CVJ ^- CD LIBRARY OF THE UNIVERSITY OF ILLINOIS AT URBANA-CHAMPAIGN BIOLOGY TROPICAL AMERICAN MYRTACEAE, II NOTES ON GENERIC CONCEPTS AND DESCRIPTIONS OF PREVIOUSLY UNRECOGNIZED SPECIES ROGERS McVAUGH FIELDIANA: BOTANY VOLUME 29, NUMBER 8 Published by CHICAGO NATURAL HISTORY MUSEUM MAY 31, 1963 BIOLOGY LIBRARY 101 BURRILL HALL TROPICAL AMERICAN MYRTACEAE, II NOTES ON GENERIC CONCEPTS AND DESCRIPTIONS OF PREVIOUSLY UNRECOGNIZED SPECIES ROGERS McVAUGH Curator of Vascular Plants, University Herbarium, The University of Michigan FIELDIANA: BOTANY VOLUME 29, NUMBER 8 Published by CHICAGO NATURAL HISTORY MUSEUM MAY 31, 1963 3 CONTENTS PAGE Introduction 395 1. Calyptranthes Sw 397 The Genus in Continental North America 397 Miscellaneous Notes 412 2. Eugenia L 413 Synopsis of Mexican Species West of the Isthmus of Tehuantepec . . 413 Notes on Miscellaneous Species 464 3. Marlierea Camb 470 4. Myrcia DC 470 Miscellaneous Notes 470 5. Myrcianthes Berg 473 Synopsis of the Genus (Eastern South America Excluded) 473 Nomenclature 476 6. Certain Eugenioid Genera 498 Revision of Myrciaria Berg, in North America and Northern South America 499 A New Species of Plinia L., from Costa Rica 505 The Genus Siphoneugena Berg 507 7. Pimenta Lindl 511 8. Psidiumi, 512 Delimitation of the Genus 512 Additions to the Genus 519 Species Described by Standley from British Honduras 522 Miscellaneous Notes 524 Index . . 530 393 Tropical American Myrtaceae, II1 The following notes have been prepared as a necessary prelimi- nary to a formal treatment of the Myrtaceae of Guatemala. There has been no comprehensive revision of North American Myrtaceae since the work of Berg (1855-62), and it was found impossible to revise the Guatemalan species without full consideration of those of Mexico, the rest of Central America, and the West Indies. The myrtaceous flora of Central America is a comparatively small one. The number of genera represented in the flora of Guate- mala, for example, is 8 (the same as in Costa Rica and Panama). The number of species is about 95 as far as known, as against 35-40 in Costa Rica and about as many in Panama. The flora of Surinam includes about 90 species. The myrtaceous flora of Peru includes 14 genera and about 175 species; that of Brazil, according to esti- mates by Berg, includes more than 600 species in 33 genera. The genera represented in Central America are all, with the pos- sible exception of Pimento,, represented in South America by more diversified populations and larger numbers of species. There seems to be little or no direct relationship between the North American and South American species, except for weedy species like Psidium guajava and a few like Myrcianthes fragrans and Myrciaria flori- bunda that are widespread in the West Indies and range to northern South America and continental North America. In general there seems to have been a strong phylogenetic con- nection between the myrtaceous flora of the West Indies and that of Central America, especially of the Atlantic lowlands that extend from Honduras to Yucatan. In all the genera having lowland representa- tives (that is, all except Ugni), some of the species are identical with West Indian species or scarcely distinguishable from them. The greatest amount of relative similarity between the two floras, as pointed out above, is where they are geographically closest. Other connections, however, are suggested between the flora of Jamaica and that of Costa Rica (in Myrcianthes), and most curiously (and 1 An earlier paper with the same title dealt primarily with the Myrtaceae of Peru (Fieldiana, Bot. 29: 143-228. 1956). 395 396 FIELDIANA: BOTANY, VOLUME 29 as yet unexplainedly) along an interrupted axis extending from Cuba to Yucatan and thence to western Jalisco and the Tres Marias Islands. There seems to be a certain amount of endemism in both western and eastern Mexico and in the Central American highlands, and a relatively great amount — perhaps a reflection of greater thorough- ness of exploration — in Pete"n and British Honduras. Any sup- posedly undescribed species from Central America should be carefully compared with West Indian material, not only because of the pos- sibility that the same species may occur in both regions, but also because of the high probability that the West Indian plant has already been named and described. This is true not only because of the many species described by Linnaeus, Lamarck, Swartz and others in the eighteenth and early nineteenth centuries, but also because of the many Myrtaceae described by Urban and others in more recent times. Between 1895 and 1931, for example, about 70 species of Calyptranthes were described from the West Indies; this is almost exactly the same number as all the species of Calyptranthes known to Berg a century ago. Notable contributions to our knowledge of the flora of Central America have been made in the last few decades by Paul C. Standley and C. L. Lundell. More than half of all the known Guatemalan species of Myrtaceae were first recognized and described by one or the other of these workers. In my own work of revision I have been aided by many friends and colleagues who have lent herbarium materials and have assisted in other ways. To the authorities at the Jardin Botanique de 1'Etat, Brussels (BR); the Botanical Museum, Copenhagen (C); Chicago Natural History Museum (F); the Harvard University Herbaria (A, GH) ; the Royal Botanic Gardens, Kew (K) ; the Herbarium of the Botanical Institute of the Academy of Sciences, Leningrad (LE) ; the Herbario Nacional del Institute de Biologia, Mexico City (MEXU); the University of California, Berkeley (UC); the United States National Herbarium, Washington (US); and the Naturhis- torisches Museum, Vienna (W), I am grateful for permission to study types and other unique and valuable specimens. Dr. C. L. Lundell has been most gracious in permitting me free access to his large collections of Myrtaceae from Pete"n and elsewhere in the Maya area, including the types and in most instances abundant duplicate materials of species he has recently described. I am also much indebted to Dr. Faustino Miranda for providing me with a study set of his own collections from Chiapas and elsewhere in south- eastern Mexico. McVAUGH: TROPICAL AMERICAN MYRTACEAE 397 In the systematic notes that follow, the terminology is intended to conform to that used in my earlier papers and also in the Flora of Peru (Field Mus. Bot. 13, pt. 4: 569-818. 1958), and a paper on Calyptranthes and Marlierea (Mem. N. Y. Bot. Gard. 10: 61-91. 1958). Some changes in concepts are noted below. 1. CALYPTRANTHES Sw. THE GENUS IN CONTINENTAL NORTH AMERICA There appear to be several centers of endemism in Calyptranthes. Few species are widely distributed. At least one species, C. fascicu- lata Berg, ranges from the Guianas to the Lesser Antilles, but no primarily South American species is known to range northward into North America or even into Colombia except Colombian Guayana. About 16 species are known from the region between the Amazon and the Orinoco (cf. Mem. N. Y. Bot. Gard. 10: 65-79. 1958), and about 20 were treated recently in the Flora of Peru (Field Mus. Bot. 13, pt. 4: 591-616. 1958). The genus is almost absent from the whole Andean region of South America and from the cis-Andean low- lands bordering the Pacific. North of the Amazon and west of the Andean foothills I know of but two local species, viz. C. killipii Standl., from the coastal lowlands about Buenaventura, Colombia, and C. meridensis Steyerm., from mountains in the State of MeYida, Venezuela. In view of the restricted distribution of a majority of the South American species, and the apparent absence of the genus from much of northwestern South America, it is perhaps not surprising to find that most of the North American species are distinctive and com- paratively distantly related. There is apparent, however, among the North American species, a strong West Indian influence. The genus Calyptranthes in the West Indies was reviewed by Urban many years ago (Bot. Jahrb. 19: 591-603. 1895), at which time he recognized about 30 species. About 70 additional West Indian species, almost all from the Greater Antilles, have since been described by Urban, Britton, and others. A new revision of all the Antillean species is much needed, but indeed there seems to be more diversity among the species of Calyptranthes in the Caribbean basin than in any other area in tropical America. Two well-known West Indian species, C. pollens and C. zuzygium, occur in southern Florida as well; C. zuzygium has not been found 398 FIELDIANA: BOTANY, VOLUME 29 elsewhere on the North American continent, but C. pattens is repre- sented in Mexico and Central America by several varieties. A species of Cuba and Jamaica, C. chytraculia, is represented on the mainland by the var. americana, which ranges from Mexico to northern Colom- bia. Other species common to the West Indies and the continent may become known as the flora is revised. The following key includes all the species known to occur natur- ally in continental North America. Descriptions of species are not included unless they are unavailable elsewhere. KEY TO THE SPECIES Leaf-bases broadly cordate or subcordate, the blades nearly sessile or subclasping, on stout short petioles 3 mm. long or less; blades usually acute or obtuse, sometimes obscurely acuminate, never caudate; young leafy branchlets 2- or 4-angIed or -winged except in C. chiapensis with leaves 12-23 cm. long. Young branchlets and inflorescence densely ferruginous-tomentose with con- torted and mostly 1- or 2-branched but nearly erect hairs 0.4-1.3 mm. long. Leaves 1.3-5 cm. wide, 4-12 cm. long, mostly lanceolate; branchlets bicari- nate or winged. Branchlets bialate, the wings terminating distally at each node between the leaf-bases; buds 5.5-7 mm. long; panicles 9-flowered; San Luis Potosf. C. hernandezii McVaugh. Branchlets bicarinate, the keels terminating distally in the median line of the leaf-bases; buds 4.5-5 mm. long; panicles 30-60-flowered; British Honduras C. bartlettii Standl. Leaves 6.5-7.5 cm. wide, 12-23 cm. long, oblong; branchlets terete or com- pressed, not carinate; Chiapas C. chiapensis Lundell. Young branchlets and inflorescence glabrous or thinly pubescent with appressed flat dibrachiate hairs 0.1-0.2 mm. long; branchlets 2- or 4-angled or -winged. Branchlets biconvex in cross section, bicarinate, the winglike angles termi- nating distally in the midlines of the leaf-bases; Guatemala (Alta Vera- paz, PetSn) C. contrerasii Lundell. Branchlets 4-angled, 2 of the angles terminating distally in the midlines of the leaf-bases, the other 2 winged, terminating distally between the leaf-bases. Plants quite glabrous; leaves ovate-oblong; Mexico (Mesa Chica, Vera- cruz) C. karwinskyana Berg. Plants thinly pubescent with minute appressed hairs; leaves lanceolate or elliptic-lanceolate; Tabasco, Chiapas, British Honduras, Pete"n. C. karlingii Standl. Leaf-bases acute or cuneate, or sometimes abruptly rounded, never cordate; blades usually definitely petiolate, often acuminate, sometimes caudate; branchlets various. Flowers 9 or fewer in each panicle, the two lateral branchlets and the terminal one each bearing a single pedicellate flower or a cluster of 2-3 sessile flowers; inflorescence 3-5 cm. long, the very slender or even capillary peduncle usually much longer than the rest of the panicle; leaves long-acuminate, 3-7.5 cm. long. Branchlets 4-angled or -winged, glabrous or essentially so; Mexico (Chiapas) to Panama C. hylobates Amsh. McVAUGH: TROPICAL AMERICAN MYRTACEAE 399 Branchlets terete, when young with abundant appressed or contorted rufous dibrachiate hairs. Branchlets tomentulose with coarse contorted hairs; leaves rounded at base; petioles 2 mm. long; axis of inflorescence filiform, bearing 3 (-5) flowers; Mexico C. tenuipes McVaugh. Branchlets appressed-pubescent with straight longitudinally oriented hairs; leaves acute at base; petioles 3-6 mm. long; peduncle up to about 1 mm. wide, the panicle with 9 flowers or fewer; Guatemala. C. paxillata McVaugh. Flowers more numerous, usually 20-30 or more in each panicle, often sessile or subsessile in small clusters near the tips; panicles as a whole usually 5 cm. long or more, twice or more compound; leaves various, 10-16 cm. long if the flowers are 9 or fewer (in C. perlaevigata with buds 7-8 mm. long). Plants glabrous or sparingly pubescent; pubescence, if any, usually consisting of minute flat appressed dibrachiate hairs 0.1-0.3 mm. long, these often nearly confined to the base of the hypanthium, the nodes of the inflores- cence, and the undeveloped vegetative buds. Panicles 15-20 cm. long; peduncles compressed, 10-12 cm. long, 3-4.5 mm. wide; leaves as far as known 11-12 cm. wide, up to 35 cm. long; in- florescence and buds thinly reddish-pubescent; British Honduras- Peten boundary C. megistophylla Standl. Panicles mostly 10 cm. long or less (if infrequently to 15 cm. long, the pedun- cles 3-5 cm. long, less than 2 mm. wide); leaves mostly 12-15 cm. long or less, in one Panamanian species up to 28 cm. Leaves 12-28 cm. long, caudate-acuminate; buds obovoid, 2-2.5 mm. long, glandular-verrucose; plants sparingly pubescent at flowering time, the panicles and the bases of the buds appressed-pubescent with reddish hairs; Panama C. tumidonodia Schery. Leaves shorter (or the plants quite glabrous) and merely obtuse or acute or short-acuminate; northern Central America to Mexico. Leaves prevailingly obovate to elliptic, obtuse or rounded at tip, 3.5- 6.5 cm. long; small connecting veins evident on the lower surface in dried leaves as a raised reticulum; branchlets not winged; pubescence appressed, rather sparse, grayish- or yellowish-white; British Honduras C. cuneifolia Lundell. Leaves prevailingly elliptic to lanceolate or ovate, acute or acuminate (if small and obtuse the branchlets winged, the small veins ob- scure, the pubescence, if any, reddish). Leaves small and obtuse, mostly 2.5-4.5 cm. long, usually narrow, often varying on the same plant from 2-5 (-7) times as long as wide; branchlets persistently and prominently corky- winged; pubescence sparse, evanescent, of rather loosely appressed yel- low- or reddish-brown hairs; Honduras to British Honduras. C. hondurensis Standl. Leaves usually larger and acute or acuminate, usually 2-3 times as long as wide; branchlets wingless or somewhat obscurely winged. Flower-buds 3-4.5 mm. long, glabrous; whole plant essentially glabrous; calyptra (2.5-) 3-4 mm. wide; persistent collar-like base of the calyx 3-4 mm. wide in fruit. Midvein narrowly impressed above; inflorescence pyramidal, many-flowered, up to 15 cm. long, the lowest branches of the panicle up to 6 cm. long, often as long as the peduncle; flowers clustered near the tips; branchlets winged; state of Mexico . . . . C. hintonii Lundell. 400 FIELDIANA: BOTANY, VOLUME 29 Midvein convex or flat above, sometimes concave, not narrowly impressed; inflorescence mostly 5-7 cm. long or less, with 5-15 flowers, the lowest branches of the panicle up to 2 cm. long and much shorter than the peduncle; flowers mostly solitary. Leaves 3-6 cm. long, obtuse to acuminate; branchlets winged; calyptra membranaceous, 2.5-3.5 mm. wide; flowers con- spicuously pedicellate and solitary; West Indies, Florida. C. zuzygium (L.) Sw. Leaves 10-12 cm. long, prominently acuminate; branchlets compressed, not winged; calyptra conic, thickened, 4 mm. wide; flowers nearly sessile in pairs at the 1-2 nodes of the panicle; Chiapas C. perlaevigata Lundell. Flower-buds [as far as known] (2?-) 2.5-3 mm. long; plants and buds glabrous or pubescent; calyptra [as far as known] 1.5- 2.5 mm. wide; persistent collar-like base of the calyx 1.5- 3 mm. wide in fruit; midvein impressed or sulcate above. Flower-buds and distal nodes of the panicle sparingly pubescent with appressed reddish or pale hairs. Leafy branchlets (at least some of them) wing-angled in most plants; inflorescence rather slender, the principal second- ary branches of the panicle less than 1 mm. wide; moun- tains of Chiapas. C. pollens var. mexicana (Lundell) McVaugh. Leafy branchlets not wing-angled; inflorescence stout, the principal secondary branches of the panicle much flat- tened, their lowest internode often 6-10 mm. long, 1.5 mm. wide; southern Veracruz(?), British Honduras, Yucatan [C. euryphylla Standl.] . . . .C. millspaughii Urb. Plants glabrous except for a few hairs on the youngest vegeta- tive parts; branchlets not winged. El Salvador and British Honduras; petioles 7-10 mm. long; veins and glands scarcely apparent in mature leaves. C. calderonii Standl. Veracruz; petioles 3.5-6 mm. long; veins sometimes incon- spicuous, but evident beneath; glands apparent, the leaves pellucid-punctate or impressed-punctate. Leaves 3-4 times as long as wide, 1.5-2.5 cm. wide, coria- ceous, impressed-punctate but scarcely pellucid-punc- tate; lateral veins strongly ascending, leaving the midvein at an angle of 45° or less; ventral surface of the petiole channeled, the thick obtuse margins bor- dering an ill-defined V-shaped or somewhat concave trough; fruit 5-8 mm. in diameter, the fruiting disk 1.5-2 mm. wide C. schlechtendaliana Berg. Leaves 1.6-3.3 times as long as wide, 1.5-4.5 cm. wide, thin, evidently pellucid-punctate but rarely or only in age impressed-punctate; lateral veins divergent, leav- ing the midvein at an angle of much more than 45°; ventral surface of the petiole broadly and smoothly concave or channeled, the angles acute and continu- ous; fruit 4-6 mm. in diameter, the fruiting disk 1.3- 1.5 mm. wide C, schiedeana Berg. McVAUGH: TROPICAL AMERICAN MYRTACEAE 401 Plants rather generally pubescent on the young growth and the inflorescence, often softly and loosely so, at least some of the hairs branched and spread- ing or erect, often up to 0.5 mm. long, at least to some extent persistent on the branchlets, the inflorescence and the fruits. Leaves prevailingly obovate to elliptic, obtuse or rounded at tip, 3.5- 6.5 cm. long; branchlets not winged; small connecting veins evident on the lower surface in dried leaves as a raised reticulum; pubescence appressed, rather sparse, grayish- or yellowish-white; British Honduras. C. cuneifolia Lundell. Leaves prevailingly elliptic to lanceolate or ovate, acute or acuminate (if small and obtuse the branchlets winged, the small veins obscure, the pubescence reddish). Leaves small and obtuse, mostly 2.5-4.5 cm. long, usually narrow, often varying on the same plant from 2-5 (-7) times as long as wide; branchlets persistently and prominently corky- winged; pubescence sparse, evanescent, of rather loosely appressed yellow- or reddish- brown hairs; Honduras to British Honduras . .C. hondurensis Standl. Leaves usually larger and acute or acuminate, if narrow usually consist- ently so; branchlets usually wingless. Young leafy branchlets softly brown- to rufous-tomentose with hairs of two types: straight and nearly erect, simple or short-branched hairs up to 2 mm. long, these standing above the tomentum of contorted and somewhat appressed hairs with two equal or strongly unequal branches; hairs of the panicles of the contorted, appressed type; buds 3 mm. long or less, obovoid, mostly short- apiculate; branchlets not winged. Leaves narrowly elliptic-lanceolate, 1-2.3 cm. wide, 4-6 times as long as wide; Tabasco, British Honduras and northern Guate- mala C. lindeniana Berg. Leaves elliptic-ovate, 2.5-6 cm. wide, 2-3 times as long as wide. Leaves mostly 4-6 cm. wide, 7-13 cm. long; pubescence pale red- dish, reddish-brown or yellowish-brown; southern Mexico to northern Colombia. .C. chytraculia var. americana McVaugh. Leaves mostly 2-4 cm. wide, 5-10 cm. long; pubescence reddish brown to dark red; Cuba and Jamaica. C. chytraculia (L.) Sw., var. chytraculia. Pubescence of leafy branchlets not as above, not noticeably different from that of the panicle-branches. Hairs coarse, 0.3-1.5 mm. long, erect and the branchlets tomentose, or appressed but then the individual hairs 0.5-1 mm. long; buds [as far as known] 4.5 mm. long or more; branchlets not wing- angled. Leaves broadly rounded at base; lateral veins impressed above; petioles 8-11 mm. long; buds ovoid, short-apiculate, 4.5- 6 mm. long; tomentum dense, yellowish-brown, the very numerous erect hairs simple or forked, up to 1.3-1.5 mm. long; Guatemala C. macrantha Standl. & Steyerm. Leaves acute at base; lateral veins not impressed above; petioles 4-7 mm. long; hairs various. Branchlets bristly-tomentose, the very coarse hollow erect pale yellowish-brown forked hairs up to 0.3-0.7 mm. high; buds unknown, the calyptra shallowly dome-shaped, 2 mm. wide, short-apiculate; Honduras C. zanquinensis A. Molina. 402 FIELDIANA: BOTANY, VOLUME 29 Branchlets clothed with appressed nearly sessile longitudinally oriented golden-brown dibrachiate hairs 0.5-1 mm. long; buds 9-14 mm. long, the trumpet-shaped calyptra 6-9 mm. long; Panama C.johnstonii McVaugh. Hairs minute, dibrachiate, often appressed, sometimes crisped, mostly 0.1-0.3 mm. long, the pubescence of the young branch- lets and inflorescence rusty-brown, yellowish brown or, especially on the leaves, nearly colorless. Peduncle 2.5-3 mm. wide just below the lowest branches; buds 4-4.5 mm. long; calyptra 2.5-3 mm. wide; hairs dark reddish brown; Costa Rica, at 2600 m C. pittieri Standl. Peduncle 1.5-2 mm. wide or less; buds 3 mm. long or less; calyptra 2.5 mm. wide or usually less; hairs coppery to yellow or gray- white. Branchlets wing-angled; leaves cuneate at base, often 4-7 cm. long, usually glandular-punctate above; inflorescence copi- ously soft-pubescent with reddish or sometimes pale hairs; West Indies; southern Florida; Isthmus of Tehuantepec; Yucatan C. pallens Griseb., var. pallens. Branchlets terete or compressed, not wing-angled; leaves acute at base, varying to somewhat rounded or occasionally to cuneate; Mexico and Central America. Inflorescence stout, the principal secondary branches of the panicle much flattened, their lowest internodes often 6-8 mm. long, 1.5 mm. wide; pubescence of very small flat closely appressed longitudinally oriented hairs; Vera- cruz; British Honduras to Yucatan; [C. euryphylla Standl.] C. millspaughii Urban. Inflorescence more slender, the internodes of the secondary panicle-branches terete or slightly compressed, 1 mm. wide or less; pubescence, at least in part, of crisped or soft hairs; mountains of Mexico and Guatemala. Leaf-blades often 7-12 cm. long, eglandular above; petioles 8-10 mm. long; branchlets terete or compressed; pan- icles often 50-60-flowered; inflorescence with often densely aggregated coppery hairs about the base of the hypanthium but otherwise rather sparingly pubes- cent; mountains of Guatemala and southern Mexico. C. pendula Berg. Leaf-blades often 3-7 cm. long, often glandular-punctate above but sometimes inconspicuously so; petioles 2.5- 6 mm. long; branchlets sometimes wing-angled. Branchlets of the inflorescence copiously soft-pubescent at least in the flowering stage, the hairs often pale and the whole inflorescence appearing somewhat grayish; leafy branchlets rarely if ever wing-angled; panicles often 50-60-flowered; Costa Rica. C. pallens var. williamsii (Standl.) McVaugh. Branchlets of the inflorescence glabrous or sparingly appressed-pubescent with minute coppery hairs; leafy branchlets sometimes wing-angled; panicles mostly 30-35-flowered; Mexico. C. pallens var. mexicana (Lundell) McVaugh. McVAUGH: TROPICAL AMERICAN MYRTACEAE 403 Calyptranthes chytraculia (L.) Sw. Prodr. 79. 1788. Myrtus chytraculia L. Syst. Nat. ed. 10. 1056. 1759. Known originally from the West Indies only, this species seems to be represented in Central America by a slightly different taxon, described below as a variety. The characteristic pubescence is so nearly identical in all these plants, and so distinctive, that there can be little doubt of close relationships among them. The mostly red- dish-brown hairs are abundant, rather soft, those of the panicles mostly much contorted, loosely appressed and with two equal or strongly unequal branches; the young leafy branches bear similar hairs, and bear also a number of fine, straight, nearly erect hairs up to 2 mm. long, these simple or with a short branch near the base. As far as I know, these two kinds of hairs, that give the young leafy branches a unique shaggy appearance, are found only in C. chytra- culia and in C. lindeniana Berg, which hardly differs except in its very narrow leaves. Some Central American specimens are virtually indistinguishable from West Indian ones, but usually the leaves of Central American plants are larger (4-6 cm. wide, 7-13 cm. long, as against 2-4 cm. wide, 5-10 cm. long, in typical chytraculia). The hairs in West In- dian specimens are often reddish-brown or dark red, whereas in Central American plants the hairs are paler, varying from reddish- brown to yellowish-brown or pale reddish. A specimen from north- ern Colombia (region of Santa Marta, H. H. Smith 897}, agrees perfectly with specimens from northern Central America. Miss Amshoff, in the Flora of Panama (Ann. Missouri Bot. Card. 45: 169. 1958), referred a single Panamanian specimen to C. chytraculia with the note that it "quite agrees with specimens from British Honduras distributed as C. millspaughii Urb." The plant of British Honduras and Central America has long been confused with Calyptranthes millspaughii Urb. (Symb. Antill. 7: 294. 1912), which was based on Millspaugh 1537 from Cozumel Island. This plant belongs to a quite different species. The pubes- cence is scanty, of short flat appressed coppery hairs, not at all like the shaggy abundant pubescence of chytraculia. The panicles are 4 cm. long or less, about 30-flowered, contrasting with the larger, many-flowered panicles of chytraculia. The taxonomy of C. chytraculia may be summarized as follows: Calyptranthes chytraculia (L.) Sw., var. chytraculia. Cuba and Jamaica; for synonymy and citation of specimens see Urban, Bot. Jahrb. 19: 597-598. 1895. 404 FIELDIANA: BOTANY, VOLUME 29 Calyptranthes chytraculia (L.) Sw., var. americana McVaugh, var. nov. A var. chytraculia foliis maioribus 7-13 cm. longis, pilis pallidioribus pallide rufis, vel rufo-flavidis usque ad fulvis, differt. Near the Atlantic Coast of Central America, mostly at elevations of 200 meters or less, often along rivers or lake shores; British Hon- duras and Department of Pete"n, Guatemala, to northern Colombia. Representative specimens examined : BRITISH HONDURAS: Maskall, P. H. Gentle 1217 (MICH); Big Fall, Belize River, Lundell 4343 (MICH) ; Cohune Ridge, El Cayo Dist., Lundell 6431 (MICH); Temash River, W. A. Schipp 1295 (MICH). GUATEMALA: Pete"n: Uaxactun, 4 Apr. 1931, H. H. Bartlett 12 W (MICH, type; UC), 24 Apr. 1931, Bartlett 12723 (MICH).— Izabal: Along Rio Frio, Steyermark 41632 (MICH). NICARAGUA: Isla de Providencia, Proctor 3414 (US); Pearl La- goon, Allen 6524 (F). Calyptranthes hernandezii McVaugh, sp. nov. Frutex 1-1.5 m. altus, ramulis bialatis, alis inter folia terminantibus; ramuli paniculaeque tomentosae, pilis flaccidis flavidis saepe inaequaliter dibrachiatis usque ad 1.3 mm. longis tectae; folia subsessilia, cordata, obtusa, 4-9 cm. longa, 2-3.7-plo longiora quam latiora; paniculae 5-6 cm. longae, 9-florae, floribus tri- geminis sessilibus, alabastris 5.5-7 mm. longis apiculatis, calyptra 3-5 mm. longa, 3.5-4 mm. lata; stylus 10 mm. longus; stamina ca. 100. A shrub 1-1.5 meters high, the branchlets bialate, the wings 0.5 mm. high, terminating distally between the leaf-bases; branchlets and inflorescence loosely tomentose with erect flaccid simple or unequally forked pale yellow-brown hairs up to 1.3 mm. long; leaves coriaceous, glabrate, nearly sessile, lanceolate to oblong or ovate on the same plant, 1.3-3 cm. wide, 4-9 cm. long, 2-3.7 times as long as wide; blades narrowed to a broad or slender but always obtuse tip, cordate at base, the stout petiole 1-2 mm. long, 2 mm. wide, ventrally concave; midvein broadly and deeply concave above; lateral veins rather close and parallel, 10-20 including some intermediate ones, inconspicuous, more evident beneath; marginal vein about equaling the lateral ones and but slightly arched between them, 1-1.5 mm. from the margin; blades green and sometimes impressed-punctate above, paler and with convex resinous glands beneath; paired panicles 5-6 cm. long, 9-flowered, the peduncle 3-4 cm. long, stout, compressed, up to 2.5 mm. wide; flowers sessile in 3's at the tip of a terminal branch 1.5 cm. long and at the tips of a pair of lateral branches 8 mm. long, 2.5 mm. wide; bracts at the base of the flower-cluster boat- shaped, scarious, deciduous at anthesis, enveloping the lateral bud, 6 mm. long; bracteoles similar, deciduous, ovate, 3 mm. long; buds 5.5-7 mm. long, broadly fusiform or the body subglobose and 4 mm. in diameter, the base tapering and sessile, the tip bearing an apiculum 1-2.5 mm. long; calyptra domelike, 3-5 mm. McVAUGH: TROPICAL AMERICAN MYRTACEAE 405 long including the apiculum; hypanthium after anthesis funnelform, 3.5-4 mm. wide at the mouth; style about 10 mm. long; stamens about 100, 7-8 mm. long, the anthers 0.5 mm. long; petals and fruit not seen. MEXICO: San Luis Potosi: Near river edge, Taman, 15 km. south- west of Tamazunchale, [elevation ca. 200 m.], 14 Apr. 1944, E. Her- nandez X. 159 (LL, type). Calyptranthes johnstonii McVaugh, sp. nov. Arbor vel frutex, ramulis teretibus, ramulis foliisque juvenilibus et inflores- centia copiose pilis flavidis, nitidis, dibrachiatis, 0.5-1 mm. longis, obsitis; folia late elliptica, 6-11 cm. longa, basi acuta, petiolis 4-6 mm. longis, nervo medio supra ad basin impresso; inflorescentiae geminatae, spicatae, 5-7 cm. longae, pauciflorae; alabastra cornuta, 9-14 mm. longa, hypanthiis campanulatis, 3-5 mm. longis; stylus 8-10 mm. longus; fructus globosus, glabrescens, ca. 1.5 cm. crassus. Shrub or tree 2.5-10 m. high, the "bark on trunk and stems pale, moderately smooth, not exfoliating" (Johnston); branchlets terete; inflorescence, and young branchlets and leaves, copiously appressed-pubescent with lustrous golden-brown dibrachiate hairs 0.5-1 mm. long; leaves coriaceous, broadly elliptic, about equally narrowed to the prominently short-acuminate tip and the acute base, (2.5-) 3-5 cm. wide, 6-11 cm. long, the margins decurrent on the inner angles of the channeled petiole 4-6 mm. long, 1-1.5 mm. thick; midvein impressed above near base; lateral veins rather numerous and parallel, 12-15 pairs, not very conspicu- ous (more so on the lower surface) ; marginal vein about equaling the lateral ones and somewhat arched between them, 1-2 mm. from the margin; leaves dark green and glandular-punctate above, sparingly pubescent and inconspicuously glandular beneath; inflorescence paired at the terminal nodes of old branchlets, spicate, 5-7 cm. long; flowers few, probably 2-3 clustered at a node 2-4 cm. above the base and 2-3 clustered at the tip of the spike; buds cornute, 9-14 mm. long, the calyptra trumpet-shaped, 6-9 mm. long, 4 mm. wide at mouth, thence abruptly contracted to a slender tapering tip 5-8 mm. long; hypanthium 3-5 mm. long, campanulate, prolonged 2.5 mm. beyond the summit of the ovary; style 8-10 mm. long; stamens about 50; petals none(?), or 1-2, narrow, 1-2 mm. long; fruit glo- bose, glabrescent, about 1.5 cm. in diameter, surmounted by a collar about 4 mm. wide, 1 mm. high; mature seeds not seen. PANAMA: Canal Zone: South of Ft. Sherman, in wet places back of the mangrove swamp, 1 Apr. 1956, 1. M. Johnston 1764 (MICH, type, in flower), 1764 (MICH, isotype, with nearly mature fruits); 3 miles northwest of Gatun Locks, edge of marsh, 9 Aug. 1955, John- ston 1571 (MICH); near Ft. Sherman, P. C. Standley 30933 (US). Calyptranthes millspaughii Urb. Symb. Antill. 7: 294. 1912. C. euryphylla Standl. Contr. U. S. Nat. Herb. 23: 1034. 1924. Examination of the type of C. millspaughii, a flowering specimen from Cozumel Island, Quintana Roo, Millspaugh 1537 (F), shows this to represent a distinctive taxon, quite different from the copi- UNIVERSITY OF 406 FIELDIANA: BOTANY, VOLUME 29 ously soft-hairy Central American plant that has in recent years passed under the name of C. millspaughii. The more pubescent plant is C. chytraculia var. americana McVaugh, well known along the Atlantic coast from British Honduras to Colombia. The true C. millspaughii seems to be known from but a single collection in addition to the type; the second collection is also a flowering speci- men, from British Honduras, Stann Creek District, Freshwater Creek, J. Kelly no. 1 (F, US). There is a strong superficial resemblance between C. millspaughii and C. pendula Berg, and likewise between C. millspaughii and the several varieties of C. pollens. From all of these C. millspaughii may be distinguished by the broad, flat, short internodes of the pan- icle branches. The hairs of the inflorescence of C. millspaughii are distinctive also; they are copper-colored, longitudinally oriented, very closely appressed and straight, the individual hairs readily observed. In C. pendula and in the varieties of C. pollens the hairs are crisped and often somewhat matted together. The type and only known specimen of C. euryphylla Standl., col- lected at Catemaco, Veracruz, in April 1894, Nelson 1+21 (US), bears half -grown fruit; flowers and buds are wanting. The leaves and the panicle-branches are sparingly pale-pubescent, the appressed hairs very like those of C. millspaughii but almost colorless. Apparently the paleness of the hairs is related to the maturity of the plant when collected, for the hairs persisting on the fruit and on some of the dis- tal nodes of the panicle are copper-colored and precisely like those of C. millspaughii. The lower internodes of the lateral panicle- branches in the type of C. euryphylla are 10 mm. long or less and about 1.5 mm. wide, exactly comparable to those of C. millspaughii. The foliage of C. euryphylla, as noted by Standley, is notable in that the leaves are ovate and rather broadly rounded at base, and the petioles are short; I suspect that this represents an individual dif- ference. In other respects the leaves of euryphylla and millspaughii are closely similar; both are impressed-punctate above, with im- pressed midvein; the lateral veins are but indistinctly seen in the rather opaque brownish lower surface of the rigid dried leaves. Allowing for the differences between flowering and fruiting speci- mens, there can be little doubt that C. euryphylla and C. millspaughii are conspecific. Calyptranthes pallens Griseb. Veg. Kar. 67. 1857. ?Eugenia pollens Poir. in Lam. Encyc. Suppl. 3: 122. 1813 (fide Urban, Bot. McVAUGH: TROPICAL AMERICAN MYRTACEAE 407 Jahrb. 19: 598. 1895). C. chytraculia 8 pauciflora Berg, Linnaea 27: 27. 1855. A species of the Caribbean basin, the nomenclaturally typical variety occurring in the West Indies, southern Florida, the Yucatan Peninsula and the Isthmus of Tehuan tepee, Mexico; other varieties, distinguished somewhat subjectively, occur in Mexico and Central America. A tree or tree-like shrub 5-8 m. high; leaves coriaceous, rather broadly elliptic, or sometimes broadest a little above or below the middle, narrowed about equally to both ends, 2-4 (-6) cm. wide, 3-7 (-10) cm. long, 2-3 (-3.5) times as long as wide; margins of the blade decurrent at base on the inner angles of the channeled petiole; midvein impressed above at least near base; lateral veins close and par- allel, obscure on both surfaces, 10-20 pairs including some poorly denned inter- mediate ones; marginal vein about equaling the lateral ones and slightly arched between them, 0.5-1.5 mm. from the margin; leaves smooth and dark green above; panicles usually bracteate and paired at the base of an otherwise leafy shoot, sometimes axillary on ordinary leafy branches, the individual panicles pyramidal, 3-4 times compound, the flowers in small clusters near the tips, not appearing closely sessile; peduncles 2-5 cm. long; buds 2.5-3 mm. long including the narrow base, obovoid, rounded and often slightly apiculate; calyptra domelike; hypan- thium after opening campanulate, 2 mm. long; style 4-5 mm. long; stamens about 50-60, up to 3.5 mm. long; petals tiny, adhering to the calyptra and falling with it; ripe fruit dark red, globose or oblate, 5-8 mm. in diameter. Young leafy branchlets 2-winged; leaves cuneate at base, usually glandular- punctate above; inflorescence copiously soft-pubescent with reddish or some- times pale hairs; panicles often 50-60-flowered; West Indies; southern Florida; Isthmus of Tehuantepec var. pallens. Young leafy branchlets not winged, or some plants with a few branchlets winged; leaves acute at base, varying to somewhat rounded or occasionally to cuneate, often glandular-punctate above but sometimes inconspicuously so; Mexico and Central America. Branchlets of the inflorescence copiously soft-pubescent at least in the flowering stage, the hairs often pale and the whole inflorescence appearing somewhat grayish; leafy branchlets rarely if ever winged; panicles often 50-60-flow- ered; Costa Rica var. williamsii. Branchlets of the inflorescence glabrous or sparingly appressed-pubescent with minute coppery hairs; leafy branchlets sometimes winged; panicles mostly 30-35-flowered; Mexico var. mexicana. Calyptranthes pallens Griseb., var. pallens. Vegetative branches prominently 2-winged, the wings terminating distally between the leaf-bases; inflorescence, branchlets, and to some extent the foliage, rather persistently appressed-pubescent with pale rusty-brown, yellowish brown or, especially on the leaves, nearly colorless dibrachiate hairs 0.1-0.3 mm. long; leaves prominently acuminate at tip, broadly cuneate at base, usually glandular- punctate above; petioles 6-8 mm. long; panicles 10-12 cm. long or less, usually with 50-60 or more flowers, the lowest branches often as long as the peduncles, the latter 2-4 cm. long; calyptra 1.3-1.7 (-2) mm. across. 408 FIELDIANA: BOTANY, VOLUME 29 West Indies and southern Florida. By Urban (Bot. Jahrb. 19: 598. 1895) reported from Cuba, Jamaica, Haiti, Cayman Islands, St. Thomas, St. Croix and Guadeloupe. I have accepted Urban's identification of this species but I have not seen the type. The nomenclature of C. pollens is somewhat involved. The name is often cited as C. pollens (Poir.) Griseb., with the implication that it was based upon Eugenia pattens Poir. The latter is included by Urban in the synonymy of C. pattens, but without the customary sign of verification [!]. Grisebach, however, in his first publication of C. pal- lens, made no mention of Eugenia pattens Poir., citing only C. chytra- culia var. 8 pauciflora Berg. Berg's variety pauciflora was based wholly upon a specimen collected in Guadeloupe by Bertero, and studied by Berg in Sender's herbarium. Urban indicated with a [!] that he had verified the identity of var. pauciflora, and that it was indeed the same as C. pattens Griseb. The type of Eugenia pattens Poir., however, was collected on the island of St. Thomas by Ledru; it may well represent the same species, as supposed by Urban, but it does not form part of the nomenclatural basis of Calyptranthes pattens Griseb. The latter is readily identified from Grisebach's de- scription in Fl. Brit. W. Ind. 233. 1860, as it is the only species with 2-edged branchlets and "cymes rusty-sericeous." The following specimens are quite indistinguishable from West Indian material of C. pattens, except that the petioles are no more than 3-5 mm. long: MEXICO: Oaxaca: Mena, C. D. Mell 7, 20 Jan. 1927 (US); Chi- vela, Mell 2275, 1 Mar. 1934 (US) ; Rincon Antonio, C. R. Orcutt 3233 (F, US). — Yucatan: 3 km. de El Cuyo en el camino a Colonia Yuca- tan, 0. G. Enriquez 479 (MICH). Calyptranthes pallens Griseb., var. williamsii (Standl.) McVaugh, comb. nov. C. williamsii Standl. Ceiba 1: 156. 1950. ?C. costa-ricensis Berg, Linnaea 27: 20. 1855. Vegetative branchlets terete or slightly compressed; inflorescence, branchlets, and to some extent the foliage rather persistently appressed-pubescent with pale rusty-brown, yellowish brown or nearly colorless dibrachiate hairs 0.1-0.3 mm. long; leaves acuminate at tip (sometimes obscurely so), the base usually not cuneate but the margins narrowed and rounded to an acute base; blades incon- spicuously if at all glandular, sometimes glandular-punctate above; petioles 4- 6 mm. long; panicles 6-8 cm. long, about 50-60-flowered, the lowest branches shorter than the peduncles, the latter 3-5 cm. long; calyptra 2-2.5 mm. wide. Mountain slopes, river banks and open pastures, at elevations of 800-2000 m., central Costa Rica, as far as known mostly in the prov- inces of Cartago and San Jose", flowering March-April, fruiting in June. McVAUGH: TROPICAL AMERICAN MYRTACEAE 409 A similarity between the West Indian pattens, and the species described as C. costa-ricensis Berg, was suggested by Urban. Berg's original description of "C. Costa-Ricensis" seems to apply satisfac- torily to the plant described above, but I have not seen the type ("Costa Rica et Veragua, de Warszewicz, . . . v.s. in hb. Berol.") and therefore prefer to base this new variety on a species that can be satisfactorily typified. Specimens examined: COSTA RICA: Rio Maria Aguilar, J. D. Smith 6499 (F); Rio Ciru- elas, massif du Barba, Tonduz 2211 (F) ; Sta. Maria de Dota, Standley 41716 (F); San Sebastian, Standley 32683 (F); San Jose", Stand- ley 34816 (F), 47371 (F); Cerro de la Carpintera, Standley 34516 (F); Mt. Carpintera, H. E. Stork 314 P-P- (MICH); Alto de Ocho- mogo, Williams & Molina 13841 (F; US, type), Williams 16001 ("probably from same tree from which type came," MICH) ; Cerro Carpintera, above Tres Rios, Williams 16141 P-P- (MICH); La Car- pintera, Cerro del Aguacate, Williams 16579 (MICH); along Rio Sucio near Corinto, Williams 16555 (MICH). Calyptranthes pallens Griseb., var. mexicana (Lundell) McVaugh, comb. nov. C. mexicana Lundell, Wrightia 2: 166. 1 May 1961. Vegetative branches winged or wingless, the wings sometimes apparent only on a few of the young branchlets of individual plants; inflorescence and branchlets very sparingly pubescent, the appressed and usually coppery hairs mostly 0.1- 0.2 mm. long, and aggregated chiefly about the base of the hypanthium and the distal nodes of the panicle; leaves acuminate at tip (sometimes obscurely so or some leaves merely obtusely rounded), acute or sometimes cuneate at base, often glandular-punctate above; petioles 2.5-5 mm. long; panicles 3-7.5 cm. long, 30- 35-flowered, the lowest branches shorter than the peduncles, the latter 2-4.5 cm. long; calyptra 2-2.5 mm. wide. Arroyos, brushy slopes, stream-sides, wet mountain forests, southern and western Mexico, at elevations from ca. 500 m. to 2600 m. MEXICO: Chiapas: Mt. Ovando, near Escuintla, E. Matuda 1838 (LL, type; MO); Tres Cruces, 2600 m., Matuda 5024 (F; LL, para- type). — Jalisco: Pihuamo, M. E. Jones 18 (F, US); 10 miles south of Autlan, Wilbur & Wilbur 1415 (MICH); San Sebastian, Hda. del Ototal, Y. Mexia 1806 (MICH, US).— Nayarit: 8 miles west of Tepic, McVaugh 18903 (MICH); Maria Madre Island, T. S. Maltby 110 (US).— -Sinaloa: Without collector or locality (US 1014144); Ixtagua, Mun. San Ignacio, J. Gonzalez Ortega 776 (MEXU). 410 FIELDIANA: BOTANY, VOLUME 29 The following specimen, superficially differing from the preced- ing ones in its long internodes and rather narrow and slenderly acuminate leaves up to 10 cm. long and 3-4 times as long as wide, is apparently merely a flowering stump-sprout or other vigorous shoot: MEXICO: Sinaloa: San Ignacio, M. Narvaez M. & A. E. Salazar 846 (US). Calyptranthes paxillata McVaugh, sp. nov. Arbor parva, ramulis subteretibus vel compressis; ramuli inflorescentiaeque, et folia juvenilia, pilis rufidulis appressis dibrachiatis subsessilibus longitudinaliter directis 0.5-1 mm. longis obsiti; folia elliptica vel elliptico-lanceolata, caudato- acuminata, 4-7 cm. longa, basi acuta vel cuneata, petiolis 3-6 mm. longis, nervo medio supra versus basin canaliculate; inflorescentiae 3-4.5 cm. longae, ca. 9-florae, e basi ramulorum foliosorum vel abortivorum geminati oriundi; pedunculi 3.5 cm. longi, statu frutescenti 1 mm. crassi; alabastra obovoideo-fusiformia, 4.5-5 mm. longa, apiculo obtuso 1-1.3 mm. longo paxillato; stylus 7-8 mm. longus. A small tree with nearly terete or compressed branchlets; branchlets, inflores- cence and young foliage covered with appressed, nearly sessile, dibrachiate and longitudinally oriented coppery hairs 0.5-1 mm. long; leaves glabrate, elliptic or elliptic-lanceolate, 1.5-3.5 cm. wide, 4-7 cm. long, 2.3-3 times as long as wide, caudate-acuminate, the slender tapering tip 1-1.5 cm. long, obtuse; base of blade acute or cuneate, the margins decurrent on the inner angles of the petiole 3-6 mm. long, 0.8-1.2 mm. thick; midvein channeled above near base; lateral veins obscure, close and parallel, about 15 on each side, including some intermediate ones; mar- ginal vein about equaling the lateral ones, nearly straight between them, 1 mm. from the margin; inflorescence of paired panicles from an abortive axillary axis or from the lowermost (bracteate) node of a leafy shoot; panicles 3-4.5 cm. long, twice compound, bearing 9 flowers or fewer; peduncle up to 3.5 cm. long, com- pressed, 1 mm. wide below the lowest branches; buds obo void-fusiform, 4.5-5 mm. long, 2-2.5 mm. in diameter, broadest above the middle, tapering to the base, rather abruptly contracted to a blunt peglike apiculum 1-1.3 mm. long; hypan- thium after anthesis funnelform, 2.5-3 mm. high, the rim prolonged 2 mm. beyond the summit of the ovary; calyptra about 3 mm. across, 2.5 mm. long including the apiculum; style 7-8 mm. long; stamens about 100, up to 4 mm. long; petals 1-2, minute, remaining attached to the calyptra and falling with it; fruit unknown. GUATEMALA: Alta Verapaz: Wet forest above Tactic, above the bridge across Rio Frio, alt. 1400-1500 m., 30 Mar. 1941, P. C. Stand- ley 90457 (US, type). This species is distinctive by virtue of its copious appressed pubescence of long hairs, the few-flowered panicles, long styles and long-apiculate fusiform buds. This peculiar "peglike" apiculum gives the name to the species; I do not find the word paxillatus in any Latin dictionary at my disposal, but as the word "paxillate" may be found in English dictionaries, it seems reasonable to provide a Latin equivalent for it. McVAUGH: TROPICAL AMERICAN MYRTACEAE 411 Calyptranthes schiedeana Berg, Linnaea 27: 28. 1855. Myr- cia aromatica Schlecht. Linnaea 13: 415, pro max. part. 1839. Not C. aromatica St. Hil. This species and the next were based by Berg upon the two col- lections assigned by Schlechtendal to Myrcia aromatica. The type of C. schiedeana was from Hacienda de la Laguna, Veracruz, Schiede (in herb. Schlecht.), and was presumably the collection upon which M. aromatica was chiefly based. Several collections by C. A. Purpus (e.g. nos. 2927, 10778), all from Veracruz, seem to represent the same species as a presumed isotype shown in Field Mus. Neg. 23392. Calyptranthes schlechtendaliana Berg, Linnaea 27: 29. 1855. Myrcia aromatica Schlecht. Linnaea 13: 415, quoad spec, in sylvis Papantlae coll. 1839. The type of C. schlechtendaliana was Schiede no. 25 (in herb. Schlecht.), collected at Papantla, Veracruz, and presumably the same specimen mentioned by Schiede on page 416 as similar to his Myrcia aromatica, described on the previous page. A photograph purporting to be C. schlechtendaliana (Field Mus. Neg. 36864) seems actually to represent a specimen of C. schiedeana. One collection, Purpus 8810 (UC, US), from Veracruz without definite locality, seems to represent C. schlechtendaliana. The dif- ferences between it and C. schiedeana are listed in the key above. As far as can be determined from the scanty material at hand, the two species are distinct. Calyptranthes tenuipes McVaugh, sp. nov. Frutex, ramulis teretibus; ramuli teneri densiuscule pilis crassis, contortis, nitidis, fuscis, dibrachiatis 0.5-0.8 mm. longis obsiti; folia ab initio subglabra, lanceolato-ovata, caudato-acuminata, 3-5 cm. longa, basi rotundata, petiolis 2 mm. longis, nervo medio supra impresso; inflorescentiae 3-4 cm. longae, ut videtur 3-5-florae, e basi ramulorum foliosorum geminati oriundi; pedunculi filiformes, statu fructescenti 0.3-0.4 mm. crassi; flores non vidi; fructus globosus, 7-9 mm. crassus, umbilicatus; seminis maturi cotyledones carnosi, contorto-plicati. A shrub 2 m. high, glabrate, the branchlets terete; young branchlets, and to a lesser extent the very young inflorescence, more or less thickly pubescent with coarse contorted lustrous reddish-brown dibrachiate hairs 0.5-0.8 mm. long; leaves glabrous from the first except for a few hairs on the lower surface near the mid- vein, the blades lance-ovate, caudate-acuminate, 1-2.3 cm. wide, 3.5 cm. long, the narrow acumen 1-1.8 cm. long, rounded and 1-2 mm. wide at its tip; base of blade rounded, the margins passing abruptly onto the inner angles of the deeply sulcate petiole 2 mm. long, 1 mm. thick; midvein impressed and sulcate above, the other veins obscure, the lateral ones about 10 pairs, the marginal vein about 1 mm. from 412 FIELDIANA: BOTANY, VOLUME 29 the margin; blades dark green and nearly featureless above, paler beneath, the glands obscure; inflorescences 3-4 cm. long, paired at the base of a leafy shoot; peduncle 2-2.5 cm. long, filiform, 0.3-0.4 mm. thick in fruit; flowers unknown, probably 3-5, on pedicels or lateral branches up to 3 mm. long; fruit globose, 7-9 mm. in diameter, surmounted by a collar about 2 mm. wide, less than 1 mm. high; seed 1, globose, lustrous, up to 8 mm. in diameter; cotyledons contorted. In the absence of flowers this species is assigned to the genus Calyptranthes on the basis of the contorted cotyledons, the collar- like remains of the calyx suggesting that the buds open by a calyptra, and the dibrachiate hairs characteristic of Calyptranthes but not found in other genera of the Myrcioideae. MEXICO: Veracruz: In a deep barranca below Atzalan, elev. ca. 1650 m., 13 Mar. 1946, A. J. Sharp 46165 (MICH, type). Calyptranthes zuzygium (L.) Sw. Prodr. 79. 1788. Myrtus zuzygium L. Syst. Nat. ed. 10. 1056. 1759. A West Indian species, known from Hispaniola, Jamaica and Cuba; also in southern Florida. Reported from "Michoacan and Tres Marias Islands" by Standley (Contr. U. S. Nat. Herb. 23: 1034. 1924) ; judging from specimens annotated by Standley, these reports are based upon M. E. Jones 18, from Pihuamo, Jalisco (not Micho- acan, as labeled by Jones), and T. S. Maltby 110, from Maria Madre Island. I should refer these specimens to Calyptranthes pattens var. mexicana, q.v.; C. zuzygium is so far unknown in Mexico. MISCELLANEOUS NOTES Calyptranthes tovarensis (Berg) Steyermark, Fieldiana, Bot. 28: 1010. 1957. Myrcia tovarensis Berg, Linnaea 27: 118. 1855. This name resulted from a misinterpretation of Myrcia tovarensis, and a misidentification of Tamayo 227, which is indeed not a Myrcia but a species of Calyptranthes. The type of M. tovarensis, however (Moritz 17 If?}, represents an undoubted and distinctive species of Myrcia,. In addition to the type collection, which I have seen at Chicago Natural History Museum, the following specimens from Venezuela seem to be representative of M . tovarensis: Colonia Tovar [the type locality], Fendler 42 (PH, NY); Caracas, Linden 13 (F). Calyptranthes vexata McVaugh, nom. nov. C. apoda McVaugh, Mem. N. Y. Bot. Card. 10: 70. 1958, not C. apoda Urban, Symb. Antill. 9: 96. 1923. I am indebted to Dr. Richard S. Cowan for calling this nomenclatural lapse to my attention. McVAUGH: TROPICAL AMERICAN MYRTACEAE 413 2. EUGENIA L. SYNOPSIS OF THE MEXICAN SPECIES WEST OF THE ISTHMUS OF TEHUANTEPEC In 1924, in Standley's Trees and Shrubs of Mexico, 24 species of Eugenia were recorded; two additional species now known to belong to Eugenia (Myrtus oaxacana and M. ledophylla) were also included in this work but under the genus Myrtus. Now after almost four decades of active botanical exploration in Mexico, approximately 60 species of Eugenia are known to occur as native plants in that country. The following synopsis includes the approximately 40 spe- cies known from west and north of the Isthmus of Tehuantepec. The myrtaceous flora of Chiapas and Tabasco, and that of the Yuca- tan peninsula, is quite a different one and as yet imperfectly known. Essentially all the known species are treated in the Flora of Guate- mala (Fieldiana, Bot. 24). From western Chiapas and Tabasco very few collections are available; it may be supposed that some new species will come to light from this region, and that some addi- tional species will be found to extend into the region from the isth- mus. Eventually it may be convenient and useful to treat all Mexican species in one publication, but it is expedient in the pres- ent state of knowledge to treat as a unit what may be called the endemic Mexican flora, as distinct from that with Central American or West Indian affinities. It is difficult to make a generally useful key to any large group of species of Myrtaceae, for several reasons. Vegetative similarities throughout the family are so great that sterile specimens may be quite unrecognizable even to one familiar with many species. Floral morphology varies little from one species to another. The most use- ful "recognition characters" (i.e., those possible to use in a key) are those of the inflorescence and those of pubescence. The inflorescence in Eugenia is almost always an unbranched axillary axis bearing one or more opposite and decussate pairs of pedicellate flowers. In any one species the length of this "raceme," the number of pairs of flowers, and the characteristics of the bracts and pedicels, are con- stant within rather narrow limits. Unfortunately for the taxono- mist, almost all the species abandon on occasion their characteristic modes of flowering. Apparently when stimulated to vigorous growth almost any axillary branch may develop not into a short leafless and completely fertile raceme, but into a leafy twig bearing normal pe- dicellate flowers at its lower nodes. Often the lowermost one or two nodes bear small deciduous bracts instead of ordinary green leaves. 414 FIELDIANA: BOTANY, VOLUME 29 When the raceme becomes a leafy branch in this way, the lowermost flowers on casual inspection may seem to be borne at leafless nodes, and the upper flowers solitary and pedicellate in the axils of ordinary green leaves. Obviously a key based on the morphology of the in- florescence must be interpreted with care. Pubescence in Eugenia provides a series of characters most use- ful to the taxonomist. Again, however, this is of limited use because the pubescence, if present at all, usually disappears soon after the plants flower, at least before the fruit matures. It would be quite feasible to construct a key to flowering material, based largely on characters of the pubescence; fruiting material of the same species would be quite indeterminable by the use of the same key. The following synopsis thus does not pretend to present a me- chanically perfect key to all the species involved. The plan is the same as that of the key in the Myrtaceae of Guatemala; those inter- ested in the exceptions and qualifications of such a key may consult that work. KEY TO THE SPECIES Species that are noticeably appressed-pubescent or tomentose (at least at flower- ing time), and have at the same time relatively slender and elongate racemes (the pairs of flowers well separated). Bracteoles thin and conspicuous, mostly glabrous or nearly so and contrasting with the heavily pubescent hypanthium, persistent in flower and fruit, 0.7- 1.5 mm. long, usually connate along the proximal margins, involucre-like; pubescence copious, appressed, of pale sordid simple hairs. Bracteoles acute or acuminate, broadly ovate-deltoid, 1-1. 5 mm. long; bracts thin, persistent, the upper ones often much the longest and 1-2 mm. long; raceme-axis usually more than 1 cm. long, with well-separated nodes; lateral veins 10-20 on each side of the mid vein [group of Eugenia biflora] E. karunnskyana. Bracteoles obtusely pointed or rounded, less than 1 mm. long; bracts ovate- deltoid, somewhat indurated at base, less than 1 mm. long; axis of the raceme rarely as much as 1 cm. long, the nodes approximate; lateral veins 4-6 (-8) on each side E. argyrea. Bracteoles not conspicuous, usually pubescent like the hypanthium and the pedicel, persistent (and then somewhat thickened at base), or deciduous at an thesis, not connate and involucre-like even if persistent; pubescence various, the hairs often reddish and often dibrachiate. Leaves narrowly lanceolate or oblanceolate, 1 cm. wide or less, (3-) 5-8 times as long as wide; hairs reddish, appressed, dibrachiate; region of Acapulco, Guerrero E. avicenniae. Leaves broader, mostly 2-4 (-7) cm. wide, 1.5-3 times as long as wide. Herbage and inflorescence densely hirsute-tomentose with dark reddish- brown erect hairs about 0.5 mm. long, these at flowering time quite obscuring the surfaces; raceme 1-1.5 cm. long; disk 2.5 mm. wide; blades 10-14 cm. long; petioles 10-15 mm. long; Veracruz. E. praeterita. McVAUGH: TROPICAL AMERICAN MYRTACEAE 415 Herbage and inflorescence paler and often sparingly hairy, the hairs cop- pery red, yellowish red, yellowish white or silvery, often closely appressed or loosely contorted and somewhat matted. Flowers small; disk in flower mostly 1.3-2 mm. wide; stamens 50-75; larger calyx-lobes 1.5-2.5 mm. wide; leaves broadly elliptic to ovate or obovate, thinly pubescent with pale appressed hairs, 1.6-2 (-2.5) times as long as wide; petioles 4-6 mm. long; raceme-axis usually less than 5 mm. long; Pacific lowlands of Mexico, to Guatemala. E. rekoi. Flowers larger; disk in flower 2-3 (-5) mm. wide; stamens 100-200; larger calyx-lobes up to 2.5-4 mm. wide and long; raceme-axis 1-2.5 (-8) cm. long, seldom shorter. Bracteoles deciduous at anthesis or before, 2.5-5 mm. long; racemes mostly 2-4 cm. long; pubescence often of loosely matted and contorted hairs, yellowish white to pale reddish brown. Leaf-margins cuneately decurrent on the petiole, the blades often appearing acute at base; leaves soon glabrate, usually glabrous beneath or nearly so, even at flowering time; hairs of the inflores- cence sparse, appressed E. salamensis var. rensoniana. Leaf-margins abruptly contracted to the petiole, the base of the blade rounded or subauriculate, often asymmetric; leaves densely and persistently pubescent or tomentulose beneath; hairs of the inflorescence abundant, stalked, forming a loose tomentum E. salamensis var. salamensis. Bracteoles persistent through anthesis and usually much longer; axis of the raceme usually less than 2 cm. long; pubescence closely appressed, silvery to pale reddish or yellowish. Leaves permanently silvery-scurfy beneath, obovate or obovate- elliptic, rounded at apex; mid vein flat or convex above; axis of raceme usually less than 1 cm. long E. hypargyrea. Leaves at maturity glabrous or with fine scattered pale hairs beneath (when young sometimes with a silvery coating), mostly elliptic with acuminate tip; mid vein impressed or channeled above; axis of the raceme usually 1-2 cm. long. Leaves broadly elliptic, 1.6-2 times as long as wide, 2.5-5.5 cm. wide, abruptly short-acuminate; style 7-8 mm. long; larger calyx-lobes 3.5 mm. long and wide; fruit with 8 prominent wing-angles, ellipsoid, 1 cm. in diameter, 2 cm. long; Jalisco and Nayarit E. pleurocarpa. Leaves narrowly elliptic, (2-) 2.3-3 times as long as wide, 1.5- 2.5 cm. wide, gradually acuminate; style 6-7 mm. long; larger calyx-lobes 2.5 mm. long and wide; fruit smooth, not winged, oblong or obovoid, 1 cm. long or a little more; state of Mexico and Oaxaca; to Central America. .E. guatemalensis. Species either glabrous or nearly so at flowering time, or the hairs small and up- standing, or if the plants heavily pubescent or tomentose then the racemes short (usually less than 1 cm. long) and the nodes approximate. Racemes slender and elongate, 1-2.5 (-4) cm. long; plants glabrous or nearly so at maturity. Plants glabrous; midvein (when dry) elevated on the upper leaf-surface; Colipa, Veracruz, known from the type only E. colipensis. Branchlets and petioles minutely hispidulous, the tender growth bearing tiny appressed coppery dibrachiate hairs also, a few of the latter persisting 416 FIELDIANA: BOTANY, VOLUME 29 near the base of the lower leaf-surface; mid vein impressed above, spar- ingly hispidulous in the furrow; a widespread species. . . .E. oerstedeana. Racemes short and crowded, if more than 1 cm. long then with numerous approx- imate nodes; flowers often fasciculate or glomerate, sometimes solitary at the lowermost nodes of new leafy branchlets and on the same plant in abbreviated racemes. Cauliflorous species, the flowers mostly long-pedicellate in clusters at usually leafless nodes on old wood, the axis of the raceme almost none, or 1-3 mm. long; coarse plants with relatively large leaves, flowers and fruits (leaves 7-20 cm. long; disk 2.5-6 mm. wide; pedicels (3-) 8-30 mm. long; fruit 1.5-2.5 cm. long. Leaves cordate-auriculate at base. Pedicels very stout, 3 mm. long; petioles 8-10 mm. long; buds 10-11 mm. long, the hypanthium canescent; disk 6-7 mm. wide; San Luis Potosf. E. xilitlensis. Pedicels slender, 8-10 mm. long; petioles 5-7 mm. long; buds 5-6.5 mm. long, the hypanthium glabrous or the base somewhat pubescent; disk 3.5 mm. wide; Veracruz E. trunciflora. Leaves acute to cuneate at base, or somewhat rounded to a cuneately de- current base. Midvein broad and flat or convex above (sometimes with narrow median channel near base); lateral veins not impressed; plants essentially glabrous; disk 3.5-4 mm. wide; larger calyx-lobes 2.5 mm. long, 3-3.5 mm. wide; buds 4.5-6 mm. long; Oaxaca to Guatemala. E. choapamensis. Midvein deeply and narrowly impressed above; lateral veins depressed below the upper surface but the veins and veinlets convex, in a reticu- late pattern; herbage and inflorescence sparingly puberulent with tiny erect hairs; disk 3.5 mm. wide; larger calyx-lobes 6 mm. long, 5 mm. wide; buds 8 mm. long; Chiapas and Tabasco. .E. teapensis. Species not cauliflorous, the inflorescences usually in the axils of new leaves, if occasionally at leafless nodes then the raceme-axis prolonged or the pedicels very short, or both; flowers relatively small, the calyx-lobes at most 3.5-4 mm. long and wide, usually much less; disk at most 3.5 mm. wide, usually much less. Plants abundantly and conspicuously hairy at flowering time, the inflores- cence and usually the branchlets hairy, setose, appressed-pubescent or tomentose, the hairs often 2-branched (the branches equal or very unequal); small erect hairs if present few, or obscured by the other pubescence Key A, page 416. Plants glabrous, puberulent or minutely hispidulous, the hairs if present 0.1-0.3 mm. long, often incurved (bracts and flower-parts sometimes longer-ciliate, or the branchlets with slightly longer hairs, but the leaves then essentially or quite glabrous) ; dibrachiate hairs none except in one species Key B, page 418. KEY A Appressed dibrachiate hairs numerous on the branchlets and pedicels (one branch of the hairs often much longer than the other); bracteoles and calyx-lobes usually pubescent to about the same extent as the hypanthium. Flowers glabrous or nearly so; plants as a whole soon glabrous, the youngest branchlets and foliage bearing small evanescent partly dibrachiate hairs; flowers usually in slender elongate racemes E. oerstedeana. McVAUGH: TROPICAL AMERICAN MYRTACEAE 417 Flowers and pedicels appressed-pubescent, usually densely so; hairs more or less persistent, the plants in age eventually glabrate. Leaves narrowly lanceolate or oblanceolate, 1 cm. wide or less, (3-) 5-8 times as long as wide; region of Acapulco, Guerrero E. avicenniae. Leaves broader, 1.5-3 times as long as wide, usually much more than 1 cm. wide. Species with small, relatively broad, short-petiolate leaves [blades 5-7 cm. long or often less, 1.5-2.5 (-3) times as long as wide; petioles 1-6 mm. long]. Leaves more narrowly elliptic, less than 2 cm. wide, 2-3 times as long as wide; petioles 1-2 mm. long; larger calyx-lobes up to 3 mm. long and wide; Sinaloa only E. inconspicua. Leaves broader, 1.5-4 cm. wide, 1.5-2.5 times as long as wide. Leaves blunt or obscurely acuminate; petiole 3 mm. long; flowers larger (disk 2.5-3 mm. wide; style 5-6 mm. long or more; larger calyx-lobes 2.5-3 mm. wide); Morelos; Mexico; Veracruz. E. yautepecana. Leaves prominently or sometimes obscurely acuminate (on short branches sometimes blunt); petioles 4-6 mm. long; flowers smaller (disk 1.3-2 mm. wide; style 4-4.5 mm. long; larger calyx-lobes 1.5-2 mm. wide); Pacific lowlands, Sinaloa to Oaxaca; Guatemala. E. rekoi. Species with larger leaves and flowers and usually longer racemes (blades 5-9 cm. long; petioles mostly 5-10 mm. long; disk 2-3 mm. wide or more; calyx-lobes 2.5-4 mm. long and wide; raceme mostly 1-2 cm. long) ; ordinarily to be sought in the key above (E. hypargyrea, E. gua- temalensis, E. pleurocarpa) See key above, page 415. Hairs of the branchlets and pedicels not dibrachiate, either simple and antrorsely appressed or spreading-ascending, or sometimes loosely contorted or erect; hypanthium often heavily pubescent and then often contrasting with the more nearly glabrous bracteoles and calyx-lobes. Flowers glomerate, sessile or nearly so, the pedicels and the axes of the racemes usually concealed by the clustered flowers; plants often somewhat hirsutu- lous, at least in the inflorescence. Leaves elliptic-oblong, rounded at apex and usually at base also, 1.6-2 times as long as wide, 3.5-6 (-8?) cm. long, short-petiolate; herbage rather generally short-hirsutulous; disk in fruit 1.6-1.7 mm. wide; bracteoles persistent, not connate, broadly rounded, 1-1.5 mm. wide, 1 mm. long; Sinaloa E. sinaloae. Leaves acute or acuminate, sometimes bluntly acuminate. Herbage at maturity nearly glabrous (or the mid veins pilose), the inflores- cence hirsute with coarse, straight or contorted, coppery hairs up to 0.5-0.8 mm. long; calyx-lobes and the connate bracteoles nearly gla- brous except the margins; leaves 3-7 cm. long, the petioles 5-8 mm. long; Sierra Madre Oriental, Veracruz to Tamaulipas. E. symphoricarpos. Herbage and inflorescence abundantly and similarly pubescent or hirsutu- lous. Petioles 3-4 mm. long; herbage and inflorescence tawny-hirsutulous with erect hairs 0.5-1 mm. long; bracteoles distinct, ovate, acute, 0.5- 1 mm. long; flowers small, the disk 1-1.5 mm. wide; eastern Mexico to Panama E. origanoides. 418 FIELDIANA: BOTANY, VOLUME 29 Petioles 5-12 mm. long; bracteoles 1-2 mm. long, connate at base and forming a broad 2-lobed involucre investing and often concealing the hypanthium; flowers somewhat larger. Young herbage and inflorescence closely appressed-pubescent with pale sordid hairs; leaves abruptly acuminate; mid vein deeply impressed above; calyx-lobes pubescent on both sides; disk 2.5 mm. wide; Oaxaca E, oaxacana. Inflorescence appressed-pubescent, the branchlets, petioles, leaf-mar- gins and midvein (beneath) short-hispidulous with erect hairs; leaves bluntly deltoid-acuminate; midvein shallowly concave in the nearly flat blade; calyx-lobes glabrous within except at tips; disk 1.5 mm. wide; western Michoacan E. alnifolia. Flowers evidently pedicellate, the raceme-axis usually evident also but some- times very short. Flowers solitary in the axils of foliage-leaves, slender-pedicellate; pubescence gray to white; Veracruz E. ledophylla. Flowers in short or elongate bracteate racemes. Bracteoles thin and conspicuous, glabrous or nearly so and contrasting with the canescent hypanthium; flowers evidently slender-pedicellate. Bracteoles acute or acuminate, 1-1.5 mm. long; pubescence gray to white; flowers mostly in racemes 2-4 cm. long [the group of Eugenia biflora}. E. karwinskyana. Bracteoles obtuse or rounded, mostly less than 1 mm. long; flowers mostly in racemes 2-6 mm. long, with several approximate nodes; young herbage coppery-pubescent, the hairs of the inflorescence coppery or yellowish E. argyrea. Bracteoles usually not contrasting with the hypanthium, either glabrous or pubescent, if glabrous the hypanthium also glabrous or nearly so, or the flowers sessile; flowers in congested racemes mostly less than 1 cm. long. Branchlets and petioles setose-hispid with pale sharp erect hairs 0.3- 0.5 mm. long (the inflorescence sparingly so); hypanthium glabrous, after flowering prolonged into a narrow neck 0.5 mm. high, 1 mm. in diameter; longer calyx-lobes deltoid-ovate, 1.5-2.3 mm. long. E. principium. Pubescence of herbage and inflorescence appressed, or in addition with numerous very short erect hairs 0.1-0.3 mm. long. Calyx-lobes 1.5-2 mm. long, membranous, reflexed from the base soon after an thesis; raceme-axis 2-4 mm. long, the pedicels filiform, 8-12 mm. long; bracteoles distinct, scarious; midvein slightly en- graved above in the nearly plane blade; plants nearly glabrous, the inflorescence and young herbage sparingly appressed-pubes- cent; coastal Michoacan E. turneri. Calyx-lobes 1-1.5 mm. long or less, neither evidently membranous nor strongly reflexed; pedicels mostly shorter; bracteoles usually some- what united by the proximal margins, involucre-like; midvein concave or sulcate. [Here are to be found exceptionally pubescent individuals of E. axillaris or E. acapulcensis.] See key below, page 421. KEY B Plants nearly glabrous, the very young herbage with small coppery appressed dibrachiate hairs, a few such hairs persisting on the lower leaf-surface and McVAUGH: TROPICAL AMERICAN MYRTACEAE 419 the mid vein near base; flowers at least partly in racemes 1-2.5 (-4) cm. long; bracteoles broad, membranous, persistent, up to 1.5 mm. long and wide; widespread in Mexico and Central America E. oerstedeana. Dibrachiate hairs none; flowers never in long slender racemes. Flowers large and nearly sessile (disk 3-4 mm. wide; larger calyx-lobes 3.5- 4 mm. wide; style 7-9 mm. long; pedicels 0.5 mm. thick, 1.5 mm. long or less); plants glabrous; Veracruz E. mexicana. Flowers much smaller, or the pedicels slender and 6-10 mm. long. Pedicels evidently slender and elongate, 6-16 mm. long; bracteoles distinct, scarious; calyx-lobes foliaceous or scarious, more or less reflexed in an- thesis; staminal ring bristly; flowers fasciculate, 1-3 pairs from a very short axis. Flowers large (disk 2.5-3.5 mm. wide; larger calyx-lobes up to 3.5 mm. wide); Veracruz to Chiapas; primarily West Indian E. rhombea. Flowers small (disk 1.5 mm. wide; larger calyx-lobes up to 2 mm. wide); western Michoacan E. turneri. Pedicels usually 5 mm. long or less (the lower ones in E. acapulcensis often 5-9 mm. long); calyx-lobes mostly somewhat thickened near base, erect and imbricate. Leaves 1.4-4 cm. long, oblong-linear, 6-8 times as long as wide, 2-4 mm. wide; plants glabrous; Oaxaca E. standleyi. Leaves broader and usually much longer, if only 2-4 cm. long then mostly 2-3 (-4) times as long as wide. Leaves small, 1-2.5 cm. long, 3-10 mm. wide, obovate to ovate or elliptic, rounded or obtusely pointed at apex; midvein slightly concave above, at least near base; herbage and inflorescence minutely hispidulous- puberulent; Tamaulipas to Oaxaca E. liebmannii. Leaves usually larger, generally of an elliptic or a lanceolate type, nar- rowed toward the acuminate or sometimes blunt apex. Leaves narrowly lanceolate or linear-lanceolate, 1 cm. wide or less, 6-8 times as long as wide; Oaxaca, Tabasco E. lindeniana. Leaves mostly 2-3 (-4) times as long as wide. Leaves flat in drying, the midvein not depressed beneath the general level of the upper surface of the blade, the vein itself flat or slightly elevated or convex, sometimes shallowly channeled especially near the base; leaves often opaque and featureless above, the veins scarcely or not at all apparent. Pedicels 3-5 mm. long; fruit about 5 mm. in diameter; hypanthium usually finely puberulent; base of the style glabrous; eastern Mexico to Honduras, little known from western Mexico. E. capuli. Pedicels 1-2.5 (-4) mm. long; fruit about 6-8 mm. in diameter; hypanthium glabrous; disk sparingly hairy about the base of the style; Jalisco, Colima, western Michoacan. E. michoacanensis. Leaves in drying curving down toward the midline, the midvein thus depressed below the general level of the upper surface of the blade, the resulting furrow concave or sometimes very nar- row and (in section) V-shaped. Midvein deeply and narrowly impressed above; leaves acutely narrowed at both base and apex, dark green and lustrous above, with inconspicuous veins; pedicels mostly 3-5 mm. 420 FIELDIANA: BOTANY, VOLUME 29 long, slender; bracteoles connate, involucrate; disk 1.5 mm. wide (2-2.5 mm. wide in fruit); fruit 10-13 mm. in diameter; Sierra Madre Oriental E. xalapensis. Midvein usually depressed in a broad concave channel, neither deeply nor very narrowly impressed; pedicels often shorter and stouter; combined characters never as above. Larger calyx-lobes 2-2.5 mm. wide; disk 2-3 mm. wide; pedicels 1-3 mm. long; fruit globose, 1-1.6 cm. in diameter; moun- tains of south-central Mexico, mostly at elevations of 1000- 2000 meters. Leaves glabrous, lustrous and nearly featureless above, the mid vein glabrous, rather narrowly impressed; disk 2.5 mm. wide in flower, 3 mm. in fruit; Jalisco; state of Mexico E. culminicola. Petioles and midveins of the leaves persistently puberulent, the midvein concave on the upper surface, often appear- ing flat because of the persistent hairs; disk 1.5-2 mm. wide in flower, 2.5 mm. wide in fruit. Herbage and inflorescence hispidulous and somewhat vis- cid-puberulent; fruit unknown; Morelos. .E. mirandae. Herbage puberulent when young, the inflorescence some- times nearly glabrous; fruit globose, 1-1.6 cm. in diam- eter; Puebla; Hidalgo; San Luis Potosf. . .E. pueblana. Larger calyx-lobes 1-2 mm. wide; disk 1-2 mm. wide; pedicels various; fruit globose, oblate or oblong, usually less than 1 cm. in diameter. Pedicels stout and short [0.4-0.6 mm. thick in flower, 1-2 (-3.5) mm. long]; midvein concave or channeled above, the sides of the channel hispidulous with short golden or pale hairs; bracteoles 0.5 mm. long, truncate or obtuse, distinct and persistent but not involucre-like; fruit ellip- soid, 7 mm. in diameter, 9-10 mm. long; state of Mexico to Nayarit E. crenularis. Pedicels, if only 1-2 mm. long, more slender; midvein flat or somewhat elevated or concave, not hispidulous as above; bracteoles usually somewhat united by their proximal margins, involucrate; fruit globose or oblate; lowlands of eastern Mexico, one species extending along the Pacific coast to Sinaloa. Leaves drying yellow-green or yellow; midvein convex on the upper surface of the leaf, or toward the base of the blade flat or even sulcate; hypanthium densely and very minutely puberulent with pale appressed hairs smaller than those of the pedicels, closely invested by and partly covered by the bracteoles; calyx-lobes finely pubescent on both sides; fruit transversely oval, 7- 8 mm. wide in the longer axis, 5-6 mm. long. E. farameoides. Leaves drying brownish green, green, brownish red or red- dish; midvein flat or sulcate above; hypanthium usu- ally glabrous, sometimes finely hispidulous like the pedicels, mostly or wholly exserted beyond the bracte- oles; calyx-lobes mostly quite glabrous; fruit globose. McVAUGH: TROPICAL AMERICAN MYRTACEAE 421 Leaves ovate or rhombic-ovate to elliptic, 3-6.5 cm. long, opaque, the veins scarcely evident above in dried leaves; petioles ventrally flat or concave, dilated into the base of the blade; pedicels in flower 3 mm. long or less; buds 1.5-2.7 mm. long; disk 1 mm. wide; style 3-4.5 mm. long; fruit 7-8 mm. in diameter; plants usually glabrous; West Indies; Yucatan Pen- insula; rare or doubtful from Veracruz to Oaxaca. E. axillaris. Leaves elliptic or obovate, or ovate, 6-12 cm. long, veiny, the lateral and marginal veins pale above in dried leaves; petioles deeply channeled, usually passing abruptly into the blade; midvein often flat and wrinkled near base, often somewhat depressed below the surface of the blade, and most of it in a furrow between two low parallel convex ridges; pedicels (at least the lower ones) 5-13 mm. long; buds 2.5-4 mm. long; disk 1.3-2 mm. wide; style (5-) 7-8 mm. long; fruit 1 cm. or more in diameter; plants often hispid- ulous; Central America; Yucatan Peninsula; Chi- apas; west of the Isthmus of Tehuantepec mostly near the Pacific Coast, north to Sinaloa. E. acapulcensis. Eugenia acapulcensis Steud. Norn. ed. 2. 1: 601. 1840. ?E. carthagenensis Jacq. Sel. Amer. 152, t. 178, f. 53. 1763. ?E. cartha- genensis /3 baruensis Jacq. I.e. 153. ?E. baruensis (Jacq.) Jacq. Coll. 3: 183. 1790. Myrtus maritima HBK. Nov. Gen. & Sp. 6: 146 [folio ed. 116]. 1823. ?E. guadalupensis DC. Prodr. 3: 275. 1828. E. mari- tima (HBK.) DC. I.e. 282, not E. maritima DC. I.e. 271. ?E. bonplan- diana Berg, Linnaea 27: 228. 1856. E. mosquitensis Berg, Linnaea 31: 257. 1862. E. deltoidea Standl. Contr. U. S. Nat. Herb. 23: 1045. 1924. E.antiquae~Ri\ey,Kew Bull. 1927:121. 1927. E. escuintlensis Lundell, Phytologia 2: 4. 1941. E. ovatifolia Lundell, Bull. Torrey Club 69: 396. 1 May 1942. E. bartlettiana Lundell, Contr. Univ. Mich. Herb. 7: 30 May 1942. E. bracteolosa Lundell, I.e. 31. E. campechiana Lundell, I.e. 32. E. sibunensis Lundell, I.e. 34. E. comi- tanensis Lundell, Wrightia 3: 12. 1961. As the above long list of synonyms suggests, the taxonomy of this species-complex is little understood and apparently involved. The members of the complex seem to be most abundant and diverse in northern Central America and adjacent Mexico, but some of them occur in the West Indies and in northern South America. What appears to be a single and rather homogeneous species along the Pacific Coast of Mexico from Guerrero westward, and also in eastern Central America, becomes in Guatemala and the Yucatan Penin- sula rather more variable. In the Yucatan Peninsula it comes into contact with the morphologically similar E. axillaris (Sw.) Willd., 422 FIELDIANA: BOTANY, VOLUME 29 the latter primarily a West Indian species, and it is of course possible that some of the puzzling variability has resulted from past or con- tinuing hybridization. The distinction between the mainland spe- cies (which may for the present be called E. acapulcensis) and E. axillaris is not always easy and is often subjective. The distinctions between the two species are set forth above in the key. My inter- pretation of axillaris is essentially that of Urban in his revision of the West Indian Myrtaceae (Bot. Jahrb. 19: 639-641. 1895). Urban interpreted a Swartzian specimen as representative of the plant that has commonly been passing as E. axillaris in Florida and the West Indies. It is a glabrous plant with relatively thick opaque ovate leaves, the midvein sulcate on the upper surface of the leaf, the flowers small and in very short crowded axillary racemes. On the continent of North America it seems to be of limited distribution; it is frequent in southern Florida, and it occurs on the Yucatan Peninsula and probably at a number of localities in the Atlantic lowlands of Guatemala and Honduras. Elsewhere in Central Amer- ica it is replaced by E. acapulcensis.1 Urban seems to have included in his concept of Eugenia axillaris not only this most widely distributed West Indian type, but also some variants with larger, thinner, more veiny and sometimes ellip- tic or obovate leaves, viz. those that on the mainland would be referred to E. acapulcensis. Throughout the West Indies these larger-leaved plants seem to be in the minority, but they are repre- sented in herbaria by numerous specimens from peripheral areas, e.g. Guadeloupe, central and western Cuba, and the Isle of Pines. On the mainland, as suggested above, acapukensis is more wide- spread and abundant, whereas axillaris is of very limited distribu- tion. The exact relation between the two is as yet not understood, and further taxonomic studies, especially in the West Indies, are much needed. On the North American continent E. acapulcensis occupies a con- siderable range, and is morphologically rather homogeneous over most of this range. The nomenclaturally typical population ranges along the Pacific coast of Mexico at least from Guerrero to Sinaloa, and is known from a large series of specimens. The following char- 1 In the past few decades many hundreds of herbarium specimens from Mexico and Central America have been determined as Eugenia axillaris without much regard for the actual identity of the plants. The name became a handy tag to apply to almost any Eugenia haying small flowers in crowded axillary racemes. Many of these misidentified specimens are now to be referred to E. acapulcensis, but any one of a score of other species may occasionally turn up among the speci- mens labeled axillaris. McVAUGH: TROPICAL AMERICAN MYRTACEAE 423 acterization is based primarily on these, but also applies to most Central American specimens from Honduras to Panama: Branchlets, petioles and inflorescence usually minutely hispidulous (some- times quite glabrous), the raceme sometimes thickly pubescent in addition with small golden hairs; leaves usually elliptic, varying to obovate or rarely to ovate, abruptly short-acuminate with a broad blunt tip, less often slenderly acuminate, or some or all leaves obtuse; blades acute or cuneate at base, or sometimes rounded in the broader leaves on the same plants; axis of the raceme mostly 3-6 mm. long. In the lowlands from Guatemala to Tabasco may be observed certain tendencies toward variation from this general pattern. Gla- brous individuals are relatively more frequent; leaves are seldom obtuse, often slenderly acuminate, more often ovate and rounded at base; relatively long and many-flowered racemes (up to about 15 mm. long) occur more frequently; occasional individuals with small flowers seem otherwise identical with the larger-flowered mem- bers of the same populations. Some of these variations may be normal for the species, some may represent unusual individuals, some may have resulted from hybridization. Some small popula- tions in this geographical area seem reasonably distinct, e.g. Eugenia flavoviridis Lundell; others (including those listed above in synon- ymy) are questionably distinct, but some of them may represent local populations. Of the species recently described from the Guate- mala-Yucatan area, E. bartlettiana (known from the type only) has rather loosely few-flowered racemes but is otherwise not unusual for E. acapulcensis; E. bracteolosa is glabrous, with rather long racemes and narrowly pointed leaves; E. campechiana is not distinctive; E. escuintlensis, known from the type only, was a glabrous plant having the flowers partly in racemes and partly solitary at the bases of new vigorous shoots; E. ovatifolia was based primarily on individ- uals with short petioles and pedicels, but the leaves, in spite of the specific epithet, are not significantly broader than the average for the species; E. sibunensis is glabrous, with slenderly acuminate leaves. I have studied the types of these species in the Herbarium of the University of Michigan. Of the earlier synonyms of Eugenia acapulcensis (listed above), E. deltoidea Standl. is known only from the type (Nelson 2292, US!). This is a glabrous specimen but otherwise very like the type of Myrtus maritima HBK. (E. acapulcensis) ; cf . Field Mus. Neg. 36894. The type localities of the two species are only a short distance apart. Two specimens from farther east on the Pacific slope of Mexico, Seler 1793 (US!) and Seler 2029 (US!), both from Juchitan, Oaxaca, were determined at Berlin by Loesener as Eugenia bonplandiana 424 FIELDIANA: BOTANY, VOLUME 29 Berg. This name is included in the above synonymy of E. acapul- censis because of the assumption that Loesener probably had access to the type of E. bonplandiana when he made the determinations of Seler's specimens. Berg based E. bonplandiana on a specimen col- lected by Humboldt and Bonpland in "Columbia" ("Eugenia lauri- folia Willd. herb. no. 9486"); if Loesener's comparison was a critical one, it may well be that the type of E. bonplandiana was a member of the acapulcensis complex, collected perhaps somewhere in north- ern Colombia along the lower Rio Magdalena, or possibly even at Cartagena, where Humboldt and Bonpland spent three weeks. Also from the vicinity of Cartagena came the specimens described by Jacquin under the names of Eugenia carthagenensis and its vari- ety baruensis. In general terms E. carthagenensis was described as having the leaves ovate, obtusely acuminate, glabrous, short-petio- late, 4 inches long; the common peduncles [i.e. presumably the axis of the raceme] short and many-flowered; the flowers 4 lines in diam- eter; the fruit globose, the size of a pea, red, finally black. The figure published with this description unfortunately shows but a single detached leaf, which could perhaps represent our plant. Urban (Bot. Jahrb. 19: 639. 1895) regarded E. carthagenensis ft baruensis as probably a synonym of E. axillaris, but if he was right in his understanding of Jacquin's concepts I suspect that the specimens from Cartagena were not true axillaris but acapulcensis. It has not been possible at this writing for me to learn the identity of the taxa described by Jacquin. An earlier name that may well prove to be synonymous with E. acapulcensis is E. guadalupensis DC. A photograph of the type, Field Mus. Neg. 33492, shows a broad-leaved plant extraordinarily like those common along the American mainland. The type should be compared with Central American material. Two species described from eastern Central America, E. mosqui- tensis Berg and E. antiquae Riley, seem to represent almost identical populations. The type of E. mosquitensis (Wullschlagel in 1855, from "Pearl Key Lagoon, Mosquito" [Nicaragua]), which I have seen through the courtesy of the authorities at Brussels, consists of a few twigs in young bud. The branchlets and pedicels are hispid- ulous, as commonly in specimens from Panama; the leaves are broadly elliptic, shortly and bluntly acuminate, rounded to a cuneate base. In my judgment this is the common plant of Panama, the same as that named E. antiquae. I have not seen the type of the latter, but judging from the original description it is the same plant. McVAUGH: TROPICAL AMERICAN MYRTACEAE 425 This was referred by Amshoff in the Flora of Panama (Ann. Missouri Bot. Gard. 45: 189. 1958) to E. acapulcensis. One species referred by Amshoff (I.e.) and previously by Standley (Contr. U. S. Nat. Herb. 23: 1046. 1924) to the synonymy of E. aca- pulcensis, proves to be very different. This is Eugenia colipensis Berg, Linnaea 29: 243. 1858. The type (from Colipa, Veracruz, Karwinsky 241, LE!) is an over-mature fruiting specimen, evidently a species close to or identical with E. oerstedeana Berg, but differing from that species in having the leaves impressed-punctate above, and the midvein in the dried leaves elevated above in a narrow ridge. In other respects, including the racemes 3-5 cm. long, the fruit about 1 cm. in diameter, the thin persistent bracteoles, the venation of the leaves, the short pubescence of the branchlets and the minute flat dibrachiate hairs at the base of the lower leaf-surface, it exactly simulates a specimen of the oerstedeana-florida alliance. Apparently nothing like it has since been collected. One species not listed above, Eugenia itzana Lundell, Bull. Torrey Club 69: 395. 1 May 1942, surely belongs with the acapulcensis com- plex, but its status is in doubt. The type (Yucatan, Chichen Itza, Lundell 7589, MICH!) is an immature flowering specimen that super- ficially suggests a small-leaved example of acapulcensis. One para- type, Steer e 1373, is scarcely distinguishable from E. axillaris. All four specimens cited in the protologue of E. itzana are small-flowered for acapulcensis, resembling in this respect E. axillaris. Hardly an independent species, E. itzana seems to fall between acapulcensis and axillaris. The very recently published Eugenia comitanensis, from near Comitan, Chiapas, E. Matuda 5753 (LL, type!), represents in my opinion a glabrous and rather narrow-leaved phase of E. acapulcen- sis. The leaf-margins are somewhat revolute, as they often are in E. acapulcensis, but quite entire. Superficially they appear to be "obscurely crenulate," as noted in the protologue. Eugenia alnifolia McVaugh, sp. nov. Frutex vel arbor parva, ramulis petiolisque, foliorum marginibus et nervo medio (subtus) breve hispidulis; folia elliptico-oblonga, 6-10 cm. longa, 1.7-2.2-plolongi- ora quam latiora, deltoideo-acuminata, petiolis 5-7 mm. longis, nervo medio supra concavo; flores numerosi glomerati sessiles, bracteolis connatis extus pubescenti- bus, hypanthio velutino: calycis lobi extus sericei, 1.5-1.8 mm. lati; discus 1.5 mm. latus; stylus 7 mm. longus; ovarium biloculare, ovulis quoque loculo ca. 5; fructus ignotus. A shrub or small tree, the branchlets and petioles, and the leaf-margins and lower surfaces of the midveins short-hispidulous with hairs 0.1-0.2 (-0.5) mm. 426 FIELDIANA: BOTANY, VOLUME 29 long; inflorescence densely appressed-pubescent; leaves elliptic-oblong, 3-5 cm. wide, 6-10 cm. long, 1.7-2.2 times as long as wide, on shoots relatively longer and sometimes obovate, but on flowering branchlets broad at the middle and about equally rounded to the bluntly deltoid-acuminate tip and the acute base, the margins at the very base cuneately but shortly decurrent on the channeled petiole 5-7 mm. long, 1-1.5 mm. thick; midvein shallowly concave above in the nearly flat blade, somewhat persistently hirsutulous at base, on the lower surface elevated nearly its whole diameter and strongly hirsutulous, the other veins and the leaf surface essentially glabrous; lateral veins 7-10 on each side in addition to some intermediate ones, inconspicuous above, elevated and rather prominent beneath, not forming a well-defined marginal vein but diminishing distally and each arching over irregularly, 2-5 mm. from the margin, to join the next succeed- ing one; leaves dark green and probably lustrous above, paler and dull and finely dark-dotted beneath; inflorescences consisting of dense many-flowered axillary glomerules 1-1.5 cm. in diameter, at leafy or leafless nodes on old wood, the racemes very short, often 2-3 at one node, each bearing 2-4 closely approximate pairs of sessile flowers; bracts persistent, very short; bracteoles membranous, appressed-pubescent without, orbicular-reniform, 1-2 mm. wide and almost as long, narrowed at base and somewhat connate, forming a broad 2-lobed involucre nearly concealing the hypanthium; buds 3.5 mm. long, obovoid, the broadly cup-shaped or hemispheric hypanthium velutinous, 1 mm. wide and almost as high; calyx-lobes in unequal pairs, broadly rounded, closely appressed-silky with- out, glabrous within except near the tips, the outer ones 1 mm. long, 1.5 mm. wide, the inner pair 1.3 mm. long, 1.8 mm. wide; disk about 1.5 mm. wide, hairy around the base of the style; style 7 mm. long or more, glabrous; stamens about 35, about as long as the style, the anthers 0.7-0.8 mm. long; petals white, elliptic-obovate, 3 mm. long, 2.5 mm. wide; ovary bilocular, the ovules about 5 in a compact group in each locule. "Cafecillo" (Hda. Coahuayula, Emrick). Michoacan: Distr. Coalcoman, Aquila, in woods, 15 Jan. 1942 (fl.), G. B. Hinton 16292 (K; MICH, type; NY; UC; US); Hacienda Coahuayula, Nov. 1906 (fl.), G. M. Emrick 74 (F); Hda. California, Mun. Apatzingan, 1200 m., Leavenworth & Hoogstraal 1^7 (F). A species somewhat resembling E. origanoides but differing obvi- ously in the distribution of the foliar pubescence, in the larger flowers, and in the differently pubescent inflorescence. Eugenia argyrea Lundell, Wrightia 3: 10. 1961. This very recently described species was known to its author from the type only. It proves to be widely distributed and appar- ently abundant in Central America; it is in fact represented in her- baria by more specimens than almost any other Central American species. In view of the distinctive appearance of the flowering speci- mens, and its wide distribution, it is indeed odd that neither Berg nor any of the more recent special students of the Myrtaceae seems to have noticed it. Among the small-flowered species of Eugenia bearing the flowers in abbreviated racemes the plant is easily recog- McVAUGH: TROPICAL AMERICAN MYRTACEAE 427 nized by the contrast between the glabrous or nearly glabrous brac- teoles and calyx-lobes, and the tawny-pubescent hypanthium. The characteristic coppery-silky pubescence of the young growth is dis- tinctive but does not long persist. A formal description of this species as I understand it can be found in the Flora of Guatemala (Fieldiana, Bot. 24) and need not be repeated here. The geographical distribution of E. argyrea is indicated by the following list of specimens examined : MEXICO: Veracruz: Santa Lucrecia, C. D. Mell 580 (F, US).— Oaxaca: Chivela, Mell 32 (US), Mell s.n., 10 Dec. 1928 (US) ; 17 km. northwest of La Ventosa, R. M. King 605 (MICH). — Chiapas: Ca- mino Berriozabal a S. Fernando, F. Miranda 5829 (MICH); Oco- zocuautla, Miranda 7454 (MICH); Siltepec, E. Matuda 1609 (F, MICH), Matuda s.n., 8 Aug. 1937 (K, MICH); Cascada, near Sil- tepec, Matuda 5074 (LL, type). GUATEMALA: Alta Verapaz: Between Semococh and Chajamayic, J. A. Steyermark 45725 (F); pantano 2.5 miles west of Cubilgiiitz, Steyermark 44309 (F). — Izabal: Puerto Barrios, Standley 25118 (US). HONDURAS: Atlantida: Standley 53775 (F), 56628 (F, US).— Yoro: J. B. Edwards P-820 (F, US).— Colon: Von Hagen 1334 (F).— Comayagua: Standley 56511 (F, US), J. Valerio R. 2892 (F), Yuncker et al. 5742 (F). NICARAGUA: Chontales: La Libertad, Standley 9028 (F). COSTA RICA: Rodeo, H. Pittier 1642 (US), 821*1 (US); vicinity of San Ramon, A. M. Brenes 5372, 6751, 14977, 20357, 20358 (all F); San Ramon, Austin Smith P2332 (UC). PANAMA: Campana, P. H. Allen 1324 (F, US) ; Canal Zone, Pittier 2567 (F). COLOMBIA: Region of Santa Marta, H. H. Smith 907 (MICH), Smith 1927 (MICH, US). Eugenia avicenniae Standl. Contr. U. S. Nat. Herb. 23: 1043. 1924. Because of the very narrow leaves this plant seems distinctive. It is presumably related to the other west-Mexican species with red- dish and somewhat appressed, dibrachiate hairs, viz. E. rekoi and E. salamensis. The species is still unknown except for a few col- lections from the vicinity of Acapulco, Guerrero, the type locality. 428 FIELDIANA: BOTANY, VOLUME 29 Eugenia axillaris (Sw.) Willd. Sp. PI. 2: 970. 1800. Myrtus axillaris Sw. Prodr. 78. 1788. This species appears to intergrade to some extent with E. aca- pulcensis Steud.; see the discussion under that species. In Mexico E. axillaris is known certainly from the Yucatan Peninsula, but rec- ords from the Mexican mainland should be regarded with suspicion. Eugenia biflora (L.) DC. Prodr. 3: 276. 1828. Myrtus biflora L. Syst. Nat. ed. 10: 1056. 1759. E. fieldingii Berg, Linnaea 29: 242. 1858. As pointed out in the Flora of Peru (pt. 4, no. 2: 684-685), this is one of the most variable species of Eugenia. Widely distributed in South America and the West Indies, it is well known in Panama and perhaps in Costa Rica, but in northern Central America it is scarcely known except in the lands bordering on the Gulf of Hon- duras. The latter region is already known for its strong phytogeo- graphic connections with the West Indies. The commonest form in Central America is one much resembling the West Indian var. virgultosa (Sw.) Krug & Urban, and also much like the prevailing form throughout much of South America, viz. a plant with lanceolate leaves 3-4 times as long as wide, long-acumi- nate and bristle- tipped, impressed-punctate on the upper surface; the style is 5-6 mm. long, the bracts of the raceme narrow and mostly 3-4.5 mm. long, the bracteoles often 1.5 mm. long, the whole inflorescence densely silky, and the flowers only about 6-8 to a ra- ceme. This plant is represented in herbaria by a number of collec- tions from Belize and Stann Creek Districts, British Honduras, and from the Department of Izabal, Guatemala. Apparently more widely distributed on the North American mainland than the above variety of E. biflora are several so-called species that are certainly closely related to one another and to E. bi- flora. A comprehensive revision of the whole species complex may indicate that the "species" of Mexico and Central America are not distinct from the inclusive biflora. All are alike in having abundant pale appressed silky or silky-strigose pubescence of simple hairs, the inflorescence an elongate raceme with compressed internodes, usu- ally 3-6 pairs of flowers, the bracts persistent and the distal ones increasingly longer, the bracteoles somewhat coherent by their basal margins, closely investing the hypanthium, scarious and often thinly strigose and so contrasting with the canescent hypanthium, the calyx- McVAUGH: TROPICAL AMERICAN MYRTACEAE 429 lobes broad, and silky or silky-strigose within. The distinctions be- tween the species appear to depend chiefly on leaf-shape, and on the presence and abundance of glandular depressions on the upper leaf-surface. Distinctions between E. karwinskyana and E. yuca- tanensis seem to be particularly weak; the few apparent differences in flower-size and in pubescence have been studied in relatively few specimens, and are not very suggestive. The following key contrasts E. biflora (sensu stricto) with the closely related "species" of the mainland: Leaves lanceolate, 3-4 times as long as wide, bristle-tipped, impressed-punctate above, the glandular depressions numerous and conspicuous E. biflora. Leaves elliptic to ovate or broadly lanceolate, not bristle-tipped (acuminate, but the tip obtuse or sometimes subacute), not at all or sparingly and obscurely impressed-punctate above. Leaves mostly 3-4 times as long as wide; eastern Mexico to Guatemala. E. karwinskyana. Leaves mostly 1.7-2.5 times as long as wide; Yucatan Peninsula. E. yucatanensis. Confined as far as known to the Yucatan Peninsula, E. yucatan- ensis Standl. seems to be frequent in the lowlands from the State of Yucatan to Pete"n. It is scarcely distinguishable from some Ja- maican material of E. biflora in which the leaves are rather broadly elliptic, obtusely acuminate, 1.7-2.5 times as long as wide, not at all or scarcely impressed-punctate above. The plants are somewhat less prominently silky than those of E. biflora, and the flowers appar- ently a little smaller. A third "species," ranging from San Luis Potosi, Mexico, to northern Guatemala, is Eugenia karwinskyana Berg, Linnaea 29: 244. 1858 (E. tabascensis Lundell, Phytologia 1: 482. 1941). This plant has ovate or elliptic leaves narrower than those of E. yucatan- ensis, obtusely acuminate, 2.6-4 times as long as wide, sometimes impressed-punctate above. The plants are more prominently pubes- cent than those of E. yucatanensis, sometimes indeed approaching E. biflora in this respect. The flowers are usually 8-10, or sometimes as many as 18, in a raceme. The type of E. karwinskyana Berg, which I have seen through the courtesy of the authorities at Leningrad, is a flowering specimen in which the inflorescence is notably canescent; the type of E. tabas- censis Lundell is less strongly pubescent and has somewhat narrower leaves but is surely conspecific. The following are to be referred to E. karwinskyana as now understood : 430 FIELDIANA: BOTANY, VOLUME 29 MEXICO: San Luis Potosi: Tamazunchale, Lundell 7113 (MICH). —Hidalgo: Km. 339 between Chapulhuacan and Tamazunchale, Lundell 12400 (MICH); Huejutla, Karwinsky 242 (LE, type).— Tabasco: Boca Cerro, Tenosique, Matuda 3561 (MICH, type of E. tabascensis) . GUATEMALA: Alta Verapaz: Region of Chelae, northeast of Car- cha, Standley 70388 (F). Another member of the species complex that includes E. biflora is apparently the following: Eugenia ledophylla (Standl.) McVaugh, comb. nov. Myrtus ledophylla Standl. Contr. U. S. Nat. Herb. 23: 1038. 1924. The type, Purpus 7804 (US !) is a silky-pubescent plant in which the leaves are small for biflora, up to about 1.5 cm. wide, 5 cm. long, lanceolate or lance-ovate, cuspidate, impressed-punctate above. The pubescence throughout is like that of biflora, as are the bracteoles subtending the buds. The type bears nearly mature fruit; another specimen from Zacuapan, Veracruz, presumably a topotype, Purpus 8053 (UC), is in bud. The flowers in both collections are borne on slender pedicels 1-1.5 cm. long, solitary in the axils of foliage leaves on new tender branchlets. It is conceivable that these specimens came from a plant or plants abnormally stimulated so as to cause excessive development of the axillary branches that normally be- come short, leafless, bracteate racemes in the species of Eugenia related to E. biflora. Whatever may prove to be the relationship of this plant, it is quite out of place in Myrtus. A synonym of E. biflora is Eugenia fieldingii Berg ["in Mexico (Fielding)"], the type of which proves to be an average specimen of E. biflora. The racemes are young but otherwise characteristic, the leaves are silky and impressed-punctate above and often cuspidate. The source of the specimen, however, can hardly be Mexico. The Fielding cited as collector was presumably Henry Barren Fielding (1805-1851; cf. Britten & Boulger, Biog. Ind. Brit. Bot. ed. 2. 107. 1931), who amassed a large private herbarium but who, as far as I can learn, did not himself collect in tropical America. In 1858 Berg published three new taxa based on American col- lections attributed to "Fielding." Only one of the specimens, the type of Myrtus acerosa (Linnaea 29: 253) was cited by number (Fielding 1448)- Berg saw these specimens "in herb. hort. bot. Pe- trop." Through the kindness of the authorities at Leningrad I have been able to study all three specimens. McVAUGH: TROPICAL AMERICAN MYRTACEAE 431 The types of E. fieldingii and Myrtus acerosa, and that of Myrcia saxicola /3 grandifolia (Linnaea 29: 219), all seem to have been col- lected by the same person. Each bears a small collector's field-label, slit to pass over the end of the twig; each bears on this label in a large smooth hand in faded black ink a short notation evidently by the collector, and in addition a serial number in another hand (evi- dently added in different ink); also the words "Hb. Fischer" in a small angular hand, and the word "Fielding" in what appears to be still another hand. Berg has added his annotation in one instance to this label, and in the other two on a separate label. The pertinent data are as follows: Name Serial No. Collector's Notations Myrtus acerosa 1448 636 Myrtaceae Bajasan Eugenia fieldingii 1449 Indana fruit eatable Myrcia saxicola /3 grandifolia 1958 205 Casapi Myrtacea From the data on No. 1449 the collector apparently was English- speaking. From the handwriting in which the field-labels are written I think there can be little doubt that these are collections of Andrew Mathews, whose principal Peruvian collections came to Kew in the 1830's. His Myrtaceae were arranged in systematic sequence, ac- cording to the then-current practice, and assigned serial numbers replacing the original field-numbers. The serial numbers assigned the Myrtaceae included at least 1447-1454, and in another series 1957 and 1958. No. 1449 was cited under Eugenia biflora in the Flora of Peru (p. 685), where the locality was given as Tarapoto. I assume that "Indana" (if I have interpreted correctly the writing on the label) is a locality near Tarapoto, and that the type of E. fieldingii is from eastern Peru. It is certainly not Mexican. No. 1958 was cited under Myrcia splendens (Sw.) DC. in the Flora of Peru (p. 661), where the locality was given as Cassapi. Pre- sumably the type of M. saxicola /3 grandifolia is a duplicate of this collection. The specimen belongs to one of the compact-leaved types of M. splendens. The type of Myrtus acerosa [Myrteola acerosa (Berg) Burret], was cited in the Flora of Peru (p. 806) as Fielding 1448. I have not seen a sheet of this number definitely attributed to Mathews, but he traveled in this part of Peru and there seems no reason to doubt that he collected this particular specimen. The few buds on the type are immature; some are seemingly 4-merous, not 5-merous as 432 FIELDIANA: BOTANY, VOLUME 29 Berg described them, and others are 5-merous. Vegetatively the plant much resembles Ecuadorean specimens of Myrteola micro- phylla var. microphylla, but as noted in the Flora of Peru the branch- lets in M. acerosa are spreading-setose and the very narrow leaves are revolute-sagittate when dry; in M. microphylla the branchlets are canescent and the leaves revolute-linear or -lanceolate. Fielding's herbarium, the principal part of which went to Oxford University, is said to have contained 2000 sheets of Mathews' Peru- vian plants, viz. the set reserved by the collector for himself (cf. Hook. Journ. Bot. & Kew Misc. 6: 283. 1854). Apparently some of these specimens, probably sent out as duplicates, found their way to St. Petersburg where Berg saw them and assumed that Fielding was the original collector. Eugenia capuli (Schlecht. & Cham.) Berg, Linnaea 27: 238. 1856. Myrtus capuli Schlecht. & Cham. Linnaea 5: 561. 1830. Eugenia schiedeana Schlecht. Linnaea 13: 418. 1839. E. contrerasii Lundell, Wrightia 2: 206. 1961. In his revision of West Indian Myrtaceae, Urban referred the Mexican E. capuli to the synonymy of E. monticola (Sw.) DC., var. latifolia (DC.) Krug & Urb. (Bot. Jahrb. 19: 636. 1895). In this course he has not been followed, as far as I can learn, by any bota- nist working with the floras of continental North America. Probably, however, E. capuli and E. monticola actually are conspecific; some Central American specimens are quite indistinguishable from indi- vidual West Indian ones. On the other hand most Mexican and Central American plants differ from typical monticola in having the leaves wider and longer (but neither so long nor so wide as described for var. latifolia!); the petioles longer; the blades relatively longer on the average because of the frequent occurrence of forms with caudate-acuminate tips; the pedicels longer; but the flowers in gen- eral smaller and the fruits smaller. Aside from these quantitative differences I cannot distinguish between West Indian and Central American specimens. Pubescence in both populations is variable; plants from the mainland are usually but not always less conspicu- ously puberulent than West Indian specimens, which may indeed be very obscurely puberulent or quite glabrous. After examination of a large series of specimens from both eastern and western Mexico, from the Yucatan Peninsula and from the adjacent lowlands in Central America, I cannot find any convincing basis for recognizing more than a single species in the population. McVAUGH: TROPICAL AMERICAN MYRTACEAE 433 From the standpoint of practical taxonomy it seems preferable to recognize E. capuli as an independent species at least until the relationships of the several West Indian populations of E. monticola are better understood. There is much as yet unexplained variation among these insular populations, whereas E. capuli is abundant and relatively homogeneous over a considerable lowland area from Ta- maulipas to northern Central America. The species complex that includes E. monticola and E. capuli has also several representatives in northern South America and in eastern Central America. A revision of the entire complex is much to be desired. The principal species that are involved are as follows: Eugenia micrantha (HBK.) DC. Prodr. 3: 282. 1828. Myrtus micrantha HBK. Nov. Gen. & Sp. 6: 144 [folio ed. 115]. 1823. Type locality, Honda, Colombia. Distinguished from West Indian monticola chiefly by the longer racemes, the axis of which may be 1 cm. long or more in micrantha, usually 3-6 mm. long or less in monticola. The pubescence in mi- crantha is short and inconspicuous. The known range extends from about the vicinity of Caracas, Venezuela, to the valley of the Rio Magdalena in Colombia. Eugenia casearioides (HBK.) DC. Prodr. 3: 275. 1828. Myrtus casearioides HBK. Nov. Gen. & Sp. 6: 145 [folio ed. 115]. 1823. Type locality, "in monte Cocollar Cumanensium," i.e. on the Orinoco. Distinctive chiefly in having the branchlets rather long-pubes- cent with spreading hairs up to 0.5-0.7 mm. long. The pedicels are short and the flowers glomerate, i.e. the axis of the raceme is very short. The species is known chiefly from northern Venezuela, and from one locality in Colombia (Norte de Santander). Eugenia pittieri Standl. Field Mus. Bot. 8: 145. 1930. Type locality, Panama. Specimens from Costa Rica and Panama, identified as E. pittieri, are essentially indistinguishable from E. monticola of Tobago, e.g. Broadway s.n., 19 Sep. 1910 (US). The branchlets are setose with coarse, antrorsely curved, sordid hairs 0.3-0.5 mm. long, these occur- ring also on the mid veins and margins of the leaves; the hypanthium 434 FIELDIANA: BOTANY, VOLUME 29 and the pedicels are nearly glabrous. The type, Pittier 44-84 (F), is in young bud but otherwise not different from other specimens (F and US). Eugenia chepensis Standl. Field Mus. Bot. 8:144. 1930. Type locality, Panama. The type of this species, Pittier 4716 (F) was the only specimen cited in the protologue, where the pedicels were described as "2- 3.5 mm. longis," and a little later as "comparatively long." Actually the pedicels are not long when compared with those of E. capuli or even with those of E. monticola; Standley was comparing them with those of E. pittieri, whose flowers he described as sessile. Amshoff, in the Flora of Panama (Ann. Missouri Bot. Card. 45: 185. 1958) describes the pedicels of chepensis as 4-10 mm. long. It is doubtful that chepensis is distinct from pittieri, or that either one is distinct from monticola. The monticola complex seems to be represented on the Isthmus of Tehuantepec rather sparingly, but interestingly enough two addi- tional species have been described from this area, both on the basis of their unusually narrow leaves. One of these is Eugenia lindeniana Berg, Linnaea 29: 240. 1858. The type collection, from Teapa, Tabasco, Linden 619 (LE, type; MICH), has narrowly lanceolate leaves 11 mm. wide or less, 6-8 times as long as wide; except for this one feature the specimen seems to be an average moderately hispidulous flowering specimen of E. capuli. Another specimen that presumably should be referred to E. lindeniana is: Tabasco: Boca Cerro, Tenosique, E. Matuda 3592 (MICH). The recently published Eugenia tenuissima Lundell (Wrightia 3: 19. 1961) was based entirely on Martinez Calderon 546, from Chilte- pec, District of Tuxtepec, Oaxaca. I have seen isotypes (MEXU and UC). The specimens bear nearly mature fruit; a great deal of the pubescence has been eroded away, but seems to be characteristic of E. capuli as I understand that species. The leaves are unusually narrow for E. capuli, 3.2-4.5 times as long as wide, and up to no more than 1.4 cm. wide. In other respects the specimens are quite typical of E. capuli; the type was compared in the protologue of E. tenuissima with E. lindeniana, which does indeed have even nar- rower leaves than tenuissima. The leaves of the type collection of lindeniana, however, although very narrow, are not acutely pointed McVAUGH: TROPICAL AMERICAN MYRTACEAE 435 but are obtuse exactly as in tenuissima. It is not likely that these two narrow-leaved extremes are specifically different from E. capuli. A second specimen, very like the type of E. tenuissima, is: Oaxaca: Upper Coatzacoalcos River, C. D. Mell 601 (F, US). In recent months Dr. C. L. Lundell has performed a signal service by calling attention to what he supposes to be distinct populations among the Eugenias related to E. capuli, specifically those species with puberulent herbage, small elliptic or lanceolate leaves having the midvein flat or convex (not sulcate or impressed) above, the flowers small in short crowded racemes and the fruit small and glo- bose. The fruits in E. capuli are typically small and slender-pedicel- late, 4-6 mm. in diameter, the fruiting pedicels 0.3 mm. thick, and the axis of the raceme rarely more than 4-5 mm. long. In Eugenia tikalana Lundell (Wrightia 2: 210. 1961) the fruits are mostly 9-12 mm. in diameter, the fruiting pedicels 0.4-0.6 mm. thick, and the axis of the longer racemes on most plants ranges from 6-10 (or up to 25) mm. long. Through the courtesy of Dr. Lundell I have seen a fine series of collections of E. tikalana from Pete"n. In E. michoacanensis Lundell (Wrightia 3: 16. 1961) the plants are superficially very similar to those of E. capuli and E. tikalana, but the fruits are intermediate in size (6-8 mm. in diameter). As apparently plants with large, intermediate and small fruits all occur in the same general areas in southwestern Mexico, and as the plants are otherwise so similar, it seems likely that these species are not genetically isolated. These large-fruited forms need further study but apparently represent one or more populations distinct from E. capuli. The small-fruited E. contrerasii Lundell, described in the same publication as E. tikalana, seems amply distinct from that species in Pete"n, where both are common. In the protologue of E. contre- rasii, however, it was compared not with E. tikalana but in a very general way with E. capuli. I do not find any way in which E. con- trerasii differs from E. capuli as that species occurs in eastern Mexico. Eugenia choapamensis Standl. Field Mus. Bot. 22: 93. 1940. The type, Mexia 9254 (F!), was from near San Juanito, Distr. Choapam, Oaxaca. The species is known from several additional collections from Oaxaca to Guatemala. It is treated at length in the Flora of Guatemala. 436 FIELDIANA: BOTANY, VOLUME 29 Oaxaca: Pluma Hidalgo, C. Leyva in 1947 (US); Distr. Tuxtepec, Chiltepec, 20 m., G. Martinez Calderon 525 (US).— Tabasco: El Azufre, Teapa, 25 Mar. 1957 (fr.), F. Miranda 8J>88 (MICH). Eugenia colipensis Berg, Linnaea 29: 243. 1858. Glabrous or essentially so in age, the racemes minutely and sparingly hispid- ulous; leaves elliptic, acute to acuminate at both ends, 3-4 cm. wide, 6-10 cm. long, the petioles 5-8 mm. long; blades impressed-punctate above; midvein ele- vated as a narrow ridge on the upper surface; racemes 2-4 cm. long, each with 4-7 pairs of flowers on straight pedicels 4-6 mm. long; bracteoles persistent, not connate, broadly ovate-deltoid, 1 mm. long; fruit about 1 cm. in diameter, sub- globose, the disk 1.5-2 mm. wide, the calyx-lobes rounded, erect, the larger ones 1.5-1.7 mm. wide. Still known from the type only, this apparently distinctive spe- cies should be sought in the Atlantic lowlands in Veracruz. See the discussion above under E. acapulcensis. Eugenia crenularis Lundell, Wrightia 3: 12. 1961. E. hintonii Lundell, Wrightia 3: 14. 1961. A shrub or small tree 3-4 m. high, the branchlets and petioles uniformly but rather sparsely hispidulous with sharp erect hairs 0.1-0.3 mm. long; axis of inflores- cence, and pedicels, similarly but often very sparsely pubescent or essentially glabrous, the bracteoles coarsely ciliate, the bracts usually strigose especially dis- tally with coarse appressed reddish hairs 0.2-0.4 mm. long; leaves thin, often crenulate, or apparently so because of differentially revolute margins, the blades lanceolate to elliptic or narrowly ovate, 1-3 cm. wide, (2-) 3-7.5 cm. long, (2-) 2.3-3 (-3.7) times as long as wide, often much narrowed toward both ends from the middle or below, the tip blunt-pointed or obscurely or sometimes narrowly acuminate, the base usually less drawn out than the apex, acute or rounded, the margins at the very base cuneately but usually shortly decurrent on the ventrally hispidulous and channeled petiole 3-6 mm. long; midvein concave or channeled above, the sides of the channel more or less densely and persistently hispidulous with short golden or pale hairs; lateral veins 8-10 on each side in addition to some poorly differentiated intermediate ones, somewhat convex on both surfaces and forming with the small veins a reticulum when dry, not forming a well-defined marginal vein but connected by a series of irregularly angled arches 0.7-2.5 mm. from the margins; blades lustrous above, paler beneath; glands usually evident as scattered convexities above, as small dark or resinous dots beneath; inflores- cence an abbreviated axillary raceme, or 2 racemes superposed, the stout axis 1-4.5 (-7) mm. long, bearing 1-3 (-5) opposite and decussate pairs of flowers on rather stout pedicels 1-2 (-3.5) mm. long, 0.4-0.6 mm. thick in flower; terminal flower often present, sessile in the bracts (simulating bracteoles) at the terminal node; bracts persistent, ovate or slightly elongate, 0.5-0.8 mm. long; bracteoles similar to the bracts, about 0.5 mm. long and wide, truncate or obtuse, persistent, distinct; buds rather closely sessile in the bracteoles, short-pyriform, 2-2.5 mm. long; hypanthium (usually) glabrous or sparingly puberulent, campanulate. 1 mm. McVAUGH: TROPICAL AMERICAN MYRTACEAE 437 long; calyx-lobes short and broad, thin and membranous, soon irregularly reflexed, the outer ones 0.5-1 mm. long, 1-1.5 mm. wide, the inner 0.8-1.3 mm. long, 1.5- 2 mm. wide; disk 1.3-1.4 mm. wide, glabrous; style 5-6 mm. long; stamens 50-60, up to 5 mm. long, the anthers 0.5-0.6 mm. long; petals white, obovate, ciliate, 3.5 mm. long; ovary bilocular, the ovules about 10 in a compact group in each locule; fruit ellipsoid, black, about 7 mm. in diameter, 9-10 mm. long, obtuse at both ends, or slightly prolonged distally to the erect or appressed calyx-lobes. State of Mexico: Dist. Temascaltepec: Temascaltepec, 18 Mar. (fl.), Hinton 405 (K); 1 Feb. (imm. fr.), Hinton 3255 (K, US); 5 May (fl.), Hinton 7679 (US); 2 June (fl.), Hinton 7695 (US, iso- type); San Jose", 21 Feb. (fr.), Hinton 291 (K); 7 Mar. (bud), Hin- ton 359 (K). — Guerrero: Dist. Mina: Manchon, 28 Sep. (imm. fr.), Hinton 9603 (MICH, isotype of E. hintonii); 21 Apr. (fl.), Hinton 10076 (US); Tierras Blancas, 6 May (fl.), Hinton 10146 (US). The following specimens appear to belong to the same species, as indicated by the quality and distribution of the pubescence, leaf- shape, texture, venation and crenulation, presence of the terminal flower, general flower-morphology, and fruit shape. They differ, however, in having the racemes longer (often 4-6 mm. long with up to 4 pairs of flowers), the pedicels 1.5-3 mm. long and apparently fleshy (shrinking in drying), the disk up to 1.5 mm. wide, and the hypanthium distinctly and rather closely puberulent. State of Mexico: Dist. Temascaltepec: San Lucas, 7 July (fl. & imm. fr.), Hinton 4294 (K, MICH, US).— Jalisco: Sierra del Halo, 23 June (fl.), McVaugh 15029 (MICH); 14 Aug. (past fl.), McVaugh 16187 (MICH).— Nayarit: 9.5 miles west of Tepic, 11 Sep. (past fl.), McVaugh 18949 (MICH). Eugenia culminicola McVaugh, sp. nov. Frutex subglaber, ramulis petiolisque breve hispidulis, demum glabris; folia ovato-lanceolata, 2.7-4 cm. longa, 2.3-2.5-plo longiora quam latiora, obtusa vel vix acuminata, nervo medio supra glabro, impresso; laminae nitidae, nervis later- alibus inconspicuis; racemi perbreves, floribus subsessilibus, bracteolis distinctis, hypanthio glabro infundibuliforme; calycis lobi rotundati, majores 0.8 mm. longi, 1.7 mm. lati; discus floris 2.5 mm. latus, fructus 3 mm.; stylus 5-6 mm. longus; ovarium biloculare, ovulis quoque loculo 6-8. Shrub 1.5 meters high, the branchlets short-hispidulous, the petioles at first sparingly so; leaves ovate-lanceolate, 1-1.8 cm. wide, 2.7-4 cm. long, 2.3-2.5 times as long as wide, blunt or bluntly acuminate at tip, acute at base, the margins more or less cuneately decurrent on the channeled petiole 3-4 mm. long; midvein nar- rowly impressed above; lateral veins 4-6 on each side, scarcely apparent in dried leaves; blades lustrous and nearly featureless above, probably paler and dull be- neath, and there obscurely gland-dotted; inflorescence a much-abbreviated axillary raceme or 2 racemes superposed, the axis up to 2.5 mm. long, bearing up to 6 438 FIELDIANA: BOTANY, VOLUME 29 closely approximate opposite decussate pairs of flowers on pedicels 1 mm. long or less (up to 2.5 mm. long in fruit); bracts indurate-persistent, ovate-deltoid, divari- cate, 0.8 mm. long; bracteoles suborbicular-deltoid, ciliate, persistent, distinct, 0.5 mm. long and 0.7 mm. wide; buds 3.5-4 mm. long, turbinate, the glabrous funnelform hypanthium 1.5-2 mm. long; calyx-lobes rounded, unequal, the smaller ones 0.4 mm. long, 1.2 mm. wide, the larger pair 0.8 mm. long and 1.7 mm. wide; disk about 2.5 mm. wide in flower (3 mm. in fruit); style 5-6 mm. long; stamens about 75; ovary bilocular, the ovules 6-8 in a compact group in each locule; fruit purple-black, ellipsoid or obovoid, substipitate, 8-10 mm. in diameter, 10-12 mm. long, crowned by the calyx-lobes up to 1.5 mm. long and 2.5 mm. wide. State of Mexico: Dist. Temascaltepec, Cumbre-Pena Blanca, Alnus woods, 1 Mar. 1936 ["3-1-36"] (bud), G. B. Hinton 8957 p.p. (TEX); in fruit, Hinton 8957 (US, TEX p.p.); in flower, with few detached fruits, Hinton 8957 (GH, type). The fruiting specimens associated with the budding plants of Hinton 8957 may conceivably belong to another species and are probably from another plant; the leaves are more veiny and the pedicels are longer (up to 8 mm.). A common and often very abundant treelike shrub in several localities in high barrancas in Jalisco, in forests of pine or fir or with various deciduous trees, is apparently the same species but is known only in fruit. Fruiting plants are quite glabrous; the dark green and lustrous leaves are up to 9.5 cm. long, on petioles 5-8 mm. long; the fruits are borne in clusters, 1-3 pairs on an axis 2-4 mm. long, the pedicels 1 mm. thick or more, 1-3 mm. long, the fruits themselves subglobose, turning red and then black, 8-15 mm. in diameter; the disk is about 3 mm. wide, the calyx-lobes persistent but inconspic- uous on the fruit, broad and low, about 1 mm. long, 2 mm. wide. Jalisco: Sierra de Cuale, southwest of Talpa de Allende, barrancas in fir zone, 1800-2250 m., 19-21 Nov. 1952, McVaugh 14301 (MICH); Sierra de Manantlan, 15-20 mi. southeast of Autlan, cloud-forests ca. 2000 m., 6 Nov. 1952, McVaugh 13944 (MICH); ca. 40 km. west of Ayutla, fir forest, elev. 2100-2200 m., 29 Nov. 1960, McVaugh 21559 (MICH). This species apparently has much in common with E. pueblana Lundell, which apparently occupies a similar ecological niche in the mountains to the east and northeast of E. culminicola. Neither spe- cies is adequately understood. The flowers of E. pueblana are smaller in several respects than those of E. culminicola, and the plants are more generally and persistently puberulent. The mid- vein in E. culminicola is somewhat more deeply impressed and is consistently glabrous. McVAUGH: TROPICAL AMERICAN MYRTACEAE 439 Eugenia farameoides A. Rich. Ess. Fl. Cub. 588. 1845. This plant seems not to have been reported in the literature ex- cept from Cuba, but it is well represented in herbaria by collections from the Yucatan Peninsula, where it has been passing under the name of Eugenia flavifolia Standl. Full description and synonymy are given in the Flora of Guatemala. The plant is also known from Veracruz and from the lowlands of Oaxaca. Dried specimens are easily recognized because of their yellow-green color suggestive of some species of Symplocos. Veracruz: Between Acayucan and Minatitlan, 22 Sep. 1957 (fl.), F. Miranda 85^6 (MICH). — Oaxaca: Cerro Campana, 400 m., 12 July 1926 (fl.), E. Makrinius 619 (US). Eugenia guatemalensis Donn. Sm. Bot. Gaz. 23: 245. 1897. Description and full synonymy are given in the Flora of Guate- mala. This species is very similar on the one hand to E. rekoi of the Pacific slope of Mexico, and on the other to the species with wing- angled fruit and similarly pale rufous-pubescent branchlets and ra- cemes, E. pleurocarpa of western Mexico and E. octopleura of Cen- tral America and the West Indies. Additional material, especially fruiting material that can be matched with flowering specimens of known origin, is much needed to clarify the relationships between these species. A few specimens from Mexico are referred without much doubt to E. guatemalensis: State of Mexico: Dist. Temascaltepec, 11 Nov. 1933 (bud), Hinton 3368 (A). — Oaxaca: Mountains, Joyacatlan, ca. 1800 m., 18 June 1894 (fl.), L. C. Smith 29 (GH).— Chiapas?: "Chiapas, etc.," Sep. (fr.), Ghiesbreght 893 (GH). Eugenia hypargyrea Standl. Contr. U. S. Nat. Herb. 23: 1044. 1924. This plant belongs to the complex of species that includes E. gua- temalensis and E. rekoi, but seemingly is distinct by virtue of its very fine silvery rather than reddish indumentum, and by its rela- tively large obtuse obovate leaves. It is known from but few col- lections; the type was from Zacuapan, Veracruz, Purpus 6171 (US!) and a recent collection is from Chiapas. Description and full synon- ymy are given in the Flora of Guatemala. 440 FIELDIANA: BOTANY, VOLUME 29 Chiapas: Bajada del Puerto de la Sepultura, carretera al N. de Arriaga, 600 m., 23 Aug. 1951 (fl.), F. Miranda 7262 (MICH). Eugenia inconspicua Standl. Contr. U. S. Nat. Herb. 23: 1043. 1924. "Compact upright shrub 8 feet high" [Palmer], the foliage nearly glabrous [from the first?], the branchlets and inflorescence [and the youngest leaves?] densely appressed-pubescent with pale dibrachiate hairs mostly 0.2-0.4 mm. long, the whole plant ultimately glabrate; leaves elliptic to elliptic-obovate, 1-1.8 cm. wide, 3-4.5 cm. long, 2-3 times as long as wide, narrowed about equally to an obtusely pointed or obscurely and bluntly acuminate tip, and to an acute base, the margins acutely or cuneately decurrent on the channeled petiole 1-2 mm. long; mid vein concave above at least near base, or shallowly channeled between two obscure parallel ridges; lateral veins 5-6 on each side, ascending, pale and readily visible on both surfaces of dried leaves, scarcely forming a marginal vein but arching over 0.5-1 mm. from the margin in a series of irregular angled arches to join the next succeeding ones; blades darker and very minutely (or obscurely) pale-verruculose above, paler beneath, with scattered small dark resinous glands apparent on both surfaces; inflorescence a much-abbreviated axillary raceme, the axis up to 2 mm. long, bearing 1-2 opposite decussate pairs of flowers on slender pedicels, these in fruit up to 4.5 mm. long, 0.6 mm. thick; bracts ovate, sub- persistent, 1 mm. long or less; bracteoles oblong, divaricate, distinct, persistent, up to 1.3 mm. long; calyx-lobes broadly rounded, concave, sparingly beset with large resinous convex glands 0.2 mm. in diameter, appressed-pubescent on both surfaces, but persistently and more densely on the inner; larger lobes up to 3 mm. long and wide; disk in fruit 2.5-3 mm. wide; immature fruit obovoid, 6 mm. in diameter, 8 mm. long, topped by the erect concave calyx-lobes. Sinaloa: Culiacan, Palmer 1786 in 1891 (US, type!). Still unknown except in the vicinity of Culiacan, this species is strikingly close to E. yautepecana, with which it was compared by Lundell in the protologue of the latter. The above description of E. inconspicua, and the one of E. yautepecana (q.v., infra), were pre- pared with awareness of the similarities between the species. It is probable that yautepecana and inconspicua are no more than sub- specifically different, but flowering specimens of the latter, and in- deed additional material of both species, are needed to resolve this question. Eugenia karwinskyana Berg, Linnaea 29: 244. 1858. Description and full synonymy are given in the Flora of Guate- mala. For a discussion of the taxonomic position of E. karwinskyana, see above under E. biflora. Eugenia ledophylla (Standl.) McVaugh. See above under E. biflora. McVAUGH: TROPICAL AMERICAN MYRTACEAE 441 Eugenia liebmannii Standl. Contr. U. S. Nat. Herb. 23: 1044. 1924. A shrub 5 meters high, persistently but minutely hispidulous-puberulent, the branchlets, petioles and inflorescence thickly beset with sharp erect hairs 0.1- 0.2 mm. long; leaves soon glabrate, at first sparingly strigose like the very young- est branchlets with coarse coppery hairs; leaves elliptic to obovate or ovate, 3- 10 mm. wide, 10-25 mm. long, 1.6-3 (-4) times as long as wide, rounded or obtusely pointed at apex, gradually rounded to a narrow base, the pale thick revolute mar- gins rather abruptly decurrent on the channeled petiole 1-1.5 mm. long, 0.5 mm. thick; blades sometimes apparently acute at base because of the inrolled margins; midvein slightly concave above at least near base; lateral veins inconspicuous, sometimes scarcely apparent above, 3-4 pairs, the lowest 1 (or 2) strongly ascend- ing, nearly parallel to the margin and 1-1.5 mm. within it, the distal veins diverg- ing at a wide angle from the midvein and joining the basal (marginal) one at or above the middle of the blade; blades smooth above or with a few convex glands, dull and paler beneath and there rather abundantly dark-dotted; inflorescence an abbreviated axillary raceme, or 2 racemes superposed, the axis 2 mm. long or less, bearing 1-4 opposite and decussate pairs of flowers on pedicels 1.5-3 mm. long; flowers sometimes solitary and opposite at the lowermost (bracteate) nodes of new branchlets; bracts dark, scarious, 0.5 mm. long; bracteoles ovate, obtuse, distinct but persistent, 0.5-0.6 mm. long; buds 2-2.5 mm. long, the short-campan- ulate hispidulous hypanthium 0.5-0.8 mm. long, abruptly expanded into the globe of the petals; calyx-lobes broadly rounded, nearly flat, fimbriate-ciliate but other- wise nearly glabrous, about 1 mm. long and wide; disk 1-1.5 mm. wide; style 4- 4.5 mm. long; stamens about 40, up to 4 mm. long, the anthers 0.5 mm. long and about as wide; petals white, prominently gland-dotted, orbicular-obovate, 3 mm. long; ovary bilocular, the ovules 10-11 in a compact group near the center of the dissepiment in each locule; fruit glandular- verru cose, globose or transversely oval, 4-5 mm. in diameter or up to 6 mm. wide, 4 mm. broad and high. — U. of Mich. Neg. 1415. Tamaulipas: 9 mi. south of C. Mante, 31 Aug. 1948 (fr.), Kenoyer & Crum 3678 (MICH); area of Sierra de Tamaulipas, between "La Chona" and Rio Sta. Olaya, elev. 50 m., 26 Sep. 1956 (fr.), F. Mar- tinez M. & G. Borja L. F-2150 (TEX).— Veracruz: Near Ebano along Rio Panuco, H. LeSueur 55-4 (TEX). — San Luis Potosi: Villa Juarez, lowland thickets, July 1937 (fl.), Lundell & Lundell 7270 (MICH).— Oaxaca: Villa Alta, Aug. 1842 (fr.), Liebmann 3969 (US, type). Plants of this species are superficially similar to very small-leaved examples of E, buxifolia; the two species may indeed be closely re- lated. Their geographical ranges, however, are quite distinct. Mor- phologically the two may be distinguished by the differences in leaf-size, by the considerably more veiny leaves of E. buxifolia and the more abundant and larger hairs; and by the flowers, which are consistently smaller in E. buxifolia. 442 FIELDIANA: BOTANY, VOLUME 29 Sterile specimens of Eugenia liebmannii may be distinguished from those of Myrtus ehrenbergii Berg, a similarly small-leaved plant of eastern Mexico, by the petioles (glabrous in M. ehrenbergii even on the young antrorsely hispidulous branchlets), by the midvein (distinctly convex on the upper surface in M. ehrenbergii) and by the yellowish resinous convex glands on both leaf-surfaces in M. ehrenbergii. The latter is a plant of desert mountains in San Luis Potosi, at least as indicated by modern specimens: San Luis Potosi: San Luis Potosi to Tampico, Palmer 112^ in 1878-79 (NY); Cerro de las Minas, Guadalcazar, 25 Oct. 1891 (fl.), P. Maury 7557 (MEXU); San Jose" Pass, 10 Oct. (fr.), Pringle 352^ (A, GH); Minas de San Rafael, May 1911, Purpus 5211 (GH, MO, UC), Purpus 8588 in 1910 (UC). Eugenia lindeniana Berg, Linnaea 29: 240. 1858. See above under Eugenia capuli. Eugenia mexicana Steud. Norn. Bot. ed. 2. 1: 603. 1840. E. macrocarpa Schlecht. Linnaea 5: 560. 1830; op. cit. 13: 417. 1839, not E. macrocarpa Roxb., 1814. Probably a tree, 5-6 meters high, glabrous except for reddish hairs on the vegetative buds; leaves elliptic-ovate, 2.5-4 (-5.5) cm. wide, 6-7.5 (-9) cm. long, 1.7-2.2 times as long as wide, the blades about equally rounded to a blunt decurved- deltoid-acuminate tip and a broadly angular or rounded base, the margins at the very base cuneate-decurrent on a channeled petiole 5-10 mm. long, 1-1.3 mm. thick; midvein impressed above as a V-shaped channel; lateral veins 8-12 on each side including some intermediate ones, inconspicuous, straight and divaricate, mostly passing directly into a weak marginal vein 1-2 (-4) mm. from the margin; blades dark green and lustrous above, paler and perhaps glaucous beneath and there finely dark-dotted; inflorescence a stout axillary raceme, the axis 3-6 mm. long (in Pringle 8192 up to 10-16 mm. long), 1-1.7 mm. thick, straw-color in dry- ing, the internodes strongly compressed and angled; flowers in 1-5 opposite and decussate pairs in the racemes, or opposite at the lowermost (bracteate) nodes of new leafy branchlets; pedicels almost none, or stout, glandular-verrucose, up to 1.5 mm. long, 0.5 mm. thick; bracts broad and scarious, membranous, (or their bases indurate-persistent), sparingly ciliate, glandular-verrucose, loosely enfolding the base of the pedicel, about 1.5 mm. long, 1.5-3 mm. wide; bracteoles glandular- verrucose, short-ciliate, broadly rounded and connate, forming a deep involucre 3 mm. long and 2 mm. wide that envelops the base of the hypanthium; buds 5- 6 mm. long, obovoid, the broadly campanulate hypanthium more or less hidden by the bracteoles; calyx-lobes broadly rounded, membranous, unequal, the outer pair 1.5 mm. long and 3 mm. wide, the inner 2-2.5 mm. long and 3.5-4 mm. wide; disk 3-4 mm. wide; style 7-9 mm. long; stamens about 150, up to 6-8 mm. long, the anthers 0.5-0.7 mm. long; petals obovate or ovate, 5-6 mm. long, up to 5 mm. wide at base; ovary bilocular, the ovules about 6-10 in a compact group in each locule; fruit unknown. McVAUGH: TROPICAL AMERICAN MYRTACEAE 443 Veracruz: Jalapa, woods, 4000 ft., 19 May 1899 (fl.), Pringle 8192 (MEXU, MICH, UC); in sylvis jalapensibus, "Schiede & Deppe" (G) ; Jalapa, Schiede 544 (NY, ex herb. Meisner., "comm. cl. Schlecht- endal 1840")- As indicated in the description above, the specimens collected by Pringle are somewhat more luxuriant, with larger leaves, longer racemes and pedicels, and slightly larger flowers than the plants collected by Schiede. All the specimens are otherwise very similar and certainly belong to the same species. Schlechtendal in 1839 reported the number of ovules as 3-5 in each locule of the ovary; this was based on the flowering specimens collected by Schiede and distributed to several herbaria including those cited above. My own examination of these same specimens indicates that the number of ovules varies from 6 to 8; in Pringle's specimens the number is about 10. The type of Eugenia macrocarpa Schlecht. was a fruiting specimen also collected near Jalapa but some years earlier; its identity is unknown, but there is no reason to suppose it was a different species. Eugenia michoacanensis Lundell, Wrightia 3: 16. 1961. Shrub or small tree up to 10-12 meters high, 12-15 cm. in diameter, the branchlets very minutely puberulent with erect sharp hairs; ventral surface of petiole, and of the mid vein near base, pubescent; axis of the raceme, and the bracts and bracteoles, sometimes strigose with lustrous reddish hairs 0.2-0.4 mm. long; leaves thin, elliptic-lanceolate, lanceolate or ovate, 1-3 cm. wide, 3-7.5 cm. long, 2.3-3 times as long as wide, usually prolonged distally to a blunt, gradually acuminate tip, the acumen broad and short or narrowly triangular and up to more than 1 cm. long; base acute (the margins often about at right angles) or somewhat rounded, the margins at the very base cuneately decurrent on the channeled peti- ole 3-6 mm. long; midvein when dry flat or broadly convex on the plane surface of the blade, pubescent at least near base; lateral veins 6-8 on each side, incon- spicuous, more apparent beneath, ascending and nearly straight; marginal vein about as strong as the laterals and but slightly arched between them, 1-1.5 (-3) mm. from the margin; blades dull green and darker above, paler beneath; glands often visible as small convexities above, as dark dots beneath; inflorescence an abbreviated axillary raceme or 2-3 racemes superposed, the axis 3-6 (-10) mm. long, bearing 5-7 opposite and decussate pairs of flowers on slender, minutely puberulent pedicels 1-2.5 (-4) mm. long; bracts persistent, broad, scarious or in- durate, 0.5-0.8 mm. long; bracteoles ovate or broader, persistent, contiguous or nearly so but scarcely connate, 0.5 mm. long; buds 2-2.5 mm. long, short-pyriform; hypanthium short-campanulate, glabrous, 0.7 mm. long; calyx-lobes broad and low or obtusely triangular, membranous, glabrous, up to about 0.7 mm. long, 1-1.3 mm. wide; disk 1-1.2 mm. wide in flower, sparingly hairy about the base 444 FIELDIANA: BOTANY, VOLUME 29 of the style; style 4-4.5 mm. long, hairy below the middle; stamens about 50, up to 4.5 mm. long, the anthers 0.5 mm. long; petals white, obovate, 2.5-3 mm. long; ovary bilocular, the ovules 5 in a compact group in each locule; fruit de- pressed-globose, ripening red, probably finally black, 6-8 mm. in diameter, the calyx-lobes appressed together over the disk and outlining a circular figure 1.5- 1.7 mm. wide. Michoacan: Dist. Coalcoman: San Pedro, 600 m., 19 June (fl.), Hinton 13812 (US, isotype); Huizontla, 12 Dec. (imm. fr.), Hinton 16212 (MICH, US); Aquila, Hinton 162U (MICH).— Colima: Foot- hills of the Nevado de Colima south of Hda. San Antonio, 11 Aug. (fl.), McVaugh 16101 (MICH); Rio Cihuatlan 13 mi. north of San- tiago, 27 July (fl.), McVaugh 15797 (MICH).— Jalisco: Puente San Pedro 5 mi. southwest of Tecalitlan, 2 Dec. (imm. fr.), McVaugh & Koelz 1307, 1308 (MICH) ; 9 mi. north of west end of Bahia Navidad, 12-13 Dec. (imm. fr.), McVaugh & Koelz 1725 (MICH); same tree, 1 Oct. (fl.), McVaugh 19717 (MICH) ; Santiago-Durazno road, 11 mi. north of Rio Cihuatlan, 1 Aug. 1957, McVaugh 15973 (MICH).— Guerrero: Nanorida(?), 1500 m., 16 June 1899 (fl.), Langlasse 1062 (US); Dist. Adama, Achotla, 570 m., 25 Nov. 1937 (fr.), Y. Mexia 8898 (US) ; Acahuizotla, 10 May 1903 (fr.), E. W. Nelson 7041 (US). This species is very close to and perhaps not distinct from E. ca- puli, which differs in its smaller fruit, longer pedicels, and usually puberulent hypanthium, but glabrous style-base; see discussion above under E. capuli. Eugenia mirandae D. Ramirez Cantu, An. Inst. Biol. Mex. 14: 487. 1944. Tree probably 5-10 meters high, viscid and viscid-pubescent, the branchlets, petioles and inflorescence densely short-hispidulous with viscid hairs 0.1-0.2 mm. long, the surfaces between the hairs apparently also viscid; leaf -margins undulate- crenate, somewhat revolute, slightly viscid-pubescent; leaves thin, elliptic or lan- ceolate, 1.5-2.5 cm. wide, (4-) 5-7 cm. long, 2.5-3.5 times as long as wide, usually narrowed about equally to the gradually and bluntly acuminate tip and to the acute base, the tip sometimes more prolonged and the base then somewhat rounded; margins at base cuneately decurrent on the channeled but nearly terete petiole 5-7 mm. long, 0.7-0.9 mm. thick; midvein impressed above in a concave channel in the otherwise plane leaf, the channel hairy; lateral veins 6-8 on each side, in- conspicuous but in dried leaves convex on both surfaces, in a reticulum including the smaller veinlets; marginal vein not well defined, about as strong as the lateral ones and formed by the strongly arched and asymmetrical continuations of these, 1.5-3 mm. from the margin; blades more or less concolorous, more lustrous above, at flowering time with numerous convex glands on both surfaces; inflorescence an abbreviated axillary raceme (or 2 racemes superposed), the axis up to 5 mm. long, bearing 1-3 opposite and decussate pairs of flowers on pedicels 1.5-3 mm. long, the terminal flower sometimes present between bracts simulating bracteoles; bracts McVAUGH: TROPICAL AMERICAN MYRTACEAE 445 broad and short, hairy near the tips, 0.5 mm. long; bracteoles scarious, persistent, distinct, broadly ovate-deltoid, 0.5-0.7 mm. long and about as wide; buds 3-4 mm. long, pyriform, the hypanthium minutely puberulent, viscid, campanulate, some- what attenuate at base; calyx-lobes broadly rounded, glabrous, up to 1.8-2 mm. wide, 0.8-1.3 mm. long; disk 1.7-2 mm. wide, hairy around the base of the style; style 5-6.5 mm. long, hairy in the proximal two-thirds; stamens about 75, about as long as the style, the anthers 0.5 mm. long; petals white, ovate, 3.5 mm. long; ovary bilocular, the ovules about 10 in a compact group in each locule; fruit prob- ably black, oblong-obovoid, 8 mm. in diameter, 12 mm. long. Morelos: Sierra de Tepoxtlan, 7500 feet, May 1900 (fl.), C. G. Pringle 8333 (MEXU, MICH, UC). The above description is based entirely on Pringle 8333. Dr. Faustino Miranda informs me that these specimens are essentially topotypes of E. mirandae. I find no very convincing differences between Pringle 8333 and specimens of Eugenia pueblana Lundell, except that the Pringle specimens are somewhat viscid almost throughout. As this is a most unusual condition in this group of Myrtaceae, it is possible that these plants are abnormally coated with insect exudations or other substance. Very recently, through the kindness of Srta. De"bora Ramirez Cantu, I have studied type-material of Eugenia mirandae. The type [Morelos, "bajando del Parque a Texpoxtlan," 4 May 1943 (fl.), D. Ramirez C. s.n. (MEXU)] is a nearly glabrous flowering specimen. I cannot distinguish it from Pringle 8333 except that it completely lacks the viscid covering described above, and the mid- veins are glabrous rather than pubescent. In this latter feature it differs from E. pueblana and resembles the newly described E. cul- minicola. From E. culminicola it differs in having smaller flowers (disk 1.5-2 mm. wide) and in the slightly longer pedicels. In the one (terminal) flower dissected the type of mirandae has about 10 ovules in each locule as in Pringle 8333, whereas in pueblana the num- ber of ovules is usually 4-5, and in culminicola 6-8. None of the three species is well known, and it is possible that these several popula- tions in the highlands of Morelos, Puebla, Hidalgo and Mexico may ultimately prove to be conspecific. Eugenia oaxacana Berg, Linnaea 30: 683. 1861. Shrub or tree, the branchlets, petioles and inflorescence, and probably the young foliage, closely (antrorsely) appressed-pubescent with pale sordid hairs about 0.3 mm. long, the leaves soon glabrate, a few hairs persistent on the lower surface; leaves thin, coriaceous, elliptic-oblong, 3.5-6.5 cm. wide, 8.5-15 cm. long, about twice as long as wide, about equally rounded to the abruptly acuminate tip 446 FIELDIANA: BOTANY, VOLUME 29 and to the acute or cuneate base, the acumen obtuse, about 1 cm. long, the mar- gins long-decurrent on the petioles 8-12 mm. long, 1.3-1.5 mm. thick; midvein deeply impressed above, elevated nearly its whole diameter beneath; lateral veins rather close and parallel, 10-12 on each side in addition to some intermediate ones, rather inconspicuous but in dried leaves convex on both surfaces, rather more prominent beneath; marginal vein about as strong as the lateral ones and forming a series of angled arches between them, 2-3 mm. from the margin; blades lustrous and yellowish green above, paler and dull and drying brown beneath, minutely dark-dotted beneath; flowers in axillary glomerules, the inflorescence a much-congested axillary raceme (or 2 racemes superposed), the axis 5 mm. long or less, up to 1.5 mm. thick, bearing 1-4 pairs of sessile flowers; bracts ovate, indurate, 1 mm. long; bracteoles persistent, very broadly rounded, gla- brous within, about 2 mm. wide and 1.5 mm. long, slightly connate, forming a laterally notched involucre closely investing the hypanthium; buds and open flowers not seen; hypanthium in post-anthesis densely pale-sericeous, subglobose, 1.5-2 mm. long and high, about as long as the prominent erect concave calyx- lobes, which are broadly rounded, strongly imbricated, pubescent both sides, the inner ones longer, the outer pair 1.5 mm. long, 2-2.5 mm. wide; disk 2.5 mm. wide; style about 5 mm. long; ovary bilocular, the ovules 7-8 in a compact group in each locule. Oaxaca: Chinantla, Nov. 1842, Liebmann 3963 (C); "Dep. Oa- jaca," sin. dat., Liebmann 3962 (C); Tepinapa, B. P. Reko 1*198 in 1920 (US). I have not located the type of this species. Berg gave the type locality merely as "in Mexico ad Oaxacam," and of the type he said merely "v. specimen unicum valde mancum in herb. Buchinger." From his detailed description I presume the above specimens belong to the same species. Eugenia oerstedeana Berg, Linnaea 27: 285. 1856. E. vincen- tina Krug & Urb. ex Urb. Bot. Jahrb. 19: 621. 1895. E. conzattii Standl. Contr. U. S. Nat. Herb. 23: 1041. 1924. E. cocquericotensis Lundell, Bull. Torrey Club 64: 554. 1937. E. petenensis Lundell, Bull. Torrey Club 69: 397. 1942. This Central American species seems to be a close relative of Eugenia florida DC., a widespread species of the lowlands of north- ern South America. The two are virtually indistinguishable except that in E. florida the inflorescence is usually a raceme bearing 4-5 (or up to 10) pairs of flowers on pedicels 3-4 (or occasionally to 7) mm. long. The terminal flowers usually abort. Occasional speci- mens have some or most of the flowers solitary in the axils of small bracts at the lowermost nodes of leafy branchlets. In E. oerstedeana the racemes are developed in relatively few plants (fewer than half the specimens examined) ; terminal flowers are often present; solitary McVAUGH: TROPICAL AMERICAN MYRTACEAE 447 bracteate flowers are usually present, either in addition to the flowers in racemes, or to the exclusion of these; the pedicels are 6-10 (or up to 25) mm. long. Typical florida apparently has not been reported from Panama, but the following specimen is to be referred there: Chiriqui: Progreso, Cooper & Slater 158 (F). The Central American and Mexican populations I interpret as a single series that has been derived from E. florida or something very like it, by the gradual restriction of the flowers to the lower axils of leafy branches, the reduction in frequency of non-leafy racemes and an increase in the frequency with which a terminal flower develops, the elongation of the pedicels and perhaps also a reduction in the number of ovules (in E. florida the usual number seems to vary from 4 to 7; in E. oerstedeana the number may be 4-6, but is often only 2 or 3) . A further specialization seems to have occurred in the some- what localized Mexican population that was described as E. conzattii; here solitary flowers and short racemes may be found on the same plant with 3-flowered "dichasia." I interpret the dichasium as a development from a solitary terminal flower, lateral pedicels having arisen from the axils of the floral bracteoles. The distinctive appear- ance of this 3-flowered inflorescence which Standley called "cymose" led him to associate this species with Eugenia [Myrcianthes] fragrans, to which it has little relationship. The following specimens from eastern Mexico show this tendency to the production of "dichasia": Oaxaca: Galeotti 2882, 2887; Mell 602; Conzatti et al. 3113 (type of E. conzattii) (all US).— San Luis Potosi: Palmer 148 in 1904 (US); R. M. King 4405 (TEX). The type of E. oerstedeana Berg, which I have seen through the courtesy of the authorities at Copenhagen, is an over-mature flower- ing specimen; the leaves have the typical venation of the florida- oerstedeana complex, and also bear on the lower surface near the base the minute sessile reddish flat dibrachiate hairs that mark these otherwise nearly glabrous species. A few of the flowers are axillary, but most of them are in racemes 1-2 cm. long, bearing 4 or 5 pairs of flowers on filiform pedicels 7-9 mm. long; the flowering disk is 1.5- 1.7 mm. wide, and the ovules are 3-4 in each locule of the ovary. The type of E. petenensis Lundell (Lundell 3746, MICH) is a flowering specimen bearing relatively few flowers and most of these solitary in the axils of bracts or ordinary leaves; the ovules appear to be 2 in each locule. 448 FIELDIANA: BOTANY, VOLUME 29 A large series of specimens is available for study, exhibiting many conditions intermediate between those in which the inflorescences are all racemose and those in which the flowers are all solitary. I can- not believe that more than one species is involved. I should rather question the propriety of recognizing E. oerstedeana as a species dis- tinct from E. florida. Some Central American specimens are scarcely distinguishable from E. florida, e.g. the following, in which the flowers are almost all in racemes, the pedicels are 3-5 mm. long or less, and the ovules 4-6: BRITISH HONDURAS: El Cayo District, Vaca, P. H. Gentle 2533 (MICH) ; Little Cocquericot, Belize River, C. L. Lundell 4090 (MICH, type of E. cocquericotensis). HONDURAS: Comayagua: Malcotal, Minas de Oro, J. B. Edwards P-195 (F). The type of E. vincentina [St. Vincent, "in sylvis ad leeward side rara 200-500 m. alt., m. April, flor.," H. H. Smith & G. W. Smith 1521 (NY!)] is a flowering specimen with new thin leaves on young shoots, racemes 3-4 cm. long, filiform pedicels 8-15 mm. long, ter- minal flowers often present. The leaves have the special features of the oerstedeana-complex, namely the sunken midvein which is hispidulous above in the furrow; small sessile dibrachiate hairs on the lower leaf -surf ace near the base of the petiole; and the slightly "engraved" arching veins on the upper leaf-surface. According to Krug and Urban the ovules are 3 in each locule. The plant is indis- tinguishable from many Central American specimens of E. oerstedeana. Eugenia origanoides Berg, Linnaea 29: 229. 1858. Here has been treated a variety of plants, all with short-veluti- nous foliage and branchlets, short petioles, and very small flowers in sessile glomerules. Such plants, varying in leaf-size and shape, in compactness of the inflorescence, and in density of pubescence, range from northern Veracruz to Colombia. The type, Karwinsky 238, from Papantla, Veracruz, which I have seen through the cour- tesy of the authorities at Leningrad, is a strongly hirsute specimen with rather broadly ovate leaves 3.5-6 cm. wide, 6.5-11 cm. long, rounded or even subcordate at the base; most other specimens as far as I know have the leaves averaging 2-3 cm. wide, and acute or merely rounded at base. The flowers in the type are not unusual except that the hypanthium is more than ordinarily hirsute. In the type the ovules are 5-6 in each locule of the ovary; in other speci- mens the number may be as small as 4, or as many as 9-10. McVAUGH: TROPICAL AMERICAN MYRTACEAE 449 Perhaps more than one species is included under what is now called E. origanoides. Variation seems to be considerable. Few specimens in the fruiting condition are available. Dr. C. L. Lundell has recently described, in the course of his intensive studies on the flora of the Maya area, a species that should be compared with E. origanoides when a thorough revision of that species is under- taken. The type of this new species, E. rubella Lundell (Wrightia 3: 18. Dec. 1961) was originally thought by Dr. Lundell to represent E. origanoides, but to be atypical in having pedicellate flowers. The specimen certainly resembles other plants of E. origanoides except that the buds are borne on pedicels up to 4 mm. long. Two addi- tional collections from the same locality in Oaxaca [District of Tux- tepee, Chiltepec, in llanos, G. Martinez Calderdn 157 (MEXU, UC, US), 218 (UC, US,)] are to be referred to Eugenia rubella if that spe- cies be a valid one, but some features of these specimens cast doubt upon that validity. The racemes in the Oaxaca specimens are at first crowded, but in most instances become 8-14 mm. long, 2-3 superposed in an axil, closely simulating the racemes of Eugenia acapulcensis. The pedicels are not 4 mm. long as in the type of E. rubella, but 1-2 (-2.5) mm. long. The pubescence of the branch- lets and the racemes (in the type of rubella and also in the Oaxaca plants) seems exactly that of E. origanoides, as does the impressed midvein of the leaves. It is possible that E. rubella is a valid species, but it may be of hybrid origin or merely the result of some abnor- mality (e.g., the developing ovaries in Martinez 218 are partially transformed into galls). Full description and synonymy of E. origanoides are given in the Flora of Guatemala. Not to be confused with E. origanoides is the Guatemalan E. chinajensis Standl. & Steyerm., a very distinctive species marked by its coarsely setose foliage, branchlets and inflores- cence, and by the broad-based subcordate and subsessile leaves. The flowers are unknown, but are probably larger than those of E. origanoides. Eugenia pleurocarpa Standl. Field Mus. Bot. 4: 243. 1929. Still unknown except from western Jalisco and southern Nayarit, this species is apparently related to Eugenia octopleura Krug & Urban, which I take to be the same as E. doubledayi Standl. and E. koepperi Standl.; this is discussed more fully elsewhere in this paper. A full description of E. octopleura may be found in the Flora of Guatemala. The prominently wing-angled fruits seem to be found 450 FIELDIANA: BOTANY, VOLUME 29 in these two species alone among the North American ones; in the absence of fruit, sterile specimens of either species may be confused with E. guatemalensis or E. rekoi. The following material of E. pleuro- carpa has been seen: Nayarit: Hills 6-7 miles south of Compostela, 900-1000 m., 11-12 July 1957 (fl.), McVaugh 15322 (MICH), 5 Sep. 1960 (imm. fr.), McVaugh 18743 (MICH); mountains above La Cucaracha, sea- ward slopes 12-13 miles south of Las Varas, 300-400 m., 20 Sep. 1960 (imm. fr.), McVaugh 19194 (MICH).— Jalisco: Quimixto, 14 km. southwest of Puerto Vallarta, 28 Mar. 1959 (fr.), A. Carter & F. Chi- saki 1238 (MICH); Quimixto, 70 m., 29 Nov. 1926 (fr.), Y. Mexia 1176 (F, type; MICH) ; vicinity of San Sebastian, 1500 m., Mexia 1446 Eugenia praeterita McVaugh, sp. nov. Frutex vel arbor tomentuloso-hirsuta, pilis erectis crebris atrorufis ca. 0.5 mm. longis; folia coriacea, supra glabrescentia nitidaque, subtus tomentulosa, elliptica, apice acuminata, basi cuneata, 10-14 cm. longa, ca. 2.2-2. 5-plo longiora quam latiora, marginibus revolutis, petiolis 10-15 mm. longis, nervo medio impresso; racemus 1-1.5 cm. longus, florum 2-5 pares gerens; pedicelli crassi recti 2-3.5 mm. longi; bracteolae divaricatae, distinctae, late ovatae, usque ad 1 mm. longae; alabastra 3 mm. longa; hypanthium campanulatum vel urceolatum, post anthesin 3 mm. longum; calycis lobi late ovato-deltoidei, obtusi, intus glabri vel subglabri, usque ad 2 mm. longi latique; discus 2.5 mm. latus; stylus 5 mm. longus; ovarium biloculare, ovulis quoque loculo 8-10. A shrub or tree, the branchlets, petioles, lower leaf-surfaces and inflorescence densely short-hirsute-tomentulose with dark reddish-brown erect hairs about 0.5 mm. long; leaves coriaceous, soon glabrous and lustrous above, somewhat persistently tomentulose beneath, elliptic, about equally narrowed to the acumi- nate tip and the acute or cuneate base, 4-6 cm. wide, 10-14 cm. long, about 2.2- 2.5 times as long as wide, the revolute margins strongly decurrent on the broadly channeled rugulose petiole 10-15 mm. long, 1.5-2 mm. thick; midvein sharply impressed above; lateral veins 8-10 on each side, evident above, more conspicuous beneath, stronger than the intermediate ones if any; marginal vein about equal- ing the lateral ones and arched between them, 1.5-2.5 mm. from the margin; blades probably darker green above, paler and minutely dark-dotted beneath; inflorescence a stout raceme (or 2 racemes superposed), the axis 1-1.5 cm. long, somewhat angular, 1-1.5 mm. thick, bearing 2-5 opposite and decussate pairs of flowers on stout straight pedicels 2-3.5 mm. long, about 1 mm. thick including the hairs; bracts suborbicular, indurated but subdeciduous, 1 mm. long; bracteoles divaricate, distinct, broadly ovate, apparently indurated and somewhat persist- ent, 1 mm. long or less; buds about 3 mm. long, ellipsoid, the calyx somewhat closed over the petals until anthesjs; hypanthium just after anthesis 3 mm. long, campanulate or urceolate; calyx-lobes broadly ovate-deltoid, blunt, incurved, glabrous or nearly so within, the larger ones 2 mm. long and wide; disk 2.5 mm. wide; style about 5 mm. long; petals 3-4 mm. long, hairy without; ovary bilocular, the ovules 8-10 in a compact group in each locale. McVAUGH: TROPICAL AMERICAN MYRTACEAE 451 Veracruz?: Amatlan, July 1842, Liebmann s.n. (C), Liebmann 3959 (C, type; US). Extraordinarily different from any other species of Mexico or Central America by virtue of the dense covering of dark reddish- brown tomentum. Eugenia principium McVaugh, sp. nov. Frutex vel arbor parva, ramulis petiolisque setoso-hispidis, bracteis bracteo- lisque ciliatis, racemis sparse hispidis; folia ovata vel elliptico-ovata, 2.2-4.5 cm. longa, (1.7-) 2-2.4-plo longiora quam latiora, apice obtusa vel obtuse acuminata, basi rotundata, petiolis 2 mm. longis, marginibus incrassatis, nervo medio im- presso, basin versus hirsutulo; racemus 2-4 mm. longus, florum 2-4 pares gerens, pedicellis (1-) 2.5-7 mm. longis; bracteolae in sicco atrofuscae, persistentes, non connatae; alabastra 4-4.5 mm. longa; hypanthium campanulatum vel urceolatum, post anthesin globosum vel obovoideum, supra ovarium constrictum; calycis lobi erecti, rosei, inaequales, plani, vix imbricati, majores basi 1.2-1.7 mm. lati, 1.5- 2.3 mm. longi; discus in hypanthio prolongato vix conspicuus; stylus 5 mm. longus vel longjor; ovarium biloculare, ovulis quoque loculo 4-9; fructus ignotus. A shrub or small tree up to 3-10 meters high, the branchlets and petioles setose-hispid with pale sharp erect lustrous hairs 0.3-0.5 mm. long; bracts and bracteoles coarsely ciliate; raceme-axis and pedicels sparingly beset with hairs like those of the branchlets; leaves coriaceous, ovate or elliptic-ovate, 1-2.5 cm. wide, 2.2-4.5 cm. long, (1.7-) 2-2.4 times as long as wide, narrowed from the middle or below to a blunt or bluntly acuminate tip, rounded at base, the margins passing abruptly onto the thick inner angles of the broadly channeled petiole 2 mm. long; margins coarsely ciliate, scarcely revolute but cartilaginous-thickened, the nearly terete vein-like thickening 0.2 mm. in diameter, evident from beneath; midvein somewhat impressed above, hirsutulous toward the base; lateral veins 5-7 on each side, inconspicuous, hardly apparent in mature leaves; marginal vein about equal- ing the lateral ones, 1-1.5 mm. from the margin; leaves smooth and opaque above, pale and almost glaucous beneath; the glands obscure on both surfaces; inflores- cence a stout much abbreviated raceme (usually without an additional raceme superposed), the axis up to 2-4 mm. long, bearing 2-4 approximate, opposite and decussate pairs of flowers on pedicels (1-) 2.5-7 mm. long; bracts drying chestnut color, indurate-persistent, convex on the backs, ovate-deltoid to oblong, 1-1.5 mm. long; bracteoles drying chestnut, persistent but distinct, broadly ovate-deltoid, divaricate and flat in age, 1 mm. long or a little more; buds about 4-4.5 mm. long, the globe of the petals abruptly expanded above the short glabrous hypanthium 0.5-1 mm. high; hypanthium campanulate or urceolate, sometimes attenuate to the base, in post-anthesis becoming globose or obovoid and surmounted by a prominent narrow neck 0.5 mm. high and 1 mm. in diameter below the prominent erect "rose-pink" calyx-lobes; lobes in unequal pairs, deltoid-ovate or elliptic and rounded, flat, scarcely imbricate even in the bud, the larger ones 1.2-1.7 mm. wide at base, 1.5-2.3 mm. long, the smaller pair deltoid, 1-1.2 mm. long and wide; disk 1.7-2 mm. wide, poorly developed or at least not evident in the prolonged neck of the hypanthium, slightly hairy about the base of the style; style 5 mm. long or more; stamens about 50, about as long as the style, the anthers 0.5-0.6 mm. 452 FIELDIANA: BOTANY, VOLUME 29 long; petals white, elliptic-ovate, ciliate, 4-5 mm. long; ovary bilocular, the ovules 4-6 in a compact group in each locule (in the type), or 7-9 in each locule (in the Panama specimens) ; fruit not seen. MEXICO: Colima: Palm [Orbignya] forest near Playa de Santiago, in dense shade, sandy soil near sea level, 21 July 1957 (past fl.), McVaugh 15581 (MICH, type). PANAMA: C.Z.: West of Limon Bay, Gatun Locks and Gatun Lake, thin soil over limestone, 2 Apr. 1956 (fl.), /. M. Johnston 1773 (MICH).— Darien: Forests around Pinogana, 16-21 Apr. 1914 (fl.), H. Pittier 6532 (US). It seems strange that this species has never before been recog- nized, as it has a combination of unusual morphological features in the hypanthium and calyx-lobes, and it seems to be of wide geograph- ical range. Except for the difference in ovule-number, which may represent an individual peculiarity, the Panama specimens are very like the type. The specific epithet refers to the great palms — since unfortunately destroyed by the hurricane of 1959 — among which the type was collected. Eugenia pueblana Lundell, Wrightia 2: 167. 1961. A tree or large shrub 2-6 meters high, up to 10-12 cm. in diameter, the branch- lets when young more or less antrorsely puberulent with short reddish hairs, the petioles and midveins of the leaves puberulent with short coarse hairs, the inflores- cence often nearly glabrous; leaves lance-ovate or elliptic, 1-2.5 cm. wide, 2.5- 5.5 cm. long, 2-3 times as long as wide, gradually acuminate (the slender tip blunt), acute to rounded at base, the margins cuneately decurrent on the slightly splayed summit of the nearly terete but shallowly channeled petiole 3-6 mm. long; mid- vein concave above, appearing flat because of the persistent stiff hairs; lateral veins about 4-6 on each side, inconspicuous, more evident beneath; marginal vein ill-defined, 1-2 mm. from the margin; blades probably darker above, drying brown- ish and dull beneath, finely verruculose above, scarcely glandular beneath; inflores- cence a much-abbreviated axillary raceme (sometimes at leafless nodes on old wood), the axis almost none, the flowers in 1-2 (-4) approximate pairs, on pedicels 0.5-1.3 mm. long, 0.4 mm. thick; bracts minute, hairy; bracteoles persistent, minute, deltoid-orbicular, ciliate, 0.3 mm. long; buds 3.5 mm. long, pyriform, the globe of the petals 2.5 mm. in diameter, much surpassing the short-stipitate minutely puberulent hypanthium 1-1.5 mm. long; calyx-lobes rounded, the larger ones 0.5-0.7 mm. long, 1.5 mm. wide; disk 1.5 mm. wide (2.5 mm. wide in fruit); style 5 (-6?) mm. long; stamens about 50, about as long as the style, the anthers 0.5 mm. long; petals suborbicular, 3 mm. long; ovary bilocular, the ovules 4-5 in a compact group in each locule; fruit globose, red, probably at length black, 1.3- 1.6 cm. in diameter, the low calyx-lobes usually persistent. MEXICO: Puebla: Wet mixed forest west of Huauchinango, 1 Nov. 1943 (fr.), C. L. Lundell 12618 (isotype, MICH).— Hidalgo: Zacual- tipan, M.Martinez, in June 1940 (GH) ; forested ravines 6 miles from McVAUGH: TROPICAL AMERICAN MYRTACEAE 453 Zacualtipan, road to Tlahuelompa, 2000 m., 14 Oct. 1949 (bud), H. E. Moore 5295 (MICH); Capatla below Alumbres near Zacualti- pan, ca. 1800 m., 21 Mar. 1946 (fr.), A. J. Sharp 46214 (MICH). The following specimens apparently represent the same species, but the leaves are up to 3.5 cm. wide, 7 cm. long, the axis of the raceme up to 3.5 mm. long, the inflorescence and the hypanthium strongly puberulent: MEXICO: San Luis Potosi: Tamasopo Canyon, 25 June 1890 (fl.), Pringle 3534 (A); Las Canoas, 25 June 1890 (fl.), Pringle 3534 (GH). The following specimens differ from typical E. pueblana in hav- ing the hypanthium more strongly puberulent, and the midvein on the upper surface of the leaf glabrous: MEXICO: State of Mexico: Dist. Temascal tepee, Calera, wet bar- ranca, ca. 770 m., 19 Apr. 1933 (fl.), Hinton 3797 (K). As noted above in the discussion of Eugenia mirandae, that spe- cies is perhaps not distinct from E. pueblana. Eugenia rekoi Standl. Contr. U. S. Nat. Herb. 23: 1044. 1924. A relatively well-collected species, ranging from Sinaloa to Oaxaca and occurring also in Guatemala. It seems to be most abundant in tropical deciduous forests, on hills near the Pacific Ocean. The fol- lowing specimens have been seen: MEXICO: Sinaloa: Los Caballos, Sierra Tacuichamona, ca. 300 m., 16 Feb. 1940 (fr.), H. S. Gentry 5643 (MICH); Hda. Cafia, San Ig- nacio, 460 m., J. G. Ortega 858 (F, US; doubtfully this species).— Jalisco: East of La Manzanilla on Bahia Tenacatita, 150-200 m., 11 Nov. 1960 (fl.), McVaugh 20966 (MICH).— Colima: Manzanillo, Dec. 1890 (fr.), E. Palmer 1020 (F, US); west side of Cuyutlan la- goon, 28 Nov. 1925 (fr.), R. S. Ferris 6107 (F, US).— Guerrero: Aca- pulco, Oct. 1894-Mar. 1895 (fr.), Palmer 357 (F, MICH, US), 21 Aug. 1935 (fl.), L. H. MacDaniels 144 (F).— Oaxaca: Cafetal Apango (Cerro Huatulco), 400 m., 18 Aug. 1917 (fl.), B. P. Reko 3356 (US, type); Distr. Tamiltepec, Playa de Minizo, 7 Dec. 1921 (fr.), Con- zatti4416 (US). Eugenia rhombea (Berg) Krug & Urb. ex Urb. Bot. Jahrb. 19: 644. 1895. E. foetida y rhombea Berg, Linnaea 27: 212. 1856. ?E. leptopa Lundell, Bull. Torrey Club 69: 396. 1942. Berg's Eugenia foetida 7 rhombea was based on two specimens, one from Florida (Cabanis) and the other from "St. Domingo" 454 FIELDIANA: BOTANY, VOLUME 29 (Bertero). In the original publication (Linnaea 27: 212. 1856) the name appears to be based directly upon E. foetida /3 maleolens (Pers.) DC., and so superfluous when published, but this is evidently the result of a typographical error on Berg's part; on the preceding page (p. 211) Berg had published his var. /3 maleolens, again based directly upon E. maleolens Pers. and with a reference to DeCandolle. Pre- sumably Berg meant to indicate (on p. 212) that his rhombea had made up a part of the more inclusive maleolens, as the latter was understood by DeCandolle. Apparently Eugenia leptopa was a mixture. As pointed out in the protologue, the paratypes differ in some respects from the type, which is a fruiting specimen with narrow leaves and with somewhat immature obovoid fruit, unlike the oblate fruit of E. rhombea. Noth- ing like the type of E. leptopa has come to my attention, but I should refer the paratypes to E. rhombea. For description of E. rhombea see the Flora of Guatemala. This species is little known in Mexico except in the Yucatan Peninsula, but the following specimen has been seen : Chiapas: Cerros de Don Ventura, northwest of Tuxtla Gutierrez, 1000 m., 30 Dec. 1948 (fr.), F. Miranda 5101 (MICH). The very similar Eugenia fiscalensis Donn. Sm. (Bot. Gaz. 54: 235. 1912) is thought to differ from E. rhombea in having slightly smaller flowers and leaves, and in having the calyx-lobes and mar- gins of the staminal disk more reflexed in anthesis. Until more ma- terial of E. fiscalensis can be obtained, its status cannot be deter- mined. The following incomplete specimen seems to represent a species much like E. fiscalensis: Oaxaca: 11 km. northwest of La Ventosa, Trans-Isthmian high- way, 50 m., 14 July 1958 (past fl.), R. M. King 538 (MICH). As suggested by their association in the key above, E. rhombea and the newly described E. turneri have many superficial similarities. The latter is known only from the type, a flowering specimen refer- able neither to E. rhombea nor to E. fiscalensis. The leaves in E. tur- neri are gradually rounded to the base, elliptic-lanceolate, 2.3-3 times as long as wide, the marginal vein relatively straight and less than 1 mm. from the margin; the axis of the raceme is usually 2-4 mm. long and, like the young branchlets, appressed-pubescent. In E. rhombea and E. fiscalensis the leaves are more abruptly rounded at base, ovate, 1.5-2 times as long as wide, the marginal veins more strongly arched between the laterals and 1-3 mm. from the margins; McVAUGH: TROPICAL AMERICAN MYRTACEAE 455 the axis of the raceme is scarcely developed (or in extreme instances up to 2 mm. long), and the whole plant is glabrous. It is probable that additional and perhaps more fundamental differences will be apparent when E. turneri is better known. Eugenia salamensis Bonn. Sm. Bot. Gaz. 27: 333. 1899. Psid- ium rensonianum Standl. Journ. Wash. Acad. 14: 241. 4 June 1924. Eugenia oaxacana Standl. Contr. U. S. Nat. Herb. 23: 1043. 1 Dec. 1924, not E. oaxacana Berg, 1861. E. tomentulosa Standl. I.e. E. pur- pusii Standl. Contr. U. S. Nat. Herb. 23: 1678. 15 Nov. 1926. E. mexiae Standl. Field Mus. Bot. 4: 243. 1929. E. hiraeifolia Standl. Field Mus. Bot. 17: 202. 1937. Apparently comprising a single species is a series of almost iden- tical populations ranging from Panama to northwestern Mexico. The oldest name for the inclusive species seems to be that of Smith; it seems odd that no member of this rather widely distributed Cen- tral American group ever came to the attention of Berg. Trees 3-5 m. high, or large shrubs, at least when young rather loosely pale- or rufous-tomentose with somewhat contorted stalked or partly sessile dibrachiate hairs; leaves rather large, elliptic-oblong or -obovate, sometimes broadly elliptic or obovate, 6-16 (-20) cm. long, blunt or rounded or sometimes short-acuminate at tip; inflorescence a stout tomentose raceme 2-4 cm. long, with several pairs of rather large flowers on pedicels often 5-10 mm. long; buds 4-5.5 mm. long; disk (3-) 4-5 mm. wide; calyx-lobes broadly rounded, tomentose on both sides; bracte- oles linear or oblong, 2-5 mm. long, deciduous at or before anthesis; fruit ellipsoid- oblong, 1.5-2 cm. long. Certain rather ill-defined regional populations can be recognized, as follows: Leaves soon glabrate, usually glabrous beneath or nearly so even at flowering time; hairs of the inflorescence appressed, sparse; leaf-margins cuneately decurrent on the petiole var. rensoniana. Leaves persistently and densely pubescent or tomentulose beneath; hairs of the inflorescence abundant, stalked, forming a loose tomentum; leaf -margins abruptly contracted to the petiole, the base of the blade sometimes sub- auriculate, often asymmetric. Leaves broadly obovate or broadly elliptic, 1.4-2 times as long as wide; lateral veins with few intermediates, widely spaced, 6-8 on each side; Costa Rica and Panama var. hiraeifolia. Leaves elliptic to elliptic-obovate, 2-3 times as long as wide; lateral veins 8-12 on each side, the number indefinite because of the presence of inter- mediate ones; Guatemala and Mexico var. salamensis. Eugenia salamensis Bonn. Sm., var. salamensis. E. tomen- tulosa Standl. Contr. U. S. Nat. Herb. 23: 1043. 1924. E. mexiae Standl. Field Mus. Bot. 4: 243. 1929. 456 FIELDIANA: BOTANY, VOLUME 29 Leaves elliptic or obovate, blunt or rounded or short-acuminate at tip, nar- rowed and then abruptly rounded to the base, scarcely decurrent, often subauricu- late, 3-7 cm. wide, 6-12 (-16) cm. long, 2-3 times as long as wide; petioles 4-10 mm. long; midvein depressed and usually tomentulose, often with narrow elevated central ridge when dry; pedicels 2-6 (-8) mm. long; bracteoles 2.5-5 mm. long; style 7-9 mm. long. Specimens examined : MEXICO: Sinaloa: Without locality, Ortega 299 (US); San Ignacio, Conzatti 299 (MEXU); Rosario to Colomas, Rose 3174 (US); San Ignacio, Hda. Cana, Ortega 858 (K). — Nayarit: Acaponeta, Rose 1476 (US, type of E. tomentulosa) , M. E. Jones 22903 (UC); Acaponeta, Tiger Mine, Jones 22902 (UC) ; 19 miles northwest of Tepic, McVaugh & Koelz 725 (MICH). — Jalisco: South of Puerto Vallarta, Mexia 1129 (F, type of E. mexiae; MICH; UC), Carter & Chisaki 1204, 1205, 1209 (all MICH).— Colima: 7 miles north of Santiago, McVaugh 15915 (MICH).— Michoacan: Dist. Coalcoman, Hinton 16124, 16266, 16274 (all K, MICH & UC).— Guerrero: Dist. Montes de Oca, San Antonio, Hinton 10578 (MICH, UC, US). GUATEMALA: Salama: Llano Grande, Bergwald, Seler 2445 (US, type).— Guatemala: Without locality, Aguilar 137 in 1939 (F).— Quich<§ : Without locality, Aguilar 893, 1176 (both F).— Jalapa: Be- tween Jalapa and Paraiso, Standley 77345 (F). Eugenia salamensis Bonn. Sm., var. rensoniana (Standl.) McVaugh, comb. nov. Psidium rensonianum Standl. Journ. Wash. Acad. 14: 241. 1924. E. oaxacana Standl. Contr. U. S. Nat. Herb. 23: 1043. 1924, not E. oaxacana Berg, 1861. E. purpusii Standl. Contr. U. S. Nat. Herb. 23: 1678. 1926. Leaves on short shoots obovate with 5-8 lateral veins, obtuse or rounded, on longer shoots elliptic, obtuse or acuminate, about twice as long as wide with 8 or more lateral veins; blades often acute at base and the margins evidently and usually cuneately decurrent; petioles 5 mm. long; pedicels 3-10 mm. long; bracteoles 2 mm. long; disk 3.5 mm. wide. Specimens examined : MEXICO: Oaxaca: San Geronimo [Ixtepec], July 1914, Purpus 7139 (F; UC; US, type of E. oaxacana Standl.). EL SALVADOR: S. Calderon 838 (US, type of P. rensonianum), 1930 (US), Renson 339 (F, US). This taxon is poorly known. It is characterized by little except the thin pubescence of the inflorescence at anthesis, and by the gla- brous or nearly glabrous leaves. McVAUGH: TROPICAL AMERICAN MYRTACEAE 457 Eugenia salamensis Dorm. Sm., var. hiraeifolia (Standl.) McVaugh, comb. nov. E. hiraeifolia Standl. Field Mus. Bot. 17: 202. 1937. Psidium rensonianum sensu Standl. I.e. 18: 779. 1937. Leaves broadly obovate, broadly elliptic or oblong-obovate, rounded or ob- scurely acuminate at tip (obtusely deltoid-acuminate on shoots), narrowed toward the base from the middle or above but abruptly contracted to the petiole and often unequally subauriculate, 4-6 (-11) cm. wide, 9-12 (-20) cm. long, 1.4-2 times as long as wide; petioles 4-7 mm. long; midvein rather deeply but broadly depressed, usually with elevated median ridge when dry; pedicels 3-8 mm. long; bracteoles 2-3.5 mm. long; style 5-6 mm. long. Specimens examined: COSTA RICA: Entre Canas y Tilaran, Guanacaste, Brenes 126 88 A (NY), 12688 (F, NY); San Pedro de San Ramon, Brenes 21873 (F); San Jose", Werckle (Herb. Nat. C. R. 16437) (F, NY); El Rodeo, Villa Colon, M. Valeria 94-4 (F) ; Camino de Conception a Desmonte, J. A. Echeverria 4-180 (F, UC); Pres du pont de Rio Grande, Chemin de fer au Pacifique, Pittier (Herb. Nat. C. R. 16405) (US). PANAMA: Monte Oscuro near Panama City, J. Zetek 3550 (F, type of E. hiraeifolia). The type of Eugenia hiraeifolia is a well-matured vigorous leafy branchlet bearing a single raceme and one fruit; the pubescence has almost all fallen, a little persisting in protected places on the leaves and the fruit; the leaves toward the end of the shoot are unusually elongated and bluntly acuminate, but I believe the specimen belongs to the same species as the Costa Rican plants cited above. Miss Amshoff has already pointed out (Act. Bot. Ne"erl. 5: 278-279. 1956) the similarity between these Costa Rican specimens and the Pana- manian E. hiraeifolia. Plants of this variety are surely conspecific with those of var. salamensis but are readily distinguished, at least in the specimens at hand, by the larger and broader, more obtuse leaves with fewer and more distinct veins. Eugenia sinaloae Standl. Contr. U. S. Nat. Herb. 23:1042. 1924. This species is known to me from the type only: Sinaloa: Guada- lupe, Rose et al. 14793 (US). Very little can be added to the original description, except to say that the pubescence is very like that in Eugenia origanoides, namely of abundant pale erect hairs, readily observed with the unaided eye. The flowers are unknown. 458 FIELDIANA: BOTANY, VOLUME 29 Eugenia standleyi McVaugh, nom. nov. Myrtus oaxacana Standl. Contr. U. S. Nat. Herb. 23: 1038. 1924. Not Eugenia oaxa- cana Berg, 1861, norE". oaxacana Standl. 1924. The short-racemose inflorescence and the bibracteolate tetramer- ous flowers establish this plant without question as a species of Eu- genia. It is apparently still known from the type only: Oaxaca: Between Juchitan and Chivela, Nelson 2631 (US). Eugenia symphoricarpos McVaugh, sp. nov. Arbor vel frutex, ramulis hispidulis; racemis foliorumque costis plusminusve pilis crassiusculis aeruginosis rectis vel contortis usque ad 0.8 mm. longis obsitis; folia elliptica vel ovata, 3-7 cm. longa, 1.5-2.8-plo longiora quam latiora, apice acuminata, marginibus in petiolum 5-8 mm. longum cuneatim decurrentibus, nervo medio supra depresso, laminis supra atroviridis lucidis, subtus pallidioribus; flores numerosi, sessiles, glomerulis axillaribus dense congesti, bracteis oblongo- lanceolatis vel ovatis obtusis 1.5-4 mm. longis, bracteolis conspicuis, vix connatis sed imbricatis, sub hypanthium tomentulosum involucrum efformantibus; calycis lobi concavi, majores 2 mm. longi latique; discus 1.5 mm. latus; stylus 7 mm. longus; stamina ca. 60; ovarium biloculare, ovulis quoque loculo (3-) 5-8; fructus ad modum Symphoricarpi glomerati, subglobosi, diametro 5 mm. Probably a small tree, somewhat pubescent when young, finally glabrate; branchlets hispidulous or antrorsely pubescent with dark reddish hairs; inflores- cence, and both sides of the midveins of the foliage leaves, more or less beset with rather coarse, straight or contorted coppery hairs up to 0.5-0.8 mm. long; leaves elliptic or ovate, 1.5-4 cm. wide, 3-7 cm. long, 1.5-2.8 times as long as wide, broadly and bluntly or rather slenderly acuminate at tip, rounded to the cuneately narrowed base, the margins cuneately decurrent on the slender chan- neled petiole 5-8 mm. long, 0.7-1 mm. thick; midvein depressed above in a V-shaped channel, elevated at least half its diameter beneath, more or less per- sistently pilose with contorted hairs; lateral veins 6-8 on each side including some intermediate ones, convex on both surfaces in dry leaves but not very conspicu- ous, straight and ascending; marginal vein about equaling the lateral ones and arched between them, 0.5-1.5 mm. from the margin; blades dark green and lus- trous above, when dry much paler and dull beneath, dark-dotted beneath; flowers numerous in dense axillary glomerules up to 1-1.5 cm. in diameter, the raceme axis (often 2 or more superposed) up to 3-6 mm. long, hirsute, bearing up to 3-6 approximate opposite decussate pairs of sessile flowers; buds 2.5 mm. long, obo- void; bracts oblong-lanceolate to ovate, blunt-tipped, glabrous or coarsely ciliate, about 1 mm. wide at base, 1.5-4 mm. long; bracteoles conspicuous, ovate, ca. 1.3 mm. long and more than 1 mm. wide, their margins overlapping but scarcely connate, the whole forming a deep involucre enveloping and nearly concealing the broadly campanulate, densely rufous-tomentulose hypanthium 0.7-1 mm. long; calyx-lobes concave, thin, glandular-verrucose, suborbicular or a little longer or shorter, the outer ones up to 1.5 mm. long and 2 mm. wide, the inner 2 mm. long and wide; disk 1.5 mm. wide, hairy around the base of the style; style 7 mm. long; stamens about 60, about as long as the style, the anthers 0.5 mm. long: McVAUGH: TROPICAL AMERICAN MYRTACEAE 459 petals white, 3 mm. long; ovary bilocular, the ovules (3-) 5-8 in a compact group in each locule; fruits in dense clusters, probably black, subglobose or obovoid, up to 5 mm. in diameter, tipped by the prominent erect thin calyx-lobes. At middle elevations, 800-1500 meters, in the Sierra Madre Oriental, flowering April-May, fruiting May-September. Tamaulipas: Sierra de Tamaulipas, Martinez M. & Borja L. F-1923 (TEX), F-2076 (TEX).— Veracruz: Orizaba, C. Mohr in 1857 (US) ; Cerro del Carrisal, Orizaba, P. Tow in Apr. 1891 (MEXU) ; Zacuapan, C. A. Purpus 2058 (GH, UC), 7523 (A, GH, NY; UC, type) ; Paso del Correa, Liebmann 3970 (C) ; Barranca near Rancho Viejo, Apr. 1934, Purpus 1625^ (A). — Locality unspecified: Mexico, Sumichrast 1955 (K). Eugenia teapensis McVaugh, sp. nov. Frutex vel arbor minuscule sparseque puberula, foliis majusculis ellipticis 6-9 cm. latis, 14-20 cm. longis, obtuse acuminatis, venis lateralibus marginali- busque et costa supra impressis, lateralibus prominentibus, intermediis vix evo- lutis; racemi crassi usque ad 3 mm. longi 4-6-flori, pedicellis tenuibus 10-11 mm. longis, bracteolis ovatis parvis persistentibus, floribus majusculis; alabastra 8 mm. longa; calycis lobi concavi, inaequales, majores 6 mm. longi obovati; discus 3.5 mm. latus; stylus 7-10 mm. longus; stamina plurima (ca. 200), antheris 1-1.2 mm. longis; ovarium biloculare, ovulis quoque loculo ca. 12. A tree or shrub, nearly glabrous, the foliage, branchlets and inflorescence sparingly beset with minuscule erect hairlike processes; bracts, bracteoles and calyx-lobes ciliate; leaves elliptic, rather large and prominently veined, 6-9 cm. wide, 14-20 cm. long, 2-2.5 times as long as wide, about equally narrowed to the bluntly acuminate tip and the acute to rounded base, the margins somewhat de- current on the corky-rimose petiole 6-8 mm. long, 2-2.5 mm. in diameter; midvein deeply and narrowly impressed above in a V-shaped cleft, elevated nearly its whole diameter beneath; lateral veins 6-10 on each side, well separated and with- out any conspicuous intermediate ones, prominent beneath, depressed below the upper surface of the blade but the individual veins and veinlets convex in a re- ticulate pattern; marginal vein about as strong as the lateral ones, similarly im- pressed above, forming strong somewhat asymmetrical arches 5-10 mm. from the margin; blades probably darker green and lustrous above and with some convex glands, the lower surface sparingly dark dotted; inflorescence an abbreviated raceme, the axis stout, dorsiventrally compressed, up to 1.5-3 mm. long, bearing 2-3 opposite decussate pairs of flowers on slender pedicels 10-11 mm. long; bracts small; bracteoles ovate, persistent, distinct, 1 mm. long; buds 8 mm. long, obovoid, the conic hypanthium 2.5 mm. long; calyx-lobes deeply concave, in unequal pairs, the smaller pair orbicular, 4-5 mm. long and wide, the larger pair obovate, 5 mm. wide, 6 mm. long; disk 3.5 mm. wide; style 7-10 mm. long; stamens about 200, up to 9 mm. long, the anthers 1-1.2 mm. long; petals white, obovate, 10 mm. long; ovary bilocular, the ovules about 12 in a compact group in each locule; fruit unknown. 460 FIELDIANA: BOTANY, VOLUME 29 Tabasco: Forets sombres de Teapa, April (fl.), Linden 623 (K, type). — Chiapas: Near El Ocote, ca. 30 km. northwest of Ocozo- cuautla, 27-28 July 1958 (fl.), E. Hern&ndez X. (Hb. Miranda 8870) (MICH). Eugenia trunciflora (Cham. & Schlecht.) Berg, Linnaea 27: 223. 1856. Myrtus (?) trunciflora Schlecht. & Cham. Linnaea 5: 561. 1830. The original collections of this species came from "sylvis umbro- sis inter Mesachica et Mapilque" and from Papantla, both localities in Veracruz, Schiede 547. I have not seen these specimens, but from the original description and that given by Berg I suppose the follow- ing belong to the same species; the leaves are mostly broadest at or above the middle; the flowers are a little more than half as large as those of E. xilitlensis (q.v.) and are borne on slender pedicels 8- 10 mm. long (the original description says "Pedunculi 3-9 lineas longi"): Veracruz: Zacuapan, C. A. Purpus 2434 (UC, US), 7663 (US). Eugenia turneri McVaugh, sp. nov. Arbor 6-7-metralis, subglabra, ramulis juvenilibus racemisque pubescentibus; folia lanceolata vel elliptico-lanceolata, 3.5-5.5 cm. longa, 2.3-3-plo longiora quam latiora, apice anguste sed obtuse acuminata, basi gradatim rotundata, petiolis 4.5-5 mm. longis, nervo medio supra inciso; racemi abbreviati 2-4 mm. longi, vel flores ramulorum novorum nodis infimis solitarii bracteatique; pedicelli filiformes 8-12 mm. longi; bracteolae scariosae, distinctae, vix deciduae; alabastra 3.5 mm. longa, supra hypanthium glabrum subglobosa; calycis lobi membranacei, post anthesin reflexi, majores 2 mm. longi; discus 1.5 mm. latus, annulus staminalis subhirsutulus; stylus 4 mm. longus; ovarium biloculare, ovulis quoque loculo 2 (-1); fructus ignotus. Tree 6-7 meters high, nearly glabrous, the vegetative buds, axis of the raceme, and the very youngest growth appressed-pubescent with pale hairs 0.1-0.3 mm. long; leaves lanceolate or elliptic-lanceolate, 1.5-2 cm. wide, 3.5-5.5 cm. long, 2.3-3 times as long as wide, the tip slenderly but bluntly acuminate, the base gradually rounded, the margins at base passing abruptly to the rather deeply channeled petiole 4.5-5 mm. long, 0.6 mm. thick; midvein slightly engraved above in the nearly plane surface of the blade; lateral veins 6-8 on each side including some intermediate ones, slightly elevated when dry but very inconspicuous on both sides; marginal vein about equaling the lateral ones and somewhat arched between them, less than 1 mm. from the margin; blades nearly concolorous and yellow-green when dry, very smooth but apparently not lustrous above, a little paler beneath and roughened by many small inconspicuous glands; inflorescence an abbreviated axillary raceme (sometimes at leafless nodes or terminal!), or the flowers solitary and opposite at the lowermost (bracteate) nodes of new leafy branchlets; raceme-axis 2-4 mm. long, bearing 2-3 pairs of flowers on filiform McVAUGH: TROPICAL AMERICAN MYRTACEAE 461 pedicels 8-12 mm. long; bracts scarious, pubescent, 0.5 mm. long; bracteoles lance- ovate, 0.5-0.7 mm. long, scarious, distinct, ciliate, subpersistent; buds 3.5 mm. long, subglobose above the short-campanulate glabrous hypanthium 1 mm. long; calyx-lobes membranous, ciliate, unequal, the outer pair broadly rounded or ellip- tic-ovate, 1.2 mm. wide, 1.5 mm. long, the inner pair suborbicular-oval, 1.7 mm. wide and 2 mm. long, all the lobes reflexed from near the base soon after anthesis; disk about 1.5 mm. wide, the staminal ring bristly; style 4 mm. long; stamens about 50, about as long as the style, the anthers about 0.7 mm. long and wide with broad basal lobes; petals white, ovate, ciliate, 3.5-4 mm. long; ovary bilocular, the ovules 2 and collateral, or only 1, in each locule; fruit not seen. For a note on the relationships of this species, see above under E. rhombea. Michoacan: 1 mile east of La Placita, near the Pacific Coast, dry stream bottom, 4 July 1950 (fl.), B. L. Turner 2073 (MICH, type). Eugenia xalapensis (HBK.) DC. Prodr. 3: 276. 1828. Myrtus xalapensis HBK. Nov. Gen. & Sp. 6: 145. 1823. Shrub or small tree 1-4 meters high, soon glabrate, the growing branchlets and the youngest petioles minutely hispidulous-puberulent, the mature leaves and the pedicels and flowers quite glabrous; leaves ovate or ovate-elliptic, 1.5- 3.5 cm. wide, 3-7 cm. long, 1.7-2.7 times as long as wide, the tip prominently acuminate (the very tip blunt), the base acute, the margins cuneately decurrent on the inner angles of the channeled petiole 4-8 mm. long; midvein deeply and narrowly impressed above; lateral veins 5-6 on each side in addition to some inter- mediate ones, inconspicuous above, apparent beneath; marginal vein about as strong as the lateral ones and little arched between them, 0.5-1.5 mm. from the margins; blades dark green and lustrous above, much paler beneath; glands appar- ent beneath as small dark or resinous dots, sometimes as convexities above; in- florescence an abbreviated axillary raceme, or some flowers opposite at the lower- most (bracteate) nodes of new leafy branchlets; raceme-axis up to 1 cm. long with 4 pairs of flowers, usually 2-4 mm. long with 1-3 opposite and decussate pairs of flowers on pedicels 3-5 mm. long; bracts indurate-persistent, blunt, oblong or ovate, 1-1.5 mm. long; bracteoles connate, oblong-ovate, obtuse, forming a deeply 2-lobed involucre with divaricate lobes 0.8 mm. long and wide; buds 3-3.5 mm. long, pyriform, the hypanthium turbinate or campanulate, often somewhat atten- uate at base, 1-1.5 mm. long; calyx-lobes suborbicular, thin-margined, the larger ones 1.3-2 mm. long and wide; disk 1.5 mm. wide (2-2.5 mm. wide in fruit); style (not seen expanded) probably about 5 mm. long; stamens 60-70, up to 4 mm. long, the anthers 0.5 mm. long; petals white, about 3 mm. wide, concave, suborbic- ular; ovary bilocular, the ovules about 10 in a compact group in each locule; fruit black, globose, 10-13 mm. in diameter, the calyx-lobes scarcely persistent. Veracruz: Xalapa, 770 hex., Bonpland (P, type, not seen; Field Mus. Neg. 36911), Schiede 543 (NY), Rose & Hay 6173 (NY), L. H. MacDaniels 802 (F), 849 (F), C. R. Orcutt 2855 (F), C. L. Smith 1507 (F), 1898 (F).— Hidalgo: Bluffs near Tutotepec, 5300 ft., 28 Aug. 1945 (fr.), A. J. Sharp 45866 (MICH); Dist. Molango, northwest 462 FIELDIANA: BOTANY, VOLUME 29 of Lake Atexca, deciduous woods, 1500 m., 9 Nov. 1946 (fr.), H. E. Moore 1953 (UC). — San Luis Potosi: 20 mi. east of C. del Maiz, oak forest, 1200 m., 9 May 1949 (fl.), McVaugh 101+85 (MICH).— Ta- maulipas: 5 mi. northwest of Gomez Farias, 3600 ft., 15 May 1949 (fl.), B. E. Harrell 179 (MICH).— State:? Joya de Salas trail above El Cielo, 1 Sep. 1952 (fr.), Sharp et al 52218 (MICH); San Luis Potosi to Tampico, Dec. 1878-Feb. 1879, Palmer 1045 (GH, NY). This rather local species was correctly interpreted by Schlechten- dal and Chamisso (Linnaea 5: 560. 1830) and by Standley (Contr. U. S. Nat. Herb. 23: 1040, in key; 1045. 1924), but the name E. xala- pensis has been erroneously applied to many specimens from the lowlands of eastern Mexico and adjacent Guatemala. Sterile speci- mens of E. xalapensis may ordinarily be distinguished from those of E. symphoricarpos (which occurs in the same general area) by the glabrous branchlets; in E. xalapensis only the very youngest grow- ing branchlets are pubescent. Eugenia xilitlensis McVaugh, sp. nov. Arbor 7-8-metralis, racemis abbreviates hirsutulis; folia lanceolato-elliptica, 16-20 cm. longa, ca. 3-plo longiora quam latiora, acuminata, basi cordato- auriculata, auriculis rotundatis 1-1.5 mm. longis; petioli 8-10 mm. longi; nervus medius supra concavus; lamina subtus venis elevatis glandulisque numerosis maximis instructa; flores 1-2, subsessiles, e nodis defoliatis oriundes, pedicellis hirsutis 3 mm. longis; alabastra 10-11 mm. longa, hypanthio canescenti, brac- teolis deciduis 3.5 mm. longis; calycis lobi maiores 7 mm. lati, 9 mm. longi; discus 6-7 mm. latus; stylus 15 mm. longus; stamina ca. 300, antheris 1.2-1.5 mm. longis; ovula quoque loculo ca. 15-20. A tree up to 7-8 meters high, the young branchlets and foliage sparingly hirsutulous, the inflorescence thickly hirsutulous with pale hairs; leaves lance- elliptic, 5.5-6.5 cm. wide, 16-20 cm. long, about 3 times as long as wide, gradually and bluntly acuminate; blades rounded from the middle or below to the cordate- auriculate base, the auricles rounded, 1-1.5 mm. long, 3-4 mm. wide; petioles 8-10 mm. long, 2 mm. thick; midvein concave above; lateral veins 10-15 on each side including some intermediate ones; marginal vein about as strong as the lateral ones and forming angular arches between them, 2-5 mm. from the margin; leaves green and smooth above, sparingly gland-dotted, the veins appearing as fine pale parallel lines; lower surface paler, more prominently dotted with glands 0.2-0.4 mm. in diameter, the veins elevated and more evident; flowers 1-2, nearly sessile at leafless nodes, the pedicels hairy, 3 mm. long, 1-1.5 mm. thick; buds 10-11 mm. long, pyriform, the cup-shaped hypanthium canescent, 3-4 mm. long; bracteoles ovate, membranous, deciduous, 2 mm. wide, 3.5 mm. long; calyx-lobes rounded, in unequal pairs, prominently glandular, thin, pu- bescent along the edges, the larger ones 7 mm. wide, 9 mm. long, the smaller 4.5-5 mm. wide, 6 mm. long; sinuses splitting a little between the lobes after anthesis (perhaps as a result of pressing in the specimen seen); disk 6-7 mm. wide; style 15 mm. long; stamens about 300, about as long as the style, the anthers McVAUGH: TROPICAL AMERICAN MYRTACEAE 463 1.2-1.5 mm. long; petals white, ovate, narrowed to the tip, ciliate-fringed, up to 1.5 cm. long or more; ovary bilocular, the ovules 15-20 in a compact group in each locule; fruit unknown. San Luis Potosi : In forest west of Xilitla, Cerro Prieto, 3800 feet, 22 Apr. 1946 (fl.), A. J. Sharp 46284 (MICH, type). This species occurs in the same general region as Eugenia trunci- flora, which it somewhat resembles. For a key to the North Ameri- can species of Eugenia with cordate-auriculate leaves, see below under Notes on Miscellaneous Species. Eugenia yautepecana Lundell, Wrightia 2: 107. 1960. A shrub 3-4 meters high, the foliage nearly glabrous from the first, the branchlets and inflorescence (and the youngest leaves) densely appressed-pubescent with pale reddish or almost colorless dibrachiate hairs mostly 0.2-0.4 mm. long, the whole plant ultimately glabrate; leaves elliptic or occasionally ovate, 1.5- 3.5 cm. wide, 3-6 cm. long, 2-2.4 times as long as wide (or leaves near the bases of branchlets 1.4-1.6 times as long as wide); blades often about equally narrowed to a blunt or obscurely acuminate tip and a rounded or acute base, the margins usually passing abruptly to the channeled petiole 3 mm. long (bases of the broader leaves rarely subcordate, or in the narrower ones rather long-decurrent) ; mid- vein broadly and shallowly concave above at least near base, or shallowly chan- neled between low parallel ridges; lateral veins 5-7 on each side, ascending, pale and readily visible on both surfaces of dried leaves, scarcely forming a marginal vein but recurving in a series of irregular angled arches 2-3 mm. from the margin to join the next succeeding ones; blades dark green and lustrous and very minutely pale-verruculose above, paler beneath, with numerous convex resinous glands usually apparent on both surfaces; inflorescence an abbreviated axillary raceme or two racemes superposed, the axis up to 3-6 mm. long, bearing 1-3 opposite and decussate pairs of flowers on pedicels 3-5 (-7) mm. long, up to 0.6-0.8 mm. thick in fruit; bracts ovate, 1 mm. long, the lower ones broader; bracteoles ovate, divaricate, distinct, persistent, 1-1.4 mm. long; buds 4 mm. long; calyx-lobes broadly rounded, concave, sparingly beset with large resinous convex glands 0.2-0.3 mm. in diameter, appressed-pubescent on both surfaces but persistently and more densely on the inner; inner lobes broadly rounded, 2.5-3 mm. wide, 2.5 mm. long, the outer often ovate, 1.5-2 mm. wide and long; disk 2.5-3 mm. wide, somewhat hairy about the base of the style; style 5-6 mm. long or more; stamens about 75, about as long as the style, the anthers 0.5-0.7 mm. long; ovary bilocular, the ovules about 10 in a compact group in each locule; fruit ellipsoid-oblong, black(?), 8-10 mm. in diameter, 10-15 mm. long, crowned by the erect concave calyx-lobes. Veracruz: Barranca de Panoaya, Aug. 1919 (fl.), C. A. Purpus 8432 (UC, US), July 1920, Purpus 8505 (UC), Purpus 8561 (UC).— Mexico: Temascaltepec, Acatitlan, 27 Dec. 1934 (fr.), G. B. Hinton 7172 (K, MICH, UC, US) .— Morelos : Cuernavaca, pedregal, 23 Sep. 1896 (imm. fr.), Pringle 7234 (MICH); Yautepec-Cuernavaca road, lava field, 2 Oct. 1943 (imm. fr.), Lundell & Lundell 12502 (MICH, US, isotypes). 464 FIELDIANA: BOTANY, VOLUME 29 NOTES ON MISCELLANEOUS SPECIES Eugenia basilaris McVaugh, sp. nov. Frutex vel arbor parva 2-4 m. alta, ramulis rufescentibus longitudinaliter rimosis, decor ticantibus; ramuli hispido-hirsuti, pilis acutis fuscis 0.5-1 mm. longis obsiti; folia elliptica vel ovata, basi plusminusve cordata; petioli crassi 3 mm. longi; nervi laterales utrinque 5-8, arcuatim ascendentes, nervum mar- ginalem vix efformantes; flores glomerati subsessiles, pedicellis 1 mm. longis subglabris; alabastra 3-4 mm. longa pyriformia, hypanthio hirsute anguste urce- olato 1.2-1.5 mm. longo; calycis lobi subglabri mox reflexi; discus 1.5 mm. latus, centre concavo, marginibus reflexis; fructus subglobosus diametro 1.7-2 cm., apice conico, disco peltato 2-2.5 mm. lato coronato; ovarium biloculare, loculis basilaribus, ovulis quoque loculo 5-7. Shrub or small tree 2-4 m. high, the younger branchlets wrinkling in sharp ridges when dry, the thin light reddish brown outer bark in mature branchlets cracking longitudinally and flaking off; branchlets, petioles, and young leaves at least on the lower surface hispid-hirsute with erect or flexuous stout sharp dark reddish-brown hairs 0.5-1 mm. long; leaves elliptic or elliptic-ovate, 3-5 cm. wide, 6-11 cm. long, 2-2.5 (-3) times as long as wide, broadly and bluntly or sometimes slenderly acuminate, rounded toward the base from the middle or below, the base itself broadly or narrowly rounded, cordate or cordate- auriculate, with a definite shallow often asymmetric sinus between the rounded auricles; petiole about 3 mm. long, 1.5 mm. thick; mid vein impressed above, elevated its whole thickness beneath; lateral veins 5-8 on each side, seen as fine lines above and often somewhat depressed when dry, prominent beneath, scarcely forming a marginal vein, but diminishing distally and arching over at a point 3-5 mm. from the margin to join the next succeeding vein; leaves often blackening in drying, darker above, not prominently glandular; flowers in axillary glomerules, the inflorescence a crowded axillary raceme or two racemes super- posed, the axis up to 3-3.5 mm. long, 4-angled, bearing up to 4 pairs of flowers on stout nearly glabrous pedicels 1 mm. long (in fruit up to 3 mm. long, 1.5 mm. thick); bracts oblong, hirsute outside, 1.5 mm. long; bracteoles like the bracts, subpersistent, not connate, 1 mm. long; buds 3-4 mm. long, pyriform, the globe of the petals broader than the hirsute narrowly urceolate hypanthium 1.2-1.5 mm. long; calyx-lobes ciliate but nearly glabrous, soon reflexed, the larger ones 1.5-1.7 mm. wide, 2-3 mm. long; disk 1.5 mm. wide, the center concave, the staminal ring hairy, the margins soon much reflexed; style about 6 mm. long; fruit subglobose, glabrous, glandular- verrucose, 1.7-2 cm. in diameter, pro- longed into a broadly conical apex surmounted by the peltate disk 2-2.5 mm. wide, the calyx-lobes radiate, rounded, glabrous within; ovary bilocular, the locules expanded near the base of the elongate hypanthium; ovules 5-7 in each locule. Specimens examined: Costa Rica: San Ramon, bois des collines, 1200-1400 m., 5 May 1913, A. Tonduz (Herb. Nat. C. R. 17662) (F, MICH, US); San Pedro de San Ramon, bois, collines, 850-1075 m., A. M. Brenes 4282 (F, type), 4267, 4323, 4538, 5022 (all F). McVAUGH: TROPICAL AMERICAN MYRTACEAE 465 This most distinctive species has been confused with E. truncata Berg, to which it bears little resemblance except that the leaves are about the same size. The flowers in E. truncata are borne on slender pedicels 10-25 mm. long. Eugenia truncata Berg, Linnaea 27: 157. 1856. E. oreinoma Berg, I.e. 158. ?E. guanacastensis Standl. Field Mus. Bot. 8: 144. 1930. According to Berg E. truncata was distinguished from E. oreinoma chiefly by the narrower and very short-petiolate leaves truncate at base. Study of the abundant Costa Rican material now available suggests that but one species is involved. The type of E. truncata, which I have seen through the courtesy of the authorities at Copen- hagen, is a strongly pubescent specimen consisting mostly of young vegetative branchlets 3-5 cm. long; the leaves are all truncate and small for the species (cf. Field Mus. Neg. 20982). An isotype of E. oreinoma (seen at F) bears relatively large broad leaves (5.6-7 cm. wide, 8.5-13 cm. long) which are merely rounded at the base. In other respects the two specimens are similar. The glabrous or nearly glabrous plant described as E. guanacas- tensis is evidently very closely related to E. truncata and may be a synonym of that species, as Standley ultimately decided. The young shoots are sometimes hirsutulous exactly as in E. truncata, (e.g. in Standley & Valerio 4424-0), or the whole plant may be glabrous. The petioles are but 3^4 mm. long as against the usual length of 5-8 mm. in E. truncata. The leaves are always auriculate-cordate. In North America the leaves in most species of Eugenia are acute or cuneate at base, or if the margins are rounded abruptly in to the petiole they are usually at least slightly decurrent on the petiole; the summit of that organ may appear to be dilated or the blade may be said to be cuneately narrowed at base. In a few species, how- ever, the blades are sometimes or always cordate or cordate-auricu- late, the margins seeming to merge with the ventral surface of the petiole while curving away from its base. In several species (e.g. E. basilaris, E. truncata) the leaves are usually truncate or sub- cordate but may be merely rounded at base; in E. origanoides and E. salamensis the leaves are infrequently subcordate; E. papalensis is known only from the type, which may or may not be a represent- ative specimen. In E. chinajensis, E. jutiapensis, E. trunciflora and E. xilitlensis, as far as known, the leaves are always cordate-auriculate. 466 FIELDIANA: BOTANY, VOLUME 29 Flowers in elongate axillary racemes 2-5 cm. long; pubescence of appressed and mostly dibrachiate hairs E. salamensis Donn. Sm. Flowers glomerate or fasciculate, the racemes so much contracted that the flowers appear to arise directly from the axils or from very short lateral spurs; pubes- cence if present mostly of more or less erect simple hairs. Pedicels slender, hispidulous (rarely glabrous), 1-2.5 cm. long; buds 5-7.5 mm. long. Hypanthium in flower glabrous or very sparingly pubescent especially at base. Lateral veins 10-12 on each side of the midvein; foliar glands up to 0.5 mm. in diameter, apparent on both sides of dry leaves, more conspicuous beneath; Veracruz E. trunciflora (Cham. & Schlecht.) Berg. Lateral veins 6-8 on each side; foliar glands mostly 0.1-0.2 mm. in diam- eter, scarcely visible in dried leaves; Costa Rica E. truncata Berg. Hypanthium in flower densely velutinous or hirsute; lateral veins 5-6; foliar glands 0.1-0.2 mm. in diameter, visible from the upper surface or not at all; El Salvador and Guatemala E.jutiapensis Standl. & Steyerm. Pedicels very short, 3 mm. long or less, or the flowers sessile and glomerate. Petioles 8-15 mm. long, the blades 12-20 cm. long. Blades rounded to truncate at base; foliar glands scarcely apparent; flowers unknown, probably large, the disk in fruit 3-3.5 mm. wide, the larger calyx-lobes 3-4 mm. long and wide; Guatemala. E. papalensis Standl. & Steyerm. Blades cordate-auriculate; foliar glands prominent beneath, 0.2-0.4 mm. in diameter; buds 10-11 mm. long; disk in flower 6-7 mm. wide, the larger calyx-lobes 7 mm. wide, 9 mm. long; San Luis Potosl. E. xilitlensis McVaugh. Petioles 5 mm. long or less; blades mostly 10 cm. long or less; buds 2-4 mm. long. Plants uniformly and coarsely but not very densely setose, the reddish or tawny hairs in no way obscuring the surface of the branchlets or the lower leaf-surface; setae on the midvein near the base of the lower sur- face of the leaf 1-1.3 mm. long; cilia of the calyx-lobes 0.5-0.7 mm. long; eastern Guatemala E. chinajensis Standl. & Steyerm. Plants finely and often densely soft-hirsutulous, the pale hairs often so numerous as to cover the branchlets and principal veins; hairs of the midvein near the base of the lower leaf-surface 0.5 (-0.8) mm. long; cilia of the calyx-lobes mostly 0.2-0.4 mm. long. Hypanthium narrowly urceolate; fruit 1.7-2 cm. in diameter; Costa Rica. E. basilaris McVaugh. Hypanthium hemispheric to subglobose; fruit 6-8 mm. in diameter; Mexico to Colombia E. origanoides Berg. Eugenia egensis DC. Prodr. 3: 281. 1828. This species has not previously been reported from North Amer- ica. It is one of the most distinctive South American species, and I think there can hardly be any doubt of the determination of the specimen cited below. The principal range of E. egensis is from the McVAUGH: TROPICAL AMERICAN MYRTACEAE 467 Guianas to eastern Colombia, south chiefly in the Amazon basin to eastern Bolivia. Costa Rica: El General, bordes del Rio Pacuare, 600 m., 17 Sep. 1942, J. Leon 1008 (US). Eugenia flavoviridis Lundell, Amer. Midi. Nat. 29: 479. 1943. E. dissitiflora Lundell, Wrightia 2: 207. Sep. 1961. E. flavida Lun- dell, Wrightia 3: 14. Dec. 1961. This appears to be a distinct taxon of the Yucatan region, super- ficially distinguished from other species with crowded racemes by its long-petiolate, elliptic (rather than ovate) leaves, by the rather sharply and narrowly impressed (rather than flat or merely de- pressed) mid vein, by the coarse ascending pubescence of the race- mes, and also by the very smooth upper leaf-surface in which the veins are scarcely apparent to the eye. Material from British Hon- duras does not seem to differ appreciably from specimens of E. dis- sitiflora recently collected in Pete"n. Of the approximately 15 speci- mens I have seen, all bear immature fruit. It is unfortunate that no flowering material is available for study. Species of this affinity are numerous in the West Indies, and it may be that the Central American plant will eventually prove to be a synonym of one of these. The type and only known specimen of Eugenia flavida (Guate- mala, Pete"n, Dos Lagunas, in low secondary forest of airfield clear- ing, E. Contreras 1693), which I have seen through the courtesy of Dr. Lundell, is a mature leafy branch from which almost all traces of pubescence seem to have been eroded. Vegetatively I cannot dis- tinguish this specimen from similarly weathered leafy twigs of E. "dissitiflora" from nearby localities in Pete"n. The fruits accompany- ing the type specimen of E. flavida have apparently been partly rotted while lying on the ground ; none is whole, but from the shape of the seeds the fruit seems to have been globose or didymous, as in E. ff dissitiflora" (E. flavoviridis). Flowering material of E. flavida is to be sought. The type of E. flavida bears several persistent raceme axes 2-4 mm. long, these with indurated ovate bracts. The racemes are shorter than those of an average specimen of E. flavoviridis but can be matched in some specimens of that species. Eugenia hondurensis A. Molina, Ceiba 1: 261. 1951. E. crassi- folia A. Molina I.e. 3: 169. 1953. E. nicaraguensis Amsh. Act. Bot. Ne"erl. 5: 278. 1956. 468 FIELDIANA: BOTANY, VOLUME 29 This appears to be a distinctive species related to the series in- cluding E. salamensis Donn. Sm., E. purpusii Stand!., E. mexiae Standl., and E. tomentuiosa Standl., but differing from all these in having the raceme reduced to a single (terminal) flower. There is much variation in leaf-form in this species, even on the same plant, the leaves on vigorous shoots often becoming large and broad. Flowering specimens have an aspect quite unlike that of fruiting specimens because the maturing branchlets and foliage lose most of the reddish tomentum that is abundant and characteristic on the young growth, and the leaves become firm and smooth. The original specimens of E. hondurensis were from mature plants with half -grown fruit (Molina 2559, US !) ; flowering specimens were subsequently described under the name of E. crassifolia. Some of the latter, including the type of crassifolia (Molina 3031, US!), bore relatively broad obovate leaves. The Nicaraguan specimens de- scribed under the name of E. nicaraguensis were all in flower or in bud ; the suite included some with very large broad leaves, and some with smaller narrow leaves. The type (Standley 94.27, F!) is a flower- ing specimen with moderately broad leaves a little larger than those of the type of hondurensis. There seems to be no basis for recogniz- ing more than one species in this group, the variation being appar- ently individual and seasonal. Eugenia laevis Berg, Linnaea 27: 177. 1856. E. subverticillaris Berg, Linnaea 29: 235. 1858 [type from Hispaniola, Poiteau (LE!)]. Previously unknown from the North American continent, this rather distinctive species can now be reported from Yucatan. The specimens cited below agree precisely with specimens from Hispani- ola, viz. var. laevis. Especially noteworthy are the peculiarly and minutely verruculose leaves, the pale dibrachiate hairs of the ped- icels and the canescent hypanthium, the concentration of brownish hairs at the tips of the rather long calyx-lobes, which are otherwise glabrous on the inner surface. MEXICO: Yucatan: Izamal, G. F. Gaumer in 1888 (F, ex herb. Kew); 3 km. de El Cuyo en el camino a Colonia Yucatan, 7 Mar. 1956, 0. G. Enriquez 477 (MICH); entre Valladolid y Temozon, 7 June 1956, Enriquez 584 (MICH). The very recently described Eugenia calciphila Lundell (Wrightia 3: 11. Dec. 1961) is known only from the type: McVAUGH: TROPICAL AMERICAN MYRTACEAE 469 GUATEMALA: Pete*n: Dos Lagunas, in ramonal, E. Contreras 1541 (LL). This specimen is in young bud, with a few open flowers. The pubescence is less abundant than in the specimens from Yucatan and Hispaniola; the pedicels are nearly glabrous and the hypanthium merely strigose. I cannot find any other significant differences be- tween this plant and other specimens of E. laevis, and I regard E. calciphila as a synonym of that species. Eugenia laevis Berg, var. gaumeri (Standl.) McVaugh, comb, nov. E. gaumeri Standl. Field Mus. Bot. 81 28. 1930. E. lundellii Standl. Carnegie Inst. Wash. Publ. 461: 76. 1935. An ample series of well-prepared specimens from Pete"n, collected by C. L. Lundell and his assistant, Elias Contreras, and made avail- able to me through the courtesy of Dr. Lundell, makes it clear that Eugenia gaumeri and E. lundellii are not distinguishable unless by leaf shape, and probably not even by that feature. The original specimens of E. gaumeri (Kancabtsonot, Yucatan, Gaumer 23843!, the type) were described as having narrow leaves (1.3-1.8 cm. wide, 2-3 times as long as wide), the blades widest at the middle or below, narrowed to an obtuse apex, the very tip commonly rounded. In contrast, E. lundellii was described (Tuxpena, Campeche, Lundell 1130!, the type) as having elliptic or obovate-elliptic or rarely ob- long-obovate leaves (1-2.5 cm. wide, 3-4 cm. long), obtuse or rounded at the apex. I do not find these differences significant; broadly obovate leaves may often be found on the same plant with rather narrowly elliptic ones, and some specimens from Pete"n (e.g. Lundell 15574, Contreras 1441) are generally narrow-leaved, the foli- age about as in some specimens from Yucatan (e.g. Lundell 7933), and from British Honduras (e.g. Gentle 1202, 1745). The population named above as Eugenia laevis var. gaumeri seems to differ from var. laevis chiefly in the shape and size of the leaves; these in var. laevis are mostly 4.5-6.5 cm. long and rather narrowly and prominently acuminate; in var. gaumeri they are mostly 2-5 cm. long, at most broadly and bluntly acuminate. Additional specimens examined: MEXICO: Yucatan: Lundell 7416. GUATEMALA: Pete"n: Bartlett 12801; Contreras 26, 1362, 1368, 1369, 1370; Lundell 15518, 15527, 15975, 16489, 16824.— Baja Vera- paz: Standley 69775 (F). 470 FIELDIANA: BOTANY, VOLUME 29 Eugenia octopleura Krug & Urb. ex Urb. Bot. Jahrb. 19: 653. 1895. E. doubledayi Standl. Journ. Arnold Arb. 11: 36. 1930. E. koepperi Standl. Field Mus. Bot. 9: 320. 1940. The type of E. doubledayi (Honduras, Siguatepeque, Standley 56063} is a fruiting specimen collected in February. Very little color is observable in the scanty pubescence persisting on the leaves. Immature branchlets, however, are closely and densely coppery- pubescent, exactly as in flowering specimens. The type of E. koep- peri (Honduras, Mt. Cangrejal near La Ceiba, Yuncker et al. 8^00) is a specimen with developing inflorescences; the reddish color of the pubescence is evident on the lower leaf-surfaces. The leaves are bluntly acuminate or even obtuse, and the petioles are 5-8 mm. long. Superficially the specimen seems unlike specimens of E. double- dayi in which the leaves are usually prominently acuminate and the petioles 8-15 mm. long. Except for these features I cannot distin- guish the two so-called species. Certain Costa Rican specimens (Zarcero, 4500 feet, Austin Smith A239) are seemingly identical with flowering specimens of E. double- dayi. Other Costa Rican material obviously of the same species (La Palma de San Ramon, 1250 m., Brenes ^625; La Palma, Prov. S. Jos^, 1460 m., Tonduz [Herb. C. R. 12558, J. D. Smith 7382}) is matched perfectly by two syntypes of E. octopleura: Duss 2759, 3270. On the basis of this evidence, E. doubledayi and E. koepperi should pass into synonymy. 3. MARLIEREA Camb. Reference: Marlierea, Sect. Myrciopsis McVaugh, Mem. N. Y. Bot. Gard. 10: 79. 1958. Through an unintentional orthographic error, the name of this well-marked section was originally published as Myrcioides (I.e. 79). It was correctly cited as Sect. Myrciopsis on pages 84, 86 and 87 of the article in which it was first published. 4. MYRCIA DC. MISCELLANEOUS NOTES Myrcia splendens (Sw.) DC. Prodr. 3: 244. 1828. Myrtus splendens Sw. Prodr. 79. 1788. M. rufidula Schlecht. Linnaea 13: McVAUGH: TROPICAL AMERICAN MYRTACEAE 471 416. 1839. M. costa-ricensis Berg, Linnaea 27: 104. 1855. M. dis- color Berg, I.e. 111. M, oerstedeana Berg, I.e. 112. M. melanoclada Berg, I.e. 113. M. plicato-costata Berg, I.e. 114. M. sartoriana Berg, Linnaea 29: 220. 1858. M. longicaudata Lundell, Am. Midi. Nat. 29: 481. 1943. M. schippii Lundell, I.e. 482. Additional synonymy, and some comments on the distribution and taxonomy of this species, may be found in the Flora of Peru (13, pt. 4: 659-661. 1958). I noted there that Myrtus splendens Sw. perhaps was a synonym of Eugenia periplocifolia Jacq. Coll. 2: 108. 1788. I now doubt that the two are synonymous, for in the description of E. periplocifolia the plant is said to be "tota glabra"; one of the most characteristic features of Myrcia splendens is the abundant pubescence, readily seen with the unaided eye. In the West Indies Myrcia splendens is represented primarily by a form with ovate or ovate-lanceolate leaves 1.5-3 cm. wide, 3-7 cm. long, 1.7-2.8 times as long as wide. In Trinidad, in South America generally, and in Central America and Mexico, the leaves are pre- vailingly elliptic-lanceolate or elliptic, 6-12 cm. long and mostly 2.5-3 (-4) times as long as wide. After the examination of a large suite of specimens I am unable to find any consistent differences between North American and South American material. Through the kindness of the authorities at Copenhagen, I have seen the types of the five Costa Rican species based by Berg upon the collections of Oersted. With the exception of the type of M. pli- cato-costata, none of the specimens seems in any way exceptional for M. splendens. The Oersted specimen of M. plicato-costata (see Field Mus. Neg. 21035) has a few veins of the longer leaves impressed on the upper surface, a feature matched in a number of other Costa Rican specimens. The specimens are otherwise so similar to M. splendens that I doubt their specific distinctness. All are represented by photographs in the Chicago Natural History Museum series ex- cept for M. oerstedeana. This is a fruiting specimen with rather narrow, caudate leaves, the blades about 2.5 cm. wide, 8 cm. long; the pubescence is about that of an average specimen of M. splendens. A related species, Myrcia mollis (HBK.) DC., known chiefly from the eastern Cordillera of Colombia, has ovate and short-petiolate leaves that are often rugose and have the veins impressed on the upper surface. The pubescence is silky, more copious than usual in M. splendens, and extending even to the inner faces of the calyx- lobes. Similar plants occur as far north as Costa Rica; it may be that M. plicato-costata Berg, mentioned above, is akin to M. mollis. 472 FIELDIANA: BOTANY, VOLUME 29 The whole species-complex to which M . splendens belongs [in- cluding such species as M. fallax (Rich.) DC., M. mollis, M. acumi- nata (HBK.) DC.] is variable with respect to the amount and qual- ity of the pubescence. The branchlets and the inflorescence are appressed-strigose or -silky, or in extreme forms merely sparingly strigose, or at the other extreme almost hirsute (cf. Flora of Peru, pt. 4: 660-661). It has been the fashion in recent years to recognize in Mexico and Central America two species of Myrcia, one with spreading pubescence, the other appressed; it is true that there are extremes of pubescence, but I cannot find any correlated characters that would seem to support a specific segregation. In an occasional specimen from Honduras and Guatemala, about half the known collections from British Honduras, and in most speci- mens from Mexico, the hairs of the young branchlets and the inflores- cence are spreading and somewhat reddish. The branchlets often appear shaggy or velvety. Professor F. Miranda informs me that in Chiapas the conspicuous reddish unexpanded terminal buds make the plant easy to recognize in the field. This is Myrcia rufidula Schlecht., which I cannot distinguish in the herbarium except in the most subjective way from less distinctively hairy specimens. Aside from the abundant Mexican material (from Veracruz, Tabasco, Chi- apas) and some specimens from British Honduras, I have seen only two specimens from Honduras, one doubtful (sterile) specimen from Guatemala (Alta Verapaz, Coban, Standley 69529], and a recent collection from Pete"n (Machaquila, C. L. Lundell 16^09). The recently described Myrcia belizensis Lundell, Wrightia 2: 213. 1961, was described by the collector as a woody vine. The author also noted the exceptionally long petioles, these 5-9 mm. long, or about twice as long as those of M. splendens. The type, which I have seen through the courtesy of Dr. Lundell, is Gentle 6571, from near San Antonio, Toledo District, British Honduras. The speci- men is not to be distinguished from other Central American specimens of M . splendens, except that the long petioles and the relatively large panicles (up to 10 cm. long) suggest Myrcia fallax (Rich.) DC., a widespread and common species of tropical South America. In M. fallax, however, the inflorescence is usually relatively stout (the peduncle 2-2.5 mm. wide just below the first node) , whereas in the type of M . belizensis the thickest peduncle is no more than 1.5 mm. wide. Typical M. fallax may yet be found in British Honduras, as have so many other Myrtaceae of the Caribbean basin, but Gentle 6571 cannot be referred to it with any assurance. On the other McVAUGH: TROPICAL AMERICAN MYRTACEAE 473 hand there seems to be small reason to regard M. belizensis as a dis- tinct species, at least not until after more is known about the taxon- omy of the complex to which M . splendens and M. fallax belong. The "vining" habit I suspect is not that of a true liana, but the habit assumed by so many woody plants in tropical forests, viz. the development of long trailing branches that depend for support upon the branches of other trees and shrubs. Similar developments are known in other species of Myrcia. 5. MYRCIANTHES Berg SYNOPSIS OF THE GENUS (EASTERN SOUTH AMERICA EXCLUDED) The work of Berg in the mid-nineteenth century laid the founda- tions for all subsequent work on the American Myrtaceae, provided a rational scheme of classification for the many taxa that had been loosely grouped by DeCandolle in the tribe Myrteae, and established the existence of many genera and other more inclusive taxa that had previously been unrecognized. One such group, Berg's second major "Subtribe," the Eugenioideae, included 15 genera having in common a fleshy large solid embryo and a short radicle ("Embryo cotylis carnosis, discretis v. margine v. omnino conferruminatis; radicula abbreviata;" cf. Linnaea 27: 4. 1855). The largest genus, Eugenia, including 473 species recognized by Berg, was characterized by the possession of 4 calyx-lobes, and the embryo was described as "exalbuminosus, carnosus; cotylis saepissime margine v. omnino conferruminatis; radicula abbreviata." Within the genus Berg recognized 8 groups (Uniflorae, Biflorae, Racemosae, etc.) based on features of the inflorescence. The seventh such group, Dichotomae (species numbered 355-407), was described as having "Pedunculi bifidi v. simpliciter aut ramosissime dichotomi, 6- -mul- tiflori, flore in dichotomiis sessili, lateralibus saepissime pedicellatis." Most of the Dichotomae seem to have been known to Berg in flower only, but in at least two species (E. dicrana, Linnaea 27: 259. 1856, and E. dichotoma, I.e. 262) he described the embryo as consisting of 2 discrete unequal cotyledons. One of Berg's new genera, Myrcianthes (Linnaea 27: 315. 1856) was recognized by him as coordinate with Eugenia, and distinguished from that genus chiefly by the possession of 5 (instead of 4) calyx- lobes. The embryo of Myrcianthes was fully described: "exalbumino- sis, cotylis 2, discretis, carnosis, piano-con vexis; radicula brevi, sub- exserta; gemmula saepe distincta, inter cotylas latens." Myrcian- 474 FIELDIANA: BOTANY, VOLUME 29 thes was also described as having "Pedunculi axillares, 1-3-flori v. dichotomi." The four species included by Berg in this genus were all natives of Uruguay or southern Brazil. It now seems odd that Berg did not call attention to the obvious similarities between his Dichotomae, a subgroup of Eugenia, and his genus Myrcianthes. The inflorescence is the same in the two groups — and quite unlike anything else in Eugenia; the cotyledons are free and plano-convex in Myrcianthes, and also in the few species of Di- chotomae in which Berg had seen fruit; the calyx-lobes are 5 in Myrci- anthes (actually sometimes 4), and 4 (or sometimes 5, as pointed out by Berg for several species) in the Dichotomae. Whatever may have been the reason, Berg apparently did not consider the possibility that the Dichotomae might well be removed from Eugenia into an- other genus. Authors following Berg have at one time or another transferred most of the species of Dichotomae to other genera, includ- ing Anamomis Griseb., Luma A. Gray, Myrceugenella Kausel, Myrci- anthes Berg, Pseudanamomis Kausel, and Pseudomyrcianthes Kausel. Plants of this general affinity are frequent along an Antillean- Andean axis, from Argentina and Bolivia to Venezuela and Colom- bia, the West Indies, peninsular Florida, and in Central America and Mexico north to Tamaulipas and Durango. Approximately 40 species are known from this range. There are isolated and distinct, but clearly related groups of species in Chile, and in the South Bra- zilian-Uruguayan-Argentinian region. Recent work indicates that the Chilean species belong to distinct and mostly endemic genera. I have already advocated (in the Flora of Peru, 13, pt. 4: 746. 1958) that the concept of Myrcianthes be expanded to include numerous Andean and West Indian species and to include Anamomis Griseb. (1860). There is no general agreement on the taxonomic status of the species of eastern South America. Because of the superficial resem- blance between the inflorescences and flowers of Myrcianthes and its relatives, on the one hand, and those of certain genera of the sub- tribe Pimentinae (e.g. Blepharocalyx and Pseudocaryophyllus) on the other, it is difficult to place an unknown flowering specimen in the proper genus. Final disposition (even to subtribe) may depend upon the availability of mature fruit and seeds. Some species are known from flowering material only. In recent years two Bergian species of Eugenia have been transferred by Burret to Pseudocary- ophyllus, and one Bergian species of Myrcianthes by the same author to Psidium. Kausel (Ark. Bot. II. 3: 504. 1956) has proposed a new McVAUGH: TROPICAL AMERICAN MYRTACEAE 475 genus, Pseudomyrcianthes, to include approximately 15 species origi- nally placed in Eugenia by Cambessedes, DeCandolle, Berg and others. On the basis of embryo characters, however, Kausel assigns Pseudomyrcianthes to the Eugenioideae proper, proposing at the same time a coordinate group Plinioideae1 to include Myrcianthes and its close Andean and Antillean relatives (which he assigns to the genera Amyrsia and Anamomis, respectively). Legrand (Not. Syst. 15: 266-267. 1958) feels that too much sig- nificance has been attached to the dichasium as a fundamental taxo- nomic character. He points out that in Eugenia pyriformis (the type species of Kausel's Pseudomyrcianthes), there is frequently a reduction to a 1-flowered form resembling species in other subgeneric groups of Eugenia. He feels that the occasional occurrence of a 3-flowered dichasium in a normally 1-flowered species (e.g. E. arbuti- folia Berg, a synonym of E. obtusifolia Camb.) does not justify its inclusion in Berg's group Dichotomae. He reminds us that Myrceu- genia ribeireana (Berg) Legr. & Kaus. was originally placed in the Dichotomae, in the absence of fruit, on the basis of its forked in- florescence. For want of adequate fruiting material of many of the species, and in view of the lack of agreement on generic limits in the spe- cies of eastern South America, I do not attempt to present even a summary of them at this time. Two species of Myrcianthes, namely M. cisplatensis (Camb.) Berg [including the genotype, M. apiculata Berg], and M. gigantea (Legrand) Legrand, seem to be generally accepted as members of the genus in good standing. Excluded from Myrcianthes have been M. brunnea Berg [Psidium ovale (Spreng.) Burret] and M. edulis Berg [considered by Legrand, An. Mus. Hist. Nat. Montevideo II. 4, pt. 11: 59. 1936, to be related rather to the genus Hexachlamys Berg (Eugenia subg. Myrcichlamys Legr.)]. Several species included by Berg among the Dichotomae (Linnaea 27: 246-278. 1856; Fl. Bras. 14, pt. 1: 306-312. 1857; I.e. 586-587. 1859) may be investigated as possible members of Myrcianthes; as far as I know their generic standing has not been challenged : E. ex- coriata Berg, E. imbricata Berg, E. ischnosceles Berg, E. itajurensis Camb., E. subamplexicaulis DC., E. ternatifolia Camb. and E. vimi- nalis Berg. Others (E. leandreana Berg, E. mutabilis Berg) have been transferred to Pseudocaryophyllus by Burret, and several (E. adamantium Camb., E. arbutifolia Berg, E. decumbens Camb., E. 1 Mentioned by name only; not properly described and not given any particu- lar status. 476 FIELDIANA: BOTANY, VOLUME 29 kochiana DC., E. lutescens Camb., E. piresiana Camb., E. pohliana DC., E. pruniformis Camb., E. pyriformis Camb.) have been assigned at least provisionally to Pseudomyrcianthes by Kausel. Legrand has pointed out (Not. Syst. 15: 269. 1958) that E. sylvatica Camb. (Pseudomyrcianthes cambessedeana Kausel) was wrongly assigned by Berg to the Dichotomae; the inflorescence is a loose raceme like that of £". florida DC. and E. gardneriana Berg. In the following pages is presented a synopsis of the species of Myrcianthes occurring along the Andes and in northern South Amer- ica, the West Indies and North America. The limits of the genus as I understand them were set forth in the Flora of Peru (Field Mus. Bot. 13, pt. 4: 746. 1958). As stated there, I believe the distinction between Anamomis and other genera, on the basis of the presence or absence of a plumule, to be a specious one; I have repeatedly observed a plumule in the mature seeds of Anamomis (Myrcianthes) fragrans and similarly in the seeds of Andean and other species of Myrcianthes. As far as I know there is no species in eastern South America that occurs also along the Andean-Antillean axis, except that a form of M. cisplatensis is found in northern Argentina. The following key summarizes the pertinent features of Myrcianthes and other genera having the inflorescence a dichasium and the two cotyledons distinct and plano-convex: Radicle about as long as the cotyledons; cotyledons plano-convex, fleshy but thin; Chile and adjacent Argentina Luma (incl. Myrceugenella). Radicle less than half as long as the cotyledon; cotyledons plumply plano-convex. Calyx-lobes caducous; inflorescence often irregularly dichotomous or umbelli- form; ovules 5-6; cotyledons united about one-third their length; 1 species, West Indies and Venezuela Pseudanamomis. Calyx-lobes persistent; inflorescence regularly dichotomous with the terminal flowers sessile in the forks; ovules usually more than 6; cotyledons united at the base only. Placenta shield-shaped with 6-12 marginal and inwardly directed ovules; calyx-lobes 5; Chile Reichea. Placenta wartlike or horseshoe-shaped, with usually 8-20 radially directed ovules in a subcapitate group; calyx-lobes usually 4 Myrcianthes. NOMENCLATURE The name Luma A. Gray (1854) is apparently the oldest name applied to any plant of this alliance. According to the recent studies by Kausel on the Myrtaceae of Chile, the seven species referred by Gray to Luma are to be distributed among three genera in two sub- tribes (see Ark. Bot. II. 3: 503. 1956, and an earlier paper, Rev. Arg. Agron. 9: 42. 1942). If Luma be typified by L. chequen (Mol.) McVAUGH: TROPICAL AMERICAN MYRTACEAE 477 A. Gray, var. a, as suggested by Kausel (Lilloa 13: 127. 1948) and accepted by McVaugh (Taxon 5: 142. 1956), then Myrceugenella Kausel (1942) falls into the synonymy of Luma. Five of the remain- ing six species of Luma (in the sense of Gray) are to be referred to Myrceugenia Berg (published 1855), according to Kausel (I.e. 50. 1942). Because of this preponderance of species of "Myrceugenia" in the original Luma, Burret in 1941 reduced Myrceugenia to the synonymy of Luma and published a number of new combinations and new species in the latter genus (Notizbl. Bot. Gart. Berl. 15: 522-535; Repert. Sp. Nov. 50: 50-55). Both Kausel (Lilloa 13: 125- 129. 1947) and Legrand (Darwiniana 11: 302. 1957) have argued against the course taken by Burret. There is no nomenclatural conflict between Luma A. Gray (1854) and Myrcianthes Berg (1856) as long as the two genera are held to be taxonomically distinct. Apparently a good case can be made for their distinctness. The species of Luma (Myrceugenella in the sense of Kausel) are all confined to Chile, or the Chilean-Argentinian region, i.e. they are geographically separated from all known species of Myrcianthes. They are like many species of Myrcianthes in having the flowers 4-merous, the bracteoles caducous and the dichasium 3-flowered or the flowers occasionally solitary. In Myrceugenella, however, the cotyledons are described by Kausel as "piano-con vexis, carnosis sed tenuibus," the radicle is about as long as the cotyledons (illustrated in Rev. Arg. Agron. 9: 44, fig. 1, B. 1942), and the insertion of the ovules is said to be basal or sub-basal (Lilloa 13: 148. 1947). On the basis of the long radicle and the relatively thin cotyledons, Kausel now regards Myrceugenella as a member of the Myrciinae ("Myrcioideae") (Ark. Bot. II. 3: 503. 1956). I cannot comment upon this disposition of the genus, but in any event Myrceugenella, or Luma, appears to be a small endemic Chilean group that for the present may be regarded as a distinct genus. The name Amyrsia Raf. (1838) has recently been taken up by Kausel (Ark. Bot. II. 3: 512-514. 1956) for several Andean species that I should refer to Myrcianthes. Rafinesque's genus Amyrsia was a mixture (cf. Taxon 5: 137. 1956), including three species of Myrcianthes in the present sense, and one species of Myrteola (sub- tribe Pimentinae) . Rafinesque gave no indication of any preference for one species or another as genotype, and my choice of lectotype was made on the basis of the original description, which permitted the positive identification of the Myrteola element because Rafinesque 478 FIELDIANA: BOTANY, VOLUME 29 FIG. 12. Dichasium and embryo in Myrcianthes. A, Seed of M. karsteniana, Garcia 94 (ca. X 17); inner face of one cotyledon showing position of hypocotyl and radicle, and small, apparently bilobed plumule at right. B, The same (ca. X 7); outer coats removed to show plano-convex cotyledons and projecting radicle. C, Seven-flowered dichasium in M. fragrans, Steere 2976 (X %). compared the genus to "Pimentus," from which he separated it be- cause of the bilocular ovary, and the many seeds "as in Myrtus." None of the species now referred to Myrcianthes could fit this de- scription, as ordinarily but one seed develops in each fruit. If Amyr- sia be typified on the basis of the Myrteola element, the genus Myrteola Berg (1856) may well be proposed for conservation, as was done in Taxon 5: 163-164. 1956. It would seem most unfortu- nate to preserve the poorly founded genus Amyrsia in any sense whatever, especially since both of its constituent elements have since become known under other names. If Kausel's proposal be McVAUGH: TROPICAL AMERICAN MYRTACEAE 479 accepted, and Amyrsia typified on the basis of the Myrcianthes ele- ment, then the latter name should be proposed for conservation. Myrcianthes Berg, Linnaea 27: 315. 1856. Anamomis Griseb. Fl. Brit. W. Ind. 240. 1860. Small or medium-sized trees with opposite or ternate leaves; flowers in 1-7- flowered (or sometimes 15-31-flowered) dichasia, these usually solitary and axillary, sometimes so numerous in the upper axils as to form a compound in- florescence, or modified into a terminal myrcioid panicle bearing the flowers in small dichasia at the tips; central flowers of the dichasia usually sessile; calyx- lobes and petals 4 (in some species regularly 5), or an occasional flower 5-merous; ovary usually 2-locular, the numerous (usually 8-20) ovules radiating from a centrally affixed placenta; seeds maturing 1 (-4); cotyledons 2, distinct, plano- convex; radicle terete, half as long as the cotyledons or less; plumule evident in the mature seed. Lectotype species: M. apiculata Berg [=M. cisplatensis (Camb.) Berg]. See Taxon 5: 143. 30 July 1956. In the following pages are contrasted and listed all the species known to me from continental North America, from South America north of the Amazon River, and from Andean South America except for the approximately 20 species treated in the Flora of Peru (Field Mus. Bot. 13, pt. 4, no. 2: 745-775. 1958). These Peruvian species are omitted, or in a few instances mentioned for comparison. Spe- cies of southeastern South America are not included; West Indian species are treated somewhat cursorily, following Urban (Bot. Jahrb. 19: 662-665. 1895). Certain species listed below as members of the genus Eugenia are probably to be referred eventually to Myrcianthes, but as these are unknown in the fruiting condition any formal transfers would seem to be premature. Inflorescence, and often the young herbage, softly tomentose or velutinous with rufous, brown or pale hairs; petioles stout, distinct from the blades, (4-) 10- 15 (-20) mm. long. Calyx-lobes tomentulose on both sides. Disk 3-4 mm. wide; buds 4-5 mm. long; flowers up to 15, the dichasium often 3 times forked; ovules 5; Bolivia M. pearcei. Disk 5-8 mm. wide; buds 7-10 mm. long; flowers mostly 3-7; ovules 15-30. Tomentum gray or yellowish white; disk 5 mm. wide; stamens 7-8 mm. long; northern Peru M. lanosa. Tomentum red; disk 6.5-8 mm. wide; stamens 10-14 mm. long. Calyx-lobes 5, 5-9 mm. wide; Argentina M. callicoma. Calyx-lobes 4, 4-5 mm. wide; Colombia M. borealis. 480 FIELDIANA: BOTANY, VOLUME 29 Calyx-lobes glabrous within; inflorescence irregularly 2-3 times forked; buds 5-6 mm. long; disk 4-5 mm. wide; ovules 10-11, marginal on the placenta; Ecuador M. irregularis. Inflorescence and herbage glabrous or often strigose or silky-strigose, the hairs often white or very pale, most abundant and conspicuous on the outer surface of the hypanthium; petioles often not well set off from the cuneate base of the blade. Leaves all or nearly all ternate; northern Andes of Venezuela and Colombia. Coarse plants; leaves 2-3 (-5) cm. long; buds 8-9 mm. long M. crebrifolia. Smaller-leaved plants; leaves 0.6-1.8 cm. long; buds 4-5 mm. long. Peduncles 3-flowered; leaves spatulate M. ternifolia. Peduncles 1-flowered; leaves obovate to suborbicular E. triquetra. Leaves opposite [for plants of Ecuador, Peru and Bolivia, see Field Mus. Bot. 13, pt. 4: 745-775. 1958]. Plants of the Guianas, Venezuela, Colombia, West Indies or North America. Inflorescence decompound, many-flowered, the forking branches forming a terminal leafless panicle. Leaves 2-4 cm. long, rigidly coriaceous; panicles 5 cm. long or less beyond the last leaves; southern Colombia and Ecuador E. pycnantha. Leaves 6-12 cm. long, thinner; panicles 6-11 cm. long; Venezuela. M. karsteniana. Inflorescence a regular axillary dichasium, once-, twice- or thrice-forked, the flowers thus mostly 3-7, occasionally only 1 or as many as 15. Inflorescence stout, glabrous or essentially so, the peduncle 2-3.5 (-5) mm. wide near the summit. Flowers very large, the buds 1.7-2 cm. long; style 2.5 cm. long; fruit 4-5 cm. in diameter; Guianas M. prodigiosa. Flowers smaller, the buds less than 1 cm. long; style about 1 cm. long or less; fruit as far as known 2 cm. in diameter or usually less; Andes from Peru northward; Panama and Costa Rica. Leaves rounded, 3 cm. long or less, the lateral veins impressed above; staminal disk glabrous; high mountains of ?Costa Rica and west- ern Panama (Allen 1563) ?E. rigidissima. Leaves elliptic to suborbicular, mostly 4-6 cm. long, the veins not impressed; disk bristly-pubescent; style 10-11 mm. long; Andes, Colombia to Peru M. rhopaloides. Inflorescence more slender, variously pubescent (especially the hypan- thium), or glabrous, the peduncle 2 mm. wide at summit or usually less; flowers smaller, the style 8.5 mm. long or less. Peduncles minutely bristly-hispidulous with erect hairs. Peduncles and branchlets similarly hispidulous; leaves often elliptic, sometimes obovate; dichasium often 7-flowered; Costa Rica and western Panama M.fragrans var. hispidula. Peduncles hispidulous, the branchlets glabrous; leaves mostly obo- vate; dichasium usually 3-flowered; Venezuela. . .M. compressa. Peduncles glabrous, appressed-pubescent or appressed-silky. Flowers 3 in a dichasium, all sessile; hypanthium and outer surface of the calyx-lobes whitened, silky; Santa Marta region, Colom- bia. . . .M. sessilis. McVAUGH: TROPICAL AMERICAN MYRTACEAE 481 Flowers 3-7 or -15, the lateral ones on definite slender branches; calyx-lobes glabrous or essentially so without. Hypanthium rather densely silky-strigose at least at base, often whitened by the dense covering of hairs; inflorescence in gen- eral somewhat strigose or hispidulous, sometimes glabrate. Flowers usually 3; peduncle mostly 2 cm. long or less, 1.5-2.5 mm. wide at summit; style 7-8.5 mm. long; southern Colombia and Ecuador M. hallii. Flowers often 7, sometimes 3; peduncle 2-4 (-6) cm. long, 1.2- 1.5 mm. wide at apex; style mostly 5-6.5 mm. long; often in lowlands, West Indies and northern Colombia and Vene- zuela to Mexico and Florida vars. of M. fragrans. Hypanthium glabrous, or sparingly appressed-pubescent especially at base; peduncles glabrous; highlands, Costa Rica, Colombia and Venezuela; Ecuador. Peduncles 1-flowered; southern Colombia and Ecuador. E. orthostemon. Peduncles mostly 3-7- (-15) flowered. Peduncles mostly 7-flowered, 3-4 cm. long; Colombia to west- ern and northern Venezuela; see also E. rondonensis. M. dugandii. Peduncles mostly 3-flowered, 2.5 cm. long or less. Costa Rica M. storkii. Colombia (region of Bogota) M. leucoxyla. Plants of Argentina (including southern Bolivia). Flowers solitary; hypanthium whitened without by the dense covering of hairs; calyx-lobes 4; style 5-8 mm. long M. mato. Flowers 3 (occasionally 1 or 7) in a dichasium; hypanthium glabrous or sparingly stiff-hairy without. Calyx-lobes 5; style 3.5-5.5 mm. long M. cisplatensis. Calyx-lobes 4; style 7-8 mm. long M. pseudo-mato. Eugenia alaternifolia [sphalm. alternifolia] Benth. PI. Hartw. 175. 1845. Colombia. Cf. M. rhopaloides (see Flora of Peru, 13, pt. 4: 773). Eugenia andina Berg, Linnaea 27: 274. 1856. Peru. Unknown (see Flora of Peru, I.e. 774). Eugenia aromatica Berg, Linnaea 27: 263. 1856. Venezuela, Caracas, Moritz 792 (Berlin). Unknown. Myrcianthes borealis McVaugh, sp. nov. Arbor rufo-tomentosa, ramulis inflorescentiisque, et foliis juvenilibus, dense pilis contortis usque ad 0.5-1.3 mm. longis obtectis; folia elliptica vel obovata, 5-9 cm. longa, 2.5-7 cm. lata, petiolis 9-17 mm. longis, marginibus callosis, 482 FIELDIANA: BOTANY, VOLUME 29 aliquantum revolutis; flores 3 (-7), pedunculo 3-5 cm. longo, 2.5-3 mm. crasso; alabastra ut videtur 10 mm. longa; calycis lobi 4, utraque superficie tomentulosi, 4-5 mm. lati; discus 6.5-7.5 mm. latus; stylus 7 mm. longus, pilosus; stamina ca. 300, filamentis 10-12 mm. longis; ovula quoque loculo ca. 30. Tree 2.5-3.5 m. high, the young leaves, branchlets and inflorescence rusty- tomentulose, the crisped, curled hairs often more than 0.5 mm. long, some to 1.3 mm.; leaves elliptic or obovate, rounded or retuse, rounded to the base or some- what acutely narrowed and the margins finally cuneately decurrent on the stout channeled petiole 2-2.5 mm. thick, 9-17 mm. long; blades 2.5-7 cm. wide, 5- 9 cm. long, 1.3-2 times as long as wide, lustrous and glabrate above except the hairy sulcate midvein, paler and somewhat pilose beneath; lateral veins 7-9 on each side, somewhat impressed above and prominent beneath, diminishing distally to a slightly less prominent marginal vein 2-7 mm. from the margin, and arched between the laterals; glands inconspicuous; margins callose, somewhat revolute; flowers usually 3 (-7), the peduncle 3-5 cm. long, 2.5-3 mm. wide near the apex, the branches 6-16 mm. long; central flower sessile, or the axis extended 8-11 mm. above the fork; bracts narrowly triangular, acute, 1.8 mm. wide at base, 6 mm. long; bracteoles linear, 2 mm. long, deciduous, with the bracts, at anthesis; buds probably about 10 mm. long; hypanthium cup-shaped, 5 mm. long, 6 mm. wide at mouth; calyx-lobes 4, broadly rounded, tomentulose on both sides, 2-3 mm. long, 4-5 mm. wide; disk 6.5-7.5 mm. wide, the staminal ring velutinous, the glabrous center 3 mm. wide; style 7 mm. long, pilose; stamens about 300, the filaments 10-12 mm. long; locules 2, the ovules about 30 in each locule, radiate in a subcapitate group affixed to the central dissepiment. — Univ. of Mich. Negs. 1331, 1347. COLOMBIA: Norte de Santander: Eastern Cordillera, road from Pamplona to Toledo, on the Maracaibo-Orinoco drainage, alt. 2500- 2800 m., thickets along stream, 28 Feb. 1927, Kittip & Smith 19882 (US, type; A; GH). Myrcianthes callicoma McVaugh, sp. nov. Arbor 6-8-metralis, rufo-tomentosa; folia elliptica vel ovata, 5-10 cm. longa, 1.5-5 cm. lata, obtusa vel obscure acuminata, glabrata, petiolis 6-10 mm. longis; venis utroque latere 10-12, supra inconspicuis, subtus paullo elevatis, venulis nee incrassatis; dichasio 3-floro (vel floribus solitariis), pedunculo 1-2.5 cm. longo, 2-2.5 mm. crasso, floribus in dichotomis sessilibus, alabastris 7-9 mm. longis; calycis lobis 5, utrinque tomentosis, 5-9 mm. latis; stylo 8 mm. longo; disco 8 mm. lato; stamina ca. 300, filamentis 10-14 mm. longis; ovula quoque loculo ca. 15-20. A tree 6-8 m. high, 35 cm. in diameter, the young growth and the inflor- escence densely tomentose with soft crisped rusty-red hairs, these very short or on the margins of the calyx up to 1 mm. long; leaves and branchlets glabrate or with some minute crisped hairs persisting; leaves elliptic or ovate, (1.5-) 2.5-5 cm. wide, (3-) 5-10 cm. long, 1.6-2.5 times as long as wide, bluntly pointed or obscurely acuminate, rounded to the base, the markedly cartilaginous and revolute margins narrowly prolonged and long-decurrent on the compressed sulcate petiole 1.5-2 mm. wide, 6-10 mm. long; midvein impressed above, prom- McVAUGH: TROPICAL AMERICAN MYRTACEAE 483 inent beneath; lateral veins 10-12 pairs with some additional weaker intermediate veins, inconspicuous above, somewhat raised beneath; marginal vein weaker than the laterals, irregularly and asymmetrically arched between them, 1-5 mm. from the margin; blades smooth on both surfaces, concolorous, slightly more glossy above, obscurely and finely gland-dotted; inflorescence an axillary dicha- sium, 3-flowered (or often 1-flowered), the peduncle 1-2.5 cm. long, compressed and enlarged distally and there 2-2.5 mm. wide; flowers sessile, or the lateral ones in the dichasium on pedicels up to 8 mm. long; bracts and bracteoles deciduous before anthesis; buds 7-9 mm. long, broadly obovoid or subglobose; hypanthium broadly obconic with straight sides 4-5 mm. long; calyx lobes 5, unequal (1 or 2 smaller than the others), tomentose on both sides, broadly rounded and often narrowed just above the base, 3-6 mm. long, 5-9 mm. wide; disk 8 mm. wide, the staminal ring hairy, the glabrous or pubescent central portion 4 mm. wide; style 8 mm. long; stamens about 300, 10-14 mm. long, the anthers 0.7-0.8 mm. long; petals obovate, about 7 mm. wide, 12 mm. long, pubescent near the tips; ovary bilocular, the ovules about 17 (15-20?) in each locule, radiate in a sub- capitate group attached above the middle of the central partition; fruit not seen. —Univ. of Mich. Neg. 1345. ARGENTINA: Tucuman: Dept. Tafi, Cerro de Taficillo, 1600 m. elevation, in quebradas, Dec. 7, 1929, S. Venturi 9859 (GH, type); same locality, VenturilOOSS, Jan. 1930 (GH). Myrtus coccolobaefolia HBK. Nov. Gen. & Sp. 6: 139 [folio ed. 110]. 1823. Colombia. Referred in Flora of Peru, pt. 4: 772 to Myrcianthes rhopaloides. Myrcianthes compressa (HBK.) McVaugh, Field Mus. Bot. 13, pt. 4: 754. 1958. Myrtus compressa HBK. Nov. Gen. & Sp. 6: 135 (folio ed. p. 107). 1823. Often nearly glabrous, the young shoots finely pale-strigose, the peduncles minutely bristly pubescent (sometimes glabrate), the petals and calyx-lobes ciliate, the hypanthium sometimes sparsely strigose; leaves obovate, up to 4 cm. wide and 6 cm. long, obtuse or rounded at the tips (on flowering branchlets often acute at both ends and more or less elliptic), the base cuneate and the margins decurrent on the channeled petiole 3-4 mm. long; blades lustrous above, smooth, finely gland-dotted especially beneath; inflorescence 3-flowered (rarely 7-flowered), the peduncle 2-3.5 cm. long, 1-1.7 mm. wide near summit, often compressed and 2-edged; style 6-7 mm. long; stamens about 175; ovules 15-20 (-25) in each locule. The plant described above bears considerable resemblance to the inclusive Myrcianthes fragrans, differing chiefly in the usually 3-flow- ered dichasium, the nearly glabrous inflorescence (except that the peduncle is usually minutely bristly-pubescent), the cuneate obovate leaves. With some doubt I refer this to M. compressa, originally described from near Cajamarca, Peru. Nothing like the type has since been found in Peru, but a number of collections from the moun- 484 FIELDIANA: BOTANY, VOLUME 29 tains near Caracas, Venezuela, appear to represent the same species and are cited below. A photograph of the type (Field Mus. Neg. 36879) shows the characteristic leaf-outline, and my notes on the type of Myrtus compressa, made in Paris in 1954, indicated that this specimen has the same minutely bristly peduncle, a feature unusual in Myrcianthes. I have never seen any plants from the Ecuadorean- Peruvian region that resembled Venezuelan plants in this respect. It is entirely possible that the type of Myrtus compressa came from somewhere in Venezuela and was later mistakenly supposed by Kunth to have come from Peru. VENEZUELA: Margen de la selva de las Flores, 1619 m., L. Wil- liams 10095 (A, F); between Agua Negra and El Junquito, 1900 m., Pittier 13822 (F, US); Galipan bei Caracas, 6000 ft., 0. Kuntze 1575 (US); Caracas, H. M. Curran & M. Raman 1124 (NY, US), Karsten s.n. (F, GH). COUNTRY UNKNOWN: Paramo de Cisne, Andre 4307, 29 Oct. 1876 (F, GH). Myrcianthes crebrifolia (Steyerm.) McVaugh, comb. nov. Eu- genia crebrifolia Steyerm. Fieldiana, Bot. 28: 1011. 1957. As suggested by Steyermark, and as indicated in the key above, this species is to be compared with Eugenia ternifolia Berg. It is also strikingly similar in most respects to M. rhopaloides, from which it seems to differ chiefly in the ternate arrangement of the leaves and in the very short fleshy petioles 2-4 mm. long and about as thick. In addition to the two collections cited by Steyermark, I should refer to this species the following: COLOMBIA: Santander: Vicinity of La Baja, 3200 m., Killip & Smith 18154 (US). — Norte de Santander: Between Mutiscua and Pamplona, 3100 m., Killip & Smith 19743 (GH). Myrcianthes dugandii (Standl.) McVaugh, comb. nov. Eu- genia dugandii Standl. Trop. Woods 52: 28. 1937. Nearly glabrous, the young herbage somewhat appressed-pubescent, the hypanthium glabrous or minutely appressed-pubescent, the calyx-lobes densely appressed-silky within, the disk hairy; leaves rigidly coriaceous, smooth and glossy above, 2-4.5 cm. wide, 3.5-7.5 cm. long, 1.3-2 times as long as wide, obo- vate or elliptic, rarely ovate, rounded or retuse at tip or obtusely pointed, the base abruptly or gradually rounded, the margins finally often cuneately decurrent on the stout petiole 1-2 mm. thick, 3-8 mm. long; inflorescence mostly 7-flowered, varying to 3-flowered or 15-flowered on the same plant; peduncles (1.5-) 3-4 (-5) McVAUGH: TROPICAL AMERICAN MYRTACEAE 485 cm. long, 1-1.7 (-2) mm. wide near summit, the lateral branches up to 1 cm. long (or nearly 2 cm. long in 15-flowered dichasia); calyx-lobes 4; disk about 3 mm. wide; style 6.5 mm. long; locules 2; ovules (12-) 15-20 (-25); fruit orange; cotyle- dons 2, plano-convex; plumule present. — Univ. of Mich. Neg. 1334. Few specimens that can definitely be referred to Myrcianthes fra- grans have been collected in South America. Many specimens from the Andes of Colombia and Venezuela, however, are similar to M. fragrans except that they are glabrous or essentially so throughout, and the dichasia tend to be 7- or 15-flowered, rather than 3-7-flow- ered as in most populations of M. fragrans. This glabrous and rather abundantly flowered plant is Eugenia dugandii Standl. It is known mostly at middle elevations, from 1500 to 2200 meters. COLOMBIA: Antioquia: Medellin, A. Dugand G. 904 (F, type); Bello, W. A. Archer 233 (US); Copacabana, Bro. Daniel 269 (US).— Tolima: Rio Paez, Lehmann 5818 (GH, US).— Huila: Cordillera Oriental, 25 km. southeast of La Bodega, E. L. Little 9056 (MICH). VENEZUELA: Me"rida: Aristeguieta 3332 (MICH), Bernardi 283 (NY), John 708 (US), 1049 (GH, US), Steyermark 55955 (F), Ta- mayo 2431 (US). — Sucre: Headwaters of Rio de Amana, Steyermark 62734 (F).— Monagas: Cerro Guacharo, Steyermark 62338 (F). In the specimens from Sucre and Monagas, the leaves are thinner and more ovate than usual, in these features suggesting West Indian specimens of M. fragrans. Myrcianthes fragrans (Sw.) McVaugh, comb, nov., var. fra- grans. Myrtus fragrans Sw. Prodr. 79. 1788. Eugenia punctata Vahl, Symb. 3: 65. 1794 (type from Santa Cruz, von Rohr, West). E. fra- grans (Sw.) Willd. Sp. PI. 2: 964. 1800. Anamomis fragrans (Sw.) Griseb. Fl. Brit. W. Ind. 240. 1860. E. triflora Sesse* & Moc. Nat- uraleza II. 1: app. 83. 1888 (type from Cuernavaca, Morelos). M. biflora, sensu Sess<§ & Moc. Fl. Mex. ed. 2. 125. 1894 (from Cordoba, Veracruz). Myrcia seleriana Donn.-Sm. Bot. Gaz. 27: 332. 1899 (type from Chacula, Guatemala, Seler 3169). E. steyermarkii Standl. Field Mus. Bot. 22: 360. 1940 (type from Guatemala near Volcan de Tacana, Steyermark 36210). E. lopeziana A. Molina, Ceiba 3: 170. 1953 (type from Honduras, Molina 3007). Leaves nearly glabrous, elliptic to ovate or obovate, 2.5-6.5 cm. long, usually about twice as long as wide, short-petioled; inflorescence more or less appressed- pubescent, the hypanthium often densely white-hairy and contrasting with the green glabrous calyx-lobes; dichasia 3- or often 7-flowered, the peduncles 2-4 (-6) 486 FIELDIANA: BOTANY, VOLUME 29 cm. long, (0.7-) 1.2-1.5 mm. wide at apex; flowers small (buds 3-4.5 mm. long, styles 6.5 mm. long or less, disk 3-3.5 mm. wide, calyx-lobes 2.5-3.5 mm. wide at base). What appears to comprise a single species, Myrcianthes fragrans, is centered geographically in the Caribbean region. The nomen- clatural type was from Jamaica. Members of the complex occur naturally on all the Greater Antilles, and to a lesser extent on other islands farther east; in the Bahamas and Florida; rather abundantly in eastern Mexico from Tamaulipas to Veracruz and in the Yucatan peninsula, in Central America and in northern Venezuela. A few collections from western and southwestern Mexico (state of Mexico to Durango) apparently belong to the same species-complex but are hard to classify because of the small number of specimens available. Additional small populations in eastern Mexico (San Luis Potosi and Tamaulipas), in Florida (Eugenia dicrana Berg, Anamomis simpsonii Small), in the Bahamas (Anamomis lucayana Britton), and larger populations in the Greater Antilles, are evidently inter- related in a way not yet understood. It would seem premature to assign varietal or subspecific names to most of these populations until their taxonomy can be more adequately worked out. Urban said of this species many years ago, "Typus polymorphus in futuro in varietates dividendus nee vix in species distinctas disjungendus." In accordance with the letter of the nomenclatural Code, a pedant may argue that Myrtus fragrans Sw. must be rejected because it was superfluous when published and therefore illegitimate. It is true that in the Prodromus Swartz cited in the synonymy of M. fragrans, without any question or qualification, "Eugenia montana Aubl. guian. 495. t. 195." In a later publication, however (Fl. Ind. Occ. 2: 914. 1800), Swartz included Eugenia montana Aubl. in the synonymy of M. fragrans with a query, and with the statement on p. 916: "Eu- genia montana Aubl. (I.e.) meae [i.e. to Myrtus fragrans] simillima est; quam vero color fructus hujus mini sit ignotus, nihil pro certo affirmare possum." I think it may be reasonably argued that Swartz never was really sure of the identity of Eugenia montana Aubl., and that even in the Prodromus his intention was merely to compare with his own new species an Aublet species that he thought might be the same. Limitations of space in the Prodromus presumably precluded the full explanation given later in the Flora. I do not see any indication that Swartz intended to replace Aublet's name by a new one; his early judgment, and also his more matured opinion, was that the Jamaican Myrtus fragrans was an independent species. I do not think that a well-established name should be upset on a McVAUGH: TROPICAL AMERICAN MYRTACEAE 487 legalistic interpretation of a technicality, perhaps merely because Swartz, in ignorance of the International Code of Nomenclature, neglected in 1788 to insert the question-mark that he included in 1800. Among the insular populations in the West Indies, it seems there are two principal tendencies: One toward a 3-flowered dichasium and smooth, elliptic, broad and obtusely pointed leaves with short petioles (e.g. in Jamaica) , and another toward a 7-flowered dichasium and glandular-verrucose, oblong-obovate, obtuse or rounded or re- tuse leaves with longer petioles (e.g. Eugenia granulata Berg, in Cuba and Hispaniola). Many exceptions to these generalizations may be observed, and it is evident that a revision of West Indian Myrcianthes is still to be desired. Most specimens of the fragrans group from Central America or Mexico, however, can be referred to one or the other of the above types. A majority of the specimens from the continent of North America, and a few from northern South America, bear a strong resemblance to the plant of Jamaica (i.e., M. fragrans var. fragrans) except the petioles are longer (often 4-6 mm. as against 2-4 mm. in the Jamaican plant) and the dichasia are often 7-flowered (usually 3-flowered in the Jamaican plant); the leaf -blades in Central American plants may be narrower than those of Jamaica, where the blades are often 1.3-2 times as long as wide. The following specimens are of the general type of M. fragrans var. fragrans: MEXICO: Veracruz: Rancho La Palmilla, Purpus 16350 (A, F); Zacuapan, Purpus 7518 (A); Mirador, Liebmann 3971 (US). — Chi- apas: Mt. Pasitar [Paxtal], Matuda 1*68 (MICH, US); Pinabeto, Motozintla, Matuda 154-81 (F, an exceptionally lanose form). BRITISH HONDURAS: Guatemala boundary, Camp 33, Schipp 1240 (A, F, GH, MICH). GUATEMALA: Huehuetenango: Maxbal, Steyermark 48904. (A, F); Chacula, Seler 3169 (a narrow-leaved form, type of Myrcia seleriana; US). — San Marcos: Volcan Tacana, Steyermark 36210 (F, type of E. steyermarkii) . HONDURAS: Comayagua, Edwards P592 (A, F, US). — Morazan: Rio Guarabuqui, Molina 3007 (F; US, type of E. lopeziana; Univ. of Mich. Negs. 1326, 1327). NICARAGUA: Near Granada, summit Mt. Mombacho, Grant 835 (A, F). 488 FIELDIANA: BOTANY, VOLUME 29 COSTA RICA: Alajuela: Zarcero, Austin Smith 608, 2687, 4,173, 4185 (all F); Zapote, Smith NY 57 9 (A, F); Llano Bonito, Smith P2391 (UC, A). Copey, Tonduz 11741 (GH). Piedra Blanca, Iscasu, F. Solis R. 480 (F). COLOMBIA: Sta. Marta, near Escalera de los Indies, H. H. Smith 434. (GH); near Masinga, Smith 402 (A, F, GH, MICH, PH, US).— Norte de Santander: Pamplona-Toledo, on divide, Killip & Smith 19917 (A, GH, US). VENEZUELA: Paraguana Peninsula, Cerro Sta. Ana, Curran & Haman 703 (GH, NY, US). Plants resembling Eugenia granulata Berg (i.e. like many speci- mens from Cuba and Hispaniola) are not commonly found in con- tinental North America. A few collections, however, are note- worthy, for their resemblance to the many-flowered West Indian populations with oblong-obovate glandular-verrucose leaves: MEXICO: Quintana Roo: Cozumel Island, Steere 2976 (MICH).— Tres Marias: Maria Madre, Nelson 4306 (F, US), Maltbyl47 (F, US). In Costa Rica and Panama there has been some confusion be- tween what I take to be M. fragrans in the broad sense, and an endemic species, Myrcianthes (Eugenia) storkii. After the publica- tion in 1930 of Eugenia storkii, Standley routinely referred to that species all specimens of "cymosely flowered" eugenias from Costa Rica and Panama. Miss Amshoff, in her treatment of the Myrta- ceae of Panama (Ann. Missouri Bot. Gard. 45: 182. 1958), referred all the Panamanian material to E. fragrans, and relegated E. storkii to synonymy without comment. Some comment seems necessary. Most specimens from both Costa Rica and Panama, at least from elevations up to 1650-1700 meters, I should refer to Myrcianthes fragrans in the broad sense. Two principal races, however, are dis- tinguishable. One of these is like typical fragrans of Jamaica, and specimens of it are cited above under M. fragrans var. fragrans; the pubescence in this race is usually appressed, and the hypan- thium rather strongly and noticeably white-hairy. In the second race the pubescence on the branchlets and inflorescence is bristly rather than appressed, but the hypanthium is only sparingly stri- gose. This is var. hispidula, described below. At higher elevations (1900-2850 meters) in Costa Rica, Myrcianthes fragrans is appar- ently replaced by another species, with somewhat larger flowers, stouter glabrous inflorescence, and broader and more obtuse leaves; this is Eugenia storkii Standl., q.v. below. McVAUGH: TROPICAL AMERICAN MYRTACEAE 489 Myrcianthes fragrans, var. hispidula McVaugh, var. nov. Ramuli inflorescentiaeque hispiduli; hypanthium basi sparsim vel vix sericeo- strigosum; folia elliptica vel obovata, 2.5-6 cm. longa, (1.6-) 2-3-plo longiora quam latiora, apice obtusa vel rotundata, basi acuta vel cuneata; pedunculus 2-3 (-4) cm. longus, ad apicem 1.2-1.5 (-1.7) mm. crassus, flores 3-7 gerens; alabastra 3-4.5 mm. longa; discus diametro 2.5-3 mm., hirsutus; ovula ovarii quoque locula 16-20. — Univ. of Mich. Neg. 1351. Costa Rica and Panama, in woods, savannahs and pastures, at middle elevations (1000-1700 meters), flowering March-June, fruit- ing July-August. COSTA RICA: San Ramon, vicinity of La Palma, A. M. Brenes 5497 (F), 5503 (F), 5659 (F), 15142 (A; F, type), 16194 (F); Cerro Jucosal, H. E. Stork 1228 (F, paratype of Eugenia storkii); Las Con- cavas, C. H. Lankester K99 (F, paratype of E. storkii); Navarrito- Cartago, R. Torres R., 18 Oct. 1926 (US); Prov. Cartago, Dulce Nombre, Standley 35929 (US). PANAMA: Chiriqui: Between El Boquete and Caldera, H. Pittier 3326 (US) ; Boquete, M. E. Davidson 460 (A, F, US), 774 (A, F, US), 821 (F), 1062 (F); valley of Rio Chiriqui Viejo, G. White 108 (F, GH); valley of Rio Chiriqui Viejo near El Volcan, P. White 211 (F, GH), 217 (F, GH). Myrcianthes irregularis McVaugh, sp. nov. Arbor velutina vel sparse hirsuta, pilis fulvo-flavidis usque ad 0.8 mm. longis instructa; folia rigide coriacea, suborbicularia, ovalia vel obovata, 2-5.5 (-11) cm. longa, 1.5-5 (-9) cm. lata, apice rotundata retusave, basi rotundata truncata vel subcordata, petiolis hirsutis 4-8 mm. longis, venis lateralibus utroque latere 5-6, subtus prominiusculis, inter se arcuatis sed venam marginalem vix formantibus; inflorescentia irregulariter 2-3-dichotoma, pedunculo 2.5-3.5 cm. longo, 2-2.5 mm. lato; discus 4-5 mm. latus, annulo staminali molliter pubescente; stamina ca. 150; ovarium biloculare, loculis intus hirsutis, ovulis quoque loculo 10-11, e marginibus placentae erectae hippocrepiformis radiantibus. A tree, densely velutinous or sparsely hirsute with yellowish-brown bristly hairs up to 0.8 mm. long; leaves rigidly coriaceous, glabrate above (at first bristly along the broadly sulcate midvein), suborbicular to broadly oval or obovate, 1.5-5 (-9) cm. wide, 2-5.5 (-11) cm. long, rounded or retuse, rounded truncate or subcordate at base, the margins abruptly narrowed to stout short hirsute petioles 1-2.5 mm. thick, 4-8 mm. long; lateral veins on each side 5-6 in addition to some intermediate ones, inconspicuous above in mature leaves, prominent and somewhat elevated beneath, diminishing distally, not forming a strong marginal vein but each over- arching 2-6 mm. from the margin and joining the adjacent lateral; glandular dots inconspicuous; immature leaves conspicuously reticulate-veined in drying; inflores- cence irregularly forked 2 or 3 times; peduncle 2.5-3.5 cm. long, 2-2.5 mm. wide near summit, compressed, the flowers usually 3 in small dichasia at the tips of 490 FIELDIANA: BOTANY, VOLUME 29 the 2-3 branches; central flowers sessile; hypanthium subglobose, thinly hirsute, 2.5 mm. wide, somewhat contracted to a neck; calyx-lobes 4, nearly equal, fleshy, concave, rounded or obscurely triangular, 1.5-1.9 mm. long, 3-3.5 mm. wide, glabrous or essentially so within; disk quadrate, 4-5 mm. wide, the staminal ring soft-hairy; style soft-hairy at base (not seen complete); stamens about 150; petals broad-clawed, finely pubescent near base, about 4 mm. broad and long; ovary bilocular, the locules hirsute on the inner surfaces, the ovules 10-11 in each locule, radiating in one plane from an axile horseshoe-shaped placenta; fruit "round, yellow," according to Acosta-Solfs, not seen. "Mate-mate." ECUADOR: Loja: Bosque de "La Mira," Hda. La Hamaca, Cata- cocha, alt. 2400-2600 m., 17 Apr. 1944, Acosta-Solis 7923 (F, type). The generic position of this species may be questioned by some. The disposition of the ovules and the irregularly forked flowering branches are somewhat suggestive of Psidium, but against this are the bilocular ovary and the 4-parted calyx. If the fruit is actually yellow, as stated by the collector, this may indicate a relationship to Psidium. Fruiting specimens should be sought for final disposi- tion of this species. Myrcianthes karsteniana (Berg) McVaugh, comb. nov. Eu- genia karsteniana Berg, Linnaea 27: 277. 1856. Myrcia karsteniana (Berg) Steyermark, Fieldiana, Bot. 28: 1017. 1957. A recent collection, Garcia 94, is the first-known specimen in the fruiting condition. In the slightly immature fruit the beanlike em- bryo separates easily into 2 plano-convex plump cotyledons; the radicle is short and erect, the plumule immature but perceptible. On the basis of this embryo-structure, the species is to be referred to Myrcianthes as I understand it. In two flowering specimens, Fendler 2321 and Allart 413, the ovules are 15-20 in each of two locules, radiate in a subcapitate group affixed to the central dissepi- ment; this seems to rule out any possibility that the species belongs with Myrcia. I have not seen the syntypes (Karsten, Moritz in herb. Berlin), but duplicates photographed in Vienna (Field Mus. Negs. 31578, 31579) certainly are the same species as the Fendler, Allart and Garcia specimens. The inflorescence in this species is not exactly like that of other species of Myrcianthes; it is in fact a myrcioid panicle (see Fieldiana, Bot. 29: 158-160. 1956) 6-11 cm. long, each main branch with 5-6 decussate nodes from which arise shorter branches, the strongly compressed internodes 2.5-5 mm. wide or the distal ones narrower, the flowers numerous, mostly in 3- or 7-flowered dichasia at the tips of the branches. McVAUGH: TROPICAL AMERICAN MYRTACEAE 491 In Eugenia pycnantha Benth., a species of western Ecuador and southern Colombia (also erroneously referred to Myrcia by Steyer- mark, I.e. 1018), the inflorescences are corymbosely aggregated toward the ends of the branches, the individual dichasia are 3 or 4 times dichotomously forked with flat joint-like more or less appressed internodes 1 cm. long or less. This is probably also best referred to Myrcianthes, but certain disposition of it must await the discovery of fruiting material. The ovules are 10-15 in each of two locules, definitely excluding this species from the Myrciinae. Myrcianthes leucoxyla (Ort.) McVaugh, comb. nov. Myrtus leucoxyla Ort. Dec. 129. t. 17, fig. 2. 1800. Myrtus foliosa HBK. Nov. Gen. & Sp. 6: 134 (folio ed. p. 106). 1823. Myrcianthes foliosa (HBK.) McVaugh, Field Mus. Bot. 13, pt. 4: 758. 1958. The name Myrtus leucoxyla Ort. seems to have been overlooked both by DeCandolle and by Berg in their general revisions of the American Myrtaceae. The type-locality according to Ortega was "in novo Regno Granatensi cum Hopea terniflora" [i.e. "in Sanctae- fidensi Australis Americes Regno, ad urbem principem, aliisque passim locis," or in other words near Bogota, Columbia, where the local name was said to be palo bianco]. Ortega's plate is a very good representation indeed of the plant long known as Myrtus foli- osa or as Eugenia foliosa (HBK.) DC., a species common in the moun- tains about Bogota. The description calls for a glabrous plant with short-petiolate, ovate, subcordate leaves with short-acuminate tip; 3-flowered axillary peduncles and short-pedicellate flowers; four- parted calyx with subrotund concave ciliate lobes reflexed in flower, and subglobose fruit. There can be little doubt that Myrtus foliosa HBK. and M. leucoxyla Ort. are conspecific. Eugenia mariquitensis Berg, Linnaea 27: 267. 1856. This was based on a specimen from Mariquita, Colombia, Lin- den 1102. I have seen isotypes in the herbaria of Harvard University and the Naturhistorisches Museum in Vienna, and I should refer this species to the synonymy of Myrcianthes rhopaloides (HBK.) McVaugh. Myrcianthes mato (Griseb.) McVaugh, comb. nov. Eugenia mato Griseb. Abh. Kgl. Ges. Wiss. Goett. 24: 125. 1879. I have not seen authentic material of this species, but Legrand studied a "cotype" (Lorentz & Hieronymus 880), and on the basis 492 FIELDIANA: BOTANY, VOLUME 29 of this and other material published a detailed description and an illustration contrasting this species with another previously unde- scribed, E. pseudo-mato (Legrand, D., Contribution al conocimiento de tres arboles argentinos de la familia de las mirtaceas. Lilloa 10: 471-482, with plate. 1944). The embryo of E. mato is described by Legrand as follows (I.e., p. 476) : ". . . cotiledones grandes, algo con- vexos, separados, con radicula diminutas exserta." This plant is very like Eugenia [Myrcianthes] pseudo-mato and other undoubted species of Myrcianthes, except that the peduncles are one-flowered. This is apparently not a particularly fundamental character, as one-flowered peduncles may be found occasionally in various species of Myrcianthes that are normally 3-flowered (see the key in Flora of Peru, 13, pt. 4: 747-749). In addition to the specimens cited by Legrand, the following may be mentioned: ARGENTINA: Tucuman: Venturi 1991 (A), 10035 (A, GH). Myrcianthes minimifolia (McVaugh) McVaugh, comb. nov. Eugenia minimifolia McVaugh, Fieldiana, Bot. 29: 213. 1956. This was treated in the Flora of Peru, 13, pt. 4: 764-765, as a species of Myrcianthes, but the formal nomenclatural transfer was not made because this species has one-flowered peduncles and at that time I had seen none except flowering specimens. In 1959 Dr. Ra- mon Ferreyra sent me fruiting specimens; in the somewhat imma- ture embryo there are two large fleshy separate cotyledons and a short radicle. This would appear to justify the transfer made above. The following additional specimens have been seen: PERU: Arequipa: Prov. Caraveli, Lomas de Pongo cerca de Acari, 700-750 m., Ferreyra 13423 (MICH), 13*81 (MICH). Eugenia orthostemon Berg, Linnaea 27: 179. 1856. This was based on Linden 807, from near Fusagasuga, Colombia. Because of the one-flowered peduncles it was placed by Berg in his large group Biflorae, not with the Dichotomae. On the basis of most of its morphological characteristics, I believe it belongs with the other Andean species of Myrcianthes, but fruit is needed to confirm this. It may be related to such species as Eugenia punicifolia (HBK.) DC. Myrcianthes prodigiosa McVaugh, sp. nov. Arbor glabra 5-15-metralis; folia rigide coriacea, elliptica vel elliptico-ovata vel -obovata, 10-23 cm. longa, 5-10 cm. lata, brevi-acuminata, basi rotundata McVAUGH: TROPICAL AMERICAN MYRTACEAE 493 usque ad acuta, petiolis 1.5-2.5 cm. longis, nervo medio supra impresso, venis lateralibus numerosis inconspicuis; inflorescentia crassa, plerumque triflora, pe- dunculo compresso 3-8 cm. longo, versus apicem 4-5 mm. lato; alabastra turbinata 1.7-2 cm. longa; calycis lobi 4 (-5), suborbiculares, 5-10 mm. longi latique; discus 12-15 mm. latus; stylus 2.5 cm. longus; stamina ca. 250; ovarium triloculare, loculis duobus minoribus, ovulis quoque loculo 10-20; fructus globosus oblatusve, dia- metro 4-5 cm.; semina 2 (-1), cotyledonibus 2, plano-convexis inaequalibus. A tree 5-15 meters high, 20-30 cm. in diameter, glabrous except for stiff sordid hairs on the vegetative buds, and for the bristly staminal ring; leaves rigidly cori- aceous, elliptic or elliptic-ovate or -obovate, 5-10 cm. wide, 10-23 cm. long, 2-2.5 times as long as wide, abruptly short-acuminate at tip, the base unequal-sided and rounded to acute, the revolute margins decurrent on the narrow inner angles of the stout petioles 2-3 mm. thick, 1.5-2.5 cm. long; midvein narrowly impressed above, elevated most of its diameter beneath; lateral veins numerous, 12-20 pairs including some intermediate ones, inconspicuous on both sides or somewhat ele- vated above when dry; marginal vein like the lateral ones and somewhat arched between them, 2-5 mm. from the margin; blades lustrous above and finely resinous- dotted (sometimes impressed-punctate), dull below, opaque and nearly featureless; inflorescences in leaf-axils, or a pair opposite at the lowermost (bracteate) nodes of a short terminal branch; peduncles stout, 3-flowered or by reduction 2-flowered, 3-8 cm. long, compressed toward the summit and there 4-5 mm. wide; terminal flower sessile or its pedicel 3-5 mm. long, the lateral pedicels 5-15 mm. long; bracteoles falling before the buds mature; buds turbinate, 1.7-2 cm. long; calyx 4- to 5-merous, the lobes 4 or 5 and subequal, or 4 with one additional smaller lobe; lobes suborbicular, fleshy with thin hyaline short-ciliate margins, 5-10 mm. long and wide, soon reflexed at anthesis; hypanthium 1 cm. long, attenuate to the base; disk 12-15 mm. wide (10-12 mm. wide in fruit), the glabrous center 6-8 mm. wide; style 2.5 cm. long; stamens about 250, about as long as the style; petals white; ovary trilocular (one locule sometimes about as large as the other two combined), the ovules 10-20 in each locule, in compact somewhat elongate groups radiating from the central axis; fruit green, globose or oblate, 4-5 cm. in diameter; seeds 1 (-2), up to 4 cm. long, with leathery testa wrinkled in drying, and 2 unequal plano-convex cotyledons. BRITISH GUIANA: Kaieteur Plateau, savanna and savanna forest, Kaieteur Falls to Mure-Mure Savanna, elev. 1400 ft., 12 Mar. 1962 (fr.), R. S. Cowan & T. R. Soderstrom 2147 (MICH). SURINAM: Tafelberg, near summit, margin of North Ridge Creek, mixed high forest, 8 Sep. 1944 (bud & fl.), B. Maguire 24657 (MICH, type); Tafelberg, 5 km. northeast of Savanna No. 2, mixed high forest, 1850 ft., 13 Sep. 1944 (old fl.), Maguire 24711 (MICH). This extraordinary species can hardly be referred to any genus except Myrcianthes, but it is unique in that genus by virtue of its very large flowers and fruits and its trilocular ovary. Myrcianthes pseudo-mato (Legrand) McVaugh, comb. nov. Eugenia pseudo-mato Legrand, Lilloa 10: 477. 1944. Pseudomyrci- anthes pseudo-mato (Legrand) Kausel, Ark. Bot. II. 3: 505. 1956. 494 FIELDIANA: BOTANY, VOLUME 29 I have not seen the type, but from the photograph of it published by Legrand, and from his excellent description and illustration, there can be no doubt of its identity. It is clearly a species of Myrcianthes; actually some Argentinian and Bolivian material can scarcely be dis- tinguished from M. discolor of the Peruvian Andes. In West's no. 8312, cited below, the cotyledons are distinct, plano-convex, the radicle half as long, the plumule evident, silky, 1.5 mm. long. The following specimens have been seen: BOLIVIA: Tarija: Rincon de la Victoria, 15 km. southwest of Tarija, J. West 8312 (GH). ARGENTINA: Tucuman: Dept. Chicligasta, Estancia Las Pavas, P. Jorgensen 13 (GH; this number also cited by Legrand). Dept. Monteros, Quebrada de Caspiuchango, Schreiter 5765 (Herb. Lillo 6672) (F, GH), Venturi 9595 (A, GH, US). Dept. Tafi, Cerro de Taficillo, Venturi 9723 (A, GH), 9998 (US). The genus Pseudomyrcianthes was distinguished by Kausel be- cause it was said to have the placenta arising near the apex of the locule, the embryo sub-homogeneous ("Embryo homogeneus later- aliter paullo retusus, fissuram interembryonariam longitudinalem praebens") and splitting only partially. My observations on speci- mens of what I take to be Myrcianthes pseudo-mato do not confirm those of Kausel on the structure of the embryo as described for Pseudomyrcianthes. I cannot attest to the importance of apical versus central placentation, but the distinction between these would seem to be somewhat subjective. Eugenia pycnantha Benth. PI. Hartw. 174. 1845. Colombia (the type from the province of Popayan, Hartweg 976) and Ecuador. This is probably a species of Myrcianthes, in general appearance not unlike M . bifurcata, and M. osteomeloides (see Flora of Peru, 13, pt. 4: 750, 752, 768, 773), and somewhat suggesting M. karsteniana (q.v. above). Fruiting material is needed for proper generic dis- position. Eugenia rigidissima Cufodontis, Archivio Bot. 9: 198. 1933. Costa Rica, the type from Volcan de Turrialba at 2500 m., Porsch 758. Standley (Field Mus. Bot. 18: 774. 1937) referred this species to the synonymy of E. storkii Standl. I have not seen the type of E. rigidissima, but it may be a distinct species. Amshoff (Ann. McVAUGH: TROPICAL AMERICAN MYRTACEAE 495 Missouri Bot. Card. 45: 184. 1958), in writing of P. H. Allen 1563, from Cerro Punta at 2000 m., Chiriqui, Panama, refers this speci- men to E. fragrans but says: "... a rather distinct looking specimen with suborbiculate, very rigid leaves with impressed nerves agreeing with Eugenia rigidissima Cuf . ..." Allen's specimen (F, US) may be described as follows: Tree 30 m.; young growth with silky-matted hairs, the year-old branchlets strigose; leaves very broadly rounded, 1.5-2.8 cm. wide, 1.8-3 cm. long, rounded or obtuse at both ends, very coriaceous, 3-5 pairs of the lateral veins impressed above, stronger than any intermediate veins and prominent beneath; buds gla- brous, 5-6 mm. long; inflorescence glabrous, the peduncles 3-flowered, 1-2 cm. long, 2-2.5 mm. wide at apex; disk glabrous, 3.5-4.5 mm. wide; stamens 175-200; ovules about 30 in each locule. On the basis of characteristics used elsewhere in separating spe- cies of Myrcianthes, this would deserve recognition as a distinct species. Eugenia rondonensis Steyerm. Fieldiana, Bot. 28: 1013. 1957. Venezuela, the type from Mt. Roraima, Steyermark 58956 (F). This is a nearly glabrous ample-leaved plant looking superficially like some form of M. fragrans or M. dugandii. The type is a speci- men in very immature bud (Univ. of Mich. Neg. 1337). The pedun- cles are relatively long and very slender (5-6 cm. long), all 3-flowered, the hypanthium glabrous like the rest of the inflorescence, the ovules about 12 in a subcapitate group in each of two locules. In the key above, this specimen would be identified as M. dugandii, except that the peduncles are 3-flowered and unusually long and slender. It does not appear to be a very distinctive species, but the available material is too immature to permit a judgment. Myrcianthes sessilis McVaugh, sp. nov. Arbor 12-metralis, innovationibus gemmisque, et petiolis, pilis appressis seri- ceis, pallidis, usque ad 0.5 mm. longis, rare praeditis; hypanthio et loborum calycis superficie exteriori similiter sed densiuscule sericeis; foliis 4-8 cm. longis ellipticis obtusis vel obtuse acuminatis, venis inconspicuis; dichasio 3-floro, pedunculo 2- 3 cm. longo, apicem versus 2-2.5 mm. lato, floribus sessilibus; stylo 7-8 mm. longo. A tree to 40 feet high (Smith), the youngest branchlets and petioles, and the leaf-buds thinly appressed-silky with white hairs up to 0.5 mm. long; hypanthium and the outer surface of the calyx-lobes similarly but conspicuously and densely pubescent; leaves coriaceous, lustrous above, elliptic, 2-3.5 cm. wide, 4-8 cm. long, 1.8-2.2 times as long as wide, the tip obtuse to acute or bluntly acuminate, the base acute, the margins decurrent on the sulcate petiole 1-1.5 mm. thick, 8-10 mm. long; midvein impressed above, prominent beneath; lateral veins 10-12 pairs, in- 496 FIELDIANA: BOTANY, VOLUME 29 definite in number with some intermediate veins, in drying raised a little on both surfaces; marginal vein about equaling the laterals and arched between them, 1-2 mm. from the margin; inflorescence an axillary dichasium, 3-flowered, the flowers all sessile, the peduncle 2-3 cm. long, compressed and enlarged distally, where 2-2.5 mm. wide; bracts and bracteoles deciduous before anthesis; buds 5-6 mm. long, obovoid; hypanthium 3 mm. long, funnelform or somewhat rounded at the base; calyx-lobes in 2 nearly equal pairs, rounded, glabrous or silky on the inner surface, 3-5 mm. wide, 2.5-3.5 mm. long; disk 4-5 mm. wide, the broad staminal ring hairy, the style-base conic, 1 mm. wide; style 7-8 mm. long; stamens 150-175, to 8-9 mm. long, the anthers 0.7-0.8 mm. long; petals oblong-obovate, ciliate-f ringed and silky on outer surface near base, 6 mm. wide and 10 mm. long; ovary bilocular, the ovules 8 in each locule, somewhat biseriate on a centrally affixed placenta; fruit not seen. COLOMBIA: Masinga, [Sierra de] Santa Marta, rare below 1500 feet, generally in wooded valleys, Dec. 22 [1898-1901], Herbert H. Smith 776 (A; F; GH, type; MICH; PH; US); Santa Marta, Smith 1735 (F; PH; US). Apparently a most distinctive species, easily recognized from the limited material now available by the abundant silky pubescence, the sessile flowers and the long-petiolate, lustrous, and coriaceous leaves. Myrcianthes storkii (Standl.) McVaugh, comb. nov. Eugenia storkii Standl. Field Mus. Bot. 8: 143. 1930. Young growth thinly silky-strigose, the 1-year-old branchlets thinly strigose or glabrate; leaves obovate or rounded to broadly elliptic, 1.5-3.5 cm. wide, 2-5 cm. long, 1-1.7 (-2.5) times as long as wide [on shoots up to 2.5-4.5 cm. wide, 4.5-7 cm. long, 1.5-1.75 times as long as wide]; blades coriaceous, rounded or bluntly pointed at apex, cuneate to obtuse (or rounded) at base, the veins not impressed above but 3-7 pairs of the laterals much stronger than the intermediate ones and prominent beneath; buds 4-6 mm. long, glabrous; inflorescence glabrous, the peduncle 3-flowered, 2-2.5 cm. long, 1.5-1.7 mm. wide at the apex; disk hir- sute, 3.5-4 mm. wide; stamens ca. 150; ovules ca. 20 in each of 2 locules. Costa Rica, in potreros and about edges of woodlands, 1900- 2850 m., flowering in April and May. COSTA RICA: Potrero below Poas, H. E. Stork 2501 (F, type); east of Irazu, Stork 2074 (F, paratype; MICH); Alajuela, south slope of Volcan Poas, Austin Smith P2401 (A, UC) ; Alajuela, Pal- mira, Smith NY 676 (A), H676 (F), P2627 (F); El Tablazo, Echever- ria 377 (F, UC); Prov. Heredia, Cerro de las Caricias, Standley & Valeria 52193 (US). Of the four paratype collections of E. storkii, only one (Stork 2074) can be confidently referred to this species. Two (Stork 1228 McVAUGH: TROPICAL AMERICAN MYRTACEAE 497 and Lankester K99) are cited above under M. fragrans var. hispidula. The fourth paratype (Standley & Torres 47426, from Fraijanes, Ala- juela, at 1500-1700 m.) is sterile (F, US) and indeterminable. Two other collections from the same general locality (Standley and Torres 47619, 47725, both US) are similarly sterile and are likewise inde- terminable. For discussion of the taxonomic position of this species, see above under M. fragrans var. hispidula. Myrcianthes ternifolia (Berg) McVaugh, comb. nov. Eugenia ternifolia Berg, Linnaea 27: 246. 1856. Venezuela, the type "in montibus Sierra Nevada," Moritz 1183. A little-known species of the mountains of northeastern Colom- bia and adjacent Venezuela. The leaves are sometimes partly oppo- site, usually ternate; the flowers in all the specimens I have seen are aggregated near the tips of the branches in abundantly produced but 3-flowered stout axillary dichasia; the total effect is that of a series of small corymbose panicles suggesting the inflorescences of Eugenia pycnantha, q.v. Berg distinguished this species from Eugenia triquetra by the spatulate leaves and the "pedunculis saepissime 3-floris." A photo- graph of a presumed isotype, however, Moritz 1183 (G-DEL), shows the peduncles chiefly 1-flowered (Field Mus. Neg. 23594). Possibly Moritz 1183 was a mixture. Berg described the ovary as bilocular and many-ovulate. Eugenia triquetra Berg, Linnaea 27: 141. 1856. Colombia and Venezuela, the syntypes from "inter Pamplona et Tapagua," Linden 726 and "in montibus Sierra Nevada, in Venezuela," Moritz 1184- A poorly understood species, treated in the Flora of Peru, 13, pt. 4: 774 with Myrcianthes. It seems to be still unknown in the fruiting condition. Eugenia turumiquirensis Steyerm. Fieldiana, Bot. 28: 1015. 1957. Venezuela, the type from Sucre, Cerro Turumiquire, Steyer- mark 62548. The type is in young bud. It suggests a moderately or perhaps more than ordinarily pubescent form of Myrcianthes fragrans. If it had come from Jamaica or Central America I should have referred it without question to M. fragrans. — Univ. of Mich. Neg. 1338. 498 FIELDIANA: BOTANY, VOLUME 29 6. CERTAIN EUGENIOID GENERA There is need for some clarification of generic lines among cer- tain taxa that seem to represent specialized lines of evolution among the Eugeniinae. In these the flowers are morphologically racemose but the axis of the raceme is usually so much reduced in length that the flowers appear to be in small axillary fascicles or glomerules; the number of ovules in each locule of the ovary is reduced at least to 4 and usually to 2; the hypanthium is prolonged into a tube or cylin- der extending well above the level of the summit of the ovary; and the combined structures of the calyx and hypanthium are otherwise specialized: The calyx is closed or nearly closed in the bud, and usually splits irregularly down toward the summit of the ovary at flowering time, and in many species the tubular part of the hypan- thium is circumscissile and cleanly dehiscent from the summit of the ovary. In recent years explorations in northern South America have brought to light a number of species, mostly as yet unstudied and undescribed, that according to current interpretations might be as- signed (and sometimes have been assigned) to Marlierea, Myrciaria, Plinia or Siphoneugena. A considerable number of flowering speci- mens is available, but fruiting specimens are poorly represented in herbaria, and it has not often been possible to associate flowering and fruiting material of the same species. Characters of the embryo are especially important in the delimitation of the genus Plinia as understood by Amshoff (Fl. Suriname 3: 97-99. 1951) and McVaugh (Flora of Peru, 13, pt. 4: 775-780. 1958). As a result of the incom- pleteness of material now available, it is sometimes impossible to assign flowering specimens to genus. As I understand the genera listed above, they may be distinguished as follows (Marlierea not included, as its relationships are with the Myrciinae; it resembles the other genera only in having 2 ovules. The key that follows is particularly applicable to plants of northern South America; the numerous West Indian species of Plinia described by Urban have not been considered, as they are mostly vegetatively very different and few of them are known except from sterile specimens) : Hypanthium circumscissile at base, cleanly deciduous (with the perianth and androecium) at about the time of anthesis, leaving a circular scar on the ovary and fruit; calyx shortly four-lobed, the lobes ciliate, probably always imbricate but often obscurely so. Ovary below the deciduous part of the hypanthium prolonged, fusiform or cam- panulate, usually contracted into a neck in bud and in flower; bracteoles scarcely connate, much shorter than the ovary; embryo as far as known McVAUGH: TROPICAL AMERICAN MYRTACEAE 499 with 2 distinct plano-convex cotyledons; ovules usually 4 in each locule; flowers mostly pedicellate; wet mountain forests, known chiefly from the Guayana region of Venezuela; Guadeloupe Siphoneugena. Ovary broadened from the base upward, not contracted below the line of de- hiscence of the hypanthium; after dehiscence broad and flat, scarcely longer than the involucre formed by the broad connate bracteoles; pedicels almost none, or very stout and scarcely longer than the bracteoles; embryo as far as known undivided ("eugenioid"); ovules usually 2 in each locule; wide- spread Myrciaria. Hypanthium not circumscissile; calyx closed in bud or sometimes lobed, in an- thesis splitting by irregular longitudinal fractures down to the summit of the ovary, some remnants usually persisting in fruit; embryo, as far as known, with 2 free plano-convex cotyledons, one above and one below; ovules 2 in each locule.. . .Plinia. REVISION OF MYRCIARIA BERG, IN NORTH AMERICA AND NORTHERN SOUTH AMERICA The genus typified by Myrciaria tenella (DC.) Berg (cf. Taxon 5: 143. 1956) is a very natural and closely related group, more easily recognized as a distinct genus than any other American group of the Eugeniinae. Berg, unfortunately, included in the genus as he first characterized it a number of species that obviously were unrelated to the rest — in fact, mostly species of Myrcia — and this error on the part of Berg may have led Bentham and Hooker to disregard the very distinct and homogeneous generic group that remains after these anomalous species have been removed. At any rate these authors, in the Genera Plantarum, relegated Myrciaria to the synon- ymy of Eugenia. Niedenzu, in the Naturlichen Pflanzenfamilien, accepted Myrciaria as a distinct genus but apparently did not un- derstand it well, delimiting it more or less in the original sense of Berg to include several species of Myrcia. Beginning with the work of Chodat and Hassler on the Myrtaceae of Paraguay (Bull. Herb. Boiss. II. 7: 807-808. 1907), in which the authors recognized 11 spe- cies of Myrciaria, the genus has been rather generally accepted by students of the American flora. The following key is intended to be applicable to all known spe- cies of Myrciaria in South America north of the Amazon River, in Central America and Mexico, and in the West Indies. Plants glabrous except the ciliate-margined bracts and bracteoles, and the inner faces of the perianth-parts; bracts membranous, ovate, sometimes nearly concealing the buds; bracteoles membranous, distinct, orbicular or sometimes pointed, 1.5-2.5 mm. wide and about as long; leaves elliptic-oblong or -ovate, 500 FIELDIANA: BOTANY, VOLUME 29 2-4.5 cm. wide, 5.5-13.5 cm. long, acuminate and acutely attenuately pointed, broadly rounded at base, on stout terete petioles 4.5-8 mm. long; glands in- conspicuous or not seen in mature leaves; veins apparent but inconspicuous above, the lower surface opaque, the veins obscure; northern Venezuela; Panama; Costa Rica M. vexator. Plants sometimes glabrous, usually at least the petioles (ventrally) and the young- est branchlets pubescent; bracts inconspicuous, much shorter than the brac- teoles; bracteoles somewhat fleshy or coriaceous, usually wrinkled in drying like the hypanthium, connate by the basal margins and forming an involucre- like cupule beneath the flower; leaves various. Leaves cordate or subcordate at base, the sinus between the basal auricles defi- nite but sometimes shallow; blades ovate to oblong, 4-8 cm. long. Leaves rounded or obtusely pointed at tips, prominently cordate, subsessile, the very stout petioles 2-2.5 mm. thick, 2-4 mm. long from the basal sinus of the blade to the base of the petiole; midvein about equally con- vex on both surfaces; plants as far as known quite glabrous; style more than 7 mm. long; Mt. Roraima M. cordata. Leaves gradually acuminate, the tip attenuate and acutely pointed; blades broadly subcordate; petioles pubescent, 1 mm. thick, 2-3 mm. long; midvein flat or somewhat convex above, convex beneath; styles probably 6-7 mm. long (ex Berg); British Guiana; Suriname M. vismeifolia. Leaves acute or cuneate or sometimes slightly rounded to base, sometimes with minute auricle-like folds at the summit of the petiole, never cordate; blades various, often elliptic to lanceolate. Leaves at apex acute (usually not prominently acuminate), mostly 6-10 cm. long; blades when dry often discolorous, the lower surface smooth and opaque, the upper dotted with large convex dark or amber-colored resi- nous glands; lateral veins near the base of the blade leaving the midvein at an angle less than 45°; small anastomosing veinlets almost none, the small veins all nearly parallel and ascending; margins at the base of the blade abruptly incurved into the petiole, often forming a pair of slightly raised auricle-like folds; style (8-) 10-11 mm. long; plants glabrous or the petioles (ventrally) and the youngest branchlets pubescent; widely distributed in the Amazon basin and the upper Orinoco drainage, from eastern Peru to Para (Santarem), north to southern Venezuela . .M. dubia. Leaves at apex mostly prominently and often caudately acuminate, incon- spicuously if at all glandular above, concolorous or essentially so; mar- gins at the base of the blade cuneately decurrent or at least narrowly acute (except in M. ibarrae with style 5-6 mm. long); style in all species (as far as known) 7 mm. long or less. Leaves (1.5-) 2.5-5 cm. wide, (4.5-) 7-14 cm. long, gradually long-acumi- nate; plants glabrous or the petioles ventrally pubescent; glands in- conspicuous or scarcely apparent on both surfaces of the leaves; lateral veins near the base of the blade leaving the midvein at an angle of 45° or more; small irregular anastomosing veinlets frequent; open flowers unknown; Para, Brazil, near the mouth of the Amazon. .M. amazonica. Leaves 0.7-2 (-3) cm. wide, 2-6 (-8) cm. long; branchlets and petioles usually abundantly and finely pubescent; glands inconspicuous, or small convex glands rather prominent beneath; lateral veins near the base of the blade varying in direction, the small veinlets few or very inconspicuous; style 4.5-6 mm. long. McVAUGH: TROPICAL AMERICAN MYRTACEAE 501 Leaves of an ovate type, mostly 2-2.5 times as long as wide, sharply acuminate and often cuspidate, broadly rounded at base, the mar- gins little if at all decurrent on the petiole; branchlets and petioles coarsely and rather sparsely hispidulous; flowers sessile; bracteoles very broadly rounded and imbricate, connate at very base only, 1.5-2 mm. long and wide, much exceeding the ovary; Guatemala (Pete"n) M. ibarrae. Leaves of a lanceolate type, mostly 2.5-3 times as long as wide, variously acuminate but the very tip usually blunt or merely acute, the base of the blade somewhat rounded below the middle, finally usually acute or cuneate with the margins cuneately decurrent on the splayed summit of the petiole; branchlets and petioles finely pubescent or rarely glabrous; pedicels usually 0.5-1 mm. long; bracteoles not im- bricate (or very slightly so in the bud stage), united by their proxi- mal margins into an oval bilobed or finally explanate involucre 2 mm. long; West Indies; eastern Mexico and Central America; Colombia; lowlands of the Guianas; Orinoco and Amazon lowlands from south- ern Venezuela to Peru and eastern Brazil (Maranhao). M . floribunda. Myrciaria amazonica Berg, in Mart. Fl. Bras. 14, pt. 1: 374. 1857. An imperfectly known species, known to me from two nearly sterile specimens only, both of these from Brazilian Amazonia: Pard: Belem, B. E. Dahlgren & E. Setta 398 (F); Insula Colares, Poeppig s.n. (W, type). The leaves seem to be distinctive as indi- cated in the key. Myrciaria cordata Berg, Linnaea 27: 337. 1856. Apparently known only from the original collections by Schom- burgk. The type was seen by Berg at Berlin, and was cited by him as from "Guiana Anglica (Rich. Schomburgk, coll. no. 957)"; another collection, not cited by Berg, is from Mt. Roraima, Schomburgk 608 (F, W). Myrciaria dubia (HBK.) McVaugh, comb. nov. Psidium du- bium HBK. Nov. Gen. & Sp. 6: 152 [folio ed. p. 121]. 1823. M. para- ensis Berg, in Mart. Fl. Bras. 14, pt. 1: 364. 1857. M. caurensis Steyerm. Fieldiana, Bot. 28: 1020. 1957. The type of Psidium dubium HBK., from near Atures on the Orinoco, is surely a Myrciaria. This is indicated by the description ("bracteis . . . ovarium subaequantibus eique adpressis. . . . Calyx superus, hemisphaericus, glaber, glanduloso-punctatus, margine quad- rilobus. . . . Stamina . . . calyci medium versus inserta . . . ovarium 502 FIELDIANA: BOTANY, VOLUME 29 . . . biloculare; . . . ovula duo in quolibet loculo, apposita . . . fructus . . . edulis"), and is apparent from a photograph (Field Mus. Neg. 36872). The species is not identifiable from the plate (t. 547bis) accompanying the original description of P. dubium. The type of P. dubium appears to represent a relatively narrow- leaved race, frequent in the Orinoco drainage, of a species found also in the Amazon drainage. The type of M. paraensis Berg has the somewhat broader leaves characteristic of most specimens from the Amazon basin, but I believe M. dubia and M. paraensis are conspecific. The following may be cited as representing the species in the Orinoco drainage or at its upper limits. VENEZUELA [possibly COLOMBIA]: Prope Atures (Missiones del Orinoco), Bonpland (P, not seen). — Bolivar: La Prision, Medio Caura, L. Williams 11691 (F, type of M. caurensis); Cerro Guaiqui- nima, Rio Paragua, Maguire 33135, 33139 (both MICH).— Ama- zonas: Rio Sanariapo, H. M. Curran 1833 (NY); Capibara, Canal del Casiquiare, Holt & Gehriger 289 (NY). Myrciaria floribunda (Willd.) Berg, Linnaea 27: 330. 1856. This, the most widely distributed species of the genus, has a considerable synonymy. The principal synonyms are listed by Ur- ban (Bot. Jahrb. 19: 657-658. 1895) and by Amshoff (Fl. Suriname 3: 108-109. 1951, and Ann. Missouri Bot. Gard. 45: 177. 1958). An additional synonym is apparently M. maragnanensis Berg, in Mart. Fl. Bras. 14, pt. 1 : 372. 1857, of which I have seen an isotype (Gardner 6023, W). Two species described by Berg, Myrciaria verticillata Berg and M. divaricata (Benth.) Berg, were based entirely or in part on Schomburgk 958. This collection seems actually to have been a mix- ture. A sheet of no. 958 (MICH) has the leaves apparently of M. floribunda and is presumably to be referred to that species, al- though the styles are about 7 mm. long. This is presumably the plant described by Berg as M . verticillata ("foliis . . . utrinque angus- tatis, apice longissime lineari-acuminatis, basi elongato-acutis. . . .")• Another sheet of no. 958 (F) has the leaves of M. dubia (HBK.) McVaugh, but the flowers are smaller than usual in that species; I have not seen the type of Eugenia divaricata Benth., but the leaves are described as being rounded at base, i.e., therefore more like dubia than floribunda. Apparently Berg had the same plant, for he de- McVAUGH: TROPICAL AMERICAN MYRTACEAE 503 scribed M. divaricata as with "foliis . . . ovatis v. ovato-oblongis, . . . acute acuminatis, basi rotundatis v. breviter acutis. ..." The plant from Jamaica described as Eugenia polyneura Urb. Symb. Antill. 5: 446. 1908 (E. pycnoneura Urb. Symb. Antill. 6: 25. 1909, not E. polyneura Koord. & Val. 1855) is apparently not spe- cifically different from M. floribunda. I have seen an isotype (F) or perhaps the actual type (Jamaica below Vinegar Hill, W. Harris 7448, ex herb. Krug et Urb.). The specimen is nearly glabrous, with but traces of pubescence on the petioles; the petioles them- selves are 7-8 mm. long on the mature leaves; the blades are 1.5- 2.5 cm. wide, elliptic-ovate and mostly acutely long-acuminate. The combination of nearly complete lack of pubescence, broad sharp- tipped leaves and long petioles is most unusual in M. floribunda, but the individual features are matched in many other South Amer- ican and West Indian specimens. I have not seen specimens of Eugenia asa-grayi Krug & Urb. [Plinia asa-grayi (Krug & Urb.) Urb.], nor of E. I'heritieriana DC., both of which species were associated by Urban (1895) with his Eugenia floribunda. The former was from eastern Cuba (Wright 1610) and the latter supposedly from the island of Tobago. Myrciaria ibarrae Lundell, Wrightia 2: 213. 1961. This recently described species is known to me only through the original specimens cited by Lundell. It seems to be abundantly distinct from M. floribunda, which is known from the same gen- eral area. Myrciaria phillyraeoides Berg, Linnaea 27: 326. 1856. This plant, based on a collection by Humboldt in "America me- ridionali," and seen by Berg in the Willdenow herbarium (No. 9574), I do not recognize from the description. It may not be a Myrciaria. If it belongs to that genus, I suspect it is M. dubia. Myrciaria vexator McVaugh, sp. nov. M. piltieri Burret ex Badillo in Pittier et al. Cat. Fl. Venez. 2: 198, nomen. 1947. Eugenia palmarum Standl. & L. O. Williams ex Allen, Rain For. Golfo Dulce 201, nomen. 1956. Arbor 10 m. alta, bracteis bracteolisque, et perianthii segmentis, ciliatis, peri- anthii segmentis intus tomentulosis; folia elliptico-oblonga vel ovata, 2-4.5 cm. lata, 5.5-13.5 cm. longa, acuminata, apicibus attenuatis acutis; lamina basi ro- tundata, marginibus plusminusve decurrentibus, petiolis teretibus crassis anguste 504 FIELDIANA: BOTANY, VOLUME 29 profundeque sulcatis 4.5-8 mm. longis; glandulae vix prominulae; bracteae ovatae, membranaceae, usque ad 3 mm. latae, usque ad 4-4.5 mm. longae, saepe alabastra fere occultantes; bracteolae rotundae, distinctae, membranaceae, 1.5-2.5 mm. longae lataeque; hypanthium supra germen 2 mm. productum; discus 1.3-1.5 mm. latus, glaber; stylus 8 mm. longus; calycis lobi 2 mm. lati, 1.5 mm. longi; petala ovalia 2 mm. longa lataque; stamina ca. 50, antheris 0.6-0.8 mm. longis; fructus globosus, atropurpureus, carnosus, diametro 2.5 cm. Attributed to Burret by Badillo, as a new species, but quite without description; described in English by Allen, under the name of Eugenia palmarum. This species is known mostly from cultivation, but has appar- ently been collected in the wild both in Costa Rica and in northern Venezuela. Specimens examined are: COSTA RICA: Puntarenas, forested hills above Palmar Norte, alt. 450 m., P. H. Allen 6331 (F). PANAMA: Summit Gardens, C. Z., J. E. Higgins 287 (type, US); Summit, J. Zetek 3520 (F; MO, sterile, probably this species; see below) . VENEZUELA: Lara: Bosques de Chirgua, entre Nonavana y Bar- quisimeto, Jose Saer 515 (F). Carabobo: Valencia, in gardens, Pittier 8154 (US); Valencia, often cultivated, Pittier 8899 (US). Myrciaria vismeifolia (Benth.) Berg, Linnaea 27: 336. 1856. The type was from British Guiana; Amshoff cites a number of collections from Suriname. In the Flora of Panama (Ann. Missouri Bot. Gard. 45: 179. 1958) Miss Amshoff refers to this species a single specimen from the Canal Zone, Zetek 3520, remarking at the same time that this differs from the typical plant of the Guianas by its "firmer leaves which are never (?) cordate at the base, its longer petioles and its slightly larger flowers." The Zetek specimen, which I have seen (MO), is sterile, but can hardly be referred to M. vismei- folia. In that species the petioles are 3-4 mm. long, ventrally some- what pubescent and flat or shallowly sulcate; the blades are defi- nitely cordate-auriculate, the auricles extending mostly 1-1.5 mm. beyond the sinus at the insertion of the petiole. In the Zetek speci- men the petioles are 5-8 mm. long, quite glabrous (like the rest of the specimen), nearly terete but with a narrow sharply indented ventral furrow; the blades are rounded at base, but not auriculate. The leaves are in fact strikingly similar to those of Myrciaria vexator, and the specimen probably belongs to that species. In any event M. vismeifolia should be excluded from the known flora of Panama. McVAUGH: TROPICAL AMERICAN MYRTACEAE A NEW SPECIES OF PLINIA L., FROM COSTA RICA 505 There seem to be no reports of any species of this genus from continental North America. A few species are known in South America, and a number have been reported from the West Indies. FIG. 13. Plinia salticola, from the type. A, Open flower; B, Bud (X 2.5); C, Enlarged tip of bud, seen from above ( X 5). The open flower is seen from the adaxial side; the observer is looking down at the tip of the raceme-axis. Opposite the open flower may be seen the bract and bracteoles of the other flower of the proximal pair; at right angles to this may be seen the bracts and bracteoles of the second (distal) pair of flowers. The small bud is seen from the abaxial side, with bract at base and the bracteoles at the sides. In the enlarged terminal view of the bud the outer calyx-lobes are shown above and below the central opening; the inner lobes overlap and are concealed; compression of the tissues in the young bud has given rise to curved ridges at right and left of the outer lobes. The following species resembles in many respects such South Amer- ican species as P. duplipilosa McVaugh, Fieldiana, Bot. 29: 224. 1956, but differs in detail from that and all other species known to me. Plinia salticola McVaugh, sp. nov. Arbor vel frutex puberula, foliorum nervo medio subtus ad basin copiose pilis pallidis appressis 1-2 mm. longis obsito; folia elliptica, 3.5-5 cm. lata, 11-17 cm. 506 FIELDIANA: BOTANY, VOLUME 29 longa, acute acuminata, basi subrotundata, petiolis 8-10 mm. longis, nervo medio supra convexo, nervis lateralibus utrinque 7-12; inflorescentia abbreviata, glom- erulis 4-floris, sessilibus, bracteatis; bracteis 2-seriatis, bracteolis membranaceis obtusis ciliatis 3 mm. longis; alabastra pyriformia 9-10 mm. longa, basi sericeo- strigosa, apice breviter 4-lobata, hypanthio supra germen 5-7 mm. producto, stylo 16 mm. longo, staminibus ca. 150; ovarium biloculare, ovulis quoque lo- culo duobus. A tree or shrub, the branchlets and petioles thickly puberulent with short pale ascending hairs 0.3-0.5 mm. long; hairs at the base of the midvein on the lower leaf-surface appressed and longitudinally oriented, up to 2 mm. long; leaves ellip- tic, 3.5-5 cm. wide, 11-17 cm. long, 2.5-3 times as long as wide, about equally narrowed and rounded to a slenderly and sharply acuminate tip 8-10 mm. long and to the obtusely rounded base, the curved margins meeting at about right angles and passing abruptly to the nearly terete petiole 8-10 mm. long, 1-1.3 mm. thick; midvein broadly convex above, persistently puberulent; lateral veins 7-12 on each side including some intermediate ones, inconspicuous and sometimes inscribed above, more prominent beneath, the proximal ones arcuate-ascending, those toward the middle of the blade straight, connected by asymmetrical arches 5- 8 mm. from the margin, a distinct but weaker submarginal vein 1-3 mm. from the margin; blades dark green and somewhat lustrous above, paler and finely dark-dotted beneath; inflorescence at leafless axils on old wood, the flowers in 4's, in sessile bracteate decussate pairs; bracts membranous, scarious, ciliate, ovate, blunt, 1.5-2 mm. wide, 2.5-3 mm. long; bracteoles like the bracts, ellip- soid-oblong, blunt, 1 mm. wide, 3 mm. long, probably deciduous after anthesis; buds 9-10 mm. long, pyriform, silky-strigose at base, rounded and obscurely 4-lobed at apex, the fusiform-ellipsoid base 2 mm. thick, 3 mm. long, the distal part turbinate, 7 mm. long and about 5 mm. in diameter, formed by the much- prolonged tube of the hypanthium; hypanthium at anthesis splitting irregularly between the calyx-lobes, and the divisions recurving and bearing out the petals and portions of the staminal ring; calyx-lobes in the bud short and broadly rounded, about 1 mm. long, appressed, the inner ones relatively shorter, 2 mm. wide, the outer ones bluntly ovate, 1.5 mm. wide at base; style 16 mm. long, much exserted; stamens about 150, borne in a zone about 1.5 mm. wide at the sum- mit of the hypanthium, strongly incurved in the bud, the anthers 1-1.5 mm. long; petals 4, broad-based, subrotund. about 4 mm. long and wide; ovary bilocular, the ovules collateral, 2 in each locule. COSTA RICA: Cataratas de San Ramon, 21 Feb. 1931, A. M. Brenes 1SW (F; NY, type). At the present time I should refer to Plinia any eugenioid spe- cies having 2 or 4 ovules in each locule of the ovary, the hypanthium somewhat prolonged beyond the ovary but not circumscissile, and the additional features outlined in the key above. The inflorescence is usually, but not invariably, glomerate with relatively large bracts and bracteoles simulating an involucre (cf. Amshoff in Rec. Trav. Bot. Ne"erl. 42: 11. 1950). More complete data on fruit-types, em- bryo-structure, and number of ovules per locule are needed before the limits of the genus can be firmly established. The species now McVAUGH: TROPICAL AMERICAN MYRTACEAE 507 assigned to Plinia on the basis of flower-characters seem to represent a somewhat heterogeneous assemblage, as I have pointed out (Field Mus. Bot. 13, pt. 4: 775. 1958), and it may be that some reconsidera- tion of generic lines will be indicated when new data become available. THE GENUS SIPHONEUGENA BERG The type species of Siphoneugena is S. densiflora Berg, described from somewhere in southeastern Brazil (cf. Taxon 5: 146. 1956; see also Field Mus. Neg. 31474). According to Berg the genus was distinguished from Myrciaria by having the locules of the ovary multiovulate, and from other multiovulate genera with a prolonged hypanthium by the circumscissile nature of that structure. The three original species were all Brazilian in origin, one from Minas Gerais and the others from unspecified localities. All three species have at one time or another been transferred to other genera; the genus Siphoneugena, as a matter of fact, hardly became known to botanists and was given short shrift. By Bentham and Hooker it was united with Eugenia; these authors regarded it as a connecting link between Eugenia and Myrciaria. By Niedenzu it was merged with Calycorectes and given the rank of a subgenus. No new taxa in the genus were described for a little more than a century follow- ing the work of Berg, until in 1957 Steyermark reported the discovery of what he took to be a new variety of Siphoneugena densiflora from the mountains of Venezuela. It now transpires that members of the same taxon occur on a number of the high mountains (tepuis) of Venezuela; in similar ecological situations in northern Venezuela; and in the Lesser Antilles. All the specimens from northern South America and the West Indies, specimens of the newly described S. reitzii Legrand (a native of southern Brazil; cf. Sellowia 8: 78. 1958), and the specimens shown in the type-photograph of S. densi- flora Berg, are so similar in general appearance and in structural details that they could all belong to the same species. Surely they are congeneric. The affinities of the taxon appear to be with Myrci- aria rather than with Plinia or with Eugenia as a whole. In view of the distinctive morphological features of Siphoneugena, and in view of its distinctive geographical distribution as now demonstrable, I suggest that it is best regarded as an independent genus, represent- ing an evolutionary offshoot from eugenioid stock, parallel to Myr- ciaria but in some respects less specialized. It is interesting phyto- geographically that Myrciaria is primarily a lowland genus, wide- 508 FIELDIANA: BOTANY, VOLUME 29 spread both north and south of the Amazon Basin, whereas Siphon- eugena is exclusively a montane genus, occupying a disjunct range that spans the Amazon lowland. Although Siphoneugena densiflora was reported by Berg as multi- ovulate, this is open to question. Modern specimens of the genus usually have 3 or 4 ovules in each of the two locules of the ovary. The locules vary from 1 to 3 or possibly 4; the ovules may be as few as 2, or occasionally as many as five. It is suggested that this vari- ation in the number of ovules reflects a less specialized condition than that in Myrciaria, in which the number of ovules is consist- ently two. There would appear to be little real affinity between this very homogeneous and relatively invariable taxon, and the more hetero- geneous assemblage that is called Plinia. The two have little in common except that in each the number of ovules is reduced. The South American species of Plinia, at least, are superficially dissim- ilar to those of Siphoneugena, and perhaps are more closely related to Calycorectes and to some of the larger-flowered species of Eugenia, differing from the former in the reduced number of ovules, and from the latter in the closed or nearly closed calyx. This feature of the closed or nearly closed calyx has led some workers to confuse the buds of Plinia with those of Siphoneugena. In the latter genus the dis- tinctive feature of the flower is the circumscissile hypanthium, which in most species terminates in four small but definite calyx-lobes. These lobes are in unequal pairs, closely incurved in bud so that to the casual observer the bud may appear closed. The lobes in the opening flower are usually not perceptibly imbricate, but the differ- ence in the form and position of the inner and outer pairs makes it clear that they are not valvate. The inner pair are somewhat broadly rounded, with membranous, long-ciliate margins that are appressed to the opening petals; the outer pair are smaller, deltoid- ovate, with thicker, often nearly erect and poorly fringed apex. These differences are subtle, in lobes that may be no more than half a millimeter long, but to one familiar with the imbricate calyx of Myrciaria it is apparent that in this feature there are strong simi- larities between the two genera. In 1895 Krug and Urban described from the mountains of Guade- loupe a species they called Marlierea dussii. Urban later transferred this to Plinia. An isotype (Duss 2750, K) is in my judgment con- specific with the Venezuelan Plinia fruticosa Steyermark, described from the States of Sucre (Steyermark 62611, F, type) and Anzoategui McVAUGH: TROPICAL AMERICAN MYRTACEAE 509 (Steyermark 61658 & 61620, both F). Neither species has anything to do with the other species of Plinia, and both are typical members of Siphoneugena. The question of specific limits in Siphoneugena is a difficult one, complicated by the peculiar isolation of individual small populations on the tepuis of Venezuela, on other similarly isolated mountain ranges in northern South America, and in the West Indies. Leaves vary from plant to plant: from small to large, from ovate to elliptic or obovate, from rounded to acute at apex, from rounded to acute to cuneate at base. A few specimens, like those of Marlierea dussii and Plinia fruticosa, have sparingly pubescent branchlets and /or young foliage. Pedicels and buds vary somewhat in length from plant to plant. These features make the assemblage of specimens superficially diverse, but, as indicated in the description below, the specimens have so many — and such specialized — features in common that their relationship is apparent. I believe these isolated but not excessively varied populations comprise a single species as indicated below: Siphoneugena densiflora Berg, in Mart. Fl. Bras. 14, pt. 1: 379. 1857. Calycorectes densiflorus (Berg) Ndzu. in Nat. Pflanzen- fam. 3, Abt. 7: 82. 1893. Marlierea dussii Krug & Urb. ex Urb. Bot. Jahrb. 19: 590. 1895. Plinia dussii (Krug & Urb.) Urb. Repert. Sp. Nov. 15: 413. 1919. Paramitranthes densiflora (Berg) Burret, Notizbl. Bot. Gart. Berlin 15: 541. 1941. Plinia fruticosa Steyer- mark, Fieldiana, Bot. 28: 1023. 1957. Siphoneugena densiflora var. tepuiensis Steyermark, I.e. 1024. A shrub up to 3 m. high (rarely a tree up to 12-13 m. high), nearly glabrous, the branchlets and petioles (and seldom the pedicels) sometimes strigose or pubes- cent; inner faces of the petals and calyx-lobes silky tomentose with crisped matted nearly white hairs; bracts and bracteoles ciliate, the inflorescence otherwise gla- brous; leaves coriaceous, rigid when dry, lustrous and dark green above, dull and paler beneath; blades mostly elliptic-ovate, short-petiolate, opaque and nearly featureless except that the midvein is prominently convex above and somewhat less so beneath; upper surface sometimes obscurely impressed-punctate, the lower often with small convex translucent glands; lateral veins close and parallel, the number indefinite, mostly 15-20, visible as fine convex lines on the lower surface of dry leaves; blades up to 10 cm. long, mostly 3-5 cm., mostly about twice as long as wide but sometimes as wide as long, or up to 2.7 times as long as wide; petioles fleshy, dorsiventrally compressed, (1.5-) 3-4 (-6) mm. long, ventrally concave, 1.5-2.5 mm. wide; inflorescence a much abbreviated raceme, the axis up to about 7 mm. long, usually much shorter, bearing up to 5 approximate de- cussate pairs of flowers; bracts broadly ovate-rounded, persistent, scarious; pedi- cels up to 5 mm. long, sometimes much shorter, strongly dorsiventrally compressed 510 FIELDIANA: BOTANY, VOLUME 29 or even 2-edged, up to 1 mm. wide; bracteoles persistent, scarious, distinct, ovate, blunt or acute, convex on the back, 0.7 mm. long, much shorter than the fusiform, urceolate or campanulate ovary; hypanthium prolonged 1.5-2.5 mm. above the ovary, usually notably flaring above the summit of the ovary and then contracted to the mouth; bud as a whole 3.5-4.5 mm. long; calyx-lobes in unequal pairs, scarcely imbricate, the outer pair deltoid-ovate, scarcely fringed, more erect, the inner lying flatter, broadly rounded, copiously fringed, the free parts 1 mm. wide or a little more, 0.5-0.7 mm. long; hypanthium funnelform at anthesis, splitting two-thirds of its length on one side, or into 3-4 irregular lobes, its margins finally rolled back (including most of the staminal ring) so the calyx and petals are re- flexed, the whole ultimately circumscissile at base; style 4-6 mm. long; stamens 75-100, borne on the margin of the hypanthial tube; petals white, obovate, small; locules of the ovary 2 (rarely 1 or 3), the ovules usually 4 (sometimes 3 or 5) in each locule, ascending from near the base of the locule; fruit globose or nearly so, purplish-black, soft and edible, probably 1 cm. in diameter or less; seed reniform, the 2 cotyledons distinct or fused along the margins only, the radicle very small. The type of Siphoneugena densiflora (as shown in Field Mus. Neg. 31474) is a flowering specimen with leaves 2-3.2 cm. wide, 5-7 cm. long; the pedicels are up to about 5 mm. long, and the styles 7 mm. long. The species was described by Berg as glabrous. In most of its parts this plant is larger than the majority of plants from north of the Amazon. The following, however, may be com- pared with it except that the pedicels are shorter: VENEZUELA: Amazonas: Serrania Yutaje, Maguire 35397, 35477 (both MICH); Cerro Sipapo (Paraque), Maguire & Politi 28228 (MICH, fr. only). — Dist. Federal: Cordillera del Avila, Steyermark 55604 (F; leaves as above, flowers smaller). Specimens in the groups listed below conform to the general de- scription given above for this species, but are divided into groups characterized primarily by features of the leaves. The following are glabrous; the leaves are elliptic or obovate, mostly obtuse, occasionally obscurely acuminate, 2.5-5 cm. long: VENEZUELA: Bolivar: Chimanta massif, Steyermark & Wurdack 729, 857, 1061 (all MICH); Ptari-tepui, Steyermark 59711 (F, type of var. tepuiensis) ; Sororopan-tepui, Steyermark 60157 (F). The following are somewhat pubescent; the leaves resemble those of the preceding group but include some specimens with more ovate leaves and some in which acuminate leaves are not rare; the flowers tend to be nearly sessile, often with a few on each plant pedicellate: GUADELOUPE: Duss 2750 (K, isotype of Marlierea dussii; most leaves acuminate). McVAUGH: TROPICAL AMERICAN MYRTACEAE 511 VENEZUELA: Anzoategui: Cerro Peonia above Santa Cruz, Steyer- mark 61658 (F; paratype of Plinia fruticosa; leaves mostly acumi- nate), Steyermark 61620 (F; paratype of P. fruticosa; leaves mostly obtuse). — Sucre: Cerro Turumiquire, Steyermark 62611 (F; type of P. fruticosa; leaves broader, ovate, obtuse to acuminate) . The above data support the hypothesis that there are regional populations, and perhaps local populations within these, of a wide- ranging species. 7. PIMENTA Lindl. The allspice of commerce, Pimenta dioica (L.) Merrill, is supposed by some authors to be confined to the West Indies. There is some evidence that the West Indian plants have smaller flowers and more slender panicles, and smaller fruits with a greater proportion of vola- tile oil, than do those of Central America. Because of these differ- ences, Standley proposed to recognize the Central American plant as a variety [P. dioica var. tabasco (Schlecht. & Cham.) Standl. Ceiba 3: 172. 1953], and Lundell recognized it as a species distinct from P. dioica [P. tabasco (Schlecht. & Cham.) Lundell, Wrightia 2: 58. I960]. Both of these names are based nomenclaturally on Myrtus tabasco Schlecht. & Cham. Linnaea 5: 559. 1830, presumably with the assumption that the name M. tabasco was based also on a Cen- tral American plant. Examination of the protologue of M. tabasco, however, shows that this was not the case. In reporting on the Mexican plants collected by Schiede and Deppe, Schlechtendal and Chamisso described a number of new spe- cies; most of these naturally enough were Mexican in origin. The protologue of Myrtus tabasco, however, begins as follows: "542 [the serial number under which Schiede's collections were reported]. Myrtus Tabasco H[erb.] W[illd.] n[o.] 9580. (Specimen Humb. 'Thermidor an 7. Cumana. n. 113. Myrtus Pimenta. Ta- basco').— Myrtus Pimenta, Pimenta de Tabasco Hispano-Mexicano- rum Schiede in schedula." It seems clear that Schlechtendal was merely taking up a name by Willdenow based on a Humboldt specimen from Venezuela, and identifying with this the new specimens from Mexico. Schiede's collection cannot be regarded as the type of Myrtus tabasco, and until and unless the identity of the Humboldt specimen can be de- termined, it should not be assumed that Central American popula- tions of allspice are members of the taxon to which the specimen belongs. Berg, who seems to have seen both the Humboldt speci- 512 FIELDIANA: BOTANY, VOLUME 29 men and those of Schiede and Deppe, regarded them as varietally different (cf. Linnaea 27: 424-425. 1856). His Pimento, officinalis /3 cumanensis, which he saw in Willdenow's herbarium, was presum- ably based on the very Humboldt specimen that formed the basis for Myrtus tabasco Schlecht. & Cham., and P. officinalis e tabasco (Schlecht. & Cham.) Berg was based nomenclaturally, but not know- ingly, on the same specimen. Berg described the Venezuelan plant as having petioles 3-4 lines long; the Central American 9-12 lines; his diagnoses are not otherwise very comparable. The leaves of cumanensis are described as "approximato-tenuissime venosis," a characterization that seems unlikely for the Central American pop- ulation. 8. PSIDIUM L. DELIMITATION OF THE GENUS As discussed in an earlier article (Fieldiana, Bot. 29: 172-173. 1956), the two large pimentoid genera Psidium and Myrtus have been separated traditionally upon the basis of calyx morphology. The two have in common a C-shaped or uncinate embryo and hard, bony seeds. The delimitation of Myrtus is particularly difficult, as pointed out by Bentham in an important paper many years ago (Jour. Linn. Soc. Bot. 10: 101-166. 1868), because the concepts of earlier authors included numerous species-groups now known to belong to distinct genera (even members of other sub-tribes); be- cause fruits and seeds are unknown in many supposed species of Myrtus; and because there is no sharp distinction between the num- ber of locules in the ovary of Myrtus (2-3) and in that of the closely related Psidium (3-5 or sometimes only 2). Since the time of Bentham, taxonomic opinion has favored the continuing dismem- berment of Myrtus, which at the time of the publication of the Genera Plantarum of Bentham and Hooker (1865) included more than 100 species in South America, tropical North America, Austra- lia, New Zealand, and one (the type species) in the Mediterranean region. As defined by Berg in his classic papers on the American Myrta- ceae (Linnaea 27: 397. 1856; Mart. Fl. Bras. 14, pt. 1: 413. 1857), the genus Myrtus was impossibly heterogeneous by modern stand- ards. The Chilean species, approximately one-third of the total, have all been transferred to other genera. About half the remainder were South American species that have recently been transferred by Burret to Psidium (Notizbl. Bot. Gart. Berlin 15: 483-485. 1941). McVAUGH: TROPICAL AMERICAN MYRTACEAE 513 FIG. 14. Floral morphology in Psidium. A, Bud of P. mouririoides, Gentle type (X 5). B, Axillary flower of P. salutare, showing open lobed calyx, Haught 2602 (X 2.5). C, Diagrammatic section of seed of P. guineense, showing curved embryo unshaded in center (X 10). D, Bud and old flower of P. anglo- hondurense, Bartlett 13060 (X 5). E, Lateral and terminal views of bud and old flower of P. biloculare, Gentle 3382 ( X 5). According to Burret, Myrtus is represented in the American flora by about 14 species only, these in Florida and the Greater Antilles. The type species of Myrtus, M. communis of Europe, has a 3- or sometimes 2-locular ovary, and most of the known West Indian spe- cies seem to have the ovary 2- or sometimes 3-locular. The generic limits of Myrtus are by no means agreed upon even yet, but the number of locules in the ovary, in the genus approximately as under- stood by Burret, appears usually to be 2 or 3, as against 3 or more in Psidium. Species having bi- or trilocular ovary, therefore, cannot 514 FIELDIANA: BOTANY, VOLUME 29 be assigned forthrightly to one genus or the other on this character alone. Bentham reached essentially this conclusion nearly a cen- tury ago (Jour. Linn. Soc. Bot. 10: 152. 1868). The generic limits of Myrtus being somewhat uncertain, at least in the American flora, it seems more profitable to concentrate upon Psidium, to determine what limits, if any, can be assigned to it, and to set forth criteria by which it may be distinguished from Myrtus or Myrtus-\ike plants. There are apparently several diverse evolutionary lines represented in the large American genus Psidium, and morphological variation among the species is extreme. Many species of Psidium known to the earlier authors had con- spicuously large flowers, the calyx closed or nearly closed in bud, and the ovary 3- to 5-locular. The best example of this is the com- mon guava, P. guajava, widespread in tropical regions through cul- tivation from the earliest times following the discovery of America by Europeans. The earliest concepts of the genus were based upon this species and a few others like it. Unfortunately none of the above characteristics is consistent in the genus, and none is distinc- tive. The range in size of flowers in Psidium is about the same as it is in other large genera (e.g., Eugenia, Syzygium, Calyptranthes) ; the calyx is perhaps more often open than closed; and a bilocular ovary occurs in a considerable number of species. Too much taxonomic importance has apparently been attributed to the most "unusual" feature of Psidium flowers, the completely closed calyx which opens at anthesis by irregular longitudinal splits. This prejudice seems to stem in part from the original concepts of Linnaeus and his contemporaries, but chiefly from the work of Berg, who (perhaps inadvertently) emphasized this one morphological fea- ture (Linnaea 27: 347 et seq. 1856; Mart. Fl. Bras. 14, pt. 1: 382 et seq. 1857). Although more than two-thirds of the species of Psidium known to Berg (58 out of 84 in his treatment of 1856) were assigned by him to infrageneric groups having the calyx open and normally toothed or lobed in bud, he nevertheless in his keys to genera omitted all mention of toothed or lobed calyces, saying without any qualifi- cation: "Calyx ante anthesin omnino clausus v. apice pertusus." [italics mine]. By thus highlighting unduly the characteristic of the closed calyx, Berg seems to have biased the interpretations of later authors. Bentham, for example, 20 years after the work of Berg (Gen. PI. 1: 694, 716, 720. 1865), groups Psidium in his key with those genera having the "calycis limbus in alabastro clausus v. sub- imbricato-lobatus" ; after this concession to the species with open McVAUGH: TROPICAL AMERICAN MYRTACEAE 515 calyx, however, he describes Calycorectes and Marlierea as having "Calyx Psidii." His idea of the "calyx of a Psidium" may be ascer- tained from the descriptions of the individual genera: Of Calycorectes he says later "Calycis limbus in alabastro clausus"; of Marlierea he says "Calycis tubus ... in alabastro clausus v. apice brevissime apertus." It may be inferred that Bentham thought of Psidium as a genus having the calyx typically closed in bud, even though he included in the genus some species with lobed or toothed calyx. A similar bias seems apparent in the work of Niedenzu (Natiirl. Pflanzenfam. 3, pt. 7: 64, 67. 1893), who describes the calyx of Psidium as gamosepalous ["verwachsenblattrig"], and who separates Psidium from Myrtus in the key because the calyx-lobes of the former are united ["verwachsen"] in the bud, as against "already free in the bud." Like Berg, however, Niedenzu stressed the feature of the closed calyx, and then assigned a large group of species to an intra- generic group (in this instance a new section, Apertiflora), charac- terized by having an open calyx. It seems clear that this one feature of the calyx is not to be thought of as having diagnostic significance in the determination of generic limits. A second characteristic of the calyx (or hypanthium), however, provides the single most dependable feature by means of which spe- cies of Psidium can be recognized. This feature was known to Berg, and mentioned by Bentham and by Niedenzu, but has perhaps been insufficiently emphasized. Berg's descriptions, not his keys, provide the following distinctions between Psidium and Myrtus: Hypanthium (calycis tubus auct.) supra germen productum . . ., in calycem abiens. . . . Calyx ante anthesin indivisus claususque v. apice pertusus v. 4-5-dentatus, demum longitudinaliter usque ad verticem germinis 2-5-par- titus, ex margine hypanthii libero oriundus Psidium. Hypanthium (calycis tubus auct.) supra germen . . . inferum vix productum; sepala 4-5, ex margine angusto, libero hypanthii oriunda, in alabastro globo petalorum incumbentia, . . . sinu acuto separata, serius non profundius partita Myrtus. The important point here is not the difference between a closed calyx and a calyx with 4-5 sepals, as has often been assumed, but (in Psidium) the initial prolongation of the hypanthium or calyx- tube beyond the ovary, and the subsequent irregular longitudinal splitting of these tissues down to or beyond the staminal ring and often to the very summit of the ovary. In Myrtus, on the other hand, the calyx-lobes are separate from the beginning, develop at about the level of the summit of the ovary, and fail to separate farther at anthesis, remaining contiguous so that the entire calyx 516 FIELDIANA: BOTANY, VOLUME 29 may persist as a unit on the fruit. Berg's words as quoted above may now be paraphrased: Hypanthium produced beyond the summit of the ovary, the stamens borne on the inner wall of the tube; calyx (whether closed or open at tip in bud) appearing tubular at base, in anthesis splitting longitudinally to the summit of the ovary into 2-5 somewhat irregular divisions, the individual lobes usually rather widely separated and often deciduous after anthesis Psidium. Hypanthium hardly if at all produced beyond the summit of the ovary, the sta- mens borne on a perigynous disk; calyx-lobes 4-5, distinct, fully formed in the bud, separated by acute sinuses, but not separating farther at anthesis or in fruit, the calyx not splitting between the lobes and often persisting as a unit at the summit of the fruit Myrtus. The validity of the above distinction seems to have been ac- cepted by Bentham, who in the Genera Plantarum used it in com- bination with other calyx-characters to separate the principal groups of genera in the Myrteae. The same distinction was apparently familiar to Urban, who, after Berg, had perhaps the best knowledge of the American Myrtaceae as a whole. In describing Psidium leiophloeum, for example, Urban wrote (Symb. Antill. 9: 460. 1928): "Ad seriem earum specierum pertinens, quae calyce ab initio aperto et lobis liberis sub anthesi dirumpendo paullo profundius separatis gaudent ideoque a Myrto vix v. ne vix quidem discernendae sunt." In another place (Symb. Antill. 9: 78. 1923) Urban stressed the importance of the same feature: "Myrtus et Psidium . . . inter sese arete affinia nee nisi aut calycis lobis ab initio liberis aut sub anthesi (supra basin saltern) diruptis in speciebus nonnullis antillanis dis- cernenda sunt." As Urban points out in the same place, species of Eugenia may be confused with Myrtus or with Psidium in the absence of seeds. The last two genera are quite different from Eugenia in the structure of the testa and that of the embryo, but in vegetative characters and those of the flowers, including those of the placentation, there may be no discernible differences. This brings us to the matter of the importance, if any, of the number of locules in the ovary of Psidium and related genera. The three principal groups of American Myrtaceae differ among themselves in the constancy of the number of locules. In two of the subtribes, the Myrciinae and the Eugeniinae, the number of locules is almost always 2, occasionally 3 or 4. In the Pimentinae, to which belong Psidium and Myrtus, the number is far more variable and often more than 2. Some genera, according to Berg, are in- deed always bilocular (e.g., Blepharocalyx) . In others (e.g., Caly- McVAUGH: TROPICAL AMERICAN MYRTACEAE 517 colpus, Abbevillea, Campomanesia, Acrandra, Britoa and various small genera), the number of locules is said to be always 4 or more, some- times as many as 11. In the two large genera having bony seeds and an uncinate-curved embryo, namely Psidium and Myrtus, the number of locules is said to vary from 2 to 7, and 2 to 4, respectively. In most of the species of Psidium known to Berg the ovary was said to be 3- or 4-locular (about 50 species); about twelve species were described as sometimes or always with more than 4 locules; and eight species as having but 2, 2-3, or 2-4 locules. Berg appar- ently knew but four species of Psidium having consistently a bi- locular ovary (cf. Linnaea 27: 359, 376. 1856; Mart. Fl. Bras. 14, pt. 1 : 604, 605. 1857) . He seems not to have regarded these species as in any way atypical. Among the species of Psidium described more recently by Urban are several with bilocular ovary. The number of locules in Psidium may thus be stated as usually 3 or 4, occasionally as few as 2 or as many as 7. Perhaps correlated with the number of locules in the ovary (i.e., with respect to 2- and 3-locular ovaries) is the type of placenta- tion. Altogether too little is known of the details of placentation in the American Myrtaceae, but in analyzing nearly 600 Australian species Bentham found that the "best generic characters" were afforded by the modifications of the ovules and their placentation, "provided too much weight be not attached to the precise form of the placenta." (Jour. Linn. Soc. Bot. 10: 119. 1868.) Berg seems to have attached little importance to the details of placentation in the subtribes Myrciinae and Eugeniinae, in which the ovary is generally bilocular. Usually he did not mention plac- entation in his descriptions of genera. Of Eugenia he said merely: "ovula sporophoris centralibus aflixa." He was impressed, however, by what he considered to be the parietal origin of the ovule-bearing structures ["sporophora"] among the Pimentinae. In Psidium he described these as "parietal, 2-7, reaching the axis, there connate or discrete and bilamellate, the lamellae reflexed, more or less pro- duced [i.e., radially into the locule, away from the axis], ovule- bearing on the inner surface ["intus"], connate in pairs, simulating central bilamellate sporophores." These structures were illustrated by Berg, for species having respectively 3-, 4- and 5-locular ovary (Mart. Fl. Bras. 14, pt. 1: pis. 41, 4.1 a, 4-2, 1857). Similar structures, apparently derived in the way described by Berg, can be seen in various species with multilocular 518 FIELDIANA: BOTANY, VOLUME 29 ovary. Even in some species with bilocular ovary (e.g., Psidium pedicellatum; see Flora of Peru, pt. 4: 799) the ovules occur in a double incurved row about the margins of an apparently peltate placenta. In other genera of the Pimentinae (e.g., Blepharocalyx) having bilocular ovary, Berg made no attempt to demonstrate a parietal origin for the sporophores. In the description of Blepharocalyx in the Flora Brasiliensis he said merely "ovula sporophoro centrali affixa" (p. 420). Berg's concept of the placentation in Psidium (and in the Pimentinae as a whole) seems to have been based pri- marily upon study of species with 3-locular or multilocular ovary. As pointed out above, he seems to have known few species of Psidium with bilocular ovary, and these from few specimens. In his descrip- tions of P. moritzianum and P. cordatum (both described as bilocu- lar), he described the sporophores as "bilamellate," but did not elaborate upon this. The ovules of Myrtus were described by Berg as "saepissime plurima, sporophoris bilamellatis, spurie centralibus affixa" [italics mine}. This seems to me to point up the question of the difference between the type of placentation in Psidium and in Myrtus, and between 2- and 3-locular ovaries. I suppose Berg meant that in Myrtus (especially in the 3-locular types) the sporophores were actu- ally parietal in origin, as in Psidium, but seemed to arise directly from the central axis. Nowhere did he specifically say, however, that the centrally attached placenta in the bilocular ovary was parietal in origin. We still do not have the information that will permit us to speculate upon the derivation of the placentae when these structures, even if bilamellate, seem to arise directly from the central dissepiment of a bilocular ovary. Comparative develop- mental studies are much needed. In the absence of such studies, variation in placental structure must be interpreted with caution. No correlation has been demonstrated between characters of the placenta and other characters in the Psidium-Myrtus complex, ex- cept that in species having a bilocular ovary the ovules are often attached in a single close sessile group near the middle of the central dissepiment, exactly as in many species of Eugenia. In summary, the genus Psidium may be recognized by its pi- mentoid C -shaped or uncinate embryo; hard or bony seeds; calyx (whether originally open or closed) splitting between the lobes at anthesis; ovary 3-4- (rarely 2-7-) locular; placentae usually bila- mellate but in bilocular ovaries the ovules often in a single cluster. McVAUGH: TROPICAL AMERICAN MYRTACEAE 519 ADDITIONS TO THE GENUS PSIDIUM The above review of the generic character of Psidium was occa- sioned by the discovery in Central America, in the region bordering the Gulf of Honduras, of a group of five species that are surely members of the Pimentinae, although all were originally described as species of Eugenia. These five species are vegetatively similar, with moderately large, coriaceous, glabrous or glabrate subsessile or short-petiolate leaves; the flowers are sessile or short-pedicellate, borne in small axillary clusters at leafless (or sometimes leafy) nodes on old wood; the calyx varies from completely closed in the bud to somewhat prominently lobed, but the lobes in all species sep- arating at anthesis, the base of the calyx splitting down to the ovary, which is bilocular (or in one species trilocular and in another unilocular); the ovules are attached in a subcapitate group to the central dissepiment (or a bilamellate placenta in the trilocular ovary) ; the seeds (known in three species only) are reniform, compressed, osseous, the embryo curved into a nearly complete ring. These plants, evidently very closely interrelated, are quite un- like any other species known from Central America, superficially suggesting in appearance some species of Psidium from the West Indies. They combine in a unique way several features that per- haps suggest evolutionary tendencies in the Pimentinae. The pro- duction of flowers at leafless nodes may be regarded as a character more specialized than the production of leafy or bracteate inflor- escences on young shoots. In these five species, otherwise so similar, may be seen an almost completely closed calyx (in anglohondurense) ; the calyx-lobes united their whole length but not closing the bud (mouririoides) ; or the calyx-lobes about half as long as the tube (biloculare) . The ovary is trilocular (in anglohondurense), or bilocular (in the other species), or sometimes unilocular (in mouririoides). It may be supposed that the bilocular condition is more specialized than the trilocular; the unilocular condition, if normally developed in this species, is presumably even more specialized. The unilocular condition is well known in Old World Myrtaceae, perhaps having resulted from a reduction series (cf. Bentham, Jour. Linn. Soc. Bot. 10: 119. 1868). It is of interest to note that the placentae are bilamellate in anglohondurense, associated with a trilocular ovary; in the other species the closely grouped ovules are associated with a bilocular ovary. Any generic distinction based on these char- acters, among species that are otherwise so strikingly alike, would seem to be specious. 520 FIELDIANA: BOTANY, VOLUME 29 The above five Central American species, which are apparently properly transferred to Psidium, may be separated as follows: Branchlets acutely 4-angled, the angles becoming winglike in the distal parts of the internodes and terminating in the stipular positions at the sides of the petioles; flowers sessile or nearly so, the pedicels 1 mm. long or less; ovary bilocular. Plants, including the inflorescence, glabrous or essentially so. Leaves subsessile, rounded to acute at base, 3-7 (-12) cm. long, mostly 1.6- 2.3 times as long as wide; style 6-8 mm. long; stamens 60-75. P. biloculare. Leaves on petioles 3-5 mm. long, rounded or subauriculate at base, 12-20 cm. long, 3-3.5 times as long as wide; style 9-10 mm. long; stamens about 200. P. musarum. Young leaves and branchlets, and the inflorescence, coarsely silky- or velutinous- tomentose; leaves subsessile, abruptly rounded or subcordate at base, 6- 10 cm. long, 2-3 times as long as wide; style 9-10 mm. long; stamens about 125 P. apodanthum. Branchlets terete or compressed, the nodes enlarged; plants glabrous; flowers pedicellate; calyx closed in the bud or with a small apical opening; style 5.5- 7 mm. long. Petioles 3-6 mm. long; pedicels 10-18 mm. long; buds 5-6 mm. long, apiculate, closed except a minute subapical pore; locules of the ovary 3 (-2), the ovules about 20 (-45) in each locule P. anglohondurense. Petioles 1-3 mm. long; pedicels 3-4 mm. long; buds 3-4.5 mm. long, subglobose or pyriform, the calyx-lobes united their whole length but leaving an apical opening 1.5-2 mm. across; locules 2 (-1), the ovules about 50 in each locule. P. mouririoides. Psidium apodanthum (Standl.) McVaugh, comb. nov. Eu- genia apodantha Standl. Field Mus. Bot. 17: 380. 1938. HONDURAS: Comayagua: Known only from the type collection, near Siguatepeque, Yuncker et al. 5766 (F, type; MICH). Psidium biloculare McVaugh, nom. nov. Eugenia gentlei Lun- dell, Carneg. Inst. Wash. Publ. 478: 216. 1937. Not Psidium gentlei Lundell, 1943. GUATELAMA: Izabal: Rio Chacon, Harry Johnson 1191 (F, US). BRITISH HONDURAS: Belize Dist., Gracie Rock, Sibun River, P. H. Gentle 1684 (MICH, type). Known also from several other localities: Stann Creek Dist., W. A. Schipp S-141 (F), Gentle 1887, 3382, 8552; Belize Dist., Gentle 9479, O'Neill 8761; Toledo Dist., Gentle 4062B, 4157, 4200, 4719. A specimen from Toledo Dist., near Condemned Branch Pine Ridge, Gentle 5358 (LL), is apparently this species but the calycine lobes are pubescent on the inner sur- McVAUGH: TROPICAL AMERICAN MYRTACEAE 521 face and the styles are 8-9 mm. long, almost as in P. apodanthum. An imperfect specimen, Montana del Mico, Dept Izabal, Guatemala, Steyermark 38544, is similar (F). Psidium anglohondurense (Lundell) McVaugh, nom. nov. Eugenia schippii Standl. Field Mus. Bot. 11 : 137. 1932. Not Psidium schippii Standl., 1931. Eugenia anglohondurensis [sphalm. anglohon- duransis] Lundell, Wrightia 2: 123. 1961. BRITISH HONDURAS: Stann Creek Dist.: Seine Bight, W. A. Schipp 669 (F, type of E. schippii; MICH); Commerce Bight, P. H. Gentle 8354 (LL, isotype of E. anglohondurensis). — El Cayo Dist.: Little Mountain Pine Ridge, H. H. Bartlett 13060 (F, MICH). —Toledo Dist.: Monkey River, Jenkins Creek, Gentle 4063 (F, LL, MICH). The type of Eugenia schippii was a specimen with half-grown fruit; Gentle 4063 and 8354 include mature fruit, and the latter has flowers as well; Bartlett's collection has flowers only. It is noteworthy that Gentle gives the common name "Wild Guava" for both his collections. Psidium mouririoides (Lundell) McVaugh, comb. nov. Eu- genia mouririoides Lundell, Am. Midi. Nat. 29: 480. 1943. BRITISH HONDURAS: Toledo Dist.: Monkey River, Jenkins Creek, Gentle 4142 (MICH, type). The one known specimen of this species is in some ways intermediate between P. biloculare and P. anglo- hondurense; e.g., the pedicels are of intermediate length and the calyx is in some ways intermediate (fig. 14). In view of the fact that all three species were collected by Gentle at the same local- ity within a period of about a month, they presumably grow to- gether or in associated habitats, and it may be that P. mouririoides is a hybrid. Psidium musarum (Standl. & Steyerm.) McVaugh, comb. nov. Eugenia musarum Standl. & Steyerm. Field Mus. Bot. 22: 358. 1940. GUATEMALA: Izabal: Rio Juyama, about 15 miles southwest of Bananera, Steyermark 39165 (F, type). Known also from Alta Verapaz. The only known specimens of P. musarum are sterile, or bear flowers just past anthesis. Confirmation of the generic status of 522 FIELDIANA: BOTANY, VOLUME 29 this species must await the discovery of fruiting specimens, but I believe a transfer to Psidium is justified on the basis of general similarity to the three species in which fruits and seeds are known. SPECIES DESCRIBED BY STANDLEY FROM BRITISH HONDURAS Four species of Psidium described by Standley in 1931, from a single locality near the coast of British Honduras, were all based on fruiting specimens. Their author, in differentiating them, empha- sized their differences in leaf-form, but did not compare them with P. guineense, to which they all bear more than a casual resemblance. Perhaps the most distinctive is the following, which seems to be known only from the type collection, Schipp 596 (F!). Psidium chrysobalanoides Standl. Field Mus. Bot. 8: 319. 1931. Plant nearly glabrous, the leaves and the strongly 4-angled branchlets merely sparingly pale-strigose; blades mostly obovate, abruptly deltoid-acuminate with acute tip, mostly cuneate at base, 2.5-5 cm. wide, 4-7.5 cm. long, 1.25-2 times as long as wide, on petioles 2 mm. long; midvein convex above; lateral veins 4-6 on each side, strongly ascending, forming with the small veinlets a prominent ele- vated reticulum on both surfaces; peduncles 1- or 3-flowered, 1.5-2 cm. long; immature fruit 1 cm. long, probably pyriform. In an area that has been rather well collected, it seems odd that this plant has not been found again. It is conceivable that it represents some complex hybrid involving principally P. guineense and perhaps also P. guajava and even P. salutare. In the absence of better material, any disposition of it can be little better than a guess. Psidium schippii Standl. Field Mus. Bot. 8: 319. 1931. This species, also apparently known from the type collection only, seems to have been based on a vigorous shoot of P. guineense Sw. The leaves in the type (Schipp 595, F!) are obovate and somewhat elongated toward the tip of the shoot. A 3-flowered and apparently normal inflorescence near the base of the specimen seems typical of P. guineense. Psidium rotundifolium Standl. Field Mus. Bot. 8: 318. 1931. Standley noted in the protologue of P. rotundifolium that the distinctive leaves were "usually large, rounded and nearly as broad as long." In his own later travels in Central America from 1940 McVAUGH: TROPICAL AMERICAN MYRTACEAE 523 to 1951, Standley collected (or annotated for others) approximately 30 specimens of what he took to be P. rotundifolium. Most of these were from Honduras; others came from Guatemala, Nicaragua, or Costa Rica. All tended to have large stiff leaves (Standley noted in one instance that the leaves were used as spoons). The blades are very broadly elliptic or obovate, up to 10-12 cm. wide and 12-16 cm. long, resembling superficially those of P. guineense but the red pubescence not obscuring the lower surface and the small veins forming an elaborate raised reticulum. Specimens with the smallest leaves look much like guineense, and the young growing tips when present are indistinguishable from those of guineense. Most sig- nificantly, to my mind, all the specimens that Standley called ro- tundifolium appear to be vigorous shoots, i.e., shoots bearing large fast-growing leaves; of the 35 sheets so named by Standley, 24 are sterile terminal shoots; 3 of the others seem to be ordinary flowering specimens of P. guineense; the other 8, including the type of P. rotundifolium (Schipp S85, F!), bear nothing that may be considered a normal inflorescence, but only scattered axillary pe- duncles. In the absence of distinctive flowering and fruiting speci- mens P. rotundifolium can hardly be recognized as a valid species; the evidence suggests that it is a vegetative form of P. guineense. Psidium Xhypoglaucum Standl. Field Mus. Bot. 8: 320, pro. spec. 1931. P. guineense Sw. X P. guajava L. (?). Most individuals of this so-called species bear a strong super- ficial resemblance to P. guineense, seeming to differ from it chiefly in the gray pubescence of the lower leaf-surface, and in the thinly pubescent, angled branchlets. Each of these latter features can be matched precisely in some individual specimens of P. guajava. The usual habitat of P. Xhypoglaucum is in disturbed places where other species of Psidium are growing. At one locality in Honduras, for example (Las Mesas, Dept. Morazan, at an elevation of about 900 meters), P. Xhypoglaucum was collected with P. guineense (Williams & Molina 1007 '4 includes both) at a spot where P. guajava also grew, and where the collectors obtained the so- called P. rotundifolium, P. salutare, and a pubescent form of P. salutare that bears some resemblance to P. guineense. The hybrid nature of P. Xhypoglaucum is suggested by its usual habitat and by its morphological intermediacy between two species that often grow together. 524 FIELDIANA: BOTANY, VOLUME 29 Specimens examined: BRITISH HONDURAS: All Pines, W. A. Schipp S99 (F, type). GUATEMALA: Jutiapa: Cerro Colorado just west of Jutiapa, Standley 76173, 76190 (both F). HONDURAS: Morazan: Standley 846, 12148, 22187, 22198, 22251, J. Valeria R. 520 (all F). MISCELLANEOUS NOTES ON PSIDIUM Psidiumxoerstedeanum Berg, Linnaea 27: 360, pro. spec. 1856. P. chiapasense Lundell, Wrightia 2: 204. 1961. The type locality of P. oerstedeanum was given as "In provincia Guanacarte [sic], floret Martio (Dr. Oersted), nee non in monte Rincon in Guatemala (Friedrichsthal, coll. no. 1226)." The Fried- richsthal specimen, according to Standley (MSS.), probably came from Volcan Rincon de la Vieja, Guanacaste, Costa Rica, not from Guatemala. Central American plants referred to P. oerstedeanum by Standley and others seem to differ from Psidium salutare (HBK.) Berg chiefly in having slightly broader and sometimes obovate or elliptic leaves, strongly pubescent stems and somewhat pubescent leaves, canescent buds and strongly pubescent peduncles, and somewhat more prom- inent veins in the mature leaves. Presumably the material upon which Berg based P. oerstedeanum was all of this white-pubescent plant. The Oersted specimen cited by Berg ("in hb. Oersted, no. 16"), which I have seen through the kindness of the authorities at Copenhagen, is of this description and so apparently is the speci- men of Friedrichsthal 1226 at Vienna (Field Mus. Neg. 31430). I find no significant differences between the type of P. chiapasense Lundell (Matuda 5769, an isotype, kindly lent by Dr. Lundell) and the syntype specimens of P. oerstedeanum. Very few specimens in herbaria resemble the original ones of Psidium oerstedeanum; the ones that do have the open calyx of oerstedeanum and salutare, and the canescent herbage and buds of oerstedeanum, are otherwise so diverse they can hardly represent an independent species. Two Mexican specimens from near Comitan, Chiapas (Matuda 15769, Carlson 1892) suggest by their appearance small-leaved gray-pubescent forms of P. guineense, a supposition strengthened by the reddish color of the pubescence of immature shoots. Lundell noted that the peduncles of his P. chiapasense McVAUGH: TROPICAL AMERICAN MYRTACEAE 525 were sometimes 2-flowered, a feature that also suggests some con- nection with P. guineense. A clue to the origin of the so-called Psidium oerstedeanum seems to lie in its frequent close association with P. salutare, a species inhabiting lowland savannahs from Mexico to the Guianas. Ordi- narily P. salutare is glabrous, with mostly lanceolate or lance-ovate narrowly pointed leaves. In various localities in Central America and southern Mexico, however, the plants are somewhat pubescent, but otherwise seemingly unchanged. At some localities, moreover, the narrow-leaved glabrous plants and narrow-leaved pubescent plants grow intimately associated with canescent plants that have even broader leaves (i.e. the "oerstedeanum" type). The type of Psidium gentlei Lundell, for example (British Honduras, Monkey River, Jenkins Creek, Gentle 4062, MICH), is a narrow-leaved plant with glabrous leaves and with but scanty pubescence on the branches; I should refer it without hesitation to P. salutare. Two other collec- tions from the same locality and taken apparently at the same time, Gentle 4062C and 4062D, have quite a different aspect, viz. that of small-leaved specimens of P. guineense; they are pubescent and broad-leaved and would be referred to P. oerstedeanum if that species were to be recognized. Two collections from Honduras (Standley 21236 and Williams & Molina 10484) are similarly mixed; each collection includes one branchlet with narrow glabrous leaves and glabrous twigs, and one or more with broader, pubescent leaves and pubescent inflorescence. These are from a disturbed pastured area to which allusion has already been made above (under P. Xhypoglaucum) , where natural hybridization is suspected. It seems likely that the broad-leaved, somewhat variable but always densely pubescent or canescent plants called Psidium oer- stedeanum have resulted from hybridization between P. salutare and some other species, perhaps P. guineense, which is common and weedy and variable throughout Central America. Psidium sartorianum (Berg) Ndzu. in Engl. & Prantl, Natiirl. Pflanzenfam. 3, Abt. 7: 69. 1893. Mitranthes sartoriana Berg, Lin- naea 29: 248. 1858. At the time of the publication of the genus Mitranthes it was assigned by Berg to the subtribe Eugenioideae. One wonders why such an astute student of the Myrtaceae as Berg did not suspect some relationship between Mitranthes and Psidium, in view of his 526 FIELDIANA: BOTANY, VOLUME 29 descriptions of the several species of Mitranthes as having a closed, circumscissile and operculate calyx, a multiovulate 3-locular ovary, and the "ovulis sporophoris bilamellatis afRxis." Apparently, how- ever, he never saw fruits or seeds of any of the original eight species of Mitranthes, and thus could never have been entirely sure of the systematic position of the genus (see Linnaea 29: 248-249. 1858, for the complete list of these species, including M. sartoriana). Most recent authors have agreed that Mitranthes belongs taxo- nomically with Psidium rather than with Eugenia, but there has been no general agreement as to the exact generic status of the group. Niedenzu (in the Natiirlichen Pflanzenfamilien) proposed a revised classification of Psidium and related genera, erecting a new infra-generic group, Psidium sect. Calyptropsidium (Berg) Ndzu., to include the one species referred by Berg to the genus Calyptropsidium, and five species, including P. sartorianum, that Berg had referred to Mitranthes. Urban (Bot. Jahrb. 19: 571. 1895) re-established Calyptropsidium as a genus, and transferred to it Mitranthes sartoriana. Standley, in the Trees and Shrubs of Mexico (Contr. U. S. Nat. Herb. 23: 1035. 1924), followed Niedenzu in treating this species as a Psidium, and in this he has been followed by most recent American workers. Burret, however, in 1941 (Notizbl. Bot. Gart. Berlin 15: 486), created a new genus, Mitropsidium, based primarily on Psidium sect. Calyptropsidium (Berg) Ndzu. but including three additional species with operculate calyx, and excluding the type species of Calyptropsidium Berg, C. friedrichsthalianum, on the grounds that in that species the calyx opens irregularly and not by a true calyptra. Burret argued that Mitropsidium constituted a taxon closely re- lated to Psidium, the two genera perhaps not entirely monophyletic, and separated somewhat artificially by the one character of the calyx; he used for support of his argument the universally recog- nized genera Calyptranthes and Marlierea, separated artificially by the same character. The most recent student of this particular group, Miss G. J. H. Amshoff, pointed out that the character of the circumscissile calyx "has to be used with circumspection as e.g. in Psidium (Calyptro- psidium) Friedrichsthalianum (Berg) Ndz. the calyx is sometimes calyptrate and sometimes vertically dehiscent" (Act. Bot. Ne"erl. 5: 277. 1956). In the Flora of Panama (Ann. Missouri Bot. Gard. 45: 198. 1958) Miss Amshoff treated Mitranthes-Calyptropsidium-Mitro- psidium sartorianum as a species of Psidium, with no comment as McVAUGH: TROPICAL AMERICAN MYRTACEAE 527 to its generic position but noting in her description that the calyx usually splits irregularly, "with 1 large more or less lid-like deciduous segment and 2-3 other minute persistent segments," and stating specifically that a neatly circumscissile calyx is exceptional. It seems probable that too much emphasis has been placed upon this one feature of the calyx in determining generic lines. It is at best a variable feature, even in individual species; furthermore it has not been shown that the species having in common the cir- cumscissile calyx have also other shared attributes to set them apart from Psidium. There seem to be no sharp distinctions in Psidium between groups of species with closed buds and those with open buds. Two Amazonian species, for example, P. guianensis Pers. and P. densicomum DC., are essentially indistinguishable ex- cept that the former has a closed bud and the latter an open, shal- lowly lobed calyx. The specific limits of Psidium sartorianum are somewhat in doubt also. Several so-called species have been distinguished from it; most of these have been based partly upon features of the calyx, but in part also on leaf-characteristics. The entire species-complex should be reviewed. The leaves are usually elliptic and short- or obtusely pointed, but in some parts of the range, particularly in Mexico from Jalisco to Guerrero, in the Yucatan Peninsula, and in Honduras, many or most individuals have long-acuminate or at least long-acute leaves, often of a lanceolate type. This local popu- lation may be appropriately designated as follows: Psidium sartorianum (Berg) Ndzu., var. yucatanense McVaugh, var. nov. A var. sartoriano foliis lanceolatis vel ovato-lanceolatis longiuscule acutis acu- minatisve differt. This variety is typified by a specimen from British Honduras, P. H. Gentle 9 (MICH), which is also the type of Psidium yucatanense Lundell (Contr. Univ. Mich. Herb. 7: 35. 1942). I cannot dis- tinguish specimens of P. yucatanense from Cuban specimens de- scribed as Calycorectes protractus Griseb., the type a specimen col- lected by Charles Wright (isotype, US!) (Griseb. Cat. PI. Cub. 284. 1866). If these Cuban and Central American plants together con- stitute a distinct species, it will be necessary to provide a new specific epithet in the genus Psidium, the name Calycorectes protractus having 76 years' priority over P. yucatanense. If, as in the present paper, this taxon is regarded as a variety of P. sartorianum, the name var. 528 FIELDIANA: BOTANY, VOLUME 29 yucatanense may be used. It should be noted that in accordance with the International Code of Botanical Nomenclature (Art. 60, 72) this is to be treated as a new varietal name, not a new combination based on P. yucatanense Lundell, which is regarded as a taxonomic (not nomenclatural) synonym of Calycorectes protractus and therefore undesirable as a basis for a name in a new status. Dr. Lundell contrasted his new species with P. sartorianum, em- phasizing especially the point that in P. yucatanense the basal part of the calyx "above the ovary splits into five irregular lobes at anthesis." I do not find this feature of diagnostic value, for in most specimens of undoubted P. sartorianum the calyx splits in a similar fashion at or after anthesis. There is no doubt, however, that all specimens from the Yucatan Peninsula have a distinctive appearance because of the slender-pointed lanceolate or lance-ovate leaves. The following belong here: CUBA: Santa Clara: Dist. Cienfuegos, Cieneguita, Combs 201 (F). YUCATAN: Chiche"n Itza, Steere 1293; Tizimin, Swollen 2502; Yakdzonoot, Lundell 7494 (all MICH). BRITISH HONDURAS: Belize Dist., Belize-Sibun Road, P. H. Gentle 9 (F; MICH, type, and type of P. yucatanense Lundell). GUATEMALA: Pete"n: Tikal National Park, E. Contreras 1340 (LL, MICH). This local population of the Yucatan Peninsula hardly seems specifically different from P. sartorianum of the Mexican mainland. A paratype of P. yucatanense, for example (Steere 1293}, is a good match for a Karwinsky specimen identified by Berg as Mitranthes sartoriana (Field Mus. Neg. 19745). In western Mexico, moreover, where P. sartorianum is abundant, individual plants of the "yuca- tanense" type, i.e., with acute and more or less lanceolate leaves, are frequent but do not seem to differ otherwise from the plants with more obtuse elliptic leaves. Some specimens are cited below: MEXICO: Jalisco: Near Bahia de Tenacatita, McVaugh 20965 (MICH) ; 4 miles north of Bahia Navidad, McVaugh 20832 (MICH). —Guerrero: Acapulco, Palmer 357 (F), MacDaniels 149 (F). Another plant that seems to have close affinities with P. sar- torianum is P. molinae Amsh. (Act. Bot. Ne"erl. 5: 277. 1956) (Hon- duras, the type from Dept. Morazan, Standley 21255, F!). This was contrasted by its author with P. sartorianum, with the state- ment that the latter species is at once distinguishable "from our new one by its calyptrate calyx and shortly acuminate leaves" McVAUGH: TROPICAL AMERICAN MYRTACEAE 529 (P. molinae having "alabastra . . . apice minute aperta, brevissime 4-loba. Calyx ... in flore adulto irregulariter 4-fissus," and the leaves obtuse). The type of P. molinae bears a few buds that open in the way described by Amshoff, i.e., the calyx is 4-lobed, open at the tip, with thick incurved-deltoid lobes 1 mm. long. One other specimen, Williams & Molina 14143 (F), also in bud, is similar to the type. The other specimens cited by Amshoff are sterile or in the fruiting condition, and apparently were selected on the basis of leaf-shape only. The leaves are indeed somewhat obtuse, but not more so than in about half the available specimens through- out the range of P. sartorianum. Miss Amshoff points out that a specimen from Oaxaca (Makrinius 609} has "the flowers of P. molinae and the leaves of P. sartorianum." A plant similar to P. molinae is P. microphyllum Britton (Sci. Surv. P. R. & Virgin Is. 6: 555. 1930), based on a cultivated plant from the Agricultural Experiment Station at Mayaguez, Puerto Rico (J. B. McClelland, July 1930, NY, type!). The leaves are relatively obtuse, and the buds apparently are never fully closed; there is an apical pore about 0.3 mm. in diameter, and minute lobe- like hairy appendages surround the pore; the calyx in an thesis splits irregularly, one lobe often much larger than the others but not forming a calyptra. I do not know how much importance is to be attached to the single feature of the calyptrate vs. non-calyptrate calyx. Super- ficially the two seem very different, but, as pointed out above, the calyx in a normally calyptrate species may sometimes open by irregular longitudinal splits. Superficial examination of a number of specimens of P. sartorianum shows a great deal of individual variation in the "neatness" with which the calyptra is cut off; sometimes it is very irregular, resembling a large calyx-lobe opposing the smaller irregular lobes around the circumference of the hypan- thium. I think it probable that this calyx-character is not a very stable one, genetically speaking, and that a form with open calyx might well appear in a large population like P. sartorianum. Addi- tional series of specimens of plants in bud and in flower are needed to establish the taxonomic position of P. molinae and P. micro- phyllum. INDEX New names in bold -face type Anamomis fragrans, 485 Calycorectes densiflorus, 509 Calyptranth.es apoda, 412 aromatica, 411 costa-ricensis, 408 chytraculia, 403 var. americana, 404 var. chytraculia, 403 5 pauciflora, 407 euryphylla, 405 hernandezii, 404 johnstonii, 405 mexicana, 409 millspaughii, 405 pallens, 406 var. mexicana, 409 var. pallens, 407 var. williamsii, 408 paxillata, 410 schiedeana, 411 schlechtendaliana, 411 tenuipes, 411 tovarensis, 412 vexata, 412 williamsii, 408 zuzygium, 412 Eugenia acapulcensis, 421 alaternifolia, 481 alnifolia, 425 andina, 481 anglohondurensis, 521 antiquae, 421 apodantha, 520 argyrea, 426 aromatica, 481 asa-grayi, 503 avicenniae, 427 axillaris, 428 bartlettiana, 421 baruensis, 421 basilaris, 464 biflora, 428 bonplandiana, 421 bracteolosa, 421 buxifolia, 441 calciphila, 468 campechiana, 421 capuli, 432 carthagenensis, 421 /3 baruensis, 421 casearioides, 433 chepensis, 434 chinajensis, 449 choapamensis, 435 cocquericotensis, 446 colipensis, 436 comitanensis, 421 contrerasii, 432 conzattii, 446 crassifolia, 467 crebrifolia, 434 crenularis, 436 culminicola, 437 deltoidea, 421 dissitiflora, 467 divaricata, 502 doubledayi, 470 dugandii, 484 egensis, 466 escuintlensis, 421 farameoides, 439 fieldingii, 428 fiscalensis, 454 flavida, 467 flavifolia, 439 flavoviridis, 467 florida, 446 foetida, 453 18 maleolens, 454 7 rhombea, 453 fragrans, 485 gaumeri, 469 gentlei, 520 granulata, 487 guadalupensis, 421 guanacastensis, 465 guatemalensis, 439 hintonii, 436 hiraeifolia, 455, 457 hondurensis, 467 hypargyrea, 439 inconspicua, 440 itzana, 425 jutiapensis, 465 karsteniana, 490 karwinskyana, 440 koepperi, 449 laevis, 468 var. gaumeri, 469 var. laevis, 468 530 INDEX 531 ledophylla, 430, 440 leptppa, 453 1'heritieriana, 503 liebmannii, 441 lindeniana, 442 lopeziana, 485 lundellii, 469 macrocarpa, 442 mariquitensis, 491 maritima, 421 mato, 491 mexiae, 455 mexicana, 442 michoacanensis, 443 micrantha, 433 minimifolia, 492 mirandae, 444 monticola, 432 var. latifolia, 432 mosquitensis, 421 mouririodes, 521 musarum, 521 nicaraguensis, 467 oaxacana Berg, 445, 455, 456, 458 oaxacana Standl., 455, 456, 458 octopleura, 470 oerstedeana, 446 oreinoma, 465 origanoides, 448 orthostemon, 492 ovatifolia, 421 pallens, 406 palmarum, 503 paplensis, 465 periplocifolia, 471 petenensis, 446 pittieri, 433 pleurocarpa, 449 polyneura, 503 praeterita, 450 principium, 451 pseudo-mato, 493 pueblana, 452 punctata, 485 purpusii, 455, 456 pycnantha, 494 pycnoneura, 503 rekoi, 453 rhombea, 453 rigidissima, 494 roepperi, 470 rondonensis, 495 rubella, 449 salamensis, 455 var. hiraeifolia, 457 var. rensoniana, 456 var. salamensis, 455 schiedeana, 432 schippii, 521 sibunensis, 421 sinaloae, 457 standleyi, 458 steyermarkii, 485 storkii, 496 subverticillaris, 468 symphoricarpos, 458 tabascensis, 429 teapensis, 459 tenuissima, 434 ternifolia, 497 tikalana, 435 tomentulosa, 455 triflora, 485 triquetra, 497 truncata, 465 trunciflora, 460 turneri, 460 turumiquirensis, 497 vincentina, 446 xalapensis, 461 xilitlensis, 462 yautepecana, 463 yucatanensis, 429 Marlierea dussii, 509 Mitranthes sartoriana, 525 Myrcia arpmatica, 411 belizensis, 472 costa-ricensis, 471 discolor, 471 fallax, 472 karsteniana, 490 longicaudata, 471 melanoclada, 471 mollis, 471 oerstedeana, 471 plicato-costata, 471 rufidula, 470 sartoriana, 471 saxicola, 431 /3 grandifolia, 431 schippii, 471 seleriana, 485 splendens, 470 tovarensis, 412 Myrcianthes biflora, 485 borealis, 481 callicoma, 482 compressa, 483 crebrifolia, 484 dugandii, 484 foliosa, 491 fragrans, 485 var. fragrans, 485 var. hispidula, 489 irregularis, 489 karsteniana, 490 leucpxyla, 491 mariquitensis, 491 mato, 491 minimifolia, 492 prodigiosa, 492 pseudo-mato, 493 rhopaloides, 481, 483 sessilis, 495 storkii, 496 532 FIELDIANA: BOTANY, VOLUME 29 ternifolia, 497 xalapensis, 461 Myrciaria amazonica, 501 zuzygium, 412 caurensis, 501 cordata, 501 Paramitranthes densiflora, 509 divaricata, 502 Pimenta dioica, 511 dubia, 501 tabasco, 511 floribunda, 502 Plinia asa-grayi, 503 ibarrae, 503 dussii, 509 maragnanensis, 502 fruticosa, 509 paraensis, 501 salticola, 505 phillyraeoides, 503 Pseudomyrcianthes pseudo-mato, 493 pittieri, 503 Psidium apodanthum, 520 verticillata, 502 anglohondurense, 521 vexator, 503 biloculare, 520 vismeifolia, 504 chiapasense, 524 Myrteola acerosa, 431 chrysobalanoides, 522 microphylla, 432 dubium, 501 Myrtus acerosa, 430 gentlei, 520 axillaris, 428 guajava, 523 biflora, 428 guineense, 523 casearoides, 433 hyppglaucum, 523 capuli, 432 molinae, 528 chytraculia, 403 mouririodes, 521 coccolobaefolia, 483 musarum, 521 compressa, 483 oerstedeanum, 524 ehrenbergii, 442 rensonianum, 455, 456, 457 fragrans, 485 rotundifolium, 522 foliosa, 491 sartorianum, 525 ledophylla, 430 var. yucatanense, 527 leucpxyla, 491 schippii, 522 maritima, 421 yucatanense, 527 micrantha, 433 oaxacana, 458 Siphoneugena densiflora, 509 splendens, 470 var. tepuiensis, 509 trtmciflora, 460 reitzii, 507 Publication 968 UNIVERSITY OF ILLINOIS-URBANA