(2 ANNALS OF THE
MISSOURI BOTANICAL GARDEN
— " РА већ) » ч и зї жээ — P еді бе ЭР © Bots » 1 tae 5 Aat me. ме. оди: бж "9 ~ ». Wb ` ч Je ag жж», + ХАТ * Л Ка md pA RR ERA, pe " =i E E 4 4 » У > РУМИ РЕ d NEU
Annals of the
Missouri Botanical
Garden
Volume XV 1928
With Fifty-three Plates and Thirty-seven Figures
Published pey. at 8 West King Street, Lancaster, Pa., by the Board of Trustees he Missouri Botanical Garden, St. Louis, Mo.
Entered as Ея — at the Post-Office " ae Pennsylvania, r the act of March 3,
Annals
of the Missouri Botanical Garden
A Quarterly Journal containing Scientific Contributions from the Missouri Botanical Garden and the Graduate Labora- tory of the Henry Shaw School of Botany of Washington Uni- versity in affiliation with the Missouri Botanical Garden.
Information
The Annals of the Missouri Botanical Garden appears four times during the calendar year: February, April, September, and November. Four numbers constitute a volume.
Subscription Price - - - $3.00 per volume Single Numbers - - - - 1.00 each The following agent is authorized to accept foreign subscriptions: Wheldon & Wesley, 2, 3 and 4 Arthur St., New Oxford St., London, W. C. 2, England.
—— (T
STAFF OF THE MISSOURI BOTANICAL GARDEN
Director, Сбковавк T. Moonx. Assistant to the Director, KATHERINE Н. Lerau.
HERMANN VON SCHRENK, Ernest 8. REYNOLDS, Pathologist. Physiologist.
Jesse М. GREENMAN, Davip H. LINDER, Curator of the Herbarium. Mycologist.
Когакр У. L. La GARDE,
EDGAR ANDERSON, Geneticist. Research Assistant
NELL С. HORNER Librarian and Editor of Publications.
BOARD OF TRUSTEES OF THE MISSOURI BOTANICAL GARDEN
President, Grorae C. Нитснсоск. Vice-President, SAMUEL C. Davis.
Second Vice-President, DANIEL К. CATLIN.
L. RAY CARTER. Рнпль C. SCANLAN. Tuomas S. MAFFITT. Jonn Е. SHEPLEY. ALBERT T. PERKINS. Евер С. ZEIBIG.
EX-OFFICIO MEMBERS:
, OR J. Ми, Chancellor of Washington University. pyrex of eet ro of St. Louis. FREDERICK Е. JOHNS GerorGE T. Moor
Bishop of the 5” of Missouri. President of ~ Academy of Sci- ence of St. Louis. Автнов А. BLUMEY President of the Board of sica of St. Louis.
DANIEL Breck, Secretary.
TABLE OF CONTENTS
A New Genus of the Acanthaceae.... . И С» е es ek АЛҒЫ Clarence E. Kobuski
A Monograph of the American Species
of the Genus Dyschoriste..... Clarence E. Kobuski Studies іп Ше Umbelliferae. I... . Mildred E. Mathias Concerning the Status of the Genus
15:01068............- а АИИ David Н. Linder Sources of Energy for Azotobacter, with
Special Reference to Fatty Acids....P. L. Gainey The Effect of Ultra-Violet Radiation
upon Higher Plants............. Ethel T. Eltinge The Problem of Species in the Northern
Blue Flags, Iris versicolor L. and Iris
virginica №............ oS REM Edgar Anderson А New Variety of Vernonia Lindheim-
ГТ EE Esther L. Larsen Dysosma: A New Genus of Berberida-
СВРСИ E Robert E. Woodson, Jr. Studies in the Apocynaceae. Й. |
Revision of the Genus Stemmadenia
и Robert Е. Woodson, Jr. Studies іп Ше Apocynaceae. III. A
Monograph of the Genus Amsonia. .
Sr. + oe ИРЕР Жака Robert E. Woodson, Jr.
PAGE
9-90 91-108
109-112
113-168
169-240
241—882
333-334
335-340
341-378
Annals of the Missouri Botanical Garden
Vol. 15 FEBRUARY, 1928 No. 1
A NEW GENUS OF THE ACANTHACEAE!
CLARENCE EMMEREN KOBUSKI Assistant, Arnold Arboretum of Harvard University Formerly Rufus J. Lackland Research Fellow in the Henry Shaw School of Botany of Washington University
A critical study of herbarium material of the genus Dyschoriste has revealed a small group of plants which possess sufficient morphological characters differing from Dyschoriste to merit generic recognition.
Apassalus? nov. gen. of the Acanthaceae. Calyx profunde 5-fidus. Corolla infundibuliforma; limbus subbilabiatus vel subaequalis; lobi rotundi, convoluti. Stamina 4, didynama, per paria lateraliter contigua vel connata decurrentia; antherae bi- loculares, basi obtusae, non acutae. Stigmatis lobus anticus obliquus vel dilatus, posticus subnullus. Capsula oblongo- linearis. Semina 2-4, plane compressa, suborbicularia.—Herbae perennes. КоПае ovatae, parvae. Flores parvae, solitarii vel in axillis fasciculati.
Type species: Apassalus diffusus (Nees) Kobuski.
KEY то SPECIES A. Capsule 2-seeded; plants covered with short, hirsute, spreading hairs; QUE PE io о kg A A. diffusus AA. Capsule 4-seeded; plants glabrous
vi mm. long, ovate-aubpotund: flowers 8-9 mm. long; КОЛОНИЈА са а, ои ни а 0007 0000 04 А ВВ. Leaves 25-45 mm. long, ovate-elliptic; flowers 11-12 mm. long; (Ah. THE a aeran EN US UI o Vos LOS . humistratus A. diffusus (Nees) Kobuski, n. comb. ELA. 2.
Dyschoriste diffusa (Nees) Urb. Symb. Ant. 7: 380. 1912.
! Issued April 30, 1928.
* Name dervied from the Greek a, without and хёссаХос, peg, on account of the absence of anther appendages. Ann. Мо. Вот. Garp., Vor. 15, 1928 (1)
[ Уот. 15 2 ANNALS OF THE MISSOURI BOTANICAL GARDEN
Dipteracanthus diffusus Nees in DC. Prodr. 11: 124. 1847.
Dyschoriste humistrata Lindau in Urb. Symb. Ant. 2: 188. 1900, not O. Ktze., namely, as to plants of Santo Domingo.
Stems somewhat tetragonal, slender, shortly hirsute, ascending from a perennial base, nodes closely placed, 1-2.5 cm. distant; leaves suborbicular-obovate, broadly obtuse at the apex, nar- rowing into a petiolate cuneate base, shortly hirsute on both surfaces, entire, 10-13 mm. long, 5-9 mm. wide; inflorescence bracteate, axillary; calyx 6-7 mm. long, lobes linear-acuminate, ciliate, 24 total length; corolla white (ex Buch) or pale lilac (ex Tuerckheim), puberulent оп the external surface, 7-8 mm. long, tube extending into a slightly ampliated throat, lobes rounded; anthers didynamous, filaments slightly pilose at the base, anther cells parallel or nearly so, truncate or rounded at the base; ovary 2-celled, glabrous, style linear, pubescent a little above the base, stigma dilated, oblique; capsule 6-7 mm. long, 2-celled, each cell containing a single seed attached by the retinaculum, both of which (retinacula) are situated on the central ridge of the com- missural surfaces; seeds flat, orbicular, becoming mucilaginous when wetted.
Distribution: Islands of Haiti and Santo Domingo.
Specimens examined:
Haiti: on rocky outcrop, dry wooded mountain slope, vicinity of St. Marc, 25-28 Feb. 1920, Е. С. Leonard 2913 (US, G); dry bank along road near Ennery, Dept. of Artibonite, 325-900 m. alt., 13 Jan. 1926, Е. С. Leonard 8823 (US); arid thickets, north- east of the N. West Indies Company, vicinity of St. Michel de l'Atalaye, Dept. du Nord, 300 m. alt., 17 Хоу. 1925, Е. С. Leonard 7093 (US); common in dry thickets, vicinity of 56. Michel de l'Atalaye, Dept. du Nord, 350 m. alt., 26 Nov. 1925, E. C. Leonard 7472 (US); Barahona, 1200 m. alt., Sept. 1911, Fuertes 1407b (FM, G, US).
Santo Domingo: Azua, March, 1913, Rose, Fitch & Russel 4072 (05).
A. cubensis (Urb.) Kobuski, п. comb. PLE
Dyschoriste cubensis Urb. Symb. Ant. 7: 381. 191.
Dyschoriste humistrata Lindau in Urb. Symb. Ant. 2:188. 1900, not O. Ktze., namely, as to plants of Cuba.
1928] KOBUSKI—-A NEW GENUS ОЕ ACANTHACEAE 3
Ruellia diffusa Grisebach, Cat. Pl. Cub. 195. 1866 (excl. syn.); балу. Fl. Cub. 97 (no. 1500). 1873.
Low-growing perennial, decumbent, occasionally rising erect, glabrous or minutely scabrous, young stems densely covered with cystoliths; leaves shortly petiolate, ovate to suborbicular, 9-12 mm. long, 5-7 mm. wide, rotund at the apex, tapering to a cuneate base, entire, densely covered with cystoliths on both surfaces, glabrous; flowers solitary, rarely in twos, bracts narrowly obovate; calyx 5-cleft, 6-8 mm. long, lobes linear-acuminate, nearly 24 total length, entire external surface covered with cystoliths, glabrous, lobes ciliated; corolla 8-9 mm. long, tube cylindrical, enlarging until ampliated throat is reached, lobes shortly obovate; stamens didynamous, adnate to the middle of the tube, anthers narrowly ovate, obtuse at the base; ovary 2-celled, style linear, nearly glabrous; capsule oblong-linear, 7-8 mm. long, glabrous, 4-seeded; seeds suborbicular, mucilaginous when wetted.
Distribution: near Cojimar, Prov. of Havana, Cuba.
Specimens examined:
Cuba: near Cojimar, Prov. of Havana, 14 March, 1906, Baker 2894 (FM); shady places in coastal sand between Rio Cojimar and Playa de Bacuranao, Prov. of Havana, 26 Dec. 1910, Wilson 9533 (G, US).
A. humistratus (Michx.) Kobuski, n. comb. EL
Dyschoriste humistrata (Michx.) O. Ktze. Rev. Gen. Pl. 2: 486. 1891.
Ruellia humistrata Michx. Fl. Bor.-Am. 2: 23. 1803; Pursh, Fl. Am. Sept. 2: 421. 1814.
Calophanes humistrata Shuttleworth ex. Nees in DC. Prodr. 11: 108. 1847; Gray, Syn. Fl. N. Am. ed. 1, 2: 324. 1878, and ed. 2. 1886; Chapman, Fl. Southeastern U.S. ed. 1, 1083. 1860, and ed. 2. 1889.
Dipteracanthus humistratus о. Fl. Southeastern U.S. ed. 2, 303. 1889.
Dipteracanthia riparius Chapman, Fl. Southeastern 0.5. ed. 2, 303. 1889.
Stems several, ascending or rising erect from a ligneous peren- nial base, 4 dm. or less high, glabrous or slightly pubescent;
[Vor. 15 4 ANNALS ОЕ THE MISSOURI BOTANICAL GARDEN
leaves ovate-elliptic to oblong-sublanceolate, 2.5-4.5 cm. long, 1-2 em. broad, obtuse to acute at the apex, abruptly attenuated at the base into a petiole which may be so short as to give the leaf a subsessile appearance or as much as 4 mm. long, glabrous or nearly so, entire or slightly crenulate margins; bracts oblong- oblanceolate, about equalling the length of the flower; flowers axillary; calyx deeply 5-parted, 9-10 mm. long, glabrous or slightly pubescent, lobes subulate-setaceous; corolla small, white, 10-11 mm. long, tube 2.5-4 mm. long; stamens didynamous (very seldom 5), filaments pubescent at point of adnation to corolla throat, anther cells obtuse or slightly mucronulate at the base; mature capsule 9-10 mm. long, glabrous, linear, 4-seeded.
Distribution: low grounds, southeastern United States.
Specimens examined:
Georgia: Lumber City on the Ocmulgee River, Telfair Co., July, 1900, C. Mohr (US, 721392); shaded places in Ogeechee River swamp, Burke Co., 5 June, 1901, R. M. Harper 769 (M, US).
Florida: fertile ground under oaks, upper St. John’s River, 1 June, A. H. Curtiss 28 (а); Hot Springs, 7 April, 1925, H. O’Neill 601 (US); Pine Island, St. John’s River, 11 April, 1911, S. C. Hood (G); swampy shore of St. John’s River, June, 1878, A. H. Curtiss 1939 (M, FM, G, US); wooded banks of the Suwan- nee River at Branford, Suwannee Co., 9 June, 1900, А. Н. Curtiss 6654, (С, M); Suwannee Co., June-July, 1898, А. S. Hitchcock 1457, 1468 (FM); damp shady places, banks of Rice Creek, Putnam Co., 26 March, 1882, C. Mohr (US 721391); Dunnellon, Marion Co., 25 Feb. 1891, L. F. & R. Ward (US, 147428); Port Orange, Volusia Co., 20 Мау, 1895, F. C. Straub 164 (а); Lake Alfred, Polk Co., 11 June, 1922, С. М. & J. К. Armstrong (М 911680); swamp, Hernando Co., June-July, 1898, А. S. Hitchcock (M 120820).
озю;
огл фет
ГУ оз. 15, 1928]
ANNALS OF THE MISSOURI BOTANICAL GARDEN
EXPLANATION OF PLATE PLATE 1 Apassalus diffusus (Nees) Kobuski Open calyx. Open corolla showing stamens. Dehiscing capsule showing seeds and retinacula. Apassalus cubensis (Urban) Kobuski
Open calyx.
Open corolla showing stamens. istil.
Dehiscing capsule showing seeds and retinacula.
Apassalus humistratus (Michx.) Kobuski Open calyx. Open corolla showing stamens.
Dehiscing capsule showing seeds and retinacula.
ANN. Мо. Вот. Garp., Vor. 15, 1928 PLATE 1
KOBUSKI—A NEW GENUS ОЕ ACANTHACEAE
[Vor. 15, 1928] 8 ANNALS OF THE MISSOURI BOTANICAL GARDEN
EXPLANATION OF PLATE PLATE 2 Fig. 1. Apassalus diffusus (Nees) Kobuski From the specimen, Fuertes 1407b, in the United States National Herbarium, Fig. 2. Apassalus cubensis (Urban) Kobuski From the specimen Baker 2894, in the Herbarium of the Field Museum.
AVAOVHINVOV ЯО SONHD MAN У—ТУИЗПЧОМ
Ann. Mo. Bor.
GARD.,
Мог.
15, 1928
PLATE 2
A MONOGRAPH OF THE AMERICAN SPECIES OF THE GENUS DYSCHORISTE!
CLARENCE EMMEREN KOBUSKI
Assistant, Arnold Arboretum of Harvard Universit Formerly Rufus J. Lackland Research Fellow in the Henry Shaw School of Botany of Washington University
INTRODUCTION
Any attempt to determine specifically herbarium specimens of the genus under present consideration formerly proved very unsatisfactory because of the inadequacy of many of the original descriptions and also because of the small representation of type or authentic material in American herbaria. These facts, along with an especial interest in the genus and its allies, led to the present study. At first it was hoped that a monographic treat- ment of the whole genus might be made. A general survey of the material deposited in American herbaria, however, showed the Old World species to be so poorly represented that it was deemed advisable to exclude them from the present discussion and to include only the American species. Later the writer plans to visit some of the larger European herbaria and to supplement this monograph by a critical study of the far-eastern species.
This investigation was made possible only through the собрег- ation of the botanists connected with the various herbaria from which material was borrowed. Sincere appreciation is due Dr. B. L. Robinson of the Gray Herbarium, W. R. Maxon of the United States National Herbarium, and D. C. Davies, Director of the Field Museum, who so willingly loaned their entire col- lections of Dyschoriste for this study. It was found necessary also to borrow types, and to obtain fragments and photographs of type collections from several European herbaria. Dr. Santiago Ramón y Cajal, Instituto Cajal, Madrid, and Professor Eduardo Balguerias y Quesada, Jardin Botanico, Universidad de Madrid,
1 Ап investigation carried out at the Missouri Botanical Garden in the Graduate Laboratory of the Henry Shaw School of Botany of Washington University, and submitted as a thesis in partial fulfillment of the requirements for the degree of doctor of philosophy in the Henry Shaw School of Botany of Washington University.
Issued April 30, 1928.
ANN. Mo. Вот. Слво., Vor. 15, 1928 3 (9)
[Vor. 15 10 ANNALS OF THE MISSOURI BOTANICAL GARDEN
very kindly furnished an excellent photograph of a little-known species, the type of which is preserved in the Madrid Herbarium. Professor Boris Fedtschenko, Jardin Botanique Principal, Lenin- grad, U.S.S.R., obligingly supplied two types essential for the completion of thismonograph. Dr. A. W. Hilland T. A. Sprague, Royal Botanic Gardens, Kew, Dr. L. Diels, Botanischer Garten und Museum zu Berlin-Dahlem, Dr. C. H. Ostenfeld and Dr. Carl Christensen, Botanisk Garten, Kgbenhavns Universitet, as well as Dr. Adele Lewis Grant, Huguenot College, South Africa, who so willingly made critical comparisons with types at the Kew Her- barium on her journey to Africa, have all contributed either directly or indirectly, in material loaned or in verification of specimens submitted for comparison. The writer takes this op- portunity to express his gratitude for their generosity and kindly assistance.
This study was made at the Missouri Botanical Garden, and thanks are due to the Director, Dr. George T. Moore, for the use of the excellent library and herbarium facilities which this in- stitution affords. Also, especial thanks are extended to the Curator of the Herbarium, Dr. Jesse M. Greenman, under whose constant guidance and supervision this work has been made possible.
HISTORY
The genus Dyschoriste was proposed by Nees in the third volume of Wallich's ‘Plantae Asiaticae Rariores’ 1 published in 1832. The genus was segregated from Ruellia on account of stamen, corolla, and fruit characters, and was based on Ruellia depressa Wallich, namely, Wallich’s No. 2379 from East India, which, however, is not conspecific with Ruellia depressa L. Two other East Indian species, D. cernua and D. litoralis, were referred by Nees in the same work to his new genus.
In 1833, only a year later, David Don in Sweet’s ‘British Flower Garden’? described the genus Calophanes. Don’s new genus was also a segregate from Ruellia, and was founded on Ruellia oblongifolia Michaux, which in turn was based on speci- mens collected in the state of хан
! Wallich, N. РІ. As. Rar. 3: 81. ?Sweet, В. Brit. Fl. Gard. II, 2: x 181. 1833.
1928] KOBUSKI—MONOGRAPH OF DYSCHORISTE 11
These two generic names were current in botanical literature for many years, as representing two supposedly distinct genera indigenous to remote regions—Dyschoriste of the eastern hemi- sphere and Calophanes of the western hemisphere. Nees, the foremost student in his time of the Acanthaceae, in his treatment of this family for Martius’ ‘Flora Brasiliensis,’ 1 in 1847, accepted Calophanes and described seven Brazilian species of this genus. The same author and in the same year elaborated the Acanthaceae for DeCandolle's ‘Prodromus’ ? and maintained both names as representing separate and distinct genera. In this work, which was the first to present а comprehensive treatment of the group, five species of Dyschoriste and twenty-seven species of Calo- phanes, as well as several varieties, were recognized.
Bentham and Hooker in the ‘Genera Plantarum, 2 1876, treated these two previously supposed distinct generic elements as congenerie, but unfortunately they took up the later name Calophanes and relegated Dyschoriste to synonymy.
Mr. C. B. Clarke, who contributed the treatment of the Acanthaceae for Hooker's ‘Flora of British India,’ * 1885, fol- lowed Bentham and Hooker’s generic interpretation of Ше group and recognized four East Indian species of Calophanes, namely, C. Nagchana Nees, C. littoralis 'T. Anders., C. vagans Wight, and C. Dalzellii T. Anders. Under C. Nagchana Nees the following species are cited as synonyms: C. depressa T. Anders., Ruellia Nagchana Ham., В. erecta Burm., R. depressa and R. cernua Nees, Dipteracanthus Nagchana Nees, Dyschoriste depressa, and D. cernua Nees.
In 1891, Dr. O. Кип ге" revived the name Dyschoriste Nees and transferred thereto several species, including Ruellia erecta Burm. which was described and illustrated in 1768, being the oldest known species of this group. Lindau, in 1895, in reviewing the Acanthaceae for Engler and Prantl’s ‘Die Natürlichen Pflanzenfamilien, * followed Kuntze in recognizing the genus Dyschoriste.
1 Martius, C. F. P. de. Fl. Bras. 9: 26. 1847.
2 DeCandolle, A. P. Prodr. 11: 106, 107. 1847.
3 Bentham, G. & Hooker, J. D. Gen. Pl. 2: 1077. 1876.
‘Hooker, J. D. ЕІ. Brit. India 4: 410. 85.
5 Kuntze, O. Rev. Gen. Pl. 2: 486. 1891.
6 Engler, А. & Prantl, K. Nat. Pflanzenfam. 4%; 302. 1895.
(Мог. 15 12 ANNALS OF THE MISSOURI BOTANICAL GARDEN
Clarke, in working up the Acanthaceae for the ‘Flora Capen- sis, 1901, took up the name Dyschoriste, thus reversing the position taken by him in Hooker’s ‘Flora of British India,’ mentioned above. He recognized D. depressa Nees as a valid species along with four other African species, one of which is D. erecta Clarke, thus apparently disregarding the D. erecta (Burm.) O. Ktze.
Several new species have been described from time to time and referred to either Dyschoriste or Calophanes, but no compre- hensive treatment of the group as a whole has been published since that of Nees in DeCandolle’s ‘Prodromus.’
Assuming that there is absolute identity and thus complete synonymy of the several elements which Clarke referred to Calophanes Nagchana in the ‘Flora of British India,’ then, as pointed out by Dr. Kuntze, the name erecta, as the oldest specific name involved, must be retained and the binomial Dyschoriste erecta (Burm.) O. Ktze. becomes the valid combination for the plant concerned and D. depressa (Wall.) Nees must be regarded as a synonym of it. Since this study has been confined to the American species we must admit the observations of Clarke and Kuntze and accept Dyschoriste erecta (Burm.) O. Ktze. as the type species of the genus until the eastern species in question can be examined.
GENERAL MORPHOLOGY
Roots.—The root system in the genus Dyschoriste is not very extensive. All the species are perennial and the roots, in turn, are of the simple fibrous form. By the casual observer, however, some of the slender underground stems of the previous year’s growth are sometimes mistaken for roots.
Stems.—There is considerable variation in the stem and its habit of growth. In all cases, the stem or stems arise from a ligneous perennial base. However, the mode of ascent varies. Many species are prostrate and the stems spread over the ground in several directions. In these cases the leaves assume a secund position. Often when the stems are short a rosette appearance is attained for the whole plant. The species D. oazacensis illus-
! Flora Capensis 51: 15. 1901.
1928] KOBUSKI—MONOGRAPH OF DYSCHORISTE 13
trates this character. However, it is not a stable specific char- acteristic. A second habit of growth is the ascending type. It is in this category that the majority of species is placed. The stem not infrequently becomes more or less geniculate; this mode of growth is very characteristic of D. Pringlei. The third habit of growth is the erect type. It is to this type that the sturdy D. hirsutissima, D. oblongifolia, D. ovata, and D. trichanthera belong. Some species may follow consistently a distinct habit of growth, while others may have the stems ascending or erect even on the same plant. One can usually associate stem growth exceeding a length of 4-5.5 dm. with the erect habit, and shorter stems with the ascending or prostrate habit. Along with this, the prostrate type will possess small leaves and the erect type will have leaves with a more extensive surface.
Several species, among them JD. oblongifolia, possess slender underground stems which are common in perennials. These stems have the appearance of roots but on close observation buds and modified leaves can be seen. After passing under the ground for some distance they come to the surface and then rise erect.
The stem may be terete or quadrangular. The latter is the more common type іп the genus. Іп the species D. quadrangularis the stem is not only angular but the angles are winged. This condition is probably brought about by the decurrent petiole of the leaf extending down the stem.
Leaves.—The leaves of the species in the genus Dyschoriste present a great variety of differences. АП species have leaves with entire margins, except D. bilabiata which is not distinctly dentate but has a decided tendency in that direction. Several species, such as D. crenulata and D. hirsutissima, show a tendency toward crenulate margins. Others combine the crenulate with the repand margin. Along with these characteristics the margin is usually ciliate. In shape the leaf varies from that of the narrow, linear D. angusta, D. Greenmanii, and D. Purpusii to that of the oblong-ovate D. quadrangularis which grows to a length of 10 centimeters. Some species have two types of leaves, the lower or cauline leaves often being larger and different in shape from the upper leaves in whose axils branches and flowers are crowded.
(Хот, 15 14 ANNALS OF THE MISSOURI BOTANICAL GARDEN
The surface of the leaf itself is usually pubescent. When the pubescence is sparse or absent, an abundance of cystoliths is usually seen. Often the cystoliths, since they lack an orderly arrangement, are mistaken for appressed hairs. This cystolithic character is not sufficiently definite for specific delimitation. Some plants have a more copious formation of cystoliths than others. This character exists on the stem and calyx as well as the leaves. The venation which is very pronounced on the lower surface varies very little within the genus. The usual form is the feather-veined type.
Inflorescence.—In all cases the flowers are axillary and sub- tended by bracts and sometimes bracteoles. Only occasionally are the flowers solitary in the axils. "There are usually several to many flowers at the node, giving the appearance of a cymose cluster as in D. quadrangularis or а capitulum as in D. capitata and D. pinetorum. D. Greenmanii is an excellent example of a species with a solitary flower at the node. The majority of species have flowers which are pedicellate, often so short, however, that a subsessile effect is presented.
Calyx.—In the calyx of Dyschoriste is found one of the most constant characters of the genus. It is usually five-parted and always persistent. The only deviation from the five-merous condition is found in D. maranhonis where the calyx-lobes are occasionally only four. In all cases the lobes are subulate- setaceous and usually ciliate. The ciliation may vary from a long whitish, flaccid pubescence to a very short hirsuteness. When the calyx proper is pubescent, the pubescence is usually confined to the nerves. The tissue connecting the lobes of the calyx is usually very membranaceous and tears apart very easily, making it difficult and quite unsatisfactory to use the ratio between tube-length and lobe-length as a character for specific differentiation. The lobes are usually quite equal in length. However, here again variation is found. Cystoliths, as in the leaves, are very abundant; but in the calyx they are frequently disposed in a more or less regular arrangement.
Corolla.—There is very little differentiation to be found in the corolla of the genus. In most cases the bilabiate type occurs. This, however, is not as distinct as in some other genera of the
1928] KOBUSKI—MONOGRAPH OF DYSCHORISTE 15
Acanthaceae. The corolla is five-lobed, the two posterior lobes being coalescent to a greater extent than the three anterior lobes. The length of the corolla varies from 10 to 17 mm., as found in D. decumbens, D. hygrophiloides, D. saltuensis, D. quadrangularis, and D. angustata, to 25-28 mm., exemplified in D. zylopoda, D. humilis, and D. ovata. One species, D. Pringlei, has а corolla measuring 35-38 millimeters long. None, however, reach the length of 70-80 millimeters as found in some of the Ruellias. The proportion between tube and throat is variable in the genus. The ventricose throat is found quite often. The condition should exist in all species because of the contiguity of the adnate fila- ments in the posterior portion of the throat and tube. This ventricosity, hence, is more pronounced in the larger-flowered species. The narrow tube of the corolla is usually slightly flared at the base to make room for the disc and ovary. The ampliation from the tube to the throat is very variable and may be abrupt or gradual according to the species. In all cases, the external surface is quite pubescent. In the species D. trichanthera the pubescence is found on the interior as well as the exterior surface of the throat.
Stamens.—The stamens are didynamous. The long and short filaments on each side are contiguous or united at the base by a membrane which extends from the point of adnation to the base of the corolla tube. A very distinctive feature of the anthers is the mucronate appendages at the base of the anther cells. These mucronate appendages are characteristic of the genus and are very easily seen with the hand-lens. Under the low power of the compound microscope they are found to be composed of several multicellular strands of cells closely compacted together. These strands of cells are easily torn apart, and a dentate or ragged appearance is given to the whole appendage. ‘This is doubtless what Nees saw when he described the appendages of D. quitensis as 2-3-toothed." А similar example was found in D. Schiedeana. On microscopic study, however, the so-called dentations were found to be nothing more than shreds of tissue torn away slightly from the compacted mass. The anther cells are usually parallel and oblong in shape. In the case of D. sagit- tata and D. maranhonis the cells are so disposed as to have a
[Vor. 15 16 ANNALS OF THE MISSOURI BOTANICAL GARDEN
sagittate appearance. In both the species mentioned the apex as well as the base is appendaged. Аз a rule the anther cells are glabrous but in the species D. trichanthera the anther cells are very pubescent. The mode of dehiscence is by a longitudinal slit on the side of the anther cell. The filaments are commonly pubescent.
Pistil.—There is little variation in the parts of the pistil. A disc is present beneath the ovary in all species. The ovary itself is two-celled, glabrous, and oblong. Little or no variation is found in the filiform pubescent style. Only the anterior lobe of the stigma is developed, and this lobe is usually linear and oblique with a flattened stigmatic surface. However, in D. hygrophiloides the stigma is curved, while in D. sagittata it is basally lobed. In D. maranhonis the stigma is reflexed.
Сарвше.--Тһе capsule of the genus is quite uniform. The constant linear, glabrous and four-seeded characters, combined with other diagnostic characters, help considerably in generic determination. Retinacula or hooked appendages on the median ridge of the valves hold the flat suborbicular seeds in place. When dry the seeds appear to have many soft, appressed hairs. These same hairs when wetted diverge, elongate, and become mucilaginous.
GEOGRAPHICAL DISTRIBUTION
The geographical distribution of the American species of the genus Dyschoriste offers very interesting problems. The accom- panying maps demonstrate very clearly that there are three distinct areas of distribution: (1) southeastern United States; (2) southwestern United States and Mexico ; and (3) South America.
Two species, namely, Dyschoriste oblongifolia and D. angusta, occur in the southeastern United States area. This area extends the width of the coastal plain from southern Virginia to southern Florida. The regions of distribution of the two species do not overlap. D. angusta is confined to the wet region of Dade and Palm Beach Counties in southern Florida, while D. oblongifolia extends northward through the remainder of the area, seeking the dry, sandy pine woods.
1928]
KOBUSKI—MONOGRAPH ОҒ DYSCHORISTE 17 ` M v їе | В % E . ^ “4 af “.. в X OE \ ( 7 з i - 1 ‘ ` тепкен => E a p^)». Ч а K ( | SALA : S MP У (c ау — A \ ! г ! ~ (4 5 р Н : 3 ча 145 4 Ири Хо / М Фа 3 ' : @ > “У ( р ~ ба МУ NN р ~ 3 5 5 СҮҮ 1 / \ 1 i я 1 1 / Aa А ; p Y 4 9” | | 1 i 4 4 MAI 2 7) d ! | : | р — Ф 4 еуі КАЗАК |Т АИК ЛАН 77707777 - 2 i ! > t 8 эе MR | A ОМ if Ae ea 1 | Бе , | H | | " x А of ПИ FT ’ ч 15 M И і } мді “> 2 7 КУ Ч Ни е - і 27 Е J ' | hi 5 s E | ~ | p , р КЕЙИ 8-1 Жл ККЕ М E ШТ E apu Pi +a! і | Ж | : E ғ; 5 FER 3 ! ERE 4 Е — 8 а à 8 в ; ч 24k H 3 aff ; ДЕ 3 Ж s / Ч в y i ® Е; х Я ЕЭ З
The second area extends from Texas and southern Arizona southward to the Isthmus of Tehuantepec and contains the
greater number of species.
Of the forty recognized species in
the genus Dyschoriste, twenty-one, or more than half, are confined
Map showing the geographical distribution of the North American species of Dyschoriste.
Fig. 1.
18
Vor. 15 ANNALS OF THE MISSOURI BOTANICAL GARDEN
os Т he K = ~ 2) а %%; = Ка 4 9 1) : р ;
1 ww
Т9 зора Боэчоненюы КЕКЕКХЕРЕКККІ ЖЕЛ
SARL |- № зра. 38 НЕ H cm
7» Б 722444
- СЗО Ира:
Fig. 2. Мар showing the geographical distribution of Ше South American species of Dyschoriste.
to this area. It is an interesting fact that the distribution of some species is almost coincident with the geological formation of the country. D. decumbens, which occurs on the plateau
1928] KOBUSKI—MONOGRAPH OF DYSCHORISTE 19
region between the Sierra Madre ranges, is an excellent example of this fact. D. hirsutissima extends from Sonora southward along the western slope of the Sierra Madre range to Oaxaca. Many species have a localized distribution only, D. Greenmanit, D. crenulata, D. saltuensis, and D. angustifolia being examples. Some of these localized areas are characterized by three or more species. Ап instance of this is а small region around Guadalajara in the state of Jalisco where four species are represented. Ап- other illustration of limited areal distribution occurs in the northern part of Oaxaca which harbors D. oaxacensis, D. angusti- folia, D. capitata, and D. hirsutissima. Many species, especially the localized ones, appear to be extremely едар с since they inhabit only regions near volcanoes. Тһе center of distribution in this second area falls within the region represented by the states of Puebla, Michoacan, and Mexico.
The third and last area, namely that of South America, com- prises more territory than either of the areas indicated above and includes the seventeen remaining species of the genus. Тће material examined in all cases was not very copious, hence an accurate range of geographical distribution of these species could not definitely be ascertained. Nearly all species appear to occur in isolated and limited areas, but the relationship between some of these areas indicates that a greater overlapping of areas would occur were it not for the paucity of herbarium material. Three species, namely, D. quitensis, D. ciliata, and D. repens, are found in Peru and the Andes of Ecuador. D. Niederleinii and D. humilis are found in Argentina. The other thirteen species inhabit northern Paraguay and southern Brazil, and it is here that the South American center of distribution occurs.
An unusual feature of the geographical distribution is the isolation of the areas defined. At present, there is no one species which connects up any two areas. There seems to be no satis- factory explanation to account for the absence of the genus between the Andes of Ecuador and the Isthmus of Tehuantepec in Mexico. Members of the genus may be found between these two remote regions, but until the entire Andean range has been explored more thoroughly from a botanical standpoint one would hardly venture an explanation of the marked discontinuous dis- tribution which the genus now presents.
(Vou. 15 20 ANNALS OF THE MISSOURI BOTANICAL GARDEN
The non-occurrence of the genus in the Mississippi Valley is equally surprising. Since the flora of this entire area is com- paratively well known, it is hardly possible that the members of the genus would be overlooked if they there existed. The only solution seems to be the possible age of the genus. Dyschoriste is probably a pre-glacial genus which, prior to the Oligocene period of the Cenozoic era, extended continuously across the southern United States. However, the Eocene and Oligocene seas encroached upon the United States in the present Mississippi Valley, thereby splitting the distribution areas of the genus into two parts.
PHYLOGENY
Because of the large number of closely allied species in the genus Dyschoriste it is quite necessary that the phylogenetic discussion of the group be made from a purely hypothetical standpoint. The fact that the discussion is confined to the American species alone seconds this consideration, since the eastern species of the genus exceed the American species in number.
On account of the three distinct geographical distribution areas, which have been discussed before, a tree method of illus- trating probable phylogenetic sequence proved unsatisfactory; hence the method used in the accompanying chart was devised finally to illustrate the apparent relationship of the species of the western hemisphere. This chart if superimposed on a map of the regions inhabited by the genus would coincide with the specific regional distribution.
It was felt reasonably certain that all species of the genus have evolved from a common ancestor designated in the chart as x From this ancestor, species and groups of species have evolved. One might ask why, since the species seem to be placed in definite groups, subgenera or sections have not been designated. This question was given much thought and consideration; it was felt finally, however, that on account of the relative uniformity of the essential morphological characters within the genus, except in the case of group II, no adequate basis exists for the designation of subgenera or sections
It may be observed that all the designated groups with the
1928] KOBUSKI—MONOGRAPH OF DYSCHORISTE 21
exception of group II originated from the common ancestor at approximately the same time. Group П, on account of the muticous character of its anther appendages, the extremely small leaves, the very small flowers, and the fruit characters, has been separated from the genus Dyschoriste and raised to the rank of a new genus which is considered as intermediate between the hypothetical type and Dyschoriste proper. In this genus, Apassalus,! containing three species, the species А passalus diffusus
Fig. 3. Phylogenetic chart of the species of Dyschoriste.
has reached the highest development in the reduction of the number of ovules to two, one being borne in each valve of the capsule. The other two species contain the four seeds which are characteristic of the majority of species in the genus Dys- choriste. Apassalus is confined to the islands of Cuba and Haiti and to, the southeastern United States.
'See Kobuski, C. E. А new genus of the Acanthaceae. Ann. Mo. Bot. Gard. 15: 1-8, pl. 1-2. 1928.
[Уоь. 15 22 ANNALS OF THE MISSOURI BOTANICAL GARDEN
Group I, involving ten species, is confined to the plateau regions of Mexico. In this instance D. decumbens, on account of its extended range and characteristic relation to all species con- cerned, is considered the base species. D. Lloydii and D. crenu- lata are species having vital characteristics similar to D. decumbens but differing sufficiently in minor characters to be considered direct descendants from the base species. A small group con- taining three species, D. Purpusii, D. Greenmanii, and D. Rosei, stands by itself. 'The highest development is reached in D. Greenmanii and D. Rosei in which cases the inflorescence has been reduced to a solitary flower at each node. АП species of the last group have very slender linear leaves.
Another branch from D. decumbens, as the chart illustrates, is the linearis-jaliscensis branch. Although separated somewhat in regional distribution these two species are closely allied through their flower and foliage structures and are undoubtedly derived from D. decumbens.
The species D. oblongifolia and D. angusta seem to be closely related to D. linearis; in fact, D. linearis was once considered а variety of D. oblongifolia. However, the two species under dis- cussion are confined to the southeastern United States in their distribution and, as the distribution map of North America shows, are not connected definitely with the Mexican-southwestern United States species. This suggests the probability that there may be a relationship between the two species in Florida and the genus Apassalus also found there and in the West Indies. The migration may have been northward through the Antilles, and a connecting link between the southeastern United States species and the southwestern United States-Mexico group may never have existed.
In group III, one finds an entirely different situation. Here we have four large, erect species, each showing a definite charac- teristic development toward advancement. Perhaps the highest development is found in D. hirsutissima which extends the length of the western Sierra Madre range and possesses а well-developed glandular pubescence. This is the only instance of this character in the whole genus. D. bilabiata and D. quadrangularis are close relatives but do not seem to have been derived from the D. hir-
1928] KOBUSKI—MONOGRAPH ОҒ DYSCHORISTE 23
sutissima line. Instead, they undoubtedly arose along parallel lines of development. In D. bilabiata we find a distinct dentation of the leaf. This group is also characterized by large, petiolate leaves, in some cases as long as ten centimeters, an unusual feature in the genus.
Group IV for the most part occurs in southwest Mexico, that is, the states of Oaxaca, Jalisco, and Michoacan.