonia 14: 53, 54. pl. 474. 1930. [D. eximia. |

Stms, J. Fumaria Cucullaria. Bot. Mag. 28: pl. 1127. 1808. [D. Cucullaria.]

Seay K. R. Revision of Dicentra Ree Brittonia 13: 1-57, 1961. [Includes illustrations, distribution maps, and review of alkaloidal prop- erties with literature citations, 10-12; treats meee as a subgenus of Dicentra. See pollen, Am. Jour. Ree 49: 362-368. 1962.]

338 JOURNAL OF THE ARNOLD ARBORETUM [ VOL, XLIII 2. Adlumia Rafinesque ex A. P. de Candolle, Reg. Veg. Syst. Nat. 2:

¢

111. 1821, nom. cons.*

Delicate, taprooted biennial vine usually with a single, indeterminate leader. eave: pinnately divided, at first rosulate, then cauline and al- ternate, the petiolules + opposite, distally tendriliform. Inflorescences many, | or 2 ina leaf axil, cymose, often branched, usually with many nod- ding, bracteate flowers. Sepals somewhat peltate. Corolla sympetalous, white to purplish, persistent and spongy, bilaterally symmetrical, basally + saccate alternate with the sepals, apically with 2 reflexed outer lobes, and 2 connate inner lobes. Stamens 6, diadelphous, but basally + mon- adelphous and united with the corolla; median stamens nectariferous basally but not spurred; pollen 3-colpate or 6-rugate [9—12-rugate]. Style narrow and persistent, terminating in a flattened, + 2-lobed stigmatic re- gion. Fruits + linear, 2-valved, dehiscence acropetal, the valves free basally. Seeds few, flattened. suborbicular, lustrous, nonanilate: Seedling cotyledons 2, petiolate. TyPE SPECIES: A. cirhosa Raf. ex A. fungosa (Ait.) Greene ex BSP. (Named for John Adlum, 1759-1836, sol- dier of the Revolution, major under the elder Adams, later brigadier gen- eral of the Pennsylvania Militia; enthusiastic advocate of American viti- culture. )

A genus, probably monotypic, of eastern North America and Korea. Adlumia fungosa, Allegheny vine, mountain fringe, or climbing fumitory,

(1904, p. 343), who stated that his specimens, insofar as he could de- termine, were identical to the American species, a conclusion substantiated by his description. Ohwi (1931, p. 387), described A. asiatica, but the description and plate are indistinguishable from the American plant. The name A, Komarovii was used by Popov (1937, posthumous The unique, sympetalous corolla of Adlumia may culminate a trend toward the tubular condition evident in the scandent species of Dicentra and in the analogous, long-spurred petals of some species of Corydalis. The affinities of Adlumia are with Dicentra subg. DactyLicapnos, which also has tendrilliform leaflets, scandent growth, and acropetal dehiscence of the fruits, the valves becoming free basally. Current growth of other species of Dicentra is not scandent, lacks tendrilliform leaflets, and is terminated by an inflorescence; and dehiscence of the fruits is basipetal, the valves re- maining attached apically only by the filamentous median carpellary vein and basally where no abscission zone forms. Stern favored the retention both of Adlumia (indeterminate-monopodial growth with axillary, cymose inflorescences, sympetalous corolla and nectariferous but nonspurred median stamen filaments) as a genus and of DactyLicapNnos (indetermi- mia was unnecessarily conserved over Bicuculla Borckh., a later homonym of Bikukulla Adans. which is equal to Fumaria Cucullaria L. and now equivalent to the type species of the conserved Dicentra Bernh. (See Rickett & Stafleu, Taxon 8: 274, 1959.)

1962 | ERNST, PAPAVERACEAE & FUMARIACEAE 339

nate-sympodial growth with leaf-opposed, racemose inflorescences, apopet-

alous corollas, elaborately spurred and nectariferous median stamen fila-

ments) as a Subgenus of Dicentra.

REFERENCES:

Under family references see GRAY.

Britton, E. G. Adlumia fungosa, Addisonia 5: 21, 22. pl. 171. 1920. [Re- see garden plants annual, attaining a length a 10 feet.

Komaroy, V. L. Baas Zn: Fl. Manshuriae. vol. 2. (In Russian.) cere Horti Petrop. 22: 337-352. 1904. [Reports discovery in 1897 in Korea of Adlumia, a and, as far as could be determined, identical to the American

Onwt., J. Adlumia gee sp. nov. Zz: Symbolae ad Floram Asiae Orientalis Il n Latin.) Bot. Mag. Tokyo 45: 387. 1931. ne the sp. for Korea, (2) biennial; does not seem to differ from A. fungos

Popov, M. G. Papaveraceae. In: Fl. URSS 7: 573-717. 1937. (In Russian.) [A. asiatica Ohwi given as perennial, 647; pl. 45, fig. 6

3. Corydalis Ventenat, Choix de Plantes. pi. 19. 1803, nom. cons.!°

Annual, biennial [or perennial, very rarely scandent] herbs, from tap- roots [tubers or rhizomes], often glaucous, usually glabrous (or very rarely with 1-celled hairs). Leaves variously pinnately or ternately dis- sected, rosulate or cauline, alternate, petiolate toward the base. Inflores- cence terminal, simple or branched, few to many flowered [or scapose] usually racemose (or very rarely cymose); flowers irregular, often horizontally oriented, bracteate. Sepals ovate, + peltate. Outer petals dissimilar, the upper conspicuously saccate, the lower sometimes + gib- bous; inner petals similar but asymmetrical, connate apically. Stamen filaments fused in 2 dissimilar phalanges, the upper median stamen spurred and nectariferous; pollen 3-colpate or 6—12-rugate. Gynoecium elongate, usually upwardly curved; style persistent [or seldom articulated and deciduous], apically flattened into a lobed stigmatic structure. Fruit erect or nodding [sometimes geniculate at the receptacle], usually elongate, 2-valved, sometimes falcate, often torulose [or very rarely inflated], dehiscence + acropetal, the valves free at the base and often transversely ribbed between the seeds. Seeds few to many, flattened, reniform or + orbicular, lustrous, and usually arillate. Seedling cotyledons 2 [or 1], usually long-petiolate. Lecrorype species: C. sempervirens (L.) Pers. (Name from Greek, korydos, or korydaillis, the crested lark of Greece, and presumably the name of a plant with a corolla spur resembling the spur of the lark.)

* When first conserved, Corydalis was attributed to Medicus who applied the name to Fumaria vesicaria L., now Cysticapnos vesicarius (L.) Fedde. To preserve the modern concept of Corydalis, the authority was changed to Ventenat who stated in a

ournefort and Gaertner. Tournefort cited Fumaria siliquosa sempervirens of Corn uti, presumably now equivalent to Corydalis sempervirens oe Pers., the lectotype species. Gaertner presumably illustrated Corydalis lutea (L.) D

340 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLUI

Fic. 3. Corydalis. a-k, C. sempervirens: a, flowering and fruiting shoot and basal leaf, X Y%: b, flower, lateral view — note peltate sepal, X 4; c tip of

stamens, the lower two-locular, the two upper one-locular, * 10; h, gynoecium

and pedicel, lateral view, X 45 i i, lobed stigmatic region, lateral view, Oe 4

fruit after dehiscence, show z persistent style, placentae, and attachment. of seeds, X 1%; k, arillate seeds the micropyle at left edge below aril, * 20

A genus of the Northern Hemisphere, probably of fewer than 300 species, concentrated in south-central Asia (where poorly known), with one species in East Africa, variously segregated without agreement into several sections and subsections.

1962] ERNST, PAPAVERACEAE & FUMARIACEAE 341

In our area, five taxa of sect. Coryparis (§ encom aa Prantl) are widely distributed (Ownbey). Corydalis sempervirens, 2n 16, pink or pale corydalis, rock fumeroot, or roke harlequin, a biennial ec anpated from northern Georgia to Newfoundland and central Alaska, tentatively re- ported from Siberia (Fl. USSR.), is distinguished by its branched, leafy shoots, pinkish flowers, and apparently unique cymose inflorescences. Four winter annuals bearing a general resemblance to C. aurea Willd., 2n = 16, the latter widespread to the west and north of our area, were defined by Ownbey (1947): C. crystallina Engelm., with conspicuously pubescent fruits, occurring in Arkansas and the states immediately west and north; C. micrantha (Engelm.) Gray subsp. micrantha, 2n = 16, slender fume- wort, with erect fruits, occurring in northwestern Arkansas northward and westward; C. micrantha subsp. australis (Chapm.) G. B. Ownbey, intergrading with subsp. micrantha, usually with + longer racemes and flowers, and longer, narrower fruits, occurring from Arkansas southeast- ward along the Gulf Coast, across Blond and north along the Atlantic Coast to North Carolina; and C. flavula (Raf.) DC., yellow fumewort or yellow harlequin, similar to C. micrantha subsp. australis, with nodding fruits, occurring from Louisiana to Missouri eastward to North Carolina and Connecticut.

The taxon Corydalis micrantha subsp. australis, first described as C. aurea var. australis Chapm., was considered conspecific with C. micrantha (Engelm.) Gray (originally C. aurea var. micrantha Engelm.), and re- described as Capnoides Halei Small (with part of the original material referable to Corydalis crystallina), later Corydalis Halei (Small) Fernald & Schubert.

Both Corydalis micrantha (including subspp. micrantha and australis) and C, flavula are complicated biologically by the occasional occurrence of individual, or whole inflorescences of, cleistogamous flowers, apparently sometimes correlated with certain vegetative peculiarities. Normal flowers in this instance are closed by the permanent apical connation of the inner petals, and the stamens shed pollen on the stigmatic mechanism to which they are appressed. In some species the stigmatic region can be exposed suddenly by probing along the path leading to the nectariferous spur.

Chromosome numbers of 2m = 12, 14, 22, 24, 34, and 36 have been re- ported.

REFERENCES:

Under family references see Gray; under Papaveraceae see FEDDE (1936). BERSILLON, G. Le renouvellement annuel du tubercule de Corydalis solida Sw.; mécanismes histogénétiques. Revue Gén. Bot. 66: 469-488. 1959. Bort, H.-G., & H. Eumxe. Alkaloide aus Corydalis angustifolia und C. de-

cipiens. ere 46: 427. 1959 Curtis, W. Fumaria glauca. Bot. Mag. 5: fl. 179. 1792. [C. sempervirens. ] FERNALD, M. L. TaonaGane and reidentifications of North American plants. Rhodora 48: 207-216. 1946. [History of Corydalis Halei, 207. Gray, A. Corydalis aurea and its allies. Bot. Gaz. 11: 188, 189. 1886. MansrFELp, R. Bemerkungen zur Liste der Nomina Generica Conservanda.

342 . JOURNAL OF THE ARNOLD ARBORETUM [ VOL, XLIII

a German.) Taxon 2: 187-190. 1953. [Ventenat, rather than Medicus, ust be the authority for proper conservation of the name Corydalis. | isan M.R. The toxicity of Corydalis Caseana, Jour. Agr. Res. 42: 239-243. 1931.

Ownsey, G. B. Monograph of the North American species of Corydalis. Ann. Missouri Bot. Gard. 34: 187-259. 1947. [Includes citation of specimens and distribution maps. |

n the cytotaxonomy of the genus Corydalis, section Eucorydalis. Am. Midl. Nat. 45: 184-186. 1951. | Reports from root tips 2” = 16 for C aurea, C. micrantha, and C, sempervirens; identical chromosomal mor-

RYBERG, M. Studies in the taxonomy and fertility of some Scandinavian Cory- dalis species of the sect. Pes-Gallinaceus Irmisch. Acta Horti Berg. 15: 207-284. 1950. [Analysis of variation and breeding behavior in northern European spp. of Corydalis; fall meiosis. |

A taxonomical survey of the genus Corydalis Ventenat with reference to cultivated species. bid. 17: 115-175. pls. eS

A morphological study of Corydalis nobilis, C. cava, C. solida, and

some allied species with special reference to their underground organs. Ibid. 19: 15-119. pls. 1-16. 1959.

SovEcEs, R. Embryogénie des Fumariacées. Les premiers ee du deével- oppement de l’embryon chez le Corydalis lutea DC. Compt. Rend. Acad. Sci. Paris 222: 161-163. 1946. La différenciation des Ras foudamient: ales du corps chez le Corydalis lutea DC. Ibid. 253-255.

SPRAGUE, T. A. The correct spelling of certain generic names — III. Kew Bull.

928: 337-365. 1928. [Review of use and spelling of the name Corydalis. |

TraBerT, C. H., & U. SCHNEIDEWIND. Uber Inhaltsstoffe von Corydalis cava Schweigger und Koerte, 2. Die Untersuchung der Samen. Pharm. Zentralh, Deutsch. 99(1): 2-7. 1960.*

Tribe FUMARIEAE 4, Fumaria Linnaeus, Sp. Pl. 1: 699. 1753; Gen. Pl. ed. 5. 314. 1754.

Low, glabrous, compact or diffuse herbs from taproots. Leaves rosulate and/or cauline, alternate, variously pinnately or ternately dissected |some- times + tendriliform]. Inflorescences many flowered, terminal or axillary, indeterminate, and racemose; flowers irregular, bracteate. Sepals small, peltate-ovate, variously dissected. Outer petals dissimilar, the upper sac- cate at the base; inner petals similar but asymmetrical, connate apically. Stamen filaments fused in 2 dissimilar phalanges, somewhat adnate later- ally with the base of the petals; upper median stamen spurred and _nec- tariferous; pollen polyporate with 6-12 apertures. Style narrow, + ascending, articulated and dehiscent, apically flattened into + 2 lobes; ovary short, + ovate; ovules 4, + reniform, Fruit subhemispheric, often with 2 shallow depressions apically, hard, indehiscent, filled with the single nonarillate seed. Seedling cotyledons linear-lanceolate, nonpetiolate.

YPE SPECIES: F. officinalis L. (Name from Latin, fwmus, smoke, perhaps alluding to the odor of the plant.) — Fumirory, EARTH-SMOKE, FUME- ROOT, FUMEWEED, FUMEWORT.

1962] ERNST, PAPAVERACEAE & FUMARIACEAE 343

A genus of cosmopolitan weeds, concentrated in the Mediterranean region, with many species and infraspecific taxa; represented in our area by Fumaria officinalis, 2n = 32 (also ?14 and Oy occurring from western Florida along the Coastal Plain into North Carolina. Other species, the differences appearing trivial in view of the apparent edaphic plasticity of the genus, may occur in our area. The environment readily modifies the character of the vegetative organs and, under some conditions, the fre- quency of cleistogamous flowers which are paler or whitish in color, re- duced in size and lacking nectaries but producing fruit. Hybrids in the Old World are uncommon, isolated, and sterile (Pugsley). Chromosome numbers of 2” = 14, 28, 32, ca. 48, 56, and 64 have been reported.

REFERENCES:

GreceER, J. Uber Kalkoxalatkristalle in der Samenschale von Fumaria officinalis L. Planta 12: 49-52. 1930 [1931].

Lerort, L. Quelques groupes infra-spécifiques luxembourgeois de Fumaria Di aerttes Ib, Soc. Nat. Luxembourg. Bull. II. 44: 180, 181. 1950 [1951].*

Pucstey, H. W. A revision of the genera Pumaria and Rupicapnos. Jour. Linn. Soc. Bot. 44: 233-355. pls. 9-16. 1919.

SouEGES, R. Embryogénie des Fumariacées. L’origine du corps de l’embryon chez le Fumaria officinalis L. Compt. Rend. Acad. Sci. Paris 213: 528-530. 194]

344 JOURNAL OF THE ARNOLD ARBORETUM [ VOL, XLIII

NOTE ON DAPHNOPSIS CRASSIFOLIA (THYMELAEACEAE)

Lorin I. NEVLING, JR.

AMONG A NUMBER OF SPECIES which were poorly or incompletely known at the time of my revision of the genus Daphnopsis (1959) was D. crasst- folia (Poiret) Meissner. The description of this species necessarily was sketchy and incomplete because of the availability of relatively few and fragmentary specimens. I was not able to see fruiting material, for ex- ample, although it had been described previously. Since my publication, a relatively large number of specimens referable to this species have been rediscovered and examined. On the basis of new data drawn from re- cently examined herbarium material, the description has been recast and completed. In addition, these collections give the first precise knowledge of the range of the species (see Map).

Daphnopsis crassifolia (Poir.) Meissn. in DC. Prodr. 14: 524. 1857.

Daphne crassifolia Poir, Encycl. Méthod. Bot. Suppl. 3: 316. 1813 (Tyee: Nectoux sn. [é]!).

Daphnopsis oe var. eggersii Krug & Urb. Bot. Jahrb. 15: 350. 1892 Type: Eggers 2317 [9 ]).

Hyptiodaphne crassifolia (Poir.) Urb. Symb. Ant. 2: 454. 1901.

Hyptiodaphne crassifolia var. 8. eggersii (Krug & Urb.) Urb. ibid. 455.

Dioecious shrubs or trees, to 4 m. tall, the young branches terete, densely ochraceous-hirsute and glabrescent, rugose. Leaves simple, approximately whorled by irregular condensation of internodes or rarely alternate on very vigorous shoots, usually 3 or 4 per whorl; leaf blades rarely narrowly elliptic to elliptic to oblanceolate to obovate or very reduced and bractlike, the expanded blades 3—6(—13.5) cm. long, 1-3.5 cm. broad, acute to acuminate or rotund and mucronulate at the apex, cuneate to subauriculate at the base, coriaceous, (sparsely) to densely ochraceous-sericeous and

neath, arcuate-ascending and forming a prominent submarginal vein; petiole 2-4 mm. long; exstipulate. Inflorescences borne terminally (but appearing axillary or extra-axillary) from young bracteate stems, umbel- liform to subracemiform, hirsute, the primary peduncle 10-40 mm. long, the rachis 1-3 mm. long, the secondary peduncles 3-7 mm. long. Flowers unisexual, 4-merous, perigynous. Staminate flowers: 2—5 per inflorescence, white; pedicel obsolete to 0.5 mm. long; calyx tube tubular to broadly obconic, 7-10 mm. long, 2—4.5 mm. broad at the orifice, hirsute without, glabrous within; calyx lobes subequal, indefinitely papillate within, 2.5—4 mm. broad; petals connate into an obscure faucal annulus; antisepalous

1962] NEVLING, DAPHNOPSIS CRASSIFOLIA 345

stamens inserted at the orifice, subexserted, the alternisepalous inserted 2-3 anthers’ lengths below the orifice, included, the anthers oblong, about 1 mm. long, 0.5 mm. broad, sessile; disc of irregular lobes, free, 1—3.5 mm. tall, glabrous; pistillode tenpin-shaped, 0.75-1.5 mm. tall, glabrous. Pistil- late flowers: 2 or 3 per inflorescence, white; pedicel 1-1.5 mm. long; calyx tube more or less tubular, 4-8.5 mm. long, 2 mm. broad at the orifice, hirsute without, glabrous within; calyx lobes unequal, indefinitely papillate within, the outer 3—4.5 mm. long, 2.5—3.5 mm. broad, the inner 2.5-3 mm. long, 2 mm. broad; petals connate into an obscure faucal annulus; stami- nodia 8, papilliform, minute; disc of 4 lobes, the lobes irregularly shaped, free, 1-2 mm. tall, glabrous; gynoecium single, tenpin-shaped, 4.5—-5.5 mm. long, the ovary 1-locular, glabrous, the single ovule anatropous, the style ca. 1.5 mm. long, thick, the stigma capitate, obscurely bilobed, in- cluded or very rarely exserted. Drupe ellipsoid, ca. 15 mm. long, 10 mm. in diameter, glabrous.

Endemic to the island of Hispaniola where it is abundant on lateritic soils, according to Ekman. It has been collected along river banks, in thickets and woods from 600 to 2000 m. altitude. Flowering is known from January through May and October; fruiting in April and May. The tough fibrous bark is used, according to Holdridge, for making rope. It is known locally as mahaut (Holdridge) and jayao (Jaquez).

Map of Hispaniola showing distribution of Daphnopsis crassifolia.

Dominican Republic. Azua: Sierra de Ocoa, San José de Ocoa, Loma del Rancho, Ekman H11641 (us, 6). BaraHona: Barahona, Fuertes 891 (a, sterile); Firme Noche Buena, Fuertes 1433 (a, NY, 6), 1498 (a, Ny, US, 2) BENEFACTOR: Piedra del Aguacate, n. of Rio Arriba, Howard & Howard 9436 (A, ?). La VecA: prope Constanza, Von Tiirckheim 3103 (cu, ?; Ny, 6, 9; us, ¢, @). Without precise locality: Pico de Igua, Jaguez 2028 (us, @). Haiti. ARTIBONITE: Massif du Nord, Hinche, Ekman H11641 (Ny, 6); Massif des Bahos, Pte- Rio de l’Artibonite, Perodin, bridge above Ingrand, Ekman H3454 (us, 6). OUEST: vicinity of Mission, Fonds Varettes, Leonard 3905

346 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIII

(GH, Ny, US, 6); Massif de la Selle, Petionville, top of Mt. Tranchant, Ekman H1159 (vs, 9): Massif des Cahas group, Las Caobas [Cahobas], Belladire [ Belladére|, Morne Lagoume-Hire (?), Ekman H35627 (ny, 6); Massif de la Selle. group Créte-a-Piquants. Port-au-Prince, Morne, Bercy, erupt. Ekman H5910 (xy, 6); Gros Cheval, Mornes aes Comnmissaires, Holdridge 951 (GH,

MO, NY, US, : ). Without precise locality: La Revellue (?), collector unknown sm. (C, 2); “Saint-Domingue,’ Nese sn. (P- ae pe of Daphne crassifolia, 3 )

The precise locality at which the holotype of this species was col-

lected by Nectoux is unknown: consequently, it has not been mapped. Nectoux was “Jardinier en chef du Jardin du Roi” (Nectoux, 1791) in what is now Haiti. I have been unable to ascertain to what extent he travelled and collected.

Ekman’s number 11641 is a split collection, part collected in the Dominican Republic and part in Haiti. Both localities have been mapped.

The included stigma is an unusual feature of the pistillate flower of this species. The only exception I have seen, in which the stigma is exserted, is on Von Tiirckheim 3103 (us). In this specimen the gynoecium is of average size, but the calyx tube is relatively short for the species, and this may account for the exsertion of the stigma.

Krug and Urban’s Daphnopsis crassifolia var. eggersi is based on an Eggers specimen (2317) which is larger in respect to petiole length and leaf size than var. crassifolia. They give the following measurements for var. crassifolia: petiole 2 mm. long; leaf blade 3-3.5 cm. long, 1.3-1.7 cm. broad. In contrast, for var. eggersii they give: petiole 3-4 mm. long: leaf blade 5—8 cm. long, 2—3 cm. broad. These are the only contrasts, for var. crassifolia was known from staminate material and var. eggerst only from fruiting material. The measurements of leaf and petiole may be applied successfully to separate Nectoux s.n. (type of var. crassifolia) from a few specimens (the holotype of var. eggersi (Eggers 2317), was destroyed at Berlin and T have not located an isotype); but many speci- mens display enough variability on a single branchlet to overlap sig- nificantly the measurement ranges given for the two varieties. I believe that the leaf variation exemplified by this species is of a phenotypic, rather than of a genotypic nature.

In 1901, when Urban erected the genus Hyptiodaphne to accommodate Daphne crassifolia Poir. (= Daphnopsis crassifolia (Poir.) Meissn.), he indicated as one of the main features of the new genus the orthotropous ovule attached basally in the ovary. Unfortunately, there was (and is) little or no information available concerning ovule position and attach- ment within the genus Daphnopsis except that the ovules are anatropous. Domke has figured (1934, figs. 39a,6,c) several ovules from plants refer- able to Daphnopsis crassifolia which show, quite strikingly, considerable variation in ovule attachment. The ovules, however, are always of the anatropous type. A pistillate flower (Holdridge 951, GH) was cleared and stained to recheck this point. The ovary wall is heavily vascularized with veinlets composed of short and sinuous tracheary elements. The ovular

pee

1962 | NEVLING, DAPHNOPSIS CRASSIFOLIA 347

trace is developed poorly, but the trace and ovule position is nearly identical with that figured by Domke in Fig. 396. If Urban based his ob- servation on sectioned material he may have sectioned the ovule at right angles to the attachment, thus leading to the false impression of an ortho- tropous ovule (demonstrated in Domke’s fig. 39c). Regardless of these difficulties, a slight shift in ovule position alone is not enough to warrant recognition of Hyptiodaphne,

LITERATURE CITED

DomkeE, W. Untersuchungen uber die systematische und geographische Glhede- rung der Thymelaeaceen. Bibliot. Bot. 27(111): 1-151.

Nectoux, M. Observations. Mém. d’agric. d’econ. rurale et domestique 110-123. 1791. | Reference incomplete; seen only as a separa

Neviinc, L. I., Jr. A revision of the genus Dies Ann. Missouri Bot. Gard. 46: 257 —358. 1959.

348 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIII

TWO NEW ASIATIC PANDANACEAE

BENJAMIN C. STONE

DuRING A SHORT visit to the Harvard University Herbarium I was enabled, through the courtesy of the directors, to study the collections of Pandanaceae. Among the specimens studied were two which represent undescribed species, one in the genus Freycinetia, the other in Pandanus. Descriptions of these follow.

Freycinetia kostermansii B. C. Stone, sp. nov. (Sect. FREYCINETIA), (Fie.1);

Planta scandens, caulibus subrobustis subteretibus, plerumque 8-9 mm. crassis, internodiis 5-15 mm. longis; foliis obovato-oblanceolatis, basi breviter contractis, apice abrupte acuminatis breviter caudatis, vulgo 10- 15 cm. longis et 12-28 mm. latis, margine basim versus serrato-dentatis, dentibus ad 2 mm. longis, et apicem versus serrato-denticulatis (dentibus ad 0.5—1 mm. longis), costa media ad apicem et in acumine denticulata; syncarpiis terminalibus cylindricis usque ad 8 cm. longis et 3 cm. latis, pedicellis in angulis et distaliter scabridulis; baccis rubris, 14-16 mm. longis, circiter 4 mm. latis, lageniformibus, apice rostrato-contractis, truncatis (areola stigmatica minima 1-2 mm. lata), parte inferiore succu- lentibus, stigmatibus (5—)6—-9, seminibus circiter 1.6—-1.9 mm. longis line- aribus rhaphide et strophiolo instructis.

Hototype. Moluccas: Morotai, 1000 m. alt., May 27, 1949, A.J. G. H. Koster- mans 1193 (A).

This new species is apparently related to Freycinetia oblanceolata Mar- telli, of New Guinea, and perhaps also to F. Aemsleyi Warb., of Borneo.

Pandanus gressittii B. C. Stone, sp. nov. (Sect. ACROSTIGMA).

P. Gressittii Merr. ex G. F. Sauer, List Pl. Lingnan Univ. Campus 94, 1947, nomen nudum.

Frutex foliis circiter 30-101 cm. longis et (7—) 11-14 mm. latis, margine robuste armatis dentibus antrorsis acicularibus, costa media dorsaliter armatis dentibus robustis antrorsis, apicem versus adpressibus; syn- carpium solitarium, ellipsoideum, circiter 5-7 cm. longum et 3.5-5 cm. diametro; pedunculo erecto bracteato; drupis anguste obclavatis circiter 14 mm. longis et 2-3 mm. latis, apice rotundato-conicis, pileo 2-3 mm. longo; stigmate breviter erecto-curvato, aceroso, circiter 2 mm. longo et 0.75 mm. lato, endocarpio supramediale, semen unicum.

DISTRIBUTION: endemic to Hainan Island.

1962] STONE, NEW ASIATIC PANDANACEAE 349

Fic. 1. FREYCINETIA KOSTERMANSIT: a, leaf. X 0.8; b, ¢, berry, lateral and top views, X 8; d, seed, X 8 and & 35. (Drawn from the hatety pe.)

Ho.Lotype. Hainan Island: small jungle stream at Liamui, 400 m. alt., June- July 1935, J. Linsley Gressitt 1185 (A). ADDITIONAL SPECIMENS EXAMINED. Lokwui, near stream, 15 May 1935, F. C. How 72350 (a); Bo-Ting, in thickets, 17 Oct. 1936, S. K. Law 28026 (A).

The existence of this species was first called to my attention by Dr. J. Linsley Gressitt, who collected the type specimen. The name conferred by Merrill and long carried in manuscript appears never to have been formally published. I take pleasure in bringing into use the name honoring Dr. Gressitt, now Entomologist and Head of the Entomology Depart-

350 JOURNAL OF THE ARNOLD ARBORETUM [ VOL, XLII

ment of the Bernice P. Bishop Museum, Honolulu, whose labors in the two fields of insect taxonomy and zoogeography of the Pacific are of more than casual interest to the botanist.

The new species may be recognized by its shrubby habit; solitary syncarp set on a long or slightly nodding peduncle furnished with bracts; and the unilocular drupes with conoid pilei about 2-3 mm. long produced into a short acerose stylose stigma slightly curved or hooked in form, with the stigmatic groove uppermost. Gressitt’s notes indicate that the type plant was about 1 m. tall, with a trunk 1.5 cm. in diameter; the fruits were green. The leaves, which are quite small and narrow for the genus, are evenly tapered to a sharp, elongated apex.

COLLEGE oF GUAM

AGANA, GUAM

oe anaadeume

a

ERNEST JESSE PALMER

JOURNAL

OF THE

ARNOLD ARBORETUM

VoL. XLII OctTOBER 1962 NUMBER 4

ERNEST JESSE PALMER, 1875-1962 CLARENCE E. Kosusxk1!

With portrait

ERNEST JESSE PALMER died at Webb City, Missouri, on February 25, 1962, in his 87th year. This marked the end of a long and in many ways remarkable career in the fields of botany and allied sciences.

He was born in Leicester, England, April 8, 1875, and, in 1878, when he was three years old, he and his sister, Louise, were brought to this country by their parents, Amos and Annie Palmer, who settled temporarily near Warrensburg, in Johnson County, Missouri. When he was 14, his father, responding to the stories of quick wealth offered by the mining opportunities in the southwestern part of Missouri, moved the family to Webb City and built a home for them at 321 South Main Street. The wealth never materialized! Throughout the rest of his life, although he spent many of his years away from Webb City, the Palmer home was maintained, and after retirement, Ernest Palmer moved with his family and lived out the rest of his life there.

In this mining area, young Palmer did find a wealth of his own liking. It was here that he began his studies of natural phenomena. In the piles of waste material from the mines, fossils of plants and animals caught his fancy, and as in all boys of his age the collecting instinct came to the fore. He developed a great keenness for unearthing most interesting artifacts. In later years at the Arnold Arboretum he was always “picking up” Indian arrowheads and other Indian deposits. This was easy for him, but not necessarily for others. I can recall his telling me of a spot in the Arboretum where arrowheads were in abundance. I hurried to the exact location but found nothing. The next day we went out together— and sure enough, he picked up several more artifacts. Eventually two attractive trays of Palmer’s finds in the Arboretum were mounted and have been on display there ever since.

His formal education was broken off suddenly while he was attending high school. A physical collapse of his father made it necessary for him to leave school and seek employment. This was undoubtedly a severe

This article on Mr. Palmer will be ene also in the current volume of Rhodora, but without the bibliography. C.E.K

352 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLII

blow to his pride because he wanted an education above all things. He loved study and research — and he intended to continue it at all costs. Just recently [I learned from an article by Marcia W. Kershaw, a Special Correspondent of the St. Louis Post-Dispatch, that Palmer managed to study Latin and Greek at old Webb City Baptist College. I had often wondered where he obtained his profound knowledge of these languages, especially Latin which he used so ably in his original technical descrip- tions of new species of plants. Mrs. Kershaw’s article brought to light many other facets of Mr. Palmer’s career which were previously unknown to me and [ am very thankful to her for her assistance.

All his life he had managed to study not only the natural sciences but also English literature, mathematics, economics, and poetry until he was considered very proficient in each of them.

The portrait which accompanies this article was taken in 1958 and shows Mr. Palmer at his desk with a copy of his published book of poems entitled, “Gathered Leaves, Green, Gold and Sere.” These poems, I understand, are truly gathered leaves, ‘leaves’ which had been lying around in various places during the years and were gathered together by his wife and published through her efforts when Palmer was eighty- three years old.

His scientific papers were noted not only for his Latin descriptions but also for the fine phrasing which he always employed. He wrote so well and so easily that often he found it necessary to delete whole paragraphs from his papers because he considered the contents “unscientific.” It was a pity because everyone would have enjoyed reading them.

Although he was dedicated to the life of a naturalist it was not until his meeting with Benjamin Franklin Bush, also from Missouri, that he tried his wings and ventured away from his home setting. When Bush died in 1937, Palmer published a resumé of his life in The American Midland Naturalist. On re-reading this biography one cannot help but be struck by the similarity in the early backgrounds, the interests, and, in fact, the entire careers of the two men.

In 1901, Bush, then a collector for the Arnold Arboretum, visited the Palmer home and spent a week there using it as a base for his botanical collections. Bush was especially interested in obtaining specimens of Crataegus for C. S. Sargent, then director of the Arnold Arboretum, and when he departed from Webb City he left with Palmer a supply of driers for further collections. Thus at the age of 25 Palmer began his lifelong interest in Crataegus and his association with Sargent and the Arnold Arboretum. For the next twelve years he collected extensively in his home area, concentrating on the flora of Missouri, and it was not until 1913, when he was 38, that he became associated with the Missouri Botanical Garden. Two years later at the age of 40 he began his many collecting trips to the Southwest under the aegides of both the Missouri Botanical Garden and the Arnold Arboretum.

Palmer began publishing as early as 1910, at which time his first paper, “Flora of the Grand Falls Chert Barrens” appeared in the Trans-

1962 | ERNEST JESSE PALMER, 1875-1962 Siok)

actions of the Academy of Science, St. Louis. During the next ten years he published seven papers, mostly on ferns. In 1920, while still a collector- at-large he sent in from Webb City to the newly organized Journal of the Arnold Arboretum his first of many papers to be published by this Harvard University publication. Two more papers were published in this journal before he came to Jamaica Plain in 1921 as a regular staff member. He retained this status at the Arnold Arboretum for twenty-seven years, until he retired in 1948.

My personal association with Palmer began in 1927. By that time he was firmly established at the Arboretum as a collector-botanist. Since he was away so much of the time on his many collecting trips and since he was still a bachelor he maintained personal quarters in a rooming house.

He was a very busy man during this period. He managed to make a collecting trip nearly every year, usually alone, occasionally with a single companion. Upon his return from these trips his time was spent identifying his specimens, making up sets for distribution, and writing occasional papers — and getting ready for his next trip. He once mentioned to me that he had hopes of collecting in all the states of the Union. At the time he had nearly accomplished his intention. If I recall correctly he said that there were only two states in which he had not made “official” collections — and, as one might suspect, one of them was surprisingly close by!

The greater portion of his collections was made in the southwestern United States. However, he always maintained his interest in the Missouri flora, and, in 1935, there appeared in the Annals of the Missouri Botanical Garden the 385-page article by Palmer and Steyermark, “An Annotated Catalogue of the Flowering Plants of Missouri.” This catalogue is more complete than is usual in such a publication. The first seventy-five pages are devoted to the botanical history, terrain, climate, plant regions, etc., of the state. Under the actual listing of the taxa may be found the habitat and a record of the counties wherein specimens of each entry had been collected. A footnote states that an earlier publication on the ferns and fern allies of Missouri published by the same authors three years earlier in the American Fern Journal should be considered supplementary to the 1935 Catalogue.

A recent letter to me from Julian Steyermark, the junior author of the above two publications, mentions that his latest publication, Flora of Missouri, which has been in press since 1959 is being dedicated to Bey), Palmer. Palmer had hoped to see the book and knew of the proposed dedication.

In 1930, when he reached the age of 55 he surprised us all by getting married. We had thought of him as a confirmed bachelor. He married Elizabeth McDougal, a bacteriologist at the Massachusetts State Lab- oratory which is located alongside the Arnold Arboretum. Naturally, a great change took place in Palmer’s life. He eventually set up a fine home in one of the houses belonging to the Arnold Arboretuin in Jamaica Plain. He and his wife had three children: Ernest Macdougal Palmer,

354 JOURNAL OF THE ARNOLD ARBORETUM [VOL. XLII

born in 1931; Grace Elizabeth Palmer, in 1932; and Theodore Windle Palmer, in 1935

Palmer possessed a physique that defied age. He was the slender, wiry type and at the time of his retirement and departure from Jamaica Plain for Webb City could outlast any of us on a collecting trip. Although I did not see him again, I understand from his long letters and from various conversations with others that he continued his same brisk manner in everything which he undertook. His letters show that when he retired he merely transferred his activities from Massachusetts to Missouri. In a letter to me in February 1955 he says: “Of course we miss many friends there [Jamaica Plain], and I miss many of the associations at the Arboretum. But I have never been sorry that I retired when I did, for my life on the whole has been quite happy here. I have never been busier at any time or interested in more things. In April I will pass another anniversary — and you can figure it out as the records show that I was born in 1875. It is hard for me to believe it as I still feel as strong in body and mind — so far as I can tell —as I did at fifty. I can easily walk ten miles any day; and I think I could climb a good sized tree, if there was anything at the top of it that I wanted badly enough. . Steyermark and I are still working actively on the Flora of Missouri .. . Specimens of Crataegus, Quercus and other genera keep coming to me for determination or revision. . . Other collections in my museum also occupy much time. . . I have agreed to talk to a joint meeting of all the Garden Clubs of Carthage at the public library. While I do not solicit or particularly enjoy speaking engagements, I fill them occasion-

t is so fine and warm today that I think I will go fossil hunting this afternoon.” For a man retired and eighty years old he certainly was enjoying life to the fullest, doing exactly what he wanted to do — and he seemed so well equipped to do it!

During his life time Ernest Palmer joined many scientific organiza- tions, among them: the American Fern Society, the American Associa- tion for the Advancement of Science, the American Society of Plant Taxonomists, the Boston Mineral Club, the Botanical Society of America, the Missouri Archaeological Society, the New England Botanical Club (President, 1944-1945), and the St. Louis Academy of Science.

BIBLIOGRAPHY *

— 1910 —

Flora of the Grand Falls chert barrens. Trans. Acad. St. Louis 19: 97-112. — 1914 —

Ophioglossum Engelmanni in Missouri. Am. Fern Jour. 4: 66-68. — 1915 —

Trailing the first settlers. Archeol. Bull. 6: 37-41.

* Compiled by Lazella Schwarten and Elizabeth M. Palmer.

1962]

ERNEST JESSE PALMER, 1875-1962

— 1916 — Catalogue of the plants of Jasper County, Missouri plants.) Ann. Missouri Bot. Gard. 3: 3

(Fernworts and flowering Texas Pteridophyta—I. Am. Fern Jour. 9: 17-22

— 1919 — Texas Pteridophyta — II. /bid. 50-56 Texas Pteridophyta — III. /bzd. 81-85

— 1920 — mes canyon flora of the Edwards Plateau of Texas. —239.

Jour. Arnold Arb ea ligneous flora of the Staked Plains of Texas. /bid. 2: 90-105.

ae — 1921 — Botanical reconnaissance of southern a The forest flora of the Ozark region. [bid. 21

Jour. Arnold Arb. 2: 129-153. 32,

— 1923 —

The flowers that bloom in the ing. Cornell Rural School Leafl

The Red River Forest at Fulton, Arkansas Jour. Arnold Arb. 4: 8-3 — 1924 — Arb. 5:

16: 1-58. Two interesting ferns from Arkansas. 108-134.

Am. Fern Jour. 14: 39-4 The ligneous flora of Rich Mountain, Arkansas and eee “Jour. Arnold — 1925 — Is Quercus arkansana a hybrid? Jour. Arnold Arb. 6: 195-200 Synopsis of North American Crataegi. Ibid. 5-12 — 1926 —

Leaves from a collector’s note book. Jour. Arnold Arb. 7: 136-145.

The ligneous flora of Hot Springs National Park and vicinity. Ibid. Supplement to the catalogue of Arkansas Plants. Trans. Acad. St. Louis 25: 91

104-135. (With Buchholtz, J. T.) — 1927 — Ferns of the Davis Mountains, western Texas On Nuttall’s trail through Arkansas

Am. Fern Jour. 17 Jour. Arnold Arb. 8: 24-55 — 1928 — A botanical trip through the Chisos Mountains of Texas. Jour. Arnold Arb 9: 153-173. Leaves from a collector’s note book. Jbid. 173-187 — 1929 —

The ligneous flora of the Davis Mountains, Texas. Jour. Arnold Arb. 10: 8 45

— 1930 — Th

Additional notes on Texas ferns. Am. Fern Jour. 20:

e spontaneous flora of the Arnold Arboretum. a fe Arb. 11: 63-119

356 JOURNAL OF THE ARNOLD ARBORETUM [ VOL, XLIII

— 1931 — A conspectus of the genus Amorpha. Jour. Arnold Arb. 12: 157-197. illust. — 1932 — Crataegus arnoldiana. Addisonia 17: The Crataegus problem. Jour. oe ie: om 342-362. The ferns and fern allies of Missouri. (With Steyermark, J. A.) Am. Fern Jour. 22: 105-122, Leaves from a collector’s note book. Jour. Arnold Arb. 13: 417-437. Notes on Ophioglossum Engelmanni. Am. Fern Jour. 22: 43-47.

— 1933 — The family Isoétaceae in Missouri. (With Steyermark, J. A.) Am. Fern Jour. 23: 65, 66

— 1934 — Adventures in fern collecting. Am. Fern Jour. 24: 104-109. The beach plum in Michigan. Jour. Arnold Arb. 15: 88. Indian relics of the Arnold Arboretum. Bull. Pop. Inf. IV. 2: 61-68. illust. Notes on some plants of Oklahoma. i Arnold Arb. 15: 127-134 Quercus ellipsoidalis in Missouri. Ibid. Trees of the Southeastern States. ae Ibid. 266.

— 1935 — Additional fern records for Missouri. (With Steyermark, J. A.) Am. Fern Jour. 25: 1-6.

An annotated catalogue of the flowering Nata of Missouri. (With Steyermark, Ann. Missouri Bot. Gard. 22: -758

uae ae i the spontaneous flora of . Arnold Arboretum. Jour. Arnold Arb. 16: 81-97.

Two new species of Crataegus from Missouri. Ibid, 353-357.

Uvidaria perfoliata in Louisiana. (With Anderson, E.) Rhodora 37: 58, 59.

— 1936 —

Adventures in fern collecting, I]. Am. Fern Jour. 26: 136-139.

Brauneria atrorubens and B. paradoxa. Rhodora 38: 197-199.

Deciduous shrubs and trees; Evergreen shrubs and conifers: Broad leaved evergreens. Landscape Gardening and Landscape Art, Garden Clubs of Ark.

. 22-39,

The hawthorns (Crataegus: over 50 species in Arkansas). Native plant material (famous in garden lore). Arkansas state program for garden clubs. (Mimeo- graphed.)

— 1937 —

Benjamin Franklin Bush. Am. Midl. Nat. 18(3): i-vi. portra

The identification of plant material at the Arnold oe. Bull. Pop. Inf. IV. 5: 13-16.

Notes on North American trees and shrubs. Jour. Arnold Arb. 18: 133-140.

— 1938 —

Additions, corrections, and revisions of the annotated catalogue of the flowering plants of Missouri. (With Steyermark, J. A.) Ann. Missouri Bot. Gard. 25: 775-794

1962 | ERNEST JESSE PALMER, 1875-1962 Sou. New varieties and forms from Missouri. (With Steyermark, J. A.) /bid. 769-773 — 1939 —

Mines and minerals of the Tri-State District. Rocks & Minerals 14: 35-49, Some woody plants of Rhode Island. Rhodora 41: 314-316

— 1940 — Adventures in fern collecting, III. Am. Fern Jour. 30: 1-9. Crataegus of Indiana. In: C. C. Deam, Flora of Indiana. 533-557.

Geodes. Rocks and Minerals 15: 120-123. A new form of Crataegus viridis Linn. (With Pickens, A. L.) Neighborhood Research 4: 61. — 1941 — Answering isolationists. Changing World 13: 10-12. — 1942 —

Panicum recognitum in Rhode Island. Rhodora 44: 227. The red oak complex in the United States. Am. Midl. Nat. 27: 732-740.

— 1943 — Quercus prinus Linnaeus. Am. Mid]. Nat. 29: 783-784. The species concept in Crataegus. Chron. Bot. 7: 373- Hise

Hed (eye Food plants in the Arnold Arboretum. Arnoldia 4: 1-7. bs. Ao?

— 1945 — Quercis Durandii and its allies. Am. Midl. Nat. 33: 514-519.

— 1946 — Crataegus in the northeastern and central United States and adjacent Canada. Brittonia 5: 471

eso) ee

Ferns and fern allies of Cumberland, Rhode Island. Am. Fern Jour. 37: 33-38.

Second supplement to the s eo ieee flora of the Arnold Arboretum. Jour. Arnold Arb. 28: 410-418

— 1948 — Hybrid oaks of North America. Jour. Arnold Arb. 29: 1-48.

— 1950 — Crataegus. In: M. L. Fernald. Gray’s Manual of Botany (8th ed.) pp. 767-801. Notes on eae EDIT Mackenzie. (With Steyermark, J. A.) Bull. Torrey Club 77: Pai

— 1951 —

Trees and plants of the Southwest Missouri Region. 21 articles appearing weekly in the Webb City Daily Sentinel Oct. 1951-Mar. 1952.

358 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIII

— 1952 —

Crataegus. In: Henry A. Gleason, The new Britton and Brown illustrated flora of the northeastern United States and adjacent Canada 2: 338-375 New pteridophyte records from Missouri. (With Steyermark, J. A.) Am. Fern

Jour. 42: 61-66.

— 1953 — A hybrid Amorpha and new forms and records from Missouri. Rhodora 55: 157-160 — 1955 — Plants new to Missouri. (With Steyermark, J. A.) Rhodora 57: 310-319.

— 1956 — Crataegus in Ohio with description of one new species. Ohio Jour. Sci. 56: 205-216.

— 1958 — Cephalopods aes the Burgner Formation in oe (With Unklesbay, A. G.) Univ. Missouri Jour. Paleont. 32: 1070-10 Plants new to “Missouri. (With Steyermark, J. A.) Balt 10: 109-120.

— 1962 — The Ozark Forest. Missouri Log (pub. annually by School of Forestry, Univ. Missouri).

1962] BRIZICKY, GENERA OF ANACARDIACEAE 399

THE GENERA OF ANACARDIACEAE IN THE SOUTHEASTERN UNITED STATES *

GEORGE K. BrizIckKy

ANACARDIACEAE Lindley, Intr. Nat. Syst. 127. 1830. nom. cons. CASHEW FAMILY)

Trees or shrubs, rarely subshrubs or vines, with resin ducts in bark and often in leaves, flowers, and fruits. Leaves alternate {very rarely oppo- site or verticillate], simple, 3-foliolate, or pinnate, deciduous or persistent, exstipulate. Flowers small, usually regular, hvpogynous [rarely perigy- nous or epigynous], usually unisexual by abortion or both uni- and bi- sexual [rarely bisexual throughout], in axillary and/or terminal thyrses or panicles, more rarely in solitary or panicled spikelike racemes; bracts deciduous or persistent, bractlets 2, deciduous or persistent, or wanting: Perianth double [rarely simple or absent], usually 5-merous. Sepals usually connate at least at base, imbricate [or valvate]. Petals usually distinct, imbricate [or valvate]. Stamens 5-10 [or very rarely more], sometimes only 1 or 2 fertile in ¢ flowers, distinct [or basally connate], reduced and sterile [or wanting] in @ flowers: anthers versatile, 2-locular at anthesis, introrse, longitudinally dehiscent. Intrastaminal (or rarely extrastaminal) nectariferous disc usually 5- or 10-lobed [sometimes pro- duced into a gynophore, rarely absent]. Gynoecium syncarpous [or apo- carpous], 3[-5]-carpellate or 1-carpellate by reduction, rudimentary [or absent] in 4 flowers; stigmas 1-3[—5]; styles 1-3[-5]; ovary 1[3— 5]-locular; ovules anatropous, apotropous, solitary in a locule, the funicle usually elongate with a basal, parietal, or apical insertion. Fruit usually drupaceous, with + resinous and sometimes waxy or oily mesocarp and crustaceous or bony endocarp (stone). Seeds with scanty endosperm or endospermless; embryo + curved [or straight]. (Terebinthaceae Juss., 1789: Spondiaceae Kunth, 1824.) Type GeNus: Anacardium

1 Prepared for a generic flora of the southeastern United States, a joint project of under the direction of Reed C. Rollins and Carroll E. Wood, Jr. This treatment are the pattern established in the first paper in the series (Jour. Arnold Arb. 39: 296- 346. 1958) and continued through those in volumes 40-43 (1959— 1962). It should be plants of ae eee with any supplementary material in brackets. References which

Since is aie to Dr. Carroll E. Wood is for his criticism and valuable

en and to Mrs. Gordon W. Dillon, for her careful help in the preparation of the manuscript.

360 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLUI

A family of about 70 genera with nearly 600 species, primarily pan- tropical, extending with several genera into the temperate zones of both hemispheres. Of the five tribes, only Rhoéae (Rhoideae) and Anacardieae (Mangifereae Marchand emend. Engler) a1 are represented in our area, the former by three indigenous genera and a naturalized species of chile the latter by a species of Mangifera naturalized in southernmost Flori Since generic lines in the oéae are often weak and differently ane preted, various authors have recognized one to five indigenous genera in our area.

The family, with a few exceptions (e.g., Pistacia L.), is considered to be entomophilous. Since the species are mostly dioecious or polygamo-dioeci- ous, cross-pollination should be the rule. Apomixis (adventitious poly- embryony) has been found in some species of Mangifera, parthenocarpy in several genera. Cytological data presently known for about 25 species in ten genera (2m = 24, 28, 30, 32, 40, 42, 48, 60) suggest polyploidy. Resin of many species is poisonous on contact.

The family is believed to be closely allied with Sapindaceae and Juliana- ceae. Wood-anatomical and palynological evidence seems to support the assumption of a close relationship with Burseraceae, from which Anacar- diaceae differ essentially in the uniovulate ovary locules and in apotropy of he ovules.

Anacardiaceae are economically important for edible fruits (e.g., Mangi- fera indica, mango: Spondias spp., mombin) or nuts (Anacardium occi- dentale L., cashew; Pistacia vera L., pistachio), for resins (e.g., Pistacia Lentiscus L., mastic) and lacquers (Rhus verniciflua), for tannins (Cotinus Coggvgria, Rhus spp., Schinopsis spp.), and for timbers (Schinopsis spp., red quebracho; Astronium spp., zebrawood or kingwood). Some species are ornamentals.

REFERENCES: Bark Ley, I’. A. A monographic study of Rhus a its ee allies in North and bernie America, including the West In . Missouri Bot. Gard. 24: 265-498. pls. 10-26. 1937. | Includes Eo ieee Rhus, Toxi- codendron, Malosma, and Actinocheita. . A key to the genera of the Anacardiaceae, Am. Midl. Nat. 28: 465- 474. 1942. [Bibliographical footnotes; Searsia Barkley and Duckera Bark- ley segregated from Rhus. | Generic key to the sumac family (Anacardiaceae). Lloydia 20: 255- 265. 1957. [Includes introductory notes on economic uses. | BLANK, H. Dermatitis from Anacardiaceae. Proc. Fla. Mango Forum 17: 24-26. oro Beas Encier, A. Uber die morphologischen Verhaltnisse und die geographische Ver- breitung der Gattung Rhus, wie der mit ihr verwandten, lebenden und aus- gestrobenen Anacardiaceae, Bot. Jahrb. 1: 365-420. . Burseraceae Anacardiaceae. Ja: CANDOLLE, A. & C. pe. Monogr. Phaner. 4: 1-500, 536-573. pls. 1-15. 1883. | Anacardiaceae, 171-500, 537,

wal wa CO >> =~ a

2] Anacaedincéas Nat. Pflanzenfam. III. 5: 138-178. 1892. Heimscu, C., Jr. Wood anatomy and pollen morphology of Rus and allied

—

1962 | BRIZICKY, GENERA OF ANACARDIACEAE 361

genera. Jour. Arnold Arb. 21; 279-291. pls. 1-3. 1940. {Includes Actino- cheita, Malosma, Metopium, Cotinus, Rhus, and Toxicodendron mparative anatomy of the Seconda, xylem in the “Gruinales” and

“Terebinthales,” of Wettstein with reference to taxonomic grouping. Lilloa 8: 83-198. pls. 1-17. 1942. [Anacardiaceae, 136-144, pls. 9-12.|

Kryn, J. M. The anatomy of the wood of the Anacardiaceae and its bearing on the Beas and relationships of the family. Diss. Abs. 13: 290. 1953.*

Martin, A. C. The comparative internal morphology of seeds. Am. Midl. Nat. 36: 513-660. 1946. [Anacardiaceae 628, 629

Merritt, E. D. Dermatitis caused by various representatives of the Ana- cardiaceae. Jour. Am. Med. Assoc. 124: 222-224.

Recorp, S. J. American moods of the family Annee Trop. Woods 60: 11-45. 1939.

SARGENT, C. S. Manual of the trees of North America (exclusive of Mexico). ed. 2. 910 pp. Boston & New York. 1922. [Anacardiaceae, 655-665. |

Sweet, H. R., & F. A. Barkrey. A most useful plant family, the Anacardiaceae. Missouri Bot. Gard. Bull. 24: 216-229. 1936

VENNING, F. D. The ontogeny of the laticiferous canals in the Anacardiaceae Am. Jour. Bot. 35: 637-644. 1948.

West, E., & L. E. Arnotp. The native trees of Florida. 212 pp. Gainesville. 1946. [ Anacardiaceae, 112-114. ]

KEY TO THE GENERA OF ANACARDIACEAE

General characters: woody plants with resin ducts in bark, leaves, flowers, and fruit; flowers small, usually hypogynous, unisexual or uni- and bisexual, in axillary and/or terminal thyrses or panicles; perianth double, usually 5-merous ; stamens 5-10; nectariferous disc intra- or rarely extrastaminal; gynoecium 1-3 | -5 |-carpellate, ses usually 1-locular by abortion; fruit drupaceous Flowers é and bisexual; petals with longitudinal ridges; disc oe eae stamens 5, only 1 or 2 fertile; gynoecium 1-carpellate; drupes large to ve large, with copious juicy flesh; leaves simple; naturalized tree . a cacal FSO TULA aye tose cts, Cn ses eke ole te ee ce 1d cola 1. Manigifera. . Flowers usually unisexual; petals not ridged; disc ee stamens

5 or 10, all fertile in ¢, all reduced and sterile in 2 flowers; gynoecium 3-

carpellate; drupes cum not exceeding 15 mm. in length, with meager dry-

He flesh; leaves variou

. Stamens 10; ae spicy-aromatic, containing oil cavities adherent to stone ; eaves pinnate, with winges: rachises; naturalized ee or small treessor southern «blOmGa. 2 4a ote ee eee en tts: . Schinus.

B. Stamens 5; drupes not aromatic, without oil cavities; leaves ae if

pinnate, rachises not winged, except in Rhus copallina.

C. Stigma 1, 3-lobed; style 1, very short, stout; drupes 10-15 mm. long, ellipsoid to obovoid, scarlet to orange, glabrous; endocarp thin, crustaceous; leaves odd-pinnate, leaflets long petioluled; A a trees or shrubs of subtropical: Floridasgnna ss .25 2.50. . Bh topium.

3-8 mm. long, subglobular or oe form (covered with glandular club-shaped hairs if red or orange); endocarp thick, bony; leaves various (if pinnate, leaflets sessile or short petioluled).

D. Styles appearing sublateral in flowers; drupes obliquely subreni-

>

>

& WM Hag

q soo) wm pe?) jo] jor) wm os al SB oO wm ww fol} en o

362 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIII

form, usually much compressed laterally, prominently reticulate, elabrous, the remnants of styles lateral; mesocarp very scanty; pedicels of aborted flowers becoming plumose-villous; seal fee innocuous trees or shrubs. ................ 4. Cotinus, . Styles terminal; drupes subglobular, usually slightly pains laterally, not reticulate, sometimes with simple and/or glandular hairs; mesocarp thickish, resinous and sometimes also waxy; pedicels never plumose; leaves 3-foliolate or odd-pinnate; trees, shrubs, or vines, some poisonous. .................... 5. Rhus.

=)

Tribe ANACARDIEAE 1. Mangifera Linnaeus, Sp. Pl. 1: 200. 1753; Gen. Pl. ed. 5. 93. 1754.

Mostly large evergreen trees, the leaves alternate [rarely verticillate | , simple, entire, membranaceous to coriaceous, petiolate, persistent. Plants polygamous, usually andromonoecious. Flowers small, ¢ and bisexual, in terminal [and/or axillary] thyrses; pedicels articulated; bracts and bract- lets deciduous. Sepals 5 (4 or 6), deciduous. Petals 5 (4 or 6), distinct [or adnate to the disc at base], the veins thickened into [1—|]3—5 prominent ridges, deciduous. Disc extrastaminal, tumid [or minute, sometimes stipe- like, or wanting], S-lobed. Stamens usually 5 [rarely 10-12], inserted just within [or on] the disc, distinct [or connate at base], usually 1 or 2 [rarely 5 or 6] fertile, the rest remaining sterile with imperfect small an- thers [or reduced to toothlike projections]; pollen 3-colpate, small to me- dium sized, ellipsoidal, finely reticulate. Gynoecium 1-carpellate (excep- tionally 2-carpellate, apo- or syncarpous), rudimentary or wanting in ¢ flowers; stigma simple; style sublateral; ovary obliquely subglobular, 1- locular, sessile; ovule solitary, funicle subbasal. Fruit a large [or relatively small] drupe of various shapes and colors; exocarp + leathery; mesocarp thick, fleshy, sweet [to acid], sometimes resinous; stone compressed, usual- ly hard, woody, coated with fibers. Seed large, compressed; testa papery; endosperm lacking (or scanty?); embryo subreniform with plano-convex, often unequal and lobed cotyledons and small ascendent radicle; germina- tion hypogeous. Type species: M. indica L. (From Portuguese, mango, the common name for this fruit, and Latin -fera, bearing, i.e., “bearing mangoes.”’)

An Indo-Malayan genus of about 40 species. Mangifera indica (includ- ing M. laurina Bl.), common mango, 2n = 40, presumably native to India (and perhaps to Indochina and the Sunda Islands), widely cultivated throughout the tropics for its edible fruits, has become naturalized in ham- mocks of southern Florida and the Florida Keys. A tree with relatively large, narrow leaves and large to very large, ovate to subreniform or sub- cylindrical, yellowish or greenish fruits sometimes flushed with scarlet or crimson, the species has been in cultivation in the tropics of the Old World for over four thousand years and shows a wide range of variation, about a thousand wild forms and cultivars occurring in India.

1962 | BRIZICKY, GENERA OF ANACARDIACEAE 363

Mangifera indica is the only economically important species. Its fruits, mangoes, are said to be eaten by at least one-fifth of the world’s popula- tion. In India the astringent flowers, seeds, and bark find medicinal appli- cation, especially in diarrhoea. Bark and leaves yield a yellow dye used locally in southern Asia. The wood, said to be of excellent quality, is used extensively in India for carpentry, cabinet-work, boat-building, e

Short-tongued insects, especially Diptera, Hymenoptera, and Coleoptera, seem to be the principal pollinators of Mangifera indica. Cross-pollination seems to be the rule. Since self-pollination has not been recorded, one may assume that the morphologically bisexual flowers are (at least in most cases) functionally (physiologically) carpellate. The somatic chromosome number 2” = 40 has been recorded for 23 grafted varieties and one wild race of Mangifera indica, as well as for M. sylvatica Roxb., M. caloneura Kurz, M. caesia Jack, and M. foetida Lour. Although no polyploid series has been found in the genus, Mukherjee (1950) suggested on the basis of his cytological analysis that M. indica and its allies are allopolyploids. No records of natural or artificial interspecific hybrids have been available, but Mukherjee (1957) remarked that “similarity in the chromosome num- ber and morphology and in the pollen size and morphology . . . indicates that there may be close compatibility among the species during hybridiza- tion and in stock-scion relationship in grafting . . .’

Apomixis (adventitious polyembryony) has been recorded in Mangifera indica and in M. odorata Griff., the embryos originating either from the nucellus or by budding from the cotyledons and hypocotyl. While Indian “seedling races” and “horticultural [grafted] varieties” are almost exclu- sively monoembryonic, the ‘‘seedling races” common in culture in some other countries (‘‘Philippine” or ‘‘Manila” races) mostly are polyembry- onic. There are, however, records indicating that some “Indian races” considered monoembryonic in India yielded a considerable percentage of polyembryonic seeds when grown in the Philippines.

Although Mangifera indica is usually innocuous and dermatitis caused by its resin is infrequent, several wild species (e.g., M. caesia, M. foetida, M. odorata, and M. lagenifera Griff.) sometimes cultivated in the Malaysian region for their edible fruits are regarded as positively poisonous.

The genus seems to be related to the Malayan Bouea Meisn. and to the tropical American Anacardium L.

REFERENCES:

The vast number of references has been reduced here primarily to those of general interest. Under family references see ENGLER (1883, pp. 195-215 & 1892, pp. 146, 147), Hermscu (1942, pp. 136, 137), and MeErRILt.

ArNpT, C. H. Notes on polyembryony and multiple aie from the seed in Mangifera indica. Am. Jour. Bot. 22: 26-28. pls. 1, 935.

BaRNES, H. V. The mango; a list of references. U. S. ae Agr. Libr. List 29. 62 pp. 1946.*

CHoupHury, K. R., & S. K. MuKkuHerjee. Floral biology in mango. Proc. Indian Sal Congr: Assoc. 42: 370, 371. 1955.*

364 JOURNAL OF THE ARNOLD ARBORETUM yOu, xn

HAMMERSTEIN, C. P. Mango dermatitis experiences. Proc. Fla. Mango Forum 19: 14-16. 1959.* [Control of allergy to mangoes.

JULIANO, J B. Origin of embryos in the strawberry mango. Philip. Jour. Sci. 54: 553-559. 1934. [See also Philip. Agr. 25: 749-760. 1937.]

. E. Cuevas. Floral morphology of the mango ee indica Li) with eer eos to the Pico variety from the Philippines. Philip. Agr, 21: 449-472 *

KENNARD, W. C. Bite of the fruit, seed and embryo of the Paneri mango. Bot. Gaz. 117: 28-32. 1955

Leroy, J. F. Sur un ‘complexe aeamicue? des manguiers et sur l’origine et la phylogénie des variétés cultivées. Revue Int. Bot. Appl. Agr. Trop. 27: 304-309. 1947.

Lurz, B. Estudos sobre a biologia floral de pee es ba es ees sum- mary.) Mus. Nac. Rio de Janeiro 26: 125- pls. 1-4.

Lyncu, S. J.. & R. O. Netson. Current methods of ees See of avocado, mango, lychee and guava in Florida. Ceiba 4: 315-337. 1956. | Mangifera indica, 320-328. |

Matix, P. C. Morphology and biology of the mango flower. Indian Jour. Hort. 14: 1-23. 1957.*

MUKHERJEE, S. Kk. anaes oe of the species of Mangifera Linn. Bull. Bot. Soc. Bengal 2: 15

The varieties of mango ca indica L.) and their classification. ibid. 2: 101-133. 1948. . The iaononite value of the anatomical Sane of inflorescence axes of citi tale L. Jour. Indian Bot. Soc. 28: 162-171. hi graph on the genus Mangifera is ant 12: 73-136. 1949. : unas investigation of the mango (Mangifera indi L. ) a the allied eu species. Proc. Natl. Inst. Sci. India 16: 287-30 The mango: its allopolyploid nature. Nature 166: 196, 197, 1950. ; Pollen analysis in Mangifera in relation to fruit-set and taxonomy. Jour. Indian Bot. Soc. 30: 49-55. 1951 . Origin, distribution and phylogenetic Sos of the species of Mangi- fera L. Tou Linn. Soc. Bot. 55 =83:

The mango: its botany, a uses and future improvement,

panne as observed in India. Econ. Bot. 7: 130-162. .

i, 2G y of some Malayan species of Mangifera. Cytologia 22: 239-

241. 1957. “TM. foetida and M. caesia, 2n = 40. |]

Mustarp, M. J., & S. J. Lyncu. Flower-bud formation and development in Manegifera indica. Bot. Gaz. 108: 136-140. 1946.

Rao, P. L. N. The essential oil of the mango flower. Sci. Cult. 11: 70. 1946.*

Rao, U.N., & S. seers AMY. A note on ses formation from the axillary buds of the mango. S. Indian Hort. 3: 30, 955.

Ramee. Po. he FOR, - W. C. Cooper. ae studies of floral induction in the Haden mango Series indica L.), Am. Jour. Bot. 36: 734-740. 1949. [See also ibid. 33: , 210. 1946.]

RvuEHLE, G. D., & R. B. .. Man growing in Florida. Fla. Univ. Agr. Ext. Bull. 174. 88 a map. 1960.

SacHar, R. C., & R. N. CHopra. A study of ae — and embryo in Mangifera L. Sone Jour: Agr. Scl..272-21

SHARMA, M. R. Studies in the family etme eres ¥ weal anatomy of the flower of Mazgifera indica L. Phytomorphology 4: 201-208. 1954.

~

1962 | BRIZICKY, GENERA OF ANACARDIACEAE 365

SincH, L. B. The mango: botany, cultivation, and utilization. 438 pp. London

& New York. 1960.* [ Bibliography, 395-420. ]

& R. N. Sincu. A monograph on the mangoes of Uttar Pradesh. 16

1. 144 pp., 80 pls.; vol. 2. 138 pp., 75 pls. Lucknow. 1956. [Descriptions of vars. accompanied by colored pls. |

Sincu, R. N. Sex ratio and fruit setting in mango (Mangifera indica L.). Science 119: 389, 390. 1954.

SPENCER, J. L., & W. C. Kennarp. Limited stigmatic receptivity may con- tribute to low fruit set in the mango (Mangifera indica L.). Proc. Am. Soc. Hort. Sci. 67: 287-289. 1956.

VENKATARATNAM, L. Hormone induced set and Perey in mango (Mangifera indica L.). Curr. Sci. Bangalore 18: 409. 1949.*

Wesser, H. J. The economic importance of apogamy in Citrus and Manegifera. Proc. Am. Soc. Hort. Sci. 28: 57-61. 1931.

Younc, T. W. Investigations of the unfruitfulness of the Haden mango in Florida. Proc. Fla. State Hort. Soc. 55: 106-110. 1943.*

Influence of temperature on growth of mango pollen. /bid. 68: 308-

SiSe 195525

Tribe RHOEAE Marchand, ‘“Rhoideae”

2. Schinus Linnaeus, Sp. Pl. 1: 388. 1753; Gen. Pl. ed. 5. 184. 1754.

Trees or shrubs [rarely subshrubs, sometimes thorny]. Leaves odd-pin- nate or rarely even-pinnate [or simple]; the leaflets membranaceous to subcoriaceous [or coriaceous |, (3)5-13[-41], usually opposite [or al- ternate], relatively small to medium sized, entire or toothed, sessile or subsessile, the rachis often winged. Plants usually dioecious. Flowers usually unisexual, small, pediceled, in axillary [and terminal] thyrses |or in contracted raceme-like inflorescences]; bracts small, bractlets 2, minute, deciduous. Sepals 5 [4], persistent. Petals 5 [4], white [or yellow], con- siderably longer than the sepals. Stamens 10 [8], in 2 series, the anti- petalous shorter than the antisepalous, inserted below and between the lobes of a saucer-shaped 10[8]-lobed intrastaminal nectariferous disc, re- duced and sterile in 2 flowers; pollen 3-colpate, medium sized, ellipsoidal, finely reticulate. Gynoecium 3-carpellate, rudimentary in 4 flowers; stigmas 3 [1], capitellate; styles 3 [1], connate at least at base; ovary sessile, 1-locular by abortion; ovule solitary, 2-integumented, with a thick nucellus, suspended from near the top of the locule. Drupe small, pea-like, bright red [pink or lavender]; exocarp thin, chartaceous, shining; meso- carp resinous, the innermost layers (adherent to endocarp) with large cavi- ties containing an aromatic oil; stone laterally compressed, cartilaginous to bony. Seed sublenticular; testa thin, membranaceous; endosperm fleshy, scanty; cotyledons flat; radicle elongated, incurved upwards. (Including Duvaua Kunth). Lecrotypr species: S. Molle L., 2n = 28, 30; see A. S. Hitchcock, im Int. Bot. Congr. 1930. Nomencl. Propos. Brit. Bot. 153. 1929. (An ancient Greek name for the mastic tree, Pistacia Lentiscus L., applied by Linnaeus to this genus.) — PEPPER-TREE.

A genus of about 28 species of warm-temperate and tropical South

366 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIII

America, from Chile, Argentina, and Uruguay, north to Venezuela and Colombia. Schinus terebinthifolius Raddi, Brazilian pepper-tree or Christ- mas-berry, with 5-13 oblong-ovate leaflets, and S. Molle, Peruvian pepper- tree, with 15-41 very narrow leaflets, are frequently cultivated in the warmer parts of the Eastern and Western hemispheres as ornamentals. Both are grown in Florida, and the former, a shrub, has become naturalized in southern Florida and the Keys where it seems to be spreading rapidly; the exact extent of naturalization needs to be determined. Schinus Molle has become naturalized in California and Mexico. The infructescences are often used as Christmas decorations. In Mexico and South America the fruits are used in various beverages, the seeds are sometimes used to adulterate pepper, and the resin, bark, and fruits are used locally in medicine.

The genus seems to be most closely related to the South American Lithraea Miers and the East Australian Rhodosphaera Engl. The group is entomophilous, and honeybees have been reported as frequent visitors of flowers of Schinus Molle. No hybrids have been recorded.

REFERENCES: Under family references see ENGLER (1883, pp. 331-345 & 1892, pp. 162-164), HeErIMScH (1942, pp. 139-141), and Recorp (pp. 38-40). BARKLEY, F. A. Schinus L. Brittonia 5: 160-198. 1944. [Monograph. | . A study of Schinus L. Lilloa 28: 5-110. 1957. [ Monograph. | CopeLAND, H. F. The reproductive structures of Schinus Molle (Anacardiaceae). Madrono 15: 14-25. 1959. [Floral morphology and embryology; 2n =

30.]

Dickson, J. D., HII, R. O. Woopsury, & T. R. ALEXANDER. Check list of flora of Big Pine Key, Florida and surrounding keys. Quart. Jour. Fla. Acad. Sci. 16: 181-197, 1953. [S. terebinthifolius, an escape on Big Pine Key,

92.]

RAINERI, L. Lo sviluppo dei fiori femminili di Schinus Molle L. (English sum- mary.) Nuovo Gior. Bot. Ital. II. 59: 49-63. 1952. [Development of the @ flowers. |

——. Lo sviluppo dei fiori maschili di Schinus Molle L. (English summary.) Ibid, 61: 409-413. 1954. [Development of the 4 flowers. |

3. Cotinus Miller, Gard. Dict. Abr. ed. 4. 1754.

Small trees or shrubs with orange-yellow wood and _ strong-smelling, resinous sap. Leaves simple, entire, medium sized [or small], membrana- ceous, deciduous, the petioles short to long. Plants usually dioecious [occasionally polygamo-dioecious or -monoecious|. Flowers small, usually unisexual [rarely also bisexual], many abortive, in loose terminal thyrses, the slender pedicels much elongated after flowering, those of abortive flowers becoming plumose-villous; bracts narrow, + scarious, in part persistent. Sepals 5, persistent. Petals 5, yellowish- or greenish-white [rarely crimson], twice as long as the sepals. Stamens 5, distinct, inserted below an annular, shallowly 5-lobed intrastaminal disc, alternate with and shorter than the petals, reduced and sterile in @ flowers; filaments

1962 | BRIZICKY, GENERA OF ANACARDIACEAE 367

subulate; anthers broadly ovate in outline, somewhat shorter than the filaments, 2-locular at anthesis; pollen 3-colpate, medium sized, sub- spherical, striate-reticulate, with round to slightly oblong, smooth germ pores. Gynoecium 3- carpellate, rudimentary in ¢ flowers; stigmas 3, small, subcapitate; styles 3, appearing sublateral, unequal, that of the fertile carpel conspicuously longer than the 2 sterile ones; ovary 1-locular by abortion, obliquely obovoid, sessile; ovule raised on a basal funicle. Drupes small (4-5 mm. long), obliquely obovate to subreniform, usually much compressed, conspicuously reticulate veined, the style-remnants lateral; exocarp membranaceous; mesocarp very scanty; stone (endocarp) bony, subreniform, 1-seeded. Seed reniform; testa thin, membranaceous; endosperm wanting (?); embryo with flat, elliptic cotyledons and an elongate radicle incurved upwards (toward the hilum). Typr sPrciEs: Rhus Cotinus L. = Cotinus Coggygria Scop. (Derived from Greek, kotinos, an ancient name for wild olive, used by Pliny for an unidentified shrub of the Apennines, but applied by some pre-Linnaean botanists, e.g., Tournefort, to C. Coggygria.) — SMOKE-TREE.

A genus of three or more species, primarily of warm-temperate Eurasia (southern France to the eastern subtropical Himalaya and central China) and eastern North America. Cotinus obovatus Raf. (C. americanus Nutt., Rhus americanus (Nutt.) Sudworth), American smoke-tree,” a tree, some- times a low shrub, with obovate-cuneate leaves turning orange to crimson in autumn, occurs sparingly on wooded, rocky cliffs and river bluffs, usually on limestone, in disjunct populations in the mountains of northern Alabama and adjacent Tennessee (Franklin County); in Daviess County, Kentucky (introduced?); in southwestern Missouri, northwestern Ar- kansas, and eastern Oklahoma; and on the Edwards Plateau, Texas. “Its rarity, discontinuous distribution, lack of related species in the New World, and occurrence as a pioneer on rocky cliffs all suggest that [it] is an old species formerly of general distribution but now approaching extinction” (Little, p. 23). Trees to about 10 m. tall have been recorded, but Sargent (1892) wrote that “during the War of Secession nearly all the large specimens were cut down for the dye which the wood yields . . .”

and monoecism, as well as flowers with 3-locular ovaries, have been ob- served occasionally. Short-tongued Hymenoptera and Diptera apparently are the principal pollinators. Parthenocarpy seems to be of common oc- currence, since many fruits fail to develop seeds. Although birds have

? Since the above was written, chromosome numbers of three taxa of Cotinus have been determined by Dr. Otto T. Solbrig, of the Gray Herbarium, from staminate material collected by Dr. C. E. Wood from the living ae of the Arnold Arbo- retum: C. obovatus (Arnold Arb. No. 1814-B, from seed, Charles Mohr, Mobile, Alabama, ae 5, 1882; Wood 9427 [aan]), 2n = 15%; C. Coggygria (Arnold Arb. No. 276, from Harvard Bot. Gard., 1876; Wood 9428 [aaH]), 2n = 15; and C. Coggygria ‘Atropurpureus’ (Arnold Arb. a 708-49-A, from seed, New York Bot. Gard., 1948; Wood 9429 [AaH]), 2n = 1

368 JOURNAL OF THE ARNOLD ARBORETUM [ VOL, XLIII

been mentioned, wind is assumed to be the principal means of seed dis- persal by transport of the whole infructescence. Asexual reproduction takes place by root- and stump-sprouts.

In spite of many similarities with Rhus, from which Cotimus differs mainly in the asymmetrical ovaries and fruits, unequal styles appearing sublateral in flower and lateral in fruit, very scanty mesocarp, and pe- culiar infructescences, the two genera do not seem to be very closely related. According to Engler, Cotinus represents a branch of a line of evolution different from but with a common origin with that of RAws. Some relationship to the South African genera Laurophyllus Thunb., Smodingium E. Mey., and Loxostylis Spreng. f. ex Reichenb, (Botryceras Willd.) seems possible.

The leaves and bark of Cotinus Coggygria are important sources of tannin and have been employed extensively in the tanning industry in southern Europe. Leaves and flowers yield an aromatic oil. The orange- yellow dye (fustic, ‘young fustic’’) extracted from the heartwood of C. Cogevgria and C. obovatus was formerly used extensively for dyeing silk, wool, etc. Both species are ornamental.

REFERENCES:

See also under family references BARKLEY (1937, pp. 300-303), ENGLER (1881, p. 378; 1883, pp. 349-352; 1892, pp. 164, 165), HermscH (1940; 1942, pp. 139-141), Recorp (p. 23), and SARGENT (pp. 657, 658).

Buck Ley, S. B. Rhus cotinoides, Nutt. Proc. Acad. Nat. Sci. Phila. 1881: 125. 1881. gece in Alabama. |

HorHAmMMer, L., H. WacNner, J. IzQuterpo, & H. Enpres. Isolierung eines aie aus Cotinus Coggygria Scop. (Rhus Cotinus L.) durch Chromatographie an Polyamid. Arch. Pharm. 291: 269-273. 1958.

Kinc, H. G. C., & T. WHire. colouring matter of Rhus Cotinus wood (young fustic). Chem. Soc. Jour. 1961: 3538, 3539. 1962.*

Litter, E. R. American ae (Cotinus obovatus Raf.), one of Oliahoma rarest tree species. Proc. Okla. Acad. Sci. 23: 21-23. 1943. [C. obovatus Raf. correct name for the sp.; distribution and a new locality in Oklahoma. |

MEEHAN, T. Sex in Rhus Cotinus. Proc. Acad. Nat. Sci. Phila. 1873: 300. 1873.

On hermaphroditism in Rhus Cotinus (the Mist Tree) and in Rhus ee (Common Sumac). Proc. AAAS 22(B): 73-75. 1874.

Osmota, N. Ku. The course of accumulation of tannides in the leaves of Cotinus Coggygria Scop. (In Russian.) Bot. ry. 43: 581-583. 1958. | Includes notes on economic importance. |

Penzes, A. Data to the ecology and taxonomy of the Cotimus genus. Acta Bot. Sinica 7: 167-169. 1958.*

SARGENT, C.S. Cotinus. Silva N. Am. 3: 1-6. pls. 98, 99. 1892.

ScuHuiz, A. Beitrage zur oe und Biologie der Bluthen. IJ. Ber. Deutsch. Bot. Ges. 395-409. 1892. [C. Coggygria, 395-401. |

ic

4. Metopium P. Browne, Civ. Nat. Hist. Jamaica 177. 1756.

Trees or shrubs with resin ducts in bark and wood. Leaves odd- pinnate, (1)3-—7-foliolate, stoutly petioled, usually clustered near the tips

1962 | BRIZICKY, GENERA OF ANACARDIACEAE 369

of branches, persistent; leaflets long-petioluled, + leathery, entire, lustrous, veins and veinlets prominent. Plants usually dioecious. Flowers small, usually unisexual; pedicels short, stout, and articulate near base, in loose, ascending, determinate axillary panicles; bracts and bractlets minute, in part persistent. Sepals 5, connate into a + cupuliform 5-lobed calyx; lobes broadly ovate to almost semicircular [or truncate], + un- equal, thin-leathery with scarious margins, persistent. Petals 5, twice as long as sepals, yellow-green, with brownish to blackish veins. Stamens 5, distinct, alternate with and shorter than the petals, inserted at base and between the glandlike lobes of an intrastaminal nectariferous disc, reduced and sterile in ? flowers; anthers oblong in outline, nearly as long as jor longer than] the subulate filaments, 2-locular at anthesis; pollen 3-colpate, medium sized, ellipsoidal, reticulate- striate, with round, smooth germ pores. Gynoecium 3-carpellate, rudimentary in ¢ flowers. Stigma 3-lobed; style very short, stout; ovary 1-locular by abortion, obovate to subglobular; ovule raised on a basal funicle. Drupe ellipsoidal to some- what obovoid, about 1-1.5 cm. long, glabrous, shining, orange to scarlet when ripe, tipped with the remnant of the style; exocarp membranaceous, mesocarp resinous, endocarp thin, crustaceous, all permanently united (coherent). Seed compressed, almost quadrangular in outline, the broad funicle covering one of its margins; testa thin, smooth, dark brown; endosperm scanty or wanting; embryo oriented eerccally, the cotyledons accumbent, flat, the radicle long, incurved upwards toward the hilum. TYPE SPECIES: Terebinthus Brownei Jacq. = M. Brownei (Jacq.) Urban. (An ancient name for some oriental plant [perhaps some species of Ferula|, yielding the gum resin galbanum, applied by Browne to his genus; etymology obscure.)

A genus of three species, of the West Indies, southern Florida, British Honduras, Guatemala, and southern Mexico. The West Indian Metopium toxiferum (L.) Krug & Urban, poisonwood or coral sumac (also hog gum, doctor gum), occurs in hammocks, pinelands, and coastal sand dunes in southern Florida (about as far north as Martin County) and on the Florida Keys. A large tree in hammocks, a shrub in pinelands, in its appearance the species resembles Bursera Simaruba, with which it is often confused.

The genus is very closely related to RAus, but the combination of the single style and stigma, the thin, crustaceous endocarp, the vertical embryo, and anatomical characters (diffuse-porous wood; abundant, often banded, vasicentric parenchyma; septate wood fibers) seems to support the maintenance of Metopium as distinct.

All species are poisonous, and all parts of the plants act as a contact skin-poison. Leaves and resin were formerly used in local medicine in the West Indies. The floral biology probably does not differ much from that of Rhus. Birds apparently are responsible for seed dispersal.

370 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIII

REFERENCES:

See also under family references BARKLEY (1937, pp. 303-309), ENGLER (1881, p. 378; 1883, pp. 367-369; 1892, p. 167), Hermscu (1940; 1942, pp. 139- Recorp (pp. 28, 29), SARGENT (pp. 658, 659), and West & ARNOLD (p. 112).

Harrar, E. S., & J. G. Harrar. Guide to southern trees. 712 pp. New York. 1946. [M. toxiferum, 435-437. ]

SARGENT, C. S. Rhus Metopium. Silva N. Am. 3: 13, 14. pls. 100, 101, 1892. [M. toxiferum; not Rhus Metopium = M. Brownei el

5. Rhus Linnaeus, Sp. Pl. 1: 265. 1753; Gen. Pl. ed. 5. 129. 1754.

Trees, shrubs [sometimes thorny], subshrubs, or vines climbing by aérial, clinging, adventitious roots, often pubescent with glandular and/or nonglandular hairs, some poisonous by contact. Leaves 3-foliolate | rarely palmately 5-foliolate] or odd-pinnate (occasionally even-pinnate) [or simple], petioled, deciduous [or persistent]; leaflets usually opposite, entire or toothed, sometimes lobed, sessile or + distinctly petioluled, rachis sometimes winged. Plants usually dioecious. Flowers small [to minute], usually unisexual by abortion, pediceled [or sessile], in termi- nal [and]/or axillary thyrses [or panicles], or in solitary or panicled racemes [or spikes]; bracts deciduous or persistent, bractlets 2, at base of pedicel, deciduous or persistent, or wanting. Sepals 5, connate at base [or higher], usually persistent. Petals 5, longer than the sepals, greenish- white to yellow [or rarely pink], inserted at base of disc, imbricate. Stamens 5, alternipetalous, inserted below a 5—10-lobed, annular, patelli- or cupuliform intrastaminal nectariferous disc, reduced and _ sterile in @ flowers; anthers ovate to oblong in outline, usually shorter than the + subulate filaments; pollen 3-colpate, medium sized [or small], ellip- soidal to spherical, reticulate, reticulate-striate or striate, more rarely smooth, with oblong, mostly irregular to ragged, germ pores. Gynoecium 3-carpellate, rudimentary in ¢ flowers; stigmas 3, capitellate, sometimes slightly depressed; styles 3, terminal, distinct or partly [rarely + com- pletely] connate; ovary 1-locular; ovule raised on an elongate ascendent basal funicle. Drupe subglobular to globular, 3-8 mm. [rarely over 1 cm.| in diameter, often somewhat laterally compressed, white, tan to brown, or red [or black], smooth or sometimes striate [rarely verruculose], glabrous or pubescent with glandular and/or nonglandular hairs; exocarp membranaceous, sometimes fragile and bursting irregularly; mesocarp thin or thick, + resinous, sometimes also waxy; stone + laterally com- pressed, smooth or ridged, sometimes bumpy, with thick, bony endocarp. Seed + laterally compressed, ovate to subreniform in outline; testa membranaceous; endosperm scanty or wanting; embryo large, generally transverse, with flat transverse cotyledons and a + elongate radicle in- curved upwards toward the hilum. (Including Lobadium Raf. |Schmaltzia Desv. ex Small emend. Greene], Toxicodendron Mill.). Lectotypr spEcIES: R. Cortaria L.; see E. L. Greene, Leafl. Bot. Obs. Crit. 1: 114.

1962 | BRIZICKY, GENERA OF ANACARDIACEAE 371

1905. (Name Latin, from Greek, rkous or rhoys, an ancient name for the Sicilian sumac, R. Coriaria; etymology obscure.)

A genus of over 150 species, primarily of the warm-temperate areas of both hemispheres, but extending into tropical and cold-temperate regions. The genus seems to be somewhat heterogeneous, and its taxonomy is difficult and confused. Both the four sections of Engler, based mainly on fruit morphology, and some fractions of the sections have been treated by Barkley as distinct genera. Although thorough investigations may perhaps justify at least some of these segregations, a complete study of the entire complex is necessary. The relationships appear to be reticu- late, with speciés showing transitions in respect to most characteristics used for delimiting genera, and many supposed differences are based on incomplete observations or on insufficient material. At present it appears preferable to regard Lobadium Raf. and Toxicodendron Mill. as sub- genera of Rhus.

Subgenus Ruus (subg. Sumac Torr. & Gray, 1838; § Trichocarpae Engler, 1881, in part), includes about ten species (primarily of warm- temperate North America and Eurasia) with flowers in terminal thyrses expanding after the leaves; thin, narrow, caducous bracts and bractlets; reddish drupes covered with + club-shaped hairs and sometimes also with slender red to colorless nonglandular hairs; exocarp adherent to the resinous mesocarp and both easily detachable from the smooth, bony stone; and odd-pinnate leaves. All (at least in our area) are innocuous trees or shrubs. The northeastern American Rhus typhina L. (R. hirta (L.) Sudw.); the wide-ranging R. glabra L.; the very local R. Michauxii Sarg., of the Piedmont of North Carolina and Georgia; and the widely distributed R. copallina L. (including R. leucantha Jacq. and R. obtusi- folia Small) represent the subgenus with us. Presumed natural hybrids of R. glabra & typhina have been known as R. glabra var. borealis Britt. (R. borealis (Britt.) Greene, R. pulvinata Greene), while R. Ashei

copallina (including vars. copallina, latifolia Engler, leucantha (Jacq.) DC., obtusifolia (Small) Fern. & Grisc., and lanceolata Gray [the last supposedly beyond our range] is much needed.

Subgenus Lopapium (Raf.) Torr. & Gray, 1838 (Lobadium Raf., 1819; Schmalizia Desv. ex Small emend. Greene, 1905; Rhus § Trichocarpae Engler, 1881, in part; including Styphonia Nutt. and Rkoeidium Greene) includes species with flowers in short, dense, spikelike racemes usually forming terminal panicles expanding before the leaves (more rarely in axillary and/or terminal panicles expanding after the leaves); leathery, broad, persistent (rarely caducous) bracts and bractlets; drupes essentially as in subg. Ruvus (but in some with the inner layers of mesocarp, striate by numerous resin ducts, remaining attached to the stone); and leaves 3-foliolate (in ours). The nearly 35 species, all innocuous shrubs [or trees], are centered in Mexico, extending north to Alberta and Quebec

oie JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIII

and south to Costa Rica. The northeastern RAus aromatica Ait. var. aromatica (Schmaltzia crenata (Mill.) Greene), fragrant sumac, with precocious flowers, is of frequent occurrence on calcareous soils through our area southward to northwestern Florida and westward and northward beyond our limits. The var. serotina (Greene) Rehd. (R. ¢trilobata Nutt. var. serotina (Greene) Barkley), with coetaneous flowers, occurring north and west of our region has been recorded from Arkansas. Hybridization between R. aromatica and the largely western American R. trilobata, as well as between some other species of the subgenus in areas of overlap, is believed possible (Barkley, 1937).

Subgenus ToxIcoDENDRON (Mill.) K. Koch, 1853, emend. Gray, 1856 ($ Venenatae Engler; § Trichocarpae Engler, as to R. trichocarpa Miq.; Toxicodendron . Mill., 1754), comprises species with flowers in axillary thyrses expanding after the leaves; caducous bracts; no bractlets; vel- low-white to -brown, usually smooth drupes, glabrous or pubescent with nonglandular hairs; usually thin, fragile exocarp which finally breaks, exposing the whitish, waxy mesocarp striate with black resin ducts; stones bony with a few riblike longitudinal ridges (or more rarely smooth); and leaves 3-foliolate or odd-pinnate [rarely simple]. The nearly 15 species, including trees, shrubs, or woody vines climbing by aérial roots and all producing a contact dermatitis, are largely of temperate North American— eastern Asiatic distribution, but with at least R. striata Ruiz & Pavon in tropical South America (Colombia, Venezuela, Peru). Our representatives are Rk. Vernix L. (Toxicodendron Vernix (L.) Kuntze), poison sumac, a shrub or small tree with odd-pinnate leaves, chiefly of the Coastal Plain: R. radicans L. (T. radicans (L.) Kuntze), poison ivy, 2n = 30; and R. Toxicodendron L. (T. Toxicodendron (L.) Britt., 7. quercifolium (Michx.) Greene), poison oak. RAus radicans is both widely distributed and vari- able, and a number of varieties and forms have been recognized. According to Gillis (1962, p. 21), the eastern Asiatic R. orientalis (Greene) Schneider is conspecific with R. radicans. The external and internal morphology of R. diversiloba Torr. & Gray, western poison oak, 2n = 30, is better known than that of other species.

The genus seems to be rather uniform in regard to floral morphology and biology. Flowers are usually unisexual and the species dioecious, but the occasional occurrence of bisexual flowers, and consequent polygamy, is probable. Various Hymenoptera (especially short-tongued bees) and Diptera have been recorded as the most frequent visitors. Chalazogamy has been found in all the species thus far investigated. The chromosome number 2m = 30 has been recorded for three or four species of subg. Toxt- CODENDRON, 2” = 32 for the North African Rhus oxyacantha Schousb. ex Cav. (Searsia Barkley; § Gerontogeae Engler). Seed dispersal by animals, especially by birds, apparently may be assumed for the genus as a whole.

Dried leaves of Rhus glabra, R. typhina, and R. copallina, and espe- cially of R. Coriaria, are important sources of tannin. Leaf galls produced on R. javanica L. (R. chinensis Mill., R. semialata Murr.) vield tannin

1962 | BRIZICKY, GENERA OF ANACARDIACEAE Bike.

which is also sometimes used in making ink. The Asiatic R. verniciflua Stokes and sometimes R. succedanea L. are sources of natural lacquer. Commercial vegetable wax is obtained from the mesocarp of fruits of the last and some allied species.

omacraces e large number of references has been reduced here primarily to those ae of general interest or dealing specifically with the southeastern United States. Under family reference see BARKLEY (1937, pp. 312-441), ENGLER (1881; 1883, pp. 371-452; 1892, pp. 167-172), HermscH (1940; 1942, pp. 139-141), Recorp (pp. 3034. 43-45), SARGENT (pp. 660-665), and West & ARNOLD (pp. 113, 114). Under Cotinus see MzEHAN (1874). ASHE, W. W. Magnolia cordata and other woody plants. Bull. Torrey Bot. Club 54: 579-582. 1927. [“ Schmaltzia “shel hyb. nov.” (= Rhus Ashei),

= Baer, J. E., & A. F. Sievers. Sumac; its collection and culture as a source of tannin U.S. Dep. Agr. Prod. Res. Rep. 8. 14 pp. 1957.* Bark ey, F. A. Five-leaved poison ivy. Biologist 17: 122-124. 1936.* . Studies in the Anacardiaceae. III. A note concerning the status of Rhus pulvinata Greene (R. glabra & typhina Koehne). Am. Midl, Nat. 1938

19: 598-600.

——.. Schmaltzia. ee 24: 647-665. 1940.

& E. BaRKLEY. A short history of Rhus to the time of Linnaeus.

Ibid. 19: 265-333. 1938.

Boyp, I. L. Germination tests on four species of sumac. Trans. Kan. Acad. Sci. 46: 85, 86. 1943

Crausen, R. T. Northeastern limits of the known range of Rhus Toxico- dendron. Torreya 41: 58, 59. 1941. [R. radicans.]

CopELAND, H. F., & B. E. Dovet. Some features of the paar of Toxicoden-

0.]

dron diversiloba. Am. Jour. Bot. 27: 932-939. 1941. [2n

Crooxs, D. M. L. W. Kepuart. Poison-ivy, poison-oak, i poison sumac: identification, precautions, eradication. U. S. Dep. Agr. Farmers’ Bull. 1972: 1-30. 1958.

Dawson, a R. The toxic principle of poison ivy and related plants. Rec. Chem. Prog. 15: 39-53. 1954.*

: ci chemistry of poison ivy. Trans. N. Y. Acad. Sci. II. 18: 427- 443. 1956.*

Driers, L. Die Epharmose der Vegetationsorgane bei Rhus § Gerontogeae Engl. Bot. Jahrb. 24: 568-647. pl. 14. 1898

Duncan, W. H., & T. J. Jones. Poisonous plants of Georgia. Bull. Univ. Ga. School Veterin. Med. 49(13): i-iv, 1-46. 1949. [R. Vernix, 17-20; R. radicans, 20. |

FERNALD, M. L. Another century of additions to the flora of Virginia. Rhodora 43: 559-630. pls. 672-692. 1941. [“Some varieties and forms of Rhus radicans and R. To: einen. 589-599. pls. 683-685; “The variations of Rhus aromatica in the Gray’s Manual range,” 599-603. pls. 686, 687. |

—— &L. Griscom. Three days of botanizing in southeastern Virginia. /bid. 37: 167-189. pls. 345-351. 1935. [Variations of R. copallina, 167, 168.]

FRANKLIN, J. J. Another plant to be labeled poisonous. Pl. Gard. 8: 271. 1952. [R. trichocarpa. |

374 JOURNAL OF THE ARNOLD ARBORETUM [ VOL, XLIII

Gitpert, E. F. Phenology of sumacs. Am. Midl. Nat. 66: 286-300. 1961.* [R. glabra and R. typhina. See also Diss. Abs. 20: 1553, 1554. 1959.] Giiuis, W. T. Taxonomic problems in poison ivy. Pap. Mich. Acad. Sci. I. 45: 27-34. 1960. . Poison-ivy and the poison-oaks. Cranbrook Inst. Sci. News Lett. 30: 98- 107. es

on-ivy in northern Michigan. Mich. Bot. 1: 17-22. 1962. [R. asda radicans var. Rydbergii, delimitation, range in Michigan. |

GREENE, E, Segregates of the genus Rhus. Leafl. Bot. Obs. Crit. 1: 114-144. 1905.

——. A study of Rhus glabra. Proc. Wash. Acad. Sci. 8: 167- 196. 1906. [Includes descriptions of 28 segregates, 24 proposed as new spp.

Grimm, J. Entwicklungsgeschichtliche Untersuchungen an RAus und Coriaria, Flora 104: 309-334. pls. 10, 11. 1912. [R. “Toxicodendron’” (R. radicans ?), 2n = 30, embryology, 311-323; R. glabra and R. typhina, chalazogamy, parthenocarpy, 323-325.

Harapa, M. On the distribution and construction of the resin canal[s] in Rhus succedanes, Bot. Mag. Tokyo 51: 846-856. 1937.

stigation on the development of resin canals and the formation of

ne in a mesocarp of fruits of the Rhus plants found in Japan. Bull. Sci. Fak. Terkult. Kyushu Univ. 8: 179-191. 9.*

Kexar, S. S. Embryology of Rhus mysurensis Heyne. Jour. Indian Bot. Soc. 37: 114-122. 1958.

LittLe, E. L., Jr. Miscellaneous notes on nomenclature of United States trees. Am. Midl. Nat. 33: 495-513. 1945. [RAus, 498-500. |

McFappen, G. H., & R. L. McMurry. Rhus glabra. Am. Jour. Pharm. 109:

McNair, J. B. Fats from Rhus [Malosma] laurina and Rhus [ Toxicodendron | diversiloba, Bot. Gaz. 64: 330-336. 1917. [Fats identical; includes fruit morphology of R. diversiloba.

Secretory canals of Rhus diversiloba. Ibid. 8: 179-191. pls. 3, 4.

ZA,

The taxonomy of poison- ee with a note on the origin of the generic

name. Publ. Field Mus. Bot. 4: 55-70. pls. 14-24. 1925.

Five-leaflet poison oak. a Torrey Bot. Club 63: 473-476. 1936. [| R. diversiloba, pinnate-leaved form. |

Meyer, A. Ueber die Entwickelung des Wachses der Frucht von Rhus Toxico- dendron Mich. Arch. Pharm, III. 15: 514-516. 1879.*

Mosivus, M. Der japanische Lackbaum, Rhus vernicifera DC. Abh. Senckenberg. Naturf. Ges. 20: 203-247. pl. 1. 1899. [Morphology and anatomy of R verniciflua. See also Ber. Deutsch. Bot. Ges. 15: 435-441. 1897. ]

MorceneverR, W. Untersuchungen tber die Beseitigung der Keimhemmung beim Samen des Hirschkolbensumachs (Rhus typhina). Arch. Forst. 5: 203-242. 1956.*

MuENSCcHER, W. C., & J. M. Kincssury. Poison ivy and poison sumac. Cor- nell Ext. Bull. 191: 1-12. 1960.

RosBertson, C. Flowers and eS 17. Bot. Gaz. 22: 154-165. 1896. [Rhus, 159-164. ]

Flowers and insects. 221 pp. Carlinville, Illinois. 1928. [R. aromatica (“R. canadensis”) R. copallina, R. glabra, R. radicans (“R. Toxicoden- dron’’), list of insect visitors, 24-26. SARGENT, C. S. Rhus. Silva N. Am. 3: 7-11, 15-28. pls. 102-109. 1892.

1962] BRIZICKY, GENERA OF ANACARDIACEAE 375

. Rhus Michauxii. Garden Forest 8: 404. 1895. [Considered to be very

poisonous, but see Warren.

. Rhus trichocarpa. Ibid. 10: 383, 384.

SCHNEIDER, C. K. Illustriertes Handbuch der Dine ances Vola2.. 21070" pp: Jena. 1907-1912

SCHONLAND, S. The South African species of Rhus L. Bothalia 3: 1-115. 1930.

Sizer, I. W., & C. E. Proxescu. Inactivation of the irritant toxicants of poison i and related compounds by tyrosinase. Science 101: 517, 518. 1945.

TapaTa, S. Ueber die Friichte und Keimpflanzen von Rhus succedanea, L. Jour. Coll. Sci. Univ. Tokyo 23(1): 1-11. 1 pl. 1907

WARAWDEKAR, S. S. Analysis of poison ivy fruit fat. Diss. Abs. 17: 2164. 1957.*

WarreEN, L. E. Rhus AC POE: A nonpoisonous plant. Am. Jour. Pharm. 1910: 499-506. 1910

376 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLII

COMPARATIVE ANATOMY OF THE LEAF-BEARING CACTACEAE, VI THE XYLEM OF PERESKIA SACHAROSA AND PERESKIA ACULEATA

I. W. Batvey !

THE MOST PRIMITIVE living representatives of the Cactaceae have been considered to be Pereskia sacharosa Griseb. and P. aculeata Mill. Such a conclusion is based largely upon the claim that the flowers of these species have “superior” ovaries (Berger, 1926; Buxbaum, 1953). It is of interest to determine whether evidence from other parts of the plants supports such a phylogenetic generalization. Evidence from the xylem is considered in this paper. Additional evidence from other parts of the plants will be considered subsequently.

PERESKIA SACHAROSA

During its earlier stages of growth, this species is considered to be a shrub which ultimately may become a small tree eight meters high. As in the case of other wide-ranging putative species of Pereskia, plants from different localities exhibit more or less conspicuous differences in their leaves, general habit of growth, spination, and floral characters. At present, it is uncertain whether such differences are produced b different environmental influences or are due to genetic differences at racial or specific levels.

My anatomical specimens of this species have been obtained from the following sources: 1) From a wild plant collected by Carenzo and Legname in Jujuy Province, Argentina; 2) from a wild plant collected by Cardenas in the Department of Cochabamba, Bolivia; 3) from a plant growing in the Missouri Botanical Garden; and 4) from a plant growing at the New York Botanical Garden. The ranges of variability in the xylem of the four plants are so closely similar as to provide no reliable anatomical criteria for differentiating one plant from the others.

In the outer secondary xylem of larger stems the vessels occur singly and in small clusters (Fics. 1, 3), but tend at times in their overall arrangement to exhibit more or less conspicuous concentric patterns (Fic, 1). The wood parenchyma is scanty paratracheal (Fic. 3), having lignified walls. The multiseriate rays are composed of cells of varying size, form, and orientation which likewise have comparatively thick lignified walls

'This investigation was financed by a grant from the National Science Founda- tion. I am indebted to the American Philosophical Society for the loan of a Wild microscope.

1962 | BAILEY, LEAF-BEARING CACTACEAE, VI 377

(Fics. 3, 5). Where crystals of calcium oxalate are present in ray cells they occur singly or as several independent ones, druses being absent. The distribution of vessels and wood parenchyma in comparable secondary xylem a roots is similar (Fic. 2) but the multiseriate rays are broader (Fics. 2, 6). The rays, as in those of the stem, may be composed, ate their radial extension, of cells with lignified walls. In some cases, however, the first-formed mmultiseraite rays in the center of the root are composed at first of cells having thin, unlignified walls. These cells differ from subsequently formed lignified ones in containing isotropic granular contents, rather than starch.

It is evident that the secondary xylem in stems and roots of Pereskia sacharosa is of a phylogenetically highly evolved structure such as occurs in trees and shrubs of normal form in a number of other dicotyledonous families. For a discussion of pitting, perforation plates, and other struc- tures visible under higher magnification, the reader is referred to the fourth paper of this series (Bailey & Srivastava, 1962

There are certain ontogenetic changes in rays that should be considered in dealing with the secondary xylem of the leaf-bearing Cactaceae. The multiseriate rays of dicotyledons commonly exhibit significant changes in passing from the first-formed to the later-formed secondary xylem (Chattaway, 1933; Barghoorn, 1940, 1941a). The first-formed part of multiseriate rays adjacent to gaps in the primary body (particularly of stems having elongated internodes) frequently tend to be narrow and vertically extensive as seen in tangential longitudinal sections (Fic. 8). During subsequent radial extension of such rays they tend to become wider and dissected sooner or later into vertically less extensive parts which ultimately may assume fusiform outlines (Fic. 6). Of the various cyto- logical changes involved in the modification of such rays the transforma- tion of ray initials into fusiform cambial initials is particularly significant. Furthermore, the dissected parts of the original rays tend to be laterally displaced during increase in girth of the cambium. In addition, the cells of the ravs generally vary more or less in form and orientation during successive modifications of the multiseriate rays, the cells in the first- ormed part of the rays tending to have a longer vertical axis (i.e., being “erect’’), whereas in subsequently formed parts of the rays they become more or less isodiametric or even radially extensive (i.e., ‘‘procumbent”’). Such changes in multiseriate rays vary in degree, and may be precocious and relatively abrupt, gradual, or considerably delayed, varying at least in part with different rates of growth and the enlargement of stems.

In roots of Pereskia sacharosa, the first-formed multiseriate rays broaden conspicuously during their radial extension outward (Fic. 2). In their first-formed part they are not only much narrower, but also more ex- tensive vertically. During their radial extension outward they tend to become dissected into lower derivatives which assume fusiform outlines as seen in tangential longitudinal sections (Fic. 6). In stems from some parts of a mature plant, the first-formed parts of the rays are narrower and more extensive vertically, but subsequent widening and dissection of

378 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLII

the rays is much less conspicuous than in roots (compare Fics. | and 2, 5 and 6). In other stems from the same plant, the first-formed rays of the secondary xylem are broader, subsequently becoming narrower (Fic. 4) and then broadening somewhat in the later-formed xylem (Fic. 3).

Such variations in the form of the multiseriate rays appear to be corre- lated with differences in the development of the primary body. The diameter of the primary vascular cylinder and pith commonly varies in different stems of the same plant from three millimeters to one centimeter. As seen in transverse sections, the two extreme forms of eusteles may be composed of the same number of fascicular and interfascicular parts. Although there is some variation in the breadth of fascicular parts, the conspicuous differences in circumference of the eusteles appears to be due largely to differences in the width of interfascicular parts (i.e., paren- chymatous gaps). In the case of the smaller young stems, cambial activity develops precociously in the parenchyma of the narrow gaps, and the multiseriate rays of the secondary xylem are relatively narrow when first formed. On the contrary, in the larger young stems the parenchymatous gaps tend to widen by division and transverse expansion of their con- stituent cells for a considerable time after cambial activity is initiated in the fascicular parts of the eustele (Fic. 4). It should be noted in this connection that the widening of the interfascicular parts and the delayed initiation of cambial activity within them leads, not only to a conspicuous increase in the circumference of the eustele, but also permits a marked increase in the diameter of the pith. Furthermore, when multiseriate rays of the secondary body are initiated they tend at first to be considerably broader than in the case of smaller stems (Fic. 4).

In both stems and roots of Pereskia sacharosa, the parenchymatous cells of the first-formed part of the multiseriate rays are of “erect” form but are shorter than the fusiform initials of the cambium. Subsequent changes in the size, form, and orientation of the ray cells, as seen in radial longitudinal sections of the xylem, are precocious and abrupt. Most of the cells become more or less isodiametric, except on the sides of the rays and their upper and lower margins where they may retain an erect form. In addition, some radial tiers of slender cells of procumbent orientation tend to occur within the rays.

PERESKIA ACULEATA

This species is widely distributed in the West Indies and in eastern and northern South America. Because of the edibility of its fruits and leaves, its ornamental value in gardens, and its use in hedges, its original range has been considerably extended by man. It is therefore difficult at times to determine with certainty whether plants now growing in the wild (e.g., Florida or Mexico) are native or were introduced and escaped from cultivation.

The species is commonly described as a shrub, frequently producing long clambering branches and forming vines three to ten meters long

1962 | BAILEY, LEAF-BEARING CACTACEAE, VI 379

(cf. Britton & Rose, 1919). The larger, older stems may have a diameter of two to three centimeters with a primary vascular cylinder and pith five to six millimeters in diameter. The ultimate terminal branches of mature plants are, in general, more numerous and more slender than those of Pereskia sacharosa, grading down in many cases to as little as two millimeters in diameter with correspondingly slender primary vas- cular cylinders and pith. Furthermore, in branches of comparable diameter, the eusteles of P. aculeata are composed of fewer fascicular strands and oe ane broader parenchymatous gaps. This conspicuous difference

n the primary bodies of the two species appears to be correlated with ce tendency of P. aculeata to have two-trace unilacunar nodes, whereas P. sacharosa commonly has four to six vascular strands at its unilacunar nodes. (For illustrations see Bailey, 1960.)

The young branches of Pereskia aculeata are commonly of two different forms: slender woody ones and broader, very succulent ones, referred to in Boke’s (1954) developmental investigations as long shoots and spur shoots respectively. In the former stems cambial activity and the forma- tion of secondary xylem are precocious, whereas in the latter stems the formation of secondary xylem is delayed and the thickness of the pith and cortex is relatively greater. At times, the most slender woody stems bear short, broad, highly succulent laterals. In such cases the slender woody stems have long internodes, whereas the succulent laterals have short internodes. These abrupt differences in external form and internal structure resemble those between long shoots (pycnoxylic”) and short shoots (manoxylic *) that occur in Ginkgo biloba L. (Gunckel & Wetmore, 1946a, 1946b) in Cercidiphyllum japonicum Sieb. & Zucc. (Titman & Wetmore 1955), and a number of other plants.

The structure of the secondary xylem of Pereskia aculeata varies markedly in different parts of a mature plant. As so frequently happens in scandent representatives of other dicotyledonous families, e.g.. Hippo- crateaceae, Icacinaceae, Schisandraceae, etc., many stems exhibit more or less abrupt transitions from first-formed, denser, more nearly normal secondary xylem to phylogenetically highly modified, softer, more porous forms of tissue. The denser, earlier-formed secondary xylem (illustrated in Fic. 7) resembles the secondary xylem of P. sacharosa (Fic. 7). The vessels, allowing for higher magnification, are of comparable size, form, and distribution. The wood parenchyma is scanty paratracheal and has strongly lignified walls. The multiseriate rays, although somewhat nar- rower than those of P. sacharosa, have thick, strongly lignified cell walls. On the contrary, the later-formed secondary xylem has much enlarged vessels, as well as radial rows of much smaller ones. In addition to scanty wood parenchyma about vessels in the denser parts, it has arcs or con- centric zones of unlignified parenchyma. The cells of the broadened rays (compare Fics. 2 and 3) are very thin, unlignified, and many of them contain druses (Fic. 16). Furthermore, the number of thick-walled libri- form fibers is greatly reduced.

? Using these terms as redefined by Titman and Wetmore.

380 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIII

In other stems, even from the same plant, the transition from normal dense to modified secondary xylem occurs at varying distances from the pith. The transition to xylem containing huge vessels may be abrupt or gradual (Fic. 11). Furthermore, the volume and distributional patterns of the unlignified ® parenchyma varies markedly from stem to stem. In the xylem illustrated in Fic. 10, the unlignified parenchyma occurs in broad concentricities alternating with zones of strongly lignified tissue. The relatively narrow multiseriate rays, unlike those of Fic. 7, have ligni- fied walls, except where they pass radially through the broad zones of unlignified parenchyma. In contrast to this, the outer secondary xylem in Fic. 11, although exhibiting a conspicuous enlargement of vessels has unlignified parenchyma that is confined to narrow arcs and concentricities.

The roots of Pereskia aculeata superficially resemble those of P. sacha- rosa, having numerous rapidly broadening multiseriate rays (compare Fics. 2 and 14), but the ray cells of P. aculeata, many of which contain druses of calcium oxalate, differ in having thin, unlignified walls (Fic. 14). It should be noted in this connection that the first-formed ray cells in the center of some roots of P. sacharosa, although having thin unlignified walls, contain no druses and little if any starch. Furthermore, in roots, as in stems (Fics. 8, 9), of P. aculeata, the multiseriate rays tend to remain vertically extensive, and in passing radially outward do not exhibit precocious and conspicuous dissection into laterally displaced parts as illustrated for P. sacharosa in Fic. 6. The unlignified rays of both te and stems of P. aculeata, not only contain abundant druses (Fic.

16) but also at times varying numbers of mucilage cells (Fic. 17).

In many stems of Pereskia aculeata, owing to the breadth of the parenchymatous gaps in the eustele, the multiseriate rays tend to be broader when first formed on in the subsequently formed secondary xylem. As indicated in Fic. 11, this reduction in ray width resembles that which occurs in very stout young stems of P. sacharosa (Fic. 4). However, in the case of P. aculeata, the cells of the first-formed ray tissue tend to resemble fusiform cambial initials in their vertical exten- sion (Fic. 12). In other words, the first-formed ray initials of the cambium resemble fusiform initials in size and orientation. Subsequent transverse divisions of the ray initials lead to the formation of less ‘erect’ ray cells (Fic. 13) and ultimately to more or less isodiametric or some- what procumbent ones. It should be noted in this connection that simi- larities between ray initials and fusiform initials occur in the first-formed cambium of some other dicotyledonous families (Barghoorn, 1941). Tyloses are of sporadic occurrence in the vessels of Pereskia aculeata. n some stems, they are not only abundant but contain starch (Fic. 15).

—

—_

“The unlignified primary walls of leaf-bearing Cactaceae give a red coloration

sections stained ae Haidenhain’s perenne and safranin. Therefore, in testing for lignificati it is advisable to use the phloroglucin- eae and Maule reactions where unlignified walls remain ieee as in Fics. 10 and

1962] BAILEY, LEAF-BEARING CACTACEAE, VI 381

DISCUSSION

In the fourth paper of this series (Bailey & Srivastava, 1962) we dis- cussed the finer structural details in derivatives of the fusiform initials of the cambium. We found that the ranges of variability of potential diagnostic criteria in the leaf-bearing genera Pereskia, Pereskiopsis, and Ouiabentia are extensive, not only in different collections of the same taxon and in the same clone grown under different environmental influ- ences, but also in different parts of the same mature plant. It is evident that the grosser anatomical features of Pereskia aculeata (e.g., size and distribution of vessels, abundance, distributional patterns. and internal structure of wood parenchyma, and the size, form and internal structure of the multiseriate rays) likewise vary markedly even in different parts of a single plant.

Anatomical surveys of the dicotyledons as a whole demonstrate that the xylem of vines and lianas commonly exhibits trends of highly ad- vanced phylogenetic modification, i.e., in comparisons with the structure of related trees and large woody shrubs of the same genus, family, or order. This obviously occurs in Pereskia aculeata in contrast to P. sacha- rosa. Thus, although P. sacharosa may have retained a form of xylem characteristic of ancestral leaf-bearing cacti, the internal structure of stems and roots of P. aculeata negates any possibility of considering this species to be the most primitive living representative of the Cactaceae. Therefore, in searching for such a primitive representative, it will be of interest in succeeding papers of this series to compare the internal struc- ture of P. sacharosa with that of other arborescent and woody, shrubby species of Pereskia.

The tendencies in the xylem of Pereskia aculeata toward the formation of parenchymatous cells having thin, unlignified walls, with concomitant inclusions of druses, are extensively developed in roots of Andean pereskias and in stems of Pereskiopsis and Quiabentia. They seem to be correlated in some manner with increasing succulence and merit detailed investiga- tion. Furthermore, the tendency in some stems of both P. sacharosa and P. aculeata toward increase in circumference of the primary vascular cylinder and in diameter of the pith after cambial activity is initiated in the fascicular parts of the stele is a phenomenon which becomes greatly exaggerated in many representatives of the Cactaceae. It is a conspicuous feature in dealing with the larger basal stems of pereskias from Southern Mexico and Central America and in stems of certain species of Pereskiopsis and Ouiabentia.

LITERATURE CITED

Bartey. I. W. 1960. Comparative anatomy of the leaf-bearing Cactaceae, L. Foliar vasculature of Pereskia, Pereskiopsis and Quiahentia. Jour. Arnold Arb. 41: 341-356.

L. M. Srivastava. 1962. IV. The fusiform initials of the cambium

and the form and structure of their derivatives. /bid. 43: 187-202.

382 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLHI

BarcHoorn, E. S. 1940. The ontogenetic development and_ phylogenetic specialization of rays in the xylem of dicotyledons, I. The primitive ray structure. Am. Jour. Bot. 27: 918-928

ky ce of the multiseriate and uniseriate rays. Am. Jour.

Bot. 28: 273-282.

. 1941b. III. The elimination of rays. Bull. Torrey Club 68: 317-325.

Bercer, A. 1926. Die Entwicklungslinien der Kakteen. Gustav Fischer, Jena.

Boxe, N. H. 1954. Organogenesis of the vegetative shoot in Pereskia. Am. Jour. Bot. 41: 619-637.

BRITTON, N. L., & J. N. Rose. 1919. The Cactaceae. Vol. I. Carnegie Inst.,

BuxpauM, F. 1953. Morphology of Cacti. Sect. IJ. The flower. Abbey Garden Press, Pasadena, Calif.

Cuatraway, M. 1933. Ray development in the Sterculiaceae. Forestry 7: 93-108.

GuNCKEL, J. E., & R. H. Wermore. 1946a. Studies of development in long shoots and short shoots of Ginkgo biloba L. 1. The origin and pattern of development of the cortex, pith and procambium. Am. Jour. Bot. 33: 285-295.

. II. Phyllotaxis and the organization of the primary vascular system;

Laie phloem and xylem. /bid. 532-543.

AN, P. W., & R. H. Wermore. 1955. The growth of long and short

pens in Corgi phuluti: Am. Jour. Bot. 42: 364-372

EXPLANATION OF PLATES

PLATE I Fics, 1-3. Transverse sections - the secondary gen: of Pereskia sacharosa. 1, Outer xylem from stem 7 cm. in diameter, & 11. 2, Xylem ar root 3 cm.

in diameter, X 11. 3, Xylem of Fic. 1 more highly acute’. x 8

PLATE II Fics. 4-6. Transverse and tangential longitudinal sections of Pereskia sacha- rosa showing structure of multiseriate rays. 4, Transverse section of inner xylem from a stem 7 cm. in diameter, * 34. 5, Tangential section from outer- most xylem of Fic. 1, & 43. 6, Tangential section of outermost xylem of Fic. 2, x 43.

PLATE III . 7-9, Transverse and tangential longitudinal sections of Pereskia aculeata from stem 2.5 cm. in diameter. 7, Transverse section of xylem, X 34. 8, Tangential section of denser xylem, X 43. 9, Tangential section of outer, softer xylem, * 43.

PLATE IV

‘ics. 10-13. Transverse and longitudinal sections of the xylem from stems of eae aculeata. 10, Transverse section treated with phloroglucin-HCl, showing concentric zones of unlignified parenchyma (white), X 34. 11, Trans- verse section of inner xylem, showing ace in ray structure and narrow arcs of thin-walled wood parenchyma, * 43. 12, Tangential section of first-formed secondary xylem, * 88. 13, Radial Bere showing transitional changes in the height of ray cells, & 88.

1962 | BAILEY, LEAF-BEARING CACTACEAE, VI 383

PLATE V Fics. 14-17. Transverse and tangential longitudinal sections of Pereskia aculeata. 14, Transverse section of root, treated with phloroglucin-HCl, showing unlignified multiseriate rays (white), X 34. 15, Transverse section of vessel, showing starch in tyloses, X 107. 16, Part of Fic. 9, more highly magnified, showing druses in multiseriate ray, X 107. 17, Tangential section of xylem, showing mucilage cells (arrows) in multiseriate ray, X 107.

Jour. ARNOLD Ars. VoL. XLIII

+e *

4,

6¢

ma) ewe?’ ‘ene a* 94 vse! vee Natt o'e que M MASALA > Xd Paree eg & ot, s00 “ | Se ketones eo! “9. Oa, Fx ey 0 9 ‘se, That e Re Al “ed

“0088 Ry F Pee ey yee

Ud EA | a Ore @,;, ‘3 ae Cees er

; = é 4

4s e ote ONS!

Spey. Seite tes

Ener

en NERO Cat

: OSI Beak: a

be die Fy } 6. VOM Oe are Ses 16-@ 9 bak i as shale RN aE Ly : 4 8.06 0 hey baie tae Por i RR oagg lt ; sine: ry soe SPEER) LS, ented ‘ae. ered LED 3 Tie SO A IN Diem NE BOTY “AO NR SR

BAILEY, LEAF-BEARING CACTACEAE, VI

PLATE II

Jour. ARNOLD Ars. VoL. XLII

‘©; —— ILS

Es

i : etienece SSeS:

=

iF AY

sesegs ame

Bes

dy

Ss

VI

)

-BEARING CACTACEAE

, LEAF

BAILEY

Jour. ARNOLD Ars, VoL. XLIII

PLaTE III

cat.

a tj

\ iy ire

ee ee

= Peon a, few te con a a ae wwe SES bee

a a ”

sod

ee we

>

es

Se ee om

r

Jour. ARNoLtp Ars. VoL. XLII PLATE IV

ee om ten To Soe eS eda <

Raemeeel fran an

i

i Trout | in oan

fA

re)

iN ss tel oli m (rary

\ ECS

7a

H

bake oe pat mwas 2 os wer: "9

bad he ,

Shed! BEA a

4 Piety) \) “3

e rt eatery et ae Oa war ee: “ratterae

nest AE cae aera

es.

MN il i

Ed

BAILEY, LEAF-BEARING CACTACEAE, VI

PLATE V

Jour. ARNoLD Ars. VoL. XLITI

BAILEY, LEAF-BEARING CACTACEAE, VI

1962 | HOWARD, GUTTIFERAE 389

SOME GUTTIFERAE OF THE LESSER ANTILLES

Ricuarp A. Howarp

IN THE cCouRSE of preliminary work leading towards the preparation of a flora of the Lesser Antilles, I have encountered several nomenclatural problems in the genera Clusia, Calophyllum, and Rheedia. The following notes are presented as discussion and clarification of these problems.

I have received most generous assistance from Miss Alicia Lourteig, Mr. George Proctor, and Mr. James Dandy in checking historic specimens and illustrations in European herbaria. It is also with gratitude that I acknowledge the support of this work through a grant from the National Science Foundation.

CLUSIA

Britton and Millspaugh (Bahama Fl. 281. 1920) and Hitchcock and Green (Prop. Brit. Bot. 160. 1929) selected Clusia major L. as the type species of the genus. More recently the /ndex Nominum Genericorum formalized this status. Unfortunately, Clusia major, proposed by Linnaeus in the first edition of Species Plantarum, was abandoned by him in the second edition and has not been used in floras or monographs in the intervening 209 years. In the interest of stability of well known specific names, one questions the value and the necessity of resurrecting such an epithet. There is, however, no option in the present rules of nomenclature, and so Clusia major L. must replace the better known Clusia alba Jacq. In the process of this investigation, it became apparent that the nomen- clature of nearly every species of Clusia in the Lesser Antilles was in- volved, and several others in the Greater Antilles presented one or more additional problems. These will be discussed in the following paragraphs.

Clusia major L.

In the first edition of Species Plantarum o 509. 1753), Linnaeus described two species of Clusia, namely, C. major and C. minor. The protologue of C. major, with the modern ees ts of its supporting literature given in brackets, is the following:

—

CLUSIA foliis aveniis. major. Clusia flore albo, fructu coccineo. Plum. gen. 21. [C. alba Jacq. | Cenchramidea arbor saxis adnascens, obrotundo pingui

folio, fructu pomiformi. Pluk. alm. 92 t. 157. f. 2.

[C. plukenetit Urb. | Terebinthus folio singulari non alato rotundo succu-

Jento, flore pallide luteo. Sloan. jam. 167. hist. 2.

p. 97 t. 200. f. 1 [C. flava Jacq.] Raj. dendr. 51 [unknown]

390 JOURNAL OF THE ARNOLD ARBORETUM [ VOL, XLIII

8. Clusia flore roseo major, fructu subviridi, Plum. gen. 21. [C. rosea Jacq. |

y. Clusia alia minor: flore albo, fructu virescente. Plum. gen. 21. [possibly C. plumieri, or not known according to Planchon & Triana’

In considering this species, the specimens in the Linnaean herbarium are of little assistance. There are five sheets included under the name of Clusia. Sheet 1224.1 does not appear to be a Clusia. Sheet 1224.2 is annotated, “Clusia alba H. Aublet vix Linnaei J. E. S.”’ The specimen is probably a Clusia. Sheet 1224.3 bears only two detached leaves, One is obviously sessile and is probably Clusia mangle L. C. Rich. ex Planch. & Triana. The other leaf has a petiole and is annotated as Clusia flava ex Herb. Jacq. The remaining two sheets, 1224.4 and 1224.5, represent Mammea americana and Chrysobalanus icaco respectively.

In 1760, in his Enumeratio Systematica Plantarum, Jacquin described four species of Clusia but made no reference to Species Plantarum. The species are briefly but validly described and each carries a reference to an illustration. Clusia rosea has the reference “Catesb. Car. 2. t, 99.”: C. alba carries “Plum, ic. 87. f. 1.°; C. flava bears “Sloane hist. Jam. 2. t. 200. f. 1.”; and C. venosa is supported by a reference to “Plum. ic. 87. f. 2.” The species published in Jacquin’s Enumeratio were normally based on material he collected in the West Indies. Mr. Dandy has pointed out in correspondence that Jacquin sometimes cited published figures from Browne, Sloane, Catesby, and others, but in doing this his intention was to provide the reader with what he supposed was an illustration of his own plant. In his later Selectarum Stirpium Americanarum Historia the Same species were usually described at length, more complete references were given, type localities were cited, and often a figure drawn from his own material was published.

Unfortunately, Jacquin’s herbarium was badly damaged while in the West Indies. Although the remains were purchased by Sir Joseph Banks and are in the herbarium of the British Museum, there is no material of Clusia available. Jacquin, in his Selectarum, in 1763, did publish complete descriptions of the four species of Clusia, along with illustrations of two of them, previously briefly described in the Enumeratio of 1760. It is necessary to accept 1760 as the date of publication of Jacquin’s species but to typify them with the data and illustrations of 1763. Three of Jacquin’s species, C. alba, C. rosea, and C. flava represent segregates from Linnaeus’ C. major. The fourth species, C. venosa, appears to be identical with Linnaeus’ C. minor and will be discussed later. In the Selectarum in 1763 Jacquin gave for his C. alba the basic polynomial “Clusia foliis aveniis” and the reference “Linn. sp. pl. I. p. 509,” as well as the reference “Clusia flore albo, fructu coccineo. Plum. gen. 21. ic. 87. | ae although he did not use the Linnaean specific epithet. Jacquin’s illustra- tion clearly indicates the same plant as in Plumier’s unpublished plate (Fic. 1) which was copied (with alterations) for the Burmann edition.

1962] HOWARD, GUTTIFERAE 391

Chl five albe fue te |

Fic. 1. Lectotype of Clusia major L. ce 85, oh aquarellé “Clusia flore albo, fructu coccineo” from Manuscript No. 6, Plumier, Botan nicum Americanum, I. Courtesy of the Muséum National ee ticle Paris.

392 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLII

There is no doubt in my mind that Jacquin was renaming Clusia major by dividing this composite species and assigning new names to all three parts of it. In the second edition of Species Plantarum, also published in 1763, Linnaeus accepted Jacquin’s divisions of Clusia major, used Jac- quin’s specific names, credited him, and cited Jacquin’s publications. By this act Clusia major disappeared and was never again used until Britton and Millspaugh cited it as the type species of the genus.

In 1860, Planchon and Triana (Ann. Sci. Nat. IV. 13: 318-376. 1860) published a detailed treatment of Clusia in a larger work considering the Guttiferae as a whole. They concluded that Clusia major should be regarded as a nomen confusum and cited C. major in part under each of the three species mentioned “by Jacquin and accepted by Linnaeus. Other monographers have followed suit.

Although Hitchcock and Green (loc. cit.) selected Clusia major as the type species of the genus, they qualified this by stating “senus C. rosea Jacq.” The Index Nominum Genericorum refers to the Hitchcock and Green selection. The qualification is obviously incorrect, for Jacquin’s application of the basic references and illustration of C. major is to C. alba Jacq.

The typification of Clusia major must rest on Jacquin’s selection of the Linnaean reference to Plumier’s work. The plant in the Linnaean her- barium annotated by J. E. Smith as “Clusia alba H. Aublet vix Linnaeus” is neither the plant Jacquin described, nor does it correspond to the earlier Plumier description. Specimens collected by Plumier exist but none is available for the genus Clusia. Planchon and Triana selected as basic material an unpublished plate of Plumier (Fic. 1) but noted that the Burmann reproduction of this plate contained inaccuracies and excluded the flower and the analytical drawing. Thus the choice of a type then is between a previously unpublished plate and the illustration of C. alba published by Jacquin. I prefer the former, now published in this paper, to detract a bit from the long established status of C. alba.

Clusia major L. (syn. C. alba Jacq.) appears to be characterized by the elongate fruits borne on a cymose inflorescence which has a short peduncle. The species is represented by recent collections of material of the Lesser Antillean islands from St. Eustatius, St. Kitts, Montserrat, and Antigua, southward to St. Vincent. Urban’s C. plukenetii is similar in having a much elongated peduncle to the cymose inflorescence and a globose fruit. Urban cited a collection from Martinique (Duss 1829) which I have not seen. I have seen more recent material from Barbados (the type locality) and St. Lucia which agrees with Urban’s description. Regret- tably, considerable variation is found in the shape of the fruits of Clusia major. Although the length of the peduncle appears to be a reliable difference between these species, additional field study is necessary to determine if two taxa are truly represented.

1962 | HOWARD, GUTTIFERAE 393

Clusia flava Jacq.

There is no confusion in the use of this name. Linnaeus (Sp. Pl. 509. 1753) included a Sloane reference (Hist. Jam. 2. ¢. 200, f. 1.) in the literature cited with the original publication of Clusia major. Jacquin (Enum. 34. 1760; Select. 272. 1763) cited the same reference under the name C. flava. In the second edition of Species Plantarum (p. 1495. 1763) Linnaeus credited the epithet to Jacquin and cited as supporting literature references to Jacquin (1763), Sloane, and Browne. Fawcett and Rendle (Fl. Jam. 5: 193, 194, 1926) selected as the lectotype the Browne specimen in the Linnaean Herbarium, Presumably this is part of sheet 1224.3 and is the solitary detached leaf on the right-hand side.

Fawcett and Rendle cited the distribution of Clusia fzva as Jamaica, Barbados, Grand Cayman, and the Florida Keys. The species is indeed well represented by recent collections from Jamaica and has been te- collected recently by George Proctor on Grand Cayman (Proctor 15141 [GH]). The occurrence of this species in Barbados is not supported by specimens in any herbarium collections I have seen. The reference to its occurrence in the Florida Keys is apparently obtained from the writings of Nuttall (N. Amer. Sylva 2: 58. pl. 77. 1859) who stated of Clusia flava, “This singular and splendid tree is a native of Jamaica, and Cayenne in South America, where it is found among rocks on the declivities of moun- tains. We have now also to record it as a native of Key West in Florida, where it has recently been found, with so many other tropical productions, by Dr. Blodgett.” The illustration given by Nuttall is clearly that of C. flava; however, there are no supporting herbarium vouchers cited and one wonders if the illustration was not made from other herbarium material. Blodgett’s collections are preserved in the herbarium of the New York Botanical Garden where there are two sheets labelled “Clusia flava” collected by Dr. Blodgett. One sheet bears the common name “Bull Bay” and is from Pine Key. The other sheet without a common name was collected on Key West. Both specimens are sterile; however, both have heavier leaf blades than does ¢’. flava and both speciniens, I believe, should be referred to C. rosea. In further reference to Nuttall’s statement, I have seen no material of C. flava from Cayenne. Clusia flava appears to be restricted to Jamaica and Grand Cayman.

Fawcett and Rendle and authors 4! other modern floras of the Antilles do not accept the two varieties of Clusia flava proposed by Planchon and Triana (q.v.).

Clusia rosea Jacq.

This species was described briefly by Jacquin (Enum. 34. 1760) with the supporting citation “Catesh. car. 2, p. 99. #. 99.” In the Selectarum (270. 1763) Jacquin supplied the additional references of “Plum. gen. 21.” and “Pluk. alm. 92. t. 157. f. 2.” and gave a full description. In the first edition of Catesby’s work (1743), the plant is described and illustrated with white petals. In the second edition (1754), the illustration shows rose-

394 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIII

colored petals. The text refers to the petals as ‘white with rose streaks.” Jacquin did not illustrate the species in his Selectarum, but he reported it to be from Santo Domingo. In 1926, Fawcett and Rendle selected as the “type” the Bahama specimen collected by Catesby and now in the herbarium of the British Museum.

In modern floras, Clusia rosea is reported to occur in Florida, the Bahamas, the Greater and Lesser Antilles, Trinidad, Central America, and Venezuela. I have not seen all of the specimens cited by various authors for this species but a study of many specimens labeled C. rosea led to the conclusion that it occurs only in Southern Florida, the Bahamas, the Greater Antilles (Cuba, Jamaica, Hispaniola, and Puerto Rico), and the Virgin Islands (St. Thomas, St. Martins, St. Jan, and Anguilla). All of the material I have seen from Trinidad is best referred to C. palmicida L. C. Rich., although I have some doubt about the application of that name. The specimens labeled C. rosea from northern South America and from Central America do not represent that species as typified by the Catesby plant from the Bahamas.

Clusia plukenetii Urb.

Both Jacquin (Select. 270. 1763) and Linnaeus (Sp. Pl. ed. 2. 1495. 1763) cited the polynomial by Plukenet in the references given for Clusia rosea, Fawcett and Rendle (FI. Jam. 5: 192. 1926) did the same. Urban, in 1908 (Symb. Antill. 5: 432.), described C. plukenetii, gave the Plukenet reference and polynomial, and cited specimens from Martin- ique, St. Lucia, and Barbados, but did not designate a type. The Plukenet reference is to a poor illustration of a plant reported to occur in Barbados. It shows a branch with alternate leaves except for two very small leaves at the apex of the stem. Mr. George Proctor has informed me that a specimen of Clusia credited to Plukenet is in the Sloane Herbarium (Vol. 95, p. 152, upper right). It consists of only three leaves but is probably the holotype of Plukenet’s polynomial and therefore of Urban’s species. The common name of ‘Balsam apple” reported by Plukenet is appropriate for the genus Clusia, This species has been discussed under C. major. It is not comparable to C. rosea, and the name should not be used in the synonymy of that species as it has been by many recent authors.

Clusia minor L.

The protologue of this species as given by Linnaeus is the following: CLUSIA foliis venosis, minor Clussa, flore roseo, minor, fructu flavescente, Plum.

gen. 21. Habitat in America meridionali, 4 Arbor foliis venosis. Racemus florum terminalis.

bho

Jacquin in his Enumeratio described Clusia venosa as “C. foliis venosis” and cited “Plum. ic. 87. f. 2.” The Plumier references used by Linnaeus and Jacquin are comparable, although the latter (Ic. 87. f. 2) expands on

1962 | HOWARD, GUTTIFERAE 395

aa Clufia fbr ree money, cei Phra Vales:

Fic. 2. Lectotype of Clusia minor L. Fic. 88, Dessin a la plume ‘‘Clusia flore roseo minor, fructu e viridi rubra” from Manuscript No. 6, Plumier, Botanicum Americanum, t. VI. Courtesy of the Muséum National d’Histoire Naturelle,

aris.

396 JOURNAL OF THE ARNOLD ARBORETUM [ VOL, XLII

the former (Gen. 21) and is accompanied by an illustration. One can conclude, therefore, that Jacquin in his Enumeratio supplied an illegiti- mate substitute name for C. minor L. In the Selectarum, however, onan refers to ‘Clusia (venosa) foliis venosis. Linn. sp. pl. 20 0 DiC DME supplements this with a vague description which some subsequent authors felt represented a different species. It is clear from Jacquin’s use of Clusia venosa in 1760 that this name must be considered a synonym of C. minor L. In the second edition of Species Plantarum (p. 1495. 1763) Linnaeus accepted the specific epithet ‘“‘venosa” given by Jacquin and abandoned his own ‘‘minor,” but did not use Jacquin’s name nor refer to Jacquin’s publication as he did for the three segregate species of Clusia major. This action seems to indicate that Linnaeus felt C. venosa Jacq. was the same as his C. minor.

Planchon and Triana typified Clusia minor by the unpublished plate numbered 88 in the Plumier manuscript. They recognized that this draw- ing was not finished in the characteristic manner of Plumier’s other draw- ings but stated that with alterations it was comparable to fig. 2 of plate 87 in the Burmann edition of Plantae Americanae. Planchon and Triana, to clarify this species, printed the original, previously unpublished descrip- tion and commentary by Plumier and compared these with existing her- barium specimens available to them. Either the original Plumier drawing, reproduced here as Fic. 2, or the Burmann version with corrections will serve to typify the species. Clusia minor L. is clearly defined in modern floras.

Clusia venosa Jacq. (1763 not 1760)

Planchon and Triana (loc. cit. 369) were troubled by the description of Clusia venosa supplied by Jacquin in the Selectarum (273. 1763). In their manuscript, they described Clusia mangle, crediting the name to L. C. Richard on the basis of a manuscript notation. In a discussion of this species they state, “D’apres le nom de Palétuvier de montagne que porte a la Martinique le Clusia venosa de Jacquin (non L.) on pourrait croie que cette espéce est identique avec celle que nous décriverons ici. Mais la description de la plante de Jacquin ne justiferait en aucun point une telle détermination.”” Planchon and Triana do not otherwise place C. venosa Jacq. (1763).

In 1893 J. Vesque (DC. Monographiae Phanerogamarum 8: 140, 57. 1893) listed both C. venosa Jacq. (1763) and C. mangle Rich. ex Planch. & Triana. For C. venosa Vesque stated, ‘Il est impossible de classer avec certitude la plante visée par Jacquin. C’est probablement une espéce de la section Anandrogyne, voisine du Cl. Mangle qui, a ce qu'il parait, porte le méme nom vulgaire de ‘palétuvier de montagne’.” Vesque placed as supporting literature the polynomial and plate references which Planchon and Triana had used in defining C. minor. Vesque did not cite Jacquin’s Enumeratio of 1760, but listed only the Selectarum of 1763. There is an implication in the work of these men that C. venosa Jacq. as used in 1760 and defined in 1763 represent two different plants. The description Jacquin

1962] HOWARD, GUTTIFERAE 397

used in 1763 does not clearly define either C. minor L. (C. venosa Jacq. 1760) or Clusia mangle L. C. Rich. ex Planch. & Triana, except as the common name applies to the later species.

Engler (Nat. Pflanzenfam. 21: 201. 1925) considered Clusia mangle to be a synonym of Clusia venosa Jacq. This decision can not be accepted. Grisebach (Flora Brit. W. Indies 107. 1859) used Clusia venosa Jacq. “exclus. syn. Plum.” and the Imray collection he cited is Clusia mangle. This treatment, too, is invalid. Clusia mangle L. C. Rich. ex Planch. & Triana is typified by a Richard collection from the Soufriére in Guadeloupe. The species is known from Martinique, Dominica, and Guadeloupe. It is characterized by long peduncled cymes, small globose fruits about 2 cm. in diameter and by subsessile or short-petioled leaves. The identity of Clusia venosa Jacq. 1763 remains unsolved but the epithet is a later homonym of Clusia venosa Jacq. 1760 which is a synonym of Clusia minor EP Otis oe

Clusia grisebachiana (Planchon & Triana) Alain

Grisebach described Tovomita clusioides for a plant from Cuba (T. clusioides Griseb. Mem. Amer. Acad. IT. 8: 166. 1860, not T. clusioides [ Choisy ] Cambessédes, 1828). Planchon and Triana recognized the earlier homonym and renamed the species in honor of Grisebach. They retained the species in the genus Tovomita, but expressed some doubt as to its proper assignment (Tovomita (?) grisebachiana Planch. & Triana, Ann. Sci. Nat. IV. 14: 284. 1860). Alain correctly transferred the species to the genus Clusia (FI. Cub. 3: 314. 1953).

Urban described Clusia krugiana (Repert. Sp. Nov. 20: 340. 1924) from Puerto Rico and C. abbottii (Symb. Antill. 1: 367. 1899) from the Dominican Republic. Schmidt indicated on the herbarium labels of several Ekman collections that he believed C. abbottii belonged in the synonymy of C. Arugiana. In as much as Schmidt’s work was never published, this lead was reexamined on the basis of more recent collections and the type collections of each species. Only minor differences in leaf size, those partly of age, separate these three supposedly endemic species and they should be considered as one.

CALOPHYLLUM CALABA

A common tree of the Lesser Antilles, often used as a wind break, has a widely used common name of ‘‘galba.”’ Regrettably, the scientific name used in modern floras is less consistent. Grisebach (Flora Brit. W. Indies 108. 1859), Urban (Symb. Antill. 8: 438. 1920), and Duss (Ann. Inst. Colon. Marseille 13: 103. 1896) use Calophyllum calaba L. Britton & Wilson (Sci. Surv. Porto Rico 5: 584. 1924) and Williams (Fl. Trinidad & Tobago 1: 62. 1929) use Calophyllum antillanum Britt. Fawcett and Rendle (Fl. Jam. 5: 200. 1926) use Calophyllum jacquinii Fawc. & Rend., while Moscoso (Cat. Fl. Dom. 378. 1943) and Leon and Alain (FI. Cub. 3: 309. 1953) use Calophyllum brasiliense Camb. var. antillanum (Britt.)

398 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIII

Standl. Recently Furtado (Gard. Bull. Straits Settlements 11: 258-260. 1941) suggested the typification of Calophyllum calaba L. and his pro- posals seem acceptable.

Linnaeus (Sp. Pl. 514. 1753) proposed the name Calophyllum calaba citing four earlier references and referring to the origin of the plant as in “Indiis,” thereby implying that the plant occurred in both the Old World and the New World. The first generic description is given by Linnaeus in 1754 (Gen. Pl. ed. 5. 229.) where only the reference to Plumier’s Calaba is given. The Plumier reference is not given as such in the first edition of Species Plantarum, but is included in the reference Linnaeus gave to his Flora Zeylanica (90. 1747). In 1763, Jacquin elaborated on Linnaeus’ description (Select. Stirp. Amer. 269. ¢. 165. 1763), citing both the works of Plumier and Linnaeus, thereby implying a New World origin. In the sixth edition of Genera Plantarum (p. 266. 1764), Linnaeus ac- cepted Jacquin’s treatment. The monographers Planchon and Triana (Ann. Sci. Nat. IV. 15: 249. 1861), Vesque (DC. Monogr. Phan. 8: 588. 1893), and Engler (Nat. Pflanzenfam. 21: 196. 1925) restricted Calo- phyllum calaba in application to plants of the New World but credit the name to Jacquin.

Hitchcock and Greene (Prop. Brit. Bot. 161. 1929) suggested that the species be typified by Jacquin’s interpretation of Linnaeus’ name. More recently, Swartz in preparing the Index Nominum Genericorum cards cited as the lectotype for Calophyllum L. (Sp. Pl. 513. 1753) “C. calaba Linnaeus vide Gen. Pl. ed. 5. 229. 1754; etiam vide M. L. Green, Prop. Brit. Bot. 161. 1929).”

The acceptable synonymy therefore is:

Calophyllum calaba L. Sp. Pl. 514. 1753; Gen. Pl. ed. 5. 229. 1754; Jacquin, Sel. Stirp. Amer. 269. ¢. 165. 1763. Calophyllum antillanum Britt. in Britton & Wilson, Sci. Surv. Porto Rico 5: 584. Calophyllum jacquinu Fawc. & Rendl., Fl. Jam. 5: 584. Calophyllum brasiliense Camb. var. guido (Britt.) er Trop. Woods 30: 6. 1932

Britton had rejected the Jacquin interpretation of Calophyllum calaba as a species of the New World and had proposed the name C. antillanum for the American elements. Fawcett and Rendle reached the same con- clusion, proposing C. jacquinii in apparent unawareness of Britton’s earlier piblicatzan. Standley felt that the West Indian plants represented only a geographical extension of a Brazilian species and proposed several varieties including the combination Calophyllum brasiliense var. antil- lanum., the Antillean material is different from that of Central and South America at the varietal level, many of Standley’s varieties must be transferred to the older specific name Calophyllum calaba L. typified by plants in the Antilles.

1962 | HOWARD, GUTTIFERAE 399

RHEEDIA LATERIFLORA

Linnaeus (Sp. Pl. 1193. 1753) based this species on the work of Plumier (Gen. 45. 1703). Plumier did not specify the country of origin of the plant, but Lamarck (Encyc. 2: 245. 1786) noted the plant to be abundant in the Cul-de-sac aux Frégates in Martinique, an area visited by Plumier. Urban cites the distribution of the species as Jamaica, Hispaniola, Mont- serrat, Guadeloupe, Dominica, Martinique, St. Vincent, and Trinidad (Repert Sp. Nov. Beih. 5: 98. 1920). In spite of a study of recent collections from Guadeloupe, Dominica, Martinique, Marie Galante, Gre- nada, and Jamaica the species remains poorly understood. There have been no recent collections from Hispaniola and even the assignment of plants from Jamaica to this species is questionable.

The synonymy of this species is the following:

Rheedia lateriflora L. Sp. Pl. 1193. 1753. Mammea humilis Vahl, Eclog. Amer. 2: 40, ¢. 20. 1798. Mammea humilis var. macrophylla (Mart.) Duss, Ann. Inst. Col. Marseille 13: 102. 1897. Mammea humilis var. vahlii Griseb. Fl. Brit. W. Indies 108. 1859. Mammea humilis var. plumieri Griseb. Ibid.

Vahl’s species is based on a Ryan collection from Montserrat. Grise- bach’s two varieties are based respectively on the Vahl and Plumier types. Grisebach described var. vahlii as shrubby with the leaves pointed at both ends. His var. plumieri was a tree with the leaves rounded or sub- cordate at the base. The specimens I have seen are variable in leaf shape and both types of leaf-bases can be found on one branch. Usually the leaves of young plants and of the lateral or axillary branches of older plants have the acute leaf bases,

Most modern workers consider Garcinia macrophylla Mart., the basio- nym of Rkeedia macrophylla (Mart.) Planchon & Triana and of Duss’ var. macrophylla, to be a distinct species. The specimens Duss cited are to be referred to Rheedia lateriflora.

400 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLII

DATES OF PUBLICATION OF THE JOURNAL LINNAEA

ROBERT C. FOSTER

THE NINETEENTH CENTURY GERMAN periodical Linnaea is of consid- erable importance to American taxonomy because of the large numbers of New World plants described in it. For some time it has been known that certain title-page dates were incorrect. Suspicion was also felt that others might well be incorrect. When the Gray Herbarium Card Index began to include varietal names from Linnaea, it became imperative to establish the dating as accurately as possible. Between my own investi- gations and those of six previous workers, it has been possible to do this for many of the volumes. It seems desirable to combine these results in one publication to obviate the necessity of searching in seven different places. The earlier references are as follows:

BLAKE, S. F., Jour. Wash. Acad. Sci. 46: 192. 1933.

Kuntze, 0. Rev. Gen. 3(2): 158, 159. 1898.

Moore, H. E., Gent. Herb. 8: 375. 1954.

STEENIS- KRUSEMAN, M. J. vAN, Fl. Males. ser. 1. 4: cxevili-cxcix. 1954.

VOLUME DATE COMMENTS AND SOURCES Volume 1. TITLE-PAGE DATE: 1826. Heft 1: 1-165 P. 65: paper dated oe 1825, Heft 2: 166-332 P. 279: ref. to Mart. Nov. Gen. fasc. 4, publ. ae Apr. 1826 P) Heft 3: 333-511 After Aug. 1826 P. 468: reviewed Pohl, Pl. Bras. Tc. ere 6; this appeared in Aug. 1826. Heft 4: 512-677 Volume 2. TITLE- oe DATE: 1827. Heft 1: 1-144 Heft 2: ee Heft 3: 277-540 After Aug. 1827 ? P. 527: reviewed Lindl. Orch.

Scelet.; copy by Lindl. to Brit. Mus., Aug. Heft 4: 541-738

Volume 3. TITLE-PAGE DATE: 1828. Heft 1: 1-114 P. 69: paper dated Aug. 1827. Heft 2: 115-198 P, 173: paper dated Dec. 1827.

Heft 3: 199-308 Heft 4: 309-440

1962 | FOSTER, THE JOURNAL LINNAEA 401

VOLUME DATE COMMENTS AND SOURCES

Volume 4. TITLE-PAGE DATE: 1829, Heft 1: 1-128 Jan. 1829 Hook. Bot. Misc. 3: 1. 1832. Heft 2: 129-288 Heft 3: 289-450 P. 358: paper dated 16 Mar. 1829. Heft 4: 451 to end P. 598: note dated Oct. 1829.

Volume 5.

TITLE-PAGE DATE: 1830.

Heft 1: 1-176 Jan. 1830

Heft 2: 177-337 Apr. 1830

Heft 3: 338-496 July, 1830

Heft 4: 497-688 Oct. 1830

Heft 5: 689-756 [1830] All dates from Van Steenis-Kruse-

man.

Volume 6. TITLE-PAGE DATE: 1831. Heft 1: 1-208 After Mar. 1831 P. 65: ref. to Mar. 1831. Heft 2: 209-384 Heft 3: 385-544 Heft 4: 545-736 Heft 5: 737-796

Volume 7.

TITLE-PAGE DATE: 1832. Heft 1: 1-144

Heft 2: 145-272

Heft 3: 273-400

Heft 4: 401-560

Heft 5: 561-688

Heft 6: 689-778

Volume 8. TITLE-PAGE DATE: 1833. 112

ref. to July, 1833.

8 Heft 4: 385-512 After July, 1833 Sys 9: ref. to 27 July 1833.

P Heft 5: 513-624 After 27 July 1833 P. Heft 6: 625-684

39 52

Volume 9 TITLE-PAGE DATE: 1834; 1835 at bottom of page.

Heft 1: 1-144 After Jan. 1834 P. 124: letter dated 20 Jan. 1834.

Heft 2: 145-272

Heft 3: 273-402 P. 331: counts 216 Annonaceae known in 1834.

Heft 4: 403-514 Early 1835 P. 506: manuscript dated 27 Dec. 1834.

Heft 5: 515-642 P. 566: notes dated 7 Feb. 1835.

Heft 6: 643-758

402 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIII

VOLUME DATE COMMENTS AND SOURCES

Volume 10. TITLE-PAGE DATE: 1836; 1835-1836 at bottom

of page. Heft 1: 1-128 After July, 1835 P. 123: paper dated 1 July 1835. Heft 2: 129-224 Heft 3: 225-368 P. 289: paper dated 25 Oct. 1835. Heft 4: 369-464 After Apr. 1836 P. 463: see date of letter. Heft 5: 465-608 After Apr. 24, 1836 P. 484: see date of letter. Heft 6: 609-758 Volume 11. TITLE-PAGE DATE: 1837. Heft 1: 1-128 P. 108: paper dated 16 Oct 1836. Heft 2: 129-256 P. 164: paper dated 23 Nov. 1836. Heft 3: 257-432 P. 280: paper dated Feb. 1837. Heft 4: 433-544 P. 486: letter dated 15 Apr. 1837. Heft 5: 545-608 Heft 6: 609-728 P. 650: paper ey ees 1837. At end of volum Pl. Ex- sicc. Afr. eae be ee dated 20 Mar. 1838. Volume 12. TITLE-PAGE DATE: 1838. Heft 1: 1-128 P. 119: paper dated Jan. 1838. Heft 2: 129-256 P. 200: paper dated 16 Feb. 1838. Heft 3: 257-352 Heft 4: 353-496 Heft 5: 497-576 Heft 6: 577-700 P. 686: ref. to plants collected in June, 1838. Volume 13. Tink sie DATE: 1839. Heft 1: 1-112

Heft 2: os 272 Heft 3: 273-336 Heft 4: 337-448 P. 377: paper dated 7 Aug. 1839. Heft 5: 449-560 Heft 6: 561-744

Volume 14. TITLE-PAGE DATE: 1840. Heft 1: 1-160 Heft 2: 161-240 F..1732 te dated 31 Dec. 1839 (Old Style). Heft 3: 241-304 Po2iT a dated 12 Apr. 1840.

Heft 4: 305-384 Heft 5: 385-528 Heft 6: 529-728 Early 1841 P. 706: bears date Feb. 1841.

1962] FOSTER, THE JOURNAL LINNAEA 403

VOLUME DATE COMMENTS AND SOURCES Volume 15. TITLE-PAGE DATE: 1841. Heft 1 60 P. 160: paper dated Jan. 1841. Heft 2: 161-288 P. 282: paper dated 15 Apr. 1841. Heft 3: 289-384 P. 384: paper dated May, 1841. Heft 4: 385-480 P. 442: paper dated 1 July, 1841. Heft 5: 481-704 P. 703: paper dated Oct. 1841. Volume 16. TITLE-PAGE DATE: 1842, Heft 1: 1-112 P. 112: paper dated + Jan. 1842. Heft 2: 113-224 Heft 3: 225-336 P. 336: note, dated June, 1842. Heft 4: 337-400 P. 388: paper dated 9 Aug. 1842. Heft 5: 401-512 P. 512: paper dated Sept. 1842. Heft 6: 513-592 Noticed, Bot. Zeit. 1: 288. 28 pr. Introduction to volume dated March, 1843. Volume 17. TITLE-PAGE DATE: 1843. Heft 1: 1-128 Vol. 16, p. iv, dated Mar. 1843, says a Ad o oh de be will soon pea Heft 2: 129-256 P. 244: ae dated May, 1843. Heft 3: 257-384 Heft 4: 385-512 P. 446: paper cites material col- lected 29 Aug. 3. Heft 5: 513-640 P. 577: paper dated Dec. 1843. Heft 6: 641-764 Early 1844 P, 676: paper dated Jan. 1844. Volume 18. TITLE-PAGE DATE: 1844. Heft 1: 1-112 P. 50: paper dated May, 1844. Heft 2: 113-256 lst paper dated Sept. 1844 Heft 3: 257-384 P. 301: paper a Oct 1844, Heft 4: 385-512 P. 505: review dated 1 Dec. 1844. Heft 5: 513-640 Mid-1845 P. 626: preface to paper dated May, 1845. Heft 6: 641-774 Py fin paper read 12 June 1845. Volume 19. TITLE-PAGE DATE: 1847. Heft 1: 1-128 Jan. 1846 Kuntze, Rev. Gen. 3(2): 158. 1898. Heft 2: 129-256 Feb. 1846 Kuntze, loc, cit. Heft 3: 257-384 1846 Before 18 Dec. 1846; noticed Bot. Zeit. 4: 875, of that date. Heft 4: 385-512 Dec. 1846 Kuntze, loc. cit. Heft 5: 513-640 Feb. 1847 Kuntze, loc. cit.

Heft 6: 641-765 Apr. 1847 Kuntze, loc. cit.

404 JOURNAL OF THE ARNOLD ARBORETUM [ VOL, XLIII

VOLUME DATE COMMENTS AND SOURCES

Volume 20.

nena — DATE: 1847.

Heft 1: 1-128 May, 1847 All dates from Van Steenis-Kruse-

Heit ace June, 1847 man.

Heft 3: 257-384 July, 1847

Heft 4: 385-512 Aug. 1847

Heft 5: 513-640 Oct. 1847

Heft 6: 641-781 Dec. 1847 Volume 21.

TITLE-PAGE DATE: 1848.

Heft 1; 1-128

Heft 2: 129-256 P. 202: paper dated Jan. 1848.

Heft 3: 257-384 Heft 4: 385-512

Heft 5: 513-640 Hefte 1-5 noticed, Bot. Zeit. 6: 870. 15 Dec. 1848.

Heft 6: 641-780 Noticed, Bot. Zeit. 7: 750. 19 Oct. 1849.

Volume 22.

TITLE-PAGE DATE: 1849,

Heft 1: 1-128 P. 113: ee ref. to Bot. Zeit. 10 Nov. 1848.

Heft 2: 129-256 P. 144: paper dated 20 Dec. 1848.

Heft 3: 257-384

Heft 4: 385-512 P. 464: paper dated 24 Mar. 1849.

Heft 5: 513-640 Heft 6: 641-768 Heft 7: 769-898

Volume 23. TITLE-PAGE DATE: 1850. Heft 1: 1-128 Heft 2: 129-256 P. 212: paper dated 30 Mar. 1850. Heft 3: 257-384 Heft 4: 385-512 Heft 5: 513-640 Heft 6: 641-770 P. 736: note dated 18 Sept. 1850.

Volume 24. TITLE-PAGE DATE: 1851. Heft 1: 1-128 Heft 2: 129-256 Heft 3: 257-384 P. 300: paper dated July, 1851. Heft 4: 385-512 Heft 5: 513-640

Heft 6: 641-804 Early 1852 P. 649: paper dated Jan. 1852. Volume 25.

TITLE-PAGE DATE: 1852.

Heft 1: 1-128 June, 1852 All i ae Kuntze, Rev. Gen.

Heft 2: 129-256 Dec. 1852 a2): 1898.

1962 | FOSTER, THE JOURNAL LINNAEA 405

VOLUME DATE COMMENTS AND SOURCES Heft 3: 257-384 Feb. 1853 Heft 4: 385-512 Apr. 1853 Heft 5: 513-640 June, 1853 Heft 6: 641-772 Dec. 1853 Volume 26. TITLE-PAGE DATE: 1853. Heft 1: 1-128 Feb. 1854 All dates from Kuntze, Rev. Gen. Heft 2: 129-256 Apr. 1854 3(2): 159. 1898. Heft 3: 257-384 Aug. 1854 Heft 4: 385-512 Feb. 1855 Heft 5: 513-640 May, 1855 Heft 6: 641-807 Sept. 1855 Volume 27. TITLE-PAGE DATE: 1854. Heft 1: 1-128 Nov. 1855 Engl. Bot. Jahrb. 19: 562. 1894. Heft 2: 129-256 Jan. 1856 Ibid. Heft 3: 257-384 Jan. 1856 Ibid. Heft 4: 385-512 Feb. 1856 Ibid. Heft 5: 513-640 [1856] Heft 6: 641-799 Aug. 1856 Gent. Herb. 8: 375. 1954. Volume 28. Tire. -PAGE DATE: 1856. Heft 1: 1-128 Aug. 1856 All dates from Blake in Jour. Heft 2: 129-256 Sept. 1856 Wash. Acad. Sci. 46: 192. 1933. Heft 3: 257-384 Jan. 1857 Heft 4: 385-512 June, 1857 Heft 5: 513-640 Aug. 1857 Heft 6: 641-767 Feb. 1858 Volume 29. TITLE-PAGE DATE: 1857-1858. Heft 1: 1-128 Feb. or later, See cover of Vol. 28, Heft 6. 1858 Heft 2: 129-256 June, 1858 Engl. Bot. Jahrb. 19: 526. 1894. Heft 3: 257-384 Sept. 1858 Engl. Bot. Jahrb. 19: 526. 1894. Heft 4: 385-512 Feb. 1859 Gent. Herb. 8: 375. eek Heft 5: 513-640 Apr. 1859 Gent. Herb. 8: 375. Heft 6: 641-764 June, 1859 Bull. Soc. Bot. Fr. 7. oe 1860. Volume 30. TITLE-PAGE DATE: 1859-1860. Heft 1: 1-128 Heft 2: 129-256 Before 25 Nov. 1859; cf. Bull. Soc. Bot. Fr. 6: 742. 1859

Heft 3: 257-384 Heft 4: 385-512 Before 23 Nov. 1860; cf. Bull. Soc. Bot. Fr. 7: 867. 1860.

406 JOURNAL OF THE ARNOLD ARBORETUM

VOLUME Heft 5: 513-640 Heft 6: 641-779 Volume 31.

Tine -PAGE DATE: Heft 1: 1-128 Heft 2: 129-256 Heft 3: 257-384 Heft 4: 385-512 Heft 5: 513-640

Heft 6: 641-751

Volume 32.

TITLE-PAGE DATE:

Heft 1: 1-128

Heft 2: 129-256 Heft 3: 257-384 Heft 4: 385-512 Heft 5: 513-640

Heft 6: 641-801

Volume 33.

TITLE-PAGE DATE:

Heft 1: 1-128

Heft 2: 129-256

Heft 3: 257-384 Heft 4: 385-512

Heft 5: 513-640

Heft 6: 641-770

DATE

Mar. 1861

1861-1862.

1863.

1864-1865.

[ VOL. XLII1 COMMENTS AND SOURCES

Before 25 Jan. 1861; me Bull. Soc. Bot. Fr. 8: 24. Engl. Bot. Jahrb. 19: A “1894.

Before 11 Apr. 1862; cf. Bull. Soc. Bot. Fr. 9: 194. 1862. Before 11 Apr. 1862;

Soc. Bot. Fr. 9: 419. 1862. Before 13 Mar. 1863; cf. Bull. Soc. Bot. Fr. 10: 122. 1863 Before 13 Mar. 1863; Soc. Bot. Fr. 10: 122. 1863.

Before 13 Nov. 1863; Soc. Bot. Fr. 10: 459. 1863. Before 11 Dec. 1863; c. Bot. Fr. 10: 539, 1863. Before 11 Dec. 1863; Soc. Bot. Fr. 10: 539. 1863. Before 1 Apr. 1864; Soc. Bot. Fr. 11: 87. 1864. Before 13 May 1864; Soc. Bot. Fr. 11: 149. 1864. Before 13 Jan. 1865; Soc. Bot. Fr. 12: 4. 1865

Before 31 Aug. 1864; reviewed in Flora, Repert., Halfbogen 3- 4, 31 Aug. 1864.

Betore 14 no 1864; reviewed in Flora, Repert., Halfbogen 5, 14 Dec

Before 24 Mar. 1865; cf. Bull. 4

Before 20 May 1865; reviewed in Flora, ae Halfbogen 10, 20 May 1865.

Before Sept.—Oct., 1865; cf. Bull. Soc. Bot. Fr. 12: 227. 1865.

1962] FOSTER, THE JOURNAL LINNAEA 407

VOLUME DATE

Volume 34. TITLE-PAGE DATE: 1865-1866. Heft 1: 1-128 Heft 2: 129-256 Heft 3: 257-384

Heft 4: 385-512

Heft 5: 513-640

Heft 6: 641-752

Volume 35. TITLE-PAGE DATE: 1867-1868. Heft 1: 1-128

Heft 2: 129-256 Heft 3: 257-352 Heft 4: 353-384

Heft 5: 385-512

Heft 6: 513-637 Nov. 1868 Volume 36.

TITLE-PAGE DATE: 1869-1870.

Heft 1: 1-128 Late 1869

Heft 2: 129-256 Late 1869

Heft 3: 257-384 Jan. 1870

Heft 4: 385-512 Apr. 1870

Heft 5: 513-640

Heft 6: 641-790 Dec. 1870 Volume 37.

TITLE-PAGE DATE: 1871-1872.

Heft 1: 1-128 [Early 1872]

Heft 2: 129-256 Apr. 1872

Heft 3: 257-432 July, 1872

Heft 4: 433-512 July, 1872

COMMENTS AND SOURCES

Before Mar. 1865; cited in DC. Prodr. 15(2): 543 1865. Before 22 Dec. 1865: cf. Bull. Soc. Bot. Fr. 12: 409. 1865. Before 22 Dec. 1865; cf. Bull. Soc. Bot. Fr. 12: 409. 1865. Before 23 Jan. 1866; reviewed in Flora, Repert. (1865), Half- bogen 3, 23 Jan. 1866. Before 11 May 1866; cf. Bull. Soc. Bot. Fr. 13: 253. 1866.

Before 13 Oct. 1867; reviewed in Flora, Repert. 13 Oct. 1867, Z

Dena:

Before 13 Oct. 1867; reviewed in

Flora, Repert. 13 Oct. 1867, 7

Dok

Before June, 1868; reviewed in

Flora, Repert. Mid-June, 1868, 26

p. 26.

Before June, 1868; reviewed in Flora, Repert. Mid-June, 1868, p. 26.

Date from Van Steenis-Kruseman.

Se in Flora, Repert. 15,

Jan. ps9:

Me in Flora, Repert. 15. Jan. 1870,

Van Steenis-Kruseman.

Van Steenis-Kruseman.

Van Steenis-Kruseman.

Jour. Bot. 10: 220. 1872. Jour. Bot. 10: 382. 1872. Jour. Bot. 10: 383, 1872.

408 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIII VOLUME DaTE COMMENTS AND SOURCES Heft 5: 513-544 Oct. 1872 Jour. Bot. 11: 61. 1873. Heft 6: 545-663 Sept. 1873 Jour. Bot, 12: 125. 1874. Volume 38. TITLE-E es DATE: 1874. Heft 1: 1-144 Dec. Lore Jour. Bot: 122 125. 1874: Fett 23 peer Jan. 1874 Jour. Bot. 12: 125. 1874.

Heft 3: 257-384 Heft. 42 395=512 Heft 5: 513-640 Heft 6: 641-753

Volume 39.

TITLE-PAGE DATE:

Heft 1: 1-128

Heft 2: 129-224 Heft 3: 225-256 Heft 4: 257-352

Heft 5: 353-448 Heft 6: 449-526

Volume 40. TITLE-PAGE DATE: Heft 1

Heft 2-3: 97-192

Heft 4: 193-288

Heft 5: 289-384 Heft 6: 385-468

Volume 41. TITLE- or DATE: Heft 1:

Heft 2: 113-192 : 193-288 | 289-384 3852570 Heft 6: 577-655

May, 1874 July, 1874

1875.

Feb.

1875

June, 1875 June, 1875

1876.

Aug. Dec.

1876 1876

ec. 1876

. 1877 PLery . 1877 oe eas

. 1878

Van Steenis-Kruseman.

Van Steenis-Kruseman

sea Bot. Zeit. 333 751. 13 Nov. 1874

Jour, Bot, 133253._1875.

Jour. Bot. 13: 253. 1875.

Bot. Zeit. 33: 455. 2 July 1875.

Noticed, Bot. Zeit. 33: 769. 19 Nov. 1875.

Noticed, Bot. Zeit. 33: 790. 26 Nov. 1875.

Noticed, Bot. Zeit. 34: 64. 28 Jan. 1876

Noticed, Bot. Zeit. 34: 352. 2 June 1876.

Noticed, Bot. Zeit. 34: 480. 25 uly

Svenson, Rhodora 41: 313. 1939.

Svenson, Rhodora 41: 313. 1939; noticed, Bot. Zeit. 35: 16. 5 Jan. 1877.

All dates from original covers bound in a volume in the Oakes Ames Orchid Herbarium Li- brary.

1962 | FOSTER, THE JOURNAL LINNAEA 409

VOLUME DATE COMMENTS AND SOURCES

Volume 42.

TITLE-PAGE DATE: 1878-1879.

12 Feb. 1878 Jour. Bot. 16: 123. 1878. Heft 2: 113-192 Noticed, Bot. Zeit. 36: 318. 17 May 1878.

Heft 3: 193-288 1879 Van Steenis-Kruseman.

Heft 4: 289-384 1879 Van Steenis-Kruseman.

Heft 5: 385-480 1879 Van Steenis-Kruseman.

Heft 6: 481-667 Nov. 1879 Van Steenis-Kruseman. Volume 43.

‘TITLE-PAGE DATE: 1880-1881.

Heft 1: 1-66 Sept. 1880 All dates from Van Steenis-Kruse-

Heft 2: 67-138 July, 1881

Heft 3-4: 139=252 Aug. 1881 Heft 5-6: 253-486 June. 1882 Heft 7: 487-554 July, 1882

Gray HERBARIUM, HarVARD UNIVERSITY

410 JOURNAL OF THE ARNOLD ARBORETUM [ VOL, XLIII

A MONOGRAPH OF THE GENUS PLATYDESMA (RUTACEAE)!

BENJAMIN C. STONE *

THE GENUS PLATYDESMA was described by Horace Mann, Jr. (1866), to accommodate an unusual rutaceous species collected by him and William Brigham ‘‘on the mountains behind Honolulu” on the island of Oahu in the Hawaiian Islands. Mann presented a rather full description of the genus and its single species, which he named Platydesma campanulata. A few years later (1869) he again wrote of the genus and species, with almost unchanged descriptions, but added a few sentences pertaining to the circumstances of the discovery, and a fine full-page half-tone plate illustrating its habit and flowers. The species was noted by Heinrich Wawra (1873) without further description or commentary. In 1888, how- ever, Wilhelm Hillebrand, in his great Flora of the Hawaiian IJslands,

added three species to the genus, two of them, P. cornuta and P. rostrata, described as new, and a third to which Hillebrand applied the name Pp. auriculaefolia, based on Pelea auriculaefolia A. Gray. As has previously been pointed out (Rock 1913, 1918; Stone 1961, 1962), Hillebrand erred in considering Gray’s species to be a Platydesma, for it is a true Pelea of the typical section. However, the specimens which led Hillebrand to include a fourth species in Platydesma are, in fact, members of the genus. Rock (1913) named the taxon represented by these specimens Platydesma campanulatum var. sessilifolia (sic). This taxon is here considered to be of specific rank, agreeing essentially with Hillebrand’s treatment, but clearly less distinct from P. spathulatum (the correct name of Mann’s P. campanulata) than either P. cornutum or P. rostratum which Hillebrand first described.

"Studies in the Hawaiian Rutaceae, VI.

* The field work for this revision was accomplished while the writer was Research Assistant, Department of Botany, University of Hawaii, Honolulu. Some studies of

National a Histoire Naturelle: Paris; and Dr. Otto Degener, Waialua, Oahu, Hawaii, who ee both herbarium material and fresh collections. I also wish to thank Dr. . Smith of the U. S. National Museum for his kind assistance. At the Bishop Museum, Honolulu, where a rather large number of collections of the various species has been accumulating, I was provided with all necessary facilities for study. I am much indebted to Marie C. Neal and to E. H. Bryan, Jr., of the Museum staff, for

R L. M

Perry provided generous assistance in my studies of the materials in the herbarium of the Arnold Arboretum and of the Gray Herbarium, for which I am most grateful.

1962 | STONE, THE GENUS PLATYDESMA 411

The name now applied to this species has a curious history. Material collected by Jules Remy was provisionally determined by H. Baillon in Paris as Claoxylon insigne sp. nov. (Claoxylon is a genus of Euphorbiaceae, the fruits of which [in the Hawaiian species] bear a slight resemblance to those of Platydesma.) Remy’s specimen bears only fruit. The name C. insigne was never validly published, but was listed (without description and only a herbarium reference) by Drake del Castillo (1890). Many years later, while revising the Hawaiian members of Claoxylon, E. E. Sherff happened to see the Remy collection and, relying all too heavily on the authority of Baillon, published the name Claoxvlon Remy Sherff sp. nov., with a full deen orion, Recently the real identity of this “Claoxylon” became apparent to Dr. Otto Degener, whe received con- firmation from the Muséum National d’Histoire Naturelle of Paris that the plant was rutaceous. A lifelike drawing of the specimen was prepared (see Fic. 4) in which the character of Platvdesma may readily be seen. The proper transfer of the name was made in Degener’s Flora Hawaiiensis (1960), where the present writer was able to show that this species is identical with the one intended by Hillebrand to include his specimens from Kohala, Hawaii, to which he had appended the name Platydesma auriculaefolia. The species is now called Platydesma Remyi (Sherff) Deg., Sherff & Stone.

Only two other names must be considered in the genus, both published by H. Léveillé, whose erratic work, discussed at some length both by Rock and Rehder in recent years, scarcely needs further mention. The first is Platydesma oahuensis Lévl. (1911), which is simply a synonym of Mann’s original species. The second, Platydesma Fauriei Lévl. (1911) is borne by a plant not a member of the Rutaceae, but of the Solanaceae, Notho- cestrum longifolium A. Gray, as was first pointed out by Rock (1914).

Recently it was found necessary to take up an older specific epithet for Mann’s original species (Stone, 1962), which must now be called Platy- desma spathulatum (A. Gray) Stone. The basionym is Melicope spathu- lata A. Gray, published a decade earlier than Mann’s species. Gray also described Melicope ? grandifolia in the sarne work (1854), which is the same species. Photographs of the type specimens were published in a preceding paper (Stone, 1962).

Platydesma H. Mann,’ Proc. Boston Soc. Nat. Hist. 10: 317. 1866; Mem. Boston Soc. Nat. Hist. 1: 529. 1869. Hillebrand, Fl. Hawaiian Is. 71. 1888. Rock, Indig. Trees Hawaiian Is. 241. 1913. Engler, Nat. Pflanzen- fam. ed. 2. 19a: 240, fig. 101, A-G. 1931.

Shrubs or small trees with opposite or subopposite, simple, punctate,

® Mann attributed feminine gender to the generic name, and was followed in this usage by most later authors. The name, however, like several others taken from Greek (such as Geniostoma) is of a class which, although ending in the apparently Latin feminine -a, takes neuter modifiers. Therefore the specific and eens epithets are corrected, where necessary, to neuter gender.

412 JOURNAL OF THE ARNOLD ARBORETUM [ VOL, XLIII

elongate, nonstipulate leaves; wood when freshly broken with a pepsin- like odor, somewhat brittle; inflorescences axillary, cymose or subsessile and contracted; flowers hermaphrodite, with 4 broad, rounded sepals in imbricated pairs; petals 4, white, but slightly imbricated; disk plane, obscurely 4—8-lobate; stamens 8, connate into a staminal tube, white, only the anthers and a short, broad, deltoid filament (adnate to the con- nective) free, the anthers narrow, elongate, subhastate at the base, with linear thecae; ovary of 4 carpels, rounded above, adnate to the apex, each with 5-8 pendulous ovules on short, broad funiculi; style central, of 4 connate elements, columnar, the linear, erect, united stigmas somewhat thickened at apex; fruit capsular, dry, 4-lobed, the carpels remaining connate, loculicidally (sometimes tardily) dehiscent, rounded or horned at apex, the endocarp thin, glabrous, finally separable, the seeds subovoid or ellipsoid with angles and faces (if any) from mutual pressures, the testa black and shining, the wall osseous, the cotyledons of the embryo thin, broad, rounded, the hypocotyl very short, the embryo enclosed by endosperm.

Type species: Platydesma campanulatum H. Mann = P. spathulatum (A. Gray) Stone.

DIsTRIBUTION: Endemic to the Hawaiian Islands; four species: Platy- desma Spathulatum, widespread in the archipelago, with insular varieties and forms; P. cornutum, endemic to Oahu; P. rostratum, endemic to Kauai; and P. Remyi, apparently endemic to the Kohala Mountains, Hawaii.

RELATIONSHIPS: Platydesma is placed by Engler in the subfamily Rutoideae, tribe Xanthoxyleae, subtribe Choisyinae, as genus 28 between Choisya HBK., a Mexican and southwestern American genus, and Dutail- /yea Baill., a small genus of two species endemic to New Caledonia. Neither of these genera, however, appears to be either very close in relationship or similar in superficial or technical characters. Medicosma Hook. f., a monotypic Australian genus, is closer in appearance, and possibly in affinity, although placed before (as genus 25) the above-mentioned genera. It is perhaps more plausible to consider that Platvdesma may be ultimately an Old World derivative, but to exclude American affinities would be premature.

EcoLocy: Platydesma spathulatum, and its subtaxa, are both more common and more widely distributed in the Hawaiian group than the other species. None, however, could be considered dominant or even frequent members of the vegetation. The species are to be found chiefly in the rainforests at moderate elevations, from roughly 2000 to 5000 feet, and also on windswept crests in low, stunted vegetation. Platydesma cornutum and P. rostratum usually occur as single plants or in small groups, often along streams or gullies under heavy shade, while P. spathu- latum may be found under open or closed canopy forests, or exposed on slopes, and can apparently tolerate drier situations. In a few localities

1962 | STONE, THE GENUS PLATYDESMA 413

(such as around Kokee, Kauai) P. spathulatum may be found so fre- quently as to constitute a considerable, though never dominant, component of the association.

Hasit: Platydesma spathulatum is a polymorphic species, but older individuals in more protected areas become small trees. In forests such as those about Kumuwela, Kauai, trees of five to eight meters in height form an understory in a forest containing species of Psychotria, Pelea, Bobea, Pleomele, Pritchardia, and other genera. In exposed, windswept areas such as open crests along the Koolau Range, Oahu. a form of the same species may be found occurring as a small shrub. Both Platydesma cornutum and P. rostratum are shrubby, with erect, usually slender stems and erect or ascending branches. They can accurately be called ‘‘Schopf- baume” with their rather large leaves clustered toward the ends of the branches (see Fic. 1). Platydesma Remyi has the habit of P. spathulatum.

LOCAL NAME AND USE: The Hawaiian name for all species is ““Pilo kea,” meaning ‘‘white pilo.” If the Hawaiians had a particular use for the plant, it is not presently known. Certain species have recently been the subject of chemical investigations by P. J. Scheuer and his associates at the Uni- versity of Hawaii. Evidence has been found that some species contain appreciable amounts of alkaloids and essential oils. In this connection, it is worthy of note that Rock (Indig. Trees 242. 1913) reported that the odor of the plant when bruised is like pepsin. The wood, when broken, emits a starchy odor.

SPECIFIC CHARACTERS: Characters of taxonomic value are found chiefly in the structure, shape, and size and presence or absence of indument of the fruits: the structure of the inflorescence; the shape and size of the leaf, especially the petiole; the habit; and, to a lesser extent, the amount of pubescence both on leaves and parts of the inflorescence. The other indigenous genera of Rutaceae, Fagara (Zanthoxylum) and Pelea, are greatly different both in appearance and in technical characters, although seedling stages of Pelea and Platydesma could perhaps be confused with one another.

KEY TO THE SPECIES

1. Carpels in fruit rounded, the capsule subglobose; habit shrubby or usually arborescent, small trees with several to many irregularly arranged spreading or ascending branches. Section PLATYDESMA.

2. Petioles conspicuous, 1-6 cm. long; fruits glabrous, roe , Maui, Hawaii (probably also Molokai). ................ Les nnn 2. Petioles obsolete or at most 5 mm. ee oe ae with a more or less persistent minute pubescence; Hawai. ............ 2. P. Remyt.,

Carpels in fruit rostrate, cornute, or sharply Peed and somewhat divari- cate; habit shrubby, with erect, sparsely branched or simple stems closely foliose toward the apices. Section CORNUTIA.

3. Carpels in fruit conic-pointed or shortly attenuate; leaves elongate, spath-

—

414 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLII

ulate, narrowed at base to a small rounded or aes lamina; Oahu. asa, A see eet ge nee ae ae ee P. cornutum. Carpels | in ‘fruit tapered into a slender, elongate beak; cee narrow oblong or subspathulate, the base broad and abruptly tauneate or obtuse; WA Gg yee oo ehh eos AGS ee Ooo and LE 4. P. vastus.

ios)

Section PLATYDFSMA

1. Platydesma spathulatum (A. Gray) B. C. Stone, Madrono 16: 165. 1962.

Melicope spathulata A. Gray, Bot. U. _ a Exped. 15: 352. 1854.

Melicope? grandifolia A. Gray, op. cit.

Platydesma begga Mann, st a Soc. Nat. Hist. 10: 317. 1866;

osto . Hist. 1: 529-531. pl. 22. 1869. Hillebrand, A

Hawaiian Is. a, es “companulata.” Heller, Minn. Bot. Stud. 1: , 1897. Rock, Indig. Trees Hawaiian Is. 241. 1913 (including f. lee Skottsberg, Acta Horti Gothob. 10: 120. 1936; ibid. 15: 388. 1944

Platydesma campanulata y var. macrophylla Hillebrand, loc. cit.

Platydesma oahuensis Léveillé, Rep. Sp. Nov. 10: 153. 1911.

A shrub or small tree to about 12 m., with spreading branches foliose toward the ends, trunk straight, to about 20 cm. thick, glabrous through- out or with fulvous pubescence on the new leaves and inflorescence (and in some varieties the mature leaves pubescent beneath); juvenile branches greenish, at last clothed with pale grayish bark; all parts copiously glandu- lar, emitting odor of pepsin; wood, when broken, with a starchy odor; leaves variable in size, distinctly petiolate, the petiole 10-60 mm. long, the blades usually 10-50 cm. long (sometimes less), 5-20 cm. wide (some- times less), spathulate, lanceolate, obovate-lanceolate, or sometimes elliptic, generally about 2.5 to 4.5 times longer than broad, coriaceous or thickly so, sometimes the margins revolute, rounded to obtuse, acute, or even slightly acuminate at apex, rounded, obtuse, or acute at base, entire, moderately to very dark green above, usually paler beneath, pinnately veined, the lateral nerves subopposite, ascending from the dorsally promi- nent ventrally sulcate costa, straight for about 34 their length, thereafter dividing but not uniting into a distinct marginal vein; inflorescences axillary, initiated among the leaves, pedunculate, cymose, usually 3—5- flowered; peduncles about as long as the petioles, nodose, with ovate- subulate or narrowly deltoid acute bracts; pedicels up to 1 cm. long, bracteolate; flowers about 15-20 mm. long, the 4 calyx lobes imbricate in pairs, green, glabrous except for the ciliolulate margins, or sparsely to densely puberulent with appressed hairs, ovate-orbicular, about 8-10 mm. long, minutely carinate apically without, more or less persistent into fruit- ing; petals 4, clear somewhat creamy white, somewhat (never widely) 7. at anthesis, slightly obovate- oblong to ovate-oblong, minutely thickened unguiculate within at apex, sometimes obscurely emarginate and mucronulate in the notch, about 14-21 mm. long; stamens 8, united into a tube, white, the tube 12-13 mm. long, the anthers connivent on short deltoid oblong processes adnate to the connective; ovary 4-locular,

1962 | STONE, THE GENUS PLATYDESMA 415

cornutum rostratum

spathulatum

“1G. 1. Leaves, fruits, and habit sketches of the four species of Platydesma; semi-diagrammatic.

416 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIII

4-lobed, at first minutely puberulous; style about 9 mm. long, whitish, slightly clavate and greenish at apex, with linear stigmas; fruit globose, slightly 4-lobed, to 26 mm. in diameter, the carpels adnate to the summit, in age slightly separating laterally, at last dehiscent; seeds 5—8 per locule, with a shining black testa, attached by short funicles to the glabrous endocarp.

Horotype: United States Exploring Expedition, 1838-1842, island of Kauai (us) (Melicope spathulata A. Gray)

Key TO SUBTAXA OF PLATYDESMA SPATHULATUM

1. Calyx glabrous or very sparsely puberulent only at extreme base. 2. Leaves glabrous beneat

Petioles slender, slonwate, usually 2-6 on long; blades ample, ob- long-spathulate or obovate-spathulate, 3-4 times longer than wide, mostly rounded at apex; Kauai, Oahu, Hawaii; perhaps also else- NPE. cc batdeimdotess aduenenes var. Spathulatum (typical form).

3. Petioles shorter, somewhat flattened, about 1—2 cm. long. 4. Leaves usually less se 15 cm. long, spathulate but tapered and

acute or even acuminate at apex, tapered at hee pale, rigid, the margins often rev a the blade usually 3-4 times longer than Wide? OA, cece vd yam Cewie te ae ce var. Pe es form b.

Leaves generally 20-50 cm. long or more, relatively broader (about 2-3 times oe than wide), generally rounded or obtuse at both pet TAAL. 644 20 dential pt-deeo% var. eee form c,

2. Leaves pubescent te densely villous at least on the midrib, the laminar surface sometimes glabrescent; blades ea dark, rather broadly

“ll

spathulata: Kauai, Oak. soc oscviega cca nerds ri dian es var. pubescens. 1. Calyx densely cinereous-puberulent; leaf-blades usually somewhat pubescent beneath; Oahu, Maui, Hawaii. ......................005. var. pallidum.

Platydesma spathulatum var. spathulatum. Platydesma campanulatum {. coriaceum Rock, Indig. Trees Hawaiian Is. 242. 1913

Glabrous throughout; petioles slender, elongate, usually 2—6 cm. long; blades ample, oblong-spathulate or obovate-spathulate, usually 3—4 times longer than wide, mostly rounded at apex, often somewhat rounded at base.

Kauai. Hanalei, Faurie 241 (A); Makaweli, Faurie 246 (4); Waimea, Faurie 245 (A, BISH), Forbes 911.K. (ptsu); K Kokee, Rock, 1956 (prstt) : Kaholua- manu, Forbes 328.K. (BISH); Rock 1968, 1976, 5338 (GH), 1978, 5344, 5969 eee) 6021 (A); Kaunuohua ridge, Kaluapuhi trail, Wichman (cB); Kokee mp, Degener 8640 (A); Kokee-Kilohana, aie 981 (BISH, GB); Lehua- bea cea Kilohana, Cranwell et al. HBS. 2938 (BIsH, GB); Halemanu, Rock 2298 (BisH), 2314 (GH); Neal, 1929, 1930 (BIsH), Kusche (GH); Kumuwela, Cranwell et al. HBS. 2841 (sisH, K), 2847 (GB), Stone 3360 (BISH, BM, CU, E GH, ae _— aia cB); Awa’awapuhi trail, Lane 56-614 (BIsH), Stone 1609 BISH 0 (BISH, P, US); Kohua ridge, Degeners & Hansen 23919 (BIsH); Tao) aed Forbes (Bist).

—

1962 | STONE, THE GENUS PLATYDESMA 417

Oahu. WatIANAE RANGE: Piko trail, Makua, eee forest reserve, Davis, 1932 (BISH). KOOLAU RANGE: Northen end: r summit 1 mi. se. of Black Junction, Kahuku, Degener & Carroll 20554 ae Kabul, Degener 8624 (a); Punalu’u and vicinity, Faurie 242 (a), Forbes. 1908, 1909 (B1sH); Rock 459, 640, 657, 8834 (BisH); MacDaniels 382 (sisH), Hume 76, 106 (BisH); Kali- uwa’a, Rock 65 (sisH); Waipio-Waiawa, MacDaniels 16, 53 (BISH); Warawa. Anon., 1930 (ex Herb. Bd. Comm. Agric. For. Terr. Hawaii) (s1sH); Kawailoa, Bryan 867 (stsH); Kahana, Lyon, 1926 (pisH); Wahiawa-Kahana, Forbes 2202.0. (BIsH); back of SALES Swezey, 1920 (s1sH); Budd & ner son 1155 (sisH); Waikane-Schofield trail, St. John 10161 (s1sH); Kipapa gulch trail, Cowan 700 (s1sH); Anahulu Grail. Degener 10092 (A). sonthern end: Pauoa, Hillebrand (s1sH); Konahuanui, Forbes, 1908 (s1sH), Heller 2373 (cH), Mann & Brigham 94 (cu, type of Platydesma pe BISH, GH, K); Mt. Olympus, Rock 10225 (BISH, GH), Forbes 16. a. (sisH); Koolau range, with- out locality, Hillebrand 242 (K); Rock 134 (GR

Hawaii. Kohala, Rock, 1957 (BisH), Rock a (cu); Hualalai (?), Rock 3800 (BIsH); Ola’a aie Forbes 652.H. (sisH); Ten-Miles, Fullaway & Gif- ford, 1919 (sis); Glenwood, Faurie 244 (a, BIsH), Rock, 1918 (BisH), 1914 (piso); Kilauea, Rock 12999 (sisH); “Sandwich Isl,” U. S. Exploring Ex- pedition (GH).

Platydesma spathulatum var. spathulatum, form b. (Fic. 2.) Shrubby, with slender branches and small, narrowly spathulate, pale, rigid leaves tapered at both ends, the margins often revolute; petiole 1-2 cm. long. This form is the one most commonly found on the summit crests of the Koolau Range, Oahu, where the vegetation is low, stunted, and windswept.

Oahu. KooLtau Rance: Punalu’u, summit of Castle trail, Stone 3551 (BISH) (see Fic. 2), 1142 (Bisu), Degener, Park & Nitta 8633 (us); Laie, Malaeka- hana trail, Degener et al. 10093 (cu); ridge above Kahana, Skottsberg 1853

BISH, GB); near summit of Poamoho trail, Degener 27328 (us); Punalu’u, Faurie 243 (a, BIsH); Rock 460 (A).

_~

Platydesma spathulatum var. spathulatum, form c.

Arborescent, with thick, often rugulose branches and ample leaves up to 50 cm. long, usually 7-25 cm. wide, obtuse at apex and rounded at base, dark green, quite glabrous.

This form appears to be limited to Kauai, where it occurs in the high forested tableland around the rim of Kalalau Valley.

Kau Honopu, Awa’awapuhi trail, Lane 56-586 (BISH); east rim of Kalalau Valley Degeners & Cadenheads 27150 (BISH, GB, K, US).

Platydesma spathulatum var. pubescens (Skottsberg) B. C. Stone, Madrono 16: 165. 1962

Platydesma campanulata var. pubescens Skottsberg, Acta Horti Gothob. 15: 388

Arborescent; petioles distinct, up to 5-6 cm. long; blades ample, rela- tively broad, pubescent beneath (on the lamina sometimes glabrescent, but

418 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIII

3 . Ao am A ao Me) ey x

Fic. 2. Habit, with flower and ae of Platydesma spathulatum var. spathu- latum form b. (From Punalu’u, Oa

the costa with a persistent villosity), broadly spathulate to subelliptic, dark above (the lower surface paler); calyx glabrous or nearly so

Horotype: Oahu: Waianae Mountains, slope of Kaala, O. Selling 3710 (GB).

Kauai. Wahiawa Mountains, Forbes 231.K. (sisH); Waimea drainage basin west side, Forbes 782.K. (p1sH); Halemanu, Rock 2318 (BIsH) ; aes an tral, Forbes 1046.K. (BisH); Kaholuamanu, Rock 5339, 10226, (BIsH), 5 Kumuwela plateau, Stone 3358 (A, BISH, BM, E, GR): Pendle: -Lihue csea

1962 | STONE, THE GENUS PLATYDESMA 419 trail, Forbes, 1909 (BIsH); Hanapepe, Faurie 240 (A). Oahu. Waianae Moun- tains, slopes of Kaala, Mokuleia, Morley, 1934 (BISH).

Platydesma spathulatum var. pallidum (Hillebrand) B. C. Stone, Madrofio 16: 165. 1962. (Fic. 3.)

Platydesma campanulata 6 var. pallida Hillebrand, Fl. Hawaiian Is. 71. 1888.

Similar in most respects to the typical variety, but with densely ciner- eous-puberulent calyx lobes (sometimes also the outer surfaces and

Fic. 3. Habit, flowers, and fruit of Platydesma spathulatum var. pallidum. (From Olinda, Maui, coll. Degener; courtesy of Dr. Degener.)

420 JOURNAL OF THE ARNOLD ARBORETUM [ VOL, XLIII

margins copiously ciliolulate), the pubescence extending to the mid- region of the petals; leaf blades pubescent beneath; hairs pale or fulvous, erect to somewhat spreading, about 0.2-0.4 mm. long, those on the calyx (and axes of the inflorescence) appressed, those of the leaves more erect.

SyNTyPEs: Oahu, Kaala, Hillebrand; Maui, Hamakua, Rev. John Lyd- gate, both formerly at Berlin, but destroyed in World War I]. Nrorype: pied (Degener 8615) (BIsH), cited below.

)

u. WAIANAE Mountains: Hillebrand, 1869 (cH); Peahinaia trail, oa a 12771 (a). Kootau Rance: Pupukea, Degener, Park & Nitta 8622 (A, NY); Poamoho trail, Degener et al. re (A). Maui. Woods near Ukulele, hee Olinda, Forbes 205.M. (stsH); Olinda, Rock 8534 (BIsH, GH) : Tapping (Degener 8615) (BISH, CU, K); Kaea, Forbes 2552 (BISH); Nahiku, Forbes 240.M. (isu); below Kula pipeline, Munro 796 (BISH) ; Kipahulu valley, ridge on left side, Forbes 1642.M. (Bis); ridge between west and east Wailuai- lai, Forbes 2664.M. (stsu); Honomanu, east bank of gulch, Forbes 2658.M. (BISH); west trail behind Hoaomani: Forbes 2587.M. (s1sH); Honomanu trail, Rock 10228 (BisH), Rock, 1911 (Brsit) : Kailua, Rock, 1908 (stsH); without definite locality other than “East Maui,” Rock, 1911 (BrsH) ; without locality, Forbes, 1920 (BisH). Hawaii. Kohala, Monekancant Rock 8367 (A, BISH); Papaikou near Honomu, Bryan, 1937 (Bist).

bo

Platydesma Remyi (Sherff) Deg., Sherff, & Stone, FI. Hawaiiensis, fam. 179. 1960. Stone, Adansonia IT. 1: 98, 99, 1961. (Fic, 4.)

Claoxylon Remyi Sherff, Publ. Field Mus. Bot. 17: 557. 1939.

ecg insigne H. Baillon ex Drake del Castillo, Tilustr. Fl. Ins. Mar. Pac. 291. 1892 (nomen nudum).

Ae see campanulatum var. sessilifolia Rock, Indig. Trees Hawaiian Is. 243. 1913.

Platydesma auriculaefolia sensu ar lebrand, Fl. Hawaiian Is. 72. 1888, not Pelea auriculaefolia A. Gray,

A sparingly branched light-wooded shrub or laxly branched small tree up to 2 or 3 m. tall, with branches loosely foliose and leaves subdistant, opposite; habit similar to that of preceding species; leaves sessile or sub- sessile, the short obsolete petiole broad, flattened, up to 5 mm. long sparsely hispidulous; blades obovate or broadly spathulate, Bene 14— 38 cm. long, 4.5--18.5 cm. broad, often about 30 « 12 cm. (2—3 times longer than broad), punctate- glandular, entire, rounded or iene obtuse to subacute at apex, narrowed and at last rather abruptly rounded, sub- truncate, or even subauriculate at base, glabrous above, sparingly puberu- lent heneath at least when young, the costa sulcate above and raised beneath; inflorescences axillary, 1-3 (—5?)-flowered, on short peduncles and axes, the pedicels up to 2 cm. long, the axes nodose with opposed ovate bracts 1-2 mm. long; flowers as in Platydesma spathulatum, but th calyx lobes puberulent on both faces; capsule about 25 mm. in Pieces minutely puberulent, with rounded carpels.

Horotype: Hawaii. Location unknown, Jules Remy 604, 1851-55 (Pp).

1962 | STONE, THE GENUS PLATYDESMA 421

3

Dbado! « Mure,

=e

Fic. 4. Holotype of Platydesma Remyt. eae by Mme. Godot de Mauroy, commissioned by O. Degener; courtesy Muséum National d’Histoire Naturelle, Paris, and Dr. Degener. Coll. J. Remy 604, Hawaii 1851-55.)

422 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIII

Hawaii. Kohala Mountains, Hillebrand 244 (cH, x; this collection is the basis of Hillebrand’s Platydesma auriculaefolia) ; upper Hamakua ditch trail, above Koiawe valley, Cranwell & Skottsberg HBS. 3192 (BISH, GB); Kohala, Forbes 499.H. (sisH); mauka of Waipio valley (inland side), Lyon, 1919 (Bis); Kohala, Rock 8354 (sis, GH), 4222 (GH, type of var. sessilifolia), Rock, 1912 (a).

This species is very close to Platydesma spathulatum, differing mainly in the obsolete petioles, the pubescence of the leaves and calyx (similar to that of P. spathulatum var. pallidum) and the puberulence of the fruits.

The drawing of the holotype given here (Fic. 4) was made at the re- quest of Dr. Otto Degener, and has been published previously in Flora Hawatiensis.

A specimen from Kauai (Alakai, MacDaniels 781) may belong here, but it is fragmentary and somewhat dubious. The locality would not be implausible in relation to the type of habitat, but the wide distributional gap would be most unusual. At present it seems apparent that this species is endemic to the Kohala region of Hawaii. Because of the number and recentness of some collections there seems to be no basis for considering it extinct, but new collections would be desirable. It may be possible eventually to show that this taxon should better be considered as a sub- species of Platydesma spathulatum.

Section CorNutTIA, sect. nov.

Carpidii maturitatem divaricati cornuti vel rostrati. Type species: Platydesma cornutum Hillebrand.

3. Platydesma cornutum Hillebrand, Fl. Hawaiian Is. 72. 1888. (Fic. a)

Melicope grandifolia sensu Wawra, Flora 56: 138. 1873, non A. Gray, Bot. U.S. Explor. Exped. 15: 354. 1854.

Erect unbranched or sparsely. branching shrub with soft wood up to 7m. high, the main stem about 2 cm. diameter, the few branches erect or ascending, closely foliose near the ends, the youngest branches green, soon clothed with a pale grayish bark, the wood when freshly broken having a starchy odor; glabrous throughout except for the leaf buds and the proximal parts of the inflorescences, these sparsely hirtellous, but glabrate; leaves petiolate, petioles short, broad, flattened, 2-10 mm. long; blades glabrous, dark green when mature, linear-spathulate or narrowly obovate, 12-40 cm. long, 5-13 cm. broad, commonly about 30 « 9 cm., tapering gradually to the base, decurrent or finally obtuse, rounded or more commonly bluntly obtuse or acute at apex, the costa sulcate above, prominent beneath, the lateral nerves similar, ascending from the costa, merging distally with the deeply arched connecting vein. Inflorescences axillary, borne among the leaves but maturing on the lower stems after the leaves fall, fasciculately cymose, (3—)9—15-flowered, the peduncle and

1962 | STONE, THE GENUS PLATYDESMA 423

axes obsolete, pedicels thus subsessile, bractlets and bracteoles at first hirtellous, pedicels glabrous or nearly so, 5-10 mm. long, flaring into the base of the calyx; sepals glabrous except for the ciliolulate margins, broadly rounded, 3-4 mm. long and 5-6 mm. broad, imbricate in pairs, the inner pair slightly smaller; petals white, glabrous except for the minutely ciliolulate margins, 9-12 mm. long; stamens 8, connate into a staminal tube, the anthers on very short deltoid filaments, unequal, 4 somewhat larger, equalling the corolla, 4 slightly shorter; tube at base very slightly connate with base of corolla; ovary 4-lobed. on an obscurely 8-lobed disk, the styles at first connate, with 4 linear stigmas, later separating as the fruit matures; capsule with thin, chartaceous walls, glabrous, about 12 mm. long, the carpels united along their mutual faces nearly to the apex or free down to the middle, the distal portions divergent, each carpel terminating in a sharply acuminate horn-like tip; carpels usually 8-seeded; seeds with a shining black crustaceous testa.

Key TO VARIETIES OF PLATYDESMA CORNUTUM

Leaves abruptly truncate or obtuse at base, petiole distinct. var. cornutum. Leaves tapered and gradually, evenly decurrent at base. ...... var. decurrens.

Platydesma cornutum var. cornutum.

Leaves tapering toward the base but at last briefly but abruptly trun- cate or obtuse, the petiole distinct.

Ho.otypre: Oahu. Without definite locality except ‘“Helemano, Wailupe, and Pauoa” (Koolau Range), Hillebrand (x; isotype, GH).

Oahu. Kootau RANCE: Northern end: Pupukea, Skottsberg 1815 (BISH, GB) ; Pupukea-Malaekahana, Lyon, 1926 (pisH); south Opaeula gulch, Pa’ala’a, St. John 10642 (a, BISH); Pupukea-Kahuku, Degener 8623 (A, cu); Paumalu, Hosaka 123 (pts); Punalu’u, Rock 8833 (BisH), Rock and Forbes, 1908 (A), Rock, 1908 (GH), Anon., 1908 (Herb. Bd. Agric. For. Terr. Hawaii) (BISH); between Kaipapau and Punalu’u, Forbes, 1908 (BisH); Kaluanui, summit and beyond, Castle trail, Punalu’u, Webster 1610 (sisH); Wahiawa gulches, Forbes 1712.0. (s1sH); southern end, Puu Konahuanui, Forbes 1010.0 (sIsH); head of Manoa valley, Rock & Shaw, 1912 (BIsH).

mountains of Oahu. There does appear, however, to between the varieties, since var. decurrens, though mostly restricted to the

of the Koolau Range nearest the Waianae mountains. It is here that intermediate forms, if any are to be found, could be expected to occur. Platydesma cornutum var. decurrens, var. nov.

Folia longe spathulata, base attenuata laminis in petiolam decurrens.

Hoioryvpe: Oahu. Waianae Range, Pu’u Kanehoa, southeastern side,

424

L O

JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIII

eaf, flower, and fruit of Platydesma cornutum var. cornutum. (From ahu; courtesy of Dr. Degener.)

1962 | STONE, THE GENUS PLATYDESMA 425

on steep moist banks of stream in valley about 200 m. below summit at altitude of about 700 m., March 26, 1960, B. C. Stone & G. Pearsall 3263 (BISH ).

Oahu. Makaha valley, ridges of Mt. Kaala, Forbes, 1909 (sts); Mokuleia valley, Forbes 1822.0. (ptsH); Pahole gulch in Mokuleia, Degener & Hatheway 20664 (pm); Makaleha valley, Russ, 1929 (sIsH); valley south-east of trail to Pu’u Kanehoa, type locality, Stone 3430 (pIsH, US); Mt. Kaala, Degener, Park, Topping & Swezey 8631 (A)

This variety is still another of the many examples of taxa endemic to the Waianae mountain range of western Oahu. This range, which is geologically much older than the larger Koolau Range. may be about the age of the island of Kauai. Several examples are known of taxa re- stricted to Kauai and to the Waianae Mountains, as well as species with distinct varieties in the Waianae Mountains and in Kauai. The specimens cited above are mostly sterile or in flower, but Forbes, 1909, bears an immature fruit. There are no apparent differences in floral or fruit charac- ters between the two varieties, but the difference in the leaves is both obvious and constant. The holotype and Stone 3430 are both from the same location, where a small group of these tall, erect, unbranched shrubs was growing in a moist, rocky habitat near a stream, in company with species of Urera, Touchardia, Cyrtandra, Alectryon, Straussia, Morinda, Pelea, and Athyrium. The flowers, like those of the next species, Platy- desma rostratum, are borne on the slender trunks and mature at some distance below the leaves. The sepals are green, the petals and staminal tube clear white.

4, Platydesma rostratum Hillebrand, Fl. Hawaiian Is. 72. 1888. (Fic. 6.)

Erect unbranched or sparsely branching shrub with soft wood, branches ascending, juvenile stems green, later clothed in thin, pale, grayish bark, trunks to one or two (or more) meters high, 1-2 cm. thick, freshly broken wood with a starchy odor; leaves clustered near the ends of the branches, opposite, subsessile or with short, flattened petioles about 3-18 mm. long; blades linear-oblong or long, narrowly elliptic to subspathulate, ae 20-40 cm. long, and 4-12 cm. broad at maturity, usually about 33 x at base broadly and abruptly truncate or obtuse (rarely Se Sy scarcely or not attenuate, at apex bluntly acute to coarsely acuminate (the tip bluntly acute) or somewhat rounded, in bud hirtellous but soon glab- rate, glabrous at maturity; costa sulcate above, raised beneath, the lateral nerves numerous, subopposite or subalternate, almost at right angles to the costa, united distally 3-10 mm. from the margin by a deeply and sym- metrically arching connecting vein; inflorescences axillary, at first hidden among the leaves, maturing below the leaves on defoliate young branches, cymose, the cymes 3—-9-flowered (and sometimes fasciculate, two or three together), less than 2 cm. long overall, the peduncle stout, 1-3 mm. long

426 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIII

and 1—1.3 mm. broad, the axes and pedicels up to 9 mm. long, the pedicels with an opposed pair of bractlets near the flaring apex at base of calyx, the axes with 1-3 pairs of decussate bractlets, these, with the bracteoles, axes, and peduncle, hirtellous when young, soon glabrescent; calyx slightly puberulent, the lobes imbricate in pairs, the outer pair larger, each lobe about 3 mm. long and 5 mm. wide, the inner pair with each lobe slightly smaller; petals clear white, glabrous, oblong, obtusely rounded at apex, about 5-nerved, valvate, about 12-13 mm. long and 5 mm. broad; stamens united into a staminal tube about 11 mm. long, anthers on brief deltoid filaments, these alternating in size, 4 about 2.5 mm. long, 4 about 2 mm. long, versatile, the anther affixed in the middle of the dorsal side; ovary 4-lobed, on a disk about $ mm. broad, the carpels free distally, about 1.8 mm. high, subconic and shortly attenuate into a brief style; capsule of 4 carpels, each with a distal rostrum about 8 mm. long, connate nearly their whole length but the beaks free; capsule about 4 mm. long, exclud- ing beaks; carpels 8-seeded; seeds black with a shining crustaceous testa.

O mm 5 O mm 5

Fic. 6. Floral details of Platydesma rostratum. (Three upper figures from Skottsberg 3100, Kauai. Petals and stamens from Rock 6081. Original.)

1962 | STONE, THE GENUS PLATYDESMA 427

HouotypPe: Kauai, without locality, Anudsen 68 (Berlin, probably destroyed).

Kauai. Waimea, Eee Kalalau trail, Cranwell, Selling & Skottsberg HBS. 3100 (BISH, GB); uohua ridge, Kaluapuhi trail, Wichman (Skottsberg 2937) (GB): eee Rock 2000, 1999 (A), 2323, 6081, 6084 (GH); Awa’awa- puhi trail, Stone 1599 (BISH, K, us): Lane, 1956 (sts); Pearsall, 1956 (sisH) ; Stone 3329 (pisH); Hanalei valley, Forbes 133.K. (BISH); upper Lihue ditch trail, MacDaniels 837 (Bis); without locality or date, Rock 17302 (a).

This species is to be found in the mesic to wet forests of the high plateau region of Kauai, where it is well distributed but not common. The leaves are usually rather dark green and glossy above, the flowers clear white. The small, erect, sparingly branched shrubs form an understory element in rain-forest associations.

The relationship of the species is clearly with Platydesma cornutum, but the two are easily distinguished by both vegetative and fruit characters. If (as seems probable) evolutionary divergence closely followed the se- quence of island formation, this species may antedate P. cornutum; or the two may be derived from a common ancestor. The relationship of these species to the arborescent, globose-fruited P. spathulatum and P. Remyi is obvious, but not particularly close, despite the small size (in number of species) of the genus.

LITERATURE CITED

DeEGENER, O. Flora Hawaiiensis. Unpaged. Privately printed. Honolulu. 1960-.

DRAKE DEL CASTILLO, E. Illustrationes florae insularum maris Pacifici. Paris. 1890-1892

Gray, A. Botany: ees ae States oe eae oes Gade Charles Wilkes... 1: 1- (Cf. pp. 352-354) 1

HILLEBRAND, W. Flora of the oe Islands. 1-673. eee London, and New York.

LEVEILLE, cs Platydesma oahuensis and P. Fauriei. Repert. Sp. Nov. 10: 153- 154. =

MANN, ae Revision of Schiedea and the Hawaiian Rutaceae. Proc. Boston

Soc. Ne Hist. 10: 317. 1866. Peis _Blayesmo, and Brighamia. Mem. Boston Soc. Nat.

Hist. 1: 529-53

Rock, J. F. The ae trees of the Hawaiian Islands. 1-518. Privately printed. Honolulu. 19

. Revisio plantarum Hawaiiensium a Léveillé descriptarum. Repert. Sp.

Nov. 13: 352-361. 1914.

. Pelea and Platydesma. Bot. Gaz. 65: 261-267. 1918.

Stone, B. C. Studies in the Hawaiian Rutaceae III]. On the New Caledonian species of Pelea and a misunderstood species of Platydesma. Adansonia 1: 94-99,

. Taxonomic and nomenclatural notes on feels esma (Hawaii) and a new name for a Melicope (Solomon Islands). Madrofio 16: 161-166. 1962. Wawra, H. Beitrage z. Flora d. Hawaiischen Inseln. Bae 56: 138. 18 73%

COLLEGE OF GUAM AGANA, GUAM

428 JOURNAL OF THE ARNOLD ARBORETUM [VOL. XLIII

THE THYMELAEACEAE IN THE SOUTHEASTERN UNITED STATES !

Lorin I. NEVLING, JR.

THYMELAEACEAE A. L. de Jussieu, Gen. Pl. 76. 1789, “Thymelaeae,’ nom. cons.

(DAPHNE FAMILY)

Erect shrubs [sometimes dwarf or climbing; trees, lianas, rarely herba- ceous annuals|; stems generally with a layer of tough cortical fibers, often with internal phloem. Leaves alternate |approximate, opposite or irregu- larly pseudowhorled|, simple, entire, laminar [to needle-like or scalelike |,

punctate]; petiolate or sessile, exstipulate. Inflorescences terminal at shoot apex, extra-axillary [axillary, or cauliflorous], pedunculate or ses- sile, racemose |spicate, umbelliform, or capituliform], simple |or com- pound]; bracts and bracteoles present or absent, [sometimes resembling an involucre|; pedicel articulated when present. Flowers bisexual [or uni- sexual by abortion and the plants polygamo-dioecious or dioecious], regular or rarely somewhat irregular. Calyx lobes free or, more generally, connate and adnate to the corolla and androecium to form a variously shaped calyx tube, open [or imbricate, valvate], 4/6, 5 or 3]-merous, often greenish yellow |but sometimes brightly colored, sometimes articu- lated at the middle]. Corolla nonpetaloid; petals adnate to the calyx, the lobes alternisepalous, free [or connate into a faucal corona, either simple or variously divided, of various shapes, or absent, generally inserted near the orifice of the calyx tube|. Androecium generally diplostemonous [rarely haplo- or very rarely poly- or hemistemonous, in @ flowers re- duced to staminodia or absent|. in 1 or 2 whorls, the upper whorl anti- sepalous and the lower alternisepalous; anthers fila mented [or sessile], longitudinally dehiscent, introrse |extrorse or horseshoe-like|, basifixed

‘Prepared for a_ biologically sla generic flora of the southeastern United States, a joint project of the Arnold Arboretum and the Gray Herbarium made ciel through the support of George R. Cooley and the National Science Founda-

tion, and under the direction of C. E. Wood, Jr., and R. C. Rollins. The scheme tstiows that outlined at the beginning Pe the series (Jour. “Arnold Arb. 39: 296-346. 1958) except that the family description and bibliography are more extensive than usual. The area covered in this, as in former treatments, is bounded by and includes North Carolina, Tennessee, Arkansas, and Louisiana. Material included in descrip- tions which is inapplicable to the species of this area is placed in brackets. References which have not been verified are designated by an asterisk. The figure of Dirca was drawn by Dorothy H. Marsh under the supervision of R. B. Channell and C. E. Wood, Jr

1962 | NEVLING, THYMELAEACEAE 429

[or dorsifixed, the connective sometimes thickened or produced beyond the pollen sacs]; pollen globose, polyporate. Disc annular [cupular, tubu- lar, scalelike, or absent], free [or sometimes adnate to the calyx tube], surrounding the base of the gynoecium. Gynoecium syncarpous, | 12-, 8-, 5-, 2- or| 1(pseudomonomeric)-carpellate, |reduced to a pistillode in ¢ flowers|; stigma usually capitate; style 1, [terminal or] eccentric, some- times obsolete, [rarely with parastyles|; ovary superior; ovules 1 in each locule, mostly pendent from the locule apex, anatropous and with ventral funicles [or rarely hemi-anatropous to nearly orthotropous], 2-integu- mented, [often carunculate, often with an obturator at the base of the style]. Fruit a berry [loculicidal capsule, nut, or drupe, often accom- panied by an accrescent calyx]. |Seeds sometimes arillate;| embryo Straight, with flat or thickened, narrow or broad cotyledons; endosperm present [or absent]. Type Genus: Thymelaea P. Miller, nom. cons.

Approximately 55 genera, with some 500 species now recognized, all tropical or temperate, with the greatest speciation in tropical and sub- tropical regions. It is expected that the number of genera will be reduced by nearly one-fifth in the near future, and a further reduction could be realized by the elimination of a number of genera of “convenience.” A single genus composed of two species, only one in our area, is native to the United States. Thymelaea Passerina (L.) Coss. & Germ., an annual weed, has been reported in Iowa and Nebraska; and Daphne Mezereum L. is a more or less common escape from cultivation in the northeastern United States.

The family is divided into four subfamilies, according to Domke’s sys- tem which is followed here. Some authors exclude the three genera of sub- fam. Gonystyloideae Domke and the three of subfam. Aquilarioideae Gilg, tribe Microsemmatideae Domke, treating them as a separate family, Gony- stylaceae Gilg, a segregation which seems both unnecessary and unwise, contradicting as it does both anatomical and palynological evidence. A third subfamily, Gilgiodaphnoideae Domke, is monogeneric. The Thyme- laeoideae, to which our genus belongs, contains the bulk of the genera and species of the family, although a number of genera still remain un- placed in the family system. American genera of uncertain position are Goodallia Bentham, Lasiadenia Bentham, and Linodendron Grisebach.

On the basis of the few and scattered chromosome-number reports, a base number of nine is assumed. The majority of the species are diploid; the only known exceptions are a triploid Daphne and polyploid series in Edgeworthia, Pimelea, and Wikstroemia. Apomixis has been demonstrated in one species of Wikstroemia and is suspected in others.

The family is of little general economic importance except, perhaps, in very localized areas. The genus Daphne is of considerable horticultural interest and is grown in many temperate regions for its sweetly scented flowers. In local areas, almost the world over, the tough extraxylary fibers of various species are employed as cordage and in the manufacture of paper or “cloth.” The heartwood of some Asian species is highly prized

430 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLII

as incense wood. The fruits, which seem to have some poisonous principle, are used as a purgative or emetic in primitive areas. Unidentified alka- loids are reported from plants of a number of genera.

REFERENCES: a Suaw, H. K. Thymelaeaceae-Gonystyloideae. Jn: C. G. G. J. VAN STEENIS, . Males. I. 4: 349-365. 19

poets A. M. Daphnes. jour, Roy. Hort. Soc. 78: 5-18. 1953. [Section-by- section discussion of spp. useful in horticulture. |

Aymonin, G. Essai d’une monographie écologique du Daphne Cneorum i Revue Gén. Bot. 66: 281-328. 1959. [Only detailed ecological study of any species in the hee with extensive bibliography. |

BANDONI, A. J., _A, O’DonELL. La anatomia de la timeleacea ‘‘Ovidia pillo- pillo,” planta ae de los Andes patagénicos. Physis 15: 377-385. 1939.

Burrows, C. J. Studies in Pimelea. I—The breeding system. Trans. Roy. Soc. New Zealand 88: 29-45. 1960.

Cortrst, R. Anatomie de quelques plantes alpines. Chapitre I. Le Daphne Blagayana Freyer. Bull. Soc. Bot. Genéve I. 33: 145-157. 1942. | Includes interesting grate |

oi de Vhistométrie dans l’examen de certaines drogues naturelles. Pharm. hee Helvet. 18: 361-368. 1943.* [ Bark of 4 spp. of Daphne studied. ]

CruicKsHANK, R. H. Chromosome numbers in the genus ‘Pimelea.’ Pap. Proc. Roy. Soc. Tasmania 87: 13-16. 1953. [Demonstration of polyploid series. |

Domke, W. Untersuchungen itber die systematische und geographische Glieder- ung der Thymelaeaceen. Bibliot. Bot. 27(111): 1-151. 1934. [Basic taxo- nomic treatment to the generic level. ]

Facertinp, F. Zytologie und Gametophyten bildung in der Gattung W7k- stroemia, Hereditas 26: 23-50.

Fucus, A. Beitrage zur Embryologie der Thymelaeaceae. Osterr. Bot. Zeitschr. 87: 1-41. 1938.

Gitc, E. Thymelaeaceae. Pflanzenfam. III. 6a: 216-245. 1894.

Goswamt, P. C. Hand-made paper industry in Tawang area of the Kameng Frontier Division. Indian Forester 87: 765. 1961. [Paper-making with Daphne cannabina. |

Guerin, P. Recherches sur la structure anatomique de l’ovule et de la graine des Thyméléacées. Ann. Jard. Bot. Buitenzorg II. 14: 3-35. 1916. [Broad study including representatives of many genera.

Hamaya, T. A dendrological monograph on the Thymelaeaceae plants of Japan. Bull. Tokyo Univ. Forests 50: 45-96. 1955. [Discussion of taxonomic snug and taxonomic treatment. |

endrological studies of the Japanese and some foreign genera of the ee Ibid. 55: 1-80. 1959. [Detailed morphological and ana-

Hernic, K. H. Studies in the floral morphology of the Thymelaeaceae. Am. Jour. Bot. 38: 113-132. 1951. [Floral morphology of 34 spp. in 11 genera of subfams. Aquilarioideae and Thymelaeoideae. |

Hiraoka, T. Somatic syndesis in Daphne odora I. The chromosome behavior in mitosis. Proc. Japan Acad. 34: 700-705. 1958.

: ic syndesis in Daphne odora IJ. The chromosome behavior in

meiosis. /bid. 706-711. [Chromosome behavior in a triploid sp.]

1962 | NEVLING, THYMELAEACEAE 431

Hou, D. Thymelaeaceae. Jn: C. G. G. J. VAN STEENIS, Fl. Males. I. 6: 1-48. 1960. [Floristic treatment, ner ng the Gonystyloideae. ]

Jones, L. R. Daphne Mezcreum in Vermont. Rhodora 2: 142. 1900.

LEANDRI, J. Structure particuli¢ére du rhizome d’un Daphne. Bull. Soc. Bot. Fr. 75: 243-248. 1928

. Recherches anatomiques sur les Thyméléacées. Ann. Sci. Nat. Bot. X. 12: 125-137. 1930. [Basic anatomical study. ]

Martin, H. The genus Daphne. Rep. 3™ Int. Rock Gard. Pl. Conf. 125-130. 1961. [Of horticultural interest. ]

MEISSNER, C. F. Thymelaeaceae. DC. Prodr. 14: 493-605. 1857. [First im- portant taxonomic resumé of the family. ]

NaKAHiIRA, K. Studies on the breeding of trees of special use. 4. Artificial hexaploid plants of mitumata (Edgeworthia papyrifera). (In Japanese; English summary.) Jap. Jour. Breed. 7: 112-118. 1957. [See also ibid. 157- 160. 1958.*]

Neviinc, L. I., Jr. A revision of the genus Daphnopsis. Ann. Missouri Bot. Gard. 46: 257— 358. 1959. [Includes key to New World genera.

Nye, H. A. A new station for Daphne. Rhodora 25: 45, 46. 1923. [Somerset Co., Maine. ]

Oxura, E., & M. Kono. Cytogenetical studies of Edgeworthia papyrifera Sieb. et. Zuce. I. Karyotype analysis de two varieties, Kochi and Shizuoka. Biol. Jour. Okayama Univ. 4: 60-66 58.

Cytogenetical sion of Daphne odora Thunb. based on its ary ote. Ibid. 5: 51-56. 1959.

PoHL, R. W. Thymelaea Passerina, new weed in the United States. Proc. Iowa Acad. Sci. 62: 152-154. 1955.

Riptey, H. N. Garu and Chandan. Jour. Straits Branch Roy. Asiat. Soc. 35: 73-82. 1901. [Important article on incense wood from Agquilaria ions Lam., A. dirta Ridl., and Wikstroemia Candolleana Meissn

STOIANOV, G. A. orca features of rose daphne anaes Cneorum) bloom- ing. (In Russian.) Priroda Leningrad 45: 114, 115.

Tuopay, D. On the behavior during drought of leaves 2 two Cape species of Barna with some notes on their anatomy. Jour. Bot. London 35: 585-601. 1921

. The geographical distribution and ecology of Passerina. Ann. Bot. 39: 175-208. 1925

VENKATESWARLU, J. Embryological studies in the Thymelaeaceae, II: Daphne cannabina Wall. and Wikstroemia canescens Meissn. Jour. Indian Bot. Soc. 26: 13-39. 1947.*

YaMaHA, G. Experimentelle zytologische Beitrage. III. Mitteilung. Uber die Wirkung einiger Chemikalien auf die Pollenmutterzellen von Daphne odora, Thunb. Bot. Mag. Tokyo 41: 181-211. pl. 4. 1927.

ZAZHURILO, K. Uber die Anatomischen Probleme in der Karpologie der Gegenwart. (In Russian; German summary.) Acta Univ. Voroneg. 7: 21-42. 1935. [Extensive bibliography. ]

1. Dirca Linnaeus, Sp. Pl. 1: 358. 1753; Gen. Pl. ed. 5. 167. 1754.

Slow-growing, deciduous shrubs with slender, flexible branches swollen at the nodes. Mature plant-parts usually containing simple or compound crystals of calcium oxalate. Leaves ovate, broadly elliptic, or obovate,

432 JOURNAL OF THE ARNOLD ARBORETUM [ VOL, XLIII

sparsely sericeous and glabrescent, the petioles calyptrate over the axil- lary bud. Flowers appearing before or simultaneously with the leaves, borne in few-flowered, nodding racemes from the axils of last years’ leaves, the primary and secondary peduncles conspicuous [or nearly obsolete]; bud scales woolly. Sepals connate into an infundibuliform or narrowly campanulate calyx tube, straight [or strongly bent], yellowish green to yellow, glabrous, the lobes connate, the sinuses shallow and becoming ob- scure with age (in our species). Petals minute, lobes inserted on the calyx tube between the insertion of the filaments. Stamens 8, in 2 closely adja- cent whorls; filaments filiform, inserted between the lower and upper one-third of the calyx tube, glabrous; anthers exserted even in bud. Disc a minute, irregularly lobed annulus surrounding the base of the ovary. Gynoecium pseudomonomeric, glabrous; stigma punctate, exserted; style filiform, glabrous; ovary 1-locular with a single anatropous ovule. Fruit a bilaterally symmetrical berry, yellow or yellowish green to white, be- coming reddish or purplish. Seeds with a fleshy outer and hard inner seed coat; endosperm scanty; embryo with thick, plano-convex cotyledons. Type species: D. palustris L. (Name from Greek mythology, after Dirce, second wife of Lycus, who was bound by Amphion and Zethus, sons (by Zeus) of Lycus’ first wife, Antiope, to the horns of a wild bull to be killed and who then was transformed into the fountain of Dirce near Thebes.) — LEATHERWOOD.

Two species of temperate North America, of obscure affinities, dis- tinguished from other American genera by the 4-merous, bisexual flower with eccentric style. Dirca occidentalis Gray is restricted to six counties in the San Francisco Bay area of California, while D. palustris ranges sporadically from New Brunswick, Quebec, and Ontario, southward to northwestern Florida and Alabama, and westward to Minnesota, Iowa, Missouri, and Oklahoma. The two generally are considered to be closely related, and in some instances their distinctness has been questioned.

Our species is generally restricted to rich woods where considerable moisture is available. It is found, however, in a variety of habitats and seems to ‘be a facultative calciphile. Although of sporadic occurrence, local populations often are extensive.

There is considerable disagreement in the literature concerning color of the fruits, which, being eaten by birds as soon as ripe, are rarely seen in the wild. The color is described as reddish or purplish in some studies and yellowish or yellow-green in others. The large majority of field obser- vations favor the yellow or yellow-green color, with the fruit often turning white and then purplish after falling or during the preparation of specimens for the herbarium. Further field observations are in order.

The extreme flexibility of the stems of Dirca (as well as of many other Thymelaeaceae), well known to field botanists, is due largely to the slight degree of lignification, as determined by chemical tests, of the secondary wood. Poor lignification is found sporadically throughout the entire family. The wood is very light, with a specific gravity of about 0.41, as compared

1962 | NEVLING, THYMELAEACEAE 433

to 0.21 in Leztneria floridana Chapm., the lightest native wood of the United States. The low density of the wood in Leitneria is due to exten- sive amounts of parenchyma, but in Dirca the bulk of the wood is com- posed of fibers. The stems of Dirca, as well as many other Thymelaeaceae, have an extensive development of extraxylary fibers which accompanies all vascularization. Although these fibers can be considered as an added supporting device, they, too, are nonlignified (in Dirca).

Fic. 1. Dirca. a-1, D. palustris: a, pee twig, X 1; b, flower, X 3; c, ae calyx tube opened lengthwise, to show insertion of stamens and minute tals ae tween filaments, X 3; d, cena in vertical section, diagrammatic,

e, fruiting twig, - Ya; f, mature berry, X 2; g, seed, with fleshy outer ‘coat removed to show bony inner coat, x 2; h, fruit, with single seed in

xs o Ow B

ection — semidiagra a ony pa seed coat (hatched), thin layer of endosperm (white) surrounding large embryo, ovular trace (broken line), X 3; i, fruit in cross section to show seed eee seed coats, endosperm, and cotyledons, xen

REFERENCES:

Under family references see Domke (p. 129), Gite (p. 239), HEINIG (1951), MEISSNER (p. 527

ANDERSON, E. Leatherwood (Dirca palustris). Arnold Arb. Bull. Pop. Inf. IV. 1: 25-27. 1933.

CHoguETTE, L. Contribution 4 l’étude du Dirca palustris L. ou bois de plomb. Trav. Lab. Mat. Méd. Paris 17(4): 1-93. 1926. [Nearly complete anatomical ne morelolon cal study plus pharmacology and comprehensive bibliography. ]

N, J. G. Leatherwood for early ee bloom. Jour. N. Y. Bot. Gard. 50: ce 59. 1949. [D. palustris as an ornamental. ]

FERNALD, M. L. The fruit of Dirca ae Rhodora 45: 117-119. 1943.

434 JOURNAL OF THE ARNOLD ARBORETUM | VOL, XLIII

Hom, T. Dirca palustris L. A morphological study. Am. Jour. Sci. 2: 177-182. 1921. [Anatomy of root, stem, and leaf. |

Howe Lt, J. T. Plantae concer —JI. Madrofio 2: 11-15. 1930. [Descrip- tion of fruit of D. occidentalis,

. Concerning fruit-color in ae occidentalis. Leafl. West. Bot. 7: 176. 1954.

Lecours, J. E. W. Le bois de plomb “Dirca palustris.” Bull. Sci. Pharmacol. 31: 112-116. 1924.*

Loew, F. A. Observations on the growth of an injured plant of Dirca palustris. Proc. Indiana Acad. Sci. 55: 48, 49. 1946. [Growth of injured plant more vigorous than uninjured. |

Minn, H. E., & B. ForperHase. Notes on western leatherwood, Dirca oc- cidentalis Gray. Madrono 3: 117-120. 1935. [Only report of this sp. repro- ducing by underground stems. |

McVaucuH, R. The fruit of the eastern leatherwood. Castanea 6: 83-86. 1941. [Best description of fruit; includes literature citations. |

Suggested pay laeeny of Prunus serotina and other wide-ranging phylads in North America. Brittonia 7: 317-346. 1952. |Dzrca, 345; ques- tions distinctness of spp.; for follow-up see VOGELMANN.

SHarp, A. J. Relationships between the floras of California and southeastern United States. Contr. Dudley Herb. 4: 59-61. 1951. [Indicates close rela- tionship of the 2 spp. of Dirca.|

Skori¢, V. Mikoriza u nekih Thymelaeacea. (In Croatian; German summary. ) Acta Bot. Zagreb. 1: 22-24. 1925. [Discussion of mycorrhiza in Dirca pa- lustris and several spp. of Daphne. ]

VoGELMANN, H. A comparison of Dirca palustris and Dirca occidentalis (Thy- melaeaceae). Asa Gray Bull. IT. 2: 77-82. 1953.

1962] CHANNELL & WOOD, LEITNERIACEAE 435

THE LEITNERIACEAE IN THE SOUTHEASTERN UNITED STATES !

R. B. CHANNELL AND C. E. Woop, Jr.

LEITNERIACEAE Bentham in Bentham & Hooker, Gen. Pl. 3: vi, 396. 1880, “Leitnerieae,” nom. cons.

(CorKwoop FAMILy)

A monotypic family distinguished by secretory canals in the pith and leaves, nonaromatic foliage, erect catkins, single-styled gynoecium with a superior, l-locular ovary, single parietal anatropous ovule, and seed with a large embryo and thin, fleshy endosperm.

1. Leitneria Chapman, Fl. So. U.S. 427. 1860.

Dioecious deciduous shrub or small tree to 6 m. tall, up to 15 cm. in diameter: current stems hairy; bark brown; wood very light in weight. Leaves alternate, 5-ranked, simple, exstipulate, pubescent; blades 10-15 cm. long, lanceolate to elliptic-lanceolate, acute, entire, somewhat coria- ceous and glossy in age, finally rugose; petioles half-cylindric, ca. 2.5 cm long; nodes with 3 traces from 3 gaps. Flowers appearing before the leaves in erect, preformed aments with weak axes. Staminate catkins curving outward, the axis lax, with 40-50 cymules in the axils of spirally deltoid- ovate scales, each cymule of (3—)10—-12(-—15) free stamens (apparently representing about 3 flowers); bracteoles and perianth absent; filaments short, slightly dilated at the base; anthers oblong, 2-locular at anthesis, slightly versatile, nearly extrorse, dehiscing longitudinally; pollen nearly globose, smooth, 3—6-colpate. Carpellate catkins stiffly erect, spikelike,

1Prepared for a generic flora of the southeastern United States, a joint project of the Gray Herbarium and the Arnold Arboretum made possible through the support

was prepared by Dorothy H. Marsh from living plants cultivated at the Arnold

by Dr. Robert K. Godfrey, of Florida State University. We are further grateful to Dr. Wilbur H. Duncan, of the University of Georgia, for information on the occur- rence of Leitneria in Georgia, and to Dr. Delzie Demaree, of Hot Springs, Arkansas, for data from that state and from Missouri.

436 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIII

relatively few flowered; carpellate flowers sessile, solitary in the axils of the spirally arranged primary bracts, each with 2 bractlets at the base and surrounded by a perianth (involucre) of (3) 4(-8) minute tepals (scales?) (2 often somewhat larger than the others); style linear-lanceolate, reddish, deciduous, the grooved stigmatic surface facing the bract; ovary superior, ovoid, green, 1-locular with a single parietal, pendulous, anatropous, 2- integumented ovule on the side toward the bract; embryo sac of the ‘Poly- gonum”’ type. Fruit an erect, smooth, oblong-ovoid, somewhat compressed, dry drupe (the thin flesh heavily reticulated with vascular bundles), chestnut-brown, green when young, with a terminal stylar scar. Seed with a thin layer of endosperm and a large, straight embryo. TYPE SPECIES: L. floridana Chapm. (Named for Dr. E. T. Leitner, —1838, a German naturalist who traveled in Florida and was killed during the Seminole War.) — CorKWOOD.

FIG Leitneria. a-k, L. eee a — shoot, X 4; b, twig with staminate catkins, Us t d, s of staminate catkin oe without an within, showing stamens, ye €, oe eer ee catkins, ~ carpellate catkin, X 3; g, carpellate { flower AeA subtending bract removed, ‘tzmatic sur- face toward viewer — note tepals and two lateral bractlets, 4; : oe flower in vertical section, ee surface to the right, De ci at top, diagrammatic, X 4; i, twig with immature fruits, X 4; j, k, Be te fruits, X 1

1962 | CHANNELL & WOOD, LEITNERIACEAE 437

A single species, Leitneria floridana, known from muddy, brackish tidal shores, river swamps, Swampy prairies, and sloughs ® in scattered localities in three widely separated areas: southern Georgia (McIntosh and Dough- erty counties) and northern Florida (Clay, Franklin and Levy counties) ; southeastern Missouri (Butler, Dunklin, Pemiscot, and Ripley counties) and eastern Arkansas (Arkansas, Clay, Craighead, Jackson, Jefferson, and Lincoln counties); and southeastern Texas (Brazoria and Chambers counties).

According to the literature, the species is variable in respect to height, leaf shape, length of catkins, and fruit size, but no comprehensive study of variation has been made, and no formal taxonomic segregates have been proposed.

Individual plants spread clonally, apparently from adventitious buds on the shallow root system. Pollination is by wind, and fruits are not set in the absence of pollination. Sometimes cultivated as an ornamental oddity in appropriate wet situations, the Missouri form is hardy as far north as Boston, Massachusetts, and Rochester, New York. The soft, only slightly porous wood, the lightest of any North American plant (sp. gr. 0.21) has sometimes been used locally for floats on fishing nets.

Generally admitted to be highly specialized (reduced) in structure, Leitneria shows few features which assist in determining its relationships. It has been associated with Ranales, Rosales, Geraniales, Sapindales, Parietales, and Myricales, and in most modern works it is placed in the “Amentiferae,” usually as a separate order in a position close to Myricales, despite various lines of evidence to the contrary. Comparative morphologi- cal, anatomical, and embryological studies indicate that the closest rela- tionships may be with the primitive Rosales (including Hamamelidaceae) or Geraniales, but more diagnostic evidence from these groups is needed. The development of the male gametophyte and fertilization have not been studied, and the chromosome number is unknown.

Vascular and other evidence suggests that the flower of Leitneria is derived from a bisexual ancestral form with a perianth of two cycles, at least one cycle of stamens, and a 2-carpellate, perhaps apocarpous, gynoe- cium. The staminate inflorescence apparently represents a many-flowered compound ament with a cymule of three florets in the axil of each primary bract.

The secondary xylem is so advanced in various respects, including the simple perforation plates and the alternate pit arrangement on the side walls of the vessel elements, that there would be little difficulty in deriv- ing it from that of the primitive types of any of the groups with which

Delzie Demaree has written to Channell (April 1, 1962) of an exceptional see of Leitneria in valley farmland on the property of Mr. Ronney G. Mattics about 214 miles east of Senath, Dunklin County, Missouri. “The area has been cleared of large Taxodium trees. Waste areas and ae rows are covered by Leitneria. These plants are a troublesome weed to this farm . I have watched this area for over ten years and before any of it was cleared tee of plants were present. The soil is sandy loam and fertile.” The plants in the area are three to five feet tall and were in abundant flower

438 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLII

Leitneria has been associated. The most distinctive anatomical feature is the presence of secretory canals in the outer part of the pith which extend in association with the vascular bundles throughout the leaves. This and the stratified Seta led Solereder to note the similarity to Dipterocarpaceae, from which, however, it must be excluded on other grounds. (See Metcalfe & Chalk: Heim; Van Tieghem & Le Conte.)

REFERENCES:

Asse, E. C., & T. T. Earve. Inflorescence, floral anatomy nee of Leitneria floridana. Bull. Torrey Bot. Clu b 67: 173-193.

BaILLon, H. Castanéacées. Hist. Pl. 6: 217-259. 1875. es 239-241; carpellate flowers incorrectly drawn. |

BENTHAM, G., & J. D. Hooker. Leitnerieae. Gen. Pl. 3: 396, 397. 1880.

CANDOLLE, C. de. Myricaceae. Prodr. 16(2): 147-155. 1864. [Leitneria, soles

Dunpar, J. Leitneria floridana. Gard. Chron. II. 47: 228. 1910. [Use an ornamental. |

ENGLER, A. Leitneriaceae. Nat. Pflanzenfam. III. 1: 28, 29. 1894; Nachtrage II- IV. 117, 363. 1897.

Heim, F. Sur le genre: Leitneria Chapm. Assoc. Fr. Avanc. Sci. Cong Marseille 1891. 16 pp. 2 pls. Paris. 1892. [Allies Leztneria with ae. daceae; carpellate flowers incorrectly illustrated, as in Baillon. |

echerches sur les Diptérocarpacées. 11 + 186 pp. pls. 1-11. Paris.

Z ' Leitneria, 175, 176, pls. 10, 11; excluded from Dipterocarpaceae; illustrations as above

Hyermavist, H. Studies on the floral morphology and phylogeny of the Amen- tiferae. Bot. Not. Suppl. 2: 1-171. 1948. [Comparative morphology; Leit- meria, 71-76, retained as order Leitneriales in Amentiferae nearest Myri- caceae.

MetTcaLFe, C. R., & L. CHALK. Leitneriaceae. Anat. Dicot. 2: 1282, 1283. 1950.

OLIVER, D. lai floridana, Chapm. Hooker’s Ic. Pl. 11: 33, 34. pl. 1044. 1867-18

PASFIELD, - H. Leitneria floridana. Gard. Chron. III. 107: 185. 1940.

PFEIFFER, W.M. The morphology of Leitneria floridana. Bot. Gaz. 53: 189-203. pls. 18-20. 1912. [Includes 2 gametophyte and embryology. |

SARGENT, C.S. Leitneria. Silva N. Am. 7: 109-112, pl. 330. 1895.

SOLEREDER, H. Systematische peas der Dicotvledonen. xvi -+ 984 pp. Stuttgart. siried | Leitnerieae, 879, 880. |

TiecHeM, P. van, & H. Le Conte. Structure et affinités du Leztneria. Bull. soc. Bot. Fr. 33: 181-184. 1886. [Assigned to Dipterocarpaceae. |

TRELEASE, W. Leitneria floridana. Missouri Bot. Gard. Rep. 6: 65-90. front., pls. 30-44, 1895. [Comprehensive morphological study; wood anatomy. |

DEPARTMENT OF BIOLOGY, VANDERBILT UNIVERSITY

THE ARNOLD ARBORETUM, HARVARD UNIVERSITY

1962] THE DIRECTOR’S REPORT 439

THE DIRECTOR’S REPORT

THE ARNOLD ARBORETUM DURING THE FISCAL YEAR ENDED JUNE 30, 1962

IN RECORDING the activities of the staff of the Arnold Arboretum during the past year, the event that stands out most vividly is the completion of the Charles Stratton Dana Greenhouses and the associated | ‘open house” days for the Friends of the Arnold Arboretum and interested public. The planning of this new facility has occupied the staff for several years and the actual construction of it for much of the present one. The contractors finished their work in March, leaving April and early May for the tasks of moving the contents of the old greenhouses and laboratory to the new location and of preparing the surrounding grounds. The complete coopera- tion and the hard work of the horticultural staff of the Arnold Arboretum, best described as dedicated, made it possible to meet our schedule. For once, the fickle New England climate cooperated with our plans, produc- ing excellent weather for the moving operations and also one of the most extended flowering seasons in the recent history of the Arboretum. The frequent newspaper notices and radio announcements of the progress of the flowering season were also helpful in giving us the largest weekly attendance since before the last war. The general response to the pub- licity culminated in an historic traffic tie-up on surrounding streets during lilac weekend with an estimated 25,000 visitors on the grounds between two and three P.M. on Sunday, May 20

Staff:

New appointments to the staff meee har year were Dr. Bernice G. Schubert and Dr. Wallace R. Ernst. Schubert, formerly with the U. S. Department of Agriculture, a te Maryland, joined the staff on January Ist as Associate Curator. Dr. Er nst, a recent graduate of Stanford University, was eee jointly with the Gray Herbarium to work with Dr. Wood on the generic flora of the southeastern United States. Two scholars were appointed Mercer Fellows during the year. Mr. Don M. A. Jayaweera, Director of the Royal Botanic Gardens, Peradeniya, Ceylon, held a Rockefeller Foundation Fellowship during the last year and was appointed a Mercer Fellow to complete his work on the genus Mussaenda. Dr. Lalit ooo Srivastava, a graduate of the University of California at Davis, was appointed a Mercer Fellow to work with Dr. I. W. Bailey on ane studies of the cambium and secondary phloem of vascular plants.

440 JOURNAL OF THE ARNOLD ARBORETUM | VOL. XLIII

The resignations of Dr. Joab L. Thomas and Dr. Burdette L. Wagen- knecht were accepted at the beginning of the year. Dr. Thomas accepted a position at the University of Alabama and Dr. Wagenknecht one at Norwich University.

It is a pleasure to record the horticultural award of the Colman Medal to Dr. Sax by the American Association of Nurserymen in recognition of the work which he accomplished at the Arnold Arboretum. Dr. Ernst was awarded the George R. Cooley Prize for the best paper presented at the annual meeting of the American Association of Plant Taxonomists. This paper, entitled “The Familial Status of the Fumariaceae,”’ summarized his morphological comparisons of this family with its relatives, the Papa- veraceae. Dr. Wyman completed his term as President of the American Horticultural Society and was elected to the Board of Directors of the Society at its annual meeting.

Horticulture:

The completion of the Charles Stratton Dana Greenhouses and the transfer of the propagation work to these new quarters summarize in large measure the activities of the staff in horticulture during the past year. Previous reports have indicated the nature of the physical plant proposed, and a full description of these greenhouses was published in Arnoldia, volume 22, combined issues 5 and 6. The new greenhouses are built on land owned by Harvard University for the Arnold Arboretum and adjacent to the city-owned land occupied by the main collections. The development consists of four units of construction, a main building with three attached greenhouses, a cold storage house built into an earthen bank, a free-standing slat house of modern design to house the Larz Anderson bonsai collection, and a pipe-frame construction to be covered with saran cloth to function as a shade house for nursery stock. In the surrounding area, over an acre of ground is devoted to nursery stock; additional space is for planned future expansion, appropriate landscaping for the buildings, bank plantings of suitable materials, as well as a demon- stration area for many varieties of such plants, a collection of genetically dwarf plants near the dJonsai collection, and a new location for the Arboretum hedge collection. The entire location is fenced and can be locked, thus affording for the first time excellent protection to the green- house and nursery areas.

The Dana Greenhouses have a main building, the headhouse, 36 * I11 feet with full basement and first floor and a smaller second floor with an apartment 22 68 feet for a resident guard. Included in the main build- ing, in addition to ample areas for the work of the propagation staff, are a small conference-lecture room, a laboratory for anatomical or cytological work, two walk-in cold rooms for controlled temperature experiments in ranges of plus 40° to minus 20° F., and abundant storage space. Three greenhouses, each 17 & 51 feet, are attached, and there is space for a fourth. Expansion of each is possible on standard modules. The heating

1962] THE DIRECTOR’S REPORT 441

Two views of the new lath house for the Japanese bonsai of the Larz Anderson Collection of the Arnold Arboretum. The lath house is opposite the main building of the Charles Stratton Dana Greenhouses and overlooks the edge collection.

442 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLII

The Charles Stratton Dana Greenhouses of the Arnold Arboretum. “a build;

AxsovE: Main building and green BeLtow: Greenhouses, shade houses, and cold storage house.

1962 | THE DIRECTOR’S REPORT 443

plant is designed to handle twice the present glass area, An auxiliary generator was installed to provide automatic take-over when the line voltage drops below 70 per cent of normal. This generator will operate both the heating and refrigeration units for an extended emergency. Ease of maintenance and operation were the primary considerations in the de- sign of the building and are exemplified by the uniform floor levels throughout the building, the ample aisles for trucks, an electric elevator for transport of supplies to the basement storage, glazed tile walls for cleanliness, chutes to an incinerator or bins for debris. washable floors which can be hosed and then dried with ‘‘squeegees,” and soil bins which are filled from the outside and are unloaded from the inside.

The cold storage house is another feature of the experimental oppor- tunities made possible through the new construction. This building, 15 & 100 feet, is of concrete block construction and is insulated with slabs of “styrofoam.” The house has heating as well as refrigeration units. One section will house the bonsai collection during the winter months, and a separate section can be used to produce an early cold season or extend a winter season for plants larger in size than are usually cared for in a greenhouse. Nursery stock normally subjected in New England to un- seasonable early warm periods followed by a late freeze now can be main- tained in a dormant condition until all frost danger is over.

The erection of the new greenhouses, named for Charles Stratton Dana, was made possible through a generous bequest by his daughter, Martha Dana Mercer. This development has met a long-standing need of the Arboretum for modern greenhouses with experimental facilities. Their completion and occupancy make possible continued contributions by the staff to the study of the ornamental plants hardy in New England.

To move from the old greenhouse area as quickly and as completely as possible required long planning and hard work on the part of the staff. During the fall, plants to be moved were planted in cans or were root pruned. Accumulations of many years were sorted and discarded or packed for the move. Soil was conditioned to receive the transplants. New equipment and supplies had to be anticipated and ordered. Finally, in March when construction was completed the move took place. During May and again during commencement week the greenhouse area was on display. New lawns were planted, newly transplanted materials had to be watered, mulches were spread, much pruning was required, and, finally, the ever-present weeds of newly developed areas required attention. At the time of this report operations are about back to normal, and much of the work which remains to be done can be fitted into a regular schedule. A special word of appreciation is due Dr. Wyman, horticulturist, who worked with the architects and contractors throughout the planning and the com- pletion of this construction; Mr. Williams, superintendent, for his own efforts combined with those of the grounds crew; and Mr. Fordham, propagator, and his staff. The good job expected was done.

The weather of the past year was extremely favorable to the living collections, The season was marred only by the passage of hurricane

444 JOURNAL OF THE ARNOLD ARBORETUM PyOL, seni

“Esther” on September 21st. Moderate damage to branches resulted from gale-force winds, which also destroyed our only specimen of Juglans mandshurica in a localized gust. A replacement specimen has since been obtained from Finland, but the loss emphasizes the value of the efforts of the greenhouse staff to propagate plants now represented in our collec- tions by single individuals which have proven extremely difficult to repro- duce by the usual propagating techniques. The heavy snow coverage of February gave adequate protection to the plants during the month of most violent weather, and little or no killing of flower buds or branches was experienced. The beneficial result of natural winter protection was revealed in a most floriferous spring season of moderate temperatures which saw the major collections remain in flower for longer periods than usual.

The labor requirements associated with the new greenhouses caused a reduction in the cultural efforts in the main collections during the year. Very little planting was done during the fall or spring season, and the regular distribution of plants to cooperating nurserymen was omitted this year. Materials for both programs are on hand and both will be reactivated during the transplanting season in the fall.

The Department of Parks and Recreation of the City of Boston con- tinued its regular attention to the Arboretum road system. A major road unit from the Forest Hills gate to the pond area and a second unit from the forsythias past the lilac collection to the rockery were resurfaced, the drains relocated, and the sidewalks repaired. This is a major improve- ment which will facilitate snow plowing during the winter and make many areas more accessible for winter-time work. With the cooperation of the representatives of the Department of Parks and Recreation, the City of Boston scheduled a hearing concerning necessary repairs to a storm sewer which passes through the new greenhouse area. Since this land belongs to Harvard University and not the city of Boston, a division of the costs of repairs has been agreed upon, and it is expected that the needed repair will be completed in the next year.

At the annual meeting of the American Association of Botanical Gardens and Arboretums, the Arboretum staff was asked to serve for another two- year period as registration authority for cultivated woody plants not represented by special societies. Although no additional registration lists of cultivars were published during the year, several lists have been com- pleted, and work is in progress on others. Dr. Wyman completed the regis- tration list for Fagus and Mr. Green that for Ulmus.

Dr. Howard, with the assistance of Miss Carroll, Miss Herron, and Mrs. Walsh, completed the compilation of a directory of botanical gardens of the world which is to be published by the International Association for Plant Taxonomy as a volume of Regnum Vegetabile. A grant from the International Union of Biological oe will assist in the publication of this directory which lists the physical characteristics, staff members, and the research and resources of over 500 botanical gardens.

During the past year the staff of the propagation department received

A WNT |

The Charles Stratton Dana Greenhouses of the Arnold Arboretum. LEFT, ABOVE: Office and conference room; BELOW: Research laboratory. RIGHT, ABOVE and BELow: Apartment for greenhouse guard.

[Z961

LYOdaa S AOLOAAIG AHL

446 JOURNAL OF THE ARNOLD ARBORETUM [ VOL, XLIII

ee 8

| | 2

Two views of the work area for plant es in the Charles Stratton Dana Greenhouses of the Arnold Arboretu

164 shipments of plant materials representing 579 species and varieties from 23 different countries. One hundred and thirty of these lots (repre- senting 458 taxa) came as plants or cuttings, while only 30 shipments (121 taxa) consisted of seeds and fruits. By contrast, 214 shipments comprising

1962 | THE DIRECTOR’S REPORT 447

852 taxa were distributed on request to botanists in nine countries. Of these, 157 shipments (645 taxa) were as plants or cuttings and 57 ship- ments (207 taxa) as fruits or seeds. In addition to these, we were able to fill 52 requests for pollen, leaf samples, soil samples, fruits, or wood speci- mens obtained from the living collections in support of research requests of scientists in thirteen countries, not including the United States.

The requirements of the projected plantings around the greenhouses and the development of new groups in the Arboretum collection necessitated the propagation of 467 taxa. Outside requests for materials from the Arboretum collections not available elsewhere numbered 62 items which have been propagated for distribution at the proper stage of development. The Arboretum staff requested the propagation of 74 taxa for taxonomic, cytological, or morphological studies. Finally, 84 ee, were handled by the propagation staff to obtain or to check propagation

Experimental work in the greenhouses, although are by the

move, continued on problems concerned with winter survival of rooted cuttings, methods of rooting species which defy normal procedures, tech- niques of handling and breaking seed dormancy, methods of obtaining more complete and uniform germination of seeds, and viability studies in the storage of scions. The results of these studies appear as notes or more comprehensive papers in Arnoldia and other horticultural journals.

Case Estates:

The Case Estates in Weston continue to serve as a nursery testing, and demonstration area, a quarantine zone, and a place to retain under condi- tions of easy maintenance plant material not desired in the main collections in Jamaica Plain. Plants propagated in the Arboretum greenhouses are held in Weston until they reach flowering size. During this period their growth habits, hardiness, and flowering characteristics can be determined. The best plants from the cultural or ornamental points of view later are planted in Jamaica Plain. Less desirable plants, the distribution of which may be restricted by law (e.g., Ribes or Berberis), or plants marginally hardy may be retained in permanent non-display plantings in Weston. Of the demonstration areas, the perennial garden, the ground cover plots, and the small stature trees attract the greatest attention from visitors. In- creased interest is seen in the test plantings of Narcissus varieties contrib- uted by Dr. Helen Scorgie and other members of the New England Section of the American Daffodil Society. Plants received from abroad, subject to plant quarantine restrictions, are maintained in special sections and screened houses on the Case Estates prior to their clearance for distribution by representatives of the Department of Agriculture. In addition, certain areas of the various plantings on the Case Estates can be used experi- mentally for trial of new horticultural practices. For example, the search continues for safe, yet effective, chemical weed killers for use in display nurseries, One of the most promising materials tried during the past year was “Simazine” which, when applied in the fall, made hand hoeing un- necessary in nursery rows until mid-June.

448 JOURNAL OF THE ARNOLD ARBORETUM (VOL: REN

f wet aoe REA

: | oo oo .

The Arnold Arboretum display of dwarf conifers at the Spring Flower Show of the Massachusetts Horticultural Society, Revere, Massachusetts, April 17-25, 1962

The grounds of the Case Estates are used for teaching activities of the staff. In addition to an “open house,” field classes have been held for the general public in the spring and the fall. Special tours are arranged for interested groups which can be shown certain plants, plantings, and prac- tices not demonstrable in Jamaica Plain, and the grounds are used for field work in biology classes of Harvard University and the Weston Public Schools. In addition, staff members of the Bussey Institution, the Depart- ment of Biology, the abot Foundation, and the Gray Herbarium, of

rvard University, have been allowed to use small plots of land for eae studies. Currently, three high-school and_private-school students have “science fair” projects under way on the grounds.

Herbarium:

During the year, 16,467 specimens were mounted and added to the herbarium, bringing the total collection to 742,811 specimens on June 30, 1962. During the same period, 10,920 specimens were received as acces- sions. Of these 9292 were in exchange, 1277 through subsidy, and the emainder as gifts or for identification. In conformity with the joint policy of having the Gray Herbarium maintain all exchanges with coun-

1962 | THE DIRECTOR'S REPORT 449

tries and institutions of the New World and the Arnold Arboretum those of the Old, the above accessions represent plants of the Eastern Hemi- sphere. All collections of cultivated plants are credited to the Arnold Arboretum and, whatever their source, are added to the horticultural herbarium in Jamaica Plain. Only 438 specimens were sent out as ex- change during the year, although many collections are being prepared for exchange in the near future.

The staff filled 117 requests for loans of herbarium material, amounting to 12,056 specimens sent to 69 institutions — 48 in the United States and 21 to other countries. For their study, the staff requested 94 loans com- prising 7278 specimens from 20 American herbaria and 22 foreign institu- tions. Outgoing loans averaged 103 specimens per loan and included materials from the Arnold Arboretum and the Gray Herbarium. Incoming loans averaged 80 specimens per loan, again emphasizing the wealth of material in our herbaria. It is of interest to note that of the outgoing loans, 29 per cent, representing 38 per cent of the specimens sent, were for the use of advanced students, the remainder for professional taxonomists.

Forty-three steel herbarium cases were purchased from the Art Metal Company for installation in the Administration Building in Jamaica Plain. Two additional cases were purchased for staff use in Cambridge. After the installation of the new cases, the entire horticultural herbarium was shifted to allow room for expansion throughout the collection and to provide case space for individual staff members and for the use of the mounters. It is eratifying that there is a gradual increase in the number of specimens of cultivated plants being sent for identification and in exchange. These have been received from many individuals in the United States and offer more exact evidence of the distribution of plants under cultivation. Our colleagues in foreign countries are also cooperating in response to our re- quest for specimens from cultivation in addition to those from the wild.

Although the publications cited in the bibliography speak for the scien- tific achievements of the taxonomists, it is also desirable to record work in progress: Mr. Green, studies in the Oleaceae, particularly Notelaea in New Caledonia, Australia, and New Zealand; Dr. Howard, studies on the anatomy of the petiole of the dicotyledons and floristic studies of the West Indies, particularly in the Guttiferae and Leguminosae; Dr. Hu, studies of the Compositae of China, as well as the Commelinaceae and Juncaceae of the same region; Mr. Jayaweera, studies on Asiatic Mussaendae and the orchids of Ceylon; Dr. Kobuski, the Theaceae of Asia, particularly the genus Ternstroemia; Dr. Nevling, studies of the Thymelaeaceae; Dr. Perry, with the assistance of Mrs. Metzger, studies of the medicinal plants of Southeast Asia; Dr. Schubert, the genus Desmodium in tropical East Africa and in Panama, as well as studies toward a monograph of the American species of Dioscorea; and Drs, Wood, Brizicky, and Ernst, studies of families and genera of seed plants in the southeastern United States. During the year Mrs. Metzger visited libraries in England and Germany, while Dr. Perry consulted libraries in New York and Washing- ton to lend completeness to many of the medical references being re-

450 JOURNAL OF THE ARNOLD ARBORETUM [ VOL, XLIII

viewed by them. Mr. Jayaweera completed revisions of the rubiaceous genus Mussaenda in India and Ceylon and in the Philippine Islands.

In addition to many visitors who studied in the herbarium during short visits, we have had two scholars working for longer periods of time. Dr Shun Ching Lee, Professor of Botany, National University of Taiwan, is a Fulbright Fellow working on a revision of his book on the forests of China. Lieutenant Robert Bird, U. S. Army, undertook special studies of the vegetation of several countries of southeastern Asia, using the library and herbarium.

Members of the botanical field trip after the 10th Pacific Science Congress at the summit of Haleakala, Maui, Hawaii.

Library:

The librarians continued their regular services during the year, since the shifting of books described in previous reports has been completed. Three hundred forty-seven volumes obtained by purchase, gift, and bind- ing were added to the library, making the total number of volumes 51,453 on June 30, 1962. A total of 624 pamphlets was also catalogued and added to the collection, making a total of 18,926. The work of cross-indexing the main catalogue continues, with 2088 such cards being added, neat the new acquisitions. Four issues, totalling 3000 cards, were added to Gray Herbarium Card Index of American Plants. The Torrey index c American Botanical Literature was enriched with the addition of 2600

1962] THE DIRECTOR’S REPORT 451

cards, and issue number 15 was added to the Index Nominum Genericorum.

In response to requests, sixty-six volumes were sent on interlibrary loan. This represents about one half the number of volumes lent in previous years. The staff is filling a larger number of requests by the use of “con- tura,” “xerox,” or microfilm reproduction methods to avoid lending old volumes, It was necessary to request only eight volumes from outside sources to meet the research needs of the staff, so extensive are the libraries of the Arboretum and the Gray Herbarium and of the neighboring depart- ments of Harvard University.

The librarian, Mrs. Schwarten, along with Dr. Howard, attended the dedication of the Rachel McMasters Miller Hunt Botanical Library at the Carnegie Institute of Technology, Pittsburgh, Pennsylvania.

Comparative Morphology:

Irving W. Bailey, Professor of Plant Anatomy, Emeritus, has continued to serve as curator of the wood collection. During the year, Professor Bailey continued his research on the leaf-bearing cacti of the genera Pereskia, Pereskiopsis, and Quiabentia. Additional preserved specimens were received from tropical America and prepared for anatomical studies of xylem and phloem. As a Mercer Fellow, Dr. Lalit M. Srivastava has worked with Professor Bailey in studying the cambium and phloem of these genera. Dr. Srivastava completed his doctoral dissertation on the secondary phloem in the Abietineae, and the manuscript has been sub- mitted to the University of California Press for publication. At present, he is continuing his studies involving ontogenetic and histochemical inves- tigations of the vascular cambium and its derivatives on other genera within the collections of the Arboretum.

Since the wood collection of the Arnold Arboretum is one of the best in existence, frequent requests, which are filled as materials are available, are received for study samples. During the past year, wood samples were sent to 26 investigators in 14 countries. Where specimens have not been sectioned previously for our own slide collection duplicate slides are re- quested in return, adding to the available slides for local study. The wood samples supplied on request are acknowledged in published papers, for in many cases supporting herbarium vouchers are preserved in the Arboretum herbarium.

Education:

No formal classes were offered by members of the staff during the past year. Informal classes on horticultural topics were represented by the field classes held in the fall and the spring at Weston and Jamaica Plain. The staff members also took part in two seminar series held weekly and bi- weekly in Cambridge and open to all students. The weekly series dealt with botanical problems in Latin America, while the biweekly one com- prised a discussion of the research projects of staff members and students of the Arboretum and Gray Herbarium. A series of lectures open to the

452 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIII

public was given evenings during the fall at the administration building in Jamaica Plain. Attendance varied at these meetings depending on the topic and the weather, It seems unlikely that evening programs in Jamaica Plain will ever be completely successful due to the remote location and difficulty of public transportation.

We again experienced an increase in the number of groups visiting the Arboretum and the Case Estates and requesting guided tours. There was an unexpected increase in the number of requests by mail for information on the Arboretum and in individual requests for information on or the location of specific plants in the Arboretum. These can be explained only in terms of the recent increased publicity given the Arboretum in news- papers, national magazines, and on local radio programs, The Arboretum’s exhibit at the Massachusetts Horticultural] Society’s Spring Flower Show was mentioned and complimented in the March 30th issue of Time, which drew attention to the dwarf plants. Life of May 4th listed the Arnold Arboretum first in a column on American arboreta and referred to our introduction of Metasequoia. The Harvard Alumni Bulletin of February 17th had a cover illustration of the Arboretum, drawing attention to the conifer collection. The New Vork Times of April 22nd published an article by Dr. Walter Hodge on botanic gardens which contained an illustration of our lilac collection and comment that “Harvard University’s Arnold Arboretum, America’s best known arboretum. has been rated since Ernest Wilson’s day as ‘America’s greatest garden’.” Dr. Wyman’s timely articles in the Boston Herald feature the plants in flower at the Arnold Arboretum and the flowering calendar in the Sunday edition of the New Vork Times lists the plants in bloom at the Arnold Arboretum each week during the spring season. Such publicity is welcome for the interest it focuses on the contributions of the staff and the educational values of the living collections.

The increase in requests for Arboretum staff members as speakers for individual garden club meetings now poses a real problem. Whenever possible requests are filled, but largely at the discretion of the staff member invited as speaker. Joint meetings of garden clubs are one way of utilizing speakers more efficiently. In order to compensate in some measure for the loss of working or research time. it is necessary to charge a standard fee for speaking engagements of staff members. Such receipts are used to further the work of horticultural education.

Dr. Howard appeared on the science lecture series of the Royal Canadian Institute in Toronto and that of the American Association for the Advance- ment of Science at Franklin and Marshall College. He gave the evening address on Hawaiian botany at the 16th Congress of the American Horti- cultural Society and at the annual meeting of the Massachusetts Dietetic Association spoke on the economic uses of plants. Following the Pacific Sci- ence Congress, Dr. Howard was invited to address an open meeting of the Garden Club of Honolulu. Dr. Wyman addressed meetings of nurserymen in Iowa and Michigan. He took part in short courses in horticulture sponsored by the Oregon State University and the University of Massa- chusetts. Dr. Wyman also appeared on the lecture program of Longwood

1962 | THE DIRECTOR’S REPORT 453

Gardens. Mr. Green described the work of the Arboretum at Pine Manor Junior College and reported on methods of vegetation mapping for the New England Botanical Club. He also talked about the plant introduction and distribution program at the annual meeting of the American Associa- tion of Botanical Gardens and Arboretums. Mr. Fordham discussed dwarf

Strollers in the Arnold Arboretum on lilac weekend, May 20, 21, 1962.

454 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIII

and abnormal conifers at the nurserymen’s short course at the Waltham Field Station and spoke on methods of accelerating seed germination at a meeting of the New England Nurserymen’s Association. Mr. Heman Howard is aiding the development of Bartlett Park in Chelmsford as a local arboretum and spoke about the selection of plants to its supporters, Dr. Wyman and Mr. Williams conducted a day-long demonstration lecture on pruning methods for the New England Electric Public Utility Services which was attended by over forty line superintendents involved in main- tenance of electric lines.

Radio programs often involving telephone interviews and audience questions have included several members of the staff. Dr. Wyman and Mr. Williams have also appeared on local television stations. The majority of these programs take place in the spring, when interest in gardening practices is highest.

Exhibits and Displays:

The living collections of the Arnold Arboretum are planted on 400 acres in Jamaica Plain and Weston. Only the professional horticulturists or the most determined visitors take time to see all of the plants of a given group. In some areas the many representatives of a group may be quite loosely associated, but more often the individual plants are widely distributed and located where they will show the best growth. A flower show, by contrast, offers an opportunity for displaying small plants or branches or portions of plants in a small area, conveniently displayed and studied. Such displays may be seen by more people in a few hours than may visit the living collec- tions in a full week. As a part of its effort in horticultural education, the staff of the Arboretum will prepare educational displays of plant materials. Regrettably, we cannot fill all requests or accept all invitations which, during the past year, were received from seven states, including Hawaii, from Canada, and from Europe. A display of ornamental fruiting shrubs and trees was shown at the Fall Show of the Massachusetts Horticultural Society attended by about 5000 people and at the comparable show of the Worcester County Horticultural Society which had about 9000 visitors. The staff cooperated with members of the Massachusetts Horticultural Society in preparing a Christmas Show which drew 3000 visitors to Horti- cultural Hall in Boston. The Arboretum received a silver medal for its display of cones and evergreens. A separate exhibit area featured a display of fruiting branches of selected hollies native to New England.

During the spring season our largest exhibit was at the Massachusetts Horticultural Society’s Spring Flower Show at Revere, Massachusetts. This display, awarded a first prize and a gold medal, featured a collection of dwarf evergreen plants which will eventually be located near the Dana Greenhouses. Eighty-six thousand people attended this exhibition, At the request of the New York Horticultural Society the Larz Anderson collection of bonsai was taken to New York for the New York International Flower Show. A final exhibit, again by request, was a demonstration of methods

1962] THE DIRECTOR’S REPORT 455

of pruning trees and shrubs at the Jordan Marsh Spring Show sponsored by the Garden Club Federation of Massachusetts and seen by an estimated 50,000 people. The various exhibits were designed and executed by Dr. Wyman, Mr. Williams, and Mr. H. Howard. Considerable time was spent this year in the preparation of permanent labels engraved in colored plastic for these displays, in an attempt to have a type of exhibit more easily assembled in the future.

: f as ‘ie a pa i all e . a

= as me Pe \\

Japanese bonsai of the Larz Anderson Collection of the Arnold Arboretum on exhibition at the International Flower Show, New York City, March 10-18, 1962.

ge

Travel and Exploration:

Pacific Science Congress, the National Shade Tree Conference, the Ameri- can Nurseryman’s Association, and the American Society of Horticultural Sciences.

Dr. Howard collected some special plant materials for his research while in Hawaii for the Pacific Science Congress. The expedition to Burma of Mr. James Keenan, of the Royal Botanic Garden, Edinburgh, Scotland, sponsored in part by the Arnold Arboretum, was completed during the year,

456 JOURNAL OF THE ARNOLD ARBORETUM [ VOL, XLII

and the arrangement of materials collected is under way. Mrs. Claude Weber made a special trip to gardens and nurseries in Ohio and Illinois to study cultivars of Chaenomeles in flower. The trip was rewarding in the discovery of several old cultivars previously not known to be still in culti- vation. Mrs. Weber also participated in a class in tropical botany which spent a month in Costa Rica. This trip, financed by a grant from the Fernald Fund, allowed her to make general and special collections of plant materials for the herbarium and special collections for several staff mem- bers with research problems involving plants in that area. Dr. Wyman travelled extensively in the United States during the year, visiting many gardens and arboreta during peak seasons to continue his studies of orna- mental plants and to obtain new plants for trial at the Arnold Arboretum. Dr. Ernst left in late June for field work in Texas with Dr. Henry J. Thompson, of the University of California at Los Angeles, related to their joint research interests.

Gifts and Grants:

The Friends of the Arnold Arboretum who contribute regularly to the work of the Arboretum responded generously to an appeal during the spring. To these contributors we again express our continuing appreciation.

Through the initiative of Mr. Seth Kelsey, of East Boxford. a member of the Committee to Visit the Arnold Arboretum, eight Massachusetts nurseries contributed over 1300 plants of various types and sizes to be used in establishing the basic landscape plantings around the Dana Greenhouses. Through their generosity, the greenhouses lost immediately a portion of the bleak appearance so characteristic of new buildings and freshly moved soils.

One group of plantings proposed for the fenced-in area of the Dana Greenhouses was a collection of native hollies with most attractive fruits. A selection was offered the Arboretum in the fall of 1961 by Mr. Wilfrid Wheeler, of Hatchville, Massachusetts. Although Mr. Wheeler died on Christmas Day, 1961, his wishes were carried out by his sons, Wilfrid, Jr., Richard, and Charles. We regret that Wilfrid Wheeler could not see his collection of outstanding cultivars in their new location, and we are grateful to his family for this gift of plants which so well represent his long interest in and his contribution to horticulture in New England.

Portions of the research of six staff members involving the services of six assistants continue to be supported by grants from the National Science Foundation, the National Institutes of Health, and the gifts of Mr. George R. Cooley.

Publications:

Four quarterly issues of the Journal of the Arnold Arboretum including most of the scientific publications of the staff were distributed during the year, as were the twelve numbers of Arnoldia which appear at irregular

1962 | THE DIRECTOR’S REPORT 457

intervals. These comprise the regular publications of the Arnold Arbo- retum. An issue of Arnoldia titled, “The Walter Street ‘Berrying’ Ground” was prepared by Mrs. Mary Lehmer, formerly of the Arboretum staff. This reviewed the history of the Walter Street Church and its adjacent cemetery now included within the boundaries of the Arboretum. Although remains are visible of the church, a memorial plaque, along with thirteen old head- stones and a crypt, mark the Revolutionary War dead and are decorated each Memorial Day by the historic commission of Boston. This article, which drew attention to a remote section of the Arboretum, is one of a projected series of articles on the history of the land we occupy. Another number of Arnoldia issued at the dedication of the Dana Greenhouses has been requested frequently by other botanical gardens and by architectural students.

Bibliography of the Published Writings of the Staff and Students July 1, 1961 — June 30, 1962

BarLey, Irvinc W. Comparative anatomy of the leaf-bearing Cactaceae, III. Form and Deve onelon of crystals in Pereskia, Pereskiopsis and Oi onic. ae Arnold Arb. 42: 334-346. 1961.

(with Srivastava, Lattr M.). Comparative anatomy of the leaf-bearing Cactaceae, IV. The fusiform initials of the cambium and the form and structure of their derivatives. Jour. Arnold Arb. 43: 1874202. 1962

BrizicKy, GEorGE K. The genera of Rutaceae in the southeastern United States. Jour. Arnold Arb. 43: 1-22. 1962.

The genera of Simaroubaceae and Burseraceae in the southeastern

Gaited States. Jour. Arnold Arb. 43: 173-186. 1962.

. The genera of Violaceae in the southeastern United States. Jour.

Arnold Arb. 42: 321-333. 1961.

. A synopsis of the genus Columellia (Columelliaceae). Jour. Arnold Arb. 42: 363-372. 1961. Taxonomic and nomenclatural notes on Zanthoxylum and Glycosmis (Ruta ceae). Jour. Arnold Arb. 43: 80-93. 1962.

ERNST, WALLACE R. Blue grama at Goleta, California. Leafl. West. Bot. 9: 180.

1961.

(with Tryon, Rotta M.). Comments on the International Organization of Biosystematists. Taxon 11: 139. 1962. FORDHAM, ALFRED J. Germination of double-dormant seeds. Combined Proc. Plant eee Soc. 1960: 206-210. 1961. GREEN, Peter S. Herbaceous aliens in the Arboretum. Arnoldia 22: 49-56. 1962 . Studies in the genus Jasminum, II. The species from New Caledonia and the Loyalty Islands. Jour. Arnold Arb. 43: 109-131. 1962. . Watercress in the New World. Rhodora 64: 32-43. 1962. ——- (with Lavener, L. A.). Catalogue of the names published by Hector Leveillé. I. Notes Bot. Gard. Edinburgh 23: 573-596. 1961. (with Tuomas, JoaB L.). The bulbiferous Ranunculus ficaria. Rhodora 63: 289-291. 1961. Howarp, RicHarp A. Botanical and other observations on Redonda, the West Indies. Jour. Arnold Arb. 43: 51-66. 1962.

458 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLII

. The Charles Stratton Dana Greenhouses of the Arnold Arboretum. Arnoldia 22: 33-47, 1962. . The correct name for “Diospyros ebenaster.” Jour. Arnold Arb. 42: 430-435. 1961.

’ e Director’s report. The Arnold Arboretum during the fiscal year ended June 30, 1961. Jour. Arnold Arb. 42: 447-467. 1961. Naming and registering cultivated plants. Boxwood Bull. 1: 41-45.

1962. (with JAYAWEERA, Don M. A.). The genus Duabanga in cultivation.

Baileve 10: 8-13. 1962.

(with oo. Don M. A.). A shrub called “Malitta”’ (Woodfordia

fruticosa (L.) Kurz). Baileya 10: 14-18. 1962

(With Nowburainig, TycHo). The i opincation of Diospyros ebenum and

Diospyros ebenaster. Jour. Arnold Arb. 43: 94-107. 1962.

JAYAWEERA, Don M. A. Some observations of abnormal flowers in the cultivated species of orchids. Orchid Rev. 69: 206-216. 1961.

(with Howarp, RicHarp A.). The genus Duabanga in cultivation.

Baileya 10: 8-13. 1962.

with Howarp, Ricuarp A.). A shrub called “Malitta” (W oodfordia ae a (L,) Kurz). Baileya 10: 14-18.

Kosusktr, CLARENCE E. Studies in the eas: XXXII. A review of the genus Ternstroemia in the Philippines. jour, Arnold Arb. 42: 263-275. 1961.

. Studies in the Theaceae, XXXIII. Variation in the fruit of Tern- stroemia kwangtungensis. Jour. Arnold Arb. 42: 426-429. 1961.

NEVLING, Lortn I., JR. On the status of Psilaea (Thymelaeaceae). Jour. Arnold Arb. 43: 220, 221. 1962.

A revision of the Asiatic genus Enkleia (Thymelaeaceae). Jour. Arnold Arb. 42: 373-396. 1961.

. A revision of the Asiatic genus Linostoma (Thymelaeaceae). Jour. Arnold Arb. 42: 295-320. 1961,

Perry, Liry M. Problems in the compilation of a native medicinal flora of southeastern Asia. Econ. Bot. 15: 241-244, 1961.

SAX, Kary. Effects of X-rays on aging of seeds. Nature 194: 459, 460. 1962.

Aspects of aging in plants. Ann. Rev. Pl. Physiol. 13: 489-506, 1962.

(with Sax, Hatty J.). The effect of age of seed on the frequency =

spontaneous and gamma ray induced chromosome aberrations. Rad.

1: 80-83. 1961. SCHWARTEN, LazELLA (with RoGERSON, CLARK T., Rickett, H. W., & BECKER Index to American botanical literature. Bull. Torrey Club 88: 126- 142, 187-213, 269-297, 370-385, 429-445. 1961; 89: 64-76,

SRIVASTAVA, Lari M. (with Batrey, Irvine W.). Compatative anatomy of the leaf- bearing Cactaceae, IV. he fusiform initials of the cambium and the form and structure of their derivatives. Jour. Arnold Arb. 43: 187-202.

1962.

d

Woop, Carroiy E., Jr. Magnolia. Encyclopaedia Britannica, 1961 ed. 14: 671, 672 G1,

(with BAKHUIZEN VAN DEN BRINK, R. C., BucHHEIM, E., Cowa AN,

R. S., and Danpy, J. E.). Appendix II. Nomina familiarum SouseHvantla, Int. Code Bot. Nomenel. 1959. 187-201. 1961.

1962 | THE DIRECTOR’S REPORT 459 Wyman, Donatp. Barberries. Gard. Jour. N. Y. Bot. Gard. 11: 123-125; 138 1.

Barberries. Arnoldia 22: 9-16. 1962. The best of the dogwoods. Am. Hort. Mag. 41: 1-10. 1962. Best trees for 1962. Flower Grower 49(1): 46-48; 66. 1962. The birches. Arnoldia 22: 17-23. 1962 . Dozen varieties include best in shrub Altheas. Am. Nurseryman 114(1): 19: 121-127. . Few mountain ash species popular despite borers. Am. Nurseryman 114(5): 12, 13; 106-112. 1961. ———. The Forsythia story. Plants & Gardens 17(4): 17-19. 1962. . Green is what you make it. Better Homes & Gardens 40(4): 122B. 62.

. Hawthorns. Arnoldia 22: 25-32. 1962. _ Horse chestnuts and buckeyes now supplanted. Am. Nurseryman 115(11): 11, 44-49. 1962. . The majestic beeches. Arnoldia 22: 1-7. 1962. Massive oaks are best long-lived ee trees for public areas. Am. Nurseryman 114(7): 12, 13, 104-110. 196 es offer wide ae of choice es ee ornamental planting. Am. ape 114(11): 13, 87-97. The Saturday ene gardener. A guide to easy maintenance. New York, Macmillan. 236 pages. ilust. 1962. _ Shrubs and trees that flower in summer. Flower Grower 48(8): 22-24.

— Ko)

a eee popular despite limitations. Am. Nurseryman 114(9): 12, 13, 106— 111. 1961.

: Weeping willows leaders in group often misnamed. Am. Nurseryman

114(3): 18, 19, 100-105. 1961.

: the Arnold Arboretum is. Green Thumb 18(9): 309-313.

1961. 19(1): 14-16. 1962.

RicHArD A. Howarp, Director

460 JOURNAL OF THE ARNOLD ARBORETUM [VOL. XLIII

Staff of the Arnold Arboretum 1961-1962

RicHarpD ALDEN Howarp, Ph.D., Arnold Professor of Botany, Professor of Dendrology, and Director.

IrvinG WIDMER BaILey, S.D., Professor of Plant Anatomy, Emeritus.

KARL SAX, S.D., Professor of Botany, Emeritus.

GEORGE KONSTANTINE Brizicky, R. N. Dr., Botanist, Southeastern Flora Project.

MicHaAeL, ANTHONY Canoso, M.S., Curatorial Assistant.*

Henry Draper, Superintendent, Case Estates.

WALLACE Roy Erwst, Ph.D., Botanist, Southeastern Flora Project.*

ALFRED JAMES ForpHAM, Propagator.

PETER SHAW GREEN, B.S., Horticultural Taxonomist.

KATHERINE ANN HERRON, A.B., Business Secretary.

HEMAN ARTHUR Howarp, Assistant Horticulturist.

SHIU-YING Hu, Ph.D., Botanist.

Don MartTINn ARTHUR JAYAWEERA, M.A., Mercer Fellow.

CLARENCE EMMEREN KosuSkI, Ph.D., Curator.*

MArGARET CATHERINE LeFAvouR, Herbarium Secretary.

SUSAN DELANO MCKELVEY, A.B., Research Associate.

Lorin Ives NEVLING, JR., Ph.D., Associate Curator.

Lity May Perry, Ph.D., Botanist.

BERNICE Gipuz SCHUBERT, Ph.D., Associate Curator.

LAZELLA SCHWARTEN, Librarian.*

Lait MOHAN SrIvVASTAVA, Ph.D., Mercer Fellow.

Ropert Grerow WILtiaAMs, B.S., Superintendent.

CarroLt Emory Woop, Jr., Ph.D., Associate Curator and Editor.

Donatp Wyman, Ph.D., Horticulturist.

* Appointed jointly with the Gray Herbarium

1962]

INDEX

461 INDEX Acarospora ee 63 BrizicKy, GEORGE K. The ie ) Adlumia, 338-339 nacardiaceae in the ieee gave ce eae 64 United States, 359 — sisalana, 64 Brizicky, GEorGE K. Genera of Ageratum houstonianum, 66 Rutaceae in the Southeastern United Ailanthus, 179-180 States, 1 Alvaradoa, 182-183 BRIZICKY, SORGE K. e Genera of Amaranthus dubius, 64 Sim aroubaceae and Burseraceae in the Amyris, 11-12 Southeastern United States, 173 Anacardiaceae in the Southeastern United Brizicky, GEorRGE K axonomic and States, The Genera of, Nomenclatural Notes on Zaathoxylum Anacardiaceae, 359-360 and Glycosmis (Rutaceae), 80 — tribe An pana reg 362 Buellia prospersa, 63 — tribe Rhoé Bursera, 184, 185-186 Anacardium, Be urseraceae in the Southeastern United Anatomy of the Leat- bearing Cactaceae, es e Genera of Simaroubaceae

The Fusiform Initials

of Cambium and the Form and Structure of Their Derivatives, 187; he Secondary Phloem, 234; VI.

The Xvlem of Pereskia sacharosa and Pereskia aculeata, 376

Annona squamosa, 65

Antilles, Some Guttiferae of the Lesser, 389

Antilles, Volcanism and Vegetation in the Lesser, 279

Argemone,

Ash, prickly

Asiatic ee

oes 330

Two New, 348

Battey, I. W. Comparative Anatomy of the Leaf-bearing Cactaceae, VI. he Xvlem of Pereskia sacharosa Pereskia Sey ‘

BaAILey, I. W., and Lair M. SRIVAST. AVA,

and

and Structure of Their Derivatives, 187; V. The ear Phloem

Balsam apple,

Bay cedar tee

Bibliography, Ernest Jesse Palmer,

ee eph Horace Faull,

354 230

Boerhaavia coccinea, Botanical and ae Observ see on Re- donda, the West Indies

Baneainy illea at a

nd, 173 Burseraceae, 183-186

Cactaceae, Comparative Anatomy

Pereskia aculeata, 376 Calamus, 33

ee ee 327 a,

ntillanum, 398 — brasiliense “antillanurm, 398 —calaba, 3 3

— jacquinii, 3 Cambium re fie Form and f ir Derivatives, The [ Comparative Leaf-bearing Cactaceae, IV,

66

Structure e203

*, Grandiflorae, — ser. Spinosae,

— arborescens, 2 —aurantiaca, 204 —chamlagu, 203 — frutescens ecu 208 — frutex, 203, 204

— frncees

204

462

Caragana age ee 209 204

, 40

Cashew Family, 35 Casuarina oe 64 Catalpa silvestril, 217 Catharanthus roseus, 66

326 Centrosema virginiana, 65 Centrostachya indica, 64 Cephalocereus ease 65 CHANNELL, R d C. E. Woop, Jr. The 2 Pn in the Southension 35

A New Genus of ophulariaceae from, 215 Chioris — 63

Choisya Chry balan icaco, 390 Citrus, 17-22

jorantiitellt: 65

oxylon ee 411, 420 —remyi, 411,

Cleome viscosa, 65

Clusia, 389

—sect. Anandrogyne, 396

—venosa, 390, 396-397

Cocos, 36

Comparative Anatomy of the Leaf- bearing Cactaceae, IV. The Fusiform

Croton flavens, 65

JOURNAL OF THE ARNOLD ARBORETUM

[ VOL. XLII Croton lobatus, 65 Cynanchum parviflorum, 66 Cyperus rung 63 — sphac 63 Pe ey gard of the Genus Vibur- um, A, 132 Daphne ear! 344 Daphne Family, 4 Daphnopsis pee (Thymelaeaceae), Note on, 344 Daphionss crassifolia eggersii, 344 Dates ublication of the Journal Linnaea, Dicentra, 336 avanne sanguinalis, 63 Diospyros ebenum and Diospyros eben-

eo The Typification of, 94 Diospyros digyna, e

ter, 100 —ebenum, 1 — glaberrima, 100

Director’s Report, The. The Arnold Ar- boretum During i Fiscal Year Ended

1962,

9

Ducal en. 412 Earth-smoke, 342

Ebony, 94

Echinocitrus, 14

eS OLF, DONALD A ms Study

the Genus Viburnum,

Elaphrium, 186

Emilia coccinea, 66

Enkleia, 220

Eragrostis ciliaris, 63

Ernest Jesse ae 1875-1962, 351

ERNST, WALLACE R. The Genera of

Fumariaceae in the

315

ae heterophylla, 65

— hirta

Fagara, 7, 8, 80

FauLtt, ANNA F. Joseph Horace Faull, 23

1870-1961, 2 Ficus citrifolia, 64 — ROBERT Dates of Publication the Journal ae Frevcneti hemsleyi ostermansii, 34 a-ahianceolata. 348

1962]

sie 334, 342-343 ariaceae in the Southeastern United a The Genera of Papaveraceae and, 31

cee eee e, —343 — tribe ees 336

Fumitory Family,

Fusiform ate of the Cambium and the d Structure of Their Deriva- tives ao Comparative Anatomy of the Leaf-bearing Cactaceae, IV, 187

Galactia, 65

— stricta, 6 Galba, 397 Garcinia macrophylla,

399 Genera of ear dince se in oe South- eastern United States, The, 3 Genera of Papaveraceae Ate ariaceae in the eae ae States, The, 315

Genera of Rutaceae in the Southeastern United States, The,

Genera or Simaroubaceae and Bursera-

he Southeastern United States,

ceae in: The,

Glaucium, 321

— flavu 6

Glycosmis (Rutaceae), Taxonomic

d Nomenclatural Notes on Zanthoxylum

and, Glycosmis, 12-13 b 90

XITl.

Studies in the Genus Ses minum Species from Caledonia and the Loyalty ae 109

Grenada, 2 se Guadeloup

Guttiferae th ae Lesser Antilles, Some,

Haplophyllum, 3

H Heterospat Hop-tree, ‘

INDEX

463

Howarp, RicH A. Botanical and Other Observations on Redonda, the West Indies, 51

Howarp, Ree A. The Director’s Re- port,

Howarp, Ricuarp A. Some Guttiferae os the Lesser ee 389

WARD, RICH A. Volcanism and Vestn in aie mec Antilles, 279

Howarp, RICHARD A., TycHo Nor- LINDH. The aden of Diospyros ebenum and Diospyros ebenaster, 94

Hyphaene, 33

eps ee crassifolia, 344

gger Hyptis aes 66 Tresine angustifolia, 64

Jasminum, Studies in ue Genus, II. The w ledonia and the

Jasminum sect. Sera: 112 Unifoliolata, 115

— brevistylis, 127

15-12 —linearifolium, 126-127

— neocaledonicum, 122-123

< isa oc E: 5 c B Yu

Jatropha gossypifolia, 65 Joseph Horace Faull, 1870-1961, 223 Justicia periplocifolia, 66

USKI, CLARENCE E. Ernest Jesse

Dae 1875-1962, 351

Lantana camara, 66

464

Lantana involucrata, 66 Leaf Base in Palms, The. ie ee and Mechanical Bidlosy od, 432

—438

Leitneriaceae in the Southeastern United States, The, 4

Leonotis nepetifolia, 66

Limeberry, 14

Limonia arborea, 90

—sect. Psilaea, 221

eit. iotands, The Species from New Caledonia and. Studies in the Genus Jasminum, II, 109

Macleaya, 322 —cordata, 317

mmea americana, 390 — humilis, 399 —-— macrophylla, 399 — — plumieri, 399

—_ hii, 399

Melicope gr randifolia, 414, 422 — spathulata, Melocactus es 65

ote on the Rela- tionships of Pinus merkusii, 10 Mirov, NICHOLAS T. Ph enology of Tropi-

Genus Platydesma

OORE, Ravmonp J. On the Origin of errren sinica,

Morphology and Mechanical aoe Its. The Leaf Base in Palms,

Nageia minor, 76

Nevis, 304

NEVLING, Lorin I., Jr. Note on eral sis crassifolia Sag ae earae

NEVLING, On eStats of Petlaen (hy melacaccse), ;

JOURNAL OF THE ARNOLD ARBORETUM

[ VOL. XLIII

NEVLING, Lorin I., Jr. The Thymelae ceae in the Southeastern United sere 428

New Caledonia and the Loyalty eee The ee from. Studies in the Genu min II, 1

Miechann: ees

NorLINDH, TycHo, oad RICHARD A. How R

e on Daphnopsis crassifolia (Thyme- laeaceae), 344 Note on the Relationships of Pinus mer-

ii, Nuées eanee 282

Oldenlandia corymbosa, 66 Oleaceae, 109 Opuntia ee 65

pe a 63 Orange, oe Origin of Abie sinica, On the, 203 Pacit, J. Shiuyinghua, of Scrophulariaceae from China, 215 PALMER, ELIzABETH, and LAZELLA een TEN. Bibliography, Ernest Jesse Palm- 354 Palmer, Ernest Jesse, 1875-1962, 351 Palms, The L eaf Base in. Its Morphology chan Biology, 23 are fn o New foes 348 Pandanus an 348

A New Genus

nd riacea cae “United : ae The Genera of, oe 315-: — subfan a ioe 316, 322

-- eta Bechacholeisiden Sly 320 —subf Papaveroideae, 317, 328 — sub onoideae, 317

. Platyste Pappophorum pappiferum, 63

Paspalu m, 63 Paulownia iverri 217 Pelea, 413

= dn icalaetott, 420 Peperomia simplex, 64 Pepper-tree, 365 Pereskia eee The Xylem of aoe sacharosa and. Comparative Anat of the Leaf-bearing Cactaceae, VI, "376 Pereskia, 189

1962]

Pereskia aculeata, 376, 378-380

— sacharosa, 234, 376-378

Pereskiopsis, 189, 381

— chapistle, 234

Phenology of Tropical Pines, 218

Phloem, The Secondary. Comparative Anatomy a the Leaf-bearing Cacta- ceae, V, 23

Phoenix, 35

Phyllanthus amarus, 65

Picramnia, 181-182

Pilea microphylla, 64

Pilo kea

Pines, Spee of Tropical, 2

eae merkusii, Note on the Ber tlie:

, 108

Pinus [group] Insignes, 108 — [group] paar 108

— elliottii

— insignis, a

— kasya, 21

_ ener 108, 219 — oocarpa

8 Pityrogramma chrysophylla, 63 Platydesma (Rutaceae), A Monograph of the Genus, 410 Platydesma, 411-413

— sect. Cornutia, 413, 422-427

—campanulata, 414 —-— macrophylla, 414 — — pallida, 4

419-420, 422

— — spathulatum, 416-417 Platystemon, 317 Plumbago scandens, 66 Podocarpus, A Taxo

South Pacific, 67 Podocarpus sect. Afrocarpus, 68 — sect. Se ere: 9 — sect. Nage — sect. Stach carpus 68

INDEX

onomic Revision of, XIII, oecbon Polypodiopsis in the

Podocarpus araucoensis, 67

4

Polypodiopsis in the South Pacific, Section. A Taxonomic Revision of Podocarpus, XIII, 67

Povanie 14-17

Poppy, 330

Poppy, California, 327

Poppy, prickly, 3

Poppy, rock, 326

Portulaca halimoides, 65

Prickly poppy , 328 i 2 Psilaea (Thymelaeaceae), On the Status 220

Psi lea es 221 Psilot Ptelea, on Pterocaulon virgatum, 66 Puccoon, 322

Quassia Family, 173 Quiabentia, 189, 381 —chacoensis, 234

Ramalina subasperata, 63 Red sandalwood, 94 est Indies, aes and

— subg. Sera 371 — subg. Rhus,

— subg. Toxtudendon 372 Richard: Ricinus communis, 65 Robinia sinica, 07 Roccella babingtonii, 63 Rock poppy, 326 R

Southeastern United States, Genera of The,

Rutaceae, 1—

—subfam. Aurantioideae, 12

466

Rutaceae subfam. Rutoideae, 5, 412 ae, 9

— tribe Xanthoxyleae, 412

Sagra, Ramon de la, 84

Sancalwood: ey 94

Sanguinaria, 322-323

Schinus, 365-366

SCHWARTEN, LAZzELLA. Bibliography, Jo-

3 d ExizasetH M.

Bibliography, Ernest Jesse Palmer, 354

Scrophulariaceae from China, Shiuying- hua, A New Genus of, 215

Section Polypodiopsis in the South Pacif- ic. axonomic Revision of Podocar- pus, XIII, 67

Setaria setosa, 63

Shiuyinghua, A New Genus of Scrophu- lariaceae from China, 215

Shiuyinghua silvestrii, 217

Shrubby trefoil, 9

Sida cordifolia, Ne

Sieve elements,

Simarouba, i He a ay

Simaroubaceae and Burseraceae in the Southeastern United Sales The Genera

of, 173 Simaroubaceae, b

—subfam. Surianoideae, 176

rie Picrasmeae, 179

be Simaroubeae, 177

ee tree, 367

se Guttiferae of the Lesser Antilles,

ee 283

outh Pacific, Section Polypodiopsis in the. A Sag mic Revision of Podo- carpus, XI

Southeastern she eee Pgs Genera of Anacardiaceae in the,

reer ee United Ae aie Genera of Papaveraceae and Fumariaceae in

the, Southeastern wna States, The Genera of Rutaceae in t Southeastern Deiter Brates, The Gen imaroubaceae and eae. in the, 173

JOURNAL OF THE ARNOLD ARBORETUM

[ VOL, XLIII

Southeastern et States, The Leitneri- aceae in

Southeastern United States, The Thyme- laeaceae in the, 4

SRIVASTAVA, Latit i. and I. W. BaILey, Cc Leaf-

and ledsciaads of Their 187;

Derivatives, . The Secondary Phloem, 234 Stark viarohes jamaicensis, 66

Staff of the Arnold ees 1961-1962,

460

Status of Psilaea (Thymelaeaceae), On the, 220

Stonr, BenyAMIN C. A. Monograph of the enue eager nag 410

STONE, BENJAMIN C o New Asiatic Pandanace as

Studies in i Genus Jasminum, IJ. The Species from New Caledonia and the Loyalty Islands, 109

Stylophorum, 323-325

Suriana,

Swallowwort, 326 Talinum triangulare, 65 182

oa i, Taxonomic and Nomenclatural Notes on

lication, oa The ae oe of Gly-

cosmis, 86; Summary,

Taxonomic sacs of . docarpus, A, XIII. Section Polypodiopsis in the South Pacific,

Tephrosia cinerea, 65

Thymelaea, 429

Thymelaeaceae in the Southeastern United

ates, The, 4

Tillandste recurvata, 64 B. The Leaf Base in Palms, Its Morphology and Mechani- cal sae : Torchwood, Torchwood amily 183

Trefoil, shrubby, 9

1962]

Trianthema portulacastrum, 65 Trichachne,

—insularis, 63 Tricholaena repens, 63

Two New Asiatic Pandanaceae, 348 Typification of Di Diospyros ebenaster, The,

Veitchia,

Viburnum, 8 Cytological Study of the Gen 13

enus, Viburnum a Lentago, 147 —sect. Megalotinus, 14 —sect. Odontotinus, 149 —sect. Opulus, 154 —sect. Pseudotinus, 146 — sect. oan 140 —sect. Tin 8 Se yaaakea new 149 —alnifolium, 146 —atrocyaneum, 148 — awabuki, —betulifolium, 14 —— ‘Aurantiacum,’ 149 — bitchiuense, 142 — x bo cL 140 —-— ‘Dawn,’ 140 — — ‘Deben,’ 140 —bracteatum, 150 — buddleifolium, 142

—x Pea 144

— burejaeticum, 14

— — ‘Park Farm meer 142 calvum, 14 —x pee ie 142, 143 — carlesii, —x are 143 —— macrocephalum, 142, 143

pubescens, 150 ee 150

Diospyros ebenum and 94

INDEX

Viburnum eee hispidum, 150, — — pilosum, Le 151 —edule, 154 —ellipticum, 151 erosum, 1

-- secre lana 151 — fragrans, 140

—— album, 140

—— candidissimum, 140 —— gran ndiflorum, 140

— icine 1 2

—xj , 143 — kansuensis, 154 —lantana, 1

44 — — ‘Aurea Marginata, 144 — — ‘Floribundum,’ 144 —— ‘Lanceolatum,’ 144 — — ‘Lees,’ 144 —-— ‘Macrophyllum,’ 144 4

— lobophyl ; —macrocephalum sterile, 144 —microphyllum

—molle, 15

— — leiophyllum, 152 —mongolicum, 144 —nudum, 14

467

151

468 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XLIII

Viburnum opulus Surely 154 Viburnum tinus hirtulum, 149 ‘Com mpactum,’ —— lucidum, 149

— ; —— — ‘Variegatum,’ 149

— — ‘Notcutt,’ 154 — — ‘Purpureum,’

——roseum, l —— variegatum, 1

—— variegatum, 155 —tomentosum, 146, 147

xanthocarpum, 155 — — mariesii,

— orientale, 1 — — sterile, 155

—ovatifolium, 152, 153 —trilobum, 155

—parvifolium, 153 ‘Andrews,’ 156

—phlebotrichum, 153 —— ‘Compactum,’ 156

— photinioides, 141 ssee s,’ 156

—plicatum, 14 — entworth,’ 156

—— glabrum, 146 — urceolatum, 14

— —‘Lanarth,’ 146 — utile,

—— lanceolatum, 146 —veitchil, 145

——ma ‘ —wilsonii, 15

—— ‘Roseum,’ 146 ae 153, 154

—— ‘Rowallane,’ 146 os andulosum, 154 — ‘St. Keverne,’ 146 ae ssei, 154

Volcanism and Vegetation in the Lesser

— ‘Lanceolatum,’ 148 the Lesser Antilles, 280; Types of Vol- — prunifolium, 147 canic Activity, 282; Volcanic Erup- —rafinesquianum, 153 tions and the Vegetation, 283; Fuma- — —affine, 153 roles and Their Effects, 289; Effects of — recognitum ‘umaroles on Vegetation, 292; Cata- — rhytidocarpum, 144 logue of the Fumarole Areas of the ached helieides, 144, 145 Lesser Antilles, 296; Summation and —rhytidophyllum, 145 Comparisons, 305 — — aureovariegatum, 145 —— lantana, 144, 145 Wedelia calycina, 66

— roset 145 West Indies, Botanical and Other Obser- — rigidum, 148 vations on Redonda, the, 51 —roseum, 146 Wightia, 217 —rufidulum, 147 Woop, C. E., Jr., and R. B. —sandankwa, 142 The Le jmeraceas in the a

sargentii, 154, 155 United States, 435

calvescens, 155 — — flavum, 155 Xanthoxylum, 7 — — ‘Puberulum’, 155 — americanum —scabrellum, 153 Xylem of Pevsken sacharosa and Peres- —schensianum, 145 kia aculeata, The. Comparative An-

atomy of fhe Leaf-bearing Cactaceae, —setigerum, 15 ava ei)

aurantiacum, 153 = sieboldit,

Zanthoxylum and Glycosmis (Rutaceae), Taxonomic and Nomenclatural Notes

—stellulatum, 145 on, 80 —suspensum, 142 eee ae = A —sympodiale, 146 — subg,

— theiferum, a

— subg. ee 7 —coriaceum, 8 === 'Prendh ae 149 Zombia, 39