WEST INDIAN VERNONtfE ACADEMICAL DISSERTATION BY E. L. EKMAN UPPSALA 1914 ALMQVIST & WIKSELLS BOKTRYCKERI-A.-B. WEST INDIAN VEKNOM ACADEMICAL DISSERTATION BY E. L. EKMAN WITH 6 PLATES By permission of the Philosophical Faculty of the University of Lund to be publicly defended in the Lecture Room of the Botanical Institution on March 2nd, 1914, at 10 o'clock a. m. for the degree of Doctor of Philosophy UPPSALA 1914 ALMQVIST & WIKSELLS BOKTRYCKEBI-A.-B. ,Ce7f ARKIV FOR BOTANIK. BAND 13. N:o 15. LfBRARY West Indian Vernoniae. Rn?ll°fi botanical By flAKOfc/v, E. L. BKMAN. With 6 plates. Communicated January 14th by G. Lagerheim and C. A. M. Lindman. Introduction. Six years ago, the author of this paper made a journey to South America in order to study its flora. In the Argentine Territorium of Misiones I had an opportunity to stay for sev- eral months, and I there brought together considerable collec- tions of plants, the determination of which has been carried on at the Natural History Museum at Stockholm. I soon found that the determination of certain genera offered great difficul- ties. My attention was especially drawn to the classification of the Vernoniae in the great Regnellian Herbarium at Stock- holm, but I soon extended my studies to include all Brazilian Vernoniae. Some taxonomic questions obliged me to examine the value of several West Indian species. While working with the Vernoniae of the West Indies, I consulted, in the first place, an important paper entitled »A Revision of the North American Vernonieae », having for its author Mr. Hen- ry Allan Gleason from Chicago. At first sight, this work inspired confidence by the apparently practical and natural arrangement of the species, and by their accurate delimi- -tation supported by geographical facts. But I soon was able -to state that the author had not the right conception of the species of Swartz, the types of which are in the Stockholm Arlciv for botanik. Band 13. N:o 1~>. 1 >- 2 ARKIV FOR BOTANIK. BAND 13. N:0 15. Herbarium. At first, I intended to publish some corrective remarks to Gleason's Revision, but I soon made up my mind to give a full account of all West Indian species of Ver- nonia, worked out upon a basis of type specimens available in European herbaria. In the course of my work I have been able, through the kindness of the curators of the Regnellian funds, to visit the following herbaria: Berlin. The herbarium of the Konigliches Museum contains the types of the species described by Lessing, Will- denow, and Sprengel, further, the rich collections of the Krug-Urban Herbarium, which through the courtesy of Prof. Ign. Urban have been sent to me on loan, and are the main basis of my studies. Brussels. The herbarium of the Jardin Botanique is of especial interest to students of the South American Ver- noniae, as it contains the types collected by Manso, Lind- berg, and others. Geneva. The Prodromus Herbarium contains the types of De Candolle. In the Delessert Herbarium I have found duplicate specimens of the types of some French and Spanish botanists, viz. L. C. Richard and La Llave et Lex- arca. The Boissier Herbarium contains a very good series of Wright's Cuban plants. London. The Kew Herbarium possesses the tj-pes of Gardner. Munich. Here the types of Martius are kept, described in the Prodromus by De Candolle. Paris. The collections of the Museum d'Histoire Na- turelle are of the greatest importance, containing the types of Lamarck, L. C. Richard, Humboldt-Bonpland-Kunth, Cassini, and others. The Cosson-Durand Herbarium, which is kept apart, contains the types of Schultz-Bipontinus. I have, further, had on loan some material of Vernonia from the British Museum and the Museums of Copenhagen, Florence, Gottingen (of especial interest, containing the types of Grisebach), Lund, St. Petersburg, Upsala, and Vienna. I beg to render my sincerest thanks to the di- rectors of all herbaria now mentioned. When quoting the specimens, I have indicated, in parenthesis, in which herbarium they are to be found. The letters in parenthesis signify: E. L. EKMAN, WEST INDIAN VERNONLE. 3 B the Berlin Herbarium. BB ... .... the Boissier-Barbey Herbarium. Geneva. Br the Herbarium of British Museum. Bss -. the Boissier Herbarium, Geneva. CD the Cosson-Durand Herbarium, Paris. D — the Delessert Herbarium, Geneva. DC the De Candolle Herbarium, Geneva. F the Florence Herbarium. G — the Gottingen Herbarium. H — the Copenhagen Herbarium. K the Kew Herbarium, London. KIT the Krtjg-Urban Herbarium, Berlin. L the Lund Herbarium. M the Munich Herbarium. P the Paris Herbarium. Petr the St. Petersburg Herbarium. Prodr the Prodromus Herbarium, Geneva. R the Regnell Herbarium, Stockholm. S the Stockholm Herbarium. LT — - the Upsala Herbarium. I regret very much that I have had no opportunity to visit the Desfontaines Herbarium in Florence, containing some important types of Cassini. Quite recently (July 1913, in Bull. Torr. Bot. Club, Vol. 40, No. 7) Gleason has described a set of species from the West Indies, of which I have not seen authentic specimens. .Has, I cannot delay the publication of this paper until du- plicate specimens are available in Europe, for I am very soon to undertake a journey to Hispaniola and eastern Brazil. Finally I beg Prof. Ign. Urban in Berlin to accept my sincerest thanks for his most valuable assistance in every possible way. Further, I am deeply indebted to Prof. C. A. M. Lindman in Stockholm for his never-failing kindness, and innumerable valuable hints and informations. General remarks. In the present paper only species of the genus Vernonia have been dealt with. The delimination of the genus adopted here, on the whole, is the same as that of Hoffman in Engler and Prantl: Die 4 ARK1V FOR BOTANIK. BAND 13. N:0 15. Naturlichen Pflanzenfamilien. A few modifications have, however, been made, chiefly according to the works of Glea- son. Like many North American botanists, he tends to split up large genera into smaller ones. I agree with him in removing Lachnorhiza A. Rich., Leiboldia Sch.-Bip., and Cyanthillium Blume (= Cyanopis Blxjme) from Vernonia, but I do not approve the segregation of Eremosis (DC). There are numerous species connecting Eremosis with Critoniopsis Sch.-Bip., which genus always has been considered a section of Vernonia. In fact, there are no characters of Eremosis that cannot be found again in certain species of other sec- tions of Vernonia. Eremosis (DC.) Gleason was character- ized by 1 — 5-flowered heads in rounded panicles, and Ver- nonia by 8 — 88-flowered heads. However, there are many true Vernonise having as few-flowered heads as have the spe- cies of Eremosis, for instance, in the sections Critoniopsis and Stenocephalia (Sch.-Bip.). As for the rounded pan- icles, Gleason himself has defined a division of Lepidaploa, Paniculataz umbelliformes, having »subumbellate, rounded panicles ». Three isolated species have been transferred in this paper, one to the genus, and two from it. Prof. Urban in 1899 described a Piptocarpha tetrantha from Porto Rico, nearly allied to Piptocarpha tri flora Bennet from Guj^ana. The latter Piptocarpha is a genuine Critoniopsis, being closely related to the type of the genus, Critoniopsis Lindenii Sch.- Bip. from Columbia. If Critoniopsis be referred to Verno- nia, Piptocarpha tri flora Benn. and Piptocarpha tetrantha Urb. must be so too. It is another question whether Cri- toniopsis inclusive of Eremosis are not to be genericalh* distinguished from Vernonia. I will return to this subject elsewhere. The two plants removed from the genus Vernonia are V. lepidota Grisebach and V. Milleri Johnston. The former, differing already in habit from all true Vernonia? of the West Indies, above all in its lepidote pubescence, is abundantly distinct in the structure of its pappus. The bristles of the inner series of pappus are very few, 5 — 6 — 7, broad, caducous, the scales of the outer series are concreted with each other, form- ing a short, cylindrical tube. The plant is related to Pip- tocoma Cass., which genus, however, has free outer scales. E. L. EKMAN, WEST INDIAN VERNONLE. 5 Most probably Vernonia lepidota Grisebach is to be made the representative of a new genus. Vernonia Milleri Johnston from the island of Margarita was described in the year 1905. It is evidently an Oliganthes, having few-flowered heads, narrow involucres, and broad, somewhat curled scales of the inner series of pappus. As for the delimitation of the sections of the genus, I have adopted the treatment given by Hoffman in Engler and Prantl: Die Naturlichen Pflanzenfamilien, with the except- ions already mentioned. Save Piptocarpha tetrantha Ur- ban, and some introduced species, all West Indian Verno- nia) belong to the section Lepidaploa (Cass.) DC. Within this section Lepidaploa the species have been before exclusively artificially grouped. The first attempt to arrange the species according to their relationships was made by Gleason in his »Revision of the North American Vernonieae». He correctly recognized several groups of species, for instance, the Fruticosce and the Argyropappce, but most of his other groups include species of different ori- gin. He also has proceeded too far in splitting up the divi- sions of Lepidaploa, placing, for instance, V. albicaulis Pers. (— V. longi folia Pers.) in a particular group Longifolice, separated from its nearest relative, V . icosantha DC. (= V. arbor escens Sw\ of this paper). Especially unfortunate is he in arranging the forms of V. divaricata Sw. , some of which he takes to be distinct species, putting them into three different species -groups (V. arborescens Gleason, not Sw. among his Arbor escentes, V. albicoma among the Divaricate, V. permollis and V. intonsa constituting a particular spe- cies-group). He is, further, absolutely wrong in assigning V. lepidota Griseb. to the Scorpioidece reductce, V. pallescens Gleason to the Havanenses, V. canescens H. B. K. to the Dep- peanaz, many other mistakes not to be mentioned. I have tried, in this paper, to arrange the species accord- ing to their relationships in subsections based in the first place on the characteristics of the inflorescense, in the second place on the pubescence of achenes and corollas and on the structure of the pappus. The arrangement proposed is by no means definitive. All delimitation of taxonomic units in Vernonia is extremely difficult, from that of the species up to the genus. 6 ARKIV FOR BOTANIK. BAND 13. N:0 15. The differences existing as to the shape of corollas, size of anthers etc. have been generally regarded as being of little importance in delimitating species and groups in Vernonia. In order to ascertain if these characters actually have no value, I have analyzed numbers of flowers of all the species mentioned in this paper, and of many South American ones. In general, the characters found agree with those of habit and inflorescence, and no striking results have been obtained. Yet they have proved to be of same value for recognizing smaller groups of the subsections and for identifying the spe- cies. More interesting is the fact stated through these exam- inations of flowers that the basal lobes of the anthers, which are said to be obtuse in Vernonia, sharply acute in Pipto- carpha, very often are acute in Vernonia too. This should mean that the only difference between Piptocarpha and Ver- nonia vanishes. I hope I shall have an opportunity to return to this in another paper. As characters of the flowers are seldom mentioned in the descriptions of Vernoniae, I have under each species' listed in this paper given its analytical distinctions. As to the species themselves, there will be found to exist a certain difference between my treatment of the West In- dian ones and that of Gleason, not only in regard to their names, but to their delimitation. In Gleason's paper the species are very small, narrowly defined and, as a rule, based upon a very meagre material; my species are in general much wider, including several forms described b}T him as proper species. Yet my work also has been compiled upon the principle: to separate what can be separated practically. It may be that Gleason has another conception of species than I have, as North American botanists on that point often differ from European ones (cfr. Otto Kuntze, Rev. gen. plant., Band III11, 1898, p. 144, the note). It is, possibly, be- yond his reach to conceive, for instance, the striking common features of his V. arborescens, V. intonsa, V. permollis, V. albicoma and V . amaranthina, which is the evidence of their being all merely forms of one species. He keeps, apparently, his mind open to the differences, not to the affinities. But he will, I am sure, not fail to find that, with the descrip- tion of new species like the above mentioned, the delimita- tion of the already existing ones will become always more E. L. EKMAN, WEST INDIAN VERNONISE. 7 difficult, and that all, at last, will end in chaos. He wouid be justified in describing his forms as species, if he could state that they are constant in the same manner as the micro- -species of Alchemilla, Hieracium, and Taraxacum, where apogamy is the basis of the Jordanismus. I do not mean, of course, that all species described by Gleason are inferior ones. On the contrary, he has had the opportunity of de- scribing some of the most striking of the West Indian species. What apogamy is, for instance, for the parthenogenetic species of Alchemilla, a geographical isolation may prove to be for the West Indian Vernonise. I have myself distin- guished V. amgustissima Wtright mscr. from V. stenophylla Less, in part owing to the fact that the former is a Cuban plant, the latter a member of the flora of Hispaniola. And another instance, V. sericea L. C. Rich., growing in Hispan- iola, has been found to differ from the Porto Rico plant in having invariantly larger anthers; it will consequently be segregated here, }7et only as a subspecies, the habits of the two plants being nearly identical. As matters now stand, the three larger Antilles, Cuba, Jamaica, and Hispaniola. have no common species of Vernonia, except the cosmo- politan V. cinerea Less. I confess to have learnt one thing from the North Ame- rican botanists — not to describe innumerable varieties under each variable species, as do many European botanists. In dealing with Vernonise, one should be satisfied to state to which species a certain plant belongs. In order to give an idea of the variability of some species, I have, however, de- scribed some of their most striking forms, though without indicating any form by a proper name. As to the history etc. of the genus Vernonia in the West Indies, I refer to the interesting exposition given by Gleason in his »Revision». Several dates on the same object will be found under the respective species in this paper. The name of each species accepted here is the earliest valid one, as governed by the Vienna rules of Nomenclature, adopted at the International Botanical Congress in Vienna 1905. 8 ARKIV FOR BOTANIK. BAND 13. N:0 15. Vernonia Schreb. Key to the West Indian sections. I. Inner involucral scales with dilatated, membranaceous, purplish-green tips. Bristles of the inner series of pappus broad, flat, caducous. Tube 7/s m length of the corolla. Sect. Stengelia. II. Tips of the scales not dilatated. Bristles of the inner series of pappus filiform or very inconspicuously flattened. Tube not over 4/s °f the corolla. A. Inner scales caducous. Sect. Critoniopsis . B. Inner scales persistent. 1. Inner bristles of pappus persistent. Achenes con- spicuously 10-costate. Sect. Lepidaploa. 2. Inner bristles caducous. Achenes faintly ribbed. Sect. Tephrodes. Sect. Stengel ia (Sch.-Bip.) Benth. Yernonia anthelmintica (L.) Willd. Conyza anthelmintica Linnjeus, 1763, p. 1207. Vernonia anthelmintica Willdenow, 1804, p. 1634. Pappi setae interiores caducae, numero circ. 30(?), fir- mas, valde complanatae, pallide straminese, exterioribus late linearibus circ. 8-plo longiores. Corolla glabra, 9 mm longa, tubo gracili, superne subito dilatato, long. 7/s corolla?, limbi laciniis brevibus, apice papillosis. Antherae fere 2 mm longae, ligula long. Vr, antherae, auriculis acutis. Hab. in Jamaica: Green Harbour, alt. 540 m, 13. 12. 94, Harris n. 5472 (BB, Br, KU, M). A well known, widely distributed species of the Old World, introduced in Jamaica. Sect. Critoniopsis (Sch.-Bip.) Benth. Yernonia tetrantha (Urb.) Ekman nov. comb. riptocarpha tetrantha Urban, 1899. p. 457; 1911, p. 620; Gleason. 1900 1), p. 160. Hab. in Porto Rico. E. L. EKMAN, WEST INDIAN VERNONIA. 9 As to the description and distribution of the species I refer to Urban, 1. c. The genus Critoniopsis Sch.-Bip. was based upon a plant from Columbia, Critoniopsis Lindenii Sch.-Bip., nearly re- lated to Piptocarpha triflora Bennet from Guyana. Now Piptocarpha tetrantha is said to be closely allied to P. triflora, and hence I conclude, without having seen specimens of Ur- ban's plant, that it is a Critoniopsis. This genus has always been considered a section of Vernonia. It differs from the genuine Vernoniae of the section Lepidaploa in having, as a rule, few-flowered heads, caducous inner involucral scales and acute basal lobes of the anthers. There exist, certainly, forms intermediate between Critoniopsis and Eremosis (DC.) Gleason, for instance, V. (Eremosis Gleason) Heydeana Coulter, so that the species of Eremosis must be referred to Critoniopsis. Some specimens of Eremosis have persistent inner scales. The Brazilian species of Critoniopsis, for instance, V. discolor Less., have throughout typical Vernonia-ini lores - cence as well as many-flowered heads. Many genuine Lepida- ploae have acute basal lobes of the anthers. It will be seen that there is no character to distinguish Critoniopsis generic - ally from Vernonia. The genuine Piptocarphae have a characteristic inflores- cence, dense axillary clusters of heads, and differ in habit widely from Critoniopsis. However, there are genuine Vernonise having the same inflorescence as Piptocarpha, for instance, V . Vauthierana DC. (referred erroneously by Baker in Fl. Bras, to Piptocarpha/). It seems to me as if Piptocarpha should be made a section of Vernonia, having been primarily con- sidered so by Lessing and De Candolle. Sect. Lepidaploa (Cass.) DC. Conspectus subsectioiium antillanaruni. I. Inflorescentia cymoso-scorpioidea, calathidiis bracteis fo- liaceis suffultis. A. Achsenia glaberrima. 1. Inflorescentia cymis elongatis. Suffrutices inferne simplices, foliis magnis, membranaceis vel subcori- aceis. — Species in Antillis majoribus (Cuba et Hispaniola), nee non in Brasilia vigentes. S u b s e c t . Sagrceance . 10 ARKIV FOR BOTANIK. BAND 13. N:0 15. 2. Inflorescentia cymis brevissimis, habitu umbellata. Frutices valde ramosi, foliis parvis, rigide coriaceis. — Species in Hispaniola endemicae. Sub sect. Buxifolim. B. Achaenia appresse sericeo-pubescentia. 1. Corollae limbus laciniis dorso pilosis. Suf frutices. Inflorescentia cymis saepissime elongatis, caulifor- mibus. Calathidia parva. — Species pro max. parte austro-americanae, una in Bequia. altera in Trinidad. Subsect. Graciles. 2. Corollae limbus laciniis dorso glabris, ipsissimo apice interdum pilo uno alterove instructis. Frutices, suf frutices vel herbae. Inflorescentia saepius di- stincta. Calathidia vulgo mediocria. ■ — Species praesertim antillanae. Subsect. Arbor escentes. II. Inflorescentia ebracteata. A. Inflorescentia cymoso-scorpioidea, i. e. calathidia om- nia sessilia. 1. Pappi setae exteriores distinctae. Suffrutices saepe scandentes. Calathidia in cymis contigua. Pappi setae albae. — Species austro-americanae, una in regione australi-antillana. Subsect. Scorpioides. 2. Pappi setae exteriores filiformes, quam maxime in- distinctae. Suffrutex elatus. Calathidia in cymis remota. Pappi setae brunneae. — Species unica, in St. Vincent obvia. Subsect. Pallescentes. B. Inflorescentia cymoso-paniculata, i. e. calathidia sal- tern inferiora pedunculata. 1. Inflorescentia convexa, ambitu rotundata. Suffru- tices elati. Calathidia parva — mediocria. Pappus sae- pissime albidus vel stramineus, rarissime purpu- rascens. — Species a Mexico meridiem versus distri- butae, nonnullae in Cuba. Subsect. PoJyanthes. 2. Inflorescentia plana vel concava, ambitu late obpy- ramidata. Suffrutices elati. Calathidia majuscula. Pappus saepe purpurascens. Corollae et antherae quam in praecedenti majores. — Species a Mexico septentrionem versus distributae, una in insulis Ba- ha mensibus . S u bs ec t. Noveboracenses. E. L. EKMAN, WEST INDIAN VERNONIiE. 11 Subsect. Sagraeanse. Sujfrutices a pice ramosi. Folia membranacea, breviter petiolata vel subsessilia, elliptica vel late lanceolata, glabra — tomentosa, satis magna. Inflorescentia cymoso-scorpoidea, cy- mis elongatis; bracteis evolutis, foliaceis. Calathidia majuscula. Involucrum campanulatum vel cylindriciim, squamis multise- riatis, imbricatis; appressis vel interdum apice recurvatis, ob- tusis vel breviter cuspidatis. Beceptaculum planum, nudum. Pappi setae interiores persistentes, finnidce, numero 40 — 70, exterioribus distinctis 6 — 10-plo longiores. Corolla 8 — 10 mm longa, glabra, tubo long. 2/3 corollce. Antherce magnce, 3 mm et ultra longce, ligula long. lj6 — -1 10 antheras. Achcenium glaber- rimum. A very natural subsection, easily recognized by the large heads, flowers and anthers, and, above all, by its glabrous achenes. It shows no relationship with other West Indian subsections, but is allied to V. glabrata Less, and other Bra- zilian species. It is a remarkable fact that the allies of V. glabrata do not occur in North Brazil nor in Guyana. Some species belonging here have been found, however, in the Andes of Bolivia, viz. V. robusta Rtjsby, V. subacuminata Hieron. (= V. obtusata Less.) and V. velascensis Hieron. Gleason, having, of course, recognized the subsection, in his »Revision» named it Fruticosce, supposing Vernonia fruticosa (L.) Sw. to belong here. Recently he has found that the latter species is to be removed from this subsection, and he now calls the subsection Sagrceance, which name I have adopted here, as I had originally used it myself just in the same sense. Nearly all species of the subsection are imperfectly known, the material collected being very meagre. Continued collections are badly needed to supply dates for a defini- tive delimination of the species, as well as for their frequence and distribution. The distinctions of the species which belong here will appear from the following key. 12 ARKIV FOR BOTANIK. BAND 13. NIO 15. Conspectus specierum lmjiis subsectionis. I. Calathidia omnia plane sessilia. - — • Species cubenses. A. Folia subtus glabra, puberula aut laxe villosa, haud tomentosa. 1. Squamae involucri omnes erectae, appresso-imbri- catae. a. Folia subtus glabra aut puberula. a. Involucrum campanulatum. Calathidia 20 — 25-flora. Folia glabra. f Folia elliptica, breviter acutata, membra- nacea, plana, tenuiter reticulato-venosa. Pappus sordide brunnescens. V. Sagrceana DC. ff Folia late lanceolata, acuta, coriacea, sub- bullata, vix reticulata. Pappus stramineus. V. Valenzuelana A. Rich. (3. Involucrum cylindricum. Calathidia 8—11- flora. f Folia subtus glabra, valde rugosa. Invo- lucrum 6 mm altum. Squamae acutissima?, interiores apice colore purpureo tinctae. V. purptirata Gleason. ff Folia subtus puberula, plana. Involucrum 8 — 9 mm altum. Squamse obtusse, brevis- sime apiculata?, interiores apice pallide brun- neae. V. angusticeps Ekman. b. Folia subtus pilis albidis, longis laxe sericeo- -villosa. Squama? dense villosa?, conspicue spi- noso-mucronat83. V. viminalis Gleason. 2. Squamae involucri saltern exteriores squarrosae, pa- ten tes — recurvatae. a. Calathidia 34 — 47-flora. Involucrum 8- — 9 mm altum, sice, ad 15 mm latum. Squamae lan- ceolato-lineares, subulatae. Pappus albus. V. aroni folia Gleason. b. Calathidia circ. 17-flora. Involucrum 7 mm altum, 8 — 9 mm latum. Squamae deltoideae, breviter mucronatae. Pappus sordide brunne- scens. V. Wrightii Sch.-Bip. E. L. EKMAN, WEST INDIAN VERNONIiE. 13 B. Folia subtus breviter, sed dense tomentosa, subdis- colora. 1. Squamae interiores obtusae vel subacutae, intermediae haud marginatae. V. fallax Gleason. 2. Squamae interiores apiculatae, intermediae conspicue marginatae. a. Folia basi attenuata. Calathidia 15-flora. V. incequiserrata Sch.-Bip. b. Folia ipsa basi rotundata. a. Folia anguste oblonga, apice rotundata. Cala- thidia 22-flora. V. lingucefolia Ekman. [3. Folia angustissime lineari-oblonga, apice brevi- ter acuta. Calathidia 11 — 13-flora. • V. aceratoides Gleason. II. Calathidia breviter pedunculata. Folia subtus argenteo- tomentosa. — Hispaniola. V. Sprengeliana Sch.-Bip. Vernonia Sagneana DC. Tabula nostra I, fig. 5 (folium). Vernonia Sagrceana De Candolle, 1836, p. 55 ; Gleason, 1906 b, p. 183; 1913, p. 321. Vernonia rigida Sw. var. Sagrceana Grisebach, 1866, p. 144. Vernonia foliosa Wright in sched. herb. Grisebach — non Gardn., nee. Sch.-Bip. Pappi setae interiores persistentes, numero fere 40, te- nues, filiformes, sordide brunnescentes. quam exteriores spa- thulato-lineares, distinctae circ. 8-plo longiores. Flores in speciminibus mihi commissis desunt. Hab. in Cuba: prope Havana, leg. Ramon de la Sagra, n. 73 (P, Prodr, orig. spec), n. 230 (P), n. 663 (Prodr), sine num. (CD); Wright n. 281 (G, KU, S), n. 284 (Bss, DC, P). The specimens collected by Wright agree in every par- ticular with those collected by Ramon de la Sagra. Vernonia Valenzuelana A. Rich. ap. Sagra. Tabula nostra I, fig. 2 (folium). Vernonia Valenzuelana A. Rich. ap. Sagra, 1850, p. 33; Gleason, 1906 b, p. 184; 1913, p. 321, 322. Vernonia rigida Sw. var. Valenzuelana Grisebach, 1866, p. 144. 14 ARKIV FOR BOTANIK. BAND 13. N:0 15. Pappi setae interiores persistentes, numero 70, tenues, filiformes, straminea?, exterioribus subfiliformibus circ. 8- -plo longiores. Corolla 9 mm longa, tubo long. 3/4 corolla?, glabro. Anthera? 3 mm longa?, ligula brevi, fere 1j9 long, an- thera?, auriculis brevibus. in apicem obtusiusculum subito contractis. Hab. in Cuba: Ramon de la Sagra, sine num. (KU, P, orig. spec.); Wright n. 284 (KU, S), n. 2785 (G, KU). Vernonia purnurata Gleason. Vernonia purpurata Gleason, 1913, p. 322. Hab. in Cuba orientali: Sierra Maestro, in monte Ji- quarito, alt. 1020 m, 18. 9. 06, Taylor n. 544. I have seen no specimens of this species. From the de- scription it comes near to V. angusticeps Ekman; the differ- ences will be discussed under that species. Vernonia angusticeps Ekman n. sp. Tabula nostra I, fig. 3 (folium et cyma). Herba suffruticosa, perennis. Caulis (pars superior tan- tum adest) erectus, satis validus, verisimiliter metralis, 4 — 5 mm diam., inferne simplex, superne ramosus, teres, stri- atus, superne pra?sertim pilis intricatis, cinereo-fuscis floccoso- -tomentosus, densiuscule foliosus. Folia patentia-deflexa, bre- viter petiolata, petiolo 3 — 4 mm longo, supra canaliculato, floccoso-tomentoso; lamina? lanceolato-oblonga?, 8 — 9 cm lon- ga?, 2,4 — 2,6 cm lata?, apice acuto-acuminata?, basi oblique truncata?, firmula?, fusco-virides, integra?, margine subpla- na?, remote denticulata?, supra glabriuscula?, subnitida?, sub- laeves, reticulata?, subtus pallidiores, piloso-puberula?, te- nuiter, tamen distincte areolata?, costa valida. subtus pulchre prominente, nervis utriusque lateris numero 12 — 14, arcuatis. exsculptis, nervulis distinctis, anastomosantibus. Inflo- icscentia cjnioso-scorpioidca. satis ampla. sublucida. cymis erecto-patulis, iter iterumque ramificatis, strictiusculis. maxi- mis 12 cm longis, 10 — 15-cephalis; bracteis foliaceis, calathi- diis multoties longioribus. Calathidia axillaria. sessilia, in- E. L. EKMAN, WEST INDIAN VERNONL&J. 15 fima remotiuscula, summa congesta, 9 — 11 -flora. Invo- lucrum cylindricum, 8 — 9 mm longum, 3 mm latum, basi rotundatum, squamis imbricatis circ. 6-seriatis, extimis ova- to-triangularibus, mucronulatis, marginibus apiceque floccoso- -tomentosis, intermediis oblongis, obtusis, intimis elongatis, ligulatis, velut intermedia? subglabris, obtusiusculis, omni- bus obsoletissime carinatis, minutissime ciliolatis, fuscis. Pappi seta? interiores persistentes, numero 40, firmse, filiformes, 6,5 mm longse, albo-stramineae, exterioribus distinctis, spathu- lato-linearibus circ. 7-plo longiores. Flores desunt, specimi- nibus visis nempe fructus tempore collectis. Acha?nium obscure tetragonum, subcompressum, inferne obsolete 10- costatum, glaberrimum, glandulosum, 3 mm longum. , Hab. in Cuba orientali: C. Weight, ami. 1856 — 7, sub num. 284 (Bss, D, DC, spec. orig.). The most striking feature of the new species is its very narrow, cylindrical involucres and, connected with this, the small number of flowers in the heads. I did not know any other species of the subsection having these characters, until Gleason in his »Studies» described a V. purpurata, having also cylindrical involucres and very few-flowered heads. Yet the two species are not identical, as proved by the follow- ing differences. The leaves of V. angusticeps are membrana- ceous, not »heavy, rigid, coriaceous », flat, not »strongly rugose above », not »shining», puberulent at the whole of the lower surface, not only »along the veins beneath »; the brac- teal leaves are present below all heads; the heads are 9 — 11- -flowered, not »8-flowered»; the involucres are 8 — 9 mm high, not »6 mm high», their scales are rounded at the tips, short- ly apiculated, not »sharply acute », brownish, not »purple- -brown at their exposed tips». I think the reason why the species has not been described long before is that it is so poorly represented in herbaria. I only saw it in those of Geneva. Yernonia viminalis Gleason. Tabula nostra I, fig. 1 (folium et cyma). Vernonia viminalis Gleason, 1906 b, p. 184; 1913, p. 321. Pappi setae interiores persistentes, numero 40, tenues, filiformes, exterioribus distinctis circ. 10-plo longiores. Co- 16 ARKIV FOR BOTANIK. BAND 13. N:0 15. rolla 9 mm longa, glabra. Anthera3 3,3 mm longae, ligula long. 1/8 antherae, auricalis ipsissimo apice obtusis. Hab. in Cuba: Wright n. 285 (B, CD, KU, S). The species is at once recognized by its peculiar pubes- cence. Vernonia aronifolia Gleason. Vemonia aronifolia Gleason, 1913, p. 323. Hab. in Cuba: prov. Pinar del Rio, prope Sumidero, in rupibus calcareis, 2 — 4. 8. 12, Shafer n. 13514. Not seen. In leaf-characters this species must be iden- tical with V. Wrightii Sch.-Bip., and in reading the de- scription I at first believed there was no difference at all. How- ever, it seems to be distinguished from V. Wrightii by its large, many-flowered heads, narrow, subulate scales and white pappus. Vernonia Wrightii Sch.-Bip. Tabula nostra I, fig. 9 (folium et cyma). Vernonia rigida Grisebach, 1862, p. 511 (»var. V. Sagrceana DC.») — non Sw- Vernonia Wrightii Schultz-Bipontinus, 1863, p. 234. Vernonia leptoclada Gleason 1906 b, p. 183 — non Sch.-Bip. Pappi seta? interiores persistentes, numero 55, tenues, filiformes, sordide brunnescentes, exterioribus distinctis 6- plo longiores. Corolla 8,5 mm longa, glabra. Antherse 3 mm longse, ligula long. 1;7 antherae, auriculis satis longis, attenuatis, ipso apice obtusiusculis. Hab. in Cuba orientali: prope villain Monte Verde dictam, Jan. — Jul. 1859, Wright n. 284 (CD, orig. spec, sub num. falso 1309, G, K, KU, P); in monte La Clarita, alt. 800 m. mens. Apr. 1889. Eggers n. 5017 (KU); Oriente, Loma Santa Teresa, prope El Yunque, in collibus siccis, 2. 12. 10, Shafer n. 7738 (KU); Oriente, Camp La Gloria, prope Sierra Moa, 24—30. 12. 10, Shafer n. 8145 (KU). In the collection of Composite; made by Wright and sent by Asa Gray to Schultz-Bipontinus for determin- ation, the labels of the numbers 284 and 1309 had been interchanged. Hence n. 1309 in the Cosson-Durand herba- rium at Paris is the same as n. 284 in other herbaria, and E. L. EKMAN, WEST INDIAN VERNONIiE. 17 this is the true V. Wrightii Sch.-Bip., as shown by the descrip- tion and the handwriting of Schultz-Bipontinus. Gri- sebach, iii his Cat. Plant. Cub., p. 144, applies the name of V. Wrightii to the plant generally numbered 1309, which is V. leptoclada Sch.-Bip. Not accepting this plant as a species, he cites it as a form of V. arbor escens Sw., though maintaining the binomial writing (»F. Wrightii Sch., forma foliis angustioribus subtus tomentosis, capitulis folio florali multo brevioribus»). He does not mention anything about V. leptoclada Sch.-Bip. Gleason, who has read the descrip- tions of Schultz-Bipontinus, but not discovered the mys- tery of the interchanged labels, in the »Revision » assigns cor- rectly V. Wrightii Sch.-Bip. to his Fruticosce (i. e. his Sagrce- ance), though as a synonym of V. Sagrwana DC. He applies, further, the name of V. leptoclada to the plant which in fact corre- sponds to V. Wrightii. The true V. leptoclada Sch.-Bip. he refers to V. gnaphalifolia A. Rich., and consequently describes the genuine V. ghaphalifolia, which thus had no name at its dis- posal, as a new species, V. sublanata Gleason. His remarks on V. gnaphalifolia A. Rich, read as follows: »Vernonia Wrightii Griseb. 1. c. ; Not V. Wrightii Sch.-Bip. », and further: »Wright's collection, 1309, included two species, one a form with broad smooth leaves, V. Sagrceana DC, and the other with tomentose leaves. The first was taken by Schultz as the type of his V. Wrightii. Grisebach, apparently consid- ering that the number included but one species, published the name for the second form, thus introducing the syno- nyms cited above ». Evidently, this is an attempt of Glea- son to explain the lack of harmony found between the descriptions of Schultz-Bipontinus and the plants cor- responding to them according to their numbers. In work- ing out his »Studies» he had learnt that his V. sublanata is identical with V. gnaphalifolia A. Rich. Thus his V. gna- phalifolia had no name, and consequently he names it V. neglecta Gleason, creating in this manner one synonym more for the true V. leptoclada Sch.-Bip. Still believing that this species belongs to the Sagrceanm, he identifies with it Shafer n. 8145, a genuine V. Wrigthii Sch.-Bip. Another plant, Shafer n. 7738, he refers to V. Wrightii, though with some reservation. However, it differs from the genuine V. Wrightii Arkiv for botanik. Band 13. N:o 15. 2 18 ARKIV FOR BOTANIK. BAND 13. NIO 15. only in having more elongated and spreading scales with recurved tips. A brief description of V. Wrightii may be of some use: Branches densely tomentose, hairs brown, spreading. Leaves obovate-oblong, 8 — 9 cm long by 4 cm wide, firm, not coriaceous, glabrous and smooth above, finely pubescent at the veins beneath, remotely denticulate with spinulose teeth at the margin. Heads about 17-flowered. Involucres 7 mm high, about 8 — 9 mm broad in press-dried specimens, its scales deltoid, spreading with short, subulate tips, brown- ish-green. Pappus 6 mm long, sordid-tawny, not at all purplish. Veriionia fallax Gleason. Vernonia fallax Gleason, 1913, p. 324. Hab. in Cuba: prov. Santa Clara, in montibus Trinidad, ad rupes, alt. 500 m, 12. 3. 10, Britton et Wilson n. 5478. I have not seen this species, but from the description it must be a rather good one, though closely related to V. incequiserrata Sch.-Bip. Yernonia inaequiserrata Sch.-Bip. Tabula nostra I, fig. 4 (folium). Vernonia rubricaulis Grisebach, 1862, p. 511 (»forma foliis latioribus») — non H. B. K. Vernonia incequiserrata Schultz-Bipontinus, 1863, p. 232 ; Grisebach, 1866, p. 144, excl. var. ; Gleason, 1906 b, p. 183; 1913, p. 321. Pappi setae interiores subpersistentes, numero 55, te- nues, filiformes, albidse, quam exteriores distinctae, spathu- latse circ. 12-plo longiores. Corolla 10,5 mm longa, tubo sensim dilatato long. 2/3 corolla?, glabro, limbi laciniis apice subtus papillosis. Antherse 3,5 mm longae, ligula long. lj7 antherse, auriculis breviter apiculato-contractis. Hab. in Cuba orientali. Sept. 1859 — Jan. 1860, Wright n. 285 prior (Bss, CD, orig. spec, D, DC, G, K, KU, P). I long endeavoured to keep within the limits of V. inacequi- serrata its two varieties, distinguished already by Grisebach, viz. var. angustifolia and var. obtusifolia. However, Gleason E. L. EKMAN, WEST INDIAN VERNONIiE. 19 describes in his »Studies» the first one as a species, which, indeed, it may be. Consequently I have to describe the second as a species, too, for it differs as amply from V. incequiserrata as does the first. It seems to me more advisable to describe a new species, even though not a very striking one, than to reduce another without very strong reasons. A few words about the characteristics of the now more narrowly delimitated V. inceqtiiserrata. Leaves broadly lan- ceolate, 9 cm long, 2,5 cm wide, narrowed at base, acute, sharp- ly serrate, fulvo-cincereous beneath, not conspicuously reti- culated. Heads 15-flowered. Involucres 7 — 8 mm long, 4,5 mm broad, inner scales somewhat elongated. An interesting form (?) of this is in the Kew Herbarium, having lanceolate, entire leaves, 7 cm long, 1,45 cm wide. Yernoiiia linguaefolia Ekman n. sp. Tabula nostra I, fig. 6 (folium). Suffrutex vel herba perennis, certe metralis et ultra. Caulis (cujus pars superior tantum adest) erectus, inferne lignosus, diam. 3,5 mm, simplex, superne ramosus, teres, striatus, inferne subglaber, superne brevissime appresso- -tomentosus, sat crebre foliatus. Folia alterna, patentia, breviter petiolata, petiolo fere 2 mm longo, tomentello; la- minae late lineari-oblongae, anguste linguseformes (unde no- men!), 9,5 cm longae, 1,8 cm lata?, apice rotundatse, obtusse, basi rotundatse, membranacese, plana?, margines versus reti- culato-rugosae, integrse, obscure crenulatae, supra olivaceo- -virides, levissime scaberulae, obsolete hispidulo-puberulae, sub- tus discolores, tomento brevissimo, rufo-cinereo, costa ner- visque utriusque lateris numero ad 15 angulo subrecto ex- euntibus, demum sursum arcuatis, margines versus reticu- latim anastomosantibus, subtus pulchre prominentibus. In- florescentia cymoso-scorpioidea, ambitu late obpjTamidata, 1,5 dm longa et totidem lata, diffusa, cymis ad 15 cm longis, inferne simplicibus, ad medium ssepe divaricatim ramosis, leviter recurvis, circ. 15-cephalis, calathidiis inferne spatio circ. 1,5 cm longo disjunctis, superne approximatis; bracteis foliaceis, inferioribus folia aBquantibus, superioribus gra- datim minoribus, calathidiis semper conspicue longioribus. 20 ARKIV FOR BOTANIK. BAND 13. N:0 15. Calathidia axillaria, sessilia, solitaria, 19 — 22-flora. Invo- lucrum late campanulatum, 6 — 7 mm altum, 5 — 6 mm latum, squamis circ. 5-seriatis, extimis brevissimis ovato-deltoideis, breviter acutis, brevissime mucronulatis, intermediis anguste oblongis, obtusis, intimis lineari-ligulatis, obtusiusculis, om- nibus erectis, appressis, imbricatis, inferne glabris, strami- neis, superne pallide brunneis, tomentellis. Receptaculum planum, nudum. Pappi setse interiores subcaducse, numero 50, tenuissimse, albidse, 6 mm longse, quam exteriores dis- tinctse circ. 10-plo longiores. Corolla 10,5 mm longa, glabra, tubo sensim dilatato long. 2/3 corollse. Antherse 3,3 mm lon- gse, ligula l/7 long. antherse, auriculis brevibus, subito contrac- ts. Achsenium (immaturum) 10-costatum, glaberrimum. Vemonia incequiserrata Sch.-Bip. var. obtuaifolia Ghisebach, 1866, p. 144. Hab. in Cuba orientali: Wright n. 285 posterior (G, spec, orig., KU, S). The species differs from V. incequiserrata in its obtuse, crenulate leaves with rounded base and in the many-flowered heads. As to its floral characters, it comes rather near to V. inazquiserrata. From V. aceratoides Gleason it is dis- tinguished by its broader, always obtuse leaves and by its many-flowered heads. A comparison of the structure of the flowers will also afford good distinctions. Vernonia aceratoides Gleason. Tabula nostra I, fig. 7 (folium). P- Vemonia incequiserrata Sch.-Bip. var. angustifolia Grisebach, 1866, 144. Vernonia aceratoides Gleason, 1913, p. 325. Pappi setae interiores subcaducse, numero 50, tenuissi- mae, albidae, quam exteriories spathulatse circ. 10-plo longi- ores. Corolla 7 mm longa, glabra, tubo -/3 l°ng- corollse. Antheree 2,8 mm longae, ligula V7 — Vs long, antheraa, auri- culis breviter apiculato-contractis. Hab. in Cuba: Wright n. 2784 (Bss, DC, G, K, KU, P). A good character of this species not mentioned by Glea- son is the small number of flowers in the heads, 11 — 13. E. L. EKMAN, WEST INDIAN VERNONL3E. 21 Very near to this species comes a plant collected by Eggers in Cuba, Loma del Jaguey, 22. 4. 89, n. 4982 (KU). It differs in having lanceolate leaves, which are very bullate and nearly glabrous beneath. The only specimen seen is very imperfect. Vernonia Sprengeliana Sch.-Bip. Tabula nostra I, fig. 8 (cyma). Eupatorium salvifolium Bertero in sched. ap. Sprengel, 1826, p. 412. Vernonia Sprengeliana Schultz-Bipontinus, 1863, p. 232; Gleason, 1906 b, p. 184; 1913, p. 321. Pappi setae interiores persistentes, numero 45—50, tenues, albidse, exterioribus distinctis 8-plo longiores. Corolla 9 — 9,5 mm longa, glabra, tubo 2/3 corollae long., limbi laciniis apice subtus papulosis. Antherse 3 — 3,5 mm longse, ligula long. Vs antherse, auriculis brevibus, brevissime acuminatis. Hab. in Hispaniola: Bertero (CD, orig. spec.); Schom- burgk (K); Mayerhoff (B, CD, KU); Picarda n. 301, 301 b (KU); prope Petionville, anno 1889, Picarda n. 885 (KU); in montibus Sierra del Palo quemado, locis calcareis, 10. 5. 87, Eggers n. 1887 (Br, D, DC, K, KU); in collibus siccis prope Gona'ives, mense Dec. 1898, Buch n. 10 (KU); ibidem, mense Nov. 1900, Buch n. 523 (KU); prov. Barahona, Las Salinas, Los Charcos, alt. 1200 m, Sept. 1911, Miguel Fuertes n. 1388 (KU). A very distinct and always readily recognized species! Though there is considerable variation as to the size of the heads and the density of the inflorescence, the material of this species must be considered as unusually uniform. Schultz-Bipontinus quotes two synonyms of the spe- cies, Eupatorium salicinum Lam. and Eupatorium salvi- folium Bertero. The latter belongs to the species and is a mere name without any description. Eupatorium salici- num Lam. is a true Eupatorium, having opposite leaves, though in shape somewhat resembling those of V. Sprengeliana, and was collected in Peru. Now Kurt Sprengel, unfortunately, in his Systema referred Eupatorium salvifolium Bertero to Eupatorium salicinum Lam., whence the quotation of Schultz- -Bipontinus. Of course, Eupatorium salvifolium Bertero 22 ARK1V FOR BOTANIK. BAND 13. N!0 15. as a nomen nudum cannot invalidate Vernonia Sprengeliana Sch.-Bip. Subsect. Buxifoliae. Frutices ramosissimi. Folia coriacea, parva. Inflo- rescentia cymosa, cytnis brevissimis, uni- vel paucifloris, cum calathidiis primariis pedunculatis axe valde abbreviato insi- dentibus, item umbellas subsimplices, depauperatas formanti- bus; bracteis foliaceis cymis sujfulcientibus. Calathidia parva, circ. 10-flora. Involucrum infundibuliforme vel anguste in- fundibuliforme-campanulatum, squamis imbricatis, subglabris. Receptaculum nudum, subplanum. Pappi setce interiores per- sistentes, rigidulce, crassiusculai, paullulo complanatce, ex- teriores breves, parum distinctce. Corolla glabra. Antherce ultra 2,5 mm longai, ligula longa, l/3 — 1/4 long, antherce. Achce- nium glaberrimum, glandulosum. A very small subsection, including only two species from Hispaniola, easily recognized by their shrubby rigid habit, the peculiar structure of the inflorescence, and by their gla- brous achenes. Gleason made this subsection a particular division of Lepidaploa: Scorpioidece aggregates; yet he refers to it in his »Revision» V. Thomce Benth., which he correctly discards in his »Studies», and V. yunquensis Gleason, which does not belong here at all. The two species of the subsection have no allies in the West Indies, nor in Central America, nor in the Andes. Possibly some Brazilian species, e. g. V. nitidula Less., may approach to the Buxifolice. Future investigations may prove this. Conspectus specierum hujus subsectionis. I. Folia subtus glabra. Calathidia 8-flora. Pappi setse stramineo-rufescentes. V. buxifolia (Cass.) Less. II. Folia discolora, subtus appresse tomentella. Calathidia 9 — 11-flora. Pappi setse stramineo-purpurascentes. V. Tuerckheimii Urb. Vernonia buxifolia (Cass.) Less. Tabula nostra I, fig. 11 (habitus). Lepidaploa buxifolia Cassini, 1823, p. 18. Proustia domingensis Sprengel, 182G, p. 502. E. L. EKMAN, WEST INDIAN VERNONLE. 23 Vernonia buxifolia Lessing, 1829, p. 313. Vemonia domingensis De Candolle, 1836, p. 30. Cacalia buxifolia Kttntze, 1891, p. 968. Vernonia montana Gleason, 1906 b, p. 191 ; 1913, p. 327. Pappi setse interiores subpersistentes, numero 50, crassi- usculse, paullulo complanatee, subrigiclse, pallicle rufescentes, exterioribus linearibus, inconspicuis 8 — 10-longiores. Co- rolla (haud evoluta) 4 mm longa, tubo amplo, glabro, long. V2 corollas, limbi laciniis ipso apice pilo uno alterove instruc- tis, subtus papulosis. Antherae 2,5 mm longse, ligula long. 1/i antherse, subenervi, auriculis obtusis. Hab. in Hispaniola: Nectoux (P); Poiteau (P); Bertero (M, Prodr, orig. spec. Proustice domingensis Spr.); prope Port au Prince in Morne Tranchant, alt. 1900 m, April. 1892, Picarda n. 1023 (KU); inter Constanza et Rio Jime- noa, alt. 1900 m, mense Junio 1910, Turckheim n. 2959 B (KU); »in cuspide montium La Hajo», mense Oct. 1910, Mi- guel Fuertes n. 597 (KU, P); inter La Brande et montem Balance, ad arbores scandens, alt. 1000 m, 15. 8. 05, Nash et Taylor n. 1756 (KU, dupl. spec. orig. V. montance Glea- son). An interesting species apparently not rare in the higher mountains of Hispaniola. The specimens cited above agree exactly with each other and with the type of Proustia do- mingensis Spr. in the Prodromus Herbarium. The type specimen of Lepidaploa buxifolia Cass, has not been seen. The good description of Cassini agrees perfectly with the species as here understood. Gleason in his »Revision » and in his »Studies » evidently has another plant in mind when describing V. buxifolia. The characteristics »Achenes pubescent; outer pappus conspi- cuous, its scales much broader than the white bristles of the inner series » (Gleason, 1913, p. 327) clearly do not apply to our V. buxifolia, of which Cassini says: »Les ovaires sont glabres . . . leur aigrette est roussatre, double: l'exterieur courte, peu distincte, composee de squamellules inegales, filiformes-laminees, subulees, denticulees ». Which species Gleason has in mind, I cannot decide; most probably it is a new one. He has, however, genuine V. buxifolia (Cass.) Less, at hand, but describes it as a new species, V. montana Gleason. 24 ARKIV FOR BOTANIK. BAND 13. N:0 15. Veruonia Tuerckheimii Urb. Tabula nostra I, fig. 10 (pars inflorescentise). Vernonia Tuerckheimii Ukban, 1912, p. 421 ; Gleason, 1913, p. 327. Pappi setae interiores persistentes, numero 65, rigidulse, crassiusculae, paullulo complanatae, stramineo-purpurascen- tes, quam exteriores inaequales, parum conspicuse circ. 8-plo longiores. Corolla 7 mm longa, tubo amplo, glabro long. z/3 corollae, limbi laciniis apice subtus papillosis, ceterum glabris. Antherse 2,5 — 2,8 mm longse, ligula longa, 1/3 long, antherse, auriculis brevissime apiculatis. Hab. in Hispaniola: Constanza, in declivibus pineti aridis, alt. 1250 m, mense Febr. 1910, Turckheim n. 2959 (D, KIT, M). A very characteristic species, easily recognized by the purple-tinged involucres and pappus. The description given by Urban is excellent. Subsect. Graciles. Suffrutices vel frutices humiles. Folia membranacea, forma et pubescentia varia. Inflorescentia cymoso-scorpioidea, cymis elongatis, scepissime cauliformibus, calathidiis axillaribus, ses- silibus, sospe geminatis; bracteis foliaceis, calathidiis longioribus. Calathidia parva. Involucrum campanulatum, squamis scepe longe vel longissime mucronatis. JReceptaculum planum, subnudum. Pappi setm interiores persistentes, tenues, albidce, exteriores distincton. Corolla vulgo 6 mm longa, limbi laciniis dorso laxe pilosis. Antheroz circ. 2 mm longce. Achcenia tenuiter appresso- -pubescentia. A very natural subsection closely related to the Arbo- rescentes, and chiefly distinguished by the pubescence of the clefts of corolla, further by the characteristics of the inflor- escence and the involucres. The range of the subsection extends from Columbia to eastern Brazil. Well-known spe- cies belonging here are V. gracilis H. B. K., V. Cotoneaster (W.) Less., V. Salzmannii DC. (not Baker in Fl. Bras.), V. araripensis Gardn., V. acutangula Gardn., V. tricephala E. L. EKMAN, WEST INDIAN VERNONLiE. 25 Gardn., etc. Two species have been found in the West Indies, viz. V. gracilis H. B. K. and V. tricephala Gardn., both in the extreme south-eastern part of the region. Conspectus specierum hujus subsectionis. I. Calathidia in axillis solitaria vel geanina. Squamae in- volucri in mucronem plus minus ve elongatum, saepe recur - vatum subito contracta3. ■ — Bequia. V. gracilis H. B. K. II. Calathidia in axillis vulgo plura, 2 — 4. Squamae invo- lucri acutissimae, in mucronem rectum sensim attenuatse. — Trinidad. V. tricephala Gardn. Yernonia gracilis H. B. K. Vernonia gracilis Humboldt, Bonpland, Kunth, 1820, p. 34; Lessing, 1829, p. 302; 1831, p. 667 (excl. var. villosa), De Candolle, 1836, p. 49; Schtxltz-Bipontinus, 1847, p. 511; Bakeb, 1873, p. 81 (quoad typum !). Vernonia elongata Willdenow in sched. ap- Lessing, 1829, p. 302. Vernonia Moritziana Schultz-Bipontinus, 1847, p. 511. Cacalia gracilis Kuntze, 1891, p. 970. Cacalia Moritziana Kuntze, 1891, p. 970. Hab. in Columbia. subsp. tomentosa Ekman n. subsp. A specie differt praesertim foliis discoloribus, subtus tomento floccoso instructis. Vernonia Cotoneaster (W.) Less. var. angustifolia Hieronymus, 1901, p. 559. Vernonia Cotoneaster (W.) Less. var. pungens Hieeonymtjs in sched. herb. Berol. Pappi setae interiores persistentes, numero circ. 25, te- nues, filiformes, albidae, quam exteriores 7-plo longiores. Corolla 6 mm longa, glabra, tubo superne dilatato 2/3 long, corollae, limbi laciniis dorso pilis longis instructis. Anthe- rae 2,7 mm longae, ligula 1/s long, antherae, subenervi, auri- culis subito contractis. Hab. in Bequia: in collibus inter frutices, alt. 150 — 300 m, Maj. 1889, Joseph Dalton et H. H. Smith n. B. 288 (K, orig. spec, subspeciei). — Insuper in Columbia. 26 ARKIV FOR BOTANIK. BAND 13. N:0 15. Vernonia gracilis H. B. K. is a plant common in the dry plains of Columbia. Like all Vernonise it is highly vari- able, especially as to the pubescence of the leaves and the length of the subulate tips of the scales. V. Moritziana Sch.-Bip. was described upon a form with very short tips. The genuine V. gracilis has leaves nearly glabrous beneath. I have based the new subspecies upon specimens having the leaves densely floccose-tomentose beneath. Just because of this pubes- cence ^Hieronymus took the same plant to be V. Cotoneaster (W.) Less., a species from Bahia, differing from V. gracilis in its smaller involucres. The material at hand of the sub- species is too meagre to give a reliable idea of its systematic rank; perhaps it is a distinct species. It is puzzling that V. gracilis subsp. tomentosa has been found to occur in Bequia, far away from its proper region, which is Columbia. Vernonia tricephala Gardn. Tabula nostra VI, fig. 5 (cyma). Vernonia gracilis H. B. K. var. villosa Lessing, 1829, p. 303; De Candolle, 1836, p. 50. Vernonia tricephala Gardner, 1846, p. 223; Baker, 1873, p. 68. Vernonia tricholepis Grisebach, 1861, p. 354, p. p. — non DC. Cacalia tricephala Kuntze, 1891, p. 971. Pappi setae interiorcs persistentes, numero 30, tenues, filiformes, albidse, quam exteriores distinctse circ. 4-plo lon- giores. Corolla 5 mm longa, tubo sensim dilatato long. 2/3 corollse, glandulis nonnullis obsito, limbi laciniis dorso pilis longis et glandulis globuliformibus instructis. Antherse 2 mm longae, ligula i/5 long, antherse, auriculis acutis. Hab. in Trinidad: Crueger, sine loco ace. (G); Cedros, loco sicco, 15. 1. 08, Broadway n. 2174 (DC, KU). Specimens from Trinidad agree exactly with the genuine V. tricephala from Brazil. Grisebach cites, 1861, p. 354, two West Indian plants as belonging to V. tricholepis DC. Neither of them belongs to that species, the former, Jamaica, Purdie, being V. se- ricea L. C. Rich., certainty not collected in Jamaica, the second is just V. tricephala Gardn. E. L. EKMAN, WEST INDIAN VERNONIiE. 27 Subsect. Arborescentes. Frutices vel suffrutices. Folia membranacea, rarius sub- coriacea, forma et pubescentia varia. Inflorescentia cymoso- -scorpioidea, cymis elongatis vel in speciebus nonnullis ab- breviatis; bracteis evolutis, foliaceis. Calathidia mediocria — par vula. Receptaculum planum, subnudum. Involucrum vulgo late campanulatum, interdum cylindricum aid late infundi- buli forme. Pappi setce interiores persistentes aut subdeciduce, numero 20 — 50, tenues aut crassiusculai, quam exteriores dis- tinctce 5 — 15-plo longiores. Corolla 5 — 10 mm longa, glabra; tubo long. l/2 — 3/4 corollas, extus interdum glandulis brevibus nonnullis instructo; limbi laciniis apice subtus papulosis, raro ipso apice pilis nonnullis munitis. Antherce 0,8 — 2,8, raro 3,2 mm longce, ligula long. 1/6 — Vs antherce, auriculis obtusis aut acutis. Achoenium sericeo-pubescens. The subsection Arborescentes as here conceived includes species of very different habits. It matches Gleason's Scor- pioidece foliatce, his smaller groups Fruticosce and Schiedeance excluded. Some of his Scorpioideos aggregatce also belong here, like all his Scorpioidece reductce. It may be, however, that the subsection, as here defined, is not quite monophyl- etic. One of the species referred to it, namely V. Trinitatis Ekman, differs somewhat in habit from the genuine Arbo- rescentes, and is strongly allied to some Mexican and South American species. I have tried to distinguish it by floral characters from the Arborescentes, but in vain. Probably it is, like some of its allies, an annual plant, differing as to that respect from the Arborescentes; but I have no evidence for this supposition. In fact, as easy as it is to distinguish the Arborescentes from other West Indian subsections, as difficult is it to determine where the line between the subsection and some South Ame- rican species is to be drawn. As far south as the Argen- tine province of Corrientes, species occur strongly recalling the Arborescentes in habit. I am firmly convinced that numbers of species from South America are to be referred in future to this subsection. With this I have stated that the Arborescentes have their nearest allies in South America, and probably the subsection 28 ARKIV FOR B0TANIK. BAND 13. N:0 15. has its origin there, though in the West Indies showing a particular and independent evolution. Within the vast assemblage of species constituting the Arborescentes smaller groups can be distinguished. They do not appear as sharply definable taxonomic units, the exist- ence of which could not be denied; on the contrary, they are but vaguely indicated, and very difficult to characterize. Sub- jective susceptibility may also play some part in the percep- tion of these delicate taxonomic sensations. However, con- sidering every method of arranging species according to their natural relationships preferable to an artificial arrangement, I have tried to give an exposition of the Arborescentes based on their supposed relationships. The species have been arranged according to characters about the value of which for recognizing groups in Vernonia very little is known, such as consistence and relative length of the pappus, pubescence of the corolla, length of the anthers etc. Or, mutatis mu- tandis, the groups recognized have been found to differ in those points. Now such characters are not convenient for identifying specimens the names of which are required. I have, therefore, given an artificial key in order to facilitate such identifications. The smaller groups concerned have not been named. They are based upon too small a material and are, as already stated, too vaguely indicated to deserve that certificate of taxo- nomic importance which a name betokens. However, a few words upon their characteristic features and their geogra- phical ranges may be advisable. 1. The allies of V. argyropappa Buek. One species of Trinidad. As already mentioned it ap- proaches very nearly to certain Mexican and South American species, the best known of which is V. argyropappa Buek. Another member of this group is the widely distributed, annual V. remoti flora L. C. Rich. By their floral characters these spe- cies cannot be distinguished from the allies of V. arborescens (L. ) Sw. The differences are chiefly habitual ones, such as larger involucres, subulate, spreading scales, the outer of which are slightly recurved, etc. The striking resemblance between V. Trinitatis and V. arborescens has made me believe that the latter has taken its origin from V. Trinitatis or from some species nearly allied E. L. EKMAN, WEST INDIAN VERNONIiE. 29 to this. This presumption is clearly supported by the geogra- phical distribution of the species in question. 2. The allies of V. arborescens (L.) Sw. Three species, V. arborescens (L.) Sw., V.albicaulis Pers., and V. borinquensis Urb. They are all shrubs, much branched and rather tall. The inflorescence is many-flowered, spread- ing or somewhat congested, especially so in V. albicaulis. The corolla is glabrous, in rare cases with a few short hairs at the tips of the clefts. The bristles of the inner series of pap- pus are fine, yet persistent, and five to seven times longer than those of the outer one. The anthers are rather large, about 2,5 mm long. The geographical range of this group extends from the island of Margarita near the coast of Venezuela over the Les- ser Antilles up to Porto Rico. V. arborescens is, may be, the primitive type of the group, being confined to the Windward Islands. From it V. albicaulis can be derived directly, struc- turally and geographically, occupying the Leeward Islands and Porto Rico. V. borinquensis I suppose to be a more recent offshoot from V. arborescens. 3. The allies of V. baharnensis Griseb. Three species of the Bahamas, V. baharnensis Griseb., V. arbuscula Less., V. obcordata Gleason, and one species from eastern Cuba, V. convplicata Griseb. They are distin- guished from V. arborescens and its allies by the dense tomen- tum of their leaves and branches and by the highly reduced inflorescence. The three Bahaman species (at least two of them) are, besides, characterized by the existence of a few long hairs at the tips of the clefts of corolla; in V. compli- cata these hairs are lacking. The group can be derived direct- ly from V. albicaulis. The Cuban species evidently derives its origin from the Bahamas, and is to be looked upon as the most recent offshoot of the Arborescens-Bahamensis-br&nch. 4. The allies of V. fruticosa Sw. Five species of Jamaica, V. divaricata Sw., V. acumi- nata Less., V. expansa Gleason, V. pluvialis Gleason and V. rigida Sw., one species of south-western Hispaniola, V. fruticosa (L.) Sw., and four little-known species of eastern Cuba, V. parvuliceps Ekman, V. yunquensis Gleason, V. leptoclada Sch.-Bip. and V. pineticola Gleason. They are distinguished from V. arborescens and its allies by the broader 30 ARKIV FOR BOTANIK. BAND 13. N:0 15. inner bristles of the pappus, which are somewhat cadu- cous, and by the shorter anthers. The Jamaican species differ from the other ones in having the leaves not whitened beneath. In his »Studies» Gleason describes four new species from Cuba, V. calophylla, V. vicina, V. calida and V. semi- talis, which are all to be placed here. As well as V. arborescens (L.) Sw. could be originated from V. Trinitatis Ekman, V. divaricata has its nearest re- lative in V. canescens H. B. K. of Venezuela. In habit the two species are strongly alike. The most characteristic fea- ture of V. canescens, the want of bracts below most of the heads, is to be seen unaltered in V. divaricata. . 5. The allies of V. membranacea Griseb. Three species of Cuba, V. membranacea Griseb., V. cras- sinervia Wright ap. Gleason and V. gnaphah folia A. Rich. They differ from the other Arborescentes in having very short outer bristles of the pappus, only Vio — Vis m length of the inner bristles. A newly described species from Cuba, V. desiliens Gleason, is to be placed here. The allies of V. membranacea closely approach to the allies of V. sericea L. C. Rich. I do not know whether they have any near relative in South America. 6. The allies of V. sericea L. C. Rich. Three species of Cuba, two of which are known to me, V. angustissima Wright, ap. Ekman and V. commutata Ekman, the third, V. corallophila recently described by Gleason, one species of Hispaniola, V. stenophylla Less., and one species of Porto Rico, St. Croix, St. Thomas and St. Jan, V. sericea L. C. Rich., with a subspecies in Hispaniola. The species in question are characterized by the shape of their involucres, their small anthers and their, as a rule, narrow leaves. 7. The allies of V. segregata Gleason. One isolated species of Cuba, of puzzling habit and mysterious affinities. As mentioned above, two keys to the species of the subsection Arborescentes are appended. The first shows the natural relationships of the species. In this key, of course, I have not given the species recently described by Glea- son, having not seen his species, nor having analyzed their E. L. EKMAN, WEST INDIAN VERNONIJE. 31 flowers and pappus. The second key is based on more superficial characters, and in this the species of Gleason have been inserted. Conspectus specierum hujus subsectionis. I. Herbae verisimiliter annuae, saepe elatae. Calathidia ma- juscula, ad 10 mm longa. Squamae involucri acutissimae, subulatae, exteriores saepe patentes — recurvatae. ■ — Species V. argyropappce Buek proximae, habitu plus quam charac- teribus distinctae, in America australi a Mexico usque ad Rio de Janeiro et Paraguay divulgatae, una in Trinidad. V. Trinitatis Ekman. II. Frutices vel suffrutices perennes, sed saepe jam primo anno florentes. Calathidia vulgo minora, ad 8 mm longa. Squamae exteriores rarius subulatae, vulgo acutae — obtusae, saepissime erectae, appressae, strictae. ■ — Species antillanae. A. Frutices vel suffrutices et interne saepe ramosi. Fo- lia lata forma variabili. Involucrum campanulatum aut subcylindricum. Squamae involucri etiam inte- riores imbricatae. Antherae majusculae, saepe 2 — 3 mm longae. 1. Pappi setae interiores exterioribus 6 — 8-plo longiores. a. Pappi setae interiores persistentes, tenues. An- therae majores, circ. 2,5 mm longae. a. Frutices plus minus elati, modice pubescen- tes. Inflorescentia satis ampla, cymis vulgo multifloris. — Species V. arborescenti (L.) Sw. proximae, in Antillis minoribus nee non in Porto Rico vigentes. | Folia membranacea. Cymae rectae vel cur- vatae, haud flexuosae. Pappus albus aut sordide stramineus. X Squamae involucri elongatae, lineares, satis laxae, patulae, exteriores subre- curvatae. Pappus albus. Antherae auriculis apice brevissime acuminatis. — In insulis ad ventum frequens. V. arborescens (L.) Sw. X X Squamae involucri breviores, lineari- -deltoideae, omnes erecto-appressae. Pap- 32 ARKIV FOR BOTANIK. BAND 13. N:0 15. pus saepe sordide stramineus. An- therae auriculis longe acutatis. — Spe- cies quam maxime variabilis, in insulis infra ventum nee non in Porto Rico frequens. V. albicaulis Pers. (cfr. etiam V. albicaulis Pers. X sericea L. C. Rich.) f | Folia pergamacea. Cymae flexuosae, rigidse. Pappus brunneus. ■ — Porto Rico. V. borinquensis Urb. (3. Frutices liumiles, copiose tomentosi. Inflore- scentia reducta, cymis uni — paucifloris. — Spe- cies V. bahamensi Griseb. proximae, in insulis Bahamensibus vigentes, una Cubensis. | Folia margine plana, tomento densissime appresso. Limbi lacinii ipso apice pilis nonnullis longis muniti. — Insulae Ba- hamenses. X Folia basi longe attenuata, supra me- dium latissima. * Folia anguste obovata, apice bre- viter acuminata vel rotundata. V. bahamensis Griseb. ** Folia late obcordata, apice emar- ginato-incisa. V. obcordata Gleason. X X Folia basi rotundata vel breviter atte- nuata, fere ad medium latissima. V. arbuscula Less. ff Folia margine grosse plicato-undulata, to- mento denso, haud appresso. Limbi la- cinii ipsa apice papillosi, pili desunt. ■ — Cuba. V. complicata Griseb. b. Pappi setae interiores subdeciduae, crassiusculae, paullulo complanatae. Antherae minores, vix 2 mm longae. — Species V. fruticosce Sw. pro- ximse, in Antillis majoribus vigentes. a. Folia concoloria. — Species Jamaicenses. f Squamae involucri interiores purpurascentes. Pappus albus. V. divaricata Sw. f f Squamae involucri brunneae. Pappus sae- pissime brunneus. E. L. EKMAN, WEST INDIAN VERNONIiE. 33 X Rami tenuiter puberuli. Cymse inflore- scentiam clistinctam formantes. Involu- crum . campanulatum, aut cylindricum. * Folia membranacea, elliptica, acumi- nata, breviter sed distincte petiolata. Inflorescentia satis ampla. Involu- crum breviter campanulatum, squa- mis 4 — 5-seriatis, satis brevibus, vix carinatis. Corollse tubus long. 3j5 co- rollse. aa. Calathidia in cymis laxe disposita. Involucrum circ. 5 mm longum. Corollse 6 — 6,5 mm longse tubus gracilis superne in limbum subito abiens. V. acuminata Less. PP. Calathidia in cymis satis approxi- mata. Involucrum circ. 4 mm longum. Corollse 5 mm longse tu- bus amplior superne in limbum sensim abiens. V. expansa Gleason. ** Folia pergamacea, acuta, subsessilia, obsolete denticulata. Inflorescentia val- de congesta. Involucrum elongatum, cylindricum, squamis 6 — 7-seriatis, e- longatis, dorso carinatis. Corollse tu- bus long, vix 1/2 corollse. V. pluvialis Gleason. X X Rami juveniles dense tomentosi. Cymse elongatse, flexuosse, inflorescentiam di- stinctam vix formantes. Involucrum in- fundibuliforme, basi cuneatum, squamis multiseriatis, exterioribus valde imbri- catis, interioribus fructus tempore elon- gatis, patentissimis. V. rigida Sw. p. Folia discoloria. — Species insularum Hispan- iola et Cuba incolse. | Folia plana. X Folia subtus laxius floccoso-tomentosa. C}^mse elongatse, ssepe cauliformes. — Hispaniola. V. fruticosa (L.) Sw. Arkiv for botanil: Band 13. N:o 15. 3 34 ARKIV FOR BOTANIK. BAND 13. N:0 15. X X Folia subtus densissime appresso-to- mentosa. Inflorescentia congesta. — Cuba. * Folia lanceolata. Inflorescentia cy- mis multifloris, arcuatis constans, calathidiis perapproximatis,parvis, 5 mm longis. V. parvuliceps Ekman. ** Folia ovata. Inflorescentia valde congesta. Calathidia majora, 10 mm longa. V. yunquensis Gleason. ff Folia valde rugosa. — Cuba. X Involucrum 5 mm longum, squamis breviter acutis, imbricatis. V. leptoclada Sch.-Bip. X X Involucrum ad 12 mm longum, squamis rigidis, patulis, longissime acutatis. V. pineticola Gleason. 2. Pappi setae interiores persistentes, tenues, quam exteriores 10 — 15-plo longiores. — Species V. mem- hranacece Griseb. proximae, in insula Cuba vigentes. a. Folia concoloria, subtus laxe puberula. V. membranacea Griseb. b. Folia subtus densius tomentosa, novella saltern discoloria. a. Folia adulta concoloria, breviter acuta. Suf- frutex humilis, 1,5 dm altus. Inflorescentia depauperata, cymis paucifloris. V. crassinervia Wright ap. Gleason. p. Folia etiam adulta discoloria, obtusa. Suf- frutex elatior, inflorescentia amplior, cymis satis multifloris. V. gnaphalifolia A. Rich. B. Suffrutices interne simplices. Folia angusta, lanceo- lata vel linearia, in V. sericea L. C. Rich, tamen vulgo latiora. Involucrum late infundibuliforme. Squa- mse involucri interiores sublaxse, omnes acutae — acu- tissimae. Pappi setae interiores persistentes, tenues, exterioribus 8 — 10-plo longiores. Antherae minimae — minutae. — Species V. sericece L. C. Rich, proximae, in insulis Cuba — Hispaniola — Porto Rico — St. Jan vi- tentes. 1. Folia concoloria. — Cuba. V. commutata Ekman. E. L. EKMAN, WEST INDIAN VERNONIiE. 35 2. Folia subtus sericeo-pubescentia — sericeo-tomentosa. a. Folia linearia, margine revoluta (cfr. V. sericece subsp.)- Calathidia 18 — 21-flora. Pappus albus. a. Inflorescentia divaricata. Squamae involucri submolles. Tubus long. 3/4 corollae. — Cuba. V. angustissima Wright ap. Ekman. (3. Inflorescentia congesta, angusta. Squamae in- volucri rigidae. Tubus long. */2 corollae. — Hispaniola. V. stenophylla Less. b. Folia latiora, vulgo anguste ovato-lanceolata. Calathidia 11 — 16-flora. Pappus saepissime brun- neus. a. Calathidia 11 — 13-flora. | Folia subtus saepe tenuissime sericea. In- florescentia satis contracta. Squamae in- volucri interiores saepe breviter acutatae. Antherae 1,2 mm longae. — Hispaniola. V. sericea L. C. Rich, subsp. racemosa (Delp.). Ekman. f| Folia subtus densius sericeo-tomentosa. Inflorescentia diffusa, cymis elongatis. Squamae involucri interiores acutae. An- therae minimae, 0,8 mm longae. — Porto Rico — Ins. infra ventum. V. sericea L. C. Rich. p. Calathidia 15 — 16-flora. — Porto Rico — In- sula? infra ventum. ^c V. Oleasonii Ekman. Artificial Key to the Species of the Subsect. Arhorescentes. I. Leaves linear to narrowly linear-lanceolate. A. Pappus white. 1. Leaves greenish beneath. — Cuba. V. commutata. 2. Leaves whitened beneath. a. Leaves revolute, sericeous beneath, a. Inflorescence spreading. — Cuba. V. angustissima. Inflorescence contracted, narrowly oblong. f Leaves 4 — 7 cm long; heads 18 — 21 -flowe- red. — Hispaniola. V. stenophylla. •j" f Leaves 1 — 3 cm long; heads 11-flowered. — Cuba. V. corollaphila. P- 36 ar'kiv for botanik. band 13. n:o 15. b. Leaves flat, tomentose beneath. — Cuba. V. gnaphali folia var. angustata. B. Pappus brown. — Hispaniola. V. sericea subsp. racemosa. II. Leaves broadly lanceolate to rounded. A. Inflorescense many -flowered, cymes elongated. 1. Leaves greenish beneath, glabrous or pubescent. a. Pappus brown or tawny, a. Scales bluntish. •j* Leaves coriaceous. Involucres obconic. — Jamaica. V. rigida. fl Leaves membranaceous. Involucres cam- panulate or cylindric. X Involucres cylindric. — Jamaica. V. pluvialis. X X Involucres campanulate. * Scales not marginated. aa. Involucres brown, 5 mm high. — Jamaica. V. acuminata. (3(3. Involucres greenish-brown, 4 mm high. — Jamaica. V. expansa. ** Scales marginated. — Cuba. V. membranacea. (3. Scales acute. | Inflorescence contracted, cymes straight. — Porto Rico — Leeward Islands. V. albicaulis. f| Inflorescence spreading. X Cymes flexuose. — Porto Rico. V. borinquensis. X X Cymes straight. * Heads 11 — 13-f lowered. Inner scales somewhat bluntish. — His- paniola. V. sericea subsp. racemosa. ** Heads 15 — 16-f lowered. Inner scales acute. — Porto Rico — Lee- ward Islands. x I7. Gleasonii. b. Pappus white. a. Scales loose, somewhat spreading, outer ones often slightly recurved. E. L. EKMAN, WEST INDIAN VERNONIiE. 37 -j- Outer scales elongated, subfiliform. — Trinidad. V. Trinitatis. •j"| Outer scales linear-lanceolate, acute. — Windward Islands, Margarita. V. arborescens. 3. Scales appressed, imbricated. -j- Cymes flexuose. — Jamaica. V. acuminata var. -{--j- Cymes straight. X Inner scales somewhat bluntish, often purplish. — Jamaica, Grand Cayman. V. divaricata. X X Inner scales acute, never purplish. — Porto Rico — Leeward Islands. V. albicaulis. Leaves whitened, or at least densely tomentose beneath. a. Pappus brown. a. Leaves sericeous beneath. Scales acute. f Heads 11 — 13-flowered. — Hispaniola — Leeward Islands. V. sericea. ff Heads 15 — 16-f lowered. — Porto Rico — Leeward Islands. X V. Gleasonii. 3. Leaves tomentose beneath. Scales obtuse. — Cuba. V. desiliens. b. Pappus white. a. Leaves flat. f Leaves loosely tomentose beneath. X Cymes many-headed, often elongated. * Inner scales often purplish. — Ja- maica, Grand Cayman. V. divaricata. ** Inner scales never purplish. — His- paniola V. fruticosa. X X Cymes few-headed, often abbreviated. * Leaves acute, only the younger ones whitened beneath. — Cuba. V. crassinervia. ** Leaves obtuse, all whitened beneath. — Cuba. V. gnaphali folia. f f Leaves densely tomentose beneath. 38 ARKIV FOR BOTANIK. BAND 13. NIO 15. X Leaves rounded or obtuse at the apex. Straggling plant. — Cuba. V. calophylla. X X Leaves acute. Erect plants. * Leaves elliptic to elliptic oblong. In- volucres 5 — 6 mm high. — Cuba. V. vicina. ** Leaves broadly lanceolate. In- volucres 3 mm high. — Cuba. V. parvuliceps. p. Leaves bullate. f Leaves acute. Cymes elongated. X Involucres to 12 mm high, scales with' filiform tips. — Cuba. V. pineticola. X X Involucres 5 mm high, scales lanceolate, acute. — Cuba. V. leptoclada. ff Leaves obtuse or rounded. Cymes ab- breviated. X Involucre densely pubescent. — Cuba. V. calida. X X Involucre thinly pubescent, or glabrate. — Cuba. V. senlitalis. B. Inflorescence few-headed, cymes reduced, or forming terminal capitate clusters. 1. Inflorescence of reduced cymes. a. Leaves flat. a. Leaves narrowed at the base, broadest above the middle. -j- Leaves narrowly obovate, rounded at the apex. — South-eastern Bahama Islands. V. bahamensis. jf Leaves broadly obcordate, emarginated at the apex. — Little Inagua Island, Ba- hamas. V. obcordata. (3. Leaves rounded at the base, or shortly atte- nuated, broadest near the middle. — New Providence Island, Andros Island, Bahamas. V. arbuscula. b. Leaves undulate or curled towards the margins. — Cuba. V. complicata. E. L. EKMAN, WEST INDIAN VERNONIiE. 39 2. Inflorescence of capitate clusters. a. Pappus brown. — Jamaica. V. pluvialis. b. Pappus white. a. Leaves densely tomentose beneath. A single terminal cluster of heads. — Cuba. V. yunquensis. (3. Leaves finely pubescent beneath. Clusters of heads both terminal and axillar. — Cuba. ; V. segregata. 1. The allies of Vernonia argyropappa Buek. Ternonia Trinitatis Ekman n. sp. Suffrutex elatus, certe metralis et ultra. Caulis (cujus pars superior tantum adest) erectus, inferne lignosus, diam. 4,5 mm, simplex, superne ramosus, teres, striatus, inferne subglaber, cortice pallide avellaneo, superne densiuscule pu- bescens, inferne nudus, apice foliatus. Folia alterna, paten- tia, subsessilia vel brevissime petiolata, petiolo ad 2 mm longo, tomentello, complanato; laminae anguste ovatse, ad 13 cm longae, 6 cm latae, infra medium latissimse, basi late rotundatse, apice longe acuminato-acutissimae, membrana- ceaa, planae, integrae vel remote denticulatae, supra olivaceo- -virides, subglabrse, sublaeves, subtus pube molli, subappressa, subvelutina, rufo-grisea instructae, costa nervisque utriusque lateris numero 8, supra impressis, subtus exculptis, nervulis obsoletis. Inflorescentia cymoso-scorpioidea, ampla, ambitu obpyramidata, ad 3 dm longa, 2,5 dm lata, diffusa, cymis elongatis, patulis, sursum arcuatis, simplicibus vel ramis nonnullis brevibus instructis, circ. 15-cephalis, calathidiis inferioribus spatio ad 3,5 cm longo disjunctis, superioribus magis approximatis, omnibus bracteis suffultis foliaceis, iis multoties longioribus. Calathidia axillaria, sessilia, solitaria vel ssepe geminata, circ. 25-flora. Involucrum infundi- buliformi-campanulatum, 8 mm altum et totidem fere latum, squamis circ. 5-seriatis, extimis ceteris dimidio brevioribus, lineari-subulatis, subpungentibus, patulis, apice subrecurvis, intermediis inferne anguste oblongis, infra medium in subu- lam longam, patulam, subrecurvam abeuntibus, pulchre ca- rinatis, intimis lineari-ligulatis, breviter acutatis, omnibus 40 ARKIV FOR BOTANIK. BAND 13. NlO 15.. pubescentibus, sordide griseo-viridibus, intimis superne pal- lide brunneis. Receptaculum planum, nudum. Pappi setae interiores persistentes, numero 35, tenues, filiformes, albi- dse, 5,5 mm longae, quam exteriores sequilongse, distinctse circ. 8-plo longiores. Corolla 8 mm longa, glabra, tubo 2f3 longitudine corollas. Antherse 2,8 mm longse, ligula x/5 long- antherae, nervo subdistincto, auriculis obtusiusculis. Achaenia (valde immatura) cylindrica, appresse sericeo-pubescentia, costis subconspicuis. Hab. in Trinidad: Lockhart (K); »E reliquis Crtjeger., Purd. etc. arranged by I. H. Hart a. 1888 », n. 2036 (KU, orig. spec). The description of this new species has been made from the specimens of the Krug-Urban Herbarium. The plant collected by Lockhart is very badly preserved, nearly all heads being discharged, without bracteal leaves. Its leaves are somewhat larger than those of the type specimen, 15 cm long by 7 cm wide; otherwise it agrees well with it. V. Trinitatis belongs to a little group of species very nearly related to each other. The best known of these spe- cies is V. argyropappa Btjek (= V. Poeppigiana DC. Prodr. p. 55, not V Poeppigiana DC. Prodr. p. 20. De Candolle in his Prodromus described two V. Poeppigiana^. Buek, having observed the fault, nullified the name of the second species, calling it V. argyropappa. This name should be retained for the species, even though the first V. Poeppigiana is now referred to Piptocarpha). The other species of the group are V. Miersiana Gardn. from eastern Brazil (= V. Salzmannii Baker in Fl. Bras., not V. Salzmannii DC), V. virens Sch.- -Bip. from Central Brazil, V. hirsutivena Gleason from Yuca- tan, and two not published species, V. Friedrichsthalii Sch.- -Bip. mscr. from Guatemala — Columbia and V. strigosa Sch.- -Bip. mscr. from Mexico. The species in question are, in fact, very closely related, and might be united into one single, though extraordinary variable, species. However, I prefer not to do so. They look, after all, rather dissimilar, they are easily characterized, and they have, moreover, their particular geo- graphical ranges. The following key gives their characteristics. I. Outer scales very long-cuspidate. A. Leaves rounded at base, velvety-pubescent beneath. — Trinidad. V. Trinitatis. E. L. EKMAN, WEST INDIAN VERNONI.E. 41 B. Leaves narrowed at base, strigose-pubescent beneath. — Guatemala — Columbia. V. Friedrichsthalii. II. Outer scales not at all, or shortly, cuspidate. A. Leaves broad, elliptical, conspicuously narrowed at base. 1. Leaves velvety-pubescent beneath, especially on the veins. — Yucatan. V. hirsutivena. 2. Leaves strigose-pubescent beneath. — Central Brazil. V. virens. B. Leaves narrowly oblong to narrowly lanceolate, shortly narrowed at base. 1. Inner scales beautifully purplish, involucres 8 — 10 mm high. — Mexico. V. strigosa. 2. Inner scales not purplish. a. Heads distant, involucres about 7 mm high, scales densely imbricated. — Peru. V. argyrojiappa. 3. Heads approximated, about 6 mm high; scales looselv imbricated. — Eastern Brazil. V. Miersiana. Besides this, there are well-marked differences in the structure of the flowers and the pappus. However, I have had a rather scanty material to examine, and the differences found thus cannot be published here. The names V. Friedrichsthalii and V. strigosa are not to be considered as published in this paper. From V. Trinitatis, or some closely related species, V. arborescens (L. ) Sw. may have originated. Especially the spe- cimen collected by Lockhart greatly resembles this species. 2. The allies of Vernonia arborescens (L.) Sw. Yernonia arborescens (L.) Sw. Tabula nostra II, fig. 1 (cyma). Conyza arborescens Linn^us, 1759 a, p. 1213 ; 1763, p. 1209, excl. syn. Brownei et Sloanei. Vernonia arborescens Swaetz, 1806, p. 1320, quoad typum Linn^i, haud quoad descript., nee patriam. Lepidaploa arborescens »Cass.» ap. Lessing, 1829, p. 302 (ap. Cassini 1823, nomen non inveni), quoad typum Linn^i. Vernonia divaricata Lessing, 1829, p. 306 — non Sw- 42 ARKIV FOR BOTANIK. BAND 13. N:0 15. Vernonia icosantha De Candolle, 1836, p. 49; Gleason, 1906 b, p. 178. Vernonia arborescens Sw. var. divaricata Grisebach, 1861, p. 353.) Cacalia arborescens Kuntze, 1891, p. 323, quoad typum Linn^ei. Vernonia ventosa Gleason, 1906 b, p. 179. Eupatorium arborescens, floribus cceruleis. Plum. Cat- pi. Amer., p. 10. Conyza foliis ovatis sessilibus acuminatis floribus ramulorum axillaribus sessilibus solitariis. Plum. ed. Burm., p. 122. — Tab. 130, fig. 2. Pappi setse interiores persistentes, numero 30, tenues, filiformes, albidse, exterioribus distinctis 5 — 6-plo longiores. Corolla 7 — 8,5 mm longa, glabra, tubo long. 1/2 corollae, limbi laciniis apice subtus papillosis, interdum pilis nonnullis brevi- bus instructis. Antherae 2,5 — 2,8 mm longae, ligula long. 1/6 antherse, auriculis breviter acuminatis. Hab. in Cuba: Havana, verisimiliter culta, de la Ossa, anno 1825 (Prodr). — Martinique: Herb. Surian, fasc. 2, n. 198 (P, orig. spec.); Herb. Jussieu, coll. haud not. (P); ex Herb. Vaillant, coll. haud not. (P); Rohr (H, S); »in caribaeis frequens», L. C. Richard (P, sub nom. Serratula solidaginoides); Isert (H); West (Prodr, S); Forsstrom (S); F. j Kohaut in Sieber, Fl. Martin, n. 190 (B, Bss, M, P, S), Fl. Martin, n. 419 (D); Plee (P, »commune a la Mar- tinique, en 5. Mai 1820», »il est tres commun a la Martinique ou il est connu sous le nom de Bois moelle»); Dtjperrey (P); M:me Riviere (P); »en abondance sur les mornes qui en- tourent le Fort Bourbon », April. 1839, A. Steinheil (P); »pres du chemin au Morne Rouge », 8. 5. 39, A. Steinheil n. 137 (P); Perrottet (D); »mornes a l'Est de Saint Pierre », Maj. 1853, Belanger n. 155 (D, F); Belanger n. 226 (BB), n. 394 (P); mense Mart. 1856, Crueger (?.K); »troisieme port, champs incultes», Nov. 1867, Hahn n. 569 (B, BB, D, DC, K, P); Hahn n. 925 (BB, DC, P); Hahn, anno 1870 (K); anno 1890, leg. Duss n. 303, 304, 305, 931, 984 (KU); ad viam inter St. Pierre et Morne Rouge, alt. 25 — 500 m, Aug. 1899, Duss n. 4069 (KU). Insuper Mignot (Bss). — Santa Lu- cia: Anderson (K); Crudy, ex. herb. Schmiedel. (M); coll. haud not., ex herb. Schreb. (M); ex herb. Schwagriciien (M). — St. Vincent: anno 1822, Guildlng (D, K, M); in silvis montium Montrose Hills, 26. 12. 89, Eggers n. 6545 (KU, P); in montibus St. Andrews, 2. 1. 90, Eggers n. 6777 (KU); H. H. et G. W. Smith n. 359 (KU); ad stationem botanicam, Powell n. 77 (KU); coll. haud not. ex herb. E. L. EKMAN, WEST INDIAN VERNONLE. 43 Fischer (Petr). — Be qui a: in declivibus montium inter frutices, alt. 300 m, Julio 1890, H. H. et G. W. Smith il B. 297 (KU). — Margarita: in monte San Juan, alt. 500 m, 19. 7. 03, J. R. Johnston n. 96 (H, KU). The history of this species goes back to the remarkable voyage of Pltjmier and Surian in 1689 — 90. Having dis- covered it, most probably, in Martinique, Plumier figured it in Plum. ed. Burm., tab. 130, fig. 2, and upon this figure LrNNiEus in Systema Naturae, 1759, p. 1213 based his Co- ny za arborescens. There is a point to be discussed in this connection. The passage in Systema Naturae, ed. X, where Linnaeus describes Conyza arborescens reads as follows: »C. (arbore- scens) fol. ovatis integerrimis subtus tomentosis, spicis recur- vatis secundis, bract, reflexis. Plum. ic. t. 130. f. 2.» The description is not merely copied from Plumier; for Llnnjeus has given here some characters of his species that must refer to another plant than that figured by Plumier, for instance the statements »fol. . . . subtus tomentosis, spicis recurva- tis ». There is full evidence as to which plant these statements refer to. In 1758 Linnaeus had purchased a set of plants collected in Jamaica by Patrick Browne, and among these plants there was a specimen of a Vernonia, the same which I men- tion in this paper, on p. 59, as V. divaricata Sw. This specimen is now in the Linnaean Herbarium in London. When I was there in 1910, I had not yet decided to study especially the West Indian Vernoniae, and I only stated that under C. arbo- rescens in the Linnaean Herbarium there were two plants: the one V. divaricata Sw. from Jamaica, labelled in Llnnjeus's handwriting »Co?iyza arborescens)} from »Br. »; the other V. scorpioides (Lam.) Pers. from Guyana. Prof. C. A. M. Lindman subsequently (in 1912) visited the Linnaean Society, and obliged me by making a brief description and an exact drawing of the two specimens in question. I thus easily ascertained that the remarks »fol. . . . subtus tomentosis, spicis recurvatis» refer to the plant collected by Browne. Another statement in the description of Conyza arborescens, »bracteis reflexis », seems to refer to the plant figured by Plumier, the specimen in the Linnaean Herbarium showing only a few small bracts. As matters now stand, I think M'e have better take the figure of Plumier connected with 44 ARK1V FOR BOTANIK. BAND 13. N:0 15. the specimen in the Surian Herbarium (P) as the type of Conyza arborescens L. rather than the plant in Linn^us's her- barium, as this only partly corresponds with his description. Linnjeus himself seems to have laid more stress upon the plant figured by Plumier, to judge by his specific name of the plant, C. arborescens, the name taken from ))Ewpatorium arbo- rescens jloribus cceruleis)}, Plum. 1. c. It is puzzling that he has not quoted the synonym in Browne, Jam., p. 313, under his C. arborescens, which he would certainly have done, if he had considered the Jamaican specimen in writing the description. It. seems as though C. arborescens had been originally based upon the figure of Plumier, and had had its description afterwards somewhat modified according to the Jamaican plant. If the name G. arborescens goes with the description, it must be the Jamaican plant which should be named V. ar- borescens (L.) Sw. The Martinique plant of Plumier should then be called V. icosantha DC. Later on in the year 1759 LiNNiEus attributed, in Pu- gillus Jamaicensium plantarum, Amoen. Acad. Vol. 5, p. 406, to his C. arborescens the synonym of Browne already mentioned, and another synonym also referring to the Jam- aican species from Sloane. In this way the incorrect application of G. arborescens to the Jamaican plant became more decisive. Olof Swartz, in fact, on his return from Jamaica, saw in London the plant of Browne and Sloane, and taking this to be the genuine Conyza arborescens, he described it in his Observationes, p. 304. On the same plant he afterwards, in Flora Indiae occidentalis, based his Vernonia arborescens. Thus the Vernonia arborescens became definitely a Jamaican plant. Consequently, Avhen it became evident that the Mar- tinique plant was different from the Jamaican one, the former was described as a new species, viz. V. icosantha DC. A general vagueness, however, as to the correct application of the name V. arborescens gradually crept in, so that nearly all species of the Arbor escentes have borne the name. Thus do matters stand up to the end of the nineteenth century. At length, the demand for a consistent nomen- clature had begun to make itself felt. The routes of the ancient travellers were studied in order to ascertain where E. L. EKMAN, WEST INDIAN VERNONIiE. 45 their plants han been collected. Thus it became obvious, in our case, that the type of V. arborescens could not be the Jamaican plant, Plumier and Surian never having visited Jamaica. In 1902 Prof. Ign. Urban succeeded in recognizing the type of V. arborescens in a plant sent by Duss from Mar- tinique, though he did not publish his discovery. I am happy to mention that on a visit to Paris I found the Conyza arbo- rescens in the Surian Herbarium, fasc. 2, n. 198. It is the same plant as that of Duss, and represents a leafy form of V. icosantha DC. The Surian plant must be considered as the type of V . arborescens (L.), being most probably the origin- al of the figure in Plum. ed. Burm. The application of the name V. arborescens to the Mar- tinique plant being thus adopted here, an interesting question remains. How to quote the author of the combination V. arborescens (L.)? The first to transfer Conyza arborescens L. into Vernonia was Swartz. But the description of his V. arborescens is incorrect, as is its distribution, Swartz having the Jamaican plant in mind. Now the question is: can Swartz be cited as author of the combination V. arbo- rescens (L.), applied to the Martinique plant? I think he can, and must, be so cited. For the operation of transferring a spe- cies from one genus into another is a matter independent of the plant at the same time described. The new combination formed belongs to the same plant to which' the original name was given; it has, in other words, the same type as this, the two names are tjTponyms in the American sense of the word. And the one who first published it is to be cited as author of the combination. Thus in our case, Swartz is to be cited for V. arborescens (L.), and this name has for its type the type of Conyza arborescens L. When Swartz refers the Jam- aican plant to his combination, it is simply a false determin- ation. I have asked Prof. Urban for his opinion on this matter, and he answered me: »Die Vernonia arborescens- -Frage ist mir vollstandig gegenwartig, da ich bereits vor vielen Jahren die Plumiersche Abbildung mit Vernonia ico- santha DC. von Martinique identifiziert habe. . . . Friiher war ich auch der Meinung, dass man auf Grund einer solchen Identifizierung nochmals die Combination vornehmen unci seinen eigenen Namen als Autor dahinter schreiben soil. Ich habe sie jetzt aufgegeben, weil sie unpraktisch und nicht 46 ARKIV FOR BOTANIK. BAND 13. N:0 15. ganz gerecht ist; denn Swartz hat doch erkannt, dass die Conyza arborescens eine Vernonia ist, und hat die Combina- tion richtig vollzogen. Was bei ihm darauf folgt: Beschrei- bung und Vaterland ist allerdings falsch. Wenn Sie genau sein wollen, konnen Sie schreiben: V. arborescens (L.) Sw. (quoad syn. Linn., non quoad descript. nee patriam. Auch die Nordamerikaner verfahren so von ihrem Princip der 'Priority in place' aus». For further illustration to this point I refer to Hitch- cock, North American species of Panicum, Contr. U. S. Nat.. Herb., Vol. 15, 1910, p. 6, and to Urban, Symb. Antill., Vol. VII, 1913, p. 342. V. arborescens (L. ) Sw. is evidently the Vernonia of the Windward Islands, being known from Martinique, Santa Lucia, St. Vincent, Bequia and Margarita. There are only two points indicating a wider distribution of the species. In the Prodromus Herbarium I found a plant from Havana collected by de la Ossa, and referred to V. arborescens (L.) Sw. (3 ovatifolia DC; it is, however, a typical V. arbo- rescens (L. ) Sw. Nothing is known about de la Ossa collecting in Martinique, so that it must be supposed that the ' plant occurred at Havana. Most probably it grew in the bo- tanical gardens of that town, of wich de la Ossa was director. Further, Gleason asserts that he has seen specimens of his V. ventosa from Guadeloupe. I venture to suppose that this statement is due to a confusion with the large-headed form of V. albicaulis Pers. common in that island. In fact, Duss n. 2812, representing this very form, was deter- mined by Gleason in a letter to Prof. Urban as V. ventosa Gleason. In his »Studies » he cites the same plant as V. ico- santha DC. (Gleason, 1913, p. 307). An interesting locality for the species is the island of Margarita near the coast of Venezuela, rather distant from its main region. It occurs there in a somewhat different form, hardly at first sight recognizable as V. arborescens. In habit it resembles a rigid V. Trinitatis Ekman, but is easily distinguished from that species by its small heads. Like many West Indian Vernonia?, V. arborescens is ex- tremely variable. I need not insist that I have repeatedly looked over the abundant material on the chance of discov- E. L. EKMAN, WEST INDIAN VERNONLE. 47 ering any constant forms, and that I have examined numbers of flowers for the same purpose. But in vain. The forms pass evidently into each other, and their particular habits are probably due to the particular circumstances in which the plants have lived. However, the types of the three different names of the species, V. arborescens, V. icosantha, and V. ven- tosa, represent, as it happens, each a particular form of the spe- cies, yet not in such a way, that the species could be divided throughout into three series each corresponding to one name. For the intergrading forms are as numerous as the typical ones. The three forms may be briefly characterized. 1. The type of V. arborescens (L.) Sw. : Leaves broad, dim. 11,5 cm in length, 5 cm in width, thinly membrana- ceous, nearly glabrous, with broadly rounded and slightly cordate base. Inflorescence very lax, the heads at a distance of 2 — 4 cm and subtended by bracteal leaves, the lower of which are nearly as large as the leaves themselves. Evidently, this is a form of damp, shady localities. So, for instance, Duss n. 305, 931, 4069. 2. The type of V. icosantha DC: Leaves large, somewhat narrowed, up to 11 cm long, 4 cm broad, attenuate at the base, firm, nearly glabrous. Inflorescence somewhat dense, the heads more crowded, 1 cm or so apart, the bracteal leaves smaller, the upper ones linear, not longer than the heads. So Eggers n. 6545, 6777, Powell n. 77, H. H. et G. W. Smith n. B. 297, etc. 3. The type of V. ventosa Gleason: Leaves small, lanceo- late-ovate, usually 6 cm long, 2 cm wide, firm, rounded or attenuate at the base, strigose pubescent, often grayish or even whitened beneath. Inflorescence generally dense, the brac- teal leaves small. So Duss n. 303, 304, 984. Hahn in 1870 represents an interesting form with leaves of V. ventosa and inflorescence of V. arborescens. Vernonia albicaulis Pers. Tabula nostra II, fig. 5 (pars. infl. et folium). Eupatorium obtusi folium Wili/denow, 1804, p. 1768. Conyza glabra Willdenow, 1804, p. 1940. Vernonia albicaulis Persoon, 1807, p. 404, n. 11 : Lessing, 1829, p. 313; Gleason, 1906 b, p. 186. 48 ARKIV FOR BOTANIK. BAND 13. N:0 15. Vernonia longifolia Persoon, 1807, p. 404, n. 12; De Candolle, 183G, p. 49; Gleason, 1906 b, p. 186; 1913, p. 326; Urban, 1911, p. 620. Lepidaploa albicaulis Cassini, 1823, p. 17. Lepidaploa lanceolata Cassini, 1823, p. 18. (?) Vernonia punctata Swartz apud Wikstrom, 1828, p. 72; Grisebach, 1861, p. 353. Vernonia emarginata Wikstrom, 1828, p. 73. Vernonia Vahliana Lessing, 1829, p. 306; 1831, p. 666; De Candolle, 1836, p. 48, p. p. Vernonia arborescens Sw. [3 ovatifolia De Candolle, 1836, p. 48, p. p. Eupatorium secundiflorum Bertero ex De Candolle, 1836, p. 48. Vernonia Thomas Bentham, 1852, p. 66; Gleason, 1906 b, p. 191. Gacalia Thomos Kuntze, 1891, p. 324. Cacalia punctata Kuntze, 1891, p. 971. Vernonia longifolia Pers. varietates a genuina, (3 Vahliana, t Sintenisii Urban, 1899, p. 456; 1911, p. 620. Vernonia Sintenisii Gleason, 1906 b, p. 187. Pappi setae interiores persistentes, numero 35 — 40, te- nues, sordide stramineae vel albidae, quam exteriores distinctae 7-plo longiores. Corolla 6 — 8 mm longa (in specimine unico anomalo corollas vidi 10,5 mm longas), glabra, tubo apice parum dilatato long. l/2 corollae, limbi laciniis apice subtus papillosis vel pilis nonnullis brevissimis instructis. Antherae 2 — 2,8 mm longae (in spec, unico anomalo vidi antheras 3,2 mm longas), ligula fere enervi, long. 1/s antherae, auriculis longe acutatis. Hab. in Porto Rico: Ledru (P, herb. Lamarck); Grosourdy, cat. n. 13 (P); prope Isabela, Jun. 1887, Stahl n. 730 (KU); inter Ponce et Penuelas, in montibus calcareis, Aug. 1888, Stahl n. 926 (KU); prope Cano Grande, anno 1876, Gundlach n. 594 (KU); prope Guanico in declivibus umbrosis montis El Maniel, 10. 2. 86, Sintenis n. 3727 (BB, D, DC, H, K, KU, M, P, S); prope Penuelas in declivibus montis Vi, 8. 7. 86, Sintenis n. 4731 (K, KU); prope Penuelas in declivibus umbrosis montis Llano, 6. 7. 86, Sintenis n. 4750 (KU); prope Rincon in montibus ad Barrio Punta, 11. 12. 86, Sintenis n. 5639 (KU); 5. 12. 02, A. A. Heller n. 6214 (D). — St. Thomas: Ledru n. 230 (P); Ravn (H); Krebs (H); Oersted n. 56 (K); Oersted sine num. (H); Eggers in Aug. 1880 (KU); Eggers n. 1170 (B, M, P); ad Bolongo, Dec. 1886, Eggers n. 34 (H); Eggers, Dec. 1887 (KU); Mart. 1874, O. Kuntze n. 131 (K); Nov. 1905, Raun- kler n. 3154 (H); ad Mandal, 5. 5. 06, Raunkijer n. 1918 E. L. EKMAN, WEST INDIAN VERNONLE. 49 (H); ad Lovenlund, 5. 5. 10, Raunkijer n. 3155 (H); coll. haud not. (P). — St. Jan: in campestribus apricis S:ti Joannis, L. C. Richard (P); in summis collibus mari imminentibus prope Brown's Bay ins. S:ti Joannis, L. C. Richard (P); ad Hermansfarm, 9. 3. 77, Eggers sine num. (H); ad »Klein Kaneel» Bay, 24. 12. 87, Eggers n. 3036 (H); in parte occi- dental!, 13. 2. 06, Raunkijer n. 3136 (H); coll. haud not. (D). — St. Croix: Rohr (Br); in montibus summis apricis S:tse Crucis, L. C. Richard (P); West (H, S); Benzon (H); Ravn (H); ad viam prope »Elizas Retreat », Dec. 1869, Eggers n. 407 (H); eo loco, Dec. 1870, Eggers sine num. (H); ad Christianstecl frequens, Jan. 1872, Eggers sin. num. (H); ad Crequis, 10. 9. 74, Eggers sin. num. (H); Paulsen n. 68 (H); in litore septentrionali, Maj. 1893, Paulsen n. 75 (H); ad Marion Hoy, 17. 1. 96, A. E. Ricksecker n. 220 (KU); ad »Northside Road», 20. 2. 97, J. J. Ricksecker n. 150 (KU); in Jolly Hill, 1. 1. 06, Raunkler n. 1919 (H); in valle »Cale- doniadalen», 2. 2. 06, Raunkler n. 3162 (H); coll. haud not. (P, herb. Jussieu, orig. spec. V. albicaulis Pers.). — St. Martin: 4. 5. 85, Suringar sine num. (KU). — St. Barthelemy: Forsstrom (Br, S); Goes (KU, S). — Saba: anno 1906, Boldingh n. 1968 B. (KU). —St. Eus- tache: in colle Signalhill, 9. 4. 85, Suringar sine num. (KU). — St. Kitts: Ryan (H); ad Sandy Point, anno 1901, Britton et Powell n. 144 (KU). — Antigua: L. C. Richard (CD, sub nom. Vernonia bipartita Rich., P); Nicholson (K); Lane n. 415 (K); in fruticetis, Wullschlagel n. 290 (G, M). — Montserrat: Ryan (H, S). — Guadeloupe: in fruticetis — silvulis redivis ins. Guadeloupe, L. C. Richard (P); Badier n. 44 (Prodr); Forsstrom (Bss, Prodr, S); Ber- tero (B, Prodr); 20. 6. 24, Perrottet n. 251 (P ); Montenegro de la Guadeloupe, 27. 6. 24, Perrottet n. 250 (P); Jan. 1843, L'Herminier (P); Krauss, anno 1818 (Prodr); Beauper- tuis (P); Duchassaing (G, K, P); Duss n. 435, 436, 437 (P); ad Morne Gobelin, Gombegre, anno 1892, Duss n. 2489 (KU); Baillif, Roulade Basse-Terre a Monteran, Vieux Fort — Capesterre; Desirade, locis siccis, alt. 100 — 315 m, anno 1892, Duss n. 2812 (KU). —Desirade: conf. adnot. sched. Duss n. 2812. — Dominica: Imray n. 294 (G); prope Wallhouse, 18. 2. 80, Eggers n. 73 (K, KU); in monte Morne Gombo, 2. 3. 80, Eggers n. 75 (K, KU); Dec. 1881, Eggers Arkiv for botanik. Band 13. N.o 15. 4 50 ARKIV FOR BOTANIK. BAND 13. N:0 15. n. 891 (K, KU); Eggers n. 1171 (B, M, P) — Sine coll. notis: Ex herb. Vaillant (P, »Conyza americana Salicis Caprece folio, flore albo. J. R. H. 455 »); Herb. Willd. n. 15144 (B, orig. spec. Eupatorii obtusifolii Willd.); Herb. Willd. n. 15622 (B, orig. spec. Cony zee glabrae Willd.); West (Prodr.); »373, in rupibus calcareis prope Mocam», Aug. 1827 (D, verisimiliter coll. Wydler); Eggers n. 196 (D), n. 416 (P). Though early collected (represented, for instance, in the herbaria of Vaillant, Lamarck and Jussieu) the species did not get its valid name, V . albicaulis Pers., till the year 1807. Certainly, there are two older binomial synonyms for it, both of Willdenow, viz. Eupatorium obtusifolium and Conyza glabra, but these specific names are invalidated by Vernonia obtusi folia Less, and Vernonia glabra Vatke, both valid names. Vernonia albicaulis Pers. was published in the same work, even on the same page as its synonym V. longi folia Pers. The latter name has generally been used for the species. When I take the first to be valid, my strongest reason is the fact that V. albicaulis precedes V. longi folia on the page, the former being numbered 11, the se- cond 12. Further, V. albicaulis represents a central- form of the species, V. longifolia an outlying one. Since Gleason in his »Revision» uses the name V. albicaulis, it will surely not appear too unfamiliar. V. albicaulis is the Vernonia of the Leeward Islands, as its near relative V. arborescens (L.) Sw. is the Vernonia of the Windward Islands. It is true that I have seen several specimens collected, according to their labels, in other islands. Collectors, however, have not always been so cautious with their statements as in our days. The labels were often writ- ten a long time after the plants had been collected, and lapses of memory were consequently likely to occur. On the distri- bution of the plants, labels may often have been interchanged. However, these specimens presumably erroneously labelled are very few compared with the abundance of material collected in the Leeward Islands. I have seen the following: Mayer- hoff, Hispaniola, in 1852 (B). In habit this plant agrees exactly with specimens of V. albicaulis from St. Thomas and it may, indeed, have been collected there. No other col- lector has V. albicaulis from Hispaniola. — - »N:o 370. Coniza E. L. EKMAN, WEST INDIAN VERNONIiE. 51 unique e Martinique)). This specimen is in the Lamarck Herbarium at Paris; no collector is given. — Martinique, Plee n. 957 (P). The specimen may be from Porto Rico, where Plee also collected. ■ — Martinique, Duss n. 302 (KU); according to the label the plant was collected in 1890, and since Duss that year collected in Guadeloupe, it is very probable that his plant came from that island, inasmuch as it agrees perfectly with other specimens from Guadeloupe, for instance, Duss n. 2489. — Martinique, Duss n. 4070 (KU). The label of this specimen reads: »N:o 4070. Pere Duss, Herbier de la Martinique. Vernonia longifolia Pers. ex Gleason in Uteris. Casser coutelas. Haut de F/2 — 3*/2 m. fl. violacees ou violet pale ou blanches. Abondant, fl. de Juin en 10-bre. Case Pilote. Alt. 2—450 m. LeAoutl899». According to this the plant should be common at Case Pilote in Martinique. But then it is puzzling that it has not been collected there more than once. Most probably the label in question originally belonged to a specimen of V. arborescens (L. ) Sw., and has been interchanged afterwards. Urban states in his Notse biographical that there are many errors of locality on the Dussian labels. — Santa Lucia: leg. Crudy (M); probably from St. Thomas, where Crudy also collected. — St. Vincent: leg. Anderson (K). The speci- men resembles those from Guadeloupe, and may have been collected there, or in another island near it. There is a specimen of V. albicaulis in the Paris Herba- rium, the locality of which »e Caroline, Th. Noisette », if meant for the North American state by this name, clearly must be an error. As indicated by the numerous synonyms, V. albicaulis is highly variable. Outlying forms of it, such as V. punctata Sw., V. longifolia Pers. and V. Sintenisii Gleason have been described as distinct species. However, it is the same case with V. albicaulis as with other Vernonia?, for instance, V. arborescens (L. ) Sw. and V. divaricata Sw. : the whole set of forms constitutes a variable, though always easily re- cognizable, species with a particular geographical range, each form considered as a species becomes impossible to delimitate. The only consistent way is therefore to take V. albicaulis as the wide species here conceived. It must be acknowledged, however, that the different 52 ARKIV FOR BOTANIK. BAND 13. NIO 15. forms of the species do not live anywhere within its range. In the northern part of it round-leaved forms with tawny pappus dominate, towards the South narrow-leaved ones with white pappus become numerous. It is a remarkable fact that V. albicaulis thus on the confines of the range of V. arborescens (L. ) Sw. becomes like that species, though al- ways recognizable by the characters of the involucres and by the acute basal lobes of the anthers. If this is due to some ancient hybridization, or if V. albicaulis and V. arbo- rescens are to be looked upon as rather recent offshoots .of one species, the southern forms being almost unaltered descend- ants of the common type, or if the fact of matter is, that either of the species (most probably V. arborescens) on its migra- tion has been altered by the force of climate into the other, the forms in question owing their existence to a particular intermediate climate, of course cannot be discussed here. It may, perhaps, be of some use to characterize the most prominent forms of the species and to give their distribution. 1. The type of V. Sintenisii (Urb.) Gleason: Leaves small, oblong (dimensions: 3X1,2 cm), obtuse or even emar- ginated, thinly pubescent with appressed hairs, somewhat seri- ceous-shining. Heads relatively small, about 8 mm long. Pappus pale straw-coloured. This is the form of Porto Rico. However, not all specimens from this island are true V. Sintenisii Gleason. Some of them agree almost exactly with the type of V. albicaulis Pers., for instance, Sintenis n. 470, others cannot be distinguished from V. emarginata Wikstr., so Stahl n. 730. 2. The type of V. albicaulis Pers. Leaves oval (dim. 5X3 cm), obtuse at the apex and shortly attenuate at base, very thinly pubescent, not shining, grayish-green or grayish- -olivaceous. Heads small, 6 — 7 mm long, in number 4 — 5 on slightly recurved cymes, somewhat apart. Pappus very pale straw-coloured. This form occurs in its most typical shape in St. Croix, where it was collected by the ancient Danish bot- anists. The original specimen at Paris came from Vahl, as did the types of Eupatorium obtusifolium Willd. and Ver- nonia Vahliana Less. To this approximate two forms, one having very large, acute leaves (dim. 13 X 6,5 cm) and larger heads (9 — 10 mm long). E. L. EKMAN, WEST INDIAN VERNONIiE. 53 collected in St. Croix and in Montserrat {Conyza glabra Willd. resembles this), the other having small, obtuse leaves (dim. 2?7 — 5 x 1,6 — 2,5 cm), which are very often curled or undu- lated at the margin. The heads of this form are small, denseby crowded towards the top of the naked peduncles, and subtended by great bracteal leaves. It is the form of St. Thomas and St. Jan. The type specimen of V. Thomce Benth., however, does not belong to this form, rather to V. albicaulis sensu strict. 3. The type of V. emarginata Wikstr. Leaves small (dim. 2,5 X 1,3 cm), thinly pubescent or nearly glabrous, obtuse or somewhat emarginated at the top, rufous-green (the branches are often rufous-tomentose). Heads small, 6 mm high or so, densely crowded in the top of naked peduncles forming a rounded inflorescence, or scattered along more elongated cymes. Pappus brown or tawny. The range of this form extends from St. Martin to Guadeloupe. How- ever, V. albicaulis sensu strict, seems to live within this range, in Montserrat. Perhaps the locality of the specimen in ques- tion, collected by West, is an error. Also of this form specimens have been seen having very large leaves (dim. 11X4,5 cm) and many-flowered inflores- cence. So Ryan and West from St. Croix (? ) and St. Kitts. 4. The type of V. longifolia Pers. Leaves lanceolate, 1 dm long, 3 — 4 cm broad, acute, glabrous or very nearly so, olivaceous above, pale and somewhat shining beneath. Heads small, 6—7 mm high, numerous in a broadly pyram- idal inflorescence. Pappus tawny, not dense. So in Gua- deloupe and Dominica. I have not seen the type specimen of V. longifolia Pers., this probably being in the Desfontaines Herbarium at Flo- rence. I therefore cannot decide if it belongs to the form just described, or to the following. The very name of the species, however, applies very well to the form described, and since it is the more common one, and always has been considered as the genuine V. longifolia, I think it must be that species. 5. The type of V. punctata Sw. Differs from the pre- ceding in having large heads and white, stiff pappus. The type specimen of V. punctata Sw. has rather narrowed leaves (dim. 6 — 7 X 1,2 — 1,4 cm) and somewhat crowded heads. Other specimens examined have wider leaves and up to 6 54 ARKIV FOR BOTANIK. BAND 13. NIO 15. cm long, curved cymes. This is V. arborescens (L.) Sw. (3. ovatifolia DC. V. punctata is known from Guadeloupe and, in a slightly different form, from Dominica. Vernonia albicaulis Pers. x sericea L. C. Rich. x Vernonia Gleasonii Ekman nov. hybr. Characteres omnes parentium mixtae. Frutex 1 — 2- metralis (Sintenis). Rami tenues, tenuissime striati, adulti glabri, avellanei, hornotini sordide stramineo-tomentosi. Fo- lia spiraliter enata, brevissime petiolata, petiolo 2 mm longo, ut rami tomentoso; laminae anguste ovatae, 3 — 5 cm longae, 1,2 — 1,5 cm lata?, apice obtusae vel brevissime acutae, basi breviter attenuatae, chartaceae, pallide olivaceo-virides, in- tegerrimae, obsolete rugosae, margine leviter revolutae, supra pilis brevibus, appressis inspersae, juniores praesertim subtus tenuiter griseo-sericeae, adultae ut in V. albicauli pilis laxio- ribus vestitae, rete nervorum subtus satis prominente. In- florescentia cymoso-scorpioidea, cymis ut in V. sericea elon- gatis, divaricatis, polycephalis ; bracteis foliaceis calathidiis triplo longioribus vel superioribus iis aequilongis. Oalathi- dia axillaria, sessilia, remota, spatio 1,5 cm longo inter se disjuncta, 15 — 16-flora. Involucrum vivum subcampanu- latum, basi rotundatum, sice, ut in V. sericea late infundi- buliforme, 5 cm altum, circ. 6 mm latum, squamis circ. 4- -seriatis, exterioribus brevissimis, deltoideo-lanceolatis, acu- tissimis, conspicue mucronulatis, inferioribus lineari-lanceo- latis, acutis, omnibus carinatis, villosis, pallide brunneis, vix nitidis, inferioribus saepe subpurpurascentibus. Pappi setae interiores persistentes, numero 35 — 45, tenues, graciles, 4 — 4,5 mm longae, sordide stramineae, exterioribus ligulatis conspicuis circ. 7-plo longiores. Corolla 6,5 mm longa, tubo glabro fere s/s l°ng- corollae, limbi laciniis apice subtus papulo- sis. Antherae 1,6 mm longae, ligula 1/G long, antherae, auri- culis longe acutatis. Achaenium 2 — 2,3 mm longum, appresso- -sericeum, costis distinctis. Planta hoc loco descripta, characteribus inter parentes plane intermedia, quamvis fertilis sit, sine ulla dubitatione origine hybrida est. Hab. in Porto Rico: prope Penuelas in declivibus E. L. EKMAN, WEST INDIAN VERNONLE. 55 umbrosis montis Llano, 6. 7. 86, Sintenis n. 4749 (D, KU, P). _ St. Jan: ad Kiistenberg, alt. 210 m, 31. 12. 87, Eggers n. 3256 (KU). — St. Croix: ex herb. Hornemann (B). — Coll. ignot. (H). In collections from the tropics hybrids are only very seldom found, apparently because the collectors have been collecting by chance, without knowing the species. In our case, however, the probability of getting the hybrid was a greater one, both the parents, V. albicaulis and V. sericea, being doubtless familiar to collectors like Eggers and Sin- tenis. In fact, not less than four particular collections of the hybrid have been made. The hybridous nature of our plant is clearly manifested by the fact, that in all characters it is intermediate between its supposed parents. It constitutes a particular tj^pe, having characters of both parents, and only of them. Since V. al- bicaulis and V. sericea have never been found to pass into each other, it cannot be considered merely as an intergrad- ing form. However, its pollen seems to be rather good. The grains are somewhat cornered and dark-coloured, and do not swell swiftly when boiled in milk acid and water. The achenes are well evoluted. The best character of the hybrid is the length of its an- thers. V. albicaulis has large anthers, 2,5 mm in length, V. sericea very small ones, only 0,8 mm long. Now the hybrid has quite intermediate anthers, 1,6 mm in length. As to the leaves it is rather variable, Eggers n. 3256 having leaves of V. albicaulis sensu strict, and Hornemann like those of V. sericea. The inflorescence resembles that of V. sericea. The involucres are larger than those of V. sericea, including 15 — 16 flowers (in V. sericea 11 — 13, in V. albicaulis 20 — 21); as to their form they are intermediate. The scales resemble those of V. albicaulis, yet the outer ones are cuspidate, as in V. sericea, and the inner ones often somewhat purplish, as in that species. The colour of the pappus is tawny. Vernoiiia borinquensis Urb. Tabula nostra II, fig- 7 (cyma). Vernonia arborescens Sw. var. Lessingiana Stahl, 1887, p. 103, p. p. non Griseb. 56 ARKIV FOR BOTANIK. BAND 13. N:0 15. Vernonia borinquensis Urban, 1903, p. 390; 1911, p. 619; Gleason, 1906 b, p. 179. Pappi setae interiores persistentes, numero 30 — 35, te- nues, filiformes, pallide brunneae, quam exteriores distinctae circ. 8-plo longiores. Corolla 8 mm longa, tubo gracili, subito in limbum amplifiato, long. 1/2 corollas, extus glandulis nun- nullis instructo, limbi laciniis apice papulosis. Antherae 2,5 mm longae, ligula long. 1/7 antherae, nervo satis perspicuo, auriculis obtusiusculis. Hab. in Porto Rico: prope Maricao, ad margines silvarum, 12. 11. 84, Sintenis n. 388 (BB, DC, KU, M); Si- erra de Juncos in graminosis apricis montis Guvuy, 28. 8. 85, Sintenis n. 2659 (KU); Lares, ad margines silvarum ad Anon, 16. 1. 87, Sintenis n. 5884 (KU); Lares, in frutice- tis ad Jobo, 29. 11. 87, Sintenis n. 6079 (KU, P); Utuado, in marginibus silvarum ad Paso-palma, 4. 3. 87, Sintenis n. 6362 (D, KU); prope Mayagiiez, alt. 150 m, 26. 1. 1900, Heller n. 4391 (KU). var. Stahlii Urb. Vernonia borinquensis Urb. var. (3 Stahlii Urban, 1903, p. 391; 1911, p. 620. Pappi setae numero 25 — 30. Antherae 2,3 mm longa?, ligula long. 78 antherae, auriculis acutiusculis; cet. ut in spe- cie. Hab. in Porto Rico: prope Bayamon, in fruticetis, Dec. 1883, Stahl n. 238 (KU); prope Baj^amon, alt. 450 m, Majo 1887, Stahl n. 667 (KU): prope Aibonito in decli- vibus, 25. 11. 85, Sintenis n. 2862 (B, BB, KU, M). Though closely allied to V. arborescens (L.) S\v. and V. albicaulis Pers., V. borinquensis at first sight amply differs in its peculiar habit due to the widely spreading in- florescense with its stiff, zigzag-bent cymes. It recalls some- what V. sericea L. C. Rich., from which it is distinguished, above all, by its nearly glabrous leaves. The variety Stahlii Urb. differs invariantly in wanting glandular points on the lower surface of the leaves, and in its nearly glabrous achenes. However, there is no difference in habit between the species and its variety. It seems to me therefore most conformable to the relationship between the two plants to list them as species and variety. E. L. EKMAN, WEST INDIAN VERNONLE. 57 3. The allies of V. bahamensis Griseb. Vernonia bahamensis Griseb. Tabula nostra II, fig. 2 (habitus et folium). Vernonia bahamensis Grisebach, 1861, p. 352; Gleason, 1906 a, p. 187 ; 1906 b, p. 190; 1913, p. 326. Cacalia bahamensis Kuntze, 1891, p. 369. Pappi setae interiores persistentes, numero 40, tenues, exterioribus distinctis circ. 7-plo longiores. Corolla 6,5 mm longa, tubo glabro, apice sensim dilatato, long. i/7 corollae, limbi laciniis apice pilis nonnullis satis longis munitis, subtus papillosis. Anthera3 2,7 mm longae, ligula nervosula V5 long, antherae, auriculis satis elongatis, obtusiusculis. Hab. in insulis Baham. Fortune Island: 4. 2. 88, Eggers n. 3832 (BB, D, H, KU, M, P, S). — Acklins Island: 9. 2. 88, Eggers n. 3893 (KU). The Vernoniae of the Bahamas have been the object of a particular study by Gleason. His exposition, entitled: »The genus Vernonia in the Bahamas », Bull. Torrey Bot. Club, Vol. 33, 1906, is accepted here as substantially correct. Yernonia obcordata Gleason. Vernonia obcordata Gleason, 1906 a, p. 187; 1906 b, p. 190. Hab. in insula Baham. Little Inagua Island: 20. 10. 04, Nash and Taylor n. 1206. I have seen no specimens of this plant. Yernonia arbuscula Less. Tabula nostra II, fig. 3 (habitus). Vernonia arbuscula Lessing, 1831, p. 664. Vernonia arctata Gleason, 1906 a, p. 185; 1906 b, p. 189: 1913, p. 326. Pappi setae interiores persistentes, numero 45, tenues, quam exteriores distinctae 7-plo longiores. Corolla 6 mm longa, tubo amplo, glabro, long 4/7 corollas, limbi laciniis apice pilis nonnullis praeditis subtusque papillosis. Antherse 2,5 58 ARKIV FOR BOTANIK. BAND 13. N.O 15. mm longae, ligula y5 long, antherse, enervi, auriculis brevibus, truncato-obtusis. Hab. in insula Bah am. New Providence Island: in pinetis, 22. 2. 88, Eggers n. 4187 (KU, M); 10. 1. 90, John et Alice Northrop n. 101 (BB, KU); prope Nassau, 5. 2. 03, Curtiss n. 65 (D, DC, H, K, KU, M,); in altoplanitie montium Blue Mountains, 23. 1. 05, C. F. Millspaugh n. 2481 (KU). In his brilliant exposition of the Vernonias known at that time Lessing describes a V. arbuscula from Mauritius. My attention was called to his plant by a statement of Lessing that it recalled V. arborescens (L.) Sw. On a visit to Berlin I examined the type of the species, and was greatly astonished to recognize in it a throughout typical V. arctata Gleason. The specimen is labelled: »Ex herb. Kunth. Vernonia arbus- cula n. sp. Isle de Bourbon. Salisbury ded. 1816 ». It is a puzzle how Mauritius (= Isle de Bourbon) came to be designed as the habitat of the plant; quite certainly it is a typical V. arctata Gleason. Though strongly resembling V. bahamensis Griseb., V. arbuscula is a good species, differing from V. bahamensis in the shape and colour of its leaves, as well as in their nerva- ture. In V. bahamensis the lateral veins are 4 — 6, ascending at oblique angles, in V. arbuscula they are only 2 — 3, issuing at nearly right angles. Vernonia complicata Griseb. Tabula nostra II, fig. 4 (habitus). Vernonia complicata Gbisebach, 1866, p. 143 ; Gleason, 1906 b, p. 189; 1913, p. 327. Pappi seta? interiores persistentes, numero 45, tenues, quam exteriores fere 7-plo longiores. Corolla 6,5 mm longa, tubo amplo, glabro, long. 5/9 corolla?, limbi laciniis apice subtus papillosis, pilis apicalibus brevissimis aut nullis. Antherse 2,3 mm longae, ligula enervi long. 1/e antherae, auriculis obtu- siusculis vel brevissime apiculatis. Hab. in Cuba: Wright n. 2790 (Bss, DC, G, orig. spec, KU, P, S); in rupibus litoralibus prope »Leeward Point » ad sinum Guantanamo, Mart. 1909. Britton n. 2225 (KU). A species of very peculiar habit due to its small, broad, curled, densely tomentose leaves and few, solitary heads. E. L. EKMAN. WEST INDIAN VERNONLE. 59 4. The allies of Vernonia fruticosa (L.) Sw. Yernonia divaricata Sw. Tabula nostra II, fig. 6 (pars inflorescentiae). Conyza arborescens Linnaeus, 1759 b, p. 406, quoad syn. Brownei et Sloanei, haud Conyza arborescens Linnaeus, 1759 a, p. 1213, quoad typum ; 1763, p. 1209, quoad syn. supra cit.; Swartz, 1791, p. 304. Vernonia divaricata Swartz, 1806, p. 1319. Vernonia arborescens Swartz, 1806, p. 1320, quoad descript. et patriam, haud quoad typ. Linnlei; Gleason, 1906 b, p. 180. Vernonia arborescens Sw. a. Swartziana Grisebach, 1861, p. 353, syn. omnibus exclusis. Vernonia permollis Gleason 1906 b, p. 181. Vernonia intonsa Gleason, 1906 b, p. 182. Vernonia albicoma Gleason, 1906 b, p. 185. Vernonia amaranthina Gleason, 1913, p. 307. Conyza fruticosa flore pallide purpureo, capitulis e lateribus ramidorum spicatim exeuntibus. Sloane, Cat. p. 124; Jam., p. 257. Eupatorium 1. Erectum hirsutum, foliis oblongis rugosis ; floribus spicatis per rainos terminales declinantes uno versu dispositis. Browne, Jam., p. 313. Pappi setae interiores subpersistentes, numero 30 — 35, crassiusculae, paullulo complanatae, albidae, exterioribus dis- tinctis circ. 8-plo longiores. Corolla fere 6 mm longa, glabra, tubo long. ya corollae, superne haud dilatato, limbi laciniis apice subtus papillosis. Antherse 1,8 mm longa?, ligula ner- vosa long. 7e antherse, auriculis apice brevissime acuminatis. Hab. in Jamaica: in herb. Linnjei; ex herb. Vaillant (P); W. Wright (S); Shakespeare (Br); Swartz (H, M, S, orig. spec. V. divaricatce et V. arborescentis sensu Swartzii); Tussac (D); Distin (K, KU); Swainson (K); prope St. Marys, Aug. 1843, Purdie (K); W. Wilson n. 238 (G, KU); Wullschlagel ii. 875 (G, M); R. C. Alexander-Prior (G); March n. 813, 1472, 1738 (G); prope Kingston ad ra- dices montium, 7. 12. 81, Lehmann n. 939 (B, BB); ad Gordon- town, alt. 450 m, 21. 1. 88, Eggers n. 3481 (BB, D, KU, M); ad Hope River, 21. 1. 88, Eggers n. 3490 (H); in Providence Road, 5. 8. 97, W. Thompson n. 6746 (KU); prope Providence, alt. 210 m, 19. 7. 1900, W. Thompson n. 7209 (KU); prope Prospect Hill, circ. 600 m alt., 13. 9. 1900, W. Thompson n. 8038 (Br, KU); in »Salt Hill Road», 13. 12. 93, Harris n. 5622 (BB, H, KU); prope Hope, 22. 12. 97, Harris, n. 60 ARKIV FOR BOTANIK. BAND 13. N:0 15. 6993 (Br, KU); prope Mona, alt. 180 m, 2. 8. 95, Campbell, n. 5870 (KU); in »Arnold Road», alt. 80 m, 25. 11. 95, Camp- bell n. 6091 (KU, dupl. orig. spec. V. intonsce Gleason); ad radices montis Long Mountain, alt. 105 m, 13. 1. 96, Campbell n. 6152 (KU, dupl. orig. spec. V. albicomce Gleason); prope Kingston in montibus humilioribus, anno 1897. O. Hansen sine num. (H, KU); Cinchona Forrest, Maj. 1903, Shreve, sine num. (KU, fragm. orig. spec. V. permollis Gleason); Vicinity of New Castle, Hardware Gap, rocky- bank, 1. 3. 08, N. L. Britton et Arthur Hollick n. 1788 (KU); coll. ignot. (P, »donne par Sir W. Hooker 1845»). — Grand Cayman: Hitchcock, fide Gleason, 1913, p. 307. As early as the year 1696 this species was introduced into the annals of Botany by Sloane in his Catalogue, and shortly after, in the year 1707, he described it accurately in his Natural History of Jamaica. Patrick Browne also gave a good description of it in Civil and Natural History of Jamaica, p. 313. I have shown while discussing the name V. arborescens (L. ) Sw. how Swartz came to apply this name to the Jamaican plant. In the same passage I have mentioned the reasons for applying the name V. arborescens to the Martini que plant. Then there arises a problem: what is to be the correct name of the Jamaican plant, the previous V . arborescens. At first I thought it must be V . intonsa Gleason, this species agreeing perfectly with the type specimen of V. arborescens Sw. How- ever, when I realized that V. arborescens sensu Swartzii, V. intonsa Gleason, V. permollis Gleason, V. albicoma Gleason and even V. divaricata Sw. all were only forms of one species, I merely had to take V. divaricata Sw. for its name. With this the problem of the synonymy of this Jamaican species had been definitively solved. It is hardly necessary to state that V. divaricata as here understood is an extremely variable species. Gleason, having abundant, new-collected material of it at his disposal, described some of its forms as new species, viz. V. permollis, V. intonsa, V. albicoma, and V. amaranthina. Of these V. intonsa is exactly the same as the type of V. arborescens Sw., V. permollis is an extraordinarily densely pubescent form, and V . albicoma a nearly glabrous one. I have not perfectly satisfied mvself about the taxonomic value of V. albicoma. E. L. EKMAN, WEST INDIAN VERNONLE. 61 the type being somewhat different from V. divaricata, above all in its elongated corollas (7 mm). It remains, however, probable that the particular habit of V. albicoma is connected with its occurrence in shade localities and, since the idea of the unity of V. divaricata Sw. as here understood seems to me very admissible, the species having no very near relative in Jamaica, I have decided to place V. albicoma under V. divaricata Sw. Besides, there are intermediate forms between V. albicoma and V. divaricata, for instance, Wilson n. 238. In the following I have tried to characterize with a few words the most prominent forms of V. divaricata. What I have said about the forms of V. arborescens (L.) Sw. and V. albicaulis Pers., might be verbatim repeated here. 1. The type of V. divaricata Sw. : Leaves nearly glabrous. Cymes widely spreading, many heads without bracts. Scales cream-coloured, inner ones not purplish, relatively long-acu- minated. Besides the type specimen in the Stockholm Her- barium, there belong here the specimens collected by Distin. A specimen collected by William Wright in the Stockholm Herbarium has shorter cymes; Hansen in 1897 has more densely pubescent leaves; Purdie in the Kew Herbarium has leafy cymes. 2. The type of V. albicoma Gleason: The most glabrous form! Inflorescence rather lax. Heads comparatively large, inner scales somewhat elongated, purplish. Corolla 7 mm in length. Campbell n. 6152. 3. The type of V. arborescens Gleason, not of Swartz: The central form, having rather tomentose leaves, shorter, arcuated, nearly leafless cymes, with somewhat aggregated heads. Inner scales purplish, shortly acuminated. To this Wilson n. 238, Eggers n. 3481 etc. 4. The type of V. arborescens Sw. : A densely tomentose form with short cymes, crowded heads, and purplish, shortly acuminated inner scales. The type specimen of V. intonsa Gleason belongs here, further Harris n. 5622. Lehmann n. 989 and Harris n. 6993 represent intermediate forms between V. arborescens Gleason and V. arborescens Sw. 5. The type of V. permollis Gleason: The most densely tomentose form with broad, rounded leaves, short, nearly bractless cymes, and very crowded heads. The scales are glabrous towards the tip, or with a tuft of hairs near it. The 62 ARKIV FOR BOTANIK. BAND 13. NIO 15. plant collected, by Swainson agrees very well with the por- tion of the type specimen seen. There also occur forms having elongated, leafy cymes, and distant heads. Such forms also may be found in other species, for instance, V. arborescens (L.) Sw. and V. fruticosa (L.) Sw. It appears very probable that they do not constitute a particular form, but are only variants of the other ones described. So, for instance, Campbell n. 5870, Thompson n. 6746, 7209, 8038. Vernonia acuminata Less. Tabula nostra II, fig. 8 (pars inflorescentise). Vernonia acuminata Lessing, 1831, p. 663; Grisebach, 1861, p. 353; Gleasox, 1913, p. 311. Vernonia divaricata De Candolle, 1836, p. 48; Gleason, 1906 b, p 185; 1913, p. 310 — non V. divaricata Sw. Cacalia acuminata Kuntze, 1891, p. 969. Pappi setae interiores persistentes, numero 30 — 35, crassi- usculae, paullulo complanatae, brunneae, quam exteriores dis- tinctae circ. 6-plo longiores. Corolla 6 — 6,5 longa, tubo graeili, glabro, subito in limbum abeunte, long. 3/s corolla?, ' limbi laciniis apice subtus papillosis. Antherae 1,7 — 1,8 mm longae, ligula apice contracta, long. Ye antherae, nervo perspicuo praedita, auriculis brevissime acuminatis. Hab. in Jamaica: Shakespeare (Br); Swartz (H, Prodr, sub nom. V. divaricate Sw., S, orig. spec. V. acuminatce Less., U, herb. Thunb.); Macfadyen n. 28 (K); Distin (K); Wullschlagel n. 876 (M); prope Moneague, Dec. 1849, R. C. Alexander-Prior sine num. (G, KU); Hart n. 665 (KU); inter Claremont et Moneague, 11. 7. 02, Fawcett n. 8405 (KU); in monte Catherine Peak, alt. 1350 m, 23. 1. 88, Eg- gers n. 3592 (KU); prope Kendal, alt. 450 m, 20. 11. 01, Harris n. 8205 (KU); in montibus prope Kingston, anno 1897, O. Hansen sine num. (H. KU). Though he had abundant material of this species at hand, Swartz did not describe it. However, he probably had this plant in mind when he wrote, Fl. Ind. occ. p. 1321: »Provenit et alia in insulis caribaeis hujus (V. arborescentis Sw.) forte varietas, foliis minoribus ovatis obtusis 1. acutis scabriuscu- lis, oculo armato hispidulis, atomis fusco-rubris nitentibus E. L. EKMAN, WEST INDIAN VERNONIiE. 63 immersis subtus adspersis, ramulis floriferis erectiusculis subcorymbosis nee divaricatis floribus confertioribus ». Spe- cimens of it have been early distributed as V. divaricata Sw. var., and so De Candolle and others were caused to take it as the genuine V. divaricata. Lessing, however, who applied this name to our V. arborescens, the Martinique plant, realized that the variety of Swartz was a new species, and described it as V. acuminata Less. The wrong idea, however, of this species being the genuine V. divaricata Sw. continued to flourish. As recently as in 1906 Gleason cites V. acuminata Less, as a synonym to V. divaricata Sw. In his »Studies» he states the two species to be different, but applies, in fact, the name V. divaricata to the genuine V. acuminata, and uses the latter name for a somewhat many-f lowered variety of the same species! I have not seen the specimen cited to his V. acuminata, Wright n. 20, which is said to have 18-f lowered heads, those of V. divaricata sensu Gleason having only 11 — 13 flowers (com- pare with this the statement of Lessing in the original descrip- tion of V. acuminata: »Capitula 15-flora», which is correct., the heads being 14 — 15-f lowered). But Harris n. 8205 in the Krug-Urban Herbarium, having 16 — 17-flowered heads, may agree with Gleason's V. acuminata, as it also agrees with typical V. acuminata in all respects, save in the number of flowers. Hence I conclude that, in spite of all, the V. diva- ricata and V. acuminata of Gleason are only one species, namely V. acuminata Less. As I have shown in another passage, V. divaricata Sw. is the species called by Gleason V. arborescens. The species varies but little, and is always easily recognized by its nearly glabrous leaves, acuminated towards both ends, by its pale brown heads in leafy cymes, and by its brown pappus. I had wondered long about the statement made bjr Grisebach in his Flora: »Pappus straw-coloured », until in the Grisebach Herbarium I saw a specimen of V. acuminata having, in fact, just that colour of the pappus. Otherwise it does not differ fom the genuine V. acuminata; also the struc- ture of the flowers is throughout the same. 64 ARKIV FOR BOTANIK. BAND 13. N.O 15. Vernonia expansa Gleason. Tabula nostra III, fig. 4 (inflorescentia). Vernonia expansa Gleason, 1906 b, p. 186; 1913, p. 311. Pappi setae interiores numero 30, ceteris ut in praeced. Corolla 5 mm longa, tubo satis amplo, glabro, in limbum sensim abeunte, cet. praeced.; limbi laciniis firmioribus. Antherae 1,8 mm longae, ligula sensim attenuata, auriculis plane obtusis. Hab. in Jamaica: prope Iroy, alt. 600 m, 6. 12. 04, Harris n. 8796 (KU, dupl. orig. spec). This species is very closely allied to V. acuminata Less., and may prove to be only a variety of this. The distinction given by Gleason as to the different shape of the leaves scarcely appears sufficient to warrant the separation. There are, in fact, specimens of genuine V. acuminata Less., Hansen in 1897, having the leaves as shortly attenuate at the base as those of V. expansa. However, V. expansa is characterized rather well by its small, greenish, somewhat crowded heads, and by its small flowers with wide tube and firm clefts. The material at hand is too scanty to give a reliable account of the taxonomic status of the plant. Vernonia pluyialis Gleason. Tabula nostra III, fig. 8 (inflorescentia). Frutex fere metralis, diffusus, ramosus. Rami satis crassi, inferne teretes, striati, superne angulati, subcom- pressi, hornotini appresse pubescentes, pilis brevissimis, sor- dide albidis; internodia saepissime brevia. Folia patentia, subsessilia vel brevissime petiolata, petiolo 1 — 2 mm longo, tenuiter pubescente; laminae rhomboideo-ellipticae, 3,5 cm longae, 1,6 cm latae, vel lanceolato-rhomboideae, 5,7 cm longae, 1,5 cm latae, apice acutae vel subacuminatae, basi attenuatae — rotundatae, sat remote glanduloso-denticulatae, firmae, planae, margine subrevolutae, olivaceo-virides, subtus pallida?, laeves, utraque facie tenuissime puberulae, glanduloso-punctatae; costa nervisque utriusque lateris numero 5 — 6, utraque facie ex- sculptis, subtus magis perspicuis. Inflorescentia cymoso- -scorpoidea, contracta, ovato-pyramidalis, ad 6 cm longa, E. L. BKMAN, WEST INDIAN VERNONIiE. 65 5 cm lata, cymis erecto-patulis, inferioribus ad 4 cm lon- gis, angulato-complanatis, basi longe nudis, apice calathidia ge- rentibus valde congesta, numero saepe 5 — 6 ; bracteis foliaceis calathidiis a?quilongis vel iis brevioribus. Calathidia sessilia, quam maxime conferta, 7 — 9-flora (sec. cl. Gleason 5 — 8- -flora). Involucrum subcylindricum, 6 mm longum, 3 mm latum, basi rotundato-attenuatum, squamis 5 — 6-seriatis orthostichiis ssepe conspicuis, exterioribus lanceolato-deltoi- deis, brevibus, carinatis, minute mucronulatis, intermediis lanceolatis, brevissime acuminatis, valde carinatis, intimis ligulato-lanceolatis, obtusiusculis, omnibus firmulis, convexu- lis, sordide brunneolis, interne glabris, nitentibus, ad apicem tenuissime puberulis. Pappi seta3 interiores subpersistentes, numero 40, firmulse, paullulo complanatse, brunnese, 4,5 mm longse, quam exteriores lineares, ingequilongse, perspicuae circ. 6-plo longiores. Corolla 6 — 6,5 mm longa, albida (Rehder), tubo angusto, C3Tlindrico, superne glanduloso, in limbum subito abeunte, long, vix xl2 corollse, limbi laciniis firmulis, patulis-recurvis, apice subtus papillosis. Anthera3 2,2 mm longse, ligula long. x/6 antheraB, satis subito contracta, nervo medio conspicuo, auriculis elongatis, obtusis. Stylus longe exsertus. Aclia3nium (immaturum) 1,8 mm longum, dense appresso-sericeum. Baccharis myrsinites Grisebach, 1861, p. 366, p. p. — non Pers. Vemonia pluvialis Gleason, 1913, p. 312. Vernonia proclivis Gleason, 1913, p. 312. Vemonia reducta Gleason, 1913, p. 313. Hab. in Jamaica: Macnab (KU, fragm. spec. herb. Kew); Purdie (KU, fragm. spec. herb. Kew); Blue Moun- tain Peak, Sept. 1885, Morris n. 2120 (KU); Hart n. 1070 (KU); in summo monte »Blue Mountain Peak», 2190 m. alt., 11. 2. 03, A. Rehder sine num. (KU); Morcis Gap, in silvis humidis, alt. 1500 m, 12. 7. 03, Geo. Nichols n. 20 (K). A species of very peculiar habit due chiefly to the den- sely clustered heads on long naked peduncles, and to the sub- -cylindrical involucres with their peculiarly imbricated scales. In fact, it resembles somewhat a Baccharis, to which genus it was referred by Grisebach. Prof. Urban, who was the first to realize the Vernonian nature of the plant, described it in manuscript as a particular Arkiv for botanik. Band 13. N:o 15. 5 66 ARKIV FOR BOTANIK. BAND 13. N:0 15. species, V. hrachypoda Urb., but afterwards published it as V. acuminata (Urban, 1903, p. 391), apparently consider- ing it to be a highland form of this species. Gleason in his »Re vision » is of the same opinion: »Lessing's V. acuminata is described with aggregated heads, resembling Nichols 20 and 120, a character possibly due merely to the habitat, since the latter grew in an altitude of 5000 — 7400 feet». In his »Studies», however, he does the species abundant justice, describing it under three different names, V. pluvialis, V. proclivis and V. reducta. Though I have not seen authentic specimens of these three species, I do not doubt that they are only forms of one species. The differences given by Gleason are merely inconsiderable ones, based on the shape of the leaves and the number of flowers in the heads. V. pluvialis is said to have (5 — )8-flowered heads and oblong-ovate to sub-rhomboid leaves, 3 — 5 cm long, 1,1 — 1,9 cm wide, V. proclivis has 8-flowered heads and elliptic-oblong leaves, which are 6 — 8 cm long and 2—3 cm wide, V. reducta is described as having 5-f lowered heads and narrowly elliptic - -oblong leaves, 4 — 4,5 cm long by 1,2 — 1,6 cm wide. The specimens examined b}^ me, clearly belonging to one single, although rather variable, species, show the following data as to the characters in question: IiEhder 9 fl., leaves 3,5 cm long, 1,65 cm wide. Hart 7 » » 3,2 » » 1,3 » » Nichols 120 9 » » 3,3 » » 1,3 » » Morris 2120 7 » » 3,4 » » 0,9 » » Macnab — » » 5,7 » » 1,5 » » It will be seen that there is considerable variation as to the width of the leaves, their shape varjdng accordingly from oblong-rhomboid with very shortly attenuate base to rhom- boid-lanceolate with long-attenuate base. As to the number of flowers I never found heads with only five, though I examined Nichols 120 cited by Gleason under V. pluvialis. The three species of Gleason were all collected in the higher mountains of Jamaica, partly in the same localities — one of the numbers cited, Britton n. 3851, is quoted both under V. pluvialis and under V. proclivis. It seems, therefore, highly probable that the three specie-, in fact, constitute one single, though very variable, species. E. L. EKMiN, WEST INDIAN VERNONIJE. 67 Vernonia rigida Sw. Tabula nostra III, fig. 1 (inflorescentia). Conyza rigida Swartz, 1788, p. 113. Vernonia rigida Swartz, 1806, p. 1322; De Candolle, 1836, p. 49; Grisebach, 1861, p. 354, syn. omnibus exclusis. Vernonia fruticosa Gleason, 1906 b, p. 182 — haud V. fruticosa (L.)Sw. Pappi setae interiores persistentes, numero 25 — 30, firmae, crassiusculae, paullulo complanatae, brunneae, quam exteriores distinctae lanceolato-spathulatae circ. 6-plo longiores. Flores desunt. Achaenia tenuiter sericeo-pubescentia. Hab. in Jamaica boreali: in montibus calcareis, petrosis rarius, Swartz (H, K, M, Prodr, S, orig. spec, U, herb. Thunb.). The species has, as far as I know, only been collected by Swartz. The problem of the affinities of this species has caused botanists much trouble. Swartz himself, in describing it, supposed it to be closely allied to V. fruticosa (L.) Sw., which, however, was not correct. Grisebach, in his Flora, referred to it not only V. fruticosa, but also the species of the subsection Sagrceana3, together with V. emarginata Wikstr. ! Gleason at first followed Grisebach, identifying in his »Re- vision » V. rigida and V. fruticosa, and placing them among the Sagrazanai. Recently he has found that V. rigida is another species than V. fruticosa, and that it is not to be placed to his SagrcEanw. However, he does not mention anything about the affinities of the species, authentic specimens of which he has not seen. Having abundant material at hand of the species in ques- tion, it was an easy matter for me to ascertain that V . rigida Sw. is quite different both from V. fruticosa, V. emarginata, and Vernonice Sagrosanai. Its relationships, however, still remained questionable, until in an examination of the unnamed Compositae of the Stockholm Herbarium I lighted upon a Ver- nonia from the West Indies collected by Swartz, and presented by him to the Swedish botanist Montin, which Vernonia had some characters of V. rigida and some of V. acuminata Less. Through this fortunate incident I was enabled to state that the nearest ally of V. rigida is V. acuminata Less. A 68 ARKIV FOR BOTANIK. BAND 13. NlO 15. few words about the characteristics of the plants in question may be of interest. V. rigida, as described by Swartz, is a woody winding shrub having flexuose, densely tomentose branches, rounded, rigidly coriaceous leaves, being 3 — 4,5 cm long by 2,5 — 3 cm wide, somewhat shining above, glandulose and finely pubescent beneath, especially on the veins. The tomentum is often more copious towards the margins of the leaves. The cymes are leafy, freely flexuose, the heads distant, sometimes two or three in the axils. The involucres are elongated, up to 15 mm high, infundibular, long-attenuate at base; the scales are imbricated in numerous series, the outer ones deltoid, carinate, nearly glabrous, erect-appressed, the inner ones elongated, spreading, the dry and open involucres thus re- sembling stars. The specimen presented to Montin agrees in leaf -cha- racters perfectly with V. rigida; may be the leaves are a little thinner, and more narrowed, 4 cm long by 2,2 cm wide. It has, further, exactly the same pappus and fruit. However, the involucres are very differently shaped. They are only 6 — 7 mm high, broadly obconic, shortly attenuate at base, the scales are shorter and more densely pubescent. Probably this is the normal form of V. rigida, the type specimens be- longing to a form with abnormally elongated involucres. Such formal imbricatissimaz occur not infrequently in South American species, for instance, in V. oligactoides Less., V. nitidula Less., and V. squamulosa Hook. Arn. yernonia fruticosa (L.) Sw. Tabula nostra III, fig. 2, 3 (habitus). Conyza fruticosa Linne, 1763,*p. 1209. ij Yernonia fruticosa Swartz, 1806, p. 1323; De Candolle, 1836, p. 65: Gleason, 1913, p. 315. Conyza frutescens Cydoinoz folio. Plum. Cat. p. 9 ; ed, Burm., tab. 95, fig. 1. Eupatorium frutescens, hederce terrestris folio, flore purpurascente. Plum. Cat. p. 9(?). Frutex vel suffrutex repens (Miguel Fuertes), aut erectus circ. metralis (W. Buch), valde ramosus, ramis saepe elongatis, divaricatis. Caulis lignosus, teres, strictus, adult us E. L. EKMAN, WEST INDIAN VERNONLE. 69 glabcr, novellus breviter albo-tomentosus (Miguel Fuertes n. 655), aut indumento nigricante (Picarda n. 884), aequaliter foliosus. Folia spiraliter enata, patentia, breviter seel dis- tincte petiolata, petiolo tomentoso, 1,5 mm longo; laminae ovatae, maxima3 2,3 cm longae, 1,3 cm latae, saepe multo minores, 1,1 cm longae, 0,8 cm latae, apice obtusae, truncatae vel etiam emarginatae, basi cordatae, membranaceae, integerrimae, ob- scure rugosae vel subplanae, margine saepe undulatae, vix revolutae, discolores, supra atro-olivaceo-virides, pubescentes, pilis e tuberculis enatis, subtus pilis longis, intricatis albo- -floccoso-tomentosae, haud sericeae, glandulosae, costa nervisque utriusque lateris numero 5 — 6, supra impressis, subtus ex- sculptis. Inflorescentia cymoso-scorpioidea, cymis aut valde elongatis, cauliformibus, leviter flexuosis, inflorescentiam dis- tinctam vix formantibus, aut brevioribus calathidiis magis approximatis, divaricatis, inflorescentiam formantibus fere 1 dm latam, subplanam; bracteis foliaceis calathidiis niul- toties longiores aut in exemplis aliis iis aequilongis. Cala- thidia axillaria, sessilia, remota, 18 — 21-flora. Involucrum campanulatum, 5 — 6 mm altum, 5 — 6 mm latum, squamis circ. 3-seriatis, satis laxis, extimis in speciminibus a cl. Mi- guel Fuertes lectis deltoideo-subulatis, conspicue carina- tis, deflexis, interioribus erectis, breviter acutatis, obscure carinatis, extimis in a speciminibus ceteris anguste deltoideis, subappressis, obscurius carinatis, intimis lanceolatis, brevissime acuminatis, omnibus firmulis, subglabris, brunneo-nitentibus. Pappi setae interiores subpersistentes, numero 25 — 30, crassius- culae, paullulo complanatae, 4 mm longae, albae, quam exteriores distinctae lanceolato-spathulatae circ. 7-plo longiores. Co- rolla purpureo-violacea, 5 mm longa, tubo amplo, sensim dilatato, glabro, long. 3/s corollae, limbi laciniis apice subtus papillosis. Antherae 2 mm longae, ligula enervi, brevi, fere Vs long, antherae, auriculis obtusis. Stylus corollae fere aequi- longus. Achaenium crasse cylindricum, 1,5 mm longum, glandulosum, sericeo-pilosum, pilis brevibus appressis, costis perspicuis. Hab. in Hispaniola: sine loc. ace, Picarda n. 146 (KU); prope Petionville, anno 1889, Picarda n. 884 (KU); Darant (Terreneuve) prope Gonai'ves, locis siccis, alt. 600 m, Nov. 1900, Buch n. 522 (KU); in locis apertis montium, alt. 350 m, Nov. 1910, Miguel Fuertes n. 655 (KU); in 70 ARKIV FOR BOTANIK. BAND 13. N:0 15. prov. Barahona, anno 1911, Miguel Fuertes sine num. (KU). Another Vernonia, the histor}^ of which begins with Plumier and Surian. Like V. arborescens (L.) Sw., it was figured in Plum. ed. Burm., and on this figure Linnjeus based his Conyza fruticosa. Since there was no specimen of it preserved, its identity remained mysterious to botanists, and the name V. fruticosa (L.) Sw. has been applied to many West Indian species, recently to V. rigida Sw. (Gleason, 1906 b, p. 182), Swartz himself comparing this species with V. fruticosa. It was not until Picarda recollected it about the year 1889 that Prof. Urban was able to identify the species, as shown by the remarks on the labels of the specimens in the Krug- -Urban Herbarium. In fact, there can be no doubt that this identification is correct. Especially Miguel Fuertes n. 655 agrees perfectly with the figure in Plum. ed. Burm. There is moreover not a word in the description to that figure, Plum. ed. Burm. p. 83, that does not apply to our plant. Its phrase name »Conyza frutescens, Cydonia? folio » is singularly apposite, the leaves resembling throughout those of Cy- donia vulgaris L., their pubescense being nearly the same. Burmann's particular phrase name (or diagnosis?) of the plant: »Conyza foliis ovatis, caule fruticoso, floribus axilla- ribus solitariis, sessilibus» and his description: »Planta fruti- cosa, ramulis flexuosis, a folio ad folium eleganter inclinatis. Folia alterna, obtusa, undulata, venosa. Flores solitarii in quavis ala sessiles» are apparently made from Plumier's figure, giving no character beyond this, and consequently agre- eing perfectly with our plant. V. fruticosa is hitherto only found in the south-western part of Hispaniola, where it appears to be rather frequent. Its nearest allies are the Vernoni* of Jamaica, as also some species of eastern Cuba, especially V. calophylla Gleason. Though the material at hand is not abundant, it indicates a certain variability of the species. Two fairly distinct forms can be distinguished. One is represented by the plants collected by Miguel Fuertes. It has long, leafy cymes not forming a distinct inflorescence; the scales of the involucres are acute and elongated. The second form has smaller leaves, shorter cymes with more approximated heads and bluntish scales. It appears to be the common one. Possibly the two E. L. EKMAN, WEST INDIAN VERNONLE. 71 plants, though not separated in leaf-characters, constitute each one species. Additional material is necessary in order to decide this question. Yernoiiia parvuliceps Ekman n. sp. Tabula nostra III, fig. 9 (inflorescentia). Suffrutex vel herba perennis, erecta, superne ramosa. Caulis (cujus pars superior tantum adest) teres, diam. 3 mm, dense breviterque cremeo-tomentosus, pilis intricatis, tor- quatis, a?qualiter foliosus. Folia spiraliter enata, patentia, distincte petiolata, petiolo ad 4 mm longo, tomentoso; lamina? ovato-lanceolata?, ad 5 mm longa?, 1,8 cm lata?, basi rotundata? vel cuneato-rotundata?, apice acuminata?, membranacea?, leviter rugosa?, margine plana? vel novella? revulata?, inte- gerrima?, discolores, supra olivaceo-virides, glabra?, la?vius- cula?, subtus dense breviterque cremeo-tomentosa?, costa nervisque utriusque lateris numero 7 — 8, supra impressis, subtus pulchre exsculptis. Im°lorescentia C3^moso-scorpioidea, subcontracta, cymis patulis, arcuatis, 4 — 5 cm longis, circ. 10-cephalis; bracteis foliaceis, inferioribus calathidiis 3— 4-plo longioribus, superioribus eadem fere a?quantibus. Ca- lathidia sessilia, conferta, parva, circ. 21 -flora. Involucrum campanulatum, basi rotundatum, 3 mm altum, 3 — 4 mm latum, squamis 4-seriatis, arete imbricatis, erecto-appressis, exterioribus brevibus, deltoideis — lanceolato-deltoideis, cari- natis mucronulatisque, interioribus lanceolato-linearibus, bre- viter acuminatis, vix carinatis, omnibus stramineo-brunneis, arachnoideo-pubescentibus. Pappi seta? interiores subcaduca?, numero fere 25, crassiuscula?, alba?, 3 mm longa?, exteriori- bus perspicuis, late linearibus fere 5-plo longiores. Corolla purpurascens, 5 mm longa, tubo angusto, cylindrico, glabro, long. 3/s corolla?, limbi laciniis apice subtus papillosis. An- thera? 1,3 mm longa?, ligula nervosa long. Ye anthera?, auri- culis obtusis. Stylus corollam a?quans. Acha?nium (imma- turum) 1 mm longum, dense sericeum, pilis brevibus, erectis, appressis, costis glandulisque vix perspicuis. Hab. in Cuba: Wright n. 2788 (Bss, DC, G, K, orig. spec, KU, P). The specimens seen of this very striking species are quite uniform, and appear to be collected in the same locality. 72 ARKIV FOR BOTANIK. BAND 13. N:0 15. It is puzzling that Gleason did not see this species well represented in herbaria, or that it has not been collected by the active Cuban explorers. In fact, it is one of the easiest recognized species of the Arborescentes-gromp and, once seen, impossible to overlook. Among the species known to me it has no near allies. According to the description, however, V. vicina Gleason approaches somewhat to our plant, but is distin- guished by its larger involucres, which are said to be »about 5 — 6 mm high». Probably there are other differences, but it is not easy to obtain an exact idea of V. vicina from Gleason 's description. Ternoiiia vicina Gleason. Vernonia vicina Gleason, 1913, p. 317. Hab. in Cuba: Oriente, Camp la Gloria, prope Sierra Moa, 24—30.12.10, Shaeer n. 8202. I have seen no specimens of this species. From the de- scription it comes near to V. parvuliceps, but has broader leaves and larger heads. Vernonia calophylla Gleason. Vernonia calophylla Gleasox, 1913, p. 317. Hab. in Cuba: Oriente, Camp la Gloria, prope Sierra Moa, 24—30.12.10, Shafer n. 8102. Not seen. The species seems to be a fairly well-marked one, related to the preceding, but differing in its obtuse, sub- -rotund leaves. Vernonia yunqnensis Gleason. Tabula nostra III, fig. 6 (folium et calathidium). Vei-nonia yunquensis Gleasox, 1906 b, p. 191 ; 1913, p. 327. Pappi setae interiores subdeciduae, numero fere 25, crassi- usculae, paullulo complanatae, albidae, quam exteriores dis- tinctae circ. 5-plo longiores. Corolla glabra, tubo ut videatur long. l/2 corolla?. Antherae 2,4 mm longae, ligula long. x/s antherae, auriculis obtusis. Achaenia dense Mtieeo-pubescentia. E. L. BKMAN, WEST INDIAN VERNONIJE. 73 Hab. in Cuba: Baracoa, in monte El Yunque, ad rupes, Mart. 1903, L. M. Underwood et F. S. Earle n. 661 (KU, fragm. orig. spec). A very curious species, differing from other Arborescentes in the characters of the inflorescence, which is described as follows: »heads about 5, sessile or nearly so in a terminal capitate cluster ». I have only seen a small fragment of the type specimen, and I therefore cannot ascertain the relation- ships of the species. As to the floral characters, however, it agrees exactly with the species allied to V. fruticosa. Ternonia leptoclada Sch.-Bip. Tabula nostra III, fig. 5 (cyma). Vernonia leptoclada Schultz-Bjpontinus, 1863, p. 133. Vemonia Wrightii Grisebach, 1866, p. 14-i — non Sch.-Bip. Vernonia gnaphalifolia Gleason, 1906 b, p. 178 — non A. Rich. Vernonia neglecta Gleason, 1913, p. 318. Pappi setae interiores subcaducse, numero 20 — 25, crassius- culae, paullulo complanatae, albidse, quam exteriores distinc- tissimge circ. 5-plo longiores. Corolla 8 mm longa, glabra, tubo 2/3 corollse long., glandulo uno alterove instructo, limbi lacimis subtus papillosis apice pilo uno alterove interdum instructis. Antherae 1,7 mm longa?, ligula fere enervi 1/6 long, antherse, auriculis apice obtusiusculis. Hab. in Cuba: Wright n. 1309 (Bss, CD, orig. spec, sub num. falso 285, D, DC, G, K, KU, P); Loma del Jagiiey, Mart. 1889, Eggers n. 4982 (K, KU, P). Through the interchange of labels already mentioned, the name of this species has been applied to the plant called V. Wrightii Sch.-Bip. in this paper. The species here dealt with has been called, vice versa, V. Wrightii. Gleason, in his »Revision», identified it, erroneously, with V. gnapha- lifolia A. Rich., and having observed his mistake, in his »Studies» named it V. neglecta. The species is easily recognized by its bullate, densely tomentose leaves and by the elongated cymes, bearing many distant, small heads. It is related to V. pineticola Gleason, as well as to V. calida Gleason and V. semitalis Gleason. The material at hand is fairly uniform, except for the shape of the leaves, which vary from lanceolate, acute, 4,5 cm long, 74 ARKIV FOR BOTANIK. BAND 13. NIO 15. 1 cm wide (Eggers n. 4982), or lanceolate, obtuse, 7,5 cm long, 2,4 cm wide (Wright n. 1309 p. p.) to ovate, acuminate, 3,8 cm long by 1,6 cm wide (Wright n. 1309 p. p.). Vernonia pineticola Gleason. Tabula nostra III, fig. 7 (calathidium et folia). Vernonia pineticola Gleason, 1906 b, p. 176. Pappi setse interiores cras-siusculae, paullulo complanatse, albidse, quam exteriores distinctae circ. 6-plo longiores. Flores in specimini mihi commisso desunt. Hab. in Cuba: Baracoa, in pinetis, Mart. 1903, L. M. Underwood et F. S. Earle n. 1341 (KU, fragm. orig. spec). Only a small fragment of this species has been seen. It is closely related to V. leptoclada Sch.-Bip., but differs abundantly in the larger involucres with elongated, spreading, sub -filiform scales. Ternonia calida Gleason. Vernonia calida Gleason, 1913, p. 318. Hab. in Cuba: Oriente, inter Sabanilla et Yamuri Arriba, 30. 1, 1. 2. 11, Shafer n. 8408. Not seen. From the description it appears to be a good species, chiefly distinguished from V. leptoclada by its con- tracted inflorescence. Vernonia semitalis Gleason. Vernonia semitalis Gleason, 1913, p. 319. Hab. in Cuba: Oriente, ad viam inter Rio Yamaniguey et Camp Toa, 22—26. 2. 10, Shafer n. 4170. This species seems to be related to the preceding one, differing above all in its nearly glabrous involucre. I did not see any specimen of this. E. L. EKMAN, WEST INDIAN VERNONLE. 75 5. The allies of V. membranacea Griseb. Yernonia membranacea Griseb. Tabula nostra IV, fig. 3, (inflorescentia). Vernonia membranacea Grisebach, 1SG6, p. 144; Gleason, 1906 b, p. 180. Pappi setae interiores persistentes, numero 40, tenues, filiformes, pallide brunneae, quam exteriores minutissimae circ. 14-plo longiores. Corolla 5,8 mm longa, tubo long. x 2 corollae, sensim dilatato, glanclulis nonnullis instructo, limbi laciniis apice subtus papillosis. Antherae 1,6 mm, longse, ligula \G long, antherae, nervo medio conspicuo, auriculis acutis. Hab. in Cuba: Wright n. 490 (G, orig. spec., KU, sine num.); prope Bahia Honda, prov. Pinar del Rio, 18 — 19. 12. 10, Wilson n. 9426 (KU). A well-marked species, which seems to be very rare. As Gleason remarks, its leaves greatly resemble those of V. albicaulis Pers. But the scales of the involucres are very differ- ent, broad, greenish-brown with stramineous margin and nearly glabrous. Structurally, it comes very near to V. gnaphali folia A. Rich., and is, in fact, closely related to that species, but only slightly so to V. albicaulis. Yernonia desiliens Gleason. Vernonia desiliens Gleason, 1913, p. 316. Hab. in Cuba: Oriente, Arroyo del Medio, ad rupes prope ripam rivuli, alt. 450—550 m, 20. 1. 10, Shafer n. 3232. I have seen no specimens of this species. From the de- scription it seems to be a good one, coming near to V. membra- nacea, from which it differs in having the leaves »closely gray- tomentose» beneath. As to the characters of the pappus it agrees very well with V. membranacea: »pappus light brown, the inner series 6 mm long, minutely barbellate, the outer series 0,5 mm. long, somewhat paler* (Gleason, 1913, p. 316). 76 ARKIV FOR BOTANIK. BAND 13. N:0 15. Vernouia crassinervia Wright ap. Gleason. Tabula nostra IV, fig. 2 (habitus). Vernonia arborescens Sw. var. divaricata Grisebach, 1866, p. 144 — non aliis locis. Vernonia crassinervia Weight ap. Gleason, 1906, p. 180. Pappi setae interiores subpersistentes, numero 30, tenues, filiformes, albidae, quam exteriores minutse circ. 12-plo lon- giores. Flores in speciminibus visis desunt. Hab. in Cuba: Wright n. 2787 (Bss, G, K, P). This imperfectly known species is very closely related to V. gnaphalijolia A. Rich., chiefly differing in its very low sta- ture and reduced inflorescence. In my notes from the Paris' Herbarium I have written as to V. crassinervia: »Strongly re- sembling V. gnaphah folia. Perhaps differing constantly in the acute leaves. Veins not prominent beneath. Scales elong- ated, pappus whiter. » Vernonia gnaphalifolia A. Rich. ap. Sagra. Tabula nostra IV, fig. I (inflorescentia). Vernonia gnaphalifolia A. Rich. ap. Sagra, 1850, p. 34; Gleason, 1913, p. 310. Vernonia sublanata Gleason, 1906 b, p. 177. Vernonia angustata Gleason, 1913, p. 309. Pappi setae interiores persistentes, numero 40, tenues, filiformes, albidae, quam exteriores indistinctae, lineares circ. 10-plo longiores. Corolla 5,5 mm longa, glabra, tubo sensim dilatato long. 3/4 corollae, extus glandulis nonnullis instructo, limbi laciniis apice dorso papillosis. Antherae 1,4 mm longae, ligula subenervi long. */5 antherae, auriculis breviter acutatis. Hab. in Cuba: prope Havana, ex herb. Pavon (Bss); ex herb. Richard, leg. Ramon de la Sagra, sine num. (KU, P, orig. spec); Ramon de la Sagra n. 42 (Prodr), n. 53 (Prodr), n. 59 (P); prope Havana, in campis siccis, Dec. 1890, Morales n. 12 (KU); ad Veclado prope Havana, Oct. 1S90, Torral- bas n. 40 (KU); Madruga, in saxosis, 28. 3. 03, N. L. Britton, E. G. Britton et J. A. Shafer n. 784 (KU, dupl. orig. spec. V. sublanatce Gleason); prov. Havana, Guanabacoa, Loma de la Yota, locis humidiusculis, Jan. 1912, Brother Leon et P. Wilson n. 11635 (KU); prov. Santa Clara, ad rupes, 8 — 9. 4. 12, N. L. Britton et J. F. Cowell n. 13322 (KU). E. L. EKMAN, WEST INDIAN VERNONIiE. 77 Having attributed the name V. gnaphalifolia to the plant here called V. leptoclada Sch.-Bip., Gleason could find no name for the genuine V. gnaphalifolia, and consequently in his »Revision» described it as a new species, V. sublanata Gleason. Prof. Urban, however, called his attention to the fact that V. sablanata was identical with V. gnaphali folia, and enabled Gleason to correct his mistake himself. The species is rather variable, both as to the shape of the leaves and the density of their pubescence. In the »Revision » Gleason described a variety angustata of the species, having narrowly oblong leaves. This variety he raises to specific rank in his »Studies ». I remember having seen elsewhere this var- iety, though I find no notes about it in my lists. It is, how- ever, not probable that it is a species. The shape of the leaves varies strongly in all Vernonise. If species were to be based only upon leaf-characters, hundreds of species could be described on the West Indian material. Evidently there must be other characters combined with the different shape and nervature of the leaves to warrant the rank of a parti- cular species. I imagine that Gleason will fail to give these for his V. angustata. 6. The allies of Vernonia sericea L. C. Rich. Ternonia commutata Ekman n. sp. Tabula nostra IV, fig. 4 (habitus). Suffrutex, certe 3,5 dm altus. Caulis (cujus pars superior tantum adest) inferne simplex, superne ramosus, lignosus, 2 mm. diam., teres, striatus, tenuiter appresso-pubescens, inferne nudus, superne foliatus. Folia alterna, patentia, subsessilia vel breviter petiolata, petiolo 1 — 2 mm longo; laminae lanceo- lato-lineares, 4,5 cm longae, 0,4 — 0,5 cmlatse, basi sensim attenu- atae, apice acutse, membranacese, integerrimse, margine revo- lutee, supra rugulosse, tuberculis ssepe piliferis scabrse, subtus in nervis appresse strigoso-pubescentes, in areolis crebriter aureo-glanduloso-punctatse, costa nervisque utriusque lateris numero 12 — 15, supra leviter impressis, subtus exsculptis. Inflorescentia cymoso-scorpioidea, ovalis, satis laxa, diffusa, cymis elongatis, divaricatis, maximis 1 dm longis, 4 — 8-ce- 78 ARKIV FOR B0TANIK. BAND 13. NIO 15. phalis, calathicliis satis remotis, bracteis suffultis foliaceis, iisdem duplo longioribus. Calathidia sessilia, axillaria, soli- taria, circ. 18-flora. Involucrum late infundibuliforme, 5 mm altum, 7 — 8 mm latum, squamis circ. 4-seriatis, satis laxis, achsenia non plane tegentibus, extimis brevibus, deltoideis, brevissime mucronulatis, intermediis lanceolato-deltoideis, in- timis lineari-deltoideis, omnibus erectis, appressis, acutis, subglabris, brunneo-nitidulis. Receptaculum planum, nudum. Pappi setse interiores persistentes, numero circ. 35, tenues, filiformes, 4 mm longse, albse, quam exteriores oblongo-ligulatse distinctae 8-plo longiores. Flores in speciminibus visis fructus tempore collectis desunt. Achaenia cylindrica, ad 2 mm longa, sericeo-pubescentia, pilis erectis, subappressis, glandulis paucis, aureis munita, costis perspicuis. Vernonia remotijlora Grisebach, 1862, p. 511 ; 1866, p. 144 {forma angustifolia) — non L. C. Rich. Vernonia araripensis Gleason, 1906 b, p- 181 : 1913, p. 308 — non Gaedn. Hab. in Cuba: Wright n. 286 (Bss, D, DC, G, K, KU, orig. spec). The new species is related to V. angustissima Wright ap. Ekman and V. stenophylla Less., but at once distinguished by its broader leaves, which are green beneath, not sericeous as in those species. Collected already by Wright, the species was identified by Grisebach with V. remotijlora L. C. Rich., a very different, annual plant, never found in the West Indies. Gleason refers it to V. araripensis Gardn. Perhaps he has arrived at this by identifying the species with V. tricholepis Grisebach (not DC!) 1861, p. 35-1, which is said by Baker, 1873, p. 69, to be identical with V. araripensis. In fact, Gleason cites V. tri- cholepis Griseb. as a S3rnonym to his V. araripensis. However, V. tricholepis Griseb. is not our plant, nor is it identical with V. araripensis Gardn., which still remains as a plant of Brazil. Vernonia angustissima Wright ap. Ekman n. sp. Tabula nostra VI, fig- 4 (habitus). Herba perennis, suffruticosa. Caulis (cujus pars superior tantum adesl ) ultra 4 dm altus, inferne lignosus, diam. 2.5 mm, E. L. EKMAN, WEST INDIAN VERNONIJE. 79 simplex, teres, striatulus, inferne subglaber, superne tenuiter appresso-tomentosus, cortice obscure fusco, densiuscule foliatus. Folia alterna, patentia, sessilia; laminae lineares, ad 4 cm longae, 1,5 — 3 mm lata?, basi attenuate, apice breviter acuminata?, membranaceae, integerrimae, margine revolutae, discolores, supra atro-virides, tuberculato-scabrae, pubescentes, subtus pilis densis, appressis sericeo-tomentosae, costa supra immersa, subtus exsculpta, nervis obscuris. Inflorescentia cymoso- -scorpioidea, ambitu late ovata, diffusula, sublaxa, cymis cliva- ricatis, leviter arcuatis, simplicibus, ad 1 dm longis, 3 — 10- cephalis, calathidiis spatio 1 — 1,5 cm longo disjunctis, bracteis foliaceis suffultis iis duplo — triplo longioribus. Calathidia ses- silia, axillaria, solitaria, circ. 21-flora. Involucrum late in- fundibuliforme, 6 mm altum, explanatum 1 cm latum, squa- mis 3 — 4-seriatis, erectis, appressis, laxis, extimis brevissimis deltoideo-linearibus, mucronatis, intermediis lanceolato-linea- ribus, intimis linearibus, acutis, basi dorso pallidis, superne violaceo-viridibus, margine pallidioribus, appresse pubescen- tibus, glandulis aureis, sessilibus sparse punctatis. Recepta- culum planum, nudum. Pappi setas interiores persistentes, numero 45, satis tenues, filiformes, albidae, quam exteriores lineari-lanceolatae circ. 12-plo longiores. Corolla 5 mm longa, glabra, tubo sensim dilatato long. 3/4 corollae, limbi laciniis subtus apicem versus breviter papillosis. Antherae lineares, 1,5 mm longae, ligula long. 1j1 antherae, auriculis sensim acu- tatis. Achaenia (immatura) dense sericeo-pubescentia, costis indistinctis. Vernonia remotiflora Grisebach, 1866, p. 144, p. p. — non L. C. Rich. Vernonia stenophylla Gleason, 1906 b, p. 181, p. p.; 1913, p. 308, p. p. — non Less. Vernonia angustissima Wright in sched. herb. Grisebach, nunc Goettingen. Hab. in Cuba: Wright n. 2786 (Bss, DC, G, orig. spec, K, KU, P, S). In general habit this species greatly resembles V. steno- phylla Less., only differing in the broader inflorescence with elongated, spreading cymes. The principal distinction is based upon the floral characters. V. stenophylla Less, has the co- rollas divided to the middle of their length, rather short, linear- -lanceolate anthers with elongated ligula and shortly acumi- 80 ARKIV FOR BOTANIK. BAND 13. NIO 15. nated basal lobes; the outer series of bristles is y8 m length of the inner one; V. angustissima has the corollas divided only to one fourth of the length, the anthers are linear with short ligula and long -acuminated basal lobes; further the outer series of bristles is only l/12 of the inner one. These structural differences are such as do not occur in one species, hardly in the same group of species. This fact together with the existence of a certain difference in habit (Gleason remarks, 1906 b, p. 181: »The single specimen examined — ■ Wright n. 2786 — does not agree in every particular with Lessing's description »), and with the different distribution has induced me to propose this new species. Vernonia corallophila Gleason. Vemonia corallophila Gleason, 1913, p. 309. Hab. in Cuba: Oriente, Guantanamo Bay, in ripa mari- tima ad rupes corallineo-calcareas, 17 — 30. 3. 09, Britton n. 1939. Not seen. The species comes near to V. angustissima and V. stenophytta, but differs from both in its small- leaves and few-flowered heads. Vernonia stenophylla Less. Tabula nostra V, fig. 1 (habitus). I'rrnonia stenophylla Lessing,' 1831, p. 6G7; de Candolle, 1836, p. 50; Gleason, 1906 b, p. 181, p. p.; 1913, p. 308, p. p. Cacalia stenophylla Kuntze, 1891, p. 971. Pappi setae interiores persistentes, numero 55, tenues, fili- formes, albidse, quam exteriores circ. 12-plo longiores. Co- rolla 4,5 mm longa, glabra, tubo long. 1/2 corollae, limbi laciniis apice subtus papillosis. Antherse lanceolato-lineares, 1,1 — 1,3 mm longae, ligula 1/5 long, antherse, auriculis satis brevibus, obtusiusculis vol brcviter acuminatis. Hab. in Hispaniola: ad rivulum »Riviere froide», Maj 1827, Jacquemont (B, ex herb. Kunth, orig. spec, P) in collibus prope Port au Prince, Eiirenberg n. 213 (B) Picarda n. 539 (KU); in »Morne Bonpere» prope Gonaives, locis siccis, mense Junio 1901, Buch n. 713 (KU); pro v. Bara- E. L. EKMAN, WEST INDIAN VERNONUE. 81 hona, Las Salinas, in ripa fluminis, alt. 100 — 150 m, Sept. 1911, Miguel Fuertes n. 1392 (KU). Two forms can be distinguished with the material at hand, one having very long-mucronated leaves and sparingly pube- scent involucral scales, which are conspicuously carinate and marginated. So Bitch n. 713. The other form, to which all other specimens seen belong, inclusive of the type of the species, has very shortly, or not at all, mucronated leaves and densely pubes- cent, obsoletely marginated scales. Additional material will certainly show that there are forms connecting these two extreme types. It is an interesting fact that there are forms of V. sericea subsp. racemosa, Turckheim n. 2905, 2906, showing such re- semblances to V. stenophylla that they might be taken for this species, were not the pappus brown instead of white. Ternonia sericea L. C. Rich. Tabula nostra IV, fig. 5 (folium et cyma). Vernonia sericea L. C. Richard, 1792, p. 112 (sphalm. 105), quoad typ. et descr., haud quoad patriam — non auct. al. Vernonia arborescens Persoon, 1807, p. 404, fide Cassini; Lessing, 1829, p. 302; De Candolle, 1836, p. 48 — non Sw. Lepidaploa phyllostaehya Cassini, 1823, p. 16. Vernonia Berteriana De Candolle, 1836, p. 52. Conyza portoricensis Bertero in sched. ap. De Candolle, 1836, p. 52. Vernonia arborescens Sw. [3 Lessinqiana Grisebach 1861, p. 353. Vernonia tricholepis Grisebach 1861, p. 354 p. p. — non DC. Cacalia sericea Kuntze, 1891, p. 971, quoad typum. Vernonia phyllostaehya Gleason, 1906 b, p. 176; Urban, 1911, p. 619. Vernonia venusta Gleason, 1906 b, p. 177; Urban, 1911, p. 619. Pappi setae interiores persistentes, numero 45 — 50, tenues, pallide brunnese, rarissimse albidse — straminese, exterioribus distinctis circ. 10-plo longiores. Corolla 5 mm longa, tubo angusto, glaberrimo, long. 3/4 corollse, limbi laciniis apice subtus papillosis. Antherse 0,8 mm longa?, ligula long. 1/e antherse, auriculis breviter acutatis. Hab. in Porto Rico: Baudin n. 292 (P, herb. Jussieu, verisimiliter spec. orig. Lepidaploce phyllostachyce Cass.); Ledru (Prodr, sub nom. V. arborescentis Sw. ); Bertero (B, Prodr, orig. spec. V. Berteriance DC); PLEEn. 205 (P); Schwan- ecke n. 95 (KU); Barbier, anno 1847 ('? P); prope Maya- Arkiv for botanile. Band 13. N:o 15. 6 82 ARKIV FOR BOTANIK. BAND 13. N:0 15. giiez, anno 1874, Krug n. 593 (KU); ibidem, anno 1875, Krug n. 592 (KU); prope Bayamon, Dec. 1883, Stahl n. 237 (KU); ibidem, Dec. 1886, Stahl n. 621 (KU); pr. Maricao ad Montoso, 20. 11. 84, Sintenis n. 389 (KU); pr. Maricao in convalli flu- minis, 22.11.84, Sintenis n. 390 (D, K, KU); pr. Mayaguez in monte Mesa, 29. 12. 84, Sintenis n. 390 b (B, KU); pr. Baya- mon in pratis ad »Pueblo viejo», 28. 3.85, Sintenis n. 1154 (KU); pr. Aibonito ad Algarrobo, 2. 12. 85, Sintenis n. 2936 (DC, KU); prope Coamo in monte Santana, 28. 12. 85, Sintenis n. 2936 b (KU); pr. Yauco in monte Duey, 5. 2. 86, Sintenis n. 3632 (KU); pr. Sabana grande in fruticetis ad ripam flum. Estero, 2. 3. 86, Sintenis n. 3929 (B, KU); pr. Rincon in silvula ad »Barrio del Pueblo », 1. 12. 86, Sintenis n. 5522 b (KU); pr. Rincon in montibus ad Puntas, 11. 12. 86, Sintenis n. 5642 (KU); pr. Utuado in fruticetis ad >>Los Angelos», L7. 1. 87, Sintenis n. 5915 (D, KU); pr. Utuado ad margines silvarum circa Santa Isabel, 8. 2. 87, Sintenis n. 6145 (KU); Heller et Heller, anno 1899, n. 84 (K), n. 183 (KU); »25 kilom. n. o. Mayaguez », alt. 325 m, 1. 2. 1900, Heller n. 4472 (KU); ad viam inter Adjuntas et Ponce, 27. 11.02, Heller n. 6136 (D, KU, fragm. orig. spec. V. venustce Gleason); coll. haud. not. (P); coll. haud not. ex herb. Ventenat (D). — St. Thomas: L. C. Richard (P, »frequens in montibus apricis incultis S:ti Thomae et in S:o Joanne »); Riedle (P); Hornbeck (S); Ehrenberg n. 9 (B); Friedrichsthal n. 99 (CD); Krebs (H, S); Oersted (H, sub nom. V. remotijlorm L. C. Rich.); Oersted n. 1, 12 (K); Hjalmarsson (S); pr. hos- pitium marinorum, 2. 12. 74, Eggers, sine num. (H); pr. »Ma- folie Gut», 22.6.76, Eggers, sine num. (KU); Aug. 1880, Eg- gers n. 301 (KU); in monte Signal Hill, alt. 460 m, Nov. 1880, Eggers n. 301 (B, BB, K, M, P); Jun. 1882, Eggers, sine num. (KU); ad-£olongo, Oct. 1885, Eggers n. 128 (H, M); ibidem, Dec. 1886, Eggers n. 128 (H); in silva, alt, 300 m, Febr. 1887, Eggers sine num. (L); sine loc. ace. Raunkler n. 3156, 3157 (H); ad Croun (?), 9. 5. 06, Raunkler n. 3160 (H); coll. haud not. ex herb. Schwagrichen (M). — St. Jan: L. C. Richard (P, »frcquens in arvis campestribus, collibus apricis insularum S :ti Thomae et S :ti Joannis — diversis anni temporibus florens » ) ; Benzon (H); Hornbeck (H); Great Maho Bay, 24. 12. 87, Eggers n. 3050 (H, KU); sine loc. ace. 13. 2. 06, Raunkleru. 3158 (H); 24. 2. 06. Raunki^er n. 3159 (H); 21 3. 06, Rain- E. L. EKMAN, WEST INDIAN VERNONI^E. 83 KUERn. 3161 (H).--St. Croix: West (H, S); Pflug (H); Ben- zon (H, S); Ravn (H); in collibus pr. »Elizas Retreat », Maj. 1871. Eggers. sine num. (H). — sine loc. ace.: Leblond n. 337 (D, orig. spec. V. sericece L. C. Rich, vel eius dupl.); ex Mus. Paris. (B); Riedle (D); ex herb. Zuccarini (M); ex herb. Schmiedel. (M); ex herb. Pa von (Bss). In the year 1792 L. C. Richard described, ver}^ briefly, some plants sent by Mr. Leblond from Guyana to the Societee d'Histoire Naturelle of Paris. Among these plants there was a Vernonia which Richard called Vernonia (err. Veronia) sericea, and characterized as follows: »fol. lineari-lanceolatis, subtus sericeo-incanis, ubique tomentosis, subintegris: floribus alternis, secundis sessilibus albis». This description, of course, is too narrow to characterize a Vernonia, and it seems there- fore as if nobody has been able to identify the species. By chance (at least so it appears) the name V. sericea was applied to a plant common at Rio de Janeiro, although this did not agree with the description, and was never collected in Guyana. When, in the summer of 1912, I examined the Vernonia? of the Delessert Herbarium at Geneva, I found a specimen of what, at that time, I called V. phyllostachya (Cass.) Gleason, labelled Vernonia sericea Rich., Leblond n. 337. At first, I could not understand the meaning of this determination, but I soon was able to ascertain that the plant in question was the type, or a duplicate of the type, of the mysterious Ver- nonia sericea L. C. Rich. The brief description of V. sericea quoted above will be found to agree perfectly with V. phyllostachya (Cass.) Gleason, except for the colour of the flowers. However, typically purple -flowered Vernonia? very often occur with white flowers. Further, the type locality of V. sericea is indicated to be Guyana, but V. phyllostachya has never been collected there. It often happened, however, in that early age of Botany that the habitats of plants described were interchanged. Since Leblond also collected in the West Indies (compare Urban, Nota? biographicae, Symb. Antill., Vol. 3) it is therefore evident that this plant came from there, and not from Guyana. Singularly enough I could not find at Paris any specimen of V. sericea collected by Leblond. The species thus renamed is one of the most richly re- presented in herbaria. It is easily recognized by its sericeous 84 ARKIV FOR BOTANIK. BAND 13. N:0 15. leaves and brown pappus. Analytically, it is characterized especially by its very small anthers. It varies somewhat as to the shape and pubescence of its leaves. Specimens from Porto Rico (Sintenis n. 389, 390 e) represent the most narrow- -leaved form (dimensions of the leaves: 7,5X0,5 cm), which is at the same time the most densely pubescent; speci- mens from St. Jan (RAUNKiiER n. 3158) have very wide and thinly pubescent leaves (dimensions: 11x5 cm). The variety mentioned by Urban, 1911, p. 619: »Var. foliis obovatis vel ellipticis, apice obtusis vel rotundatis. An species propria? », Sintenis n. 4749, is most probably a hy- brid between V. sericea L. C. Rich, and V. albicaulis Pers. (compared V. Gleasonii Ekman, on p. 54 of this paper). A remarkable variety of V. sericea having white pappus was described by Gleason in his »Revision» as a distinct species, V. venusta. I have been able to examine a specimen with white pappus, Sintenis n. 390 b in the Berlin Herbarium, which agrees perfectly with the small portion of type specimen of V. venusta seen. Now, no other difference than that of the colour of the pappus could be found to exist between V. sericea and V. venusta, even on the most careful analysis. Moreover, intermediate specimens have been seen, having stramineous or brownish-stramineous pappus, for instance, Sintenis n. 390 b in the Krug-Urban Herbarium, Sintenis n. 3929 in the Berlin Herbarium. I cannot see any reason for not reducing Gleason's species. Specimens of V. sericea L. C. Rich, collected by Oersted in St. Thomas have been referred to V. remotiflora L. C. Rich. and recently cited as that species from the West Indies (Glea- son, 1906 b, p. 175). True V . remotiflora does not occur there at all. Another specimen of V. sericea labelled »Jamaica, Purdie », but certainly not collected there, has been identified by Grise- bach as V. tricholepis DC. (Grisebach. 1861. p. 354), and by Baker in Flora Brasiliensis as V. araripensis Gardn. Neither species occurs in the West Indies. The geographical range of the species extends over Porto Rico, St. Thomas, St. Jan and St. Croix. It has also been indicated from Hispaniola. However, the plant in question in some points differs from the type of the species, and I have E. L. EKMAN, WEST INDIAN VERNONI^E. 85 found these differences striking enough to warrant its assign- ment as a subspecies of V. sericea: subsp. raeemosa (Delponte) Ekman n. subsp. Tabula nostra V, fig. 4 (habitus). Folia quani in specie tenuius pubescentia. subtus sordide grisea, raro subglabra; inflorescentia magis contracta, cymis brevioribus; squamae involucri brevius acuminata?, subvio- laceo-brunnea?. Pappi seta? interiores 35 — 45 exterioribus 8-plo longiores. Antherse 1,2 mm longse. Vemonia raeemosa Delponte, 1854, p. 396. Hab. in Hispaniola: Swartz (S); Poiteau (B); Ehrenberg n. 9 (B, p. p.), n. 150 (B); Picarda n. 145 (KU); prope Cadets, anno 1889, Picarda n. 883 (KU); prope Mira- goane, Jan. 1891, Picarda n. 189 (KU); prope Santiago in Questa da Piedra, alt. 300 m, Eggers n. 2370 (K, KU); prope Santiago in silva ad Lopez, alt. 350 m, 8. 6. 87, Eggers n. 2370 b (KU); ad Morne Bellance prope Gonaives, in silvis apertis, alt. 600 m, Oct. 1901, Buch n. 832 (KU); prope Con- stanza in pinetis, alt. 1200 m, Febr. 1910, Turckheim n. 2905 (D, KU, M); prope Constanza ad ripam rivuli, alt. 1190 m, Febr. 1910, Turckheim n. 2906 (Br, KU, S). Though in its typical forms well distinguished from V. sericea, the subspecies cannot always be recognized on habitual characters only. Especially the specimens collected by Swartz, which Wikstrom took to be V. divaricata Sw., strongly resemble V. sericea. Yet I think the larger anthers will prove to be a good analytical character of the subspecies. Its particular distribution may prevent confusion with V. sericea. The material at hand is extraordinarily heterogeneous. An exclusive type is Turckheim n. 2905 with narrowed, revolute leaves (dim. 4 X 0,3 — 0,5 cm) and very dense inflorescence. The plant collected by Buch has leaves resembling those of V. albicaulis Pers. or V. membranacea Griseb., and unusually evoluted floral branches. All specimens seen from Hispaniola have been found to belong to the subspecies except Mayerhoff in 1852 (B). However, this plant may have been collected in another island, 86 ARKIV FOR BOTANIK. BAND 13. N:0 15. for instance, in St. Thomas. This supposition is made more probable by the fact that another Vernonia, V. albicaulis, which certainly does not live in Hispaniola, is in the Berlin Herbarium labelled: Mayerhoff, Sto. Domingo in 1852. It represents the form of V. albicaulis common in St. Thomas. The type specimen of V. racemosa Delponte has not been examined. Since the description of this species agrees exactly with the Hispaniolan plant, and the type was collected there (by Bertero), there can be no doubt as to its identity. In the Lamarck Herbarium at Paris there is a specimen of the subspecies labelled: »Conyza de Mr. De Badier, de la Guadeloupe, n. 44 ». In the Prodromus Herbarium, however, the same label is affixed to V. arborescens Sw. (3. ovatifolia DC, i. e. to V. albicaulis Pers. — a fact illustrating how plants have been confused in ancient times. 5. The allies of Vermonia segregata Gleason. Vernonia segregata Gleason. Tabula nostra V, fig. 3 (inflorescentia). Vernonia segregata Gleason, 1913, p. 327. Hab. in Cuba: Oriente, Camp la Gloria, prope Sierra Moa, 24—30. 12. 10, Shafer n. 8051 (KU). I have seen only very imperfect portions of this puzzling species. It is readily recognized by its peculiar inflorescense : capitate clusters of heads subtended by bracteal leaves, the clusters both terminal and axillar. As to the leaves and the involucres it recalls the species near to V. havanensis DC, viz. V. hieracioides and V. cubensis, but differs widely in the in- florescence. Subsect. Scorpioides. Frutices elati, sazpe scandentes. Folia lata, scepe magna, basi scepe contracto-attenata. Inflorescentia cymoso-scorpioidea, cymis elongatis. midtifloris, simplicibus vel parce dichotome ramosis, calathidiis sessililnis. contiguis; hravteis nullis. Ca- E. L. EKMAN, WEST INDIAN VERNONI^. 87 lathidia parva. Receptaculum margine alveolarum lacerato- -fimbriatum. Pappi setce interiores persistentes, firmulce, jili- formes aut paullulo complanatce, numero 30 — 40, albidce, exte- riores distinctce. Corolla glabra, aut limbi laciniis dorso pilosis. Antherce 1,2 — 2,4 mm longce, ligula long. 1/3 — */„ anthers. Achai- nia tenuiter pubescentia. To this subsection I refer four species, V . scorpioides (Lam.) Pers., V. brachiata Benth., V. megaphylla Hieron., and a new species from the catingas of Bahia, collected by Ule, n. 7236. V. brachiata and V. megaphylla are closely allied, as are V. scorpioides and the new species from Bahia. In the West Indies only V. scorpioides occurs. It is found there in the islands of Tobago, Trinidad and Margarita, the flora of which greatly resembles that of South America. Yernonia scorpioides (Lam.) Pers. Tabula nostra VI, fig. 2 (inflorescentia). Conyza scorpioides Lamaeck, 1786, p. 88. Vemonia scorpioides Persoon, 1807, p. 404; Lessing, 1829, p. 282; 1831, p. 658; De Candolle, 1836, p. 41; Schultz-Bipontinus, 1847, p. 510; Gbtsebach, 1861, p. 353; Baker, 1873, p. 101; Gleason, 1906 b, p. 193. Vemonia subrepanda Persoon, 1807, p. 404. Vemonia toumefortiodes Humboldt, Bonpland, Kunth, 1820, p. 34; Lessing, 1829, p. 281; 1831, p. 657; De Candolle, 1836, p. 41. Lepidaploa scorpioides Cassini, 1823, p. 16. Vemonia centriflora Link et Otto, 1828, t. 55. Staehelina solidaginoides Willdenow herb. ap. Lessing, 1829, p. 281. Vemonia flavescens Lessing, 1831, p. 657; De Candolle, 1836, p. 41. Ghrysocoma repanda Vellozo, 1825, p. 306 (sec. ed. Netto); 1827, vero edita 1835, t. 13.1 Vemonia longeracemosa Mart. ap. De Candolle, 1836, p. 42. Vemonia lanuginosa Gardner, 1846, p. 219. Vemonia cincta Grisebach, 1879, p. 162. Cacalia cincta Kuntze, 1891, p. 969. Gacalia scorpioides Kuntze, 1891, p. 971. Cacalia tournefortioides Kuntze, 1891, p. 971. 1 The first edition. (1825) of the descriptions to Vellozo's Flora flumi- nensis is an extremely rare work. It was reprinted in 1881 (in Archivos do Museu Nacional do Rio de Janeiro). The plates to the descriptions, the well-known Florae fluminensis Icones, were not to be had in European libraries before 1835. From the descriptions it is quite impossible to iden- tify the plants concerned. Like Otto Kuntze I prefer to consider the species of Vellozo as published in the year 1835. 88 ARKIV FOR BOTANIK. BAND 13. N:0 15. Pappi setae interiores numero 30, persistentes, tenues, filiformes, albida3, exterioribus subfiliformibus, distinctis circ. 5-plo longiores. Corolla fere 6 mm longa, tubo glabro, sensim dilatato, long. 2/3 corollae, limbi laciniis dorso plus minusve pilosis, pilis longis, erecto-appressis. Antherae 2 mm longae, ligula 1/3 — 1/i long, antherae, auriculis brevibus, obtusis. Hab. in Tobago: prope Frenchfield in silv. mont. juxta Hope River, alt. 600 m, solo argillaceo, 26. 10. 89, Eggers n. 5615 (K, KU, P); in collibus ad Bacolet, 19. 11. 89, Eggers n. 5885 b (KU); Kings bay, alt. 60 m, 26. 4. 96, Seitz n. 47 (KU); Spring Road prope Scarborough, 12. 3. 09, Broadway n. 3080 (KU). — Trinidad: Cedros, ad ripam maris, 14. 1. 08, Broadway n. 2168 (KU, S). —Margarita: El Valle, 20. 7. 01, Miller et Johnston n. 134 (KU); ibidem, 30. 7. 01, Miller et Johnston n. 133 (KU); »Juan Griego trail», 2. 6. 03, John- ston n. 323 (KU). — Insuper in America austral! a Nicaragua usque ad Uruguay divulgata. A variable, though always easily recognized, species, abundantly represented in all herbaria. The West Indian spe- cimens agree throughout with South American ones. The syno- nyms cited all belong to South American specimens, and should be discussed in a work on the Vernoniae of the corresponding regions. Subsect. Pallescentes. Suffrutex erectus. Folia membranacea, elliptica, basi longe attenuata, apice longe acuminata, glabra, margine remote dentata. Inflorescentia cymoso-scorpioidea, cymis elongatis, sparse dicJio- tome ramosis, calathidiis parvulis, sessilibus, remotis; bracteis nullis. Involucrum infundibuliforme, squamis pauciseriatis, acutiusculis, extimis patulis, interioribus erectis. Beceptaculum subalveolatum, subnudum. Pappi setce interiores persistentes, numero 50, tenues, filiformes, brunnem, quam exteriores valde indistinctce, incequilongce, filiformes vulgo 8-plo longiores. Co- rolla 5 mm longa, glabra, tubo gracili, cylindrico, in limbum su- bito abeunte, long. 3/o corolloe, limbi laciniis patentibus, dorso apice papulosis. Antherce 1,5 mm longce, ligula long. lf6 antherce, auricidis obtusis. Achcenium tenuissime pubescens. This subsection includes only one species from St. Vincent. It is distinguished from other West Indian subsections by its E. L. EKMAN, WEST INDIAN VERNONIiE. 89 pappus, the bristles of the outer series being extremely indis- tinct, filiform. In general habit it recalls the Scorpioides, from which it differs, at first sight, in the colour of the pappus. Yernonia pallescens Gleason. Tabula nostra VI, fig. 3 (inflorescentia). Vernonia pallescens Gleason, 1906 b, p. 192. De characteribus analyticis, vide supra. Hab. in St. Vincent: in silvis montium, alt. 600 — 900 m, Juli. 1899, H. H. et G. W. Smith n. 922 (K, KU). From a remark on the label of the specimen cited (KU): »Vernonia = St Vincent, Anderson » it appears as if already Anderson had collected the species. Subsect. Polyanthes. Frutices vel suf frutices scepissime elati. Folia membranacea, forma et magnitudine varia, indumento numquam sericeo. In- florescentia cymoso-paniculata, scepe ampla, pluriflora, rotundata, cymis brevibus, irregulariter ramosis, calathidiis inferioribus scepe (in spec, antillanis semper) pedunculatis; bracteis nullis. Calathidia parva — mediocria, ad 25-flora. Receptacidum sub- planum. Pappi setce inter iores persistentes, f Hi formes, tenues, exterior es subdistinctce. Corolla glabra. Anther ce minor es. Achce- nia tenuissime pubescentia. The subsection thus defined includes a great number of species from the Andes (V. baccharoides H. B. K. and its allies), from Brazil (V. polyanthes Less, and its allies), and from Mexico (V. lanceolaris DC. and its allies). It matches the species- -groups Deppeance of Gleason, V. canescens H. B. K. exclu- ded, further, his Havanenses except V. pallescens Gleason, his Paniculatce dichotomce, and, may be, some of his Panicu- latce umbelliformes . The West Indian representatives are out- lying species of apparently very ancient origin. They may have originated from Columbia, where allied forms still live. The West Indian species can be divided into two smaller groups, one including V. havanensis DC. and its allies, the second V. menthcefolia (Poepp. ) Less. To the latter species Gleason attributes a higher taxonomic rank, making it the 90 ARKIV FOR BOTANIK. BAND 13. NIO 15. representative of a certain division of the section Lepidaploa: Paniculatce dichotomce, according to the structure of its inflor- escence. I think Gleason, having recently come across a member of the Havanenses having the same inflorescence, viz. V. hieracioides Griseb. (= V. Orientis Gleason, 1913), will not now deny the relationship between V. menthcefolia and Vernonice havanenses. In fact, V. havanensis highly resembles V. menthcefolia, and some specimens of it can only be distin- guished from that species by the characters of the petioles. V. havanensis and its allies, V. cubensis Griseb. and V. hieracioides Griseb., have the leaves attenuate into the petioles, the upper ones even without distinct petioles, while in V. menthcefolia, all leaves have very distinct, often very long petioles. As to the structure of the involucres and flowers, etc., no differences have been found between V. menthcefolia and the Vernonice havanenses. Conspectus specieruiu hujus suDsectionis. I. Folia basi in petiolum brevem, ssepe indistinctum atte- nuata. Species V. havanensi DC. proxima?. A. Folia tenuiter membranacea, subglabra, subhevia. Calathidia ultra 18-flora. Involucrum 6 — 7 mm al- tum. Pappi setae alba?. Corolla 7 — 8 mm longa. An- thera? circ. 3 mm longa?. V. havanensis DC. B. Folia firmula, laxissime induta, scabriuscula. Cala- thidia infra 14-flora. Involucrum 3 — 5 mm altum. Pappi seta? brunnescentes. Corolla 6 — 6,5 mm longa. Anthera? circ. 2 mm longa?. ]. Folia anguste rhomboideo-elliptica, indumento brunnescente, vix nitida. Inflorescentia vulgo habitu scorpioidea, cymis elongatis, multifloris, divaricatis, calathidiis brevius pedunculatis. Ca- lathidia 11 — 14-flora. Squama? involucri ap- pressa?, obtusiuscula? vel breviter acuminata?, sa- tis molles. V. cubensis Griseb. •1. Folia obovata vel late elliptica, rigida, nitidula, indumento nigricante. Inflorescentia habitu pa- niculata, cymis brevibus, paucifloris, calathidiis primariis longe pedunculatis. Calathidia 6 — 7 E. L. EKMAN, WEST INDIAN VERNONIiE. 91 (5< — 10)-flora. Squamae exteriores patulae, firmae, omnes conspicue cuspidatae. V. hieracioides Griseb. II. Folia basi in petiolum semper distinctum subito contracta. Inflorescentia ei V. hieracioidis similis. Calathidia magni- tudine V. havanensis. V. menihce folia (Poepp. ) Less. 1. The allies of V. havanensis DC. Vernonia havanensis DC. Tabula nostra V, fig. 6, 7 (inflorescentise). Vernonia havanensis De Candolle, 1836, p. 37; Gleason, 1906 b, p. 192; 1913, p. 329. Vernonia stictophylla Wright, 1869, p. 176. Cacalia havanensis Kuntze, 1891, p. 970. Pappi setae interiores persistentes, numero 40 — 50, fill- formes, tenues, quam exteriores lineares, breves 12 — 15-plo longiores. Corolla circ. 8 mm longa, tubo glabro, sensim dila- tato, long. 2/3 corollae, glandulis sessilibus, globuliformibus obsito, limbi laciniis apice subtus papillosis, longitudinaliter striatis. Antheree 3 — 3,3 mm longse, ligula enervi long. l/6 antherse, auriculis attenuatis, obtusiusculis. Hab. in Cuba: Ramon de la Sagra, n. 11 (P),n. 37 (Prodr), n. 67 (P, Prodr), n. 247 (P), n. 279 (Prodr), n. 331 (Prodr); Linden n. 45 (CD); Wright n. 1306 (KU, fragm. inflorescen- tiae), n. 2792 (Bss, K, una cum V. menthcefolia Less.!, P); Loma pelada, in silvis humidis, umbrosis, Wright n. 3596 (?) (K, KU). The numerous specimens collected by Ramon de la Sagra agree well with each other and with Wright n. 2792 in the Kew Herbarium. The specimen from Loma Pelada, which is the V. stictophylla Wright, appears to be only a shade form of the species, being collected in »damp, shady woods ». It has greenish involucres, those of genuine V. havanensis often being purple-tinged. Gleason in his Revision made no distinction between the three species of Vernonim havanenses , I suppose he had not seen the genuine V. hieracioides Griseb. at that time. In his »Studies» he has three genuine Vernoniai havanenses, viz. V. havanensis DC, V.Ottonis Sch.-Bip. and V.Orientis Gleason. 92 ARKIV FOR BOTANIK. BAND 13. N:0 15. Notwithstanding the different names, his species are the same as here accepted, recognized already by Grisebach. I scarcely venture to suppose, that his remark under V. havanensis: »The specimens at hand fall into two groups, the first with leaves long- -attenuate at base, almost sessile, and thin in texture; the second with leaves cuneate into a distinct petiole, and firm in texture », should be due to some confusion with V. menthce- folia; but this species can greatly resemble V. havanensis. Vernoiiia cubensis Griseb. Tabula nostra V, fig. 2 (folium et inflorescentia). Vernonia cubensis Grisebach, 1866, p. 144. Vernonia Ottonis Gleason, 1913, p. 330 — non Sch.-Bip. Pappi setae interiores persistentes, numero 30 — 35, fili- formes, tenues, sordide straminea?, quam exteriores anguste lineares 6-plo longiores. Corolla 6 — 6,5 mm longa, tubo sensim dilatato, glabro, long. 2/3 corolla?, glandulis sessilibus, globuli- formibus paucis prsedito, limbi laciniis subtus glandulis in- structis. Antherae 2,1 mm longa?, ligula long. Ye antherae, ener- vi, auriculis attenuatis, obtusiusculis. Hab. in Cuba: Wright n. 1306 (KU, S); n. 2791 (Bss, DC, G, orig. spec, KU, P); prov. Pinar del Rio, prope Pinar del Rio, 24. 10. 04, Baker et Abarca n. 3421 (KU); prov. Pinar del Rio. prope Guane, in palmetis, 1 — 4. 12. 11, Shafer n. 10646 (KU). Gleason applies (1913, p. 330) to this species V. Ottonis Sch.-Bip., as well as V. hieracioides Griseb. However, the type specimen of the former is typical V. menthaijolia (Poepp. ) Less., and the latter is the same as V. Orientis Gleason. Wright's number 1306 includes both V. cubensis Griseb. and V. hiera- cioides Griseb. Probably the specimens so numbered in the Gray Herbarium, which was consulted by Gleason, are all V. cubensis Griseb. The three species of Vernonice havanenses are by no means very good ones. V. cubensis Griseb. in many respects is inter- mediate between V. havanensis DC. and V. hieracioides Griseb. Some of the specimens cited above are suspected not to be quite correctly determined. E. L. EKMAN, WEST INDIAN VERN0NLE. 93 Veriionia hieracioides Griseb. Tabula nostra V, fig- 5 (habitus). Vernonia hieracioides Grisebach, 1862, p. 511. Cacalia hieracioides Kuntze, 1891, p. 970. Vernonia Orientis Gleason, 1913. p. 330. Pappi setse interiores persistentes, numero 35, filiformes, tenues. straminese, quam exteriores latae, distinct se circ. 10-plo longiores. Corolla 6 mm longa, tubo glabro, sensim dilatato, long, ultra 2/3 corolla?, glandulis globuliformibus rarissimis vel nullis, limbi laciniis striatulis, apice subtus papillosis. An- therae 2,3 mm longse, ligula brevi, enervi, long. 1/7 anthera3, auriculis brevissime acuminato-apiculatis. Hab. in Cuba orientali: prope villam Monte Verde dictam, Jan.— Jul. 1859, Wright n. 1306 (Bss, CD, sub nom. V. di- chroopappce Sch.-Bip., D, DC, G, orig. spec, KU, ex herb. Grisebach, ex herb. Alex. -Prior, P). Though I have not seen the type specimen of V. Orientis Gleason, I have identified this species with V. hieracioides Griseb. The most important features of the new species are the particular structure of the inflorescence and the involucres, as well as the few-flowered heads. Now V. hieracioides has exactty the same inflorescence and involucres, as shown by the type specimen in the Grisebach Herbarium and by its descrip- tion: »panicula fastigiata aphylla ramis ejus glandulo-pilosi- usculis dichotomis vel apice brevissime scorpioideis . . . squa- mis glabriusculis mucronato-acutis patulis». But V. hiera- cioides is said to have 10-f lowered heads, and V. orientis only 5-flowered ones? However, the type specimen itself was found to have 6 — -7 -flowered heads, and hence I suppose the numbers found by Grisebach and Gleason to be extreme. 2. The allies of V. menthaefolia (Poepp.) Less. Vernonia menthaefolia (Poepp. ap. Spr.) Less. Tabula nostra VI, fig. 1 (ramus floriger). Ewpatorium menthce folium Poeppig ap. Sprengel, 1826, p. 412. Vernonia menthce folia Lessing, 1829, p. 268 ; De Candolle, 1836, p. 38; Grisebach, 1862, p. 510; Gleason, 1906 b, p. 197; 1913, p. 331. Vernonia Ottonis Schultz-Bipontinus. 1847. p. 50S. Vernonia Grisebachii Schultz-Bipontinus, 1863, p. 231. 94 ARKIV FOR BOTANIK. BAND 13. N:0 15. Pappi setae interiores persistentes, numero 40 — 50, fili- formes, tenues, exterioribus subfiliformibus circ. 12-plo longi- ores. Corolla 6,5 mm longa, tubo sensim dilatato, glabro, long. 3/s corollae, limbi laciniis subtus apice glandulosis, glandu- lis parvulis, globuliformibus, sessilibus. Antherae 2,2 mm longae, ligula enervi, long. Ye antherae, auriculis brevissimis, apiculato-acuminatis. Hab. in Cuba: Poeppig (B, orig. spec. Eup. menthi- folii Poepp., Bss, F, KU, P); prope Fran que ville, Ramon de la Sagra, sine num. (B); Ramon de la Sagra n. 60 (Prodr), n. 79 (P, Prodr), n. 400 (Prodr), n. 596 (Prodr), sine num. (B); Mr. Leman, anno 1825 (Prodr); prope Catefal Fundador, 17. 1. 39, Otto n. 35 (B, CD, orig. spec. V. Ottonis Sch.-Bip.); Read (K); Wright n. 282 (D, DC, G, KU, P, S), n. 1305 (Bss, DC, KU, S), n. 2792 (Bss, CD, orig. spec. V. Grisebachii Sch.- -Bip., DC, G, K, KU, P, S); prope S:ta Maria, in fruticetis, mense Febr. 1889, Eggers n. 4626 (B, KU, P); Arroyo Gallego, mense Febr. 1889, Eggers n. 4647 (KU, M, P); Cieneguita, prope Cienfuegos, 2. 10. 96, Combs n. 706 (B, K, KU); Colonia Limones prope Cienfuegos, 2. 2. 03, Pringle n. 92 (D, KU); prope Hava- na, 30. 1. 05, Curtiss n. 623 (D, DC, H, K, KU, M); prope Ma- nagua, 23. 1. 05, Baker et Abarca n. 4456 (KU); prope San- tiago de las Vegas, 2. 2. 05, Abarca n. 4505 (KU); ibidem, 6. 2. 06, Baker n. 2643 (KU); Camaguey, prope La Gloria, 25. 1. 09, Shafer n. 28 (KU); prov. Pinar del Rio, inter E] Rosario et Los Palacios, ad marginem viae, 30. 1. 12, Shafer n. 12030 (KU); prov. Santa Clara, prope »Sancti Spiritus », in fruticetis, 15—24. 2. 12. Shafer n. 12113 (KU). By examining the type specimens I have been able to state the identity of the synonyms cited above. V. Ottonis Sch.- -Bip. was based upon a specimen with very small heads and rounded, tomentose leaves, V. Grisebachii Sch.-Bip. represents a form with larger heads and more narrowed, thinly tomentose leaves. In the Copenhagen Herbarium there is a specimen collected by Liebman, labelled: Mexico: Dep. Oaxaca, Oct. 1842, Lieb- m an" n. 356. The locality is clearly an error, and since Lieb- man also collected in Cuba, the specimen may have originated from that island. B. L. EKMAN, WEST INDIAN VERNONIiE. 95 Subsect. Noveboracenses. Suffrutices elati. Folia membranacea. scepe angnsta, in- tegerrima vel margine scepe serrulata. Inflorescentia cymoso- -paniculata, plana vel concavo-depressa, ambitu obpyramidata, cymis satis brevibus, exterioribus inferioribus longioribus, scepe sursum arcuatis; bracteis nullis. Calathidia scepe majuscula. Involucrum late campanulatum. Receptaculum planum, nudum. Pappi setce interior es persistentes, numero 40 — 70, tenues, fili- formes, scepe purpurascentes, exterioribus subindistinctis, ince- quilongis, anguste linearibus vulgo 8-plo longiores. Corollce satis longce, 8 — 10 mm in speciebus examinatis, glabrce, tubo 2/3 long, corollce. Antherce ultra 3 mm longce, ligula 1/.0 — 1/6 long, antherce, auriculis obtusis. Achcenia subglabra. This subsection includes several Mexican species, and all North American ones. It matches Gleason's Paniculatce verce, except his Alamaniance. It differs from the Polyanthes chiefty in its flat or depressed inflorescence, larger corollas and anthers, and nearly glabrous achenes. Only one West Indian species belongs here, viz. V. insu- laris Gleason, described as from the Bahama Islands. Evi- dently the origin of this species can be traced back to Florida, where related species are common. Vernonia insularis Gleason. Vernonia insularis Gleason, 1906 a, p. 184; 1906 b, p. 226. Hab. in insulis Baham. Great Bahama, N. L. Britton et C. F. Mlllspatjgh n. 2392, 2392 a; Brace n. 1385. — Abaco: Brace n. 1835. I have seen no specimens of this species. Sect. Tephrodes DC. Yernonia cinerea (L.) Less. Conyza cinerea Linnjeus, 1753, p. 862. Vernonia cinerea Lessing, 1829, p. 291. Pappi setae interiores valde caducae, numero circ. 30 (?), tenues, alba?, quam exteriores spathulata?, fimbriolatae circ. 96 ARKIV FOR BOTANTK. BAND 13. N:0 15. 10 — 12-plo longiores. Corolla tenuis, 4,5 mm longa, tubo sen- sim dilatato 3/4 long, corolla?, glabro, limbi laciniis dorso api- cem versus pilis nonnullis instructis. Antherae brevissimae, 0,8 mm longae, ligula enervi fere 1L long, antherae, auriculis acutatis. Hab. in insul. Baham. New Providence, 15. 2. 07, N. L. Britton n. 5374 (KU). — Cuba: prov. Pinar del Rio, prope Los Palacios, 3—4. 1. 12, Shafer n. 11(356 (KU.) — Ja- maica: Irwin hills, in cultis, anno 1849, Wullschlagel n. 885 (M); anno 1850, Pv. C. Alexander-Prior (G); ad Gor- donstown, alt. 300 m, 21. 1. 88, Eggers n. 3421 (Bss, D, KU, P); ibidem, alt. 360 m, 21. 1. 88, Eggers n. 3421 a (KU); Banana River, alt. 360 m, 21. 1. 88, Eggers n. 3468 (KU); Hope, 22. 10. 97, Harris n. 6865 (KU); Queen Anna's Bay, in margine viarum, 19. 3. 97, Churchill sine num. (KU); in montibus humilioribus prope Kingston, 17. 7. 97, O. Hansen sine num. (KU). — • Hispaniola: prope Port au Prince, Mart. 1892, Picarda n. 987 (KU); Cap Haytien, 16. 7. 87. Eggers n. 2843 (KU); prope Labrante, April. 1889, Buch n. 42 (KU). — St. Thomas: 10. 2. 06, Raunkler n. 1917 (H). — St. Croix, coll. haucl not. (H). — Guadeloupe: Le Baillif, ad ripam maris, anno 1892, Duss n. 2484 (KU); prope Basse- -Terre, anno 1892, Duss n. 2517 (KU). — Grenada: ad ripam maris frequens, anno 1893, G. W. Smith n. 106 (KU); 9. 3. 95, Broadway n. 432 (KU); St. Georges, Coopers Hill, 4. 3. 05, Broadway n. 2542 (D). — Tobago: ad stationem botani- cam, 11. 4. 09, Broadway n. 2542 (KU, S). A common, intertropical weed! Species excluded, Vernonia cliinensis (Lam.) Lessing, Linnsea, Band 6, 1831, p. 105. Hab. in Guadeloupe: prope Gombegre, locis incultis sparsa, Juni. 1900, Duss n. 4032. Gleason in his »Studies » lists this well-known cosmopol- itan species under its old generic name, Cyanthillium Blume, thus indicating the plant to be generically different from Vernonia, chiefly because of its very caducous, uniseriate pappus. I agree with him in assigning generic rank to the E. L. EKMAN, WEST INDIAN VERN0NIJ3. 97 plant. As for the specific name used by Gleason, C. chinensis (Lam.), I cannot join him. Lamarck describes, 1786, p. 83, Co- nyza chinensis as follows: 4. Conise de Chine. Conyza chinensis, Lin. » etc. He does not intend to describe a new species, believing himself to have the Linnaean plant before him. Now Conyza chinensis L. belongs to a widely different genus, being the type of Blumea chinensis DC. Lamarck made a false identification. I think the specific name chinensis cannot therefore be applied to our plant. The name to be used is »patula», from Conyza patula Ait. (1788). Its correct name would be, then, Cyan- thillium patulum (Ait.). However, as I have not seen the type of Conyza patula, nor studied sufficiently the relations of Cy- anthillium to Vernonia, I do not wish to be responsible for this combination. Yernonia lepidota Grisebach, 1866, p. 145. Hab. in Cuba: prov. Baracoa, in summo monte El Yun- que, 14. 6. 61, Wright n. 2789 (Bss, G, orig. spec, KU). Receptaculum alveolatam, marginibus alveolarum sca- riosis, dentatis. Pappi seta? interiores satis caducse, numero 5 — 6 — 7, crassse, complanatae, haud tortse, albidse, seta? exte- riores alte coalitae, patellam formantes cylindricam, margine laciniosam. Corolla 7 — 8 mm longa, tubo sensim dilatato, extus glanduloso, long. x/2 corolla?, limbi laciniis dorso glandulis instructis, apice recurvis, glandulosis. Anthera? omnino Ver- nonia3, 2,4 mm longa?, ligula subenervi, crassa, l/5 long, an- thera?, auriculis brevibus, oblique subemarginato-truncatis. Stylus Vernonio3. Achsenium compressiusculum, glaberrimum, irregulariter 10-costatum. As already stated (page 4), this species is to be removed from the genus Vernonia, chiefly by the characters of its pappus. Vernonia Milleri Johnston, 1905, p. 698. Hab. in Margarita: El Valle, 31. 7. 01, O. 0. Miller et J. R. Johnston n. 254 (KU, dupl. orig. spec). The plant clearly belongs to Oliganthes Cass. Avkiv for botanik. Band 13. N:o 15. 98 ARKIV FOR BOTANIK. BAND 13. N:0 15. Bibliography. Baker 1873. J. G-. Baker, Composite, pars I, in Mart. Fl. Bras., fasc. LXII, 1873. Bentham 1852. G. Bentham, Compositse centroamericanae, in Videnr skabelige Meddelelser fra den Naturhistoriske Forening i Kjoben- havn, Nr. 5 — 7, 1852. Cassini 1825. A. H. G. Cassini, Lepidaploa, in Dictionnaire des sciences naturelles, ed. F. G-. Levrault, tome XXVI, Strasbourg et Paris, 1823. De Candolle 1836. Aug. Pyr. De Candolle, Prodromus systematis naturahs Kegni Vegetabilis, pars V. Paris, 1836. Delponte 1854. J. B. Delponte, Stirpium exoticarum rariorum vel forte novarum pugillus. Mem. B. Ace. delle Sci. Torino, Ser. 2, Vol. XIV, 1854. Gardner 1846. George Gardner, Contributions towards a Flora of Brazil. Hooker's London Journal of Botany, Vol. 5, 1846. Gleason 1906 a. Henry Allan Gleason, Tbe genus Vernonia in the Bahamas. Bull. Torr. Bot. Club, Vol. 33, 1906. 1906 b. Henry Allan Gleason, A Bevision of the North Ame- can Vernonieae. Bull. New York Bot. Garden, Vol. 4, No. 13, 1906. 1913. Henry Allan Gleason, Studies on the West Indian Ver- nonieae, with one new species from Mexico. Bull. Torr. Bot. Club, Vol. 40, No. 7, 1913. Grisebach 1861. A. H. B. Grisebach, Flora of the British West In- dian Islands, Part. IV and V. London, 1861 (tit. 1864). 1862. A. H. B. Grisebach, Plantse Wrightianae e Cuba Orien- tali, Pars II. Mem. Americ. Acad. Arts. Sci., Ser. II, Part. II, No. XVIII, 1862. 1866. A. H. B. Grisebach, Catalogus plantarum Cubensium. Leipzig, 1866. 1879. A. H. B. Grisebach, Symbolae ad Floram argentinam. Abhand. Konigl. Gesellsch. Wissenschaft. zu Gottingen, Band 24, 1879. Hieronymus 1901. G. Hieronymus, Plantae Lehmannianse praeser- tim in Columbia et Ecuador collectae, Compositae II. Englers Bot. Jahrb., Vol. 23, Heft V, 1901. Humboldt, Bonpland, Kunth 1820. Nova genera et species plantarum quas in peregrinatione ad plagam oequinoctialem orbis novi colle- gerunt, describserunt, partim adumbraverunt Am at. Bonpland et Alex, de Humboldt. Ex schedis autographis Amati Bon- plandi in ordine digessit Carol. Sigismund. Kunth. Paris, 1820. Johnston 1905. J. B. Johnston, New plants from the Islands of Mar- garita and Coche, Venezuela. Proc. Am. Acad. Arts. Sci., Vol. XI, No. XXIV, 1905. Kuntze 1891. Otto Kuntze, Bevisio generum plantarum, Part. I — II. Wiirzburg, 1891. E. L. EKMAN, WEST INDIAN VERNONL.E. 99 Lamarck 1786. J. B. A. P. M. Lamarck, Encyclopedic methodique, Botanique, pars II. Paris, 1786. Lessing 1829. Chr. Fr. Lessing, De Synanthereis Herbarii Kegii Berolinensis Dissertatio prima. Linnaea, Band 4, 1829. — 1831. Chr. Fr. Lessing, De Synanthereis dissertatio quarta. Linnaea, Band 6, 1831. Link et Otto 1828. X. F. Link et F. Otto, Icones plantarum selecta rum horti regii Berolinensis, fasc. X. Berlin, 1828 (not seen) Linn^us 1753. Carl Linnaeus, Species plantarum, ed. I. Holmiae 1753. 1759 a. Carl Linn^ius, Systema naturae, ed. X. Holiniae, 1759 1759 b. Carl Linnaeus, Pugillus Jamaicensium plantarum. Amoe nitates Academicae, Vol. V, Holmiae, 1759. 1763. Carl Linn^us, Species plantarum, ed. II. Holmiae, 1763 Persoon 1807. Chr. H. Persoon, Synopsis plantarum. Paris, 1807 L. C. Richard 1792. L. C. Richard, Catalogus plantarum, ad socie tatem, ineunte anno 1792, e Cayenna missarum a Domino Le Blond. Act. Soc. Hist. Xat. Paris, Vol. I, 1792. Sagra 1850. Ramon de la Sagra, Historia fisica, politica y natural de la isla de Cuba, Vol. XI, Synipetalae etc. par A. Richard. Paris, 1850 (not seen). Schultz-Bipontinus 1847. C. H. Schultz-Bipontinus, Compositse, in »Beitrage zu einer Flora der ^Equinoctial- Gegenden der neuen Welt», von J. F. Klotsch. Linnaea, Band 20, 1847. 1863. C. H. Schultz-Bipontinus, Adnotationes in Cassiniaceas Wrightianas Cubenses, a cl. Grisebach determinatas, in Journal of Botany, ed. Seemann, Xo. XI, 1863. Sprengel 1826. Kurt Sprengel, Systema Vegetabilium, Vol. III. Gottingen, 1826. Stahl 1887. A. Stahl, Estudios sobre la Flora de Puerto-Rico. Vi- burn. — Gesner. 1887. (not seen). Swartz 1788. Olof Swartz, Xova genera et species plantarum seu Prodromus descriptionum vegetabilium, maximam partem in- cognitorum quae sub itinere in Indiam occidentalem annis 1783 — 87 digessit Olof Swartz. Holmiae, Upsaliae et Aboae, 1788. 1791. Olof Swartz, Observationes botanicae quibus plantae Indiae occidentalis aliaeque systematis vegetabilium ed. XIV illust- rantur earumque characteres passim emendantur. Erlangen, 1791. — 1806. Olof Swartz, Flora Indiae occidentalis, Vol. III. Erlangen, 1806. Urban 1899. Ign. Urban, Species novae praesertim portoricenses, in Symb. Antill., Vol. I, Xo. 5, 1899. — — 1903. Ign. Urban, Xova genera et species II, in Symb. Antill., Vol. Ill, Xo. 5, 1903. — 1911. Ign. Urban, Flora portoricensis, in Symb. Antill., Vol. IV, fasc. IV, 1911. 1912. Ign. Urban, Xova genera et species V, in Symb. Antill., Vol. VII, 1912. Vellozo 1825. Jose Marianno Vellozo, Florae Fluminensis, seu 100 ARKIV FOR BOTANIK. BAND 13. N:0 15. descriptionum plantarum praefectura Fluminensi sponte nascen- tium liber primus, ed. Netto. Arch. Mus. Nac. Rio de Janeiro, Vol. V, 1881. Vellozo 1827, vero edita 1835. Jose Marianno Veleozo, Florae Fluminensis Icones, Vol. VII. Paris, (1827) 1835. Willdenow 1804. C. L. Willdenow, Caroli Linn^i Species planta- rum. Berlin, 1804. Wikstrom 1828. Joh. Em. Wikstrom, Ofversigt af on Guadeloupe's Flora. Kongl. Vet. Akad. Stockholm Handl. for ar 1827, printed 1828. Wright 1869. C. Wright, Revisio catalogi Grisebachiani, in Anal. Acad. Cienc. med., fisic. y nat. de la Habana, 1869 (not seen). E. L. EKMAN, WEST INDIAN VERNONI.E. 101 Explanation of the Plates. (All figures are in natural size.) Plate 1. Fig. 1. Vernonia viminalis Gleason. Leaf, lower surface; branch of inflorescence. Wright n. 285 (S). » 2. Vernonia Valenzuelana A. Kich. Leaf, lower surface. Wright n. 2785 (G). » 3. Vernonia angusticeps Ekmax n. sp. Leaf, lower surface; branch of inflorescence. Wright n. 282 (DC). » 4. Vernonia inwquiserrata Sch.-Bip. Leaf, lower surface. Wright n. 285 prior (KU). » 5. Vernonia Sagrceana DC. Leaf, lower surface. Wright n. 281 (G-). » 6. Vernonia linguwfolia Ekman n. sp. Leaf, lower surface. Wright n. 285 post. (G). » 7. Vernonia aceratoides Gleason. Leaf, lower surface. Wright n. 2784 (G). Vernonia Sprengeliana Sch.-Bip. Branch of inflorescence. Miguel Fuertes n. 1388 (KU). Vernonia Wrightii Sch.-Bip. Leaf, lower surface; branch of inflorescence. Wright n. 284 (G). Vernonia Tuerckheimii Urb. Part of the inflorescence. Turckheim n. 2959 (KU). Vernonia buxifolia (Cass.) Less. Flowering branch. Turck- heim n. 2959 B (KU). Plate 2. 1. Vernonia arborescens (L. ) Sw. Branch of inflorescence. Duss n. 305 (KU). 2. Vernonia bahamensis Griseb. Leaf, lower surface; flowering- branches. Eggers n. 3893 (KU). 3. Vernonia arbuscula Less. Flowering branch. Curtiss n. 65 (KU). 4. Vernonia complicata Griseb. Flowering branch. Britton n. 2225 (KU). 5. Vernonia albicaulis Pers. Leaf, lower surface; flowering branch. RAUNKiiER n. 1919 (H). 6. Vernonia divaricata Sw. Part of the inflorescence. Wilson* n. 238 (G). 7. Vernonia borinquensis Urb. Part of the inflorescence. Hel- ler n. 4391 (KU). 8. Vernonia acuminata Less. Part of the inflorescence. Harris n. 8205 (KU). » 8. » 9. » 10. » 11. 102 AUKIV FOR BOTANIK. BAND 13. N:0 15. Plate 3. Fig. 1. Vernonia rigida Sw. Flowering branch. Swartz (S). » 2. Vernonia fruticosa (L. ) Sw. Upper part of the plant. Buch n. 522 (KU). » 3. Vernonia fruticosa (L. ) Sw. Branch of inflorescence. Mi- guel Fuertes sine num. (KU). » 4. Vernonia expansa Gleason. Inflorescence. Harris n. 8796 (KU). » 5. Vernonia leptoclada Sch.-Bip. Branch of inflorescence. Wright n. 1309 (KU). » 6. Vernonia yunquensis Gleason. Leaf, lower surface; a single head. Underwood et Earle n. 661 (KU). » 7. Vernonia pineticola Gleason. Two leaves, lower and upper surface; a single head. Underwood et Earle n. 1341 (KU). » 8. Vernonia pluvialis Gleason. Inflorescence. Rehder sine num. (KU). » 9. Vernonia parvuliceps Ekman n. sp. Two leaves, upper and lower surface; upper part of the plant. Wright n. 2788 (KU). Plate 4. Fig. 1. Vernonia gnaphali folia A. Rich. Inflorescence. Morales n. 12 (KU). » 2. Vernonia crassinervia Wright ap. Gleason. Upper part of the plant. Wright n. 2787 (G). » 3. Vernonia membranacea Griseb. Inflorescence. Wright n. 490 (G). » 4. Vernonia commutata Ekman n. sp. Upper part of the plant. Wright n. 286 (G). » 5. Vernonia sericea L. C. Rich. Leaf, lower surface; branch of inflorescence. Eggers n. 3050 (H). Plate 5. Fig. 1. Vernonia stenophylla Less. Upper part of the plant. Pi- card a n. 539 (KU). » 2. V. cubensis Griseb. Leaf, lower surface; upper part of the plant. Wright n. 2791 (KU). » 3. Vernonia segregata Gleason. Inflorescence. Shafer n. n. 8051 (KU). t> 4. Vernonia sericea L. C. Rich, subsp. racemosa (Delp. ) Ekman. Inflorescence. Turckheim n. 2905 (KU). » 5. Vernonia Meracioides Griseb. Upper part of the plant. Wright n. 1306 (G). E. L. EKMAN, WEST INDIAN VERNONLE. 103 Fig. 6. Vernonia havanensis DC. Branch of inflorescence. Wright n. 1306 (KU). » 7. Vernonia havanensis DC. Inflorescence. Wright n. 3596 (! KU). Plate 6. Fig-. 1. Vernonia menthcefolia (Poepp. ) Less. Flowering* branch. Curtiss n. 623 (H). » 2. Vernonia scorpioides (Lam.) Pers. Inflorescence. Seitz n. 47 (KU). » 3. Vernonia pallescens Gleason. Inflorescence. H. H. et G. W. Smith n. 992 (KU). » 4. Vernonia angustissima Wright ap. Ekman n. sp. Upper part of the plant. Wright n. 2786 (G). » 5. Vernonia trieephala Gardn. Flowering branch. Broadway n. 2174 (KU). 104 ARKIV FOR BOTANIK. BAND 13. N:0 15. Register. (Synonyms are in italics.) Pag. Bacchar is Myrsinites Pers., misapplied by Griseb 65 Cacalia acuminata OK 62 arborescens OK 42 bahamensis OK 57 buxifolia OK 23 cincta OK 87 gracilis OK 25 havanensis OK 91 hieracioides OK 93 Moritziana OK 25 punctata OK 48 scorpioides OK 87 sericea OK 81 stenophylla OK 80 Thomce OK 48 tournefortioides OK. 87 tricephala OK 26 Chrysocoma repanda Vell 17 Conyza anthelmintica L 8 » arborescens L 41 » » » misapplied by Swartz 59 » cinerea L 95 >> fruticosa L 68 » glabra Willd. 47 » portoricensis Bert. ap. DC 81 » rigida Sw 67 » scorpioides Lam 87 Eupatorium menthce folium Poepp. ap. Spr 93 » obtusifolium Willd 47 » salvifolium Bert. ap. Spr 21 » secundifolium Bert. ap. DC 48 Lepidaploa albicaulis Cass 48 » arborescens Less 41 » buxifolia Cass 22 » lanceolata Cass 48 » phyllostachya Cass 81 » scorpioides Cass 87 Piptocarpha tetrantha Urb • • • 8 Proustia domingensis Spr 22 Staehelina solidaginoides Willd. ap. Less 87 Vernonia aceratoides Gleason 20 acuminata Less 62 albicaulis Pers 47 albicaulis Pers. X sericea L. C. Rich 54 albicoma Gleason 59 amaranthina Gleason 59 angustata Gleason 76 angusticeps Ekman n. sp 14 angustissima Wright ap. Ekman n. sp 78 anthelmintica (L. ) Willd 8 araripensis Gardn., misapplied by Gleason 78 arborescens (L. ) Sw 41 » » » misapplied by Persoon 81 » » » » » Swartz 59 E. L. EKMAN, WEST INDIAN VERNONIJE. 105 Pag. Vernonia arborescens (L.) Sw. var. divaricata Griseb 42 » » » » » Lessingiana » 81 » » » » » ovatifolia DC 48 » » » » » Swartziana Griseb 59 » arbuscula Less 57 » arctata Gleason 57 » aronifolia Gleason 15 » bahamensis Griseb 57 » Berteriana DC 81 » borinquensis Urb. 55 » » » var. Stahlii Urb 56 » buxifolia (Cass.) Less 20 » calida Gleason 74 » calophylla Gleason 72 » centriflora Link et Otto 87 » chinensis (Lam. ) Less 96 » cincta Griseb 87 » cinerea (L. ) Less 95 » commutata Ekman n. sp 77 » complicata Griseb 58 » corallophila Gleason 80 » Cotoneaster (W. ) Less. var. angustifolia Hieron 25 » » » » var. pungens Hieron 25 » crassinervia Wright ap. Gleason 76 » cubensis Griseb 92 » desiliens Gleason 75 » divaricata Sw , 59 » » » misapplied by De Candolle 62 » » » » » Lessing 41 » doming ensis DC 23 » elongata W7illd. ap. Less 25 » emarginata Wtikstr 48 » expansa Gleason 64 » fallax Gleason 18 » flavescens Less 87 » joliosa Wright ap. Ekman 13 » fruticosa (L. ) Sw. 68 » » » » misapplied by Gleason 67 » xGleasonii Ekman nov. hybr 54 » gnaphalifolia A. Rich 76 » » » misapplied by Gleason 73 » gracilis H. B. K 25 » » » subsp. tomentosa Ekman nov. subsp 25 » » » var. villosa Less 26 » Grisebachii Sch.-Bip 93 » havanensis DC 91 » hieracioides Griseb 93 » icosantha DC 42 » insequiserrata Sch.-Bip 18 » » » var. angustifolia Griseb 20 » » » var. obtusifolia » 20 » insularis Gleason 95 » intonsa Gleason 59 » lanuginosa Gardn 87 » lepidota Griseb 97 » leptoclada Sch.-Bip. 73 » » » misapplied by Gleason 16 » linguffifolia Ekman n. sp 19 » longeracemosa Mart. ap. DC 87 » longifolia Pers 48 ArTeiv fur botanik. Band 13. N.o 15. 7* 106 AEKIV FOR BOTANIK. BAND 13. NIO 15. Pag. Vernonia membranacea Griseb 75 menthsefolia (Poepp. ap. Spk. ) Less 93 Milleri Johnst 97 montana Gleason 23 Moritziana Sch.-Bip 25 neglecta Gleason 73 obcordata Gleason 57 Orientis Gleason 93 Ottonis Sch.-Bip 93 » » misapplied by Gleason 92 pallescens Gleason 89 parvuliceps Ekman n. sp 71 permollis Gleason 59 phyllostachya Gleason 81 pineticola Gleason -. 74 pluvialis Gleason 65 proclivis Gleason 65 punctata Sw. ap. Wikstr 48 purpurata Gleason 14 racemosa Delp 85 reducta Gleason 65 remotiflora L. C. Rich., misapplied by Griseb .-78, 79 rigida Sw 67 » » var. Sagrazana Griseb 13 » » » Valenzuelana » 13 rubricaulis H. B. K., misapplied by Griseb 18 Sagraeana DC 13 scorpioides (Lam. ) Pers 87 segregata Gleason 86 semitalis Gleason 74 sericea L. C Rich •• • • 81 » » » » subsp. racemosa Ekman n. subsp 85 Sintenisii Gleason 48 Sprengeliana Sch.-Bip 21 stenophylla Less 80 stictophylla Wright 91 sublanata Gleason 76 subrepanda Pers. . 87 tetrantha (Urb. ) Ekman nov. comb 8 Thomce Benth 48 toumefortioides H. B. K 87 tricephala Gardn 26 tricholepis DC, misapplied by Griseb 26, 81 Trinitatis Ekman n. sp 39 Tuerckheimii Urb 24 Vahliana Less u 48 Valenzuelana A. Rich 13 ventosa Gleason 42 venusta Gleason 81 vicina Gleason 72 viminalis Gleason 15 Wrightii Sch.-Bip 16 » » misapplied by Griseb 73 » yunquensis Gleason 72 Tryckt den 13 februari 1914. Uppsala 1914. Almqvist & Wiksells Boktryckeri-A.-B. ARKIV FOR BOTANIK. Bd 13. N:o 15. 1. Vernonia viminalis Gleason. 2. V. Valenzuelana A. Rich. ap. Sagi 5. V. Sagraana DC. — 6. V. linguaefolia Ekman n. sp. 9. V. Wrightii Sch.-Bip. - 10. V. Tuerc PI. 00 " f A v^v V; ;k .-***> s %s.l - 3. V. angusticeps Ekman n. sp. — 4. V. inaequiserrata Sch.-Bip. — V. aceratoides Gleason. — 8. V. Sprengeliana Sch.-Bip. — nii Urb. — 11. V. buxifolia (Cass.) Less. Cederquists Graf. A.-B., Sthlm. ARK1V FOR BOTANIK. Bd 13. N:o 15. 1. Vernonia arborescens (L.) Sw. — 2. V. bahamensis Gi 5. V. albicaulis Pers. — 6. V. divaricata Sw. PI. 2. t>. — 3. V. arbuscula Less. — 4. V. complicata Griseb. . V. borinquensis Urb. — 8. V. acuminata Less. Cederquists Graf. A.-B., Sthlm. ARKIV FOR BOTANIK. Bd 13. N:o 15. 1. Vernonia rigida Sw. — 2. V. fruticosa (L.) Sw. — 3. V. fruticos 6. V. yunquensis Glcason. - - 7. V. pineticola Gleason. PI. 3. L.) Sw. — 4. V. expansa Gleason. — 5. V. leptoclada Sch.-Bip. V. pluvialis Gleason. — 9. V. parvuliceps Ekman n. sp. Cederquisls Graf. A.-B., Sthlcn. ARK1V FOR BOTANIK. Bd 13. N:o 15. Vcrnonin gnaphalifolia A. Rich. — 2. V. crassiner 4. V. commutata Ekman n. PI. 4. Wright ap. Gleason. — 3. V. mcmbranacea Griseb. — 5. V. sericea L. C. Rich. Cederquists Graf. A.-B., Sthlm. ARKIV FOR BOTANIK. Bd 13. N:o 15. 1. Vernonia stenophylla Less. — 2. V. cubensis Griseb. - - 3. V. segregal 5. V. hicracioides Griseb. — 6. V. hav PI. 5. I 4. V. sericea L. C. Rich, subsp. racemosa (Delp.) Ekman. — - 7. V. havanensis DC. Cederqnists Graf. A.-B., Sthlm. ARK1V FOR BOTANIK. Bd 13. N:o 15. hW/4 3 » „■ >4 !>.. V ", 1. Vernonia menthaefolia (Pocpp. ap. Spr.) Less. — 2. 4. V. angustissima Wright .ip. E PI. 6. • 4: V. scorpioides (Lam.) Pers. - - 3. V. pallescens Gleason. n n. sp. — 5. V. tricephala Gardn. V^ \ ^ V Ceclerquists Graf. A.-B., Sthlm. New York Botanical Garden Library QL313.Ve75E4 gen Ekman, E. L./West Indian Vernoniae 3 5185 00088 7032