Annals of the Missouri Botanical 3" Garden Volume XLII I 35 With 34 plates, 27 figures, and 60 maps Published quarterly at Galesburg, Illinois, by the Board of Trustees of the Missouri Botanical Garden, St. Louis, Mo. Entered as second-class matter at the post-office at Galesburg, Illinois, under the Act of March 3, 1879. Annals of the Missouri Botanical Garden A Quarterly Journal containing Scientific Contributions from the Missouri Botanical Garden and the Henry Shaw School of Botany of Washington University in affiliation with the Missouri Botanical Garden. Information The ANNALS OF THE Missoumi BOTANICAL GARDEN appears four times during the calendar year: February, May, September, and November. Four numbers constitute a volume. Subscription Price _______610.00 per volume Single Numbers ________ 2.50 each Contents of previous issues of the ANNALS оғ THE Missour!I BOTANICAL САКРЕМ are listed in the Agricultural Index, published by the H. W. Wilson Company. TABLE OF CONTENTS Selaginella rupestris and Its Allies Rolla M. Tryon A New Pellaea from South Africa Alice F. Tryon Studies on Asiatic Relatives of Maize Nalini Nirodi Lichenological Notes on the Flora of the Antarctic Continent and the Subantarctic Islands. I-IV Carroll W. Dodge and Emanuel D. Rudolph Three New Annonaceae from Panama... Robert E. Fries The Botanical Catalogues of Auguste де St. Hilaire... John D. Dwyer Comparison of Juniper Populations on ап Ozark Glade and Old Fields :Marion Trufant Hall Tassel Modifications in Zea Mays Norton H. Nickerson and Ernest E. Dale A New Species of Doryopteris from Surinam Karl U. Kramer and Rolla M. Tryon, Jr. A Revision of the Genus Celastrus Ding Hou The Food of a Hindu Village of North India (Reprinted from Bulletin No. 2, Bureau Statistics and Economic Research, United Provinces, Allahabad, 1937) Charlotte Viall Wiser, Ph.B., M.Sc. General Index to Volume XLII PAGE 1- 99 101-102 103—130 131-149 151-152 153—170 171-194 195-212 213-214 215—302 303—412 413—418 MISSOURI BOTANICAL GARDEN — STAFF Emeritus Director ORG МООКЕ Director Ерсак ANDERSON Assistant Director Носн C. CUTLER CARROLL W. DODGE, СЕОВСЕ B. VAN SCHAACK, Mycologist Acting Curator of Herbarium JuLian А. STEYERMA Honorary Research pepe FREDERICK С. MEYER, Ковевт E. Woopson, Ја., Senior Taxonomist Henry М. ANDREWS, Dendrologist Paleobotanist AricE Е. Tryon, Кола M. Try Research Associate Assistant Curator “at the јонм D. Dwyer, bari ren? Associate HucH i ades. NELL С. Horne Curator Museum of Librarian and Editor Sa Plants of Publications BOARD OF TRUSTEES residen Јонм $. ГЕНМАММ Vice-President DANIEL К. CATLIN Second Vice-President EUGENE PETTUS LEICESTER B. Faust RicHarp J. Lockwoop DUDLEY FRENCH Henry B. PFLAGER Henry Ниснсоск A. WESSEL SHAPLEIGH Ковевт Brooxincs SMITH EX-OFFICIO MEMBERS ARTHUR С. LICHTENBERGER, STRATFORD LEE MORTON, Бир of s Diocese of President i the му of Miss Science of St. Loui James F. Morr Етнам А. H. SHEPLEY, President of the Board of Chancellor of Washington Education of St. University Кахмомр К. Tuck Mayor Fi ха City of St. reet HucH C. Cutter, Secretary Uu Number 1 Annals of the Missouri Botanical Garden Selaginella rupestris and Its Allies 2; a = ‘Rolla M. ле ГС o e Е. Tryon 101-102 | Annals of the Missouri Botanical Garden А Quarterly Journal containing Scientific Contributions from the Missouri Botanical Garden and the Henry Shaw School of Botany of Washington University in affiliation with the Missouri Botanical Garden. Information The Аммлт$ or THE MissOURI BOTANICAL GARDEN appears four times during the calendar year: February, May, September, and November. Four numbers constitute а volume. | _ Subscription Price 2% $10.00 per volume | 250 eh Сойыл of previous issues cf the ANNALS OF THE Missourr BOTANICAL a are listed in the Agricultural Index, published by the H. W. Wilson Annals of the Missouri Botanical Garden Vol. 42 FEBRUARY, 1955 No. 1 SELAGINELLA RUPESTRIS AND ITS ALLIES* ROLLA M. TRYON, JR. These are small plants, hardly exceeding a few inches in height, yet they are a striking part of the vegetation of the semi-arid and subalpine regions of western North America, covering barren soil and festooning rock ledges. The forty-three species included in this study comprise the section Tetragonostachys of subgenus Selaginella. Thirty-two species occur in America, eight in Africa and Madagascar, and the remainder in Asia. Superficially they appear similar and nondescript but under magnification they reveal structures of beauty in design and symmetry which amply reward the student who takes more than a casual interest in them. Selaginella rupestris and its allies occupy a unique place in xeric ecology for they are vascular plants adapted to being completely desiccated and reviving a few hours after moisture becomes available. Study of their physiological response to arid conditions would be of interest. In a number of ways the group offers data of value in geographic and evolutionary studies. $. rupestris is of particular cyto- logical and geographic significance because it has both sexual and apogamous races, the latter occupying a much larger territory than the former. In this species, as well as the others, details of the life-cycle, especially the means of fertilization and of dispersal, are poorly understood. Aside from botanical collectors and browsing by deer in lean seasons, they are relatively unaffected by animals or man and the habitats which most of them occupy are relatively little affected by fire, lumbering, grazing or cultivation. Studies of the group may be of comparative value in relation to species in which the activities of man have played a critical role. The principal effort in the present study has been directed.toward a definition of the species and groups of species within a coherent framework of characters, to place the nomenclature on a firm foundation through reference to holotype material and to provide keys and illustrations to facilitate accurate identification. It is not to be expected that the keys and descriptions can be used by a student of the vascular plants without some introductory study of a few species, and in this *Issued March 24, 1955. (1) [Vor. 42 2 ANNALS OF THE MISSOURI BOTANICAL GARDEN is true of most genera. One can not successfully divorce the use of a monograph from specimens and it is not my intention that this treatment be wholly inde- pendent of such material. THE PRESENT CLASSIFICATION The taxonomy of the Selaginella rupestris group is beset with a number of difficulties which must be recognized and resolved before an adequate classification can be developed. The first, and perhaps most important, is the relative plasticity of the characters, not only of the vegetative parts, but of the strobili, sporophylls and spores as well. This does not mean that taxonomically useful characters do not exist but that considerable caution must be used in choosing those of specific value. Relatively few species possess distinctive traits; rather most of them must be defined in terms of a particular combination of characters. There is evidence that there has been a great deal of parallel development in the group and that the potential variations of different structures have been realized many times. This situation has made it troublesome to find the morphological coherence of a species and the differentiating characters. Another difficulty is the very numerous char- acters that present themselves for consideration. In many groups, students prob- ably feel the characters are too few and wish for more but a superabundance of them can be equally trying. The inadequate nature of most herbarium specimens has been a source of confusion for nearly all of the early collections and many contemporary ones are fragmentary. Complete mats of some fifteen species were collected in the field and notes were made on local variations of habit and habitat. With the information obtained from these studies, it was possible to understand the relation of herbarium fragments to the living plants. The minute nature of the characters has undoubtedly resulted in many of them being overlooked or misin- terpreted. For example, the deciduous setae of S. eremophila were not detected until many years after the species was described. А relatively high magnification is needed to see the structures accurately. For general observation of the leaves and sporophylls 30 diameters is recommended, 40 diameters for the smaller details. Inasmuch as the species must largely be defined in terms of combinations of characters rather than distinctive ones, a set of specific characters for each species does not exist by itself but only in relation to other species. One set of characters may distinguish a species from its close relatives, additional ones, or another set, may distinguish it from less closely related species. It is obvious, therefore, that before a species can be worked out, the groups of species must be defined to estab- lish the circle within which character comparisons will be made. The following method was used to form a basis for grouping the species into natural units. Some twenty-five presumed species of which complete material was available early in the study were scored for about thirty characters. These included some from all parts of the plant. Their value was not known although only qualitative ones were used. Each species was compared with the others for the number of 1955] TRYON—SELAGINELLA RUPESTRIS AND ITS ALLIES 3 characters in common. Those that showed definite affinities with each other were grouped together. Ап analysis was then made for characters which would dis- tinguish between the groups. After a certain amount of modification and rear- rangement, the series and their characters were established. Such a method has a number of limitations that might make the results obtained inaccurate but when used with discretion it undoubtedly has value. The final test of the classification was the integration of some fifteen species not originally included. These took natural places within the system in regard to their geography, morphology and phylogeny. Although many characters had been used by previous authors for the definition of species these were at first disregarded to remove all prejudice as to their possible value. The numerous available characters were reduced to a manageable number in two ways. It was assumed that lineal measurements should be avoided, if at all possible; that qualitative differences should be used for the primary characters of species. Quantitative characters of real value could be used as expressed by ratios. Structures that were variable on individual stems and within individual mats would not be reliable for taxonomic use and many characters were rejected for this reason. When the number of possibly useful characters was reduced to convenient propor- tions, several species were chosen to be analyzed for the characters of specific value. These were species such as S. cinerascens, S. Bigelovii, S. oregana and S. tortipila of which I had ample material and which have a distinctive morphology, a compact and natural range and a rather defined ecology. Each was examined to determine the structures that were sufficiently stable to be used as specific characters. The species were then compared with each other in terms of this final set of characters and it was found that the entire set was usable to express species differences. Such characters were then used to define the species within each series. Most of the species thus derived were found to have the following attributes: a set of suffi- ciently stable characters so that the “population” showed morphological cohesion; complete morphological distinction from other species; a degree of morphological cohesion and distinction that is qualitatively and quantitatively comparable to other related species; a natural geographic distribution, allopatric or largely so from related species; and a definite local ecology although this might be rather broad within the whole range. These characteristics were used as criteria for determining the category in difficult and perplexing cases. The "good" species of the section, therefore, defined themselves, their attributes were determined and the other species were defined comparatively in similar terms. Subspecific categories were defined in relation to the species. e category subspecies is used in S. arenicola where the major variants have not completely developed the properties of species. The variety is three species but only one of these may be considered a proper example of the use of that category. That is in 5. mutica where the two variants, although definite, have the characteristics of species poorly developed. In the case of S. Wightii the two varieties are recognized largely for practical reasons, the material available being [Vor. 42 4 ANNALS OF THE MISSOURI BOTANICAL GARDEN insufficient to establish the proper category. In S. densa a most interesting but quite unsatisfactory taxonomic condition exists. The three variants are quite dis- tinct morphologically and geographically over large areas in the northern part of the range of the species. Here they would certainly rank as subspecies and perhaps even as species although the characters that separate them are not quite as well developed as in other species in the series. But the varieties all grow in the large central portion of the range where they exhibit such complete intergradation in characters that only one highly variable species could be recognized. I have chosen to use the category variety for them, placing perhaps a bit more emphasis upon their intergradation than upon their distinctiveness. The species descriptions have all been taken directly from the keys and are uniform only within each series. All of the characters mentioned in the species key are included in each description. Although this is a departure from the usual procedure, it is necessary in order to emphasize the set of characters important within an evolutionary line. A number of characters used in this treatment are peculiar to the genus or at least not commonly used in vascular taxonomy. For these I have tried to use suitable descriptive terms or phrases often without intend- ing precise definition. I quite agree with Mr. Weatherby's views that: “Ап attempt to find other characters which, even if seemingly incapable of altogether definite statement, could be more easily seen and better relied upon, seemed worthwhile."! I have used many of the characters that he introduced and some new ones. Although it was not my original purpose to enter into the general classification of the genus in any way, it finally did seem necessary to give formal recognition to the major groups of species. The erection of the series has caused in turn some slight readjustments within the subgenus. These changes are presented in the synopsis of the subgenus. THE BACKGROUND OF THE STUDY L. M. Underwood was the first botanist to give serious attention to the Selag- inella rupestris group. His first paper, in 18982, clearly expressed the condition of the classification at that time: “Тһе variations of the species of Selaginella with many-ranked leaves have long been a puzzle to botanists... . Two clearly marked species from North America have been separated from the tangle already; .... there remain . . . . the widely varying forms that for the past forty years have found an unsatisfactory resting place under S. rupestris”? Underwood described a total of nine new species in unraveling "the tangle”; I maintain seven of them. He was in correspondence with Hieronymus at Berlin and they exchanged specimens of many species in the group. Hieronymus did the Selaginellaceae for Engler & Prantl's *Natürlichen Pflanzenfamilien'? and prior to and following this work described many new species in the genus. In the S. rupestris group ће рго- 1 Тошг. Arn. Arb. 25:409. 1944, Bull. Torr. Bot. Club 25:125—133. 1898. 01. 2 *Nat. Pflanz. 14:621-715. 19551 TRYON-—SELAGINELLA RUPESTRIS AND ITS ALLIES 5 posed twenty-eight species of which I recognize eleven. Hieronymus deserves credit for his general position that a multiplicity of species existed and for his attempts to work them out. His work suffered for reasons that were largely either beyond his control or else in harmony with the times. His material was scrappy and inadequate, sometimes hopelessly so. Also his concepts of the natural distribu- tion of species, particularly those of the United States, were not well developed. Finally, he relied largely upon quantitative characters of the leaves, setae and cilia, many of which have proved to be of little value. The considerable number of his species that are valid is probably due to the fact that he did have material of many species. The last worker to describe many species was Maxon? who published eight new ones from the United States. АП but three of these I recognize as valid and two of those three are recognized as varieties. Maxon not only described most of the new species remaining in the United States but in his treatments of Pteridophyta for various floras of the western United States he supplied keys for their identifica- tion. The work of Underwood and Hieronymus was largely consolidated and systematized by Maxon and a knowledge of most of the species of the United States was placed on a firm foundation by him. His knowledge and judgment of the species were excellent but probably because his work in the genus was primarily floristic he separated the species on the most convenient characters and did not develop a set of characters useful for the whole group. Also, he relied in his descriptions too strongly, in my opinion, on quantitative characters. Prior to the studies of Underwood, Hieronymus and Maxon, the species were poorly understood. The most common policy was to place all material under the name of Selaginella rupestris; a few authors segregated varieties or forms of it; a very few segregated species. This early work is reviewed although it did not play an important role in the classification of the group. The species named Lycopodium rupestre by Linnaeus? was, as many he treated, previously well known. Among several earlier figures of the species, that of illenius' is the best. It presents an excellent illustration of the plant from Virginia. Linnaeus combined Virginian and Siberian plants in the same species and this broad range was enlarged upon by later authors until finally S. rupestris was credited with essentially a world-wide distribution. At that time it include many diverse species. Beauvois® segregated Selaginella and other genera from Lycopodium in 1805. Не placed Lycopodium rupestre in his genus Stachy- gynandrum. Spring, Milde and A. Braun were the principal monographers of Selaginella in the nineteenth century and they recognized, to a greater or lesser extent, the variability within S. rupestris. Spring? grouped his material on the “Мон of them in Hedwigia © 290-320. 1900. 5 Мозг of them in Smiths. Misc. Coll. 725. 1920. 85р. РІ. 2: а 1753. THist. Musc. t. 63, fig. 11. 1741. "Pro. perte 101. 180 v. Mém. Acad. Roy. Belg. 24:57. 1850. [Vor. 42 6 ANNALS OF THE MISSOURI BOTANICAL GARDEN basis of habit, the rather short-stemmed northern species forming his S. rupestris var. borealis and the generally elongate and lax southern ones his var. tropica. Milde!? used an essentially geographic segregation, recognizing as formae of S. rupestris material from several countries and regions. Of the ten forms he lists, I recognize seven as species. A. Braun! also recognized variants of S. rupestris as they occurred in Africa, using some obvious characters of the leaves, cilia and setae. Primarily due to floristic work, by the time of Underwood's treatment the following species were usually recognized as distinct from S. rupestris: S. Dregei Presl, S. żortipila A. Br., S. oregana D. C. Eaton and S. echinata Baker. The present authority on the genus as a whole, A. H. G. Alston, has treated a number of species in the S. rupestris group, especially as they have occurred in his regional treatments. Outside of the United States and Mexico the species are rather few and he nowhere has had to deal with a sufficiently large number to seriously engage the problem of specific characters. His work may be ranked with that of Maxon, placing our knowledge of the species of South America, Africa, Madagascar, India, China and northeastern Asia on a firm foundation. It will be noted that the species of Mexico had not been treated. This was probably due to the feeling that there was not sufficient material available to make it possible to deal successfully with the many species suspected of being present. The first worker to accept this difficult problem was C. A. Weatherby!? 13. In his two papers he was the first, from the point of view of this treatment, to derive a general set of specific characters and to start grouping the species into natural units; he was also the first to understand the special nature of the S. rupestris group and to adapt his classification to it. Мг. Weatherby's publications on Selaginella epitomize rather well his publications in systematics as a whole. His output was not voluminous but he had a strong predilection for the poorly under- stood groups and in these his abilities resulted in treatments of basic and permanent value. It will be apparent to students of his papers that the present study is largely an extension of his work. This is fitting for my interest in the group stemmed from his. MorPHOLOGY AND Lire HISTORY Most of the species have prostrate, superficial stems that are variously disposed to form a mat of more or less distinctive type. The stems of a mat may be inter- laced and intricate as in S. Underwoodii (fig. 40) and S. cinerascens ог discrete as in S. densa (fig. 47). The branching may produce a very compact mat as in S. Watsonii (fig. 49) or a loose, open one as in $. mutica. Growing conditions affect the form of the mat so that it is not stable in most species. Edges of cliffs, steep slopes, boulders and seepage crevices modify the habit of growth. In addi- 10 11 19551 TRYON-—SELAGINELLA RUPESTRIS AND ITS ALLIES 7 tion, some species may have the central stems crowded and rather erect and the peripheral ones closely prostrate and widely creeping. In S. tortipila the two types were described as separate species. А similar condition exists in S. Hansenii although fortunately they were not segregated. In these species the size of the mat is in- creased by growth of the apical buds. The older portions of the stems die pro- gressively toward the apical buds and strobili. The dead portions may be rather evenly distributed throughout the mat or, especially in the species with discrete branches, they may be in the center or at one side (fig. 47). A strobilus-bearing branch dies completely since the strobili, with rare exception, are determinate. Vegetative apices are theoretically indeterminate although actually they do occas- ionally die. The species of Arenicolae, with rhizomes or basal branch buds, form various shaped clumps or mats depending upon the extent to which the substrate permits or hinders the growth of the rhizomes or basal buds. $. arenicola, which usually grows in open sand, has quite symmetrical mats. In the species with rhizomes, as S. Weatherbiana, the rhizomes grow forward, occasionally branching or producing erect, aerial stems. These aerial branches produce a branch-system on which strobili are eventually produced and a short time later the whole branch-system dies. Most of the apices die due to the determinate strobili produced although the vegetative tips also die. This may be due to the death of the portion of the rhizome beneath the aerial branches or because of the distance the apices have grown from the nearest roots. New aerial branches produced by the younger parts of the rhizome perpetuate the plant. The leaves of the rhizome are generally similar to those of the aerial stems (compare figs. 7 and 8) but are usually broader, thinner, have a shorter seta and cilia, have a less-developed dorsal groove and the living ones are pale tan ог a very pale green. $. arenicola and S. rupincola lack rhizomes but have short buds at the base of the erect aerial stems (figs. 1,5). These buds become active upon the death of the branch-system; each developing into a new branch- system. This results in a rather bushy habit unless, as is often the case in S. rupincola, growth is modified by rock crevices. The stems often branch dichotomously but this is not the predominant type of branching in any of the species. The branching pattern is formed by laterals variously subordinate to the main stem, distance between branches and the angle at which they depart from the stem. Although this is obviously different in some species the degree to which it may be modified by growing conditions makes it an unreliable taxonomic character. A branch will remain subordinate to the stem upon which it is borne until the two apical buds are sufficiently removed from each other and then it will grow more rapidly. This seems to occur at about the time the stem and branches are separated by the death of the older portions. Branches of the second and third order are also subordinate to their respective main stems. In three species, 5. utahensis, S. leucobryoides and $. asprella, the stems are fragile when dry and are easily broken even with careful handling. The anatomical nature of this character has not been investigated. [Vor. 42 8 ANNALS OF THE MISSOURI BOTANICAL GARDEN During periods of unfavorable growing conditions the stems become dormant. It is probable that many species have the ability to remain alive in the dormant ‘state for many months. A specimen of S. Watsonii was planted three months after it had been pressed and dried and it put forth new growth. Part of a mat of S. densa var. densa was planted six months after it had been collected and stored as an herbarium specimen and it also grew. Various changes take place when the stem becomes dormant. In many species the branch tips curl upward; in Eremo- philae they become involute. A considerable portion of the stem is involved and the dormant stems form ringlets in S. оғерапа. This curling of the stems recalls the behavior of S. lepidophylla and its relatives, the commercial Resurrection Plant. The leaves also change position, becoming more or less closely appressed to the stem. These changes are particularly marked in the strongly dorsiventral species in which the spreading lateral leaves almost completely enfold the upper ones. However, the erect upper leaves do not perceptibly change in position. In some species as S. tortipila the leaves are usually rather closely appressed in the growing state and change position very little when desiccated. The difference in position of the leaves of a species in the growing and dormant states is not sufficiently constant for gen- eral application. This behavior, however, has a distinct advantage to the taxonomist for it renders herbarium material as useful and as valid as living material for most characters of the plant. Conclusions as to the relation of living to herbarium material were confirmed by comparisons of specimens collected from species grow- ing in the greenhouse. There is no difference between the living dormant plant and the herbarium specimen prepared from it. Soaking the specimen in water will restore it in the same way as watering the living plant. Specimens prepared from actively growing plants will assume the characters of the dormant state upon drying with the exception that portions subject to pressure in pressing will be held in more or less the original position. I have not measured the growth of the stem of any species in its native habitat but measurements of material grown in the greenhouse indicate that it usually amounts to 1—4 cm. a year. During 100 days in the greenhouse, S. arenicola ssp. Riddellii grew 30 mm. and $. viridissima grew 15 mm. although this may not reflect the relative growth rates of the species in their natural habitats. A speci- men of $. fortipila (Hunnewell 9478, GH) appears to show annual growth of 13 mm. on some of the stems. The leafy stem is radially symmetrical in many species, that is, the leaves are disposed in the same manner on all sides of the stem at a given place and they are of the same size, shape, color and texture. This is true of the species that have erect stems and many of those that are prostrate. Most of the species with stems rather closely appressed to the ground are more or less dorsiventral. Extreme dorsiventral species such as S. Landii (figs. 62, 63) and S. echinata have the under leaves tightly appressed to the stem, the upper erect, the lateral ascending-curved, the under leaves longer than the upper, different in shape, thinner, tan or brownish, and the setae also differ. All intermediate conditions occur, some species having 1955] TRYON—SELAGINELLA RUPESTRIS AND ITS ALLIES 9 one or more dorsiventral characters so that there is not a sharp line of distinction between radial and dorsiventral symmetry. Leaves are the most useful part of the plant for purposes of taxonomy and a number of characters of the position, base, apex, shape, texture, cilia and seta have been used in this treatment. The base of the leaf, in some species, is abruptly adnate to the stem and distinct from it in color. In other species the leaf-base is decurrent, gradually blending into the stem in form and color. Red leaves often occur in 8. Sartorii, 8. Steyermarkii and 8. Hansenii, especially in the area between the oldest living leaves and the youngest dead ones. (fig. 23, the dark portions of the leaves are red). These are often a bright red or tinged with purple. They perhaps develop color in a manner similar to the sugar maple and other deciduous trees in northern regions. A few double leaves have been observed. The leaves are firm in texture with the exception of 8. oregana in which they are sufficiently soft and thick so that they partly collapse in drying. The usual sequence of leaf development appears as follows. At the beginning of the growing season relatively few green leaves are present at the apex of the stem but renewed growth of the apical bud produces new leaves and a conspicuous zone of green leaves develops. After a time the oldest of these die; those that were green at the start of the growing season die first. Under rather uniform conditions the death of the older leaves proceeds at a rather constant rate and probably at a rate similar to the production of new leaves so that a zone of green leaves of rather uniform length is maintained. During 100 days in the greenhouse, leaves died on 11 mm. of stem in 5. mutica and on 39 mm. іп 5. arenicola ssp. Riddellii. Toward the end of the growing season, or with the advent of a dry period, the leaves con- tinue to die, although no new ones are produced, until sometimes only a few green leaves near the bud remain. The apparent inhibition of growth of lateral branches has been discussed. In some species the leaves are affected in a similar manner. In $. cinerascens, for example, the leaves on the main stem are the longest and those on the primary, secondary and tertiary branches progressively shorter. However, once the branch apex is separated from the stem apex by the death of the older portions, the leaves that are produced are as large as those on a main stem. The strobilus is nearly always determinate but vegetative growth from its apex does occur. This is not uncommon in $. arenicola ssp. Riddellii and 1 have seen two other examples of the condition: 5. Х meomexicana (Slater 3, US) and S. arizonica (А. 8 В. A. Nelson 1158, MO). This condition was noted in S. densa planted in the greenhouse in a dormant condition that produced vegetative tips on two strobili upon renewal of the growth of the mat. I have seen one strobilus that was dichotomous (Soxman 351, US), а specimen of S. X meomexicana. In most species the sporophylls die progressively upward after the death of the leaves on the leafy branch beneath. In S. arenicola ssp. arenicola and ssp. acanthonota, and to some extent in $. Dregei and S. tortipila, the apical sporophylls die first and the basal last, and this occurs prior to the death of the leaves beneath so that the death of the strobilus is independent of the progressive death of the leaves. * [Vor. 42 10 ANNALS OF THE MISSOURI BOTANICAL GARDEN The sporophylls are hinged by their stalks to the strobilus axis and when the spores mature and the sporangia open they move outward presumably in response to their water content. This movement evidently aids in the dispersal of the spores. Although I am not certain, the two valves of the sporangium may also move. I have observed іп S. rupestris that megaspores will frequently remain in the basal sporangia after all of the others are shed, apparently because the leaves directly beneath do not allow the sporophylls to move sufficiently. Occasionally spores are found in dead strobili, and thus their liberation from the sporangium is not always effected I have seen sporelings in only a single collection and germinated megaspores were never observed on the specimens. Megaspores were germinated on filter paper in a petri dish and then were dried. The triradial split in the spore coat and the gametophyte were still evident; thus if germinated megaspores had occurred in the collections they could have been recognized. Lyon !* reports germination of the megaspores within the sporangium, and fertilization of the gametophyte prior to dispersal from the sporangium. I have observed nothing to support this but since the species do have natural ranges, many of them over considerable areas, it would seem that the normal life cycle is completed. Possibly germination and fertilization occur at rather rare intervals when the microclimate is particularly favorable. This may occur in the sporangium, as Miss Lyon observed, or after the spores have been shed. Since the stems of most species are quite tough and difficult to break, and are usually anchored at frequent intervals by the rhizophores and roots it does not seem likely that vegetative reproduction could account for the distribution of the species. In the group of species with fragile stems, S. utahensis, S. leucobryoides and S. asprella, it is not unreasonable to consider that fragments might be dispersed by the wind although actually they are all rather local endemics. Their restricted ranges, however, may be due to other causes and they may still possess an effective means of dispersal. In S. rupestris, which is apogamous throughout much of its range, fertilization is not a factor and the megaspores may be effectively dispersed immediately upon release from the sporangium. S, rupestris has the widest distribution of any species in recently available areas and its range may be taken to illustrate the effectiveness of megaspore dispersal. I have seen sporelings growing in the soil among the stems of a mat of S. sibirica (Calder & Billard 2005, MO). These are about 1 cm. long and inside the megaspore a foot can be seen from which are produced one to three roots and an erect stem. This stem does not bear leaves for the first third to half of its length. The first leaves are thinner, more elongate at the base and more widely spaced than the adult leaves. Those at the tip of one of the sporelings are of the adult type in aspect and disposition; the other two sporelings have only juvenile leaves. Two rhizophores had been produced on each sporeling, the first in the axil of the second or third leaf. 14 Вос. Gaz, 32:124-141. 1901. 19551 TRYON-—SELAGINELLA RUPESTRIS AND ITS ALLIES 11 The chromosome number has been determined in S. Weatberbiana (Tryon & Tryon 5077, 2n = 18), S. mutica (Tryon 8 Ттуоп 5073, п = 9, 2n = 18) and in S. Bigelovii (Tryon & Tryon 5054, n = 9). The sporophytic counts were made from the tips of rhizophores and the gametic ones from the contents of the mega- sporangia at meiosis. I am indebted to Dr. Amy Gage Skallerup for the consider- able work involved in making these counts. These numbers agree with the ones for S. selaginoides (S. spinulosa), S. belvetica and S. denticulata given by Miss Manton! as n = 9 PHYLOGENY The phyletic chart portrays the general relationships of the series and the species within them. "There is not sufficient evidence of the actual course of evolution to more than indicate in very general terms the origin of the series. The characters that are of particular importance in determining a primitive or advanced phyletic position are the nature of the leaf-base, abruptly adnate or decurrent, the presence or absence of rhizomes, stolons or basal buds and the symmetry of the leafy stem. Arenicolae is placed as primitive because it exhibits some characters of Section Selaginella, the most primitive section of the subgenus. The comparison is par- ticularly effective between S. arenicola and S. rupincola of Arenicolae and the two species of Section Selaginella, S. selaginoides and S. deflexa. These four species have erect strobilus-bearing branches and have rhizophores borne only at the base of the stem. S. rupincola is not entirely uniform in the latter character in that occasional prostrate stems may have rhizophores borne throughout. $. deflexa is similar to S. arenicola and S. rupincola in that the vegetative branches are rather small and resume active growth upon the death of the strobilus-bearing branch above them. Thus $. атетсоја and S. rupincola are the most primitive types in the section. This does not mean that they are actually ancestral to the species of the series or that they represent the transition from Section Selaginella to Section Tetragonostacbys; but rather that they retain the ancestral characters to a greater degree than the other living species. Eremophilae, on the basis of its leaf-base characters, is con- sidered to be the most advanced and Sarforii and Rupestres occupy а mid-position. Due to the limitations of space and the diversity of Sartorii it has not been possible to portray accurately the phyletic level of Rupestres. The radially symmetrical species of Sartorii are considered as more primitive than Rupestres, those with dorsi- ventral leafy stems and symmetrical strobili essentially on the same level and the species with dorsiventral strobili more advanced. It will be noted that the habit of growth is uniform in the Sartorii, Rupestres and Eremophilae and distinct from that in the Arenicolae. The leaf-base characters behave in a different manner; they are stable and distinctive in Sartorii, Rupestres and Eremophilae but are variable in Arenicolae. The series Sartorii and Rupestres are regarded as differentiating from Arenicolae or its ancestors, the characters of 15 Manton, I. Problems of cytology and evolution in the Pteridophyta. p. 259. 1950. [Vor. 42 12 ANNALS OF THE MISSOURI BOTANICAL GARDEN ` 5 * ` ее 28. DREGEI ` ` = à 27. PROXIMA T 24. CAFFRORUM 42. PARISHII ` ike 26, NIVEA і 23. NJAMNJAMENSIS A STROBILUS 41. EREMOPHILA i DORSIVENTRAL STROBILUS 9 ARISONICA 22. INDICA SYMMETRICAL 39. PERUVIANA 2tL.CARINATA -, 2 HANSENII pape ct 19. STEYERMARKII EREMOPHILAE о (LEAF ВАЗЕ ABRUPTLY ADNATE ON vol кш LEAFYSTEM | 18. WRIGHTII D UPPER SIDE OF STEM, STRONGLY горива DECURRENT ОМ UNDER BIER ум 55 22 е 2, (ne d n STEM RADIALLY 4 з РАСЕЛ к x. (4 X o oe ш, , Hus SYMMETRICAL 1 г ' * 16. SHAKOTANENSIS TEA Ч ы = 15. MUTICA 37. LEUCOBRYOIDES , 13. МАСКАТНЕКА , 14. ЕХТЕМЗА 36. m " 22-7 10. WIGHTII 12. ARSENEI 35. ge ! 33. DENSA ; 11. CINERASCENS о 9. SARTORII 34. ыз, 5 8. SELLOWII т 32. RUPESTRIS SARTORII (LEAF BASE ABRUPTLY ADNATE) 31. OREGANA CONTINUED GROWTH z 30. UNDERWOODII BY AERIAL STEMS 29. VARDEI | а «соо нар CONTINUED GROWTH 2 Y RHIZOMES, STOLONS RUPESTRES OR BASAL BUDS + 7. ТОКТЇР (LEAF ВАЗЕ DECURRENT) е es 6. VIRIDISSIMA , 4 5. WEATHERBIANA L4 4. BALANSAE я 2. BIGELOVII А 3. збі : UA. 1. RUPINCOLA ие ARENICOLAE , (LEAF BASE — ADNATE ; OR DECU ) Phyletic chart of subgenus Selaginella section Tetragonostacbys. 1955] TRYON—SELAGINELLA RUPESTRIS AND ITS ALLIES 13 the leaf-base becoming fixed in each series and the same type of growth developing independently. The Eremophilae were derived from the Sartorii by further special- ization of the leaf-bases. The development of the dorsiventral leafy stem has occurred independently in Sartorii and Rupestres. $. Dregei, with its radially symmetrical leafy stem and unilateral strobilus, poses a problem. Perhaps the radial symmetry is primitive and S. Dregei should be placed apart from S. echinata and $. proxima, or perhaps, as the chart implies, it may be derived from species with dorsiventral leafy stems. The evolutionary relation of the species is usually not known although in a few cases it is fairly clear that one is ancestral to another, or more likely, that the two have had a common ancestor. Within each series the most specialized species are considered advanced. The species that exhibit few if any special structures and which may have characters in common with a less advanced series are considered primitive. GEOGRAPHY AND ECOLOGY Section Tetragonostachys (Map 1) occurs in temperate and tropical North and South America; it is absent from the Amazon basin, Central America and, with few exceptions, the colder regions. Its distribution is not as extensive in the Old orld. In Asia it occurs in India and the Himalayan region and in Japan and Siberia. In Africa there is a single species in the region of the Atlas Mountains; the others occur south of the Sahara, with a better representation in eastern than in western Africa. ‘Three species grow in Madagascar. The section is notably absent from the Malaysian-Australian region, from the Pacific Ocean and from Eurasia. Series Arenicolae (Map 2) extends more or less laterally across the southern United States and adjacent Mexico with an isolated species in northwestern Africa —a distribution not unlike that of the genus Platanus. Sartorii (Map 3) has a range much like that of the whole section except for its absence from boreal regions and from northwestern Africa. Rupestres (Map 4) is predominantly North American with a single species in northeastern Asia and one in the Himalayan region. It is the only series that has a distribution of any extent in areas of Pleistocene continental glaciation. The ranges of S. densa (Map 43), S. sibirica (Map 47) and S. rupestris (Map 42) are partly in these areas. S. rupestris occupies more glaciated territory than the others and it is undoubtedly significant that there it is apogamous. Eremophilae (Map 5) is New World with essentially an Andean distribution. It is absent in the American tropics from northern Peru to southern Mexico. Most of the forty-three species of the section grow in the United States and Mexico. The two countries have a total of thirty species which represent all but two of the species of the New World. The United States has twenty-one species of which ten are endemics, while Mexico has sixteen species and five endemics. The southwestern United States—Texas to southern California—is the richest area [Vor. 42 ANNALS OF THE MISSOURI BOTANICAL GARDEN 14 A 5. SERIES EREMOPHILAE СЕТА {УРИ LAE| | |- И Т ad NE ЯЗ II aN З m SERIES ARENI PIT i1 | 3. SERIES SARTORII SERIES RL E dc 4 ECT OI TEIL ТТ € - 19551 TRYON—SELAGINELLA RUPESTRIS AND ITS ALLIES 15 with a total of seventeen species. 'Теп species, four of them endemic, grow in California and eight species in Texas. There are three species in Canada, two in Alaska, one in Cuba, one in Guatemala and four in South America. Africa has five species, four of them endemics, Madagascar three endemic species, India and the Himalayan area three species, two endemic, and northeastern Asia two species, one of them endemic. Most of the species have continuous ranges, minor discontinuities probably being due to insufficient collecting. The important and apparently authentic examples of discontinuous range are the following. Both S. Sellowii (Map 17) and S. Sartorii (Map 18) are disjunct between Mexico and northern South America and the former species has an isolated station in Cuba. 5. peruviana (Map 52) is similar but the disjunction is wider. S. Wightit (Map 19) grows in southern India and Ceylon and in Africa; S. rupestris (Map 42) has an isolated station in southern Greenland. Most of the species of each series are allopatric in their distribution and, with two exceptions, the relatively few cases of sympatric species are ecologically sep- arated. The examples I know in which two species of the same series actually grow together are S. densa and S. rupestris in Canada and S. peruviana and S. arizonica in Texas. The latter pair may sometimes grow in the same mat. There are two important ecological features of the species. They are mostly plants of dry habitats and they usually do not thrive in the presence of other plants. Consequently they are most abundant in regions of moderate rainfall, or in locally dry habitats in moist areas, and they grow in rocky, gravelly or sandy places, or on cliffs or barren soil where there is a minimum of other vegetation. One species, S. oregana, is usually an epiphyte and another, S. extensa, occasionally is. Many of the species grow in acid soil or on acid rocks; a few of them will also grow on basic rocks; а single species, 5. Wrightii, is definitely a calciphile. There is little evidence that they are important pioneers, being notably few in glaciated areas and in rocky places recently disturbed by fire or lumbering. Rather their chief role seems to be to fill the niche of locally transient but regionally permanent open, xeric habitats. The plants are not especially adapted either for the prevention of water loss or for its storage. The minor adaptations that are present in the form of fleshy an cutinized leaves are not sufficient to prevent frequent drying. The ability to survive desiccation undoubtedly resides in unusual physical and chemical properties of the cell contents. The extensive network of roots, borne close to the surface, enables the plants to utilize rapidly even small amounts of moisture. The species probably can grow in relatively moist places, as a few of them do, but because of their inability to compete with other plants, these habitats are not often available. They are restricted to the open habitats, which in a given area are usua ly the most xeric, where their special ability to survive desiccation allows them to be successful. ГУ ог. 42 16 ANNALS OF THE MISSOURI BOTANICAL GARDEN DISTRIBUTION MAPS A solid dot is used on the maps to indicate the location of a collection seen; in a few cases a circle is used to represent a specimen from a general region. In S. arenicola, half-dots are used to represent intermediates between the subspecies. Literature records, indicated by an X, have been used when they significantly aug- ment the range and when there is no doubt about the identity of the species. Except for S. Balansae, all of the literature records have been taken from the various papers of A. H. G. Alston. ACKNOWLEDGMENTS I am indebted to the curators of the herbaria of the following institutions for the generous loan of material Botanisches Museum, Berlin-Dahlem, British Museum (Natural History), Chicago Natural History Museum, Gray Herbarium, Department of Botany, University of Minnesota, New York Botanical Garden, Muséum National d'Histoire Naturelle, Paris, United States National Herbarium and Department of Botany of Yale University. The abbreviations for these and other herbaria are taken from Index Herbariorum, part I (Regnum Vegetabile, vol. 2). I wish to thank Mr. A. H. G. Alston and Mr. J. A. Crabbe of the British Museum, Mr. F. Ballard of Kew, Mme. Tardieu-Blot of Paris, Dr. H. Reimers of Berlin, and Mr. С. V. Morton of the United States National Herbarium for courtesies extended during the course of this study. SYSTEMATIC ACCOUNT SYNOPSIS OF SELAGINELLA SUBGENUS SELAGINELLA!$ SELAGINELLA Beauv. Prod. Aethog. 101. 1805, conserved name. Type species: Selaginella selaginoides (L.) Link. SUBGENUS SELAGINELLA (Section Homoeopbyllae Spring, in Mart. Fl. Bras. 1:118. 1840. Homotropae А. Br. Ind. Sem. Hort. Bot. Berol. 1857: Append. 11. 1857, without rank. Subgenus Euselaginella Warb. Monsunia 1:100. 1900. Subgenus Homoeophyllum (Spring) Hieron. in Engl. & Prantl, Nat. Pflanz. 1*:669. 1901.). Sporophylls uniform or similar, leaves uniform or gradually dimorphic. Type species: Selaginella selaginoides (L.) Link. SECTION SELAGINELLA (Cylindrostachyae А. Br. loc. cit., without rank. Section Cylindrostachys (А. Br.) Hieron. loc. cit.). Strobilus cylindrical, leaves spirally arranged. Type species: Selaginella selaginoides (L.) Link. SPECIES: 1. S. selaginoides (L.) Link, 2. S. deflexa Brack. 16Partly adapted from Walton & Alston, in Verdoorn, Man. Pterid. 503. 1938. БАК ТЫ лы, 2% s ы о 19551 TRYON—SELAGINELLA RUPESTRIS AND ITS ALLIES 17 Section Tetrastichus (A. Br.) Tryon, stat. nov. (Tetrastichae A. Br. loc. cit., without rank). Strobilus tetragonous, at least the lower leaves decussate. Type species: Selaginella pumila (Schlecht.) Spring. $РЕСГЕ$: 1. 5. pygmaea (Kaulf.) Alston (S. pumila (Schlecht.) Spring), 2. S. gracillima (Kze.) Alston (S. Preissiana Spring), 3. S. uliginosa (Labil.) Spring. SECTION TETRAGONOSTACHYs (A.Br.) Hieron. loc. cit. ( Tetragonostacbyae A. Br. loc. cit., without rank). Strobilus tetragonous, leaves spirally ar- ranged. Type species: Selaginella rupestris (L.) Spring. Series Arenicolae Tryon. Type species: Selaginella arenicola Underw. SPECIES: 1. S. rupincola Underw., la. S. meomexicana Maxon, 2. $. Bigelovii Underw., 3. 8. arenicola Underw., За. ssp. Riddellii (Van Eselt.) Tryon, 3b. ssp. arenicola, 3c. ssp. acanthonota (Underw.) Tryon, 4. S. Balansae (A. Br.) Hieron., 5. S. Weatherbiana Tryon, 6. S. viridissima Weath., 7. S. tortipila A. Br. Series Sartorii Tryon. Type species: Selaginella Sartorii Hieron. SPECIES: 8. S. Sellowii Hieron., 9. S. Sartori; Hieron., 10. $. Wigbtii Hieron., 10a. var. Wightii, 10b. var. Phillipsiana, 11. S. cinerascens A. A. Eaton, 12. S. Arsenei Weath., 13. S. macrathera Weath., 14. S. shakotanensis (Franch. ex Takeda) Miyabe & Kudo, 15. S. Wallacei Hieron., 16. $. mutica О. С. Eaton ex Underw., 16a. var. limitanea Weath., 16b. var. mutica, 17. S. extensa Underw., 18. S. Wrigbtii Hieron., 19. 5. Steyermarkii Alston, 20. S. Hansenii Hieron., 21. 5. carinata Tryon, 22. S. indica (Milde) Tryon, 23. S. njamnjamensis Hieron., 24. S. caffrorum (Milde) Hieron., 25. S. echinata Baker, 26. S. nivea Alston, 27. S. proxima Tryon, 28. $. Dregei (Presl) Hieron. Serres Rupestres Tryon. Type species: Selaginella rupestris (L.) Spring. species: 29. S. Vardei Lév., 30. S. oregana D. C. Eaton, 31. 5. Underwoodii Hieron., 32. $. rupestris (L.) Spring, 33. $. densa Rydb., 33a. var. scopulorum (Maxon) Tryon, 33b. var. densa, 33c. var. Standleyi(Maxon) Tryon, 34. $. sibirica (Milde) Hieron., 35. $. Wat- sonii Underw., 36. $. utabensis Flowers, 37. S. leucobryoides Maxon, 38. S. asprella Maxon. Serres Eremophilae Tryon. Type species: Selaginella eremophila Maxon. species: 39. S. peruviana (Milde) Hieron., 40. $. arizonica Maxon, 41. 8. и Maxon, 42. S. Parisbii Underw., 43. S. Landii Greenm. & Pfei [Vor. 42 18 ANNALS OF THE MISSOURI BOTANICAL GARDEN KEY TO THE SERIES OF SECTION TETRAGONOSTACHYS a. Stems erect or ascendent with rhizophores produced only at or near the base, rarely one or a few stems of a group prostrate with rhizophores pro- duced nearly throughout; subterranean rhizomes or ground-level stolons or basal branch buds ргеепе — . — — —— — Series Arenicolae, p. 18 a. Stems prostrate to decumbent, or irregularly ascendent, with rhizophores produced at or near the apex, at least of the primary stems (the branches may rarely be erect), or epiphytic and long-pendent with rhizophores pro- duced only at the base; rhizomes and stolons absent, basal branch buds rarely and irregularly present. b. b. Leaves with the base abruptly adnate and distinct from the stem in color!* on all sides of the stem, or especially on the branches, those on the main stem parely dechrment а. -- Series Sartorii, р. 31 b. Leaves with the base decurrent on all sides of the stem! or at least those on the underside strongly decurrent. c. с. Upper leaves with the base usually decurrent; leafy stem radially symmetrical to definitely dorsiventral, the zone of green leaves!? about equal on all sides of the stem; branch tips straight or slightly curled in the dormant condition.” Series Rupestres, p. 58 · Upper leaves with the base abruptly adnate, distinct from the stem in color (rarely a distinct stem-ridge may be present); leafy stem strongly dorsiventral, the zone of green leaves!? well developed on the upper side of the stem, very short to absent on the under side; branch tips involute in the dormant condition... Series Eremophilae, p. 76 о Series Arenicolae Tryon, ser. nov. Rhizomata vel gemmae brevia simplicia ad bases caulium aeriorum hae saepe stolones producentes praesentia. Caules erecti vel ascendentes. Apices ramorum recti vel leviter curvati statu inerte. Caules frondosi radialiter symmetricales. Folia base decurrente vel abrupte adnata cum caule distincta colore. Typus: Selaginella arenicola Underw. Plants terrestrial; rhizomes present, their leaves different from those of the aerial stem, thinner, with shorter setae, generally broader, with shorter cilia, with a less-developed dorsal groove, the living ones pale tan to a very pale green, or short, simple buds present at the base of the aerial stems, these sometimes elongated to form stolons, or both rhizomes and basal buds present. Stems erect or ascendent, with rhizophores produced only at or near the base, rarely one or a few stems (especially іп S. rupincola) prostrate with rhizophores produced generally through- out; stems short to moderately so, forming distinct groups or in the species with "То be observed between the first and third bran ®This character may be obscured by death of th ches back from the apex of the stem. e leaves during a long dormancy. 1955) TRYON-—SELAGINELLA RUPESTRIS AND ITS ALLIES 19 rhizomes usually forming tall, loose mats. Branches short to moderately long, usually remote, discrete or intricate; branch tips straight or slightly curved in the dormant state. Leafy stems radially symmetrical, the leaves equal in position, length, shape and texture on all sides of the same portion of the stem, rarely slightly dorsiventral in position on prostrate stems; zone of green leaves equal on all sides of the stem or rarely slightly longer on the upper side. Leaves with the base abruptly adnate and distinct from the stem in color, to decurrent; dorsal groove evident from base to apex, or in S. fortipila absent or poorly developed; setae, if present, relatively straight and stout, without a modified tip, persistent, or in S. tortipila with a filiform, tortuous, usually deciduous tip. Arenicolae is a homogeneous series although the species are quite distinct and show considerable diversity. The species are held together by the characters of the limited growth of the aerial branches and the rhizome or bud growth at the base of the branches. The leaf-base which is such an important character in the other series is variable here. In S. rupincola and S. Bigelovii, the only two species that show a close af- finity, the leaf-base is abruptly adnate to the stem and distinct in color. In $. Balansae, 8. Weatherbiana and S. viridissima it is decurrent toward the base of the branches and tends to be abruptly adnate toward the tip. In S. arenicola and $. tortipila the leaf-bases are decurrent. The least specialized species appear to be S. rupincola and 8. arenicola, while the most specialized, those with modified setae, are certainly S. fortipila and $. viridissima. Мо species shows an evident relation to a species of any other series. KEY TO SPECIES a. Leaves with the base abruptly adnate and distinct from the stem in color on all branches. b. Rhizomes qu cilia of the leaves of a branch-system usually predomi- nantly or entirely piliform, about У, as long as the width of the blade or longer, rarely only a few piliform; leaf-apex flat to rounded. Ari- zona to Texas, Tamaulipas to Sonora and southward... 1. S. rupincola, p. 20 Rhizomes present; cilia of the leaves of a branch system usually pre- dominantly or entirely dentiform, about % to № as long as the width of the blade, rarely some or most piliform and longer; веди usually carinate to broadly rounded. California and Baja California ....... 2. $. S. Бірге; p. 22 a. All or at least the basal leaves of the aerial branches and of the rhizomes (when present) with the base decurrent. c. c. Setae of the leaves stout, relatively straight, without a modified tip, persistent, or setae absent; dorsal groove evident from the base to the apex. d d. Leaves setate. e. BÉ [Vor. 42 20 ANNALS OF THE MISSOURI BOTANICAL GARDEN e. Rhizomes absent, short stolons occasionally present. Texas to Florida and North Carolina... — 3. S. arenicola, p. 25 e. Rhizomes present. f. f. Leaf-apex flat to slightly rounded; setae tawny to whitish. Northwestern Africa 4. S. Balansae, p. 27 f. Leaf-apex strongly carinate; setae lutescent to greenish-white. New Mexico and Colorado 5. S. Weatberbiana, p. 28 d. Leaves muticous 6. S. viridissima, p. 28 c. Setae of the leaves usually with a long, filiform, tortuous, often partly or wholly deciduous tip, rarely the tip only irregularly flexuous, or the setae without a modified tip; dorsal groove absent or developed only in the mid-portion of the blade, rarely moderately developed but not extending to the base or apex 7. S. tortipila, p. 29 1. SELAGINELLA RUPINCOLA Underw. in Bull. Torr. Bot. Club 25:129. 1898. (Lectotype: Wooton 124 NY! marked by Underwood as type although he did not cite it as such. Paratypes: Palmer 92 К; Toumey US!; Wright 2116 NY}, US! cited by Underwood as 2106). Figs. 1, 2. Map 6. Selaginella Chrismarii Hieron. in Hedwigia 39:299. 1900, as Chrismari. (Holotype: von Chrismar B!; Purpus 3156 Е, GH, MO, US is identical). Selaginella Chrismarii var. Karwinskyana Hieron. in Hedwigia 39:300. 1900. (Lectotype: Palmer 554 B! chosen because of its wide distribution although at least at GH and U it is mixed with S. peruviana, Paratypes: Karwinsky B!; Schaffner 11 B!). Selaginella Chrismarii var. Neeana Hieron. in Hedwigia 39:300. 1900. (Holotype: Née Selaginella rupestris (L.) Spring var. rupincola (Underw.) Clute, Fern Allies, 142. 1905. Rhizomes absent; aerial branches with buds present at their base, erect or ascendent, sometimes decumbent at the base, rarely one or some of the branches of a plant prostrate. Leaves with the base abruptly adnate, distinct from the stem in color; margins ciliate, the cilia usually piliform, V4 as long as the width of the blade or longer, rarely some or most leaves with the cilia predominantly dentiform and shorter; apex flat to rounded to rarely carinate; setae milk-white and opaque or sometimes with a greenish-lutescent base, to whitish-lutescent and translucent. Megaspores rugose to rugose-reticulate, rarely nearly smooth on the outer face, pale orange. S. rupincola is closely related only to S. Bigelovii; the two may be separated by the characters in the key. The long spreading cilia on the leaves are most char- acteristic but infrequently they may be rather short. In a very few specimens that I have seen the stems are prostrate but these may be distinguished from species of other series by the basal buds. А park supervisor in Arizona informed me that in times of drought this plant was browsed by deer. Dry open places, on ledges and in crevices of cliffs, on rocky slopes or in talus, or in gravel, usually from 1000 to 2000 m., rarely up to 3500 m. It is apparently confined to igneous rocks. 1955] TRYON—SELAGINELLA RUPESTRIS AND ITS ALLIES ZI 5 Texas to Arizona, south to Puebla. Representative specimens: UNITED STATES. TEXAS: Cory 48388 (GH); Warnock ты (СН, ^ 2003 р, МО). NEw MEXICO: Organ Mountains, Dona Ana Со., 5800 t 189 I24 (K, MO, NY, US); Wrigbt 2116 (sometimes = 2106) (GH, К "МО, МУ ГОЗ). ARIZONA: Ferriss 26 (NY, US); Maguire 10021 (GH, US); Santa Са Мой», April 5, и, Tago, (US); R. M. 8 A. F. Tryon 5050 (ARIZ, B, BM, DS, F, FI, GH, (BM, F, GH, MO, US). баня Соми и; 1848, von bots (B). JALIsco: Nov. 16, 1930, M. E. Jones (ВМ, GH, МО, US). DISTRITO FEDERAL: S eder. II 2. СН, US). PUEBLA: Purpus 3156 (BM, Е, СН, МО, US). la. SELAGINELLA Х NEOMEXICANA Maxon, in Smiths. Misc. Coll. 72°:2. 1920, as S. neomexicana. (= S. mutica X rupincola). (Holotype: Wooton іп 1909 US! fragment GH!. Paratypes: Wooton in 1902 US!, in 1904 US!, in 1907 US! fragment MO!). Map 7. The recognition of S. Х пеотехісапа as a putative hybrid is based on its abortive spores and intermediate morphological characters. The sporangia and [Vor. 42 22 ANNALS OF THE MISSOURI BOTANICAL GARDEN spores are variously abortive in all of the specimens examined. Мо mature mega- spores have been seen and in only one collection (Wooton in 1904) have I seen microspores of a size that might indicate maturity. These are probably also abortive, however, since they average some 22 p in diameter, while the microspores of $. rupincola and S. mutica!? range in size from 38 to 64 д and from 30 to 53 р respectively. Although abortive sporangia and spores might be due to a number of causes, specimens with this character are intermediate in various degrees between the two proposed parental species. The erect to ascendent branches with buds at their base are characters of S. rupincola while the commonly linear-lanceolate leaves usually with a strongly carinate apex suggest S. mutica. Most specimens are rather inter- mediate but some, such as Wherry in 1925, are rather close to S. mutica while others, such as Wooton on January 9 and on September 28, 1909, are close to S. rupincola. The distribution of the hybrid is within the range of both parent species and it is known to grow with both at some localities. On open or shaded igneous rocks at about 1000—2000 m. Texas to Arizona. Какай зреситепз: р STATES. TEXAS: Slater 2, 3 (US); о 5557 apis бойна 351 (05); ap 26, 1925, Wherry (US); Hinckloy 3404 (GH, US). мех M gs March 3, 1907, oton ; MO, US); Organ Mountains, Dona. Ana Co., (600 Та. э, 1905, Wo ooton (GH ‚ 05); Sept. 28, 1902, Wooton (NY, e Sept. HS ber Wooton (US). ARIZONA: Paradise, Cochise Co., 1904, Ferriss (GH, US). N SELAGINELLA BicELovm Underw. in Bull. Torr. Bot. Club 25:130. 1898. (Holotype: Bigelow NY!. Paratypes: Coville & Funston тог NY!; Henshaw NY! US!; Parish & Parish 671 NY!; Underwood NY!). Fig. 3. Map 8. Rhizomes widely creeping; aerial branches with buds often present at their base, erect, rarely some decumbent at the base. Leaves with the base abruptly adnate, distinct from the stem in color; margins ciliate, the cilia usually dentiform, about % to % as long as the width of the blade, less often some or most of the leaves with the cilia piliform and longer; apex strongly carinate to broadly rounded, or predominantly so; setae usually milk-white and opaque, to whitish, lutescent- or greenish-white, translucent. Megaspores rugose to rugose-reticulate, rarely nearly smooth on the outer face, lemon-yellow to pale orange. S. Bigelovii is closely related to S. rupincola from which it differs by the pres- ence of rhizomes in addition to the less important leaf characters mentioned in the key. It has been reported to be browsed by cattle in times of drought.?9 This species usually grows in open places, on serpentine, sandstone or igneous rocks, in crevices or shallow depressions, on cliffs, at the base of boulders or in so Тгуо on, А. F. in Ann. Мо. Bot. Gard. 36:418, 419. 1949: Wiggins, I. L. in Amer. Fern Tour 22:15. 1932; 19.55] TRYON-—SELAGINELLA RUPESTRIS AND ITS ALLIES 23 gravelly soil, less often in shaded or moist situations or in clay soil; it occurs from about sea level up to 2000 m. California and Baja California. Representative specimens: UNITED STATES, CALIFORNIA: Abrams 3126 (GH, MO, NY), 3402 (Е, СН, MO, NY US); Lt. Whipple's Exped., March 16, 1854, Bigelow (GH, NY, US, м Clokey 5165 (F, GH, NY); Coville & Funston ror (GH, NY, US); Fosberg 697 (F, MO, NY, US); Heller 7276 (GH, МО, NY, US), 7610 (F, GH, MO, NY, US), ог (Е, ВН, МО, , , ) ; ы 434 in 1875 (Е, MO, МҮ); Parish 9 Parish 671 (Е, МХ); К. М. 8 А. Е. Tryon 5053 deer MO, NY, P), 5054 (F, K, MO, UC), 5056 (B, СН, MO, US); Jan. 1889, Under- woo Mexico. BAJA CALIFORNIA: Ferris 8470 (NY, US); Wiggins 4212 (F, GH, NY, US). 3. SELAGINELLA ARENICOLA Underw. in Bull. Torr. Bot. Club 25:541. 1898, based on S. arenaria Underw. Map 9 Rhizomes absent; aerial branches with buds present at their base, which are sometimes elongated to form stolons, strictly erect, ascendent to decumbent at the base. Leaves with the base decurrent, blending into the stem in color or sometimes rather abruptly decurrent on the apical portion of the stem; margins rarely eciliate, usually ciliate, the cilia piliform, about М, as long as the width of the blade, less often shorter or dentiform; apex nearly flat to carinate; setae usually milk-white to tawny and opaque, rarely whitish with a lutescent base and translucent. Mega- spores rugose, rugose-reticulate to rugose-tuberculate on the commissural face, less prominently marked to smooth on the outer face, white to pale orange. The three subspecies differ in characters of essentially specific value but there are too many intermediates to allow them full recognition. These characters are those mentioned in the key to subspecies. In addition, the subspecies also tend to differ in various leaf characters but these are too variable within each subspecies to Бе of taxonomic value. Ssp. Riddellii has the longest leaf-base and ssp. acanthonota has the shortest; ssp. acanthonota has the longest setae and ssp. Réiddellii the shortest; ssp. acanthonota has the stoutest leafy stem with the most leaves; ssp. Riddellii has the most slender stem with the fewest leaves; and ssp. acanthonota has the most cilia on a leaf while ssp. Riddellii has the fewest. Subspecies arenicola occupies an intermediate position in the range of variability of all of these characters. Subspecies Riddellii is evidently the most primitive of the subspecies. The manner of death of the strobilus is the same as that in all other species of the series and its spores are less distinctive. Ssp. acanthonota, with aerial rhizophores, is the most advanced. S. crenicola differs from S. rupincola, the other member of the series that lacks rhizomes, in having the base of the leaves, at least on the basal portions of the stems, decurrent rather than abruptly adnate. It grows in open habitats or in light shade, in sand or in weathered soil pockets of acidic rocks. It probably has considerable ability as a colonizer; at least I have seen it in Florida growing on the graded sides of relatively recent roads. North Carolina to Texas. [Vor. 42 24 ANNALS OF THE MISSOURI BOTANICAL GARDEN KEY TO SUBSPECIES a. Aerial branches erect; all rhizophores subterranean, branched into roots near the base. b. b. Basal sporophylls of a strobilus die after the leaves beneath the strobilus; outer face of the megaspores usually rugose | | | За. ssp. Riddellii, p. 24 b. Basal sporophylls of a strobilus die before the leaves beneath the strobilus; outer face of the megaspores usually smooth ______36. ssp. arenicola, p. 26 a. Aerial branches erect to decumbent at the base; all or many of the rhizo- phores aerial and unbranched toward the base; basal sporophylls of a strobilus die before the leaves beneath the strobilus; outer face of the megaspores usually smooth 3c. ssp. acantbonota, p. 26 3a. SELAGINELLA ARENICOLA ssp. Riddellii (Van Eselt.) Tryon, comb. nov. Map 10. b ow Riddellii Van Eselt. in Contrib. U.S. Nat. Herb. 20:162. 1918. (Holotype: ow 7 US!. Paratypes: Drummond 352 NY}; Holmes & Fetherolf US!; Jermy - МОЇ, 051; Lindbeimer 76 MO!; Long in 1900 МУ, in 1901 NY!; Plank on Aug. 1 1892 NY!, on Aug. 18, 1892 ES Revercbon 1632 MO!, NY!, US!; Riddell 16 NY Thurow US!, 8 US!). Basal buds sometimes elongated to form stolons; aerial branches erect; rhizo- phores subterranean, branched into roots near their base. Strobilus usually borne on a long leafy branch, often with vegetative apical growth; basal sporophylls dying after the leaves beneath the strobilus. Megaspores with the outer face usually rugose to rugose-reticulate, rarely smooth. The following material is intermediate between ssp. Riddellii and ssp. arenicola (Мар 10): Louisiana: Dorman іп 1935 (СН, US); Wherry in 1937 (US). ALABAMA: Harper 37 (GH, MO, NY, US), 3716 (MO, NY, US); Wolf in 1933 (US); Leeds in 1934 (NY); Gattman in 1938 (Е). GEORGIA: Harper 1800 (Е, СН, МО, NY, 05); Pyron & McVaugb 3051 (US); Duncan 8683 (МО). АП of the material cited above from Alabama is from Eight-mile Creek and St. Bernard, Cullman Co., and has most of the megasporangia abortive. This is suggestive of тоя perhaps with S. rupestris which is known from Franklin Co., but it may be due to other causes. Subspecies Riddellii grows in open areas or in open woods, in sand or in crevices or shallow depressions of granite, sandstone or conglomerate outcrops, or at the base of boulders Alabama to Оби and Texas. Representative specimens: Омітер STATES. АТАВАМА: Harper 3403 (GH, МО, NY, US), 3762 (GH, MO, МУ, US). Louisiana: Red River, Hale (GH, US). ОКТАНОМА: 1928, Draper & Gould (US); Waterfall 6426a (GH). ‘texas: Cutler 3220 (GH, MO); Drummond 352 (GH US); Jan. 1903, Holmes & Fetherolf (US); Jermy 342 (MO, US); Lindheimer 76 ( MO, US) ; March, 1900, W. H. Long (NY); Feb. 1901, W. H. Long (NY); E. J. Palmer 9540 (МО, US), 13411 (MO, US), 33364 (GH, MO); Aug. 10, 1892, Plank (NY); Aug. 18, 1892, Plank (NY); Reverchon 1632 (F, GH, MO, NY, US); Riddell 16 (NY): 19551 TRYON-—SELAGINELLA RUPESTRIS AND ITS ALLIES 25 T SERT A9. до Y E „= INTERMEDIATE! RIDDELLII - ARENICOLA — 11. S. ARENICOLA SSP. S) (Мог. 42 26 ANNALS OF THE MISSOURI BOTANICAL GARDEN Oct. 26, 1915, Thurow (GH, NY, 2 near Prairie View, Waller Co., Jan. 5, 1911, Thurow 7 (US); Thurow 8 (US); К. M. 8 A. Е. Tryon 5030 (B, GH, MO, US), 5033 (BM, F, P, MO, NY, US). 3b. SELAGINELLA ARENICOLA ssp. arenicola. Fig. 4. Map 11. Selaginella arenaria Underw. in Bull. Torr. Bot. Club 25:129. 1898, not Baker 188 i? (Lectotype: Underwood 1355 NY! was marked by Maxon as type as to six of the cim Perret Underwood ы 355a NY!; Chapman NY!. Nasb 1449 is excluded as а type, it is ssp. acanthonota). Basal buds not forming stolons; aerial branches erect; rhizophores subterranean, branching into roots near their base. Strobilus usually borne on a long leafy branch, very rarely with vegetative apical growth; basal sporophylls dying before the leaves beneath the strobilus. Megaspores with the outer face smooth to rarely rugose. Specimens that are intermediate with ssp. Riddellii have been mentioned under that subspecies. The following specimens are those intermediate with ssp. acanthonota (Map 11): GEORGIA: Pyron t$ McVaugh 3101 (MO, US), 3114 (US), 31758 (US). rronmma: A. A. Eaton in 1903 (US); Underwood 1355 (NY in part, US). Subspecies arenicola grows in open places among shrubs in white sand or rarely in crevices or on flat exposures of sandstone or granite. Georgia and Florida. Representative specimens: UNITED STATES. GEORGIA: A. Н. Curtiss 6714 (GH, MO, NY, US); gat 1854 (МО, NY, US), 1860 (Е, GH, MO, NY, US); Pyron 8 McVaugh 3048 ( ЗУ (МО, US); Tracy 3510 (МО, US). FLORIDA: 1840, Chapman (NY); July, Bu Padi (NY, US); Tryon 5014 (B, BM, GH, МО), 5016 (F, MO, NY, P, US), 5017 (K, MO, UC); Eustis, Lake Co., Jan. 14, 1891, Равене 1355 (NY, as го specimens marked x), 13554 (МҮ). У 3с. SELAGINELLA ARENICOLA ssp. acanthonota (Underw. ) Tryon, comb. nov. Fig. 5. Map 12. guo ird bres ко in Torren 2:172. 1902. (Holotype: Williamson NY! ragm І, иті Selaginella r nn d катары» уаг. dcn (Underw.) Clute, Fern Allies, 142, 264. "ROSE funiformis Van Eselt. in Proc. Biol. Soc. Wash. 30:161. 1917. (Holotype: Mobr US!. Paratypes: Chapman, Biltmore Herb. 3432b US!; Huger NY!; Small 8 Carter 1013 GH!, МУ!; Small З Wilson 1762 NY!; Small, Dots & Small 3349 NY!). опорно ee Van Еген. in Contrib. U. S. Nat. Herb. 20:165. Me not Hieron: Holotype: Nash 1449 051. Paratypes: Eaton СНІ; Rapp U Seid Hondas Maxon in Amer. Fern Jour. 11:1. 1921, baa on X bumifusa Van Basal buds not forming stolons; aerial branches erect to usually decumbent at the base; rhizophores mostly or entirely aerial and unbranched toward their base. Strobilus borne on a very short to moderately long leafy branch, without vegetative 19551 TRYON—SELAGINELLA RUPESTRIS AND ITS ALLIES 27 apical growth; basal sporophylls dying before the leaves beneath the strobilus. Megaspores with the outer face smooth. Specimens intermediate with ssp. arenicola have been discussed under that sub- species. In the several leaf characters mentioned in the discussion of $. arenicola the ones characteristic of ssp. acanthonota tend to be more distinctive in the north- eastern portion of the range, particularly in North Carolina. Ssp. acanthonota grows in open places among shrubs in white sand, rarely on sandstone. North Carolina to Florida. Representative specimens: UNITED STATES. NORTH CAROLINA: Blake 11481A (СН, US); Curtis (NY); Heller 14101 (Е, MO); Wilmington, July, 1892, Williamson (NY, US). SOUTH CAROLINA: Hermann 10008 (Е, GH, MO, NY, US). ceorcia: Harper 1852 (Е, GH, MO, МУ, US), 1957 (F, GH, МО, NY, US), 1987 (GH, MO, NY, US); Hermann 10128 (F, NY, US); Tryon 5027 (B, ВМ, GH, К, MO, P, UC, US). FLORIDA: Chapman (Biltmore Herb. US), 17 (F, МО, NY, US); Jan.-Feb., 1902, Huger (NY, US); Carabelle, March 15, 1898, Mohr (NY, US); vicinity of Eustis, Lake Co., July 16-31, 1894, Nash 1449 (F, GH, MO, NY, US); E. Palmer 660 in 1874 (GH, MO, US); Sept. 1902, S. Rapp (NY, US); Small 8 Carter 1013 (Е, GH, NY, US); Small, Carter & Small 3349 (NY, US); Small 8 Wilson 1762 (NY, US); Tracy 7554 (Е, GH, MO, NY, US); Tryon 5012 (BM, DS, GH, MIN, МО), 5021 (В, BM, Е, GH, К, MO, МУ, P, UC, US), 5026 (ARIZ, MICH, MO, RM, US), 5003 (CU, FLAS, MO, PH, POM, US, WS). 4. SELAGINELLA BALANSAE (A. Br.) Hieron. in Hedwigia 39:318. 1900. Fig. 6. Map 13. Selaginella rupestris (L.) Spring var. Balamsae A. Br. in Kuhn, Fil. Afr. 212. 1868, egory taken from annotation of type by A. Br. (Holotype: Balansa B!. Peters, Mozambique, is excluded as a type; it is S. Dregei). Selaginella rupestris ssp. Balansae (A. Br.) [incorrectly attributed to Hieron. by] Jahand. & Maire, Cat. Pl. Maroc. 1:11. 1951 Rhizomes short, creeping; aerial branches with buds present at their base which are sometimes elongated to form stolons, erect or ascendent, rarely some decumbent. Leaves with the base decurrent and blending into the stem in color on the basal portion of the stem, rather abruptly adnate on the apical portion; margins ciliate, the cilia dentiform, about № or less as long as the width of the blade; apex nearly flat to slightly rounded; setae tawny to whitish, subopaque to translucent. Mega- spores rugose on the commissural face, less prominently rugose-reticulate on the outer face, pale yellow to pale orange. The unusual distribution of the series Arenicolae results from the inclusion of this North African species. However, the range of the series is not unlike that of the genus Platanus. Superficially S. Balansae bears a close resemblance to 8. rupin- cola but in technical characters it is probably closest to S. arenicola. The localities taken from the literature (Map 13) are from Jahandiez and Maire, Cat. Pl. Maroc. пио 1831; This species grows in exposed rocky places at 200-1200 m. [Vor. 42 28 ANNALS OF THE MISSOURI BOTANICAL GARDEN Morocco and Spanish Morocco. Representative specimens: Morocco: Gorges де l'Oued-Ghaghaia, au sud de Maroc, June 3, 1867, Balansa (B, K, P, NY, US); De Retz 31681 (MO); 1908, Gandoger (MO); 1935, Gattefossé (US): 1871, Hooker (NY); R. Maire 1533 (P). SPANISH Morocco: Quer I (BM). 5. SELAGINELLA WEATHERBIANA Tryon, in Amer. Fern Jour. 40:69. 1950. (Holotype: Standley 4558 МО!. Paratypes: Drouet & Richards 3337 МОБ Ewan 14370 МО!; Macbride 2651 МО!; Standley 4158 US!, 4626 MO! NY! US!; Underwood NY!; Underwood & Selby 2 МУ!). Figs. 7, 8. Мар 14. Rhizomes widely creeping; aerial branches with buds rarely present at their base, erect or ascendent, rarely some decumbent. Leaves with the base decurrent and blending into the stem in color on the basal portion of the stem, less decurrent, sometimes rather abruptly adnate on the apical portion; margins ciliate, the cilia dentiform, about 1% as long as the width of the blade; apex carinate; setae lutescent- to greenish-white, translucent. Megaspores rather coarsely rugose- reticulate, rarely nearly smooth on the outer face, pale yellow to pale orange. The rhizomes, the strongly carinate leaf-apex and the greenish-white or lutescent setae are characteristic of this species. It often grows in close association with S. Underwoodii and, largely on the basis of fragmentary material, was long confused with it. Selaginella Weatberbiana grows on exposed or shaded rocky hillsides and cliffs, in crevices and on ledges of igneous rock at 2000—3200 m. Colorado and New Mexico. Representative specimens: UNITED STATES. coLorapo: Ewan 14370 (MO); Macbride 2651 (MO); R. M. 8 A. F. Tryon 5077 (ARIZ, B, BM, CU, DS, F, FI, GH, K, MICH, MIN, MO, NY, P, PH; POM, RM, UC, US, WS); Sept. 1901, Underwood (NY); Underwood & Selby 2 (NY). NEW MEXICO: Arséne 17817 (US), 17080 (US); Drouet 8 Richards 3337 (F, MO); , 0 ft., July 25, 1908, Standley 4558 (F, MO, 58 (US), 4626 (MO, NY, US), 14692 (F, US). 6. SELAGINELLA vIRIDISSIMA Weath. in Jour. Arn. Arb. 24:326. 1943. (Holo- type: Stewart 2204 СНІ. Paratype: Johnston 8683 СНІ). Fig. 9. Map 15. Selaginella Соту: Weath. in Amer. Fern Jour. 36:51. 1946. (Holotype: Cory 44831 GH! ragment US!). Rhizomes widely creeping; aerial branches with buds rarely present at their base, erect or ascendent, occasionally some decumbent or prostrate. Leaves with ciliate, cilia dentiform, % to l6 as long as the width of the blade; apex broadly rounded to carinate, muticous. Megaspores rugose, pale orange. the other species of the series. An additional character is the very slender stems. 19551 TRYON—SELAGINELLA RUPESTRIS AND ITS ALLIES 29 From additional material it is evident that 5. Согу? can not be maintained on the basis of habit; both the Mexican and Texan material have erect or ascendent branches and some prostrate ones in the same mats. S. viridissima forms large, loose mats, rarely partly pendent, on shaded or locally sheltered cliffs and ledges of igneous rock. Southwestern Texas здає Coahuila. Specimens examine UNITED States, TEXAs: Correll 13627 (MO); Chisos Mountains, Brewster Co., July 10, 1944, Cory 44831 (СН, US) ; May 24, 1928, Murray (Е); В. M. 8 A. Е. Tryon 5015 (ARIZ, x BM, DS, F, FI, GH, K, MICH, MO, NY, P, PH, RM, UC, US); Warnock 21694 (U COAHUILA: Jobnston 8683 (GH); Kenoyer 8 Crum 2620 (GH); Cafion de H). Mexico. Calabasa, north wall of Sierra Maja, Oct. 27, 1941, Stewart 2204 (G ~ 15. S. VIRIDISSIMA 2; 7. SELAGINELLA TORTIPILA А. Br. in Ann. Sci, Nat. V. 3:271. 1865. (Lecto- type: Rugel B!. Paratype: Curtis B!). Fig. 10. Map 16. . В atic (L.) Spring var. tortipila (A. Br.) Underw. Our Nat. Ferns, ed. 3, шы Sherwood Underw. in Torreya 2:172. 1902, (Lectotype: 5 Sherwood in 1902 NY!. Paratypes: Donnell-Smith in 1882 NY!; comm. Sherwood in A 1902 NY!; Sherwood іп 1901 NY!. Donnell-Smith in 1881 NY! is excluded as a type, it cor- ponds to the type of S. tortipila). Selaginella rupestris var. Sherwoodii (Underw.) Clute, Fern Allies, 142. 1905. [Vor. 42 30 ANNALS OF THE MISSOURI BOTANICAL GARDEN Rhizomes widely creeping; aerial branches with buds sometimes present at their base, erect or ascendent, occasionally some decumbent. Leaves with the base decurrent and blending into the stem in color; dorsal groove absent or developed only in the mid-portion of the blade; margins eciliate to ciliate, the cilia dentiform to piliform, up to % as long as the width of the blade; apex carinate to strongly carinate; setae with a filiform, tortuous to irregularly flexuous, often deciduous, tip, rarely lacking a modified tip, lutescent- to greenish-white, translucent to sub- opaque. Megaspores rugose to tuberculate, rarely smooth on the outer face, pale yellow to pale orange. The two distinctive characters of $. tortipila, the poorly developed dorsal groove and the tortuous, filiform seta tip, are somewhat variable. Rare y on some leaves of a stem the dorsal groove may be quite well developed or the setae may be nearly straight and stouter than usual. The strobili are usually very short, about 5 mm. long or less. The basal sporophylls of a strobilus usually die after the leaves beneath; occasionally they die before. $. arenicola is the only other species in the series in which the sporophylls die before the leaves beneath the strobilus, this being a character of ssp. атетсоја and ssp. acantbonota. I am indebted to Dr. A. J. Sharp for adequate collections of this species showing the rhizomes, a character only suggested by the many other collections available. Wherry?! pointed out, on the basis of field observation, that specimens de- scribed and identified as S. Sberwoodii were portions from the center or the upper part of a mat, or from mats growing in unusually exposed situations. The type and similar specimens of S. fortipila, on the other hand, represent material from the outer or lower portions of a mat or from a mat growing in a shaded place. S. tortipila usually forms large, compact mats on exposed granite or granite- gneiss cliffs or rocks; less often in thin soil in depressions of rocks or in shaded localities. It grows from 900-1500 m. in North Carolina to as low as 90 m. in Georgia. North Carolina to Georgia. Representative specimens: UNITED STATES. NORTH CAROLINA: D. $. & Н. B. Correll 7897 (GH, US); 1882, Donnell-Smith (GH, MO, NY, US); 1898, Huger (B, NY); Hunnewell 9478 (GH, US), 9479 (GH, US); Oosting 3640 (MO); Broad Кізег,22 July, 1841, Rugel (B, NY, US); Highlands, Macon Co., 1902, Sberwood (NY); April, 1902, comm. Sherwood (NY). SOUTH CAROLINA: Curtis (B); 1881, Donnell-Smith (Е, СН, МО, МУ, US). GEORGIA: Duncan et al. 9084 (GH, MO); Hermann 10108 (Е, СН, MO, NY, US); Pyron 8 Mc- Vaugb 2056 (US). 23 Jour. South. Appalach. Bot. Club 1:65—69. 1936. "Wherry, E. T. in Jour. South. Appalach. Bot. Club 1:68. 1936, says that this species grows се ню Broad River only at Chimney Rock, Rutherford Со. Presumably Rugel made his collec- tion there. 1955] TRYON-——SELAGINELLA RUPESTRIS AND ITS ALLIES 31 Sers Sartorii Tryon, ser. nov. Rhizomata et stolones nulla, gemmae breves simplices ad bases caulium rare praesentes vel absentes. Caules prostrati vel ramis rare ascendentibus vel erectis. Apices ramorum recti vel involuti statu inerte. Caules frondosi valde dorsiventrales vel radialiter symmetricales. Folia base abrupte adnata cum caule distincta colore vel folia caulium primariorum rare decurrentia. Typus: Selaginella Sartorii Hieron. Plants terrestrial or, in S. extensa, sometimes epiphytic; rhizomes and stolons absent; short, simple, basal buds present only in 8. Dregei, and then rarely. Stems prostrate with rhizophores produced generally throughout or, in $. Dregei, often irregularly ascendent, or in S. proxima and S. Wigbtii the primary branches tending to be erect or ascendent and dendroid with rhizophores few or absent among the branches; stems moderately long to long, or in S. macrathera and S. carinata, short, or in S. extensa, very long; usually forming loose, open, less often compact mats. Branches long to short, remote to approximate, usually intricate; branch tips straight or, in the dorsiventral species, involute in the dormant state; the older primary branches of the main stem (excluding dichotomous branches) once to three, rarely four times pinnate. Leafy stems radially symmetrical, the leaves equal in length, shape, position and texture on all sides of the same portion of the stem, to strongly dorsiventral with the under leaves appressed, the upper erect, the under longer than the upper and different in shape and texture; zone of green leaves equal on all sides of the stem to much longer on the upper side in the dorsi- ventral species. Leaves with the base abruptly adnate, distinct from the stem in color, or those on the main stems slightly decurrent; apical portion plane or in- curved or, in S. extensa, revolute; setae straight or nearly so, or in S. Dregei and S. proxima, strongly curved, entire, or in the upper leaves of S. echinata, stellate or substellate. Strobilus with the sporophylls in four ranks, radially symmetrical or dorsiventral, those on the under side longer and broader than those on the upper side or, in 8. Dregei, with the sporophylls in two ranks on the under side of the branch tip. This is the most heterogeneous series consisting of three species groups that in a more elaborate classification might be given formal recognition. The least special- ized and presumably most primitive group, $. Sellowii to $. extensa, has the leafy stems and strobili radially symmetrical. The first three species, 5. Sellowii, 8. Sartorii and S. Wightii, are quite generalized in their characters, the others are specialized in one or more ways. The next group, S. Wrightii to S. indica, has dorsiventral leafy stems and radially symmetrical strobili. All of the species are specialized in some character. The last and most marked group, S. njamnjamensis to S. Dregei, has the leafy stems radially symmetrical or dorsiventral and the strobili dorsiventral. These are probably the most highly evolved species in the section and among them S. echinata of Madagascar stands as the most specialized. character of the dorsiventral strobilus has been observed very rarely in S. mutica of this series and in S. Parisbii in Eremophilae. [Vor. 42 32 ANNALS OF THE MISSOURI BOTANICAL GARDEN Nearly all of the species of the series are quite distinctive and many of them are exceptionally so. The few close relations that afford а critical separation are S. Sellowii and S. Sartorii, S. njamnjamensis and S. caffrorum and perhaps S. ciner- ascens and S. Arsenei. The following specimen may represent a new species of the Sartori but the material is sterile and thus inadequate: M'pala ad litt. occid. lacus Tanganyika, 1898, Rev. Guillemé 8 (P). It may bea variant of S. Wightii var. Phillipsiana to which it is evidently related. Its most distinctive character is the margin of the leaf that is smooth or bears a few minute, dentiform cilia. An annotation label by Hieronymus notes it as a new species bearing the name of the collector. KEY TO SPECIES а. Sporophylls in four ranks, the strobilus radially symmetrical. New orld, Asia, 1 species in eastern Africa. b b. Leafy stems radially symmetrical or nearly so, or dorsiventral only in the position of the leaves, the upper and under leaves equal in length and shape or very nearly so on the same portion of the stem. c. c. Apical portion of leaves essentially plane or incurved. d. d. Apex of the blade plane or gently beveled in profile, flat to rounded. e. e. Cilia, especially of the leaves of the main stems, piliform, or dentiform only toward the apex of the blade; or if all cilia dentiform then the setae milk-white, opaque; or leaves eciliate. f. f. Leaves definitely setate. g. 5. Older primary branches once to rarely twice pinnate; apex of the leaves usually convexly acute to acuminate. New World. h. h. Leaf-bases glabrous; setae usually milk-white, opaque, if whitish to greenish-lutescent and subtranslucent (Brazil) then slightly attenuate. Mexico, Cuba, Ven- ezuela to Argentina and Brazil 8. S. Sellowii, p. 34 . Leaf-bases predominantly pubescent, or if rarely mostly glabrous then the setae strongly attenuate; setae whitish to greenish to lutescent, translucent to subopaque. Mexico, Colombia and Мепехчеја_______9. S. Sartorii, p. 5. Older primary branches twice to three times pinnate; apex of the blade long-acuminate. India and Africa |... 10. S. Wightii, p f. Leaves, especially of the growing tips of the secondary branches, muticous or with a slightly — apex. Southern California and Baja California 2 . S. cinerascens, р. 40 e. СШа of the leaves dentiform; setae ber] or greenish- to lutescent-whitish, translucent to subopaque. Mexico. i. Е" w ON v ~ 1955] TRYON—SELAGINELLA RUPESTRIS AND ITS ALLIES 1. Setae about 14 or less as long as the blade... 12. S. Arsenei, p. i. Setae % to usually % as long as the Ыа4е_13. S. macratbera, р d. Apex of the blade predominantly truncate to abruptly beveled in profile, especially those of the growing-tips, broadly rounded to carinate. j. j. Apex of the sporophylls carinate; leafy stem somewhat dorsi- ventral in the position of the leaves. Тарап 14. S. shakotanensis, p. Apex of the sporophylls rounded to broadly rounded, or if carinate then the leafy stem quite or essentially radially sym- metrical. k. k. Leaves ligulate to ligulate-long-triangular, definitely setate. — . . Montana to British Columbia and California.15. 5. Wallacei, р. k. Leaves oblong- to ovate-lanceolate, muticous or nearly so in Colorado and Utah, setate зошћуага_______16. S. mutica, p. с. Apical portion of many leaves revolute. Mexico. .....17. S. extensa, р. b. Leafy stems strongly dorsiventral, the under leaves with the blade defi- nitely longer than the upper on the same portion of the stem, or dif- ferent in shape, or in both length and shape. 1. Sporophylls long-ciliate toward the base; under leaves linear; setae arising very evenly in form and usually in color from the blade-apex or the leaves muticous 18. S. Wrigbtii, p l. Sporophylls eciliate to short-ciliate toward the base. m. m. Setae arising rather abruptly in form and ids from the blade- apex. New World. n. n. Apex of the blade flat to rounded; some red leaves usually present, especially among the under ones. o. o. Apex of the blade of the upper leaves long-acuminate; under leaves papyraceous. Guatemala and eru Mexico. . S. Steyermarkii, р o. Apex of the blade of the upper leaves acute to convexly acute; under leaves herbaceous. California... 20. S. Hanseni, p. n. Apex of the blade carinate; no red leaves present. Ecuador... 21. S. carinata, p. m. Setae arising evenly in form and usually also in color from the blade-apex; under leaves papyraceous. India... 22. S. indica, p. a. Sporophylls in two ranks, or in four ranks and the strobilus dorsiventral, the sporophylls of the under side longer and broader than those of t upper side. Africa and Madagascar. p. р. Sporophylls in four ranks; longest cilia of the leaves up to У; as long as the width of the blade. q. q. Setae whitish to tawny or lutescent, entire. Africa. r. ~ > ~ = > “- ~ + ч + „49 49 50 52 [Vor. 42 34 ANNALS OF THE MISSOURI BOTANICAL GARDEN . Under leaves herbaceous, the base abruptly adnate, distinct from the stem in color; setae poorly differentiated from the blade-apex in form and color 23. S. njamnjamensis, р. Under leaves papyraceous to papyraceous-herbaceous, the base often decurrent, blending into the stem in color on the main stem; м мл ~ ^d setae predominantly whitish, rather abruptly distinct from the blade-apex in form and color 24. S. caffrorum, p. 5 w q. Setae milk-white, opaque, entire, or those of the upper leaves, espe- cially at the branch-tip, whitish, substellate, soon deciduous. Mada- gascar. s. s. Setae of upper leaves substellate, soon deciduous_25. S. echinata, p. 54 s. Setae entire, persistent. t. t. Setae straight, V4 to № as long as the blade ___26. S. nivea, p. 5 t. Setae strongly curved, № to У» as long аз the blade____ 27. S. proxima, p. 5 p. Sporophylls in two ranks; мра cilia of the leaves У; as long to as long as the width of the blade; setae usually strongly curved. Africa 28. S. Dregei, p. 57 MA о w SELAGINELLA SELLOWI Hieron. in Hedwigia 39:306. 1900, maintained over S. Arecbavaletae and S. montevideensis by Alston, in Physis 15:252. 1939. (Lectotype: Sellow in 1821 B! fragment NY!; Ule 232 US is nearly identical. aratypes: De Moura 008 B; Rio de Janeiro, Саш сђана B; Riedel 7 В; Schwache 949 B; Lbotzky 7 B). Fig. 11. Map 17. на нны (L.) Spring f. amazonica Milde, ЕП. Europ. Atlant. 263. 1867. Holo- mboldt 8 Bonpland ВІ; Peru, Galathea Exped. US 619317 is an excellent id бік рани f. brasiliensis Milde, Fil. Europ. Atlant. 263. 1867. (Lectotype 821 B! ragment NY!. Paratype: Sellow B! fragment NY!. Moritz D вы as а type, it is S. Sartorii). Selaginella rupestris var. brasiliensis AME) Hieron. in Engl. Bot. Jahrb. 22:417. 1896, t S. brasiliensis (Raddi) A. B Selegiuella montevideensis Hieron. in H in gia 39:309. 1900. == Sellow бр B! fragment NY!; Gibert 1323 has been асан at МУ as а close match, Рагасуре Gaudichaud B!). ЕЕ amazonica (Milde) Hieron. in Hedwigia 39:310. 1900, not $. amazonica Selaginelle Arechavaletae Hieron. in = 39:311. 1900. (Holotype: Arechavaleta 472 B! fragment NY!; Herter, Pl. Urug. 2077 is a close matc і Mildei Hieron. in „Tag & Prod. Маг. Pflanz. 15:671. 1901, based on S. amazonica (Milde) Hi pig cue pauciciliata Hira. in Hedwigia 58:284. 1916. (Holotype: Wright 1820 B!, ‘20’ cited by Hid is actually 1820 as indicated by an extra label on the det at G Stems with the older primary branches once to rarely twice pinnate; leafy stems radially symmetrical to rarely somewhat dorsiventral in position, length and shape of the upper and under leaves. Leaves herbaceous, none red, subulate-long- 19551 TRYON-—SELAGINELLA RUPESTRIS AND ITS ALLIES 35 triangular to broadly ligulate to lanceolate; base abruptly adnate, distinct from the stem in color or sometimes rather decurrent on the leader stem, glabrous; margins eciliate or usually ciliate, the cilia usually piliform, dentiform only at the apex of the blade, occasionally all dentiform, the longest cilia about % to rarely У; as long as the width of the blade; apex usually acute to convexly acute, some- times acuminate, slightly to definitely rounded, plane to rather abruptly beveled in profile; setae arising evenly or abruptly in form and color, V, to nearly № as long as the blade, slightly attenuate, milk-white and opaque to greenish- or whitish- lutescent and subopaque to translucent. Sporophylls with the margins eciliate to moderately long-ciliate toward the base; apex rounded. Megaspores rugose to rugose-reticulate on the commissural face, rugose-reticulate to smooth on the outer face, yellow to orange-yellow. This species and the next, S. Sartorii, are not well separated by distinctive morphological characters although in the regions where they both occur, Mexico, Colombia and Venezuela, the distinctions are adequate. Each is sufficiently vari- able, however, so that they overlap in the total range of variation of the characters. I would regard the two as something more than subspecies and yet less than species as judged by the attributes of the other species of the series. However, since they are clearly separable it seems best to give them specific status. S. Sellowii is char- acterized by glabrous leaf-bases and milk-white, opaque setae, while S. Sartorii has pubescent leaf-bases and whitish to greenish-lutescent and translucent to sub- opaque setae. Some Brazilian material of S. Sellowii is exceptional in having setae characters like S. Sartorii and rarely some Mexican material of S. Sartorii will have glabrous leaf-bases. S. Sellowii is composed of three major variants all of which intergrade rather freely. The value of the characters and particularly their degree of correlation is less than in S. mutica, for example, where varieties are maintained. For these reasons, the variants are not accorded a status. The most common variant has the leafy stem radially symmetrical, the leaves narrowly long-triangular to long-triangular with the apex plane in profile, with few or no cilia and with the setae milk-white and opaque. This ranges from Bahia, Brazil, to Argentina and Bolivia with slight variants extending northward to Mexico and Cuba. The material from Ecuador and northern South America has the leaves piliform-ciliate; that from Mexico and Cuba has the leafy stem more slender. The type specimens of S. pauciciliata, S. Sellowii, $, Arechavaletae and S. montevideensis are this variant although those of the last three are somewhat intermediate with the next. The second variant is usually slightly dorsiventral; the leaves are broadly ligulate to ligulate-lanceolate with the apex slightly beveled in profile, with num- erous cilia and with the setae whitish and often subopaque. It grows from Rio Grande do Sul, Brazil, to Argentina and Peru. The type specimen of 5. Mildei belongs here. [Vor. 42 36 ANNALS OF THE MISSOURI BOTANICAL GARDEN The third variant is similar to the second but the leafy stem is very slender; the leaves are not as broad and the setae are greenish- to whitish-lutescent and translucent to subopaque. It is known from the state of Rio de Janeiro, Brazil, and from Rio La Plata (Selding, GH). It is evident that Hieronymus essentially based S. Sellowii on S. rupestris f. brasiliensis, the epithet brasiliensis having been previously used at the specific level. However, he did not do it precisely so I am choosing the same type specimen for both names. This species and S. Sartorii are both quite generalized in their char- acters and probably represent the basic New World type of this series. The localities taken from the literature (Map 17) are from Alston, in Physis 15:253. 1939. On exposed or wooded rocky bluffs and cliffs or among stones, less often in sandy or clay soil, usually at about 1000—2000 m. but ranging from sea level to 0 m. ; Central Mexico; Cuba; Venezuela to Colombia, Argentina and eastern Brazil. Representative specimens: арда Mayari, Holquin, Aug. 17, 1859, 1860, С. Wrigbt 1820 (В, СН, К, MO, NY, ; XICO. DISTRITO FEDERAL: Козе & Painter 68 35 (US). PUEBLA: Arsène 801 (US), 10639 (US). veracruz: Liebmann 2062 (US). VENEZUELA: Tamayo 298 (US). Cotomsia: Lehmann KKo7 (US), BT776 (US). Ecuapon: Jameson 426 (US); Sodiro 47/1 (В) Peru: Bües 608 (US); Herrera 3009a (US); т ripam fluminis Магайоп, Humboldt 8 е (В). IVA: Balls 6229 (US); Eyerdam 25041 (Е, GH); Steinbach 82 Е, СН, МО), Мар (Р.С GH, МО); В. $ . Williams 1400 po (05). is E ARGENTINA: қысын 20454 (US); Burkart бор (СН), 3308 ЖӨН), 7128 (СН), 10113 (МО); Burkart & Troncoso 11174 (МО); Lossen 246 (Е Т. Меует ise (GH), 3687 с US); Schulz 753 (СН); L. B. Smitb 4679 qu а 1144 (СН US), 1866 (СН, US). PARAGUAY: Fiebrig 4907 (СН, US); Hassler 12535 (US); Jorgensen 4073 (F, NY); Lindman А2310 (US). Urucuay: Montevideo, June, 1876, Arechavaleta 472 (В, NY); Montevideo, | Gaudichaud (B); Gibert 1323 (NY, US); Hee 7611 B (GH, МО, 78856 (Е), 04884 (МО (MO); Montevideo, Sellow 670 (B, Вклтп.: Archer 6 Gebrt 118 (US); 0% 14 (08 X Glaziou 3306 (US); Jürgens 29 (NY), 308 (US), 309 (US); Dunes A213 (US); Litzelberg 18870 (NY); Riedel 7 (GH); Rose & Russell 19937 (NY, US); Sellow (B, NY), Praia de San Diego, 1821, Sellow d. р Ule 445 d Wilkes, U.S. So. Pacific Expl. Exped. (GH, NY, US, # (GH). 9. SELAGINELLA ФАКТОКИ Hieron. in Hedwigia 39:304. 1900. (Lectotype by Weath. in Jour. Arn. Arb. 25:418. 1944: Sartorius ВІ. Paratypes: Liebmann in Aug. 1841 B; Schiede in May, 1829 B). Fig. 12. Map 18. Selaginella Sartori var. oregonensis Hieron. in са 2. 305. 1900. (Holotype: “Lyall” B! — NY!, collector and locali Selaginella Sartorii var. venezuelensis Hieron. in Hedwigia 39; 305. 1900. (Holotype: Moritz 370 ВІ). 1955] TRYON-—SELAGINELLA RUPESTRIS AND ITS ALLIES 5 : 0 " LJ 5 ? -.. гі е 4 4-2 : ~ Ра dh ue Epi UE S ues 37 [Vor. 42 38 ANNALS OF THE MISSOURI BOTANICAL GARDEN Selaginella rupestris (L.) Spring var. Sartorii (Hieron.) Frye, Ferns Northwest, 31. 1934. Selaginella porrecta N'eath. in Jour. Arn. Arb. 25:416. 1944. (Holotype: Pennell 17198 fragment СН!. Paratypes: Bartlett 10355 US!, 11039 US!; Orcutt 1008 US!, 1142 US!, 11514 US, 5125 US!; Ortega 5310 US!; Pennell 17753 US!; Tharp 1792 US!. Schaffner 933 YU! is excluded as a type; it is S. peruviana). Selaginella Hintonii Weath. in Jour. Arn. Arb. 25:418. 1944. (Holotype: Hinton 8423 GH! fragment US!). Stems with the older primary branches once to rarely twice pinnate; leafy stems radially symmetrical except in the position of the leaves, or very slightly dorsi- ventral in the position and shape of the leaves. Leaves herbaceous to herbaceous- papyraceous, often some red, ligulate to ligulate-long-triangular; base abruptly adnate, distinct from the stem in color or sometimes slightly decurrent on the leader stems, usually pubescent, rarely glabrous; margins ciliate, the cilia piliform, or dentiform only toward the apex of the blade, the longest cilia V4 to У» as long as the width of the blade; apex acuminate to convexly acute, flat to slightly rounded, plane to gently beveled in profile; setae arising rather evenly in form but less so in color, 1⁄4 (rarely less) to nearly У» as long as the blade, slightly to strongly attenuate, whitish to greenish to lutescent, subopaque to translucent. Sporophylls with the margins short- to rather long-ciliate toward the base; apex slightly rounded to carinate. Megaspores prominently and irregularly rugose to rugose- reticulate on the commissural face, more prominently marked to nearly smooth on the outer face, yellow to orange-yellow. S. Sartorii is not strongly separated from S. Sellowii. The Venezuelan specimens of S. Sartorii, for example, are identified by the pubescent leaf-bases, a character that is usually variable within a species. They also differ from the S. Sellowii of that area, however, in the color and opaqueness of the setae. The relations of the two species are discussed in greater detail under S. Sellowii. Some of the Colombian and Venezuelan specimens such as Lindig 1523, Amór- tegui AOI and Tamayo 1753 have some of the leaves with a rather decurrent leaf- base. Hinton 3453 (BM) has the leaves almost entirely dentiform-ciliate but otherwise it agrees with S. Sartorii. Like S. Sellowii, this species is a complex but it is not as variable nor are the extremes as well marked. I am unable to maintain S. Hinfonii and S. porrecta on the basis of characters of specific value in other species in the series. Without more adequate collections it is not clear whether or not they represent geographic variants. The type specimen of S. Sartorii var. oregonensis at Berlin incorrectly bears a Lyall label; the collector and source of the material are not known. Exposed cliffs and rocky slopes of igneous rock, sandstone or limestone at 870—2000 m. Mexico, Colombia, Venezuela. Representative specimens: 1955] TRYON-——SELAGINELLA RUPESTRIS AND ITS ALLIES 39 Mexico. TAMAULIPAS: Bartlett 10355 (US), 11039 (US); Meyer 8 Rogers 3112 (МО); Stanford et al. 2187 (MO). NUEVO LEON: Orcutt 1008 (US), 1142 (US), 11514 (US); "Alamar", Pablillo, southeast of ба 1650-1700 m., July 2-3, 1934, Penneil 17108 (СН, US); Tharp 1702 (US). зам LUIS POTOSI: Orcutt 5125 (US); Pennell 17753 (US). DURANGO: Ortega 5310 (US ). MEXICO: Ypericones, Temascaltepec, Sept. 7, 1935, Hinton 8423 (GH, US). veracruz: Mirador, T (B); 1841, Liebmann (GH); — 8463 (GH, MO, US). oaxaca: Andrieux 2 (GH); Мое; 1566 (СН). COLOMBIA: ак Абі (US); Гтаг 1523 (СН); dete (YU ние Рагато de Mucuchies, Moritz 370 (В, К, YU); Tamayo 1753 (US), 1773 (US). | 10. SELAGINELLA У\Уснти Hieron. in Hedwigia 39:319. 1900. (Lectotype: Wight r5 B!. Paratypes: Walker В; Wight 283 B, 3104 ВІ). Map 19. Stems with the primary branches tending to be ascendent and dendroid, the older primary branches twice to three times pinnate; leafy stems radially sym- metrical or nearly so. Leaves herbaceous to herbaceous-papyraceous, none red, long-triangular to ligulate-long-triangular; base abruptly adnate, distinct from the stem in color, glabrous; margins ciliate, the cilia piliform, or dentiform toward the apex, the longest cilia / to У; as long as the width of the blade; apex long- acuminate, flat to slightly rounded, plane to slightly beveled in profile; setae arising evenly іп form and often in color, № to nearly У; as long as the blade, slightly to strongly attenuate, whitish-lutescent, translucent to subopaque. Sporophylls with the margins eciliate to short-ciliate toward the base; apex broadly rounded to carinate. Megaspores rather finely rugose-reticulate on the commissural face, more coarsely and less prominently marked on the outer face, pale orange-yellow. S. Wigbtii is the most generalized Old World species in the series and probably - represents the basic type for that region as S. Sellowii and S. Sartorii do for Tropical America. It is most closely related to S. mjamnjamensis. The African and Indian plants are very closely related. On the basis of rela- tively few collections it seems to be distinct but not in characters of specific value. Ample material may indicate a need to change the varietal status given the African plant here. Southern India and Ceylon, eastern Africa. KEY TO VARIETIES _ à. Setae predominantly strongly attenuate; sporophylls eciliate to only slightly ciliate toward the base 10a. var. Wigbtii, p. 39 a. Setae stout to predominantly slightly attenuate; sporophylls predominantly short-ciliate toward the Базе. 10b. var. Phillipsiana, р. 40 10a. SELAGINELLA Млонти var. Wightii. Fig. 13. Map 19. Vetus Wigbtii ve vetusta Hieron. іп Hedwigia 39:320. 1900, ex char. (Holotype: Mauritius, Sieber ? B). The localities taken from the literature (Map 19) are from Alston, in Proc. Nat. Inst. Sci. India 11:215. 1945. Although I have not seen the type of var. [Vor. 42 40 ANNALS OF THE MISSOURI BOTANICAL GARDEN vetusta said, probably in error, to come from Mauritius, the description does not adequately separate it from var. Wigbtii. In addition, I have seen Gamble 17258 (P) identified by Hieronymus as var. vetusta and it is typical var. Wightii. Rocky places, 700—1700 m. Southern India and A Specimens examine Innia: Gamble 17258 (P); Wallich 2188/3 (BM, US); Peninsula Ind. orientalis, Wigbt 15 (B, BM, К, P), Wigbt 283 (B, K), 3194 (B, GH Crrton: У. Ferguson 233 (US); Thwaites 1414 (Р); Walker (GH). 10b. SELAGINELLA Млонти var. PHiLLIPSIANA Hieron. in Hedwigia 39:320. 1900. (Holotype: Phillips B!). Selaginella Phillipsiana (Hieron.) Alston, in Jour. Bot. 77:222. 1939. The localities taken from the literature (Map 19) are from Alston, in Jour. Bot, 77:222. 1939. On stones and rocks, 1200—1500 m. British Somaliland to Tanganyika. Representative specimens: RITISH SOMALILAND: J. B. Gillett 4681 (К), 4681A (К); Glover & Gilliland 480 Ou 771 (BM); Godman 6 (BM); Darra-as, 5000 ft., June, 1895, Mrs. Lort Pbillips (B, K, P). E Napier 2054 (K). ТАМСАМУІКА: Ublig 845 (ВМ). 11. SELAGINELLA CINERASCENS А. А. Eaton, in Fern Bull. 7:33. 1899. (Holo- type: Kimball GH!). Fig. 14. Map 20. Selaginella bryoides Underw. Our Nat. Ferns, ed. 6, 13 1900, not (Kaulf.) Hieron., 1901, not Lycopodium bryoides Nutt. ex Baker, ips considered as based on Selag- inella cinerascens A. A. Eaton Stems with the older primary branches once to usually twice to rarely three times pinnate; leafy stems radially symmetrical or nearly so. Leaves herbaceous to herbaceous-papyraceous, none red, linear, acuminate to broadly lanceolate; base abruptly adnate, distinct from the stem in color or sometimes rather decurrent on the leader stem, usually glabrous, sometimes pubescent; margins rarely eciliate, usually ciliate, the cilia usually piliform, or dentiform only toward the apex, the longest cilia up to 14 as long as the width of the blade; apex acute to nearly obtuse, flat to slightly rounded, plane in profile; setae absent, the apex not or slightly modified, or if present, arising rather abruptly in form and color, up to 14 as long as the blade, stout, whitish, opaque to translucent. Sporophylls with the margins eciliate to very short-ciliate toward the base; apex broadly rounded to subcarinate. Megaspores slightly rugose on the commissural face, more prominently marked on the outer face, with an equatorial ring, pale yellow. 19551 TRYON-—SELAGINELLA RUPESTRIS AND ITS ALLIES 41 S. cinerascens is quite closely related to the next species, S. Arsenei. The two may be distinguished by the cilia which are predominantly piliform in S. cinerascens and dentiform іп S. Arsenei. There is a close resemblance to S. nivea of Madagascar but this does not extend beyond characters of habit and general aspect. This species affords the best illustration of the difference in length of leaves on the stem and branches, a character presumably due to some type of inhibition of growth. The leaves on the main stem are the longest, those on the primary branches are shorter and those on the secondary branches are the shortest. S. cinerascens is one of the few species that is not closely associated with rocky habitats. It grows on hillsides and slopes, among or beneath shrubs and small trees, often in clay soil, sometimes in rocky soil, from sea level to about 200 m. California and Baja California. Representative specimens: UNITED STATES. CALIFORNIA: Abrams 3399 (Е, СН, MO, NY, US); April 12, 1918, Betbel (GH, MO, HE National City (San Diego), May, 1897, L. F. Kimball (GH, NY); — 10765 (NY, Ч XI BAJA CALIFORNIA: Ferris 8474 (NY, US); Wiggins 4213 (GH, NY, US). 12. SELAGINELLA ARSENEI Weath. in Jour. Arn. Arb. 25:417. 1944. (Holotype: Arsène 10641 US! fragment СН!. Paratypes: Arsène 0083 US!, 10643 US!). Fig. 15. Map 21. Stems with the older primary branches once to twice pinnate; leafy stems radially symmetrical Leaves herbaceous, none red, ligulate to ligulate-long- triangular; base abruptly adnate, distinct from the stem in color, glabrous to pubescent; margins ciliate, the cilia dentiform, the longest cilia about Уз as long as the width of the blade; apex acute to acuminate, rounded, plane or nearly so in profile; setae absent, the apex slightly modified, or if present, arising evenly in form and usually in color, about № as long as the blade, stout, greenish- to lutescent-whitish, subopaque. Sporophylls with the margins very short-ciliate toward the base; apex broadly rounded. Megaspores slightly and finely rugose- reticulate, lemon-yellow. The differences between S. Arsemei and the closely related S. cinerascens are discussed under S. cinerascens. S. Arsenei grows in rocky places, at one locality at 2700 m. Central Mexico. Specimens examined: Мехі SAN LUIS IE : T. C. & E. M. Frye 2083 (Е). QUERETARO: Querétaro, 914, Aaa 10641 (GH , US), 9983 (US), 10643 (US). нірлісо: Hitchcock ё Stan- fave 7248 (US). GUERRERO: 8 Е. M. Frye 3149 (US). 13. SELAGINELLA MACRATHERA Weath. in Jour. Arn. Arb. 24:326. 1943. (Holo- _ type: Johnston 8067 GH!). Fig. 16. Map 22. Stems short, with the older primary branches once to rarely twice pinnate; leafy stems radially symmetrical. Leaves herbaceous-papyraceous, none red, sub- [Vor. 42 42 ANNALS OF THE MISSOURI BOTANICAL GARDEN ulate; base abruptly adnate, distinct from the stem in color, glabrous to slightly pubescent; margins ciliate, the cilia dentiform, or short-piliform only at the base, the longest cilia V4 or less as long as the width of the blade; apex acuminate to bluntly acuminate, broadly rounded, plane or nearly so in profile; setae arising abruptly in form and color, 35 to usually 24 as long as the blade, filiform, white or whitish, subopaque to translucent. Sporophylls with the margins short-ciliate toward the base; apex broadly rounded to carinate. Megaspores slightly to mod- erately finely rugose on the commissural face, moderately rugose on the outer face, yellow-orange. Although known from only a single collection there is no doubt that this is a valid species. The dentiform cilia of the leaves and the very long setae distinguish it from all others. Collected on ledges on north-facing lava cliffs in Chihuahua, Mexico. Specimens examined: Mexico. CHIHUAHUA: Sierra del Virulento, 2-3 miles east of Rancho Virulento, ca. 65 miles south of Ojinaga, Aug. 11, 1941, I. M. Johnston 8067 (GH, MO). 1955] TRYON-—SELAGINELLA RUPESTRIS AND ITS ALLIES 43 14. SELAGINELLA SHAKOTANENSIs (Franch. ex Takeda) Miyabe & Kudo, in Jour. Fac. Agric. Hokkaido Imp. Univ. 26 (Fl. Hokkaido & Saghal. I):64. 1930. Fig. 17. Map 23. Xi мн rupestris (L.) Spring var. ze Franch. ex Takeda, in Bot. Mag. Tokyo 7. 1909. (Holotype: Faurie 0805 SA Stems with the older primary branches once to twice pinnate; leafy stems some- what dorsiventral in the position of the leaves. Leaves fleshy to fleshy herbaceous, none red, ligulate to ligulate-lanceolate; base abruptly adnate, distinct from the stem in color, or sometimes rather decurrent on the leader stems, glabrous; margins ciliate, the cilia piliform, or dentiform only toward the apex, the longest cilia М to 15 as long as the width of the blade; apex convexly acute to obtuse, rounded to carinate, abruptly beveled to truncate in profile; setae arising abruptly in form and color, about У; as long as the blade, attenuate, whitish, translucent. Sporophylls with the margins eciliate to long-ciliate toward the base; apex narrowly carinate. Megaspores finely and very slightly rugose-reticulate, orange. S. shakotanensis may be distinguished from S. sibirica of Rupestres, which overlaps its range, by the shorter setae and the abruptly adnate leaf-bases. The setae are about № as long as the blade in this species while they are № as long or longer in S. sibirica. ‘The main stem may have some of the leaves with а decurrent base but on the branches the leaf-base is abruptly adnate. 5. shakotanensis bears a rather close resemblance to S. mutica var. limitanea from pen it may be sep- arated by its slightly dorsiventral leafy stem. Rocky alpine habitats; one collection was made at 2000 m. Japan. Specimens examined: ЈАРАМ: Aug. 20, 1898, Faurie (P), 1577 (P), Shakotan, Yezo Island, Shiribeshi, pe 9, iss, Faurie 9895 (P, МО), 10543 (P, MO), 13151 (P, MO); Ц. S. Nat. Herb. sh 10953 15. SELAGINELLA WALLACE! Hieron. in Hedwigia 39:297. 1900, maintained over S. montanensis by Broun, Index N. Am. Ferns, 159. 1938. (Holotype: Wallace B! fragment NY!). Fig. 18. Map 24. Selaginella montanensis Hieron. in Hedwigia 39:293. 1900. (Holotype: A. 8 A. Krause B! fragment NY Selaginella rupestris (L. ) Spring f. Wallacei (Hieron.) Clute, in Fern Bull. 16:52. 1908. Selaginella т Mq f. montanensis (Hieron.) Clute, in Fern Bull. 16:52. 1908, as montan Selaginella pend var. columbiana . Jones, in Univ. Mont. Bull. Biol. бег: 15:8. 1910. (Holotype: M. E. Jones, eg Lake, near Weiser, Г Selaginella rupestris var. Wallacei (Hieron.) Frye, Ferns Northwest, 33. 1934. Selaginella Rosendablii Hieron. ex Frye, Ferns Northwest, 33. 1934, in synon. (Evidently ased оп Rosendabl 86. Selaginella W allacei f. санне (М. Е. Jones) Broun, Index М. Am. Ferns, 159. 1938. Selaginella rupestris var. montanensis (Hieron.) 69 attributed to Frye by] Broun, Index М. Am. Ferns, 159. 1938, in syno. [Vor. 42 44 ANNALS OF THE MISSOURI BOTANICAL GARDEN Stems with the older primary branches once to twice, rarely three times pinnate; leafy stems radially symmetrical to rather dorsiventral in position, length and shape of the upper and under leaves. Leaves herbaceous to fleshy, none red, ligulate to ligulate-long-triangular; base abruptly adnate, distinct from the stem in color, sometimes rather decurrent on the leader stems, glabrous or pubescent; margins eciliate or ciliate, the cilia dentiform to usually piliform, the longest cilia 14 to М as long as the width of the blade; apex acute to obtuse, broadly rounded, abruptly beveled to truncate in profile, or predominantly so; setae arising abruptly in form and color, 15 to 1⁄4, rarely 14, as long as the blade, rather stout, milk-white and opaque to whitish or greenish-white and translucent. Sporophylls with the margins eciliate to short-ciliate toward the base; apex broadly rounded to subcarinate. Megaspores rugose to rugose-reticulate, usually with an equatorial ring, pale orange. S. Wallacei is one of the most variable species in habit and general aspect. It is also, of the species I know sufficiently, the one with the widest range of habitats and to a considerable extent its variability is correlated with its ecology. Specimens from very damp, shady situations have long stems that form loose mats and the leaves are somewhat distant. Those from dry, exposed rock or gravelly turf have short stems that form a compact mat and the leaves are close. Of the species that grow within its range it sometimes resembles S. densa var. scopulorum of Rupestres. It may be separated by the remote branches, S. densa having approximate ones, and by its abruptly adnate rather than decurrent leaf- bases. Suskdorf 8563 СН has а strobilus 9 cm. long, the longest I have seen in any species. Selaginella Wallacei grows on dry, exposed cliffs, talus slopes, rocky bluffs and gravelly soil, or on moist, shaded rocks, bluffs and cliffs, in crevices or covering various types of igneous and sedimentary rocks, from sea level to 2000 m. Western Montana to southern British Columbia and California. Representative specimens: ADA. BRITISH COLUMBIA: Hitchcock 8 Martin 7367 (MO, NY); Macoun 86372 (сн, NY, Us) ‚ 86376 (GH, NY, US), 86377 (GH, NY, US); Rosendabl 864 (BM, > > > UNITED STATEs. MONTANA: Aug. 13, 1908, M. E. Jones (BM, MO, US); “Crossing”, i : Kirkwood & Zeller 748 (GH, MO, US). овесом: Thompson 4002 (MO, NY, US), 4029 (MO, NY, US), 4102 (MO, US); Oregon, Capt. Wallace (B, NY, YU); Wilkes Exped, "Wash. & Ore. Terr." (GH, NY, US, YU). (GH, MO, NY, US); Heller 5024 (GH, MO, NY, US). 16. SELAGINELLA MuTica D. C. Eaton ex Underw. in Bull Torr. Bot. Club 25:128. 1898. (Lectotype by Weath. in Jour. Arn. Arb. 25:413. 1944: Meeban YU! fragment NY!. Paratypes: M. A. C. Livermore МУЉА С. Compton NY; E. J. Spence NY!; Toumey on Aug. 28, 1894 МУ!; Brandegee 19551 TRYON-—SELAGINELLA RUPESTRIS AND ITS ALLIES 45 NY!. Bigelow NY! and Mex. Bound. Survey NY! are excluded as types; they are intermediates between var. limitanea and var. mutica). Stems with the older primary branches once to twice pinnate; leafy stems radially symmetrical. Leaves fleshy to fleshy-herbaceous, none red, ligulate- lanceolate to lanceolate-elliptical to lanceolate-ovate; base abruptly adnate, distinct from the stem in color, or slightly decurrent on the leader stems, pum or pubescent; margins ciliate, the cilia entirely piliform, the longest cilia V4 to 14 as long as the width of the blade, to entirely dentiform and shorter; apex obtuse to convexly acute, strongly carinate, subtruncate in profile; setae absent, the apex being scarcely modified, or if present, arising abruptly in form and usually in color, up to У; as long as the blade, stout, whitish to greenish-white, subopaque to translucent. Sporophylls with the margins eciliate to long-ciliate toward the base; apex Carinate to narrowly carinate. Megaspores rugose-reticulate to coarsely rugose on the commissural face, less marked to smooth on the outer face, pale orange. The variation of S. mutica, as pointed out Бу C. A. Weatherby,? is rather parallel to that of S. Underwoodii which has a very similar range. However, in S. mutica the material of Texas, southern New Mexico and southern Arizona is more distinctive and the line of demarcation stronger than in $. Underwoodii, so that I am recognizing two varieties. The characters of seta length and ciliation of the sporophylls correlate to a high degree and separate the major components of the species. The southern var. limitanea is itself variable but the extremes, one with short, dentiform, ascending cilia on the leaves, and the other with long, piliform, patent cilia, are too often connected by intermediates and the characters vary considerably in some individual mats. The muticous incurved leaves of var. mutica give the stems a reptilian aspect under low magnification. In Colorado I have seen S. mutica growing with S. Underwoodii in the same niche. From a number of such cases it appeared quite clear that although the mats of the two species were originally nearly equal in size, 5. mutica gradually occupied more and more of the niche and finally entirely replaced 8. Underwoodii. Rarely a strobilus may be slightly dorsiventral. Colorado and Utah to Texas and Arizona. KEY TO VARIETIES a. Leaves setate; sporophylls with the cilia mostly or entirely ascending, all or some dentiform _________16а. var. limitanea, р. 46 - Leaves muticous or nearly so; у with the cilia entirely ог рге- 16b. var. mutica, p. 46 трн ту spreading, piliform—-————~—-————--—— 33 Jour. Arn. Arb. 25:415. 1944. [Vor. 42 46 ANNALS OF THE MISSOURI BOTANICAL GARDEN léa. SELAGINELLA MUTICA var. LIMITANEA Weath. in Jour. Arn. Arb. 25:414. 1944. (Holotype: Goodding US! fragment GH!. Paratypes: Ferriss GH!; Hinckley 1155 US!; Ingram 2723 US!; Moore & Steyermark 3046, in part, GH!; E. J. Palmer 31951 US!; Slater US!; Standley in 1906 US!; Wooton in 1899 US!, in 1903 US!, in 1906 US!, in 1909 US!). Map 25. Selaginella mutica var. texana Weath. in Jour. Arn. Arb. 25:414. 1944. (Holotype: M teyermark 3196 GH! fragment US!. Paratypes: Moore & Steyermark 3046, in part, СНІ; Е. J. Palmer 30871 СНІ; Sperry 428 US!. Tracy & Earle 275 GH! US! an ex. Bound. Survey YU fragment NY! are excluded as types; they are inter- mediate between var. limitanea and var. mutica). Leaves short-setate; cilia mostly or entirely ascending, piliform to dentiform. Sporophylls with the cilia mostly or entirely ascending, all or some dentiform. Sheltered or moist cliffs and rocky hillsides, on igneous rocks, 1300-2400 m. Southwestern Texas to southeastern Arizona. Representative specimens: 30871 (GH, MO, US), 31951 (МО, US); Sperry 428 (US). NEW MEXICO: mountains south of Deming, Oct. 4, 1937, Goodding (GH, US); Dec., 1924, Slater (US); Sept. 10, 1899, Wooton (US), April 18, 1903 (NY, US), June, 1906 (US), Jan. 9, 1909 (US). ARIZONA: March, 1904, Ferriss (GH). 16b. SELAGINELLA MUTICA var. mutica, Fig. 19. Map 26. Selaginella Watsonii mutica (D. C. Eaton ex Underw.) Clute, Fern Allies, 144, 264. 1905. Leaves muticous or nearly so; cilia entirely or predominantly spreading, pili- form. Sporophylls with the cilia entirely or predominantly spreading, piliform. The following specimens are intermediate between var. mutica and var. limitanea: Tracy 8 Earle 275 (BM, F, GH, MO, NY, US); Mex. Bound. Survey, Parry et al. (MO, NY); New Mexico, 1853, Bigelow (NY). Exposed or sheltered rocky bluffs, cliffs and ledges, in crevices or soil pockets or usually covering rocks, on igneous rocks, sandstone or rarely limestone, 1450— 4300 m. Colorado and Utah to Texas and Arizona. Representative specimens: U STATES. COLORADO: 1874-1878, Brandegee (NY); А. С. Compton (NY); 1. M. Jobnston 2423 (Е, СН, NY, US); M. A. C. Livermore (NY rado, 1871, T. Meeban (NY, YU); E. J. Spence (NY); R. M. 8 А. F. Tryon 5071 (CU, MIN, MO, PH, POM, UC, US), 5072 (B, MO, US), 5073 (F, MO, NY, P, UC), 5075 (ARIZ, DS, FI, K, MICH, MO, RM); E. P. Walker 365 (СН, NY, US). utan: Grabam 7592 (US), 10004 (US); Maguire 18450 (GH). TEXAS: 1921, Goodding (BM, GH, NY. US); R. M. 8 A. F. Tryon 5083 (BM, GH, MO, 05); Waterfall 4811 (GH, MO, NY), 5392 (GH, МО, NY); Wright 2115 (СН, NY). NEW MEXICO: Arsène 8 Benedict 16643 (Е, US); Drouet & Richards 3311 (Е, GH, MO, NY, US); Standley 5199 (MO, NY, US). arizo (NY) 19551 ТКҮОМ---<5ЕГ.АСІМЕІЛ,А RUPESTRIS AND ITS ALLIES 47 со ЗОНА Р IGHTII 28. S. WR 17. SELAGINELLA EXTENSA Underw. in Bull. Torr. Bot. Club 25:131. 1898. (Holotype: Pringle 3900 NY. Paratype: C. Müller NY). Fig. 20. Map 27. Selaginella rupestris mexicana Conzatti, Fl. Tax. Mex. 1:150. 1939, without Latin descrip- tion, not var. mexicana Milde, 1867. (Based on Pringie 3900). Plant terrestrial or sometimes epiphytic. Stems very long to long, branches long, remote, with the older primary branches once to twice pinnate; leafy stems radially symmetrical or nearly so. Leaves herbaceous, none red, subulate to ligu- late-long-triangular to ligulate-lanceolate; base abruptly adnate, distinct from the stem in color, usually glabrous, sometimes pubescent; margins ciliate, the cilia dentiform, the longest cilia less than V4 as long as the width of the blade; apical portion usually revolute; apex acuminate, rounded, plane or nearly so in profile; muticous, the apex subulate, greenish-hyaline. Sporophylls with the margins eciliate to short-ciliate toward the base; apex carinate. Megaspores finely rugose- reticulate on the commissural face, the equatorial region strongly and irregularly Papillate, granular on the outer face, pale orange. S. extensa is casually epiphytic and rather similar in habit to the pronounced epiphyte $. oregana of Rupestres. The revolute apical portion of many of the leaves, the muticous essentially unmodified leaf-apex, the habit and the unusual рарШате megaspores combine to make this a very distinctive species. On cliffs or in woods, covering limestone rocks or pendent from tree trunks, 870-1700 m. [Vor. 42 48 ANNALS OF THE MISSOURI BOTANICAL GARDEN Eastern Mexico and Jalisco. Specimens examined: EXICO. TAMAULIPAS: Sharp et al. 50255 (MO, US). sAN Luis Potosi: McVaugh 10446 (US); Pennell 17947 (US); Las Canoas, Aug. 21, 1891, Pringle 3000 (Е, СН, К, МО, US); Sharp 46264 (US). veracruz: Bourgeau 2541 (СН, TE Mobr 12 (YU); Purpus 6052 (F "Сн, МО, US); Sharp сес (US). нірлісо: T. C. 8 Е. M. Frye 2537 (NY, US); Hoogstraal & Chase 7308 (Е, MO, US); Kenoyer 8 Cram 3910 (GH); Sharp 45864 (US). PurBLa: Sharp 45375 (US). улллзсо: McVaugb 11820 (US). 18. SELAGINELLA Монти Hieron. in Hedwigia 39:298. 1900. (Holotype: Wrigbt 828 B!). Fig. 21. Map 28. Selaginella rupestris (L.) Spring f. Wrigbtii (Hieron.) Clute, in Fern Bull. 16:52. 1908. Stems with the older primary branches once to twice pinnate; leafy stems usually strongly to rarely moderately dorsiventral in position, length and shape of the upper and under leaves. Upper leaves ligulate-long-triangular to long-tri- angular; apex acuminate to convexly acute. Under leaves herbaceous, none red, longer to only slightly longer than the upper, linear; apex acuminate, flat to slightly rounded, plane or nearly so in profile. All leaves with the base abruptly adnate, distinct from the stem in color, or the under leaves with the base rather decurrent, pubescent or glabrous; margins ciliate, the cilia piliform, or dentiform only toward the apex, the longest cilia 14 to № as long as the width of the blade; setae absent, the apex being hardly modified, or if present, arising very evenly in form and usually in color, rarely up to 14 as long as the blade, stout, whitish- lutescent to lutescent, translucent to opaque. Sporophylls with the margins long- ciliate toward the base; apex broadly rounded to carinate. Megaspores prominently rugose, pale to bright orange. Of the definitely dorsiventral species of the series, $. Wrightii is the least dorsi- ventral. The linear under leaves, the long-ciliate sporophylls and the poorly dif- ferentiated setae are the primary characters of the species. This is the only species for which I have sufficient information to class definitely as a calciphile. Exposed or shaded ledges and pockets of limestone, 800-2300 m. Texas and New Mexico to southern Mexico. Representative specimens: UNITED States. TEXAS: E. J. Palmer 11389 (СН, 2 NY, US); Rose & Fitch 17973 (MO, NY, US); Tharp 43-514 (МО, NY, US); В. M. 8 A. Е. Tryon 50 5034 (BM, GH, MO, P, Ыз, 5015 (B, BM, Е, СН, MO, МУ, P, UC, US), 3037 (aniz, FI, GH, K, Mexico коти Bartlett 10561 (Е, US). Nuevo LEON: Barkley 16144M (GH, MO); Harvey 988 (СН, US); C. H. Muller 2638 (GH, US). солнипл: фаза ston 8607 (СН, МО), 88 54 (СН), 9106 ЖЕН. МО, US). сніноаноа: Harvey I (СН, US); E. Palmer 455 іп 1885 (СН, US, YU). saw Luts РОТОЧ: А. A. Lundell $^ (US). zacatecas: Lloyd & Kirkwood or Kirkwood) 20 (MO, US), 142 (GH, MO). HIDALGO: Kenoyer 1053 (GH). puesta: Е. B. Copeland 174 (US). cuxmmERo: Hinton 9678 (GH, US). oaxaca: Pringle З и 1300 (СН). 19551 TRYON-—SELAGINELLA RUPESTRIS AND ITS ALLIES 49 19. SELAGINELLA SrEvERMARKII Alston, in Ann. & Mag. Nat. Hist., XII, 7:638. 1954. (Holotype: Steyermark 50501 ВМ!. Paratypes: Steyermark 47115 BM!, 51292 BM!). Figs. 22, 23. Map 29. Stems with the older primary branches once to twice pinnate; leafy stems strongly dorsiventral in position, length and shape of the upper and under leaves. Upper leaves subulate to subulate-long-triangular; apex long-acuminate. Under leaves papyraceous to herbaceous-papyraceous, usually some red, longer than the upper, subulate-lanceolate to ligulate-lanceolate; apex long-acuminate, flat or very slightly rounded, plane in profile. All leaves with the base abruptly adnate, distinct from the stem in color, pubescent to rarely glabrous; margins ciliate, cilia piliform, the longest cilia V4 to № as long as the width of the blade; setae usually arising evenly in form and abruptly in color, И; to 14 as long as the blade, slightly to strongly attenuate, white to whitish-lutescent, translucent to subopaque. Sporo- phylls with the margins eciliate to short-ciliate toward the base; apex rounded to carinate. Megaspores finely to moderately rugose-reticulate, to nearly smooth on the outer face, more prominently and coarsely marked in the equatorial region, pale orange. This species and the next, S. Hansenii, are the only ones that usually have some red leaves on the stem. In S. Sartorii such leaves are present infrequently. They tend to occur particularly among the oldest living leaves and the youngest dead ones and often are more common on the under side of the stem. The long-acuminate apex of the upper leaves and the thin under leaves afford an ample separation from $. Hansenii. This species has the most robust leafy stems of any in the series and material collected in the growing state presents particularly attractive sprays. It is one of the very few species I know to be casually culti- vated; the Standley collection cited below was from the garden of Mariano Pacheco. Rocks and banks, 1630-2000 m. Guatemala and adjacent Mexico. Specimens examined: Mexico, cHiAPAs: Matuda 1846 (US). GuATEMALA: Aguilar 1420 (Е, US); Hatch 8 Wilson 283 (US), 325 (US); Haupt 105 (US); Maxon & Hay мо (05); — 63102 (F, ine Steyermark 47115 (BM), between San Sebastián Hacienda and large рећазсо above n, Dept. Huehuetenango, 2000—2200 m., 1942, Syl 50501 (BM), Sici pee (BM); Tuerckbeim 8844 (GH, US); L. О. Williams 14209 (US). 20. SELAGINELLA НАМЗЕМИ Hieron. in Hedwigia 39:301. 1900, as Hanseni; maintained over S. Bolanderi. (Lectotype: Hansen 878 В!. Paratypes: Hille- brand B!; A. A. Eaton B!). Fig. 24. Map 30. есен MH Hieron. in Hedwigia 39:300. 1900. (Holotype: Bolander 4511 B! ra бања a аа (L.) Spring f. Hansenii (Hieron. ) Clute, in po Bull. 16:52. 1908. Selaginella rupestris £. Bolanderi (Hieron.) Clute, in Fern Bull. 16:52. 1908 Selaginella rupestris var. Hansenii (Hieron.) Jepson, Man. Pl. Calif. 2 1923. Selaginella rupestris var. Bolanderi (Hieron.) Jepson, Man. Pl. Calif. 42. 1923. [Уог. 42 50 ANNALS OF THE MISSOURI BOTANICAL GARDEN Stems with the older primary branches once to usually twice to rarely three times pinnate; leafy stems strongly dorsiventral in position, length and shape of the upper and under leaves. Upper leaves long-triangular to ligulate to lanceolate- long-triangular; apex convexly acute to acute. Under leaves herbaceous, usually some red, longer than the upper, linear-lanceolate to lanceolate; apex convexly acute to acuminate, flat to slightly rounded, plane to gently beveled in profile. All leaves with the base abruptly adnate, distinct from the stem in color, glabrous or pubescent; margins ciliate, the cilia piliform, or dentiform only toward the apex, the longest cilia 1⁄4 to rarely nearly У» as long as the width of the blade; setae arising rather abruptly in form and in color, 15 to nearly У; as long as the blade, stout to attenuate, milk-white to whitish, translucent to opaque. Sporophylls with the margins eciliate to short-ciliate toward the base; apex broadly rounded to carinate. Megaspores rugose to rugose-reticulate on the commissural face, less prominently rugose-reticulate to smooth on the outer face, with an equatorial ring, emon-yellow. S. Hansenii is related to the previous species, S. Steyermarkii, by the dorsiventral leafy stem and the commonly occurring red leaves. These characters, however, may be the result of parallel evolution. 'The red leaves tend to be more common on the under side of the stem and in the area where living and dead leaves adjoin. e branches at the center of a mat tend to be rather erect and compact, those оп the periphery prostrate, long-creeping and remote. This variation within a mat, together with its associated characters of number and spacing of the leaves, is similar to that found in S. tortipila of Arenicolae. Open or lightly shaded bluffs or cliffs and surfaces of igneous rocks, 330- 1350 m. Central California. Representative specimens: UNITED STATES. CALIFORNIA: Auburn, Placer Со. April 10, 1865, Bolander 4511 (B, MO, NY, US); 1892, A. A. Eaton (B, NY); Puhe Cabin, Mokelumne River, Amador and Calaveras counties, April, 1893, G. Hansen 878 (B, MO, NY, US); Heller 2583 (F, GH, MO, NY, US), 5466 (GH, MO, NY, US), 8173 (F, GH, Mo, NY, US), 10715 (GH, MO, US), rr802 (F, GH, MO, US); 1863, Hillebrend (В); R. М. ДА. Е Tryon 5067 (ARIZ, B, BM, CU, DS, Е, FI, СН, К, MICH, MIN, MO, ме, P, POM, RM, UC, US, WS) 21. SELAGINELLA carinata Tryon, spec. nov. Figs. 25, 26. Map 51. Caules breves, ramis brevibus propinquis; caules frondosae positione foliorum superorum et inferorum valde dorsiventrales sed forma minus. Folia carnosa, apice carinato vel valde carinato, aspectu laterali truncato vel abrupte declivi, base abrupte adnata cum caule distincta colore, setis abrupte productis colore sensim diversis. Sporophylla marginibus ad basem eciliatis. — Megasporae laeves flavae. Typus: Rose, Pachano & Rose 23039 (US) Stems short; branches short, approximate; older primary branches once to usually twice, sometimes three times pinnate; leafy stems strongly dorsiventral in position and less so in the shape of the upper and under leaves. Upper leaves к DES QU adora erik Re hene eis 19551 TRYON-—SELAGINELLA RUPESTRIS AND ITS ALLIES 51 ligulate-long-triangular; apex convexly acute. Under leaves fleshy, none red, lanceolate to ligulate-lanceolate; apex acuminate to convexly acute, carinate to strongly carinate, truncate to abruptly beveled in profile. All leaves with the base abruptly adnate, distinct from the stem in color, glabrous; margins ciliate, the cilia piliform toward the base, dentiform above, the longest cilia V4 or less as long as the width of the blade; setae arising abruptly in form and evenly in color, / to № as long as the blade, rather stout at the base, tapering evenly to the tip, greenish- to lutescent-whitish and subopaque at the base, milk-white and opaque toward the tip. Sporophylls with the margins eciliate toward the base; apex strongly carinate. Megaspores smooth, yellow. S. carinata is a distinctive species and not evidently related to any others. The very short stems and branches and the fleshy leaves with carinate and truncate apex set it off from the other species of this series. Some strobili are slightly dorsi- ventral, particularly toward the base, apparently a development parallel to that in the African and Madagascar species. Central Ecuador. Specimens examined: Ecuapor: Vicinity of Nabón, Sept. 26, 1918, Rose, Pachano t$ Rose 23039 (СН, US). [Vor. 42 52 ANNALS OF THE MISSOURI BOTANICAL GARDEN 22. SELAGINELLa indica (Milde) Tryon, comb. nov. Fig. 27. Map 32. me че rupestris (L.) Spring f. indica Milde, Fil. Europ. Atlant. 262. 1867. (Ноіо- ooker omson Selaginells lowsipile of authors, z not Hieron., for example, of Alston, in Proc. Nat. Inst. Sci. India 11:214. 194 Stems with the older primary branches once to twice pinnate; leafy stems strongly dorsiventral in position, length and shape of the upper and under leaves. Upper leaves subulate-long-triangular to long-triangular; apex acuminate. Under leaves papyraceous, none red, longer than the upper, long-triangular to ligulate- long-triangular to subulate-lanceolate; apex acuminate, flat, plane in profile. A leaves with the base abruptly adnate, distinct in color, glabrous to slightly pubes- cent; margins ciliate, the cilia piliform toward the base, dentiform above, the longest cilia V4 to У» as long as the width of the blade; setae usually arising evenly in form, less so in color, 1⁄4 as long as the blade, strongly attenuate, tawny to lutescent-whitish, translucent. Sporophylls with the margins eciliate to short- ciliate toward the base; apex rounded to carinate. Megaspores rather finely rugose- reticulate on the commissural face, less prominently marked on the outer face, pale orange. This species was known as S. longipila from the time of the description of that species by Hieronymus, since the specimens were cited as from the Himalayas. However, the labels on Hieronymus’ material were erroneous, the specimens being S. densa of North America. S. indica is readily distinguished from the other Indian species, S. Wigbtii, by its dorsiventral leafy stem and thin under leaves. The localities taken from the literature (Map 32) are from Alston, in Proc. Nat. Inst. Sci. India 11:214. 1945. Cliffs, rocky banks and ledges, 700—2800 m. Nepal, Bhutan, northern and western India. Specimens examined: МЕРАТ: Polunin 1478 (BM). BHUTAN: Griffith 2871 (BM); Ludlow et ai. 17039 (BM), 18606 (BM). Innia: Duthie 3727 (BM, US); Khasia, 5000 ft., J. D. Hooker 8 T. Thomson (GH, NY, P, YU). 23. SELAGINELLA NJAMNJAMENSIS Hieron. in Hedwigia 39:312. 1900. (Holo- type: Schweinfurth B! fragment NY!). Fig. 28. Map 33. Stems with the older primary branches once to usually twice pinnate; leafy stems dorsiventral in position, length and slightly in the shape of the upper and under leaves. Upper leaves lanceolate to long-triangular; apex acuminate. Under leaves herbaceous, none red, longer than the upper, linear-lanceolate to lanceolate to long-triangular; apex acuminate, flat to slightly rounded, plane in profile. АП leaves with the base abruptly adnate, distinct from the stem in color, glabrous to pubescent; margins ciliate, the cilia usually piliform toward the base, dentiform above, the longest cilia М, to nearly У; as long as the width of the blade; setae arising evenly in form and color, V, or less as long as the blade, stout to attenuate, 19551 TRYON-—SELAGINELLA RUPESTRIS AND 11$ ALLIES 53 tawny to lutescent-whitish, subopaque to opaque. Sporophylls on the upper side of the strobilus lanceolate, acuminate, on the under side ovate-long-triangular to narrowly ovate, acuminate, slightly longer and broader; margins short-ciliate toward the base; apex broadly rounded to carinate. Megaspores granular to granular-tuberculate, yellow. This species is the least dorsiventral of the related species in Africa and Mada- gascar. Although sufficiently distinct from S. caffrorum, both species tend to overlap in the range of variation of the individual characters. The herbaceous under leaves with the base abruptly adnate and distinct in color and the poorly differentiated setae are the primary characters of S. mjamnjamensis, On and between exposed rocks, 800-1200 m. Central and southeastern Africa. Specimens examined: Me MA Supan: Dandy 510 (BM); Hoyle 520 (BM). FREN REQUE AFRICA: Chevalier 6668 (ВМ,Р); Le Testu 4704 (ВМ); Tisserant 3567 (BM CENTRAL AF FRICA: Njam-Njam, 1870, Schweinfurth (B, К, NY, P). РОКТОСЕЗЕ East AFRICA: Sim бо 05 (Р). 24. SELAGINELLA CAFFRORUM (Milde) Hieron. in Hedwigia 39:313. 1900. Fig. 29. Мар 34. еа қора (1-4. Spring Ма жашы Milde, Fil Europ. Atlant. 262. 1867. Terra Caffrorum, зна s нр м var. incurv ГА. "ы г. in Kuhn, Fil. Afr. 213. 1868, category taken rom annotation of Welwitsch 169 by A. Br. (Lectotype: Drége, Lycopodium rupestre а В. Paratypes: Ecklon 9 Zeyher 7 В; Drége, ни rupestre aa В; Welwitscb 169 В!; SU ne d Steudner). Selaginella rupestris var. incurva f. abyssinica A. Br. in Kuhn, Fil. Afr. 213. 1868, nomen n A single ділі rong was provided for "forma есе ыы abyssinica”. I have n Quartin-Dillon, P, but Steudner, the other specimen c Selaginella rupestris var. ^ nube + реа А. Br. in Kuhn, Fil. i 213. 1868. (Holo- e: Welwitsch 169 Selaginella pareria var. der f. capensis A. Br. in Kuhn, Fil. Afr. 213. 1868, nomen um (see ssinica DH ien irae Hieron. in Иные 39:314. 1900, epithet from S. ^g, ra t у age А. Br. (Lectotype: Drége, Lycopodium rupestre а В. Paratypes: Zeyher 7 B; See: Lycopodium rupestre B; R. Baur 1110 B; Sonder BI; Сыбан В; ebmann 3924 B, 4005 В; Sutherland B). Selaginella 4 Quartiniana Cufodontis in Phyton 4:178. 2962; nomen nudum, based on S. rupestris var. incurva f. abyssinica А. Br., nomen n Stems with the older primary branches once to Ro twice to rarely three times pinnate; leafy stems dorsiventral in position, length and shape of the upper and under leaves. Upper leaves linear-long-triangular to ligulate-long-triangular; apex acuminate to convexly acute. Under leaves papyraceous to papyraceous- herbaceous, none red, longer than the upper, lanceolate to ligulate-lanceolate to triangular; apex acuminate to convexly acute, flat to slightly rounded, plane or nearly so in profile. All leaves with the base abruptly adnate, distinct from the (Мог. 42 54 ANNALS OF THE MISSOURI BOTANICAL GARDEN stem in color, or the under leaves with the base slightly decurrent and blending in color, glabrous or rarely pubescent; margins ciliate, the cilia piliform, or sometimes dentiform toward the apex, the longest cilia М to / as long as the width of the blade; setae arising rather abruptly in form and color, 1⁄4 to У; as long as the blade, rather stout, whitish to whitish-tawny or whitish-lutescent, translucent to subopaque. Sporophylls on the upper side of the strobilus ovate-lanceolate to long- triangular, on the under side lanceolate-ovate to broadly ovate, acuminate, longer and broader; margins short- to long-ciliate toward the base; apex rounded to carinate. Megaspores rugose-reticulate to granular-rugose on the commissural face, more finely and less prominently marked on the outer face, lemon-yellow. The thin under leaves and well-differentiated setae are characteristic of S. caff rorum and set it off from the related S. njamnjamensis. In addition, S. caffrorum usually has some of the under leaves on the main stem with decurrent bases. Dry or moist rocky places, 600—2000 m. Anglo-Egyptian Зна to Angola and Union of South Africa. Specimens examined NGLO-EcYvPTIAN SUDAN: Jackson 1116 (BM); MacLeay 114 (BM). TANGANYIKA: ae 2431 Antunes 543 (P); Dehindt 543 (K, P); Lopolo, Huilla, 5000 ft., April, 51 ‚ NY, Р). м оғ SOUTH ARICA, NATAL: Rudatis 857 (P). ORANGE FREE STATE: Rebmann 3924 Em 2M (NY, Р). BASUTOLAND: Dieterlen 117 (P). САРЕ ОҒ coop HOPE: Т. ooper У); en 4000-5000 ft., Lycopodium rupestre a, Drége (GH, K, P, US); oe s Zeyber (B, P) ; Sonder (B). 25. SELAGINELLA ECHINATA Baker, in Jour. Linn. Soc. 22:536. 1887. (Нојо- type: Baron 4226 K!). Figs. 30, 31. Мар 35. Stems with the older primary branches once to three times, usually twice, pinnate; leafy stems dorsiventral in position, length, shape and in setae characters of the upper and under leaves. Upper leaves ligulate-long-triangular; apex acute; setae arising abruptly, mostly soon deciduous, usually substellate, divided into two to several filiform branches, whitish, translucent to opaque. Under leaves ћег- baceous-papyraceous, none red, longer than the upper, ovate-lanceolate; apex acuminate, flat to slightly rounded, plane in profile; setae arising very evenly in form and color, У, or less as long as the blade, persistent, stout, lutescent to tawny, opaque to subopaque. All leaves with the base adnate to abruptly adnate, distinct from the stem in color, glabrous; margins ciliate, the cilia piliform, the longest cilia М to usually У; as long аз the width of the blade. Sporophylls оп the upper side of the strobilus broadly ovate-lanceolate, obtuse, on the under side very broadly ovate-triangular, acuminate, longer and broader; margins long-ciliate toward the base; apex broadly rounded to carinate. Megaspores tuberculate, yellow. S. echinata is the most distinctive of the three species of Madagascar. It is the most strongly dorsiventral species and in addition has the truly unique character of substellate setae on the upper leaves. Also some leaves and sporophylls have 19551 TRYON—SELAGINELLA RUPESTRIS AND ITS ALLIES 55 a unique whitish to tawny, long-appressed pubescence on the back, which among the specimens I have seen is most pronounced in Perrier $262. Madagascar, 500—2000 m. нове ехапипед: ASCAR: Baron 4226 (ВМ, К); 1879, Cowan (BM); Humbert 2823 (Р), > жу SUR (P), 11778 (P); Humbert З Swingle 4791 (P); Perrier 1168 (P), 8 (P), 8262 (BM, P), 8311 (BM, P), 8312 (P), 8313 (BM, P), 8324 (P), 8346 (BM, Р). РОВОМ _ ч 38. S. DF лан $ 26. SELAGINELLA NIVEA Alston, in Perrier, Cat. Pl. Madagascar, 71. Feb. 1932, nomen nudum; in Dansk Bot, Ark. 7 (C. Chr., Pterid. Madagascar):194. June, 1932. (Holotype: Perrier 8303 BM!. Paratype. Perrier 8305 ВМ!). Fig. 32. Map 36. Stems with the older primary branches once to twice pinnate; leafy stems dorsi- ventral in position and shape of the upper and under leaves. Upper leaves linear- lanceolate to lanceolate; apex acute to acuminate. Under leaves papyraceous, none red, triangular to ligulate-triangular; apex acute to acuminate, flat to slightly rounded, plane in profile. All leaves with the base abruptly adnate, distinct from the stem in color, glabrous; margins ciliate, the cilia piliform toward the base, dentiform above, the longest cilia 1⁄4 or less as long as the width of the blade; setae arising evenly in form, abruptly in color, V4 to № as long as the blade, stout, milk- white, opaque. Sporophylls on the upper side of the strobilus lanceolate-ovate, on the under side broadly ovate, acuminate, longer and broader; margins long- [Vor. 42 56 ANNALS OF THE MISSOURI BOTANICAL GARDEN ciliate toward the base; apex broadly rounded to carinate. Megaspores finely granular to nearly smooth, light yellow-orange. This species has a remarkable superficial resemblance to $. cinerascens although in technical characters there is no relationship between them. S. nivea may be separated from the next species, S. proxima, by the straight and relatively short setae in addition to the characters of habit. S. mivea is closely prostrate with the branches once or twice divided. $. proxima has the primary branches tending to be ascendent and dendroid and commonly three, rarely four, times divided. Madagascar, among bushes. Specimens examined: МАРАСАЗСАВ: Humbert 11612 (P); Humbert 8 Swingle 5518 (BM, СН, US); Gn of Ampanihy, June, 1910, Perrier 6303 (BM, P), Perrier 6305 (BM, 72 ' 18702 ( 27. SELAGINELLA proxima Tryon, spec. nov. Figs. 33-35. Map. 37. Caules ramis primariis saepe dendroideis; caules frondosae positione longitudine et forma foliorum superorum et inferorum dorsiventrales. Folia base abrupte adnata caule distincta colore, marginibus ciliis longissimis longitudine 74—12 laminae latitudinis, setis 144-1 laminae longitudinis valde curvatis juventate praesertim cretaceis caecis. Strobili dorsiventrales sporophyllis stichorum duorum dorsalium deltoideis duorum ventralium ovato-deltoideis longioribus et latioribus. Megasporae leviter rugoso-reticulatae vel laeves citreo-flavae. Typus: Humbert 6 Swingle 5705 (US). Stems with the primary branches tending to be ascendent and dendroid, the older primary branches twice to usually three to rarely four times pinnate; leafy stem dorsiventral in position, length and shape of the upper and under leaves. Upper leaves lanceolate to ligulate-lanceolate; apex acute to obtuse. Under leaves herbaceous, none red, longer than the upper, triangular to ligulate-lanceolate; apex acute to obtuse, flat, plane in profile. All leaves with the base abruptly adnate, distinct from the stem in color, glabrous to pubescent; margins ciliate, cilia pili- form, the longest cilia М to У as long as the width of the blade; setae arising evenly to abruptly in form, abruptly in color, № to У» as long as the blade, stout at the base, usually with a filiform tip, strongly curved, especially those of the branch-tips, milk-white, opaque. Sporophylls on the upper side of the strobilus deltoid to long-deltoid, on the under side ovate-deltoid, longer and broader; margins short- to long-ciliate toward the base; apex broadly rounded to carinate. Mega- spores very slightly rugose-reticulate to smooth, lemon-yellow. The strongly curved setae suggest a relationship to S. Dregei while in other characters $. proxima is more closely related to the preceding species, S. mivea. The differences are discussed under that species. A bush xerophile, 20—1200 m. Madagascar. Specimens examined: 1955] TRYON—SELAGINELLA RUPESTRIS AND ITS ALLIES 57 Mapacascar: Humbert 6557 (P), 7089 bis (P), 12439 (P), 12706 (P), 13338 (P), 14158 (P); Humbert t$ Swingle 5581 (P), vicinity of Fort Dauphin, near Bevilany, 200- 300 m., Sept. 14, 1928, Humbert 9 Swingle 5705 (GH, P, US). 28. SELAGINELLA ПКЕСЕГ (Presl) Hieron. in Hedwigia 39:515. 1900. Fig. 36. Map 38. Lycopodium Dregei Presl,in Abhandl. Bóhm. Gesell. Wissen. V, 3 (Bot. Bermerkungen): 83, reprint 153. 1844. (Holotype: Drége, Lycopodium rupestre b PRC. Drege, odium rupestre a is excluded as a type; it is S. caffrorum). Selaginella rupestris (L.) Spring f. Dregei (Presl) Milde, Fil. Europ. Atlant. 262. 1867. Selaginella rupestris var. recurva A. Br. in Kuhn, Fil. Afr. 213. 1868, category taken from annotation of Welwitsch 48 and Drége, Lycopodium rupestre b by A. Br. (Lectotype: rége, Lycopodium rupestre b В!. Paratypes: Welwitsch 48 B!, 49 В). Selaginella rupestris var. recurva f. Dregeana А. Br. in Kuhn, Fil. Afr. 214. 1868, based on Lycopodium Dregei Presl as to Drége, Lycopodium rupestre b. Selaginella rupestris var. recurva f. Welwitschiana A. Br. in Kuhn, Fil. Afr. 214. 1868: (Lectotype: Welwitsch 48 B!. Paratype: Welwitsch 40 B). Selaginella Dregei var. pretoriensis Hieron. in Hedwigia 39:317. 1900. (Lectotype: Кеђ- mann 4333 В!. Paratype: Wilms 181 я Selaginella Dregei var. Rebmanniana Hieron. in Hedwigia 39:317. 1900. (Lectotype: mann 5576 B!. Paratype: Braga у Selaginella Dregei var. Васђтаттапа Hieron. in Hedwigia 39:317. 1900. (Holotype: Bachmann 9 B!). Selaginella Dregei var. Petersiana Hieron. in Hedwigia 39:317. 1900. (Holotype: Peters ! Selaginella Dregei var. Hildebrandtiana Hieron. in Hedwigia 39:317. 1900. (Lectotype: Hildebrandt 2363 B!. Paratypes: Fischer 627 B, fragment NY!; Stublmann 910 B, 43 . Selaginella Dregei var. Welwitschiana (A. Br.) Hieron. in Hedwigia 39:318. 1900. — Selaginella grisea Alston, in Jour. Bot. 77:222. 1939, based on S. Dregei var. Hilde- brandtiana Hieron. Stems often irregularly ascendent, buds rarely present at the base of the branches, older primary branches once to twice, rarely three times, pinnate; leafy stems radially symmetrical to somewhat dorsiventral in position and sometimes also in shape of the upper and under leaves. Leaves herbaceous to herbaceous-papy- raceous, none red, subulate to long-triangular to ligulate-long-triangular; base abruptly adnate, distinct from the stem in color, glabrous; margins ciliate, the cilia piliform, the longest cilia V2 as long to as long as the width of the blade; apex acuminate to long-acuminate, flat to slightly rounded, plane in profile; setae arising evenly in form and usually in color, 14 to usually У, as long to as long as the blade, attenuate, usually strongly curved, especially those of the growing-points, milk- white and opaque to rarely tawny-whitish, subopaque. Sporophylls in 2 ranks on the under side of the branch-tip; margins eciliate to long-ciliate toward the base; apex broadly rounded. Megaspores slightly tuberculate on the commissural face, similar to slightly rugose on the outer face, with an equatorial ridge, yellow to orange. The unilateral strobilus of this species is one of the most unique characters of any in the section. There are two rows of sporophylls on the under side of the [Vor. 42 58 ANNALS OF THE MISSOURI BOTANICAL GARDEN branch tip while vegetative leaves replace the usual two rows on the upper side. This character is sufficient to identify the species in spite of considerable variation in other characters such as the color, length and degree of curving of the setae, the leaf shape, the symmetry of the leafy stem, the number and length of the cilia on the leaves and the shape of the sporophylls. Except that the dorsiventral speci- mens have the broader under leaves, none of the variations correlate with each other or with a coherent distribution. Some sterile material with straight setae such as Peters, Mozambique, is difficult to place. However, such specimens as Moss 18088 BM which is otherwise similar, are fertile and can be placed with certainty. S. Dregei is variable in habit, the stems varying from closely prostrate to loosely and irregularly ascending. In the latter case the rhizophores may be unusually long. Rarely, some stems will have short branches at the base which will actively grow upon the death of the portions above. This character is otherwise found only in Arenicolae where it is well developed and constant in such species as S. rupincola and S. arenicola. The localities taken from the literature (Map 38) are from Alston, in Jour. Bot. 77:222. 1 Exposed or sheltered rocky places, on or among igneous or sandstone rocks, 700—2300 m. Southern and eastern Africa. Representative specimens: UGANDA: Беремо 2619 (Џ КЕ 5). ENYA: Fischer 627 (В, NY); Egu, between Duruma and Taita, Jan. 1877, Hildebrandt 2363 (B, BM, K, NY, P). ANGOLA: Exell & Mendonca 169 i M); Pedras de Guinga, ls Andongo, 3800 ft., March, 1837, Welwitsch 48 (B, К BECHUANALAND: Burcbell 231 5 (cin, Union оғ SOUTH AFRICA. TRANSVAAL: Leendertz 935 (BM, P), 2568 (Е); Aapies- river, да арія 4333 (В, К), навик Rebmann 5576 (B, d Wilms ж-з (МУ, МАТ. Wels 1844, Lycopodium rupestre b, Drége (B, K ); У. Gerrard (P); J.M - Wood 11950 (F). ORANGE FREE STATE: Rebmann 3040 (Р). сарЕ Е GOOD HOPE: Pordolusd, Қан Bacbmann 0 (B, P). Series Rupestres Tryon, ser. nov. Rhizomata et stolones nulla, gemmae breves simplices ad bases caulium rare praesentes vel absentes. Caules prostrati, ramis rare pendulis. Apices ramorum recti vel leviter curvati statu inerte. Caules frondosi leviter dorsiventrales vel radialiter symmetricales. Folia base decurrente vel valde decurrente. Typus: Selaginella rupestris (L.) Spring. Plants terrestrial or, in $. oregana, usually epiphytic; rhizomes and stolons absent, short, simple basal buds absent or, in $. oregana, occasionally present. Stems prostrate with rhizophores produced generally throughout or, in S. oregana, 19551 TRYON-—SELAGINELLA RUPESTRIS AND ITS ALLIES 59 the branches usually pendent with rhizophores only on the prostrate main stems. Branches straight or slightly curled in the dormant state or, in S. oregana, strongly curled; branch tips straight or slightly curved in the dormant state. Leafy stems radially symmetrical, the leaves equal in position, length and shape on all sides on the same portion of the stem, or dorsiventral in position, the under leaves loosely appressed, the upper erect-ascending, and sometimes also in length, the under slightly longer than the upper; zone of green leaves equal or nearly equal on all sides of the stem. Leaf-bases strongly decurrent on all sides of the stem or those on the upper side decurrent. Setae of the sporophylls terete or oval at the base or often, in $. densa and S. sibirica, with the base strongly broadened and flattened. Megaspores 4 in a megasporangium; rarely 1—2, or in S. rupestris commonly 1-2. The Rupestres is а rather homogeneous series. Although the species are suf- ficiently distinctive they do not differ from each other by as many characters as the members of the other series. The last three species, $. utahensis, S. leucobryoides and S. asprella, share the peculiar character of easily fragmenting stems and this is also present, although to a lesser degree, in S. Watsonii. These four species form the only definite group within the series. S. Vardei, with some of the leaves rarely with abruptly adnate bases, may illustrate a transition from Arenicolae. The dorsiventral S. densa appears to be an example of parallel evolution and not related by that character to Eremophilae or the dorsiventral Sarforii, S. rupestris is notable for its development of apogamy and $. densa and $. sibirica rarely show a tendency toward such a development. KEY TO SPECIES a. Dry leafy stems persistently whole, not fragmenting; stems forming cushion mats with discrete branches, or spreading mats with intricate branches, or pendent, forming festoons, or if forming cushion mats with intricate branches then the apex of the upper leaves fleshy. b. b. Stems forming flat mats with discrete branches or spreading mats with intricate branches, or pendent, forming festoons; apex of the upper leaves herbaceous to slightly fleshy, plane to abruptly beveled in profile; or if the apex of the upper leaves fleshy and truncate in profile then the setae of the leaves 14 to over № as long as the blade and the broadest sporophylls about 4 times as broad as the leaves. c. с. Stems elongate, branches long and remote, intricate; apex of „а upper leaves plane in profile; broadest зрогорћу 5 mo 4 times a broad as the leaves. China... 9. S. Verde | p. 61 с. Stems short, the branches short, approximate, Tubus or the apex of the upper leaves beveled to truncate in profile; or the broadest sporo- phylls about 2 times as broad as the leaves. d. d. Plants usually epiphytic, the stems pendent with rhizophores only at or near the base; leaves adnate to the stem for V4 to usually У to nearly № their length; branches strongly curled in the dormant state, Coastal Washington to northern California ..30. S. oregana, р. 61 [Vor. 42 60 ANNALS OF THE MISSOURI BOTANICAL GARDEN d. Plants terrestrial, usually with rhizophores throughout; upper leaves adnate to the stem for 1 to rarely М their length; branches not or slightly curled in the dormant state. e e. Apex of upper leaves nearly plane to abruptly beveled in profile, or if truncate then the setae lutescent and the stems forming compact flat mats with discrete branches. f f. Upper and under leaves essentially or quite equal in length on the same portion of the stem; stems forming open, spreading mats with intricate branches; base of setae of the sporophylls not or slightly broadened and flattened. g . Broadest sporophylls about 2 times as broad as the leaves; sporophylls and leaves eciliate or with cilia strongly ascend- ing and dentiform toward the apex. Texas to Arizona, north to Wyoming. Lun 31. S. Underwoodii, p . Broadest sporophylls about 4 times as broad as the leaves; sporophylls and leaves with cilia spreading to laxly ascend- ing and piliform toward the apex. Georgia to Greenland, west to Arkansas, Nebraska and Alberta. .32. S. rupestris, p. f. Upper and under leaves unequal in length on the same portion of the stem, the under definitely longer; stems forming com- pact mats with discrete branches; base of setae of the sporo- phylls often strongly broadened and flattened. Texas to Arizona; California; north to Saskatchewan, Alaska and British Columbia 33. S. densa, p. e. Apex of the upper leaves truncate in profile; setae white to tawny; stems forming open, spreading mats with intricate branches. Yukon to Siberia and Јарап 34. S. sibirica, p. b. Stems forming rounded cushion mats with intricate branches; apex of the upper leaves fleshy, truncate or subtruncate in profile; setae of the leaves 15 to V4, rarely 14, as long as the blade; broadest sporophylls about 2 times as broad as the leaves. California to Oregon and Montana 35. S. Watsonii, p. a. Dry leafy stems readily fragmenting; stems forming cushion mats with usually intricate branches; apex of the upper leaves herbaceous to slightly fleshy. Southwestern Utah to southern California. h. Setae not forming a conspicuous tuft at the growing-tip, 14 or less as long as the blade. i. i. Setae essentially absent or, if present, then smooth, whitish to greenish- or lutescent-whitish, subopaque. Southwestern Utah and southeastern Nevada 36. S. utabensis, p. 7 i. Setae scabrous, white, usually opaque. Southeastern California 37. S. leucobryoides, p. 7 09 99 ON ~ о ы ON с ~ = N N + ^ 1955] TRYON-—SELAGINELLA RUPESTRIS AND ITS ALLIES 61 h. Setae forming a conspicuous tuft at the growing-tip, 14 to more than V5 as long as the blade, scabrous, white, usually 2. Southern California . . S. asprella, p. 75 29. SELAGINELLA VARDEI Lév. Cat. Pl. Yun-Nan, 172. 1917. (Holotype: Maire 56 E) Fig. 37. Map 39. Stems long, forming open, spreading mats; branches long, remote, intricate, not fragmenting when dry. Upper and under leaves about equal in length on the same portion of the stem; leaves subulate to subulate-long-triangular; base pubescent, that of the upper leaves adnate to the stem for 1⁄4 or less their length; margins ciliate, the cilia strongly ascending toward the apex of the blade; apex of the upper leaves herbaceous, flat to slightly rounded, Вам or nearly so in profile; setae form- ing conspicuous tufts at the dry branch-tips, № to № as long as the blade, usually slightly scabrous, tawny to whitish-tawny, translucent. Sporophylls with the cilia dentiform and ascending toward the apex, the broadest about 4 times as broad as the leaves. Megaspores unknown On the basis of its generalized characters, this species is placed as the most primitive in the series. One specimen examined had several of the leaves abruptly adnate and distinct in color at the base, characters of the previous series, Sartorii. In the material examined only microsporangia were found and these occurred in the basal sporophylls as well as those above. The explanation of this in terms of the reproduction of the species is not clear. It suggests that the stems may be monoecious or dioecious but fertilization would be extremely unlikely under such circumstances. The localities taken from the literature (map 39) are from Alston, in Bull. Fan Mem. Inst. 5:267. 1934 Exposed rocky places, 1500—3800 m. Szechwan, Yunnan and Tibet. Specimens examined: HINA. san dfe AN: H. Smith 2394 (ВМ); E. H. Wilson 5411 (P). YUNNAN: J. У. C. J. Gregory (BM); Siao-ou-long, 2900 m., Oct. 1912, Мате 56 (ВМ). тівет: Lud- low et al. boe Хам), 5404 (ВМ), 14233 (ВМ). 30. SELAGINELLA ОКЕСАМА D. C. Eaton, in Wats. Bot. Calif. 2:350. 1880. (Lectotype Бу Weath. in Jour. Arn. Arb. 25:411. 1944: Kautz YU!. Para- type: Summers 2200 YU!). Fig. 38. Map 40. Plants epiphytic with rhizophores only at the base of the stems, less often terrestrial with rhizophores borne throughout. Stems very long to long, forming pendent festoons; branches long, remote, intricate or discrete, strongly curled in the dormant state, not fragmenting when dry; or terrestrial stems forming an irregular mat with intricate branches. Upper and under leaves essentially or quite equal in length on the same portion of the stem; leaves long-triangular to ovate- triangular, excluding the adnate base; base glabrous, rarely pubescent, that of the upper leaves adnate to the stem for У, to nearly У; their length; margins eciliate to ascending-ciliate toward the apex; apex of the upper leaves slightly fleshy, [Vor. 42 62 ANNALS OF THE MISSOURI BOTANICAL GARDEN rounded to narrowly carinate, plane to gently beveled in profile; setae not or hardly forming conspicuous tufts at the dry branch-tips, about № as long as the blade, smooth, greenish, whitish or lutescent, translucent to rarely subopaque. Sporophylls . eciliate toward the apex, the broadest about 2 times as broad as the leaves. Мера- spores rugose-reticulate with thin rugae on the commissural face, less marked on the outer face, pale yellow. S. oregana is the only species of the section that is commonly an epiphyte and its long pendent branches are characteristic. Perhaps in relation to the length of the branches, the leaf-base is unusually long and the desiccated branches curl to form ringlets. The last two characters are sufficient for the identification of the occasional plants that grow on soil or rock. he specimen collected by Scouler (335, GH, NY) and frequently cited can not be accepted as coming from Observatory Inlet as stated on the label since this locality is so far north of the otherwise known range. In megaspore characters S. oregana shows a relationship to S. Underwoodii, which occasionally has pendent branches although it is not an epiphyte. Pendent from mossy trunks or branches of trees, particularly Acer macro- phyllum, or on shaded rocky banks, sea level to 200 m. Coastal Washington to northern California. Representative specimens: UNITED STATES. WASHINGTON: А. А. 8 E. С. Heller 4002 (MO, NY, US); Otis 1399 (YU); Thompson 4115 (MO, NY, S» Часы (МО, МУ, US). Le Eastwood 12185 (Е, СН); T. Howell 696 (GH, US); Н. E. Parks 24274 (Е, СН, MO, NY); H. Е. 8 S. T. Parks 24119 (Е, СН, NY, US). 31. SELAGINELLA UNDERWoopn Hieron. in Engl. & Prantl, Nat. Pflanz. 1*:714. 1901, based on S. rupestris var. Fendleri Underw. Figs. 39, 40. Map 41. a rupestris (L.) Spring var. ин Underw. in Bull. Torr. Bot. Club 25:127. Lectotype кен Weath. in Jour. Arn. Ar b. 25:412. 1944: Fendler 1024 NY!. ratypes: Baker 2 NY!; Wooton in 1892 NY! ам Fendleri за Ніегоп. іп Hedwigia 39:503. 1900, not Baker, 1887. Selaginella Underwoodii var. сены Weath. in Jour. Arn. Arb. 25:412. 1944. (Holotype: Metcalfe 276 GH! fragment US!. Pn ypes: Ferriss in 1904 GH!; Good- ding 5 US! fragment GH!; Maguire et al. 11745 GH!; Metcalfe 711 US! fragment СНІ, 991 GH!). Stems long or moderately long, forming open, spreading mats, rarely pendent and forming festoons; branches long and remote or moderately so, intricate, not fragmenting when dry. Upper and under leaves equal or subequal in length on the same portion of the stem; leaves subulate to linear to ligulate-long-triangular; base glabrous to pubescent, that of the upper leaves adnate to the stem for about И; their length; margins ciliate, the cilia ascending toward the apex; apex of the 19551 TRYON-—SELAGINELLA RUPESTRIS AND ITS ALLIES 63 A 2 | LES Gy Ly Sd У; n "4L S UNDERWOODII ^ ` - ~ [Vor. 42 64 ANNALS OF THE MISSOURI BOTANICAL GARDEN upper leaves herbaceous to slightly fleshy, slightly rounded to narrowly carinate, plane to abruptly beveled in profile; setae sometimes forming conspicuous tufts at the dry branch-tips, № as long to nearly as long as the blade, usually smooth, some- times scabrous, whitish, greenish-white to lutescent, rarely milk-white, usually translucent to rarely opaque. Sporophylls eciliate or with cilia dentiform and ascending toward the apex, the broadest usually 2 to rarely 3 times as broad as the leaves. Megaspores rugose-reticulate оп the commissural face, more prominently marked, often with thin rugae, on the outer face, pale orange. On the basis of the material I have studied I am not able to recognize var. dolichotricha. Те is a considerably less well-defined entity than others I have гес- ognized as varieties. S. Underwoodii from northern Arizona, northern New Mexico and northward is rather uniform in having short and well-differentiated setae and mostly short cilia on the leaves. However, to the south this extreme is also found with the phase that has long and poorly differentiated setae and long cilia. The three characters do not correlate as well as one would like in this area. This is a close parallel to S. mutica but in that species the differentiation has proceeded further and varieties are recognized. S. Underwoodii often grows with S. mutica and as mentioned under that species the S. Underwoodii in such mixed mats eventually disappears. Moist or shaded cliffs or rocky slopes, rocky alpine meadows, in crevices or on ledges or among rocks, usually on granitic or other igneous rocks, less often on sandstone or limestone, 800—4000 m., usually from 2000—3000 m. Texas to Arizona, north to Wyoming. Representative specimens: UN STATES. WYOMING: Payson 2503 (US). coronapo: C. Е. Baker 2 (NY, СЕН C. S. Crandall 1054 (МУ, US); Jobnston 2424 (GH, 95), 2425 (GH, NY, US); R. M 8 А. Е. Tryon 5074 (ARIZ, B, BM, CU, Е, FI, СН, К, MO, NY, P, РИ, ge UC, US, WS), 5076 (BM, DS, GH, MIN, MO, POM, US), 5078 (B, BM, F, GH, MICH, MO, AR e NEW MEXICO: Árséne 15872 (Е, МО), 17063 (Е, 05); 1874, Fendler 1024 (В, Е, СН, MO, NY, US); Mogollon Mountains, Mogollon Creek, Socorro Co., 8000 ft., July ЈЕ 1903, Metcalfe 276 (GH, MO, NY, US); Metcalfe 711 (GH, МО, NY, US), 901 (Е, GH, МО, NY, US); Richards & Drouet 363 (F, GH, MO, NY); March 1, 1892, Wooton (NY). ОКТАНОМА: Aug. 14, 1937, Wherry (US). TEXAS: Ferris & Duncan 3588 (NY, US); Hinckley тот (HL И 1156 sue) ARIZONA: Darrow & а 2781 (СН, MO, US); March, 1904, riss (GH); Goodding 5 (GH, US), 213 (GH, NY); MT et al. 11745 (GH, DD. - Phillips Же (СН, МО, 0$); Pultz 8 Phillips 2749 (GH,US). 32. SELAGINELLA RUPESTRIS (L.) Spring, in Mart. Fl. Bras. 12:118. 1840. Figs. 41, 42. Map 42. Lycopodium rupestre L. Sp. Pl. 2:1101. 1753. eee Kalm LINN!. Kamtchatcha, Steller LINN! is excluded as a type; it is S. sibiric Stachygynandrum rupestre (L.) Beauv. Prod. Ааа. 110. 1805. Selaginella Bourgeaui Hieron. іп Hedwigia 39:295. 1900, as Bourgeauii. (Lectotype: end үк» on Aug. 15, 1857 В!; Black | Hills, Hayden MO! is an excellent match. Рага- geau on Aug. 14 1857, in part, B! fragment NY!). uoc. бно Y зон (Hieron.) Clute, in Fern Bull. 16:52. 1908. 19551 TRYON-—SELAGINELLA RUPESTRIS AND ITS ALLIES 65 Stems long to rather short, forming open, spreading mats; branches long and remote to moderately short and subapproximate, intricate, rarely only slightly so, not fragmenting when dry. Upper and under leaves equal or subequal, rarely the under slightly longer, on the same portion of the stem; leaves subulate to ligulate- long-triangular; base usually pubescent, rarely glabrous, that of the upper leaves adnate to the stem for 14 to 1⁄4 their length; margins ciliate, the cilia spreading toward the apex; apex of the upper leaves herbaceous, rounded, gently to abruptly beveled in profile; setae forming conspicuous tufts at the dry branch-tips, nearly V2 to 34 as long as the blade, scabrous, usually strongly so, milk-white to tawny, opaque to rarely subopaque. Sporophylls with the cilia piliform and spreading toward the apex, rarely laxly ascending, the broadest about 4 times as broad as the leaves. Megaspores usually 1—2, less often 3 or 4 in a sporangium, rugose to rugose- reticulate, more finely marked on the outer face, bright orange. S. rupestris is the only species that is certainly apogamous. Та the Appalachian mountain region as far north as southern Pennsylvania most of the material has four megaspores in a sporangium and microsporangia are present in the strobilus. Such plants are presumably sexual. In some plants the spore number in the sporangia of а strobilus may vary from 1 to 4. Those having 1-2 megaspores in the sporangia and a few microsporangia are less common. Throughout the rest of the range the most frequent type bears only megasporangia with either 1 or 2 megaspores. This must be an obligate apomict and it is the only kind in a broad band on the northern periphery of the range. From Missouri to Michigan and Vermont there rarely occur specimens that bear strobili having a few micro- sporangia. The widespread apogamy may explain the uniformity of S. rupestris over its broad range. All other wide-ranging species are much more variable. The occasional occurrence of 1—2 megaspores in sporangia of S. densa and S. sibirica may also indicate apogamy. S. rupestris most closely resembles S. densa var. densa and it can best be sep- arated by its radially symmetrical leafy stem. In addition, in the area in Canada where the two grow together, all of the S. rupestris have one or two megaspores in а sporangium while $. densa has, with rare exception, four. When S. rupestris grows in nearly рше! D and other vegetation is sparse it will form circular, or in age, ring-shaped m Exposed or less us ied cliffs, rocky bluffs, of acidic igneous or sedimentary rocks, gravel or sandy soil, up to 1900 m. Central and eastern United States to northeastern Alberta, Quebec and Green- 4. Representative specimens: GREENLAND: July 13, 1946, Т. W. Bócher (GH). NADA. Nova 5СОТІА: Fernald & Long 23098 (СН, TL, гв Victorin 8421 СА (МО, 05); Victorin et al. 4201 (МО, US). ontario: С. Е. 8 С. К. Jennings 7362 US), 7380 (GH, US); July 20, 1899, Umbach (MIN, NT. US); Van Eseltine 501 (GH, MANITOBA: Fort Ellice, Aug. 15, 1857, Bourgeau (B), Aug. 14, 1857 (B, NY); о 137 (BM); Масоип 8 Herriot 70372 (В, Е, СН, МУ). SASKATCHEWAN: [Vor. 42 66 ANNALS OF THE MISSOURI BOTANICAL GARDEN Каир 6305 (СН, NY), бого (GH, NY). ALBERTA: Каир & Abbe 4451 (GH, NY), 4608 (GH, NY). UNITED STATES. MAINE: Hodgdon (Pl. Exsicc. Gray. 611) (Е, СН, MO, NY, US). VERMONT: Sept. 11, 1920, Dutton (F, E MO); Eggleston 21909 (GH, NY NECTICUT: 1859, D. C. Eaton (Е, СН, MO). NEw хокк: Robinson 8 Maxon 15 (ОН, МУ, US), 32 (Е, СН, MO, NY, US). РЕММ5УГУАМЛА: Heller 8 Halbach 706 (СН, MO, US). VIRGINIA: е 8 Steele 166 (GH,MO,NY,US). NORTH CAROLINA: Biltmore Herb. 34328 (NY, US). SOUTH CAROLINA: Mackenzie 2007 (МО, 05); Clausen 8 Trapido 3663 (NY, US). GEORGIA: Е 198 (US); Correll 6615 (СН, МО, 05); Натрет 215 (СН, МО, NY). АТАВАМА: Harper 3403 (Е, MO). MicHicaN: В. W. Chaney 181 (Е, GH, св К. М. ӘР. F. Tryon 4695 (СН, MO). INDIANA: June 16, 1900, Umbach (GH, US). TENNESSEE: Сай трет (Curtiss, М. „Аш Р]. 3796) m СН, МО, МУ, US). wis- CONSIN: R. . F. Tryon 5005 (B, ‚ СН, K, ). ILLINOIS: А. Chase 1584 (Е, ае Ever 32650 (MO). MINNESOTA: Bergman "1315 56 (СН, МУ, 05); Моује 483 (Е, ). MISSOURI: Bush 4744 (GH, MO, МУ, US), 5210 (СН, NY, US). == = di Balmer 35561 (Е, СН, MO, МҮ). зоџтн pakora: Murdoch 4304 (Е, ‚ US). NEB : Kiener 11323 (US), 23095 (МО). Kansas: 1886, Е. М, Plank 14 (СН). ОКТАНОМА: ЖЕ ЭР, 835 (СН, МО, МУ). 33. SELAGINELLA DENSA Rydb. in Mem. М. Y. Bot. Gard. 1:7. Feb. 15, 1900. (Holotype: Havard NY! marked as type by Rydb., fragment US!. Paratypes: Newberry NY!; Williams 534; Tweedy 172; Missoula, Mont. 1898, Williams 8 Griffitb; Silver Bow Co., Mont., Mrs. Jennie Moore; the last four collections probably at MONT. Watson in 1880 GH! is excluded as type material; it is S. Wallacei). Map 43. Stems short, forming flat cushion mats; branches short, approximate, discrete, not fragmenting when dry. Under leaves definitely longer than the upper on the same portion of the stem, or sometimes equal to the upper on the assurgent branch- tips; leaves ligulate to ligulate-lanceolate to ligulate-long-triangular; base glabrous to pubescent, that of the upper leaves adnate to the stem for 14 to V4 their length; margins eciliate or with the cilia ascending to spreading toward the apex; apex of the upper leaves herbaceous to fleshy, slightly to broadly rounded, plane to truncate in profile; setae forming conspicuous tufts at the dry branch-tips, У, to 34 as long as the blade, smooth to scabrous, milk-white and opaque to tawny or whitish and subopaque to lutescent and translucent. Sporophylls eciliate or with the cilia dentiform to piliform and ascending, rarely laxly ascending, toward the apex, the broadest 3 to 4 times as broad as the leaves; seta base often strongly broadened and flattened. Megaspores prominently and coarsely rugose-reticulate to slightly rugose, most prominently marked in the equatorial region, pale to bright orange. S. densa is one of the most complex species of the section. The morphological extremes of the three varieties are amply distinct, but from Montana south to Colorado all occur and with an increasing number of intermediates. In Colorado nearly all possible intermediate conditions between the three may be found and over a third of the collections I have seen from there are such intermediates. To the south, in the mountains of Arizona, New Mexico and Texas, there is less variation. War. Standleyi and var. densa are lacking and there are no intermediates with var. Standleyi. The material there is either typical var. scopulorum or inter- mediates of it varying toward var. densa. 19551 : TRYON—SELAGINELLA RUPESTRIS AND ITS ALLIES 67 It is not possible to withhold recognition from the three extremes since they do have distinctive characters and distribution in the northern part of the range of the species. At the same time, it must be admitted that identification, particularly of the material from Colorado, is sometimes rather arbitrary. The intermediates show all transitions between the two extremes in respect to the varietal characters. Some specimens will be intermediate in one character, others in two or three. Some will have some of the leaves and sporophylls intermediate, others typical of one variety; or some leaves and sporophylls will be typical of one variety, others of them typical of another variety. The intermediate specimens have, for the purposes of discussion, been divided into two kinds. Those that are more or less halfway between the typical state of two varieties are called intermediate. Those that depart from the typical state of a variety to a lesser degree are mentioned as specimens showing a tendency toward another variety. Such specimens can be identified with one variety quite satis- factorily but are not entirely typical of it. Of the three varieties, var. densa most closely resembles $. rupestris and the differences are discussed under that species. Rarely S. densa may have some mega- sporangia with two megaspores but this has not been found with sufficient reg- ularity within a strobilus to indicate functional apogamy as in S. rupestris. Southwestern Manitoba to southern Alaska, south to Texas, Arizona and northern California. KEY TO VARIETIES a. Apices of the leaves and sporophylls plane to abruptly beveled in profile; setae of the leaves white or whitish, or lutescent only at the base, opaque to subopaque. b. b. Sporophylls eciliate toward the apex 33a. var. scopulorum, p. 67 b. Sporophylls dentiform to piliform-ciliate to the арех __336. var. densa, р. 68 a. Apices of the leaves and sporophylls predominantly or all truncate in pro- file; setae of the leaves whitish- to greenish- to entirely lutescent, trans- lucent 33c. var. Standleyi, p. 71 / 33а. SELAGINELLA DENSA var. scopulorum (Maxon) Tryon, comb. nov. Fig. 43. Map 44. Lycopodium bryoides Nutt. ex Baker, Handb. Fern Allies, 35. 1887, in synon. Placed here on the basis of the specimen so labeled at Kew! (fragment NY!) which agrees with Ba Selaginella scopulorum Maxon, in Amer. Fern Jour. 11:36. 1921. (Holotype: Stand. Heacock 2 earns US!; Merrill 8 Wilcox 1218 US!; Shaw 308 US!, 902 US!, pee а Ts № 5; 09, d US!, 16216 US!, 16255 USt, "6288" US!, 16378 x 17055 US!, her US!, 17979 US!, оа 081, 18185 USt; Suksdorf 8834 091; Ulke on Aug. $s зн Umbach 856 US!). АА ру је а nig А. Е Sia in Amer. Fern Jour. 11:37. 1921, in synon. (Evidently based on Shaw pe and Buen 235). [Vor. 42 68 ANNALS OF THE MISSOURI BOTANICAL GARDEN Apex of leaves plane to abruptly beveled in profile; setae white, or lutescent only at the base, opaque, rarely subtranslucent. Sporophylls eciliate toward the apex; apex plane to gently beveled in profile. Маг. scopulorum is apparently the least specialized of the varieties of S. densa. Intermediates with var. densa in the central and southern Rocky Mountains are frequent but intermediates with var. Standleyi are rare. The following are examples of intermediates between var. scopulorum and var. densa. New Mexico, Arsène & Benedict 16370 (US); Arizona, Phillips & Reynolds 2000 (GH); Colorado, Underwood & Selby 133 (NY); Utah, Hermann 5063 (MO). Examples of specimens of var. scopulorum with a tendency toward var. densa are: Colorado, Knowlton 82 (US), McKelvey 4695 (GH); Utah, Harrison 8 Larsen 7891 (MO); Arizona, Phillips & Reynolds 2900 (US). Colorado, Rydberg & Vreeland 6588 (NY) represents var. scopulorum with a tendency toward var. Standleyi. Intermediates between var. scopulorum and var. Standleyi are mentioned under that variety while specimens of var. densa with a tendency toward var. scopulorum are mentioned under var. densa Usually in rocky alpine tundra, also cliffs, talus slopes, on ledges, among boulders or in thin soil over rocks, on igneous or sedimentary rocks, 700-4660 m. Alberta to British Columbia, south to Texas, Arizona and northern California. Representative specimens: ANADA. ALBERTA: Scamman 2700 (GH); Rosendahl 1074 (US). BRITISH COLUMBIA: Heacock 235 (BM, GH, MO, NY, US); Hitchcock 8 Martin 7435 (GH, MO, МУ); VUE 398 (BM, GH, МО, NY, US), 902 (BM, GH, NY, US), тобо (BM, GH, MO, NY, U STATES. MONTANA: Standley 153184 e 15508 (US), vicinity of Cracker Lake, Glacier National Park, 1740-1920 m., July 15, 1919, Standley 15732 (US), Stand- ley 16216 (US), 16255 (US), 16288 (US), dd (US), 17055 (US), 171642 (US), 17979 (US), 18050 (US), 18185 (US); Aug. 25, 1917, Ulke (US); Umbacb 856 (F, ). чүомімс: Gooddin ғ 483 (F, GH, MO, NY, US); Mearns 4274 (US); Merrill & Wilcox 1218 (СН, NY, US). COLORADO: Cox 290 (F, MO); Murdocb 4787 Е, US). : E. B. 8 L. B. Payson 4037 (СН, MO). Texas: Moore & Steyermark 3245 (US). | NEW MEXICO: oe 1172 з Е, GH, MIN, MO, NY, US). ARIZONA: 33b. SELAGINELLA DENSA var. densa. Figs. 44, 45, 47. Map 45. олу ape longipila Hieron. in Hedwigia 39:291. Dec. 28, 1900. (Lectotype: Bourgeau . Kew Exsicc. no. 1239, wrongly labeled "Himalaya") B!. Paratype: "Herb. Griffith, Bhotan" B! fragment NY!, collector and locality unknown). prime ді кесе (L.) Spring f. longipila A. Br. ex Hieron. in Hedwigia 39:291. 1900, Selaginella. үз сыы Hieron. in Hedwigia 39:294. Dec. 28, 1900. (Holotype: Engel- mann B!). 1955] 69 TRYON-—SELAGINELLA RUPESTRIS AND ITS ALLIES Cu MI D 54 т ~ ^44. 5. DENSA VAR. 75 ""SCOPULORUM | | TJ a а i 7 а / 43. 5. ПЕМЗА 45. S.DENSA VAR. DENSA 46. S.DENSA VAR. [Vor. 42 70 ANNALS OF THE MISSOURI BOTANICAL GARDEN Selaginella Haydenii Hieron. in Hedwigia 39:296. = 28, 1900, as Haydeni. (Lecto- type en B!. Paratype: Lyall B! fragment Selaginella rupestris var. densa (Rydb.) Clute, Fern Allies, 142. 1905. Selaginella rupestris f. Engelmannii (Hieron.) Clute, in Fern Bull. 16:52. 1908. Selaginella rupestris f£. Haydenii (Hieron.) Clute, in Fern Bull. 16:52. 1908. Apex of leaves plane to abruptly beveled in profile; setae milk-white and opaque to whitish and subopaque. Sporophylls with the cilia dentiform to piliform and ascending, rarely laxly ascending, toward the apex; apex plane to abruptly beveled in profile The misapplication of S. longipila to the Himalayan S. indica, because of a mixture of labels has been discussed under S. indica. The type specimens аге. actually 5. densa var. densa but fortunately Rydberg’s well-known name has a few months priority. Intermediates with var. scopulorum are mentioned under that variety and specimens of var. Standleyi that show a tendency toward var. densa are mentioned under var. Standleyi. Specimens that are intermediate with var. Standleyi are such as: Colorado, Aug. 26, 1896, Holm (MO), Sept. 12, 1905, S. Shaw (GH), July 6, 1937, Wherry (US). Specimens of var. densa that show a tendency toward var. scopulorum are such s: Colorado, Е. E. & E. S. Clements 532 (ОН,МО, МУ, 05) ; New Mexico, Arsène 8 Benedict 15877 (Е, US), Fendler 1025 (ВМ, СН, MO T n example of var. densa with a tendency toward var. Standleyi is: Empire, Colorado, Aug. 27, 1874, Engelmann (B, MO, US). The following specimens are rather intermediate between all three varieties: Wyoming, Hermann 4665 (MO); Colorado, Ewan 12699 (Ewan); New Mexico, Arséne & Benedict 18056 (Е). Prairies, alpine meadows or dry rocky places, acidic rocks or sandy soil, 1100— 4000 m. Southeastern Manitoba to British Columbia, south to New Mexico and Arizona. Representative specimens: NADA. MANITOBA: Macoun & Herriot 70373 (Е, СН, NY). SASKATCHEWAN: Boiv 8 єл 8686 (MO); Fort Carlton, March 29, 1858, Bourgeau (Herb. Kew Exsicc M 5553 (МО); S. Brown 128 (GH, MO, US); Malte 8 Watson 1220 (СН), BRITISH COLUMBIA: Calder & Savile 7677 (MO), 9221 (МО). м TATES. "Oregon", 49 М. Lat., 1858-59, Lyall (B, NY). NoRTH DAKOTA Aug. 12, 1204, Lunell (GH, MIN, US), Sept. 10, 1908 (МҮ, US). SOUTH DAKOTA: Black Hills, 1853—54, F. V. Hayden (B, MO, NY, US); Hayward 315 (F, NY). MONTANA: Little Rocky Милене 5 Sept. 1889, Havard (NY, US); Newberry (NY); Rydberg 8 Bessey 3517 (Е, GH, MIN, NY, US); В. S. Williams 534 (US). wvoMiNe: А. Nels 8781 (Е, СН, MO, NY, US). согокаро: Johnston 3806 і 3297 (US), 3898 (US); Кий 8 Vreeland 6500 (NY); І. & К. Williams 2126 (СН, MO). 1955] TRYON-—SELAGINELLA RUPESTRIS AND ITS ALLIES 71 У 33с. SELAGINELLA DENSA var. Standleyi (Maxon) Tryon, comb. nov. Fig. 46. Map 46. а Standleyi Maxon, in Smiths, Misc. Coll. 725:9. 1920. (Holotype: Standley 228 US!. Paratypes: Brown 95 pee Standley 15363 US, 16970 US!, 17483 US!, xen US!; Ulke on Aug. 25, 1917 US!). Apex of leaves predominantly or entirely truncate in profile; setae whitish- to greenish-lutescent to lutescent, translucent. Sporophylls eciliate or with the cilia dentiform or piliform and ascending toward the apex; apex truncate, rarely abruptly beveled in profile. Intermediates of var. Standleyi and var. densa and material of var. densa with a tendency toward var. Standleyi are mentioned under var. densa. Specimens of var. scopulorum with a tendency toward var. Standleyi are mentioned under var. scopulorum. Specimens of var. Standleyi with a tendency toward var. densa are such as: Montana, Standley 16970 (US), Van Schaack 27867; (MO); Colorado, Johnston 3899 (GH). Specimens intermediate between var. Standleyi and var. salina are: Mon- tana, Aug. 11, 1931, J. H. Schaffner (BM); Colorado, Underwood & Selby 133 (NY), Vasey (MO). Rocky alpine meadows, dry rocky slopes and cliffs, acidic rocks, 1500-4660 m. Southern Alaska to Colorado. Representative specimens: ALASKA: А. & А. Krause 162 (B). ALBERTA: S. Brown (GH, MO, US). вкітізн COLUMBIA: Aug. 30, 1904, TES. MONTANA: Barkley 1725 (GH, M , US); Standley 15363 (US), vicinity of Sexton Glacier, Glacier National Park, jv “ 1919, Standley 17228 к. Standley dir T 18136 (US); Aug. 25, 1917, Ulke (US). СОГОКАРО: Jo. 903 (NY, U 34. ied sIBIRICA (Milde) Hieron. in Hedwigia 39:290. 1900. Fig. 48. Map 47. Selaginella rupestris (L.) Spring £. sibirica Milde, Fil. Europ. Atlant. 262. 1867. (Lecto- type: Unalaska, Chamisso. Paratypes: Ajan, Tilling; Dahuria ad flumen Ingoda, Pallas). көне pie се f. amurensis Milde, Fil. Europ. Atlant. 262. 1867. (Holotype: Amur, Sternen rupestris 6. manchuriensis Milde, Fil. Europ. Atlant. 262. 1867. (Holotype: Khalkyli, M | чизи Schmidt Hieron. in Hedwigia 39: = 1900. (Lectotype: Schmidt B! frag- nt NY!. Paratype: Chamisso B! fragmen !). Selaginella Schmidtii var. Krauseorum Hieron. i in Hedwigia 39:293. 1900. (Lectotype: A. Krause 53 B! fragment NY!. А. 8 A. Krause 162 В! is excluded as a type; S. densa var. Standleyi). Selaginell rupestris f. Schmidtii (Hieron.) Clute, in Fern Bull. 16:52. 1908. Stems long to moderately long, forming open, spreading mats; branches long and remote to moderately short and subapproximate, intricate, not fragmenting when dry. Upper and under leaves equal or subequal in length, or the under [Vor. 42 72 ANNALS OF THE MISSOURI BOTANICAL GARDEN slightly longer, on the same portion of the stem; leaves linear to ligulate-long- triangular; base usually glabrous, less often pubescent, that of the upper leaves adnate to the stem for М; their length; margins ciliate, the cilia spreading to laxly ascending toward the apex; apex of the upper leaves fleshy, broadly rounded to carinate, subtruncate to usually truncate in profile; setae forming conspicuous tufts at the dry branch-tips, 14 to 95 as long as the blade, scabrous, milk-white to white to tawny, opaque to translucent. Sporophylls eciliate or with the cilia piliform to dentiform and ascending toward the apex, the broadest 3 to usually 4 times as broad as the leaves; seta base usually strongly broadened and flattened. Megaspores rugose to rugose-reticulate, pale yellow to pale orange. The Alaskan and Yukon material is relatively uniform while that from Asia is less so. In addition to the typical form with short, milk-white setae a phase occurs in Asia with longer and tawny setae, and also occasional specimens bear strobili having 1 to 2 megaspores in a sporangium. This is quite parallel with S. densa var. densa, the reduced number of megaspores being so rare that it is not possible to be certain of apogamy. S. sibirica is most similar to S. densa var. Standleyi from which it may be sep- arated by the white to tawny rather than lutescent setae and the intricate rather than discrete branches. It is separated from S. shakotanensis, which also grows in Japan, by the leaf-bases that place the two in different series. The seta length likewise distinguishes these species—in S. sibirica they аге 16—36 as long as the lade while in S. shakotanensis they аге V5 as long The only sporelings observed in this study were seen in the soil of a mat of S. sibirica collected by Calder & Billard (2005 MO). These are discussed in some detail in the introduction. Dry, rocky, open places ог on cliffs, 130-2400 m. Yukon and Alaska to Trans Baikal, Manchuria and Japan. 2. specimens: NADA. YUKON: Calder 9 Billard 2005 (MO), 4454 (MO), 4601 (MO, US); Tarle- ton pc (NY, US). ALASKA: UNALASKA, Chamisso (B, о, Flett 1520 (NY, US); А. Е. 6 В. Т. Porsild 690 ео МО, US); Scan 1981 (G O). BEHRING STRAITS: Emm ше ees 21, 1881, A. 8 A. Krause 53 (B, NY); C. Wright (GH, NY, US, NION OF SOVIET SOCIALIST REPUBLICS, KAMTCHATKA: Eyerdam (F); Novogeblenov 541 (US). 5АСНАТЛМ: Fr. Schmidt (B, СН, NY, US). TRANS BAIKAL: Ingoda river TCI): Fischer ЈАР м: Faurie fe зы (Р), 8406 (МО, Р), 0687 (МО, Р), 13003 (Р). 35. SELAGINELLA WarsoNm Underw. in Bull. Torr. Bot. Club 25:127. 1898, as Watsoni. (Holotype: Watson in 1869 NY!. Paratypes: Watson in 1868 NY!; Hansen 879 NY!; Coville & Funston 2071 NY!; Brewer 2103 NY!). Figs. 49—51. Map 48. Stems short to moderately long, forming rounded cushion mats; branches short, subapproximate to moderately long and remote, intricate, not readily fragmenting when dry but rather easily broken by hand. Upper and under leaves equal in 1955] TRYON- —SELAGINELLA RUPESTRIS AND ITS ALLIES 73 length, or the under slightly longer, on the same portion of the stem; leaves ligulate to broadly ligulate to ligulate-long-triangular; base glabrous to slightly pubescent, that of the upper leaves adnate to the stem for М to 14 their length; margins eciliate, rarely with ascending cilia toward the apex; apex of the upper leaves fleshy, narrowly to broadly carinate, subtruncate to truncate in profile; setae not or hardly forming conspicuous tufts at the dry branch-tips, 15 to 1⁄4, rarely 14, as long as the blade, usually smooth, sometimes slightly scabrous, greenish-white to greenish- lutescent to whitish-lutescent, translucent to subopaque. Sporophylls eciliate toward the apex, the broadest about 2 times as broad as the leaves. Megaspores rugose to rugose-reticulate, pale orange. S. Watsonii is characterized by its fleshy leaves which are usually eciliate and short-setate. In habit it forms dense rounded mats with intricate branches. The fragile nature of the stem of the next three species is poorly developed in S. Watsonii. 'The stems can be easily broken by hand but specimens do not fragment in packets or with ordinary handling. It seems probable that the next three species have been derived from S. Watsonii or a common ancestor. Exposed or shaded cliffs, talus slopes or rocky alpine meadows, in crevices or on boulders or among rocks, usually related to igneous rocks, rarely to limestone, 1800—4300 m., usually 3000—3700 m. Southwestern Montana to northeastern Oregon, south to Utah and southern California. 4 NUS ig HKA О AF FA 3 ANSP 1% [Vor. 42 74 ANNALS OF THE MISSOURI BOTANICAL GARDEN Representative specimens: UNITED STATEs. MONTANA: i D. Howe 66 (US). џтан: G. J. Goodman (GH, MO, NY, US); M. E. Jones 1246 (NY, US); Misano & Richards 13166 (GH, МО); Е. B L. В. Payson 4917 (GH, MO, 224; 5074 (СН, МО, NY, US); Rydberg 8 Carlton 6566 (GH, NY, US); Cottonwood Саћоп, 9500 ft., July, 1869, 5. Watson 1370 (NY, YU). NEVADA: Heller 11054 (F, GH, MO, NY, U 5); Maguire 21090 (СН, МУ, US); Sept., 1868, S. Watson 1370 (МУ, US). OREGON: July 1, 1931, Wherry (US). CALIFORNIA: R. M. & A. F. Tryon 5060 (ARIZ, B, BM, CU, DS, F, FI, GH, K, MICH, MIN, MO, NY, P, PH, POM, RM, UC, US, WS). 36. SELAGINELLA UTAHENSIs Flowers, in Amer. Fern Jour. 39:83. 1949. (Holo- type: Cottam 5644 UT fragment US!. Paratype: Cottam 8817 UT). Fig. 52. Map 49. Stems moderately short, forming rounded to flat cushion mats; branches mod- erately short to moderately long, approximate to nearly remote, intricate, readily fragmenting when dry. Upper and under leaves equal or subequal in length on the same portion of the stem; leaves subulate to ligulate-lanceolate to ligulate-long- triangular; base usually glabrous, sometimes pubescent, that of the upper leaves adnate to the stem for 1⁄4 their length; margins eciliate to ascending-ciliate toward the apex; apex of the upper leaves slightly fleshy, broadly carinate, gently to rather abruptly beveled in profile; setae not forming conspicuous tufts at the dry branch- tips, less than 14 as long as the blade to essentially absent, smooth, whitish to greenish- to lutescent-whitish, subopaque. Sporophylls eciliate or with the cilia dentiform and strongly ascending toward the apex, the broadest 2 to 3 times as broad as the leaves. Megaspores slightly rugose-reticulate to slightly rugose, yellow-orange. S. utabensis may be separated from the next two species, that also share the character of readily fragmenting stems, by its leaves which are muticous or have short, smooth setae. The anatomical basis of the fragmenting stems has not been investigated. In mounted material the parts will be held together if glue or soil Ids the roots and rhizophores firmly but in packets or in unmounted material ordinary handling of the sheet will soon reduce the stems to small pieces. Ledges and crevices of sandstone cliffs, 1500-2300 m. Southern Nevada and southwestern Utah. Specimens examined: Ом STATES. UTAH: y Mountain, Zion National Park, 6, һар ft., April 5, 1931, Ww. P. Cities 5644 (MO gee ; аё & Рейет 16082 (NY); Eastwood 8 Howell 11 59 (US); Flowers 3249 (MO). NEVADA: June 28, 1930, E. Jaeger (U 5). 37. SELAGINELLA LEUCOBRYOIDES Maxon, in Smiths. Misc. Coll 725:8. 1920. (Holotype: Munz & Harwood 3789 US!. Paratypes: Coville t$ Funston 628 US!; Munz, Johnston & Harwood 4226 US!). Fig. 53. Map 50. Stems short, forming flat or rounded cushion mats; branches short, approxi- mate, intricate or discrete, readily fragmenting when dry. Upper and under leaves 1955) TRYON-—SELAGINELLA RUPESTRIS AND ITS ALLIES 75 equal in length, ог the under slightly longer, оп the same portion of the stem; leaves linear to ligulate-ovate; base glabrous to pubescent, that of the upper leaves adnate to the stem for V, or less their length; margins eciliate to ascending-ciliate toward the apex; apex of the upper leaves herbaceous, slightly to broadly rounded, gently to abruptly beveled in profile; setae not forming conspicuous tufts at the dry branch-tips, 14 or less as long as the blade, scabrous to slightly scabrous, milk- white and opaque, rarely white and translucent. Sporophylls eciliate or with the cilia dentiform and ascending toward the apex, the broadest about 3 times as broad as the leaves. Megaspores slightly rugose to rugose-reticulate, to nearly smooth on the outer face, pale orange. The short, white, scabrous and usually opaque setae are characteristic of 5. leucobryoides. Те shares the peculiar character of fragile stems with S. utahensis and S. asprella. Rocky slopes, in crevices or among rocks, 900—2300 m. Southeastern California. Specimens examined: UNITED STATES. CALIFORNIA: Coville 5 Funston 628 (NY, US); Coville & Gilman III (US); J. T. Howell 3989 (Е, GH, MO, US); Bonanza Mine, Providence Mountains, 2800 ft., March 30, 1920, P. A. Munz 8 В. D. Harwood 3789 (Е, GH, NY, US); Munz, Johnston 8 Harwood 4226 (US). 38. SELAGINELLA ASPRELLA Maxon, in Smiths. Misc. Coll. 725:6. 1920. (Holo- type: Johnston 1815 US!. Paratypes: Johnston 1505 US!, 1807 US!). Fig. 54. Map 51. Stems short to moderately short, forming rounded or flat cushion mats; branches moderately short and remote to short and subapproximate, intricate, rather readily fragmenting when dry. Upper and under leaves essentially equal in length on the same portion of the stem; leaves linear-lanceolate to lanceolate-ovate to lanceolate-long-triangular; base pubescent, rarely glabrous, that of the upper leaves adnate to the stem for about 14 their length; margins eciliate to ascending-ciliate toward the apex; apex of the upper leaves herbaceous, broadly rounded to carinate, nearly plane to truncate in profile; setae forming conspicuous tufts at the dry branch-tips, 14 to over № as long as the blade, scabrous, white and translucent to rarely milk-white and opaque. Sporophylls eciliate or with the cilia piliform to dentiform and strongly ascending toward the apex, the broadest about 3 times as broad as the leaves. Megaspores prominently and coarsely rugose-reticulate, most prominently marked in the equatorial region, pale yellow to pale orange. S. asprella is characterized by its long, scabrous, white, usually translucent setae. Of the three species with readily fragmenting stems, it is the furthest re- moved from the presumed ancestor, S. Watsonii. : Open areas in rocky soil or in crevices at the base of boulders, igneous rock, 1800—2700 m. Southern California. Specimens examined: [Vor. 42 76 ANNALS OF THE MISSOURI BOTANICAL GARDEN UNITED STATES. CALIFORNIA: J. T. Howell 5031 (US); Jaeger 276 (US); I. M. Jobnston 1595 (MO, US), 1807 (MO, US), West end of Ontario Peak, San Antonio Mountains, 6000 ft., March 25, 1918, I. M. Jobnston 1815 (MO, US) ; Munz 7612 (NY), 9683 (US), 17165 (GH); Sept. 14, 1921, Saunders (US); April, 1906, Streeter (NY, US); R. M. 8 A. F. Tryon 5059 (ARIZ, B, BM, CU, DS, F, FI, GH, K, MICH, MIN, MO, NY, P, PH, POM, RM, UC, US, WS). Series Eremophilae Tryon, ser. nov. Rhizomata et stolones et gemmae nulla. Caules prostrati. Apices ramorum involuti statu inerte. Caules frondosi valde dorsiventrales. Folia supera base abrupte adnata caule distincta colore, folia infera base valde decurrente. Typus: Selaginella eremophila Maxon. Plants terrestrial; rhizomes, stolons and basal buds absent. Stems prostrate or assurgent at the branch tips, with rhizophores produced generally throughout; stems long to rather short, forming open or usually rather compact, rounded or usually flat mats. Branches moderately long to short, remote to subapproximate, usually intricate, branch tips involute in the dormant state. Leafy stems strongly dorsiventral, the under leaves appressed, the upper erect, the under longer, usually thinner and different in shape from the upper; zone of green leaves much longer on the upper side of the stem, very short to absent on the under side. Upper leaves with the base abruptly adnate, distinct from the stem in color, under leaves with the base very strongly decurrent. The species of Eremophilae form a homogeneous series and one in which there is increasing specialization in the dorsiventral habit. S. peruviana is strongly dorsi- ventral in position, length and shape of the upper and under leaves; in S. arizonica they are also different in texture. In S. eremophila the setae are highly specialized and the leaves of S. Parishii are muticous. Finally, in S. Landii, the under leaves are unusually long in relation to the upper and they are setate while the upper leaves are muticous. The species are all quite distinct with the exception of the first two, S. peruviana and S. arizonica, which are rather closely related. KEY TO SPECIES a. Allleaves setate (sometimes only in the bud) ; under leaves less than twice as long as the upper. b b. Setae mostly or entirely persistent, stout, straight. c. | с. Under leaves subulate, acuminate, broadest at or very near the base, rarely linear-lanceolate, not appreciably thinner than the herbaceous upper leaves, setae 0.5—1.0 mm. long. Oklahoma to New Mexico; Mexico; Реги то Argentina —39. S. peruviana, p. 77 · Under leaves lanceolate to lanceolate-ovate, broadest above the base, papyraceous, the upper fleshy, setae 0.1—0.3 mm. long. Texas to Arizona; Baja California .. — — — . ж arizonica, p. 78 b. Setae mostly early-deciduous (sometimes present only in the bud), deli- cate, filiform, tortuous 41. S. eremophila, р. 80 о 19554 TRYON—SELAGINELLA RUPESTRIS AND ITS ALLIES 77 a. Upper leaves, or all leaves muticous. d. d. All leaves muticous; under leaves about twice, or less, as long аз the upper; upper leaves lanceolate to lanceolate-triangular - 42. S. Parishii, р. 80 d. Upper leaves muticous; under leaves predominantly short-setate, the setae persistent; under leaves over twice as long as the upper; upper leaves long-deltoid or with parallel sides toward the base 43. S. Landii, p. 81 39. SELAGINELLA PERUVIANA (Milde) Hieron. in Hedwigia 39:307. 1900. Figs. 55, 56. Map 52. Lycopodium ciliatum Ruiz ex Spring, in Nouv. Mém. Acad. Roy. Belg. 24 (Monog. Fam. Lycopod.):55. 1850, in synon. (evidently based on Ruiz 98 B!). Selaginella rupestris (L.) Spring f. peruviana Milde, Fil. Europ. Atlant. 265. 1867. (Holo- type: Ruiz 98 B!; the lower right-hand specimen of Wilkes Exped. 1, Peru, GH! is Schaffner vel Sartorius B; Habn in 1868 B; Mexico City, Scbmitz а, 00 Chrismar B. Vancouver Island, J. H. Henry B should be excluded as a type; it 15 probably S. Wallacei). Selaginella peruviana var. Dombeyana Hieron. in Hedwigia 39:308. 1900. (Lectotype: Paratypes: Mandon 84 В!; Dombey 14 B! sterile; Bolivia, March 20, 1892, О. Kuntze B; Argentina, Hieronymus & Lorentz 162 B; Argentina, Jan. 1874, entz © Hieronymus В). Selaginella Sbeldonii Maxon, in Proc. Biol. Soc. Wash. 31:171. 1918, as Sheldoni. (Holo- type: Sbeldon 233 US!. Рагагурез: F. C. Greene on June 10, 1918 US!, on May 4, 1918 US!; Jermy 343 US!; Havard US!; Wooton US!). Upper leaves herbaceous, or fleshy only at the base, linear, acuminate, to linear- long-triangular; under leaves herbaceous, subulate, acuminate, broadest at or very near the base, to rarely linear-lanceolate, a little longer than the upper on the same portion of the stem; leaves setate in the bud, setae persistent (part may be abraded), stout, straight, those of the under leaves 0.5-1.0 mm. long, rarely some 0.3 mm. ong. Megaspores rather coarsely rugose-reticulate on the commissural face, less prominently marked on the outer face, yellow to bright orange. S. peruviana is the least specialized member in the series and for this reason it is considered as the basic type. Some four collections of the many examined (S/an- ford et al. 118 MO, Steyermark 52276 F, Атзепе 7087 US, Sharp 44257 US) show a tendency toward S. arizonica in having either relatively thin under leaves or short (0.2-0.3 mm.) setae. The primary characters of S. peruviana are the narrow under leaves, the long setae and the similar texture of the upper and under leaves. The localities taken from the literature (map 52) are from Alston, in Physis 13:23). 1939. Bluffs, rocky slopes or on ledges and in crevices of shade, igneous rocks or sandstone, less often in sandy | the United States, 1300-3200 m. in Mexico and 1900—4000 m. in cliffs, exposed or under light or clay soil, 600-3000 m. in South America. [Vor. 42 78 АММАТ$ OF THE MISSOURI BOTANICAL GARDEN Oklahoma and New Mexico to Puebla; Peru to Argentina. Representative specimens: UNITED STATES. OKLAHOMA: Goodman 2337 (GH, MO, NY, US); May 4, ae Е.С смене (05). June 10, 1918 (US); Quanah Mountain, Indian тан July 28, 1891, baec 233 (US). Texas: July, 1885, Havard (US); к 343 {М О, 0$ 2; ыы. 7 (GH, МО, US); С.Н. мв 8256 ры МО, МУ, US); В М. ЗА Е. Ттуоп 5032 (BM, CU, DS, GH, MIN, | PH, POM, US, WS); March 7 7% 1918, "Wooton (US). NEW MEXICO: Arséne 18050 E US), bs 3 (F, US); Arséne & Benedict 168 38 (F, US). Mexico. Aschenborn 65 (B). солнипл: Stanford et al. 118 (GH, MO, NY, US). US). MEXICO: Schaffner 933 (GH, YU). DISTRITO FEDERAL: Schaffner 19. с ВМ). Peru: Cook & Gilbert 158 (US), 240 (US), 305 (US), 608 (US); bey 14 (B, P); Soukup 1885 (GH, ees ae (MO) ; Stork & Horton 9388 (F); E xen Ruiz 08 (B, e Vargas 3141 (F), (Е). LIVIA: Vicinity of La РА 10,000 ft., 1889, Bang 111 (В, Е, СН, К, MO, P, US); тайды 4299 іп 1919 (Е, МО, US); Eyerdam 24708 (Е); Mandon 84 (B), 1531 (B). ARGENTINA: Burkart 10112 (MO); Hieronymus & Lorentz 162 (P); Lossen 426 (F, MO) 40. SELAGINELLA ARIZONICA Maxon, in Smiths. Misc. Coll. 725:5. 1920. (Holo- type: Shreve US!. Paratypes: Thornber 315 US!; Cook US!; Parish 8513 US!; Р. Е. Mohr US!; Toumey US!; Goodding 722 US!; Bailey in 1913 US!, on Nov. 9, 1913 US!). Figs. 57, 58. Map 53. Upper leaves fleshy, lanceolate to linear-lanceolate; under leaves papyraceous, lanceolate to lanceolate-ovate, broadest above the base, about as long as to slightly longer than the upper on the same portion of the stem; leaves setate in the bud, setae predominantly (or many) persistent, stout, straight, those of the under leaves 0.1—0.3 mm. long. Megaspores rather finely rugose-reticulate on the com- missural face, less prominently marked on the outer face, pale orange. S. arizonica is characterized by its short setae and thin and broad under leaves. It is rather closely related to S. peruviana and some specimens of that species approach $. arizonica in having relatively thin under leaves or short setae. In Texas the two sometimes grow in the same mat but these mixtures are readily separable. Some strobili of A. & R. A. Nelson 1158 MO have vegetative growth beyond the tip, a condition not uncommon in 5. arenicola ssp. Riddellii. Exposed rocky places, in crevices or on ledges, or in gravel, usually on igneous rock, rarely on limestone, 600-2000 m Southwestern Texas, Arizona and йаг Sonora, Baja California. Representative specimens: Unitep States. TEXAs: Cory 6878 Оп), x (GH, US); diede! 2147 (GH); June 6, DA Orcutt (GH, US); R. M. ВА. F. 5048 (AR ‚ MICH, МО, ARIZONA: Nov. 9, 1913, V. Mj TUS), 1915, V. Ds (US); 1913, O. F. Cook (US); Goodding 722 (GH, NY, US); April, 1873, Р. Е, Mobr (US); Nelson Trail, Santa Catalina Mountains, July 28, 1914, Shreve (СН, МО, NY, US); Thorn es (US); April 3, 1894, Toumey (US); В. M. 8 А. Е. Tryon 5052 (B, ВМ, Е, СН, К, O, NY,P, pS B Mexico ONORA 2701 (B, US); Wiggins 8330 (US). BAJA CALIFORNIA: Oct. 26, 1930, M. E. e А (BM, MO, US). 19551 TRYON-—SELAGINELLA RUPESTRIS AND ITS ALLIES 79 ET і УДА ~ 2. Та et Be rm 31 Ө | dl p p im і «ЦО» | Ғата т, T^ M : 1 mew ox m ре, < сле ; ay | uU. [Vor. 42 80 ANNALS OF THE MISSOURI BOTANICAL GARDEN 41. SELAGINELLA EREMOPHILA Maxon, in Smiths. Misc. Coll. 725:3. 1920. (Holo- type: Jobnston 1047 US!. Paratypes: Mearns 3162 US!; Betbel US!; Saunders in 1908 US!; Hall US!; Parisb бттт US!; Dudley US). Fig. 59. Map 54. Upper leaves fleshy, lanceolate-triangular to ligulate-triangular; under leaves papyraceous, to herbaceous at the apex, ligulate-lanceolate to lanceolate-ovate, about as long as to half again as long as the upper on the same portion of the stem; leaves setate in the bud, setae entirely deciduous to casually persistent, delicate, filiform, tortuous. Megaspores rugose-reticulate with irregularly projecting rugae on the commissural face, rugose-reticulate on the outer face, pale yellow. The tortuous, delicate and early deciduous setae set this species off from all others. The setae are attached only on the young leaves in the apical bud and are so inconspicuous that they had been overlooked until C. A. Weatherby?* noted them in 1943 Jaeger in 1934, Arizona, is unusual in that the setae are only slightly tortuous and less delicate than normal. Open rocky and sandy deserts, in sheltered places in rock crevices or in sand or gravel at the base of boulders, 130—1000 m. Southwestern Arizona, southern California, Baja California. Sopra specimens: STATES. ARIZONA: Dec. 17, 1934, Jaeger (US). CALIFORNIA: April 17, 1918, Bethel (US); Dec. 25, 1903, Dudley (05); March, 1919, Hall (US); Palm Canyon (Palm prings), Riverside Co., Jobnston 1 1047 (US); Mearns 3162 (NY, US Parish 1200 Bii b 45342 (MO); 1903, Saunders (NY), 1908, y, (US); R . М. & A. F. Try 7 (B, BM, F, GH, MO, NY, P, UC, ert Wolf 846 I (GH). не BAJA CALIFORNIA: Wiggins 0081 (US). 42. SELAGINELLA ParisHu Underw. in Bull. Torr. Bot. Club 33:202. 1906. (Holo- type: Palmer 306 NY. Parisb 1200 and Saunders in 1903, also cited by Underw., were excluded by Maxon, in Smiths. Misc. Coll. 725:4. 1920; they are S. eremophila. Figs. 60, 61. Map 55. Upper leaves fleshy, lanceolate to lanceolate-triangular; under leaves papy- raceous, lanceolate-ovate, up to about twice as long as the upper on the same portion of the stem; leaves muticous, the slightly modified apex acute to obtuse. Megaspores rugose-reticulate with irregularly projecting rugae on the commissural face, rugose-reticulate on the outer face, pale yellow. he muticous leaves are sufficient to set this species off from the others of the series. In a few specimens some strobili are dorsiventral. Crevices of sandstone or slate rocks, 1700-2300 m. Coahuila and adjacent Zacatecas. Specimens examined: Mexico. солнопл: Nil 105 Pd 3453) (US); June, 1909, Nil (US); Pennell 17272 (US). zacatecas: Chaffey (US); near Concepción del Oro, E. Palmer 306 in 1904 (F, GH, MO, US). 24 Amer. Fern Jour. 33:115. 1943. 19551 TRYON-—SELAGINELLA RUPESTRIS AND ITS ALLIES 81 43. SELAGINELLA ГАМОП Сгеепт. & Pfeiff. in Ann. Mo. Bot. Gard. 5:205. 1918. (Holotype: Barnes € Land 2024 MO!). Figs. 62, 63. Map 56. Upper leaves fleshy, long-deltoid, or with parallel sides toward the base; under leaves papyraceous, lanceolate-ovate to ovate, acuminate, more than twice to 2! times as long as the upper on the same portion of the stem; leaves muticous to short-setate, the upper muticous with the slightly modified apex obtuse, the under muticous or usually with a short, stout seta up to 0.3 mm. long. Megaspores slightly and finely rugose on the commissural face, slightly rugose, granular, to nearly smooth on the outer face, yellow-orange. This is the most specialized species of the series in that it is the most strongly dorsiventral. The short-setate under leaves and the muticous upper leaves are sufficient to characterize it. Dry rocks and on boulders, 1700—4000 m. Nayarit to Puebla. Specimens examined: · Mexico. пл 546 "а Jones 23495 (МО, US). уливсо: Barnes & Land 153 (Е), йе Estebán Moun . 32 km. from Guadalajara, 1908, Barnes t$ Land 2024 (GH, O, US); а Жа qoM (СН, US); Rose & Painter 7499 (US). МОКЕТО5: Matuda 26352 (MO, US). PUEBLA: Kenoyer 25 (US). DUBIOUS AND REJECTED NAMES Lycopodium strutbioloides Nutt. ex Baker, Handb. Fern Allies, 35. 1887, in synon., not Pr. 1825. The brief description is not sufficient to place this name although it suggests S. Watsonii. Selaginella rupestris (L.) Spring var. borealis Spring, in Nouv. Mém. Acad. Roy. Belg. (Monog. Fam. Lycopod.) 24:57. 1850. Indirectly based on a variety of specimens representing many species. Selaginella rupestris var. brevipila A. Br. in Ann. Sci. Nat. V, 3:270. 1865, in synon. Some of the specimens cited under the accepted name, S. rupestris, are S. Sellowii, others are S. Sartorii. Selaginella rupestris var. longipila Fourn. Mex. Pl. 1:146. 1872, nomen nudum. Selaginella rupestris var. longipila subvar. glaucina Fourn. Mex. Pl. 1:146. 1872, nomen nudum. Wright 1820 із S. Sellowii. The other specimens cited un- doubtedly represent other species. Selaginella rupestris var. longipila subvar. viridis Fourn. Mex. Pl. 1:146. 1872, nomen nudum. Bourgeau 2541 and Bottéro 78 are S. extensa; the other speci- mens cited probably represent other species. Selaginella rupestris var. tropica Spring, in Nouv. Mém. Acad. Roy. Belg. (Mohog. Fam. Lycopod.) 24:57. 1850. Indirectly based on a variety of specimens representing many species. 82 [Vor. 42 ANNALS OF THE MISSOURI BOTANICAL GARDEN Selaginella strutbioloides (Presl) Underw. Bull. Torr. Bot. Club 25:132. 1898. к ~ hà реа» [2] (Lycopodium strutbioloides Presl, Rel. Haenk. 1:82. 1825.) Maxon (Amer. Fern Jour. 11:35—36. 1921) has discussed the misapplication of Presl’s name by Underwood to Selaginella oregana. In addition to the сћаг- acters of size mentioned by Maxon, I may add other reasons supporting his position. First, Selaginella oregana is not known to grow on Vancouver Island; the nearest station for it is about 160 miles south of Nootka Sound, the locality cited for Lycopodium strutbioloides. Second, Presl would certainly have placed a specimen of Selaginella oregana, with its definite strobilus, under his heading "Spicis sessilibus" rather than under "Capsulis axillaribus" where he placed Lycopodium strutbioloides and other species without strobili. Finally, Greville and Hooker, in Bot. Misc. 3:105. 1833, say of Lycopodium strutbioloides Presl: "We have examined Haenke’s specimen, named by Presl himself, in the Lambertian Herbarium; and find it to be in no respect different from L. laxum of that author, which also exists in the same collection." Lycopodium laxum is placed by Herter, Index Lycopod. 55. 1949, as a synonym of Urostachys carinatus (Desv.) Hert. (Lycopodium carinatum Desv.), a species of south- eastern Asia. This evidently represents another example of a mixture of Haenke's labels. LIST OF SPECIES й T RUPINCOLA Underw 11. S. ciNERAsCENS А. А. Eaton . 5. Х NEOMEXICANA | Md 12. S. ARSENEI си : E Віскі оуп "Un derw. 13. S. MACRATHERA Weath. S. ARENICOLA Underw. 14. S. SHAKOTANENSIS (Franck. ex Takeda) За. S. ARENICOLA ssp. RippELLH (Van Miyabe & Eselt.) Tryon 15: S. Аа Ніегоп. 3a-b. Intermediate between ssp. 16. $. mutica D. C. Eaton ех Underw. RIDDELLII and ssp. ARENICOLA 16a. S. MUTICA var. LIMITANEA Weath. 3b. S. ARENICOLA ssp. ARENICOLA 16a-b. Intermediate between var. 3b-c. Intermediate between ssp. LIMITANEA and var. MUTICA ARENICOLA and ss 16Ъ. S. MUTICA var. MUTICA BORDER 17. S. EXTENSA Unde 3e 5. ен de E ACANTHONOTA 1% S Х/віснти Hiero S; Bazaxsas (A. Be) Hie 19. S. STEYERMARKII Alston . WEATHERBIANA 20. S. Hansenn Hieron. $. nmaa We 21. S. CARINATA Tryon Б я iki 22. S. пимса (Milde) Tryon S. SELLown Hieron. 23. S. NJAMNJAMENSIS Hieron. S. актови Hieron 24. S. CAFFRORUM (Milde) Hieron. S. Wicun Него 25. S. EcHINATA Baker 10a. S. WicHTH var. WIGHTI 26. S. МІУБА Alston 10b. S. WicHTH var. PHILLIPSIANA 27. S. PROXIMA Tryon Hieron. 28. S. ОвЕСЕ! (Presl) Hieron. 1955] TRYON-—SELAGINELLA 29. S. VARDEI Lév. 30. S. ОКЕСАМА D. C. Ea 31 NDERWOODI Hieron 32. S. RUPESTRIS (L.) Spring 33. S. DENS. 3a. S. DENSA var. SCOPULORUM (Maxon) Tryon 33a-b. Intermediate рет уаг. чена and v NSA 33a-c. dedi мени var. SCOPULORUM and v STANDLEYI 33a-b-c. — rr between var. ORUM, Var. DENSA “~ var. STANDLEYI RUPESTRIS AND ITS ALLIES 85 33b. S. DENSA var. DENSA 33b-c. Intermediate between var. DENSA and var, STANDLEYI 33c. S. DENSA var. STANDLEYI (Maxon) ~ мл a siBIRICA (Milde) Hieron. Watsonin Underw. UTAHENSIS Flowers LEUCOBRYOIDEs Maxon ASPRELLA Maxon PERUVIANA (Milde) Hieron. w w us N |] $t (^ t^ t^ t^ (л th (o л t^ t^ ANDII Greenm. & Pfeiff, INDEX TO EXSICCATAE Abrams 3726 (2); 3309 (11); 3402 (2); 4716, 5464 (20); 6296 (15); бит 7355» 7486 (2); 7723, 70000 (20). Abrams & McGregor 760 (35 Aasi & Wiggins 207, 341 (2). Aguilar 7429 (19). Alexander & Kellogg 2062 Allard 3570, 4570b, id! 5488, 6475 (32). Amable 1383 (39). Amórtegui Aór (9). Anderson, А. P. 1804, 1975 (7). I427, 5706 (34). (24). Antunez & Dekindt 3040 (24). Archer 6764, 7129 (35). Archer & Gehrt 118 (8). Arechavaleta 472 (8). Arsène 801 (8); 3453 (42); 9983 (12); 10205 (18); 10630 (8); 10640 (1); 10641 (1 (1); UIT (12); 2); 106 17777 (332-b, 33b); 17817 (5); 17929 (5); 17930 (31); 17931 (5); 17945, 17063, 17972, 17073, 17974, 17075 17976, de 17978 (31); 17080 (5); 17082, 18049 (31); 18050 (39); 18052 (31); лоб (39); 18478 (31); 18613 (39); 18654, 18666, AV 19377 (31); 20174 (39); 20985 ( ; 20987, 20988, 20080, 20090, e. 709); 2 (3За-Б); 21118 (16Ь); 21120 Eu Arséne & Benedict 15872, 15873, 15874, 15875 (31); soy (5); 15877, 15878, nh 6152 (33b); 16370 ); 16644 (33a); 16645, 16838 (39); 16884 (31); 18053 (31); ара, 18055 (33а); 18056 (33а, с Aubineau 25 (31). Austin 241 (20 Bachmann Bacigalupi Baker, C. F 2 9 (28). 1015 (16b); 2750 (15). 2 (31); 16, 2746 (15); 5234 Barclay 1568 (15). Barkley 1725 (33a, 33c) ; 16144M (18). Barkley & Diettert 02 (15). Barnes & Land 153 (43); 2334 (1); 2024 (43). Bartlett 808 (32); 10355 (9); 10529 (1); РА (18); 10876 (1); 11039 (9); 20450 (8). ВагПеу & Жалы 167 (32). Bartram 40). Beattie 5283 (15 tle 1566, 3066. í Benedict T OD; 3 (5); 4, 5, 9, 13, 2285 (31). [Vor. 42 84 ANNALS OF THE MISSOURI BOTANICAL GARDEN Bergman 3156, "ito Веупсћ 108 ( Biltmore Herb. БЕЯ зга (32); 3432Ь (3c). Blake, S 44, 746, 916 (2); 3031, 4762 (32); 10088 (0); е (35); 10351 (15); 11481 А ( Blomquist 50 (3c). Blomquist & Correll 4715, 4716 (7). Bogusch (15 Boivin & Gillett 8686 (33b). Bolander 4511 (20); 6247 (15). (28). 9 d р [=] M Brandegee, T. 5 ras 14426, 14473, 14475 (3c). Breitung 4024, 5553 (33b); бото (32). Brewer 1597 (2); 1726, 2103 (35). Баш 855 (335). Brow 95 (33c); 128 (33b). Buchholz 2 (39); 4 (16a). Buchtien 3123, pas (39). Bües 608 Burkart ia 3308, 7128 (8); 10112 (39); 10113 (8). Burkart & Troncoso 11174 (8). Bush 174, 835, 2819, 4744, 5069, 5210. 5397 (32). Butler 874 (15). Butters & Abbe 172, 1044 (32). Mund & Rosendahl B s 2955, 3352, 3595, 4278 (32). Butters, et d. 398 (32); 592 (33a). Cain 143 (33b). Calder 2810 (32). Calder & Billard 2995, 4454, 4601 (34) Calder & Savile 7647, 7677 (33b); 7732 (15); 7945 (33a-b); 8072, 8366, 8540 (15); 9221 (335); 10914, 11135, 11323, 11484, 11634 (33а 4 (15): 1029 — 1237 (32). Camp & fané 2 Campbell & Canfield 531 (16b). Cardenas 162 (39). 1). 7488, 7492 (31); 7403, 7494 (225); 7759 pé 7947 (18); 8375 ОН 8384 (16b). Chaffey 58 (42 Chaney 181 (32). Chapman 34325 (3c). Chase, А. 1584, 1062 (32). Chase, V. Н. 318 (32); 7676 (18). о ant 23). Р 8668, 0628 (31); 10034 (40); 10036 ak Clausen 3674, 3679 (7). Clausen & Торо 3663 0 2). Clements, Е. E. 458,1 270: 01); 3312 Clokey 4849; 5164, 5165 (2). Clokey & Anderson 0508, 6532 (2). Colby 3364 (32 Cole 5135 (32). Conard 174 (30); 246 (15, 30); 1-030 (32 Constance & Rollins 008 (15 es 158, 240, sea 248, 305, Core 3869 (32). Correll 6160 (3c); 6615 (32); буті (7): 10857 (32); 13627 (6); 13804 (165); 14004, 15000 (31); 15025, 15035 (1а); 15238 (3а). Correll, D. S. & Н. В. 7705 (32); 7907, 012 Correll & McFarlin 6234 (3c). Cory 888 (1); vin (3a); 6878 (40); 8444 (18); 9077 (39); 10294, 11650 ба); ный ера 18310 (18); 30143, 0071 (18, 39); 38235, 38438, ut бес 40232 (40); 41099 Ga); 42743 (18); 44831 (6); 48388 сова 3674, 3754, 4458 (35); 5644 Coville 121 (2); 186 (20); 850 (30). unston IOI (2); 628 (37); 2071 (35). СоуШе & Gilman 325. сие 43 (335); ўя 455 (15); 579 cx 200 (33a) Crandal 1054 (31). Cronquist ee Go 5641 (15). Cu үнө pd ne: (32); 6714 (b). Ou E. RU. 2580 (32); 3220 (3а). Dandy 510 (23). Daniels 151 (31 Darker 3579 (15); 5861 (35). Darrow 2847 (35); 3631 (16b). Darrow & Philips 2777, 2781 (31).. 19551 TRYON-—SELAGINELLA RUPESTRIS AND ITS ALLIES 85 Davis, К. J. 947 (15). rend 21584 (32); 58692, 63923, 64053 раса & Peilr 16114 (33b); 76982 36). Dekindt 543 (24). етагее 3251В ) Diehl 2517, 2565 (39) Dieterlen 117 (24). ixon 303 (168); 435 (35). нераў C H Dombey "Tr 5 Doty бел a: 1868 (30). Пгаке 3053 (22 Drouet & te 4959 (33a-b); 4064 16b). Drouet, F. & 4051 (2). Drouet Б” Richards 3311 (16b); 3337 (5); ров" 2% an 58, 3350, or 3635, 33 3 687 (1); 4097 (16 Drushel 11128 (32). Drummond 352 (3a) Dudley 1405 (35 Duncan 0» (3с); 8683 (3a-b); 10467 (7); 11190 (32). Duncan & Smith 9339 (7) Duncan, e Dusén 1 Duthie 3727 (22). Eames, A. ]. 9081 (32). Eastwood 639I (2); 12185 (30); 12301 (2); 13207 (15); 15495 (33b) ; 17775, 177754 Eastwood Howell 1150 (36); 3324 (33a); 8808 (2). (28). Eggleston 2199, 2905 (32); 10609 (15); 39): 17 Ehlers 1274 (32). Epling 7033, 7284, Qu (33а). Evers 32650, 32668 (32 Ewan 2751 (35); 4240 (2); 11203 (33b); 11665, 11730, 11827 (33a); 12047 (33а-Ь); 12659 (33b) ; 12600 a еј ; 13341, 13569 (33а); 13267 (33b) ; 14048 (33а); 14370 (5); 14050, 15302 (335); 154714 т 18377» 18305 (33а); 18480 (33c); 18586 (33 16 Exell & ан 160 (2 Eyerdam 24798 (39); 25041 (8). Farwell 060, 845b, 9273 (32). Fassett 2076, 2760 (32). Fasse тс & Wibon VERA Faurie 52, 1577 (14); ipe; 9687 (34); 95, 10543, 13151 (14 Fellows 6923 (32). Fendler 1024 a: I025 (33b). (10a). 3062, 3063, 3064, 24818 gri (32). 26 (1). Ferris, R. S. f (20); 2090, 8470 (2); ния (11 Ferris uncan 2406 (39); 3588 (31, 22 irt 3589 (39); из (40); 3593 (39 nald & Pease Fernald, et al. , Ferris & Duthie 94r (53a ‚ 33 2); T DA 234 (34); 1702, a P (15); 3092 (532). відніме. 32 z 9 d Fogg 15041 Fosberg Або оз: 697 (2). (3с Friesner 10208, 11235 (3 2). Frye, & E. M. 2537 (17); 2983, 3149 (12). Galathea rd 3188 (8). Gale я ке) Garre С 26 4000, 5699, 7164 (35). Gates 1522 Gattinger [Rs за ). Gentry 1104 (1, 39); 8004, 8255 (1). gd 1043 (28). ). Шеге, J. B. 4681, 4681 Giles, J. M. 6022 (33b); Pj (15). G ( ilman m Glaziou y^ Gleason 0401, 2 9613 Men Glover & Gilliland 480, 771 (10b). 86 Godfrey GH1657 (28); 3981 (32); 4677 28). 4 (2); 5 ia 58 (16b). Goodman 687 (33b); 1858 (35); 2337 Goodman & Hitchcock 1514 (35). rman 675, 4594 (15). 42 24). Graham 7502 (16b); 8405, 8521 (35); 10004 ant, С. bog 1197, 2456 (2); 6400 (35); Mo as 7144 (35 Grant, G. B & Wheeler 126, 6144 (2). Grassl 5830, gee (32). Gray., Pl. Ехясс. біг (32); 1305 (3c). Greenman 4612 e d (32). M. T. 4607, 4721 Griffith 2871 (22). Haberer 1355 (32). agenah 1160 (32). Hall, C. С. 155, 173, 177, 178 (35); 184 (2). Hall, E. 694 ( Hanna 472 Ө: 1210 (335); 1425 33 (33a Hansen D (20); 879 (35). Harding Harper 5 Ы 10, 16, 47 бен 27 (3a-b) ; 215, 34-3 3 ); 1800 (За-ђ); ( 1852 (3c); 1854, 1860 (3b); 1957, 1987 (3c); 3116 (3a-b); 3403 (3a, 32); 3762 (3a Harris, B. Е 708 (18). Harris, С С. & 5. E 3340 (15): 3735 30). Harrison 0160 (35). Harrison & ges 7634 (35); 7801 (33a) Hartman 5 1 327 Harvey 288, 1173, 1265 (18); 1454 (39); ; 1722 (1). 14542 (1); 1507 (1 Hassler 1310, 125351 (8). Нас tes 283, 325 (19). Haupt 105 (1 Hayden, A. 0129 (32). ГУ ог. 42 ANNALS OF THE MISSOURI BOTANICAL GARDEN Haynes vA de 2640, 2711 (41); 2785, 3016 (11 Hayward 315, 90 907 (33b); 1684 (32); (33b). ). Heller 214 (32); 5024, 5201 (15); (20); 7167 (35); 7276 (2); 7583 (20); 7610 (2); 8173 (20); 8604 (2); 8995, 9145, 9380 (35); 10715 (20); 11054 (35); 11802 (20); 14101 At c). Heller, A. A. & E. G. 4002 (30). Heller & Halbach Шы (32). Henderson 505, 4062, 5071 (15). Hendrix 105 (2). H K. 0080, 0084, 0085, 0086 Henshaw 197 (2). Hermann 1851 (32); 4665 (33a-b-c); 4062, 4095 (35); a (33a-b); 5594 (33а); 7534, 7621, 8745, 8774 (32); 1000. ; 10070 (3b); 10108 (7); Неггега 30092 (8); 3084 (39). Herter det ZARO, pere 99600 (8) Hildebrandt 2363 (28 Hill, R ) Hinckley - IOII (31); 1155 (16a); 1156 (31); 1157 (39); M 2147 (40); 3256 (39); 3404 (1); (40). Hinton 3453 (9); 6078 [n 8423 (9). Hitchcock, А. 5. 22546 Hitchcock, C. L. & мева 4688 (33а); 4705 (15); 7367 (15, 33а). Hitchcock, С. L Mullick 8125 (33а); 5499, 9048, es (15); 9252, 9613 a Hitchcock, C. L. & Stanford 6780 (40); 7245 (12 Hodgdon 2). Hoogstraal & Chase 7235 (18); 7308 (17). Hoover 1399 (35); gu (20); 4900 (2); 5085 On 5308 ( House з 4197, 1274, 4307 (32); 4082 99 15); 10279, 18128, 21482, я 27978. 270% (32). owe, М. А. 000 (30 Howe, T. D 66 (35). (11); 3080 (37) i 4557 (15); 4692 (20); 5031 (38); 7600 (15); 13521 (33а); 16761, 26066 (35); 1955] TRYON——SELAGINELLA RUPESTRIS AND ITS ALLIES 87 Howell, T. 606 29: 520 4791 (25); 5518 (26); 5581, 5705 (27). Hunnewell 8365 (15); 8366 (30); 9478, 4 ; 10863 32); 15709 (33b) ; 15711 (332); EA 18335 (32). Ilis 1516 (32); 3020 (15); 3088 (33b); wor (332) ; 43994, 44782 (33a); 46400 m " Clebsch 22. DE Ingram 2723 (16а). Innes & Warnock 479 (1); 606 (18). Jackson, J. К. 1045 (23); 1116 (24). Jaeger 276 (38); 548, 551 (2); 563, 571 35 Јатезоп 426 (8). Jennings, O. E. & С.К. 6661, 7362, 7389 4 8173, 9411, 9490 (15); 14518 Jermy 342 (3а); E (39 Johnston, L М. 1044 (2); 1047 (41); 1595, 1807 b 1808 (2); y (38); 27 1825, 1826, 1827, 1828 (2); 2420 3006, 3 3908 (33b); 7208 (18); a Шу, бр (18); 8683 (6); 8854, Johnson & Muller 307 (1); 651, 1192 18). Jon i. G. N. 5539 (33c); 5610, 5772, 6287 (15); 18062 (32). Jones, G. N. & F. a» Jones M. E. 485 708 (5): 1246 (35); 3023 (2); ps (16b); sisi 5037, an 23495 (43); 26032 (1); 27018A (1). а (33b). Jones, М. E. & Eggleston 10506 (2). [нке эы, 4073 (8). Јаћо тоо (39). Jürgens 29, 308, 309 (8). Kaessner 860 ( Kearney i 0) : 984 (32). k 4458 (35). wem Kemp 30 (33b); 52 (15). Kenoyer 25 ADT 1053 (18); 1566 (9); 172 4 ( Кепоуег & Crum yi (18); 3910 (17). Kew Exsicc. 1239 e 3435 (335); Mis 6479 (5); 8247, 0288 (33b); 10280 (31); 11323, 23605 (3 2). Killip 1112, 2266, 31001, 36944 (32). Kirkwood 1415 (15). Kirkwood & peer 1718 (15). Knight 5206 (32 Knobloch 36 (1); 5044, 5053 (39); 5557 (1a); 5573, 5928, 5036 (1); 5038, 5969 Кпор# 335 (2). M Е.Н. 62, 103 (33а); 141 (31). .& А. 53 (34); 162 (33c). ко 204 (15). Каса 512 ( Тагзеп 5008 њу 5 (35). 7 (30). Leeds 76 (32); 149 (1); 375, 378, 379 380, 381, 386, 388, 380, 392, 307, 398, 399 (3c); 837 (32); 3208 (3c). Leendertz 935, 2508 s ККо7, BT776 (8). Leiberg 3222 (2); 5535 (31). ; 1540 (18). Little 4106 (40). Lloyd, F. E. (incorrect as C. E.) 20 (18); 504 (1, 39); 505 (1). Lloyd & Kirkwood 20, 142 (18). Lodewyks ee (32). Lorentz 442 Lossen 246 (з); 426 (39). Ludlow, et AVE jd 5494, 14233 (29); 17039, T (22 Lundell, A. A. 56 (1, 39); 96 (18). (Мог. 42 88 ANNALS OF THE MISSOURI BOTANICAL GARDEN Lundell, C. A. & A. А. 9037 (3a). Lützelberg 18870 (8). Lyon 32) Lyonnet 2562 m Macbride yr (5); 7878 (2). MacDaniels 8 (39). MacDougal их (19): MacLeay 114 (24 Mackenzie, H. 15). Mackenzie, K. к” 5 (33а); 2007, 3601 2 МасМ ап & Sheldon es Maguire 10021 (1 0055 iem 18450 (16b) ; 19321, 19378, 219873, 21000 (35). г ов 22607, ‚№ e (1). Maguire & Piranian 11371 (16b). Maguire & Richards 13166 (35). Maguire, et al. 3962, 3063 (35); 4359 om нев 14454 пил, 14691, Maire, 4 3 (4). Malte & Watson 1229 (33b). Mandon 84, 1531 (39). Marcelline 2905 (31). Мапе 486. (40). Marr 722 (15). мии, HI 33424 (20); 11278 (35). Мазеу 3220 Matuda 1846 um. 26116 (1); 26350 (9); 26352 (43); 20353 Ж У. Matuda & Gold 23572 Maxon 145, 6293, Hilos (32); 10833, I0841, 10843 (3c). Maxon & Hay 3 (19). McAllister 2 (18); 24 (39); 5 (33b). McCabe 5655, vi 7108 (15). McDougal 481 мт 0374 бы, 10269 (3b); 10271 Mela tosh 13. 1999 (39); 2705, 2706 (2); 4695 (33а); 5102, 5109 (2). McMurphy 453, 531 McVaugh 4577, de ві M 7675 (18); 7898 (162); n 10446, 11829 (17); 14660 ( Mearns 225 (1); 237 (16b); 540, 685 (1); 2044 (33а); 2687 (1); 2701 (40); 3162 (41); 3589 (332); 3881, 4060 (2); 4274 (332); 4401, 4736 (33b). Meislahn 138 (3b). Merrill & Wilcox 1218 (33а). Metcalf, F. P. 1400 (32). қаз О.В. 276, 711, 001 (31); 1172 MAE E (3a). Meyer, F. С. 188, 272 1062 (15); 1243 (33а); 522 (30 ыы 3112 (9). 7 (8). Milburge 3 Millspaugh 4502 (2). Mohr 12 Moldenke 882 (эс), ооге, А. 100, 537, 1561 (32). Мооге, Н.Е. 5374 (1). Моог 3699 (17); 3926 9). 13 00 teyermark 3046 (162); 3047 (39); 3106 (16а); 3245, 3745 (33а). e, J. У 7 (32); 122 (33b); 767 10547 (32). Moritz 088 (28). = 2151 (33b); 2752b (33а). Moxley 345, 346, 347, 421, 876 (2). 99 (32). #83 0 2). Mueller, Мб, DNO XE Muenscher 061 E 9673 (15 Muenscher, W. M. W. авт (32); 14526 (39); m Y: 1 unro IO2IA, 1777, 188 3 (28 Munz 6726 (2); 7930 d»; s (38); 9079 (20); 9683, 17165 (38). Munz & Eggleston а (2). Munz & Harwood 9 (37). Munz & Johnston 252 (35). Munz & Keck Pew (41). Munz, et al. 6 (37). Мань = dp 4787 (33a). Murie 1232 (30). Napier 2054 (10b). Nash 1449 (3c). 19551 TRYON—SELAGINELLA RUPESTRIS AND ITS ALLIES 89 Nelson, A. 1345 (33a, 33b); 2448 (33b); 6368 (33a); 7811 (33a-b, 33b); 8781 35]. 10676 (16b); 11217, 11217а Nelson, А. & E. i (33a). elson, A. 1158 (40); 1448 (1); ' 5765, $766 ei 5782b (33a). Nelson, Ј. С. 3876 (15). 405 (32). Novogablenov 541 (34). Nuttall, L. W. 553 (2). O'Neill 7580 (3c). Oosting 53 (32); 3640, 34422 (7). Orcutt 1080, 1097 (18); 1098 (9); тоова (18); 1102 (39); 1142 (9); 1151 (18); 1151а (9); OJ (39); 1552, 1586, 1944 (1); 1989, 2218 (18); 2974 (1); 3084, 4268 (39); 5125 (9). Ortega 5310 (9). tis 1073, 1074, 1130, 1148, 1149, 1150, Hen 1159, 1117 i „1157 1158 (15); 1242 27081 Оуег 9 (30 11008 (32 и (33b); 13678, 13679, 13680, nr. 15883 (32). Palmer, E. 38 in „1908 (1, 39); 92 in 1906 с 554 in 1896 (1, 39); 660 in 1874 О с Palmer, 7 (3 4). & Sing 186, 214 (32). ‚$. B. 1200 (41); 4728 (2); 5052 (35); 6110 ҮТ біп (41); 8513 (40); 9134 (11); 10764 (2, 11); 10765 (11); 19180 (15); 109190 (41). Parish, S. B. & У. В. буг (2). Parks, H. B. 37921, 73920 (18). Parks, H B zt 30). Parks, H S. T. 24110, 24119 (30). Parks, H. 5 ие (15). Parodi 14328 (8). Parsons 16 (15). Payson 2503 (39%; 2814, 2 (79, (33Ь). Раузоп, Е. В 4037 (33a, 33a-b); 4453 (33a); ain 5074 (3 5). Pease 115, 12050, 17967 (32); 22208 (33а); ping te 22419 (33a); 22499 (33с); 24195 (32 Pease & Bean 23334, 23581, 26103, 26118 32). Реазе & Ogden 25070, 25114, 25140 (32). Peattie 809, 7). Peck 15996a (15). Harrison 6084 Bes (40); У (1). 3 іо (18); 17198 (9); 17272 (42); 17476 (39); 17753 (9); 17947 (17); 18126 (39); 18178, 18646, 19533 ресів: 1168, 8262 (25); 8303, 8305 (26); n 8312, 8313, 8324, 6). eterson 14 (39). Phelps 65, (32). Philips, E. A. 775 75 (1). Philips, W. S. 2280 (1); 2588, 2604 (40); 2848 (31); 2861 Me 2866 (51); 2930 (40); 2050, 2981 (31 Т. К. Ate (33b). 5 (1). Philipe W. S. & Reynolds 2900 (33a, -b). бы, 231 (15). Piper 3524 (15); 3802 (30); 4964 (15); (332). 6059 Pob. y Polunin, О. 1478 (22). nE .E. & R. T. 150, 322, 690, 1305 34). Porter 1121 (33a). Porter & Rollins 5819 (33а). Pringle 224 (40); 3900 (17); 10823 (43). Pringle & Conzatti 1300 (18). Pultz & Phillips 2711, 2720 (1); 2749 (31). Purpus, a x 3156 (1, 18); 6052 (17); $463 Purpus, i + 3114 ; 3119 313 3155 (3с); 3175 (3b); 31752 (3b, 3b-c). Quer Г (4). 90 Каир 6305, 6411, 6607, бого, 8184 (32). Rau ap & Abbe 3010 (33c); 4451, 460 кын 284 (32). Кеед, Е. 6180 (2). eed, Е. М. 2244 (2); 2466 (35); 4070 (2). Reeve 100 (32). Rehmann 3924 (24); 3949 (24,28); 4005 (24); 4333, 5576 (28). Reko 5240 (1). Rendle зот (28). кке 15 (18); 1456 (39); 1632 Richs i & Drouet 363 (51). һы 16, 1707 (3а). 79 (1). Rose, З М. & Fitch 17073 (18); 18171 Rose, т М. & Раниег 0536 (39); 6694 (1); оз (8); 44 т n Rose, J. N. & J. S. Rose, J. N. & Mrs. AST D 19468 39 (32). Rose, J. М. & Russell 19937 Rose, J. N. et al. 8268 (39); co 12857 (1); 23039 (21). Rose, L. S. 42%» 40016 (2); 45039 (20); 453. 41). Rose-Innes 4102 6 (18). Rose-Innes & Warnock 783 (3a). Rosendahl 448 (32); 864 (15); 1074 (33a); 2662 (32 Rosendahl & Butters 4564 (32). et e al 1615 О 9 (32). Rouss 2 Rude 857 (24). Ruiz 08 (3 Ruth 204, 706, 749 (32). паи 1184 (33b). Rydberg & Везеу 3517, 3518, e. PU. Rydberg & Carlton 6310, 6566, 718 7555 (35). Rydberg & Garrett 8666, 8007 (33b). Rydberg & Vreeland 65 71 А 6580 (33b); 6500 (16b, зі); 6501 (31). [Vor. 42 ANNALS ОЕ THE MISSOURI BOTANICAL GARDEN St. John, H. 8668 (30). Sandberg, et al. 54; 54а, 54b (15). Scamman 878, 1687, чой, 2101 (34); 3а). $ 10 (39); Б (1): 093 (39). Schallert anr (30). Sch (32). Sharp 44126 (39); 44168, 45375, 45864, 264 (17). Sharp, et al. 50255 (17). Sharsmith 2076, 3127 ME 3973 (33a). 1060 (33 Shaw p we Shear 33a); 4616 (тв). Sheldon H (32); 8083 (15). Shoop 553 ( Shreve oes 5334, 6743, 0041 (1). im 6095 (23). I: 2,1 (ia). mall & Carter 1013 (3c). З all & DeWinkeler 9733, 9768 (3c). Small & Heller 214 (7). S TER 47. Soukup 1885, 3571 (39 Soxman 101 (3a); 351 > 449 (3а); 39), Т ( $репсег 261, МЕ 14686 (41); 1646 (11); 1697, 1007, 2124 (2). Sperry 427 (39); 428 en); ; T733 (39); (18 Steinbach 8254, 8574 (8). eid 2204 (6). Stewart & үн аи 2122 (1). tides 736, 779, 15644, 15809, 20198, 27409, 27855, к бін 47115, 50501, 51202 (19); Stanford, et al. 15318 (Hb ); 15318а (33a); 15320, 15345 (15); 15363 (33c); 15396, 15732, 1955) TRYON-——SELAGINELLA RUPESTRIS AND ITS ALLIES 91 15508, 16216, 16255, 16288, 16378 (33а); 16583 (15); 16970 (33с); 17055 (33a); 17164 (15); 17164a (33a); 17228 (33c); 17321 (33b); 17463 (33c); I7714 (15); 17979 18050 (33a); 18136 (33c); 18185, І (33а); 40380 (18); 52533 (3c); Parte (19); 73379, 92638 Enu & Bolman 10830 (31); 11965 (32 Stork & Horton 0388, 10710 (39). Suksdorf 4 (15); 47, 225 (33b); 296 (15); 753 (20); 1364, 2085, 2644, 2645, б (15); 6735, 6736, 6811 (30); 7412, a (12); 8834 (33а). Summers Sven MI. > У а (32). Swynnerton Міла (28). Татауо 208 521 1753, 1773 (9). Tarleton 188 (34). Taylor, В. С. = (32). Taylor, L. Е. 2144 (18). Taylor, М. 419 (18). Taylor, T. M. С. 1090, 1135, 1290 (15); 2014 (33a-b); 2015, 2023 сөп 208$ (15); 9704 (33а); 46024 (15). Taylor, T. M. C. & Pillsburt 46118 (15). Taylor, T. M. C., et al. 71, 144 (32). Tharp 1702 (9); 43-513, 43-514 (18); 44027 (3a). Thompson, J. W. 2010, 2615 (15); 3113 (30); 4002 20, юл; Ре, s Mec 6797, 7458, 8128, 8747 (15); 99 (30); 10124 (15); 10583 (33a); epu 12752 OM 13565 (33a). Thornber 315 40). Thorne = ов); 9394 (3c). Th 3a). . M. & Earle 275 (16a-b). Train 371 (35). 21 (2). Trelease 420 (1); 2641 (15); 2642 (34). R. M. Fors 2685, 3439, 4761, p (32) » 5009, O, 5011, 5012, Ин 5014, 5016, der? ass 5018, 5019, 5021, 5024, 5026, 5027, 5093 (3c). Tryon, К 5005 (32); 5015 (6); 5030 (3a); 5031, 5032 (39); 5033 (3a); 5034, 5035, 5037 (18); 5039 (39); 5043 (40); 5047 (39); 5048 (40); 5050, 5051 (1); 5052 (40); 5053, 5054, Hee (2); 5057 (41); 5050 (38); Z eas ; 5063, 5067, por (20); Tweedy 36I (33a). nop 26 (10Ь). Ule 2t Ulke py G3c eJ. Umbach $5, 117, 856 (33a); 5106 (32). Underwood 1355 (3b, 3b-c); 13552 (3b). Underwood & Selby 1 (33b); 2 (5); 133 (33a-b, 33a-c). Underwood & Seymour 2805 (32). Van Eseltine 240, 501, 502 (32). Van Schaack get у, Sec): 2712, 2786 (33a); 2917 (16b). Van Schaack & fera 2934 (33b). І (39). Vargas 79 (16b) Walker, E. H. 329, 2069 (32). Walker, E. P. 365 (16b). Walpole 2147 (30). Wallich 2188-3 (10a). Vishe Vorhies War e, V. A. 520 (20). Warnock T25 9% 479 Es (1); 924 = Watson, P. Watson, S. a ай 1370 (35). [Vor. 42 92 ANNALS OF THE MISSOURI BOTANICAL GARDEN Webster 13 (15). Williams, R. S. 534 (33b); r400 (8); Weintraub 724 (32) 2649 (39). Welsh 18080 (24). Williamson 1017, 2317 (32). Welwitsch 48, 40 (28); 169 (24). Wilms 1814 (28). Wetherill 272 (165). Wilson, E. H. 5411 (29). Wheeler, L. С. 2081, 3036, 3122 (33a); Wilson, Р. 37 (32). je 9 (15). Winter (32). Wheeler, W. А. 672, 1058 (32) Wolf, C. B. 3976, 5335 (35); 8461 (41). Wherry & Adams 27 Wolf, J. 997 (33a). White 3962 (1) Wol£, S 1324, 3115 (За) Whitford 47 (33b) Wolley-Dod 193 (2). Wiegand & nini о (3c). Wood, C. E. 3547 Q2). Wiggins 4212 (2, 11); 4213 (11); 6495 Wood, C. E. & Clement 7204, 7503 (3c); 14 6764 (35); 7205, 7407 (1); 8174, 7571 (3b). 8330, н (40); 9937 (2); 9081 (41); 11189 және = "Rollins 450 (1); 484, 561 Wight. 14, 283, 3194 (10a). Wild 1218, 131 8 Wilkens 1801 (16а); 20234 (39); 2422 (1 Williams, Г. 1305 (33а); 14299 (19). Williams, Г. & В. 2126 (33b). оод, ). М. о (28). Ж E det 1834 (33a). Mis ooton 1 Wright, e Gar (18); 1820 (8); 2106 = 2116; 2115 (16b); 2116 (1). Young 2r (39). Zeller, S. M. & E. B. 748 (15). INDEX TO NAMES Accepted names are in Roman es deis and rejected names are in italic; names of subgeneric categories are in CAPIT. ENICOLAE оо эу ЈЕ CYLINDROSTACHYAE . зо 22 16 CYLINDROSTACHYS __ ЗЕ 16 ЕКЕМОРНШАЕ _ ыл Е У 76 EUSELAGINELLA | о Р: НОМОЕОРЕТЕБАЕ 7 2% НОМОВОРНУЈЈИМ · 2” HOMOTRDEAER _ uas 6 ја ae bryoi ше 220-0244 -67 cannatum = нанні 222-242 сиинин оазе ий ВМ 2233000 3 22 24 lrum и 82 тирене 64 sirutbioloides ______ 81, 82 T jocis E 58 SART о 17, 31 дан є ELLA (Secti 16 SE ELAGINELLA (Subgenus) — _ 16 Selaginella пьете 26 amazonica 34 Arecbavaletae 34 arenaria 26 arenicola 23 ssp. acenthonota 26 ssp. arenicola 26 ssp. Riddelli ___ 24 arizonica .. 78 rsenei 41 Aschenbornii 77 asprella 75 alansae 27 Bigelovii 22 Bolanderi 49 Bong с... 64 bryoides ______ 40 caffrorum . 53 capensis __ 53 carinat 50 Chrismarii 20 var. Karwinskyana 20 var. Neean 20 cinerascens 40 columbiana . 67 i | 28 deflexa 16 19551 TRYON-—SELAGINELLA RUPESTRIS AND ITS ALLIES 93 densa 66 proxima 56 var. densa 68 pumila 17 var. scopuloru 67 РУШИ а. 17 var. Standleyi 71 Quartiniana 53 denticulata 11 Riddellii 24 regei 57 Rosendablii 43 var танні. с а rupestris 64 var. Hidebrandéiens інн З 57 f. amazonica 34 var. Petersiana 57 f. amurensis 71 var. pretoriensis 57 f. Bolanderi 49 var. Rebmanniana 57 f. rgeaui 64 var. Welwitschiana __________ 57 f. brasiliensis 34 echinata 54 f 53 Engelmannii 68 f. Dregei 57 eremophila 80 f. i ii 70 extensa 47 f. Hansen 49 ndl 62 f. Haydenii 70 floridana 26 f. indica 52 uniformis 26 f. longipila 68 gracillima 17 f. manchuriensis 71 isea 57 f. montan 43 Hansenii 49 f. peruviana 77 Haydenii ____ 4: f. Schmid tii 71 helvetica 11 f GNE. о. od 71 intonii E f. Wallacei 43 bumifusa еВ f. Wrightii ____- - 48 ica - 54 KS. Balane SL o 27 dii 81 var. acanthonota -....-------- 26 lepidophylla 8 var. Balensae -- 27 eucobryoides 74 vu BENE со Sur 49 longipila 52, 68 var. borealis 81 macrathera 41 ға дани 34 Mildei _ 34 vet а 9 montan 43 var. — и ПРЕ 43 montevideensis 34 var. и чс се A | тиіс. 44 var. Peli а у CUM _ 62 var. limitanea 46 var. Hansenii 49 var. mutica 46 var. incurva 53 ar. texana 46 f. abyssinica - 53 mutica X rupincola 21 . angolensis 53 neomexicana 21 capensis 53 nivea - 55 var. longipila 81 njamnjamensis 52 subvar. glaucina 81 oregana 61 subvar. viridis 81 ishii 80 mexicana -. 47 pauciciliata 34 var. mexicana 77 peruviana 77 var. montanensis 43 var beyana 77 var. peruviana 77 Phillipsiana 40 var. e d e 57 ecta 37 f. 57 Preissiana 17 f. Welwitschiana —— 37 [Vor. 42 94 ANNALS OF THE MISSOURI BOTANICAL GARDEN var. rupincola 20 tortipila 29 _ var. Sartorii 37 uliginosa 17 var. oo 43 Underwoodii 62 var. Sherw 29 var. dolichotricha 62 var. вза 29 utahensis 74 var. fropica 82 Vardei 61 var. Wallacei 43 viridissima 28 rupincola 20 W allacei 43 Sartorii 36 f. columbiana 43 var. oregonensis 36 Watsonii 72 var. venezuelensis 36 nie 46 Schmidtii 71 Weatherbiana 28 var. Krauseorum Z1 Wightii 39 Т6 lor: 67 var. Phillipsiana _ 40 em те var. vetusta 39 28 Wightii 39 ума је 43 e ла а д Sbeldonii 77 Wrightü н Sherwood 29 Stacbygynandrum rupestre ______------- 64 ыг 71 TETRAGONOSTACHYAE ._____-- aah spinulosa 11 TETRAGONOSTACHYS 4447 5 ата еуі 71 TETRASTICHAE 17 Steyermarkii 49 TETRASTICHUS 17 struthioloides 82 Urostachys carinatus — 82 EXPLANATION OF PLATE PLATE 1 Series ARENICOLAE Fig. 1. S. rupincola, buds at ground level, Tryon & Tryon 5051 (MO), X 10. Fig. 2. S. rupincola, portion of aerial stem, Wooton 124 (NY), X 10. Fig. 3. 5. Bigelovii, branching tvm and basal portion of aerial stems, Tryon 9 Tryon 5053 (MO), Fig. 4. S. с се ssp. d stems with basal buds and rhizophores, Tryon 5016 (М Fig. 5. 5 и ssp. acanthonota, basal portion of stems, Tryon 5009 (МО), uio. 6. S. Balansae, portion of aerial stem, Balansa (P), Х 10. Fig. 7. S. Weatberbiana, portion of rhizome, Tryon t$ Tryon 5077 (MO), X 10. 8. S. Weatberbiana, portion of aerial stem, Standley 4558 (MO), X 10. Fig. 9. S. viridissima, portion of aerial stem, Stewart 2204 (GH), X 10. Fig. 10. S. tortipila, portion of aerial stem, Rugel (US), X 10. РТАТЕ 1 ANN. Мо. Вот. Garp., Vor. 42, 1955 e Dm а. — T ая c'e лы a Ss ya 5 | EET АНИНА TRYON — SELAGINELLA RUPESTRIS AND ALLIES из e» — = м < a e < Uv ~ ба "is 2 RJ А, M ба + “ч ~ at -. 2 | Uv | = О мч ра E Савр., Vor. 42, 1955 Bor. ANN. Mo. [Vor. 42, 1955] TRYON—SELAGINELLA RUPESTRIS AND ITS ALLIES 95 EXPLANATION OF PLATE PLATE 2 Series SARTORI Fig. 11. $. Sellowii, portion of stem, upper side, Venturi 1866 (US), X 10. Fig. 12. S. Sartorii, portion of stem, upper side, Liebmann (GH), X 10. Fig. 13. S. Wigbtii var. Wigbtii, portion of stem, upper side, Ferguson (US), X 10. Fig. 14. S. cinerascens, portion of stem, upper side, Kimball (GH), X 10. Fig. 15. S. Arsenei, portion of stem, upper side, Атзепе 10641 (US), X 10. Fig. 16. S. macratbera, portion of stem, upper side, Johnston 8067 (СН), X 10. Fig. 17. 8. sbakotanensis, strobilus, Faurie 0805 (MO), X 10. Fig. 18. S. Wallacei, portion of stem, upper side, Wallace (NY ex B), X 10. Fig. 19. S. mutica var. mutica, portion of stem, upper side, Meeban (YU), X 10. Fig. 20. S. extensa, portion of stem, upper side, Pringle 3900 (GH), X 10. Fig. 21. S. Wrightii, portion of stem, under side, Tryon & Tryon 5034 (MO), X 10. Fig. 22. S. Steyermarkii, upper leaves, Hatch t$ Wilson 325 (US), X 10. Fig. 23. S. Steyermarkii, under leaves, the dark areas are colored red, Hatch 6$ Wilson 325 (US), Х 16. [Vor. 42, 1955] ANNALS OF THE MISSOURI BOTANICAL GARDEN EXPLANATION OF PLATE PLATE 3 Series SARTORI Hansenii, portion of stem, upper side, Tryon & Tryon 5067 (MO), X 10. carinata, stem with branches, under side, Rose et al. 2 3030 (US) X 1.25. carinata, portion of stem, upper side, Rose et al. 23039 (US), X 1.25. indica, portion of stem, under side, Hooker t$ Thomson (2). x № njamnjamensis, portion of stem, under side, Schweinfurth (P), X 10. caffrorum, portion of stem, under side, T. Cooker 662 (NY), X 10. echinata, upper leaves from stem apex, Perrier 8313 (BM), X 14. ecbinata, under leaves from stem apex, Perrier 8313 (BM), X 14. . nivea, portion of stem, upper side, Humbert & Swingle 5518 (US), X 10. aie ч ж аан а . broxima, portion of stem, upper side, Humbert & Swingle 5705 (US), 10. S. proxima, 2. from upper side of strobilus, Humbert б Swingle 5705 (US), X S. proxima, 2. from under side of strobilus, Humbert t$ Swingle 5705 (US), X 10. S. Dregei, portion of stem, lateral view, Rodin 4403 (MO), X 10. Pay A қ ғ “Хақ un - id < > =. < Uv ~ с Е ©) 53 А. MJ fc ч м - - = — 3 м fx) ny | Z о 5 | н ANN. Мо. Bor. GARD., Vor. 42, 1955 un [éa] - ча ~ < а e < т ~ ба ~ ©) цу А, кә ба < + һә ы = ~ У м E v) | 4 о ж са ~ ANN. Mo. Вот. GARD., Vor. 42, 1955 [Vor. 42, 1955] TRYON-—SELAGINELLA RUPESTRIS AND ITS ALLIES 97 ‚ 46. EXPLANATION OF PLATE PLATE 4 Series RUPESTRES S. Vardei, portion of stem, upper side, E. H. Wilson 5411 (P), X 10. $. oregana, portion of stem, Kautz (YU), X 10. S. Underwoodii, portion of stem, under side, Fendler 1024 (MO), X 10. S. Underwoodii, portion of mat, Arséne 17963 (F), X 1. S. rupestris, portion of stem, upper side, Leeds 76 (NY), X 10. S. rupestris, basal portion of strobilus and leaves, J. K. Small on Aug. 18-27, 1894 (NY), X 10. S. densa var. scopulorum, portion of strobilus, Standley 15732 (US), X 10. S. densa var. densa, portion of stem, upper side, Boivin &$ Gillett 8686 (MO), 10. S. densa var. densa, portion of stem, under side, Boivin 8 Gillett 8686 (MO), X 10. S. densa var. Standleyi, portion of strobilus, Standley 17228 (US), X 10. [Vor. 42, 1955] ANNALS OF THE MISSOURI BOTANICAL GARDEN EXPLANATION OF PLATE PLATE 5 Series RUPESTRES . 47. $. densa var. densa, portion of mat; the dark area to the right is a region of dead vegetative branches, Van Schaack & Freytag 2034 (MO), X 1. S. sibirica, portion of stem, upper side, Calder t$ Billard 4454 (MO), X 10. S. Watsonii, portion of mat, Tryon & Tryon 5060 (MO), X 1. S. Watsonii, portion of stem, upper side, Watson 1370 (NY), X 10. S. Watsonii, portion of strobilus, Watson 1 379 (NY), X 10. S. utabensis, portion of stem, upper side, Flowers 3249 (MO), X 10. S. leucobryoides, portion of stem, lateral view, Munz & Harwood 3789 (NY), X 10. S. asprella, portion of stem, upper side, Johnston 1815 (US), X 10. an {ч d = Q ^ - Uv ~ Re ~ т” Ы ~ ~ Re - ~ ~, Rd / TRYON — SELAGIN РО КУМ Др YN арок Ф 7 NS Т2 Амм. Мо. Вот. Савр., Уог. 4 7 22) — "| i < a e « со) ~ >, Зм т RJ А. һә ба Ж м =. >~ < - S о e о ж са н ANN, Мо. Bor. Сакр., Vor. 42, 1955 [Vor. 42, 1955" TRYON-—SELAGINELLA RUPESTRIS AND ITS ALLIES 99 EXPLANATION ОЕ PLATE PLATE 6 Series EREMOPHILAE S. beruviana, tip of stem in dormant state, Tryon & Tryon 5031 (MO), X 10. S. peruviana, portion of stem, under side, Soukup 1885 (US), X 10. S. arizonica, portion of stem, upper side, W. S. Phillips 2588 (US), X 10. S. arizonica, portion of stem, under side, Sbreve (US), X 10. S. eremophila, leaves from apical bud of stem, Johnston 1047 (US), X 14. S. Parisbii, portion of stem, upper side, and strobilus, E. Palmer 306 (GH), X 10. S. Parishii, portion of stem, under side, E. Palmer 306 (GH), X 10. S. Landii, portion of stem in dormant state, upper side, Pringle 10823 (GH), 10. S. Landii, portion of stem, under side, Pringle 10823 (GH), X 10. А NEW PELLAEA FROM SOUTH AFRICA* ALICE F. TRYON PELLAEA rufa sp. nov., А. Е. Tryon. Rhizoma gracile repens dichotomum paleis concoloribus rufis elongato-lanceolato-triangularibus acumine filiformi basi cordata. Folia 12-30 cm. longa stipite rufo rachide rufa apicem versus flexuosa. Laminae elongato-triangulares bi-tripinnatae segmentis ellipticis vel ovatis vel nonnullis ternatis coriaceis nervis immersis. Sporangia brevissime stipitata. Sporae pallide luteolo-fuscae triplantae laeves. Typus: Compton 16402 (US). Rhizome slender, cord-like, dichotomously branched, long creeping. Scales of the rhizome and stipe-base concolorous, ruddy-tan, elongate lanceolate-triangular, sparsely dentate, the tip filiform, the base cordate, scales surrounding the meristem pinkish or red. Fronds 12-30 cm. long, approximate, the buds paleaceous. Stipe and rachis convex or plane on the upper surface, nearly glabrous, ruddy-tan to red becoming darker and gray with age, the upper portion of the rachis usually flexuous. Blade 10 cm. long and 2 cm. broad to 25 cm. long and 5 cm. broad, bi-tripinnate, elongate-triangular, the pinnae ascending at a broad angle to the rachis, the rachises somewhat flexuous. Segments 3 mm. long and 2 mm. broad to 10 mm. long and 6 mm, broad, elliptical or oval (some ternate), retuse, coriaceous, the veins im- mersed and obscure, the margin reflexed or revolute, border lutescent, crenulate, the young segments reddish: Sporangia with short stalks less than one-fourth the capsule length. Spores 64 per sporangium, tetrahedral, pale yellowish-brown, essentially smooth. Specimens examined: Cape Province: Rock crevices, Ngaap Kop, Laingsburg Dist., Nov. 6, 1944, R. H. Compton 16402 (US); Same locality, Dec. 1, 1941, R. H. Compton 12619 (US); Foothills of Witteberg, April 19, 1925, В. H. Compton 296r (К); Rocky hillsides among succulents, The Great Karoo, near Matjiesfontein, Jan. 24, 1948, Robert |. Rodin 3342 (UC, K, MO); Karoo, Groote Fontein, Mr. Dickson, Lady Barkly recd. 5/75 (K); Locis rupestribus montosis, siccis in dumetris, Kendo, alt. 3000—4000, 1838, Drége (ВМ, К, МО, P). Additional localities from the Cape Province have been cited by Compton! from—Whitehill Ridge; Klipbank, Beaufort West Div.; and by Alston and Schlepe? from— Sutherland: Klein Roggeveld, Schietfontein. This fern of the South African Karroo has had an elusive record for some hun- dred years, It is not surprising for the species is strange among African ferns having its closest ally, Pellaea myrtillifolia in the xermorphic flora of central Chile. These and two species in the southwestern United States form an alliance so close 1Тгапѕ. Roy. Soc. South Africa 19:270. 1931. 2 Jour. South African mend 18:170. 1952. * This in connection = a grant from the Bache Fund of the National Academy Of 5 „зоне Te - баз ч of desert fer: Issued March 24, 1955. (101) [Vor. 42 102 ANNALS OF THE MISSOURI BOTANICAL GARDEN that they were long considered as a single species. Another member of the genus, P. ovata, with a widespread Cordilleran range is more distantly related. Affinities of these species are easily seen in the unique, reddish color of the stipe and rachises and of the segments particularly on the immature frond. They are likewise ex- pressed in the elongate-triangular form and generally tripinnate division of the blade, in the elliptical, retuse form of the segments and elongate, basally cordate scales. Р. andromedaefolia of California and P. intermedia of southern Arizona, New Mexico, Texas and adjacent Mexico are distinguished from the Chilian and African plants in having the rhizome scales with a sclerotic central stipe. Тһе rugose spores and compact, multicipital rhizome of P. myrtillifolia distinguish the Chilian plant from P. rufa. The first collection, made by Drége was included by Kunze? under Allosorus andromedaefolius along with collections from California and Chile. Hooker retained these under Pellaea andromedaefolia. The Chilian species was segregated in Kuhn’s publication on the Mettenius manuscript.®. Later authors regarded the African record as erroneous and it was excluded until reinstated by Marloth’, Compton and more recently by Alston and Schlepe as P. andromedaefolia. The range of the species is local largely within a radius of 40 miles in the vicinity of Whitehill. Several of the collections, including the type, were made by Professor Compton whose regard for this area has resulted in an extensive flora of the Whitehill District and a Botanical Reserve for the preservation of Karroo plants. The species occurs on dry, rocky hillsides at an elevation of 3,000—4,000 feet under 5—6 inches of rainfall. It is apparently able to make rapid growth during the January rains and to tolerate long periods of desiccation. Adaptations to this environment in other genera as Mesembryanthemum, Crassula, Helichrysum and Stapelia have resulted in the striking xeromorphic flora of the Karroo. Related species of Pellaea in Chile and the Southwestern United States exist under environmental conditions of essentially the same extremes. The distribution of such closely allied species in three distant regions is not a unique one. Suc relationships particularly between South Africa and South America have been treated at considerable length in studies on the origin of the flora of South Africa as well as in systematic treatments of several plants and animals as the earthworms Acanthodrilus, the rosaceous genus Acaena, and Menodora of the Oleaceae. The range of the latter? is nearly identical to these species of Pellaea. Evidence from these closely related species of ferns supplies additional data оп the floristic relation- - ships of three continents. SLinnaea 10:503. 1836. ыа Filicum 2:149. 1858. 5Linnaea 36:85. Раз Kapland іп Wiss. Ergeb. Deutschen Cum auf "Valdivia" 2?:276. 1908. TSteyermark, іп Ann. Mo. Bot. Gard. 19:100. 1932 ANN. Мо. Вот. GARD., Vor. 42, 1955 PLATE 7 Pellaea rufa: A, habit X 3⁄4; B, segment with vein-ends exposed; C, rhizome scale; D, spore. MISSOURI BOTANICAL GARDEN STAFF soe > Emeritus Director ae I RUE a UN Сеоксв T. Москве | | : СЗУ Director TE 5: о ANDERSON | CR cod л Pert Дон. нее B4 gee CARROLL У. DODGE, | , в В. VAN. ниві = Pe Mycologist ‘Honorary Cur: еа f the Hert Ei \ | ary Research me > Am que ANDREWS, Paleobotanist ALICE F. TRYON, Котлл M. TRYON, ЕА с ae Азова Assistant Curator of Ње =? br E Herbarium Носн C. CUTLER Curator Museum e Economic Plants ДИ зе аан ас. за. Inu ан. сыр Жы са і сот УАК E, ЕД ос > им x Studies on Ашы Relatives of f aise i: 3 Volume XLII Number 2 Annals of the Missouri Botanical Garden pem МАХ, 1955 | s у QD a ниві Nirodi | 105-130 —— Notes. on the Flora of the Antarctic > Continent а and the Subantarctic I mds. LIV. DET Carrol ой W. Dodge and E "Emanuel ty рий - 131-149 wos ee Annals of the Missouri Botanical Garden A Quarterly Journal containing Scientific Contributions from the Missouri Botanical Garden and the Henry Shaw School of Botany of Washington University in affiliation with the Missouri Botanical Garden. Information sourt BOTANICAL GARDEN appears four times ALS OF THE Mi during the calendar year: February, May, September, and November. Four constitute а volum oss per volume 0 each Subscription Price _ Single Numbers _____ of the ANNALS OF THE Missour! BOTANICAL — are аба in the Agricultural Index, published by the Н. W. Wilson of previous issues Annals of the Missouri Botanical Garden Vol. 42 MAY, 1955 No. 2 STUDIES ON ASIATIC RELATIVES OF MAIZE! NALINI NIRODI? INTRODUCTION The five oriental genera of the Tripsaceae—Coix, Trilobachne, Polytoca, Scler- achne and Chionachne—are especially worthy of an investigation more detailed than has hitherto been accorded them, not only because of their close relationship to maize but in their own right. They form a series of primitive forage plants, grain crops, and weeds, concerning which very little, aside from taxonomic descrip- tions, has appeared in print. As thorough as the study of maize has been, it is not improbable that a better knowledge of its Old World relatives will go some way towards helping to solve the big problem it still presents, that of its origin. Coix is a cereal so ancient that it is mentioned in Vedic literature, was culti- vated by the Aryans on the hill-slopes of the Himalayas, and at the present day is an important food grain of some of the aboriginal tribes of India (Watt, 1904). Watt infers, judging by the amount of attention the plant received in early botanical writings such as those of Pliny, Rumphius, Loureiro, and Gerarde, that formerly it was more extensively cultivated. Wester (1920), however, believes that though the hard-shelled forms of Coix were known in Europe as far back as the time of Pliny, the soft-shelled kinds do not appear to have been known to European writers until the seventeenth century. The Arabs are responsible for its introduction into Spain and thereby to the West in general, and also for the picturesque name by which it is commonly known in Europe, Job’s Tears (Watt, 1904). Though a small and well-defined genus it is made up of ill-defined and poorly understood taxa, largely because of the occurrence of numerous intergrading forms. Therefore, those who have worked on the genus (Watt, Vallaeys, and Mimeur) do not agree too closely about which taxa should be allotted specific rank. In Table I are listed in chronological order as many basonyms in the genus Согх, as . H . . + i iv i dd Бру о assets ft) dude of Dooce Рідну Holder of a Special Fellowship, the Pioneer Hi-Bred Corn Co., Johnston, Тота. (103) [Vor. 42 104 ANNALS OF THE MISSOURI BOTANICAL GARDEN TABL NOMENCLATURE IN THE GENUS COIX е ~ и = | щ ЕС кі» са 1. С. Lacryma Jobi L. Sp. Pl. 972. 1 = 2 Lacryma Jobi 'Thunb. Fl. Јар. 37 1 ~ see 3. C. arundinacea Lam. Encyc. Méthod. 3:422. 1789 1 = nite 4. С. agrestis Lour. Fl. Cochin. 2:551. 1790. 7? 4а. С. Lacryma-Jobi 1. agrestis (Lour.) Backer, Handb. Fl. Java 2:33. = 4 .... 4Ъ. С. es forma "о (Lour.) OE — to dd by] Vall ae in Bu : sAn Belge 39:255. 194 aks 4b 5. C. pendula Salisb. Prod. Е 1 A 6. C. exaltata је а. Eclog Gram. « 13-1820. 1 P 7. C. gigantea К x Roxb. Hort. bas oe 66. 1814 7 Ў Z. 7a. C. Lachry шы Теё. var. ы (Roxb.) Stapf e ex Hook. Fl. Brit. Ind. 7:100. я E i 8. C. aquatica Roxb. in Fl. Ind. 3: 571. 8a 8c 8a 8a. C. gigantea Koen. ex Roxb. var. iss а. Watt, Agr. Ledger 11:198. 1904 $a 8c 82 85. С. Танаа 4. aquatica (Roxb.) Backer, Handbk. Fl. Java 2:33. =з EM 8c --- 8c. С. Lacryma-Jobi forma aquatica (Roxb.) [incorrectly attributed to orae by] Va llaeys, lc. 1948. Ge $6 |: 9. C. chinensis Tod. in fades Sem. Hort. Bot. Pan. [Palermo], 5. 1861. toi р 10. C. Ma-yuen эү in Bull. Soc. Nat. Acclimat. II, 7:442—444; 1881; tos | Tob | = 10a. C. Lacryma-Jobi var. "Ma-yuen" (Roman.) Stapf ex Hook. Fl. Brit. Ind. 7:100. 1896. 10a 10b 10a 10b. C. Lacryma-Jobi for plates (Roman.) [incorrectly attributed to Stapf by] V slash, La з 105 | . 11. С. Lachryma var. stenocarpa а i in Hook. Ic. Pl. III, 8: pl. 1764. й 1888. 11 2594 Ha. С: ме (Oliya) Balansa in Jug de Bot. [Paris] 4:77. 11 116 |... Hb 4-Jobi forma stenocarpa (Oliver) [incorrectly інно t безі by] sla 120-255: us 11b - 12. C. puellarum Bal 15 Kad aa 13. C.* tubulosa Hack. ex Wa rb. in Bot. Jahrb. Hs 260. 1891. 11 14. С. lingulata Hack. in Ont: Bot. Zei чен эме 1891. 7 + E 15. C. Lacryma-Jobi var. monilifer Watt, тя 15 15а | 15 15a. С. Lacryma-Jobi forma monilifer ar [incorrectly attributed to Watt by] Vallaeys, Lo 256 ч 152 - 16. C. Іастута-)офі var я Watt, 1. 16 T 16 Е-е; пева А Koord. in Bull. Jard. Soe ашы ПІ, 1:191. 1918. RE 17 18. C. palustris Koord. l.c. 18b | ...- 18a. C. oo Jobi 4 palustris (Koord.) Backer, l.c oe 18b | ...- 18b. С. Lacryma-Jobi э тесу fris (Koord.) [incorrectly attributed to Backer by by) Aman 1. 5 i. В. 19. С. poilanei Mimeur in Rev. ide Bot. Appl. 31:199. 1951. ори TM 20. C. lacryma-jobi var. major Mimeur, |. c. 202 ‘ 22 20 21. С. lacryma-jobi var. minor Mimeur, Í. с . 204. эө - 21 * Here spelled Coyx. delimited here, as were found. These bear the unmodified numbers 1 to 21. The list includes also a selection from the many combinations which have been made in this genus. These follow their respective basonyms and bear the same number modified by a letter. А number occurring in a column under one of the names, 1955] NIRODI—ASIATIC RELATIVES OF MAIZE 105 Watt, УаПагув, or Mimeur, indicates that the corresponding author equates the opposite name to that bearing the number given. The best-known and most widely spread species is Coix Lacryma Jobi L. Originally native to southern and eastern Asia, it has today virtually run wild throughout the tropical and warm temperate regions of the world. The other species have a comparatively narrow distribution. С. gigantea Koen. ех Roxb. is found chiefly in the eastern and central parts of India, in Ceylon, Burma, and in Japan. Из variety aquatica has been reported in Indochina, western and southern India, Burma, and Japan. C. ouwehandii Koord. is endemic to Sumatra. Another endemic species is the one set up by Mimeur, C. poilanei, found in Laos. Vallaeys (1948) suggests the Malay Archipelago as the center of origin of Coix, taking into consideration the multiplicity of varieties present there, some of which are not found elsewhere in the wild state. Accordng to Burkill (1935), “it seems early to have become a cereal of some importance in the hills of Indochina and southern China, and it is probable that the beginnings of the cultivated races should be sought for in that part of the world." As an article of diet, Coix used to be known as being most important during times of famine, helping to tide over seasons when the rice crop failed. Its nutritive importance, however, has impressed many investigators, and of recent years it has been gaining in popularity as a cereal and has come to be more generally cultivated. Wester was responsible for popularizing it in the Philippines where it is known as Adlay. Since thea similar campaigns have been set in motion in Java, Ceylon, the Belgian Congo, Guatemala, Nicaragua, Costa-Rica, Bolivia, Peru, Chile and Brazil. Analyzed along with certain other famine foods, its energy value was found to be the highest of all those studied (Paton and Dunlop, quoted by Watt, 1904). Wester (1920) records the results of analysis of the Bukidnon form of adlay as closely approximating wheat in starch and protein content and exceeding it in fat. Comparing it with two other great cereals of the world, he says, "with a greater protein and fat content than either rice and corn it is a more complete human food than either of these cereals." An idea of its productiveness is obtained from Pieris' statement (1936) that a crop of adlay from an acre of land will provide sufficient food for a family o four people for about one year. Duthie (1888), speaking of its value as a fodder, says it is largely eaten by cattle and reported to be very fattening. Vallaeys (1948) writes of its curious use as “the object of practices and beliefs held in magic” and also of the preparation of a refreshing drink from the grain which acts as a diuretic and а depuratory. In hospitals in China it is said to form a good and cheap substitute for barley in the making of gruel. It is known to be used in the elimination of stones from the bladder (Burkill, 1935). A decoction of the roots is given to children as a vermifuge (Ridley, 1906). It is best known, however, for its use for decorative purposes, in the making of bead-baskets, rosaries, and various articles of personal adornment. Seeds of the wild forms with the hard, lustrous shells are the ones thus employed, usually those [Vor. 42 106 ANNALS OF THE MISSOURI BOTANICAL GARDEN belonging to var. stenocarpa, which has elongated, cylindrical false fruits, and those of var. /уріса, with false fruits of the characteristic tear-drop shape. This list of the uses of Coix could be appropriately closed with a quotation from Vallaeys. Wherever man has lived it is possible to discover vestiges of ancient cultures and of plants whose use is lost. Speaking of the regrettable tendency to limit the number of products useful as food, he writes, "Cette limita- tion, cette ‘standardisation’ ne peuvent étre considérées comme un progrés. Voila pourquoi dan certain pays, on regard la coix comme une 'solution' a des problemes d'importance capitale d'ordre diététique, economique et social". Both Polytoca and Chionachne are generic names which owe their origin to Robert Brown (Bennett, Horsfield & Brown, 1838). Many species have been described for both genera as a result of which the literature has become cluttered with synonyms. In “А contribution to the knowledge of the Indian Maydeae" (1931), Henrard made an attempt to clarify their taxonomy. The following table is based on Henrard’s and Pilger's treatments. It is constructed in the same way as Table I. TABLE II NOMENCLATURE IN THE GENUS POLYTOCA 1 215 Be две 1. Apluda digitata L. f. Suppl. 434. 1781. la 1 Ja la. Polytoca digitata (L. f.) Henr. > мт = Herb. Leid: 67:10. 1931. Та Та 2. Coix beferoclita Roxb. a Ind. 3 la | — 3. Polytoca bracteat . in i s & pee Pl. Javan. Rar. 20. ор. 5. 1838. Та | 12 4. Cyatborbacbis Wallichiana Nees ex Steud. Syn . Pl. Gram. 1:403. 18 4a | 4a 4a. Polytoca Wallicbiana (Nees) Benth. in ere ur. € fg pes 52; не. 4а | 4а 5. Sclerachne cyatbopoda Muell. Fragm. Р 1873. Ба 1 = 5a. Polytoca €t “Os ll.) Е. es i dex н ~ :1849. 1902. $a | 5a 6. Polytoca m > enth., 1. c. 1881. 6 6 7. Chiona be: Mas ii Balansa in Jour. кен ea на ы, 2%, 1890. 74 te 7a. Polytoca Massii шом) $сһепсК ех 1931. 7а 78 8. Polytoca javanica Henr. in Blumea 3: t Lin 8 8 NOMENCLATURE IN THE GENUS CHIONACHNE > "|. | 1. Coix атипатасеа Koenig ex Willd. Sp. Pl. 4:203. 1805. За | — 2. Coix barbata Roxb. Fl. Ind. 3:569. 1832. ja | 34 3. Coix Koenigii Spreng. o 1:239.: 1825, За За 3a. Chionachne Koenigii (Spreng.) Thwaites, Enum. Pl. Zeyl. 357. 1864. за | 32 4. Coix crypsoides Muell. i in Bot. Zeit. 19:334. 1861. за | = 5. Polytoca semiteres Benth. ex Hook. Fl. Brit. Ind. 7:101. 18 $a. — 5a. Chionachne semiteres (Benth. ex Hook.) Henr. in Meded. Rijks. Herb. Leid. 67: Jet 2 a 5a 5a 6. Chionachne biaurata Hack. in Philipp. Jour. Sci. Suppl. 1:263. 1906. 6 6 7. Сы Hubbardi iana Henr. in Blumea 3:162. 1938. 7 7 1955) NIRODI—ASIATIC RELATIVES ОЕ MAIZE 107 The species of both Polytoca and Chionachne range in distribution from India to Australia. Polytoca Massii is found in Tonkin; P. Wallicbiana in northeastern India and Pegu; P. digitata ranges from India to southern China and Java; P. cyathopoda is confined to northern Australia and Queensland; P. macrophylla grows in the Malay Archipelago, New Guinea, and the Louisiade Archipelago; P. javanica has been reported only from Java by Henrard. Chionachne Koenigii is distributed throughout India and in Tonkin, Ceylon, the Celibes, and Queensland; Ch. biaurata is endemic to the Philippines; Ch. semi- teres is found in the Deccan and Burma; Ch. Hubbardiana is endemic to Queens- land. The only economic use reported for these grasses is as fodder. Chionachne Koenigii, which is widely spread over India, is used as a fodder when young (Duthie, 1888). MATERIALS All the plants used in this study were grown from seed at the Missouri Botanical Garden, both outdoors and in the experimental greenhouse. Through the Plant Introduction Station, Beltsville, Maryland, seeds of various varieties of Coix were obtained from India, Ceylon, the Philippines, Central America, and the Belgian Congo, and large numbers of plants were raised. Since Coix forms a polyploid series with haploid chromosome numbers of 5, 10 and 20, it was decided to choose one representative of each of these numbers for cytological study, as follows: (1) A species with a haploid chromosome number of 5, bearing small, hard, chalk-white false fruits, kindly procured for our purpose from Cuttack, India, by Dr. N. Krishnaswami, Cytogeneticist, Agricultural College, Coimbatore, India. (2) “Blue Adlay", a variety of Coix Lacryma Jobi L., with а haploid chromo- some number of 10, obtained from Trinidad through the courtesy of Dr. William L. Brown, Geneticist, Pioneer Hi-Bred Corn Company, Тоза. (3) A species with a haploid chromosome number of 20, bearing greenish- brown false fruits, collected on the Western Ghats (the mountain range that ex- tends for 800 miles along the southwest and west coast of India), also sent by Dr. N. Krishnaswami. Specific or varietal names have not been appended to (1) and (3), because their morphological characters do not conform exactly to any of the descriptions of Coix species met with in the literature. One species of Chionachne—Ch. Koenigii (Sprengel) Thwaites (п = 10), sent by Dr. N. Krishnaswami from the Millet Breeding Station, Coimbatore, India, is included in this study as is also one species of Polytoca—P. macrophylla Benth. (n = 20)—obtained from Trinidad by Dr. William L. Brown. Full descriptions of the five taxa are given below, and inclusive herbarium specimens are being prepared for distribution. No members of the genera Sclerachne and Trilobacbne were available for cyto- logical or morphological study. [Vor. 42 108 ANNALS OF THE MISSOURI BOTANICAL GARDEN The following is a key to distinguish the three genera: I. Female spikelets enclosed in чи covering formed by modified indurated leaf sheath............ Coix II. "t spikelets enclosed in hardened covering yid — by indurated glume I . Terminal spikes male, vn; spikes mixed or Polytoca b. АП spikes = flowers of both sexes Chionachne (1) Coix from Cuttack, India.— | Root-system very much branched with extremely fine roots. Culm erect with numerous basal leaves. Leaf-blades linear, very narrow, up to 5 mm. broad, the longest leaves attaining a length of about 60 cm.; prominent white midrib; at union with sheath a well-marked suture which is narrow in the middle and fans out on either side; upper leaf surface sparsely set with hairs each arising from a colorless, crateriform gland. Sheaths ribbed, dotted on the outer face with gland- ular hairs similar to those of the blade. Inflorescences axillary, on triangular peduncles, as many as 8 arising from one node; false fruit small compared to most other varieties of Coix, 8 mm. long and 5 mm. in diameter, flattened on one side, often constricted towards the top, turbinate, chalk-white, hard. Male portion of the inflorescence up to 3.5 ст. long. Glume I 1 cm. long, lanceolate-ovate, slightly falcate, emarginate, the margins reflexed, the glume thereby 2-keeled, en- closing glume II, the keels winged, one of them throughout, the other only in its upper half, the wings serrate. Glume II about the length of glume I, acute, entire, margins infolded. This description comes fairly close to that given by Mimeur (1951) for her species, Coix poilanei, endemic to Laos. The finely branched root-system, the num- erous basal, narrow, linear leaves, and especially the dimensions and color of the false fruits, coincide with her description of C. poilanei. She says, however, that “les feuilles sont en continuité directe avec les gaines dont elles ne se distinguent pas”, whereas in our plant the blades are separated from the sheaths by a distinct suture. The glands on the sheath are colorless and not brown as stated by Mimeur. Also, for the male spikelets she reports “une seule fleur par épillet,” while the spike- lets of our plant exhibit the 2-flowered condition. In referring to her species, she remarks that it is "trés proche de Coix aquatica” (С. gigantea var. aquatica Watt). The chromosome number of the plant under discussion is п = 5, which is the same as that determined for var. aquatica by Mangelsdorf and Reeves (1939). (2) "Blue Adlay”, Coix Lacryma Jobi L., from Trinidad. This is the wild “Сох Lacryma-Jobi proper” of Watt to which he gave the varietal name typica and of which he speaks as "the normal form" referring to the false fruits. Stem erect. Leaf-blades lanceolate, broad, up to 3.5 cm. in width, in length variable, up to 56 cm., auriculate, glabrous, “except for the double row of ascending teeth along each of the veinlets of the upper surface—a peculiarity that gives the texture of the leaf the appearance of being embroidered and makes it backwardly hispid” (Watt, 1904). Writing further about this character, Watt says: "all the 1955] NIRODI—ASIATIC RELATIVES OF MAIZE 109 forms of this species are at once recognized by the texture of the leaf and its peculiar scabrosities.” Ligule membranous; union of blade with sheath marked by a broad suture. Sheaths quite smooth. False fruits ovoid-oblong, 10—11 mm, long, 7 mm. in diameter, hard, smooth, polished, bluish-white. Male portion of the in- florescence up to 3.5 cm. long, spikelets in two's or three's. Glume I 7 mm. long, lanceolate-ovate, emarginate, strongly keeled, winged in the upper half, wings serrate. Glume II about the length of glume I, acute, margins infolded, entire. Text-fig. 1. Coix poilanci Mimeur; fig. 2. «Віце Adlay"; fig. 3. Coix gigantea Koen: ex Roxb: All about 14 natural size. (3) Coix from the Western Ghats, India.— Stem perennial, erect. Leaf-blades linear-lanceolate, between 2 and 2.5 cm. broad and up to 30 cm. long; both upper and lower surfaces sprinkled with hairs each arising from a crateriform gland; margins serrate. Sheath also with crateri- [Vor. 42 110 ANNALS OF THE MISSOURI BOTANICAL GARDEN form glands bearing hairs, but here the glands of a dark reddish hue; union of blade with sheath marked by a broad suture. Ligule membranous; just above ligule, at base of leaf, a tuft of stiff hairs similar to those on the rest of the leaf. Inflorescences axillary, 2 or more spikes emerging from the sheath of a short leaf, the spike emerging first, being borne on a longer peduncle than that of the others; peduncles flattened on one side, glabrous; false fruits elongate, pyriform, 1 cm. long, 5-6 mm. in diameter, mouth oblique, pale green with brown markings, slightly flattened on one side. Male spikes up to 4 cm. long, drooping, spikelets mostly in 3’s, the lateral ones sessile and the middle one pedicellate, each 2-flowered, Glume I 9 mm. long, ovate-oblong, emarginate, the margins reflexed, the glume thereby 2-keeled, enclosing glume II, the keels winged throughout, margins of the wings serrulate. Glume II winged, acute, margins entire. The description above corresponds in general to that given for C. gigantea by Roxburgh (1832) and that by Watt (1904). The important differences are discussed below: 1. Roxburgh describes the sheaths as "smooth", and Watt as "quite glabrous and eglandular”. In the specimens under discussion, however, the sheaths are sprinkled with sharp hairs arising from crateriform glands of a dark reddish color. 2. The leaves are not "narrow" and “linear” as in Watt’s description but "about 1 inch broad" as in се $. They are evidently ес to variation, for Watt speaks of “. іг J. E. Smith's sheet of C. gigantea . . . . which has gland- dotted broad be чё 3. With regard to i dii false fruits, Watt describes them as "flattened on one side, and with 2 furrows longitudinally on the flat face . . . ." While the ones described here are flattened on one side, no furrows are present, nor are any furrows mentioned by Roxburgh or by Mimeur (1951). The name Coix gigantea is retained for the entity used in this study. The un- certainty of its correct application is indicated by a question mark. The chromo- some number of "2n == 40” tallies with that reported for C. gigantea by E. К. Janaki Ammal (Darlington & Janaki Ammal, 1945). (4) Polytoca macropbylla Bentham.— Stem robust, erect. Leaf-blades up to 55 cm. long, between 2.5 and 3.5 ст. broad, glabrous, lanceolate, acuminate, cordate at base; margins dentate; midrib prominent; union with sheath marked by a broad suture. Sheath glabrous. Ligule narrow, membranous. Female spikelets 1 cm. long, variable in glume I which may be entire or toothed at summit or have 2 lateral teeth, sometimes 1. Маје spikelets in pairs, each pair made up of 1 pedicellate and 1 sessile spikelet. Glume I up to 11 mm. long, acute, asymmetrical in the pedicellate spikelet, margins infolded with a broad wing along the fold on one side, wing serrate, other side merely folded slightly, lanceolate-ovate in the sessile spikelet, margins infolded, winged along the fold symmetrically on both sides. Glume II laterally compressed, folded along the 19551 NIRODI—ASIATIC RELATIVES OF MAIZE 111 Text-fig. 4. Polytoca macrophylla Benth. About l4 natural size. middle, keeled along the fold in the pedicellate spikelet; keel winged in upper half, wings serrate, very slightly keeled in the sessile spikelet. The chromosome number agrees with Avdulov's determination (1931) of 2n — 40. (5) Chionachne Koenigii (Sprengel) Thwaites.— Stem perennial, erect, freely branching at the base. Leaf-blades about 40 cm. long (said to attain a length of 2 feet, Ranga Achariyar, 1921), up to 1.2 cm. broad, linear, with a prominent midrib; union with sheath well-marked by a suture. Sheaths with scattered tubercle-based hairs at the base, sheath margins hairy. Ligule membranous; nodes densely ringed with hairs. Inflorescences axillary, most of them clustered at the ends of very long peduncles which are flattened or slightly concave on one side; leaf, in the axil of which several peduncles arise, with a sheath about 2.5 cm. long, the blade, although shorter than normal, fairly well-developed; each spike in a cluster enclosed in its lower portion by a bladeless awned sheath, 1–1.5 cm. long; 1, sometimes 2, of the lowermost spikelets female, borne in pairs of which one spikelet is sessile and the other pedicellate, the pedicel partly fused with the rhachis; rhachis pubescent. Female spikelets oblong, 5.5 mm. long, еп- [Ver. 42 HZ ANNALS OF THE MISSOURI BOTANICAL GARDEN v | / Fig. 5. Chionachne Koenigii (Spreng.) Thwaites. About 14 natural size. closed in glume I which becomes shiny and hard at maturity; at the union of the edges of the glume a groove through which the rhachis passes. Male spikelets 2- flowered. Glume I up to 8 mm. long, acute, pubescent, the margins infolded, the glume winged along the fold on both sides in the sessile, but only on one side in the pedicellate, spikelet; wings serrate. Glume II acute, laterally compressed and folded along the middle with a pronounced keel along the upper half of the fold in the pedicellate spikelet, only slightly keeled in the sessile. 19551 NIRODI—ASIATIC RELATIVES ОЕ MAIZE 113 Сутотосу А number of cytologists have definitely established the existence of the chromo- some numbers in multiples of 10 in two of the three genera of the New World Maydeae, Euchlaena and Zea. Tripsacum is a well-known exception; all the species and varieties have 18 pairs of chromosomes or multiples of this number. Among the Oriental Maydeae the gametic or reduced number most often encountered is also 10; this is evident from the following table: $3 85 88 | 85 OZ | Az 1. Coix Lacryma Job -- 20 | Kuwada, 1915 2: state гла Га enth. -- 40 | Avdulov, 1931 ET bryma bib set “Oliver -- 20 | Reeves and Mangelsdorf, 1935 4. Sclerachne punctata Brown -- 20 | Reeves and Mangelsdorf, 1935 5. С. Lacryma-Jobi var. "Ma-yuen" (Roman.) Stapf -- 20 | Mangelsdorf and eeves, 1939 6. Polytoca Meri Stapf [Chionachne Koenigii — 20 | Mangelsdorf and pre е) Thwaites] Reeves, 1939 Z2 BARS EA oxb. 5 10 y eg ep Reeves 8. C. gigantea Koen. ex Roxb. — 120,40! E. a Janaki And, 9. С. poilanei Mimeur (?) 5 10 Nirodi 1955 The numbers in all the above taxa except 3, 4, and 7 have been confirmed in this study. С. aquatica (Roxb.) and C. poilanei Mimeur (?) are exceptional in having the gametic number 5, which is the lowest found in the Gramineae and reported so far only for the unrelated genera Briza and Anthoxanthum and for five species from two of the sub-genera of the Sorgbastreae. METHODS For cytological study entire male spikes were fixed in a solution of glacial acetic acid and absolute alcohol 1:3. Two days to a week later they were transferred to 70 per cent alcohol and stored under refrigeration. Young spikes that had barely begun to emerge from the ensheathing spathe were found to be at the right stage for meiotic studies. Temporary smears were stained either with aceto-carmine or with aceto-orcein. A few were made permanent by using the dry-ice freezing technique described by Conger and Fairchild (1953). ile pachytene smears of Coix were being examined to investigate the possi- bilities for making chromomere analyses, stray cells from somatic anther-tissue were noticed showing well-stained and well-spread diploid chromosomes. were assumed to be tapetal cells. Proof of the assumption was obtained when young anthers fixed in a solution of 3 parts absolute alcohol to 1 of glacial acetic acid were embedded in paraffin, sectioned and stained with crystal violet. [Vor. 42 114 ANNALS OF THE MISSOURI BOTANICAL GARDEN Smears made for the observation of tapetal chromosomes were treated in a slightly different way from smears made for meiotic studies. The contents of ап anther were squeezed out in a drop of stain and well stirred with an iron needle. The debris was allowed to remain. After the placing of the cover-slip the slide was heated over a spirit-lamp and pressed vigorously between blotters to release the tapetal chromosomes from the nuclear membrane and spread them evenly. CHROMOSOME MORPHOLOGY The tapetal cells of Coix, Polytoca and Chionachne go through the process of endomitosis. At the time the pollen mother cells are undergoing pachytene the tapetal cells have already passed from the uninucleate to the binucleate condition, and smears made at this stage show many of them in endoprophase. Speaking of endoprophase stages in tomato tapetum, Brown (1949) remarks that they are “valuable for a study of somatic chromosome morphology". The same can be said of the three genera of grasses dealt with in this paper. Observations on chromo- some morphology were based almost entirely on tapetal cells from anther smears. At about the middle of endoprophase, the chromosomes reach a particularly favor- able degree of contraction and stain deeply enough to be identified individually. Sectioned material shows them surrounded by a nuclear membrane but in smearing the membrane may be lost. Coix poilanei Mimeur (?).—This species, together with Coix gigantea var. aquatica Watt, occupies the lowest position in the polyploid series formed by Согх. Its somatic chromosome number determined both from root-tips and tapetal cells was found to be 10. At very early endoprophase the chromosomes are very much attenuated and are up to five times as long as they are at the end of this phase. They are very loosely coiled, and though they often overlap they can easily be traced throughout their length (pl. 8, fig. 1). A characteristic of tapetal chromosomes in Coix is their striking differentation into regions of different staining capacity. Extremely lightly staining areas occur which are very constant in their location at a particular stage. These stand out in good contrast to the deeply staining portions, thus making the chromosomes easy to identify. Usually, and especially when they occur at the ends of the chromo- somes, these achromatic regions are visibly split, with the daughter chromatids often divergent. A point to be noted at this stage is that quite often along the achromatic regions the chromomeres are visible, and two of these may be seen side by side in each diverging half of a chromosome indicating that it is made up of four chromatids (pl. 8, fig. ЗА). “Achromatic” as used here is a relative term. These regions appear pale in comparison with the remaining portions of the chromo- some. e chromomeres in the achromatic regions often stain clearly enough to be distinctly visible. e five chromosomes of the haploid set have been designated A, B, C, S and D, $ standing for the satellite chromosome (pl. 8, figs. 2-6). Chromosome А has а 1955] NIRODI—ASIATIC RELATIVES ОЕ MAIZE 115 very long achromatic region at one end, and Chromosome B has shorter achromatic regions at both ends. Chromosome C has two achromatic regions, one in each arm midway between the end and the centromere. Chromosome S is recognized by the presence of the satellite. In its elongated state it also exhibits an achromatic region in either arm but this may disappear on contraction. Chromosome D is very like Chromosome S but devoid of the satellite and shorter. Figures 2—6 of pl. 8 show tapetal chromosomes in varying stages of contrac- tion. Even when quite strongly contracted they can be told apart——Chromosome A and B by the difference in number and location of the achromatic regions, Chromosome S by the satellite, Chromosomes C and D by their difference in length, Chromosome D being shorter. When the achromatic regions do not occur at the ends of the chromosomes their contraction leaves wide gaps in the body of the chromosomes. Figure 6 of pl. 8 shows a late endometaphase with most of the chromosomes divided throughout but held together at the centromeres. The diverging arms give them a characteristic cruciform appearance. ing to the fact that there may be two or even three places along the chromo- somes that show clear spaces and look like centromeres, the true position of the spindle-attachment region was ascertained from this stage and from root-tip squashes. Root-tip chromosomes examined at anaphase were found to be bent into V's with equal or slightly unequal arms indicating that all the chromosomes had median or sub-median centromeres. Plate 10, fig. 2, shows one large tapetal nucleus containing the tetraploid number of chromosomes after the completion of one endomitosis. Since the chromosomes are in early endoprophase it is probably entering a second cycle. It was not ascertained whether a second endomitosis is completed nor what degree of ploidy a tapetal cell is capable of attaining. When the chromosome number for Coix aquatica Roxb. was reported by Mangelsdorf and Reeves (1939) to be 2n — 10 they observed that "no variation in the number was found". In C. poilanei Mimeur (2), though the tapetal cells most often showed 10 chromosomes, occasionally cells were found where 12 and rarely 11 were present. Though the nuclear membrane is ruptured on smearing, it is possible to recognize the contents as belonging to a single nucleus. Thus the possibility that the extra chromosomes might have been displaced from a neighbor- ing nucleus due to the pressure exerted while smearing is eliminated. Besides, when pollen mother cells were examined, occasionally 6 bivalents were found at diakinesis; fig. 1 of pl. 11 shows an MI telophase where 6 chromosomes have been distributed to either pole. Due to the way in which the anthers were collected it was not pos- sible to say with certainty whether this variation in number was between plants or between spikelets of one plant. Coix Lacryma Jobi L.—The tapetal cells of three kinds of this species were smeared to see if varietal differences were manifest in the chromosomes. The first was "Blue Adlay", i. e. var. typica from Trinidad (described above). The other two were two kinds of var. "Ma-yuen". Cultivated forms of C. Lacryma jobi have been placed under this varietal name. Cultivation makes itself evident by [Vor. 42 116 ANNALS OF THE MISSOURI BOTANICAL GARDEN turning the stony capsule spathe soft and brittle and striated. One of the two kinds mentioned came from the Philippines and has chalk-white, brittle, and very faintly striated capsular spathes. The shape is that of var. typica. The other has dark brown, brittle and strongly striated capsular spathes, somewhat elongated and with a mouth larger than usual. For convenience these shall be referred to as var. “Ма-учеп (1)” and var. "Ма-упеп (2) ". A general similarity is obvious in the chromosome sets of these three kinds of Coix. Оп the whole, the chromosomes are smaller than in the 10-chromosome species. Another difference is that achromatic regions, when they occur, are usually found only towards the ends of the chromosomes. Therefore, the pronounced gaps that are left in the body of the chromosomes after a certain degree of contraction described in the 10-chromosome species are rarely met with here. The attempt to discern homologues among the somatic chromosome sets of var. typica, var. "Ma-yuen (1)” and var. “Ма-уџеп (2)" was confined to those chromosomes provided with pronounced achromatic ends as they helped as markers. They have been marked by small letters of the alphabet (pl. 8, figs. 7 and 8; pl. 9, fig. 1). The nucleolus chromosomes are very similar in all three, with distal ends achromatic and split. In var. typica and var. "Ma-yuen (2)" both nucleolus chromosomes in а diploid set possess satellites but in var. "Ма-ушеп (1)" a satellite was seen on only one of them. Its homologue is seen to be in connection with the nucleolus but no satellite was observed. As in the 10-chromosome Coix, the centro- meres are all median or sub-median which agrees with Taylor's observations (1925) on root-tip chromosomes. This is fairly clear in the tapetal chromosomes (pl. 8, figs. 7 and 8; pl. 9, fig. 1) but much more evident in pl. 11, fig. 7, where meiotic chromosomes of var. Курса are seen in second metaphase. The chromatids prior to separation are held together at the centromere which stands out markedly. Coix gigantea Koen. ex Roxb. (?).—Here the chromosomes, in spite of being twice as numerous as in the varieties of C. Lacryma Jobi, are no smaller. In fact, two pairs approach A and B in size, the two largest chromosome pairs in the 10- chromosome Coix. Coix poilanei? (pl. 8, fig. 3), the varieties of C. Lacryma Jobi (pl. 8, figs. 7 and 8), and С. gigantea (?) (pl. 8, fig. 2) all show their respective tapetal chromosomes in endoprophase at more or less the same degree of contrac- tion. By comparing these figures it can be seen that many of the chromosomes of C. gigantea (?) are intermediate in size between those of C. poilanei (?) and of the 20-chromosome Coixes. The centromeres of C. gigantea chromosomes stand out more clearly than those of any of the other kinds described so far. Again, all of them are either median or sub-median. This is confirmed by fig. 1 of pl. 13, where the chromosomes are in prophase of the second meiotic division. The chromatids are held together only at the centromeres and the four diverging arms of each chromosome are seen to be approximately equal. Due to the much larger number of chromosomes present in C. gigantea? it was difficult to obtain cells which showed them spread at an appreciable distance from 1955] NIRODI—ASIATIC RELATIVES ОЕ МАШЕ #17 one another. The cell shown in fig. 2 of pl. 9 was stained in aceto-orcein and does not show the nucleolus. Diakinesis stages clearly indicate the presence of two pairs of nucleolus chromosomes. From observations of carmine-stained tapetal cells it was seen that two of the nucleolus chromosomes were similar to the ones indicated by arrows (pl. 9, fig. 2). Although only one pair of chromosomes has been indicated (by X’s) in var. "Ma-yuen (2)" and in С. gigantea? as being prob- ably homologous, a general similarity between the genomes of the 20- and the 40- chromosome Coixes is very much in evidence. The characteristic achromatic regions are found in practically all the chromosomes of C. gigantea? and, as in the varieties of C. Lacryma Jobi, are mostly confined to the ends. The achromatic ends are visibly split and often divergent. Occasionally, tapetal cells were noticed which had two extra chromosomes. It was not ascertained how often this occurred in the tapetal cells but when studying the meiosis, counts were made in a hundred pollen mother cells at diakinesis. Of these only two were found with an extra bivalent and one with one bivalent missing. The variation seems to be within plants. Polytoca macrophylla Benth.—The tapetal chromosomes of this species differ from those of Coix in that there is no conspicuous differentiation into chromatic and achromatic regions. They stain more or less uniformly, as can be seen in pl. 10, fig. 3, where they are in endoprophase. Two nucleolus chromosomes with satellites can be observed in the same figure. Figure 4 of this plate shows them at endometaphase, more condensed and already divided as shown by the longitudinal split apparent along each chromosome. Conspicuous constrictions indicate the position of the centromeres. Plate 10, fig. 5, shows a late endometaphase. The divided halves of each chromosome are seen to diverge and are held together only at the centromere. This is a stage corresponding to the one in C. poilanei? (pl. 8, fig. 6). Such cruciform configurations due to union only at the centromeres are sometimes characteristic of colchicine-treated plants undergoing c-mitoses. Levan (1938) describes it in Allium. Brown (1949), speaking of endomitotic cycles in the tapetum of tomato, says that in some respects they resemble mitoses upset by treatment with colchicine. No cruciform stage, however, nor any evidence of protracted union at the centromere was observed by him. It is not known whether the union at the centromere in Polytoca macrophylla and Coix poilanei? is much protracted but the cruciform configurations form a definite stage in the endo- mitotic cycle of the tapetum in these two instances. In the contracted condition at endometaphase (pl. 10, fig. 4), Polytoca chromo- somes bear a likeness to somatic chromosomes of corn treated with paradichloro- benzene. This produces effects very similar to those of colchicine, one of which is to shorten the chromosomes much beyond their normal length. Chionachne Koenigii (Sprengel) Thwaites.—In this species of Chionachne the tapetal chromosomes, like those of Polytoca macrophylla, stain more or less uni- [Vor. 42 118 ANNALS OF THE MISSOURI BOTANICAL GARDEN formly. Plate 9, fig. 3, shows them in endoprophase. The presence of satellites on the nucleolus chromosomes is doubtful What seemed like satellites were only observed in one cell in which the chromosomes were in early endoprophase and were thin and long. They were not observed either in root-tip chromosomes or in pollen mother cells at diakinesis. The centromeres, though not marked by соп- spicuous constrictions as in Poly£oca, are easy to locate. At endometaphase, the chromosomes are very much contracted and the beginnings of a longitudinal split can be observed in some of them. MEIOSIS Observations were made on pachytene smears of Coix, Polytoca and Chionachne. The pachytene stage in Coix was found to be markedly different from that of Polytoca and Chionachne in that conspicuous knobs were observed in Сойх that were absent in the other two genera. The part played by chromosome knobs in the determination of relationships between maize and its relatives is only too well known. Considering this feature of chromosome morphology, it could be assumed that of the three genera Coix is the most closely related to the New World Maydeae. Longley (1941) pointed out the similarity between the chromosomes of Zea and those of Coix. Comparing the bivalents of corn and Согх, he showed that the variation in size in the two sets of bivalents was very similar. A general similarity seems even more apparent on comparing the pachytene stages of a 20-chromosome Coix with those of corn. A glance at fig. 7 of pl. 10, however, shows that of the six knobs observable only one is internal and the rest terminal, a condition at variance with that in corn where the majority of knobs are internal In this respect, therefore, Coix seems to resemble Euchlaena or Tripsacum more than corn (knobs in Euchlaena are both internal and terminal and in Tripsacum the majority are terminal). Figure 7 shows a pachytene smear from a variety of C. Lacryma Jobi obtained from the Sugar Cane Breeding Station, Coimbatore, India. From a study of its gross morphology this plant is classified as var. typica. Yet the number of its knobs differs widely from that of var. typica from Trinidad ("Blue Adlay"). This variation is quite in accordance with the situation in corn. Longley (1939), writing about such variation in knob number in corn, observes that it is "frequently very pronounced when plants of varieties from different geographical regions are compared". In corn, knob-forming centres can exist but not be evident unless a knob is formed at the knob-forming point, an occurrence which depends on several factors, one of them being the amount of knob material available (Longley, 1939). Figure 6 of pl. 10 shows a single chromosome of “Blue Adlay", with an elongated, terminal knob. In appearance and position it is exactly like one of the terminal knobs of var. typica from Coimbatore. It may well be that the position and number of knob-forming centres are similar in the two plants but that knobs make their appearance at more centres in var. typica from Coimbatore than in var. typica from Trinidad. The knobs of 1955] NIRODI—ASIATIC RELATIVES ОЕ MAIZE 119 var. typica from Coimbatore vary in appearance from large, somewhat elongated ones to smaller, more rounded ones (pl. 10, fig. 7). The pachytene chromosomes of C. poilanei? were in too tangled a state to allow proper observation. The knobs in C. gigantea? (pl. 12, fig. 1) are small, rounded and mostly terminal Мо knobs were observed in Polytoca macrophylla and Chionachne Koenigii. If а chromomere analysis of Coix chromosomes were undertaken it would seem best to start with varieties of C. Lacryma Jobi. The pachytene stages are easy to handle and the chromosomes spread out well enough to be examined. The chromo- some number being the same as in corn, similarities with and differences from corn chromosomes would be easier to identify. Only late diplotene stages of C. poilanei?, “Blue Adlay” and Polytoca macrophylla were observed. These showed a maximum of three chiasmata per bivalent in the two Coixes and up to five in Polytoca (pl. 9, fig. 4; pl. 13, fig. 3). In C. poilanei? the chromosome pairs could be identified at this stage and have been marked in the figure with the same letters used to identify tapetal chromosomes. By late diakinesis in C. poilanei (?) (pl. 9, fig. 5), all ex- cept the satellite bivalent are completely terminalized and lie side by side or are connected at the ends by slender filaments. The bivalent of satellite chromosomes is still connected at the satellite end and free at the other, forming a У. АП five bivalents could easily be identified even in this contracted phase and have been indicated by letters as in previous figures of C. poilanei (?). In "Blue Adlay", five rings and five V's were observed to make up the ten bivalents (pl. 11, fig. 2); the achromatic ends mentioned in describing the tapetal chromosomes are still noticeable and are especially clear in the distal end of the satellite bivalent. Diakinesis in C. gigantea (?) deviates somewhat from the perfectly normal for- mation of bivalents exhibited by C. poilanei (?) and “Blue Adlay". One association of 4 chromosomes was seen in some pollen mother cells and some showed two rings of 4 (pl. 12, fig. 5) but the majority had 20 bivalents. These three kinds of be- havior were found in pollen mother cells from the same spikelet. Evidently, the two sets of 4 chromosomes that tend to form rings possess some incipient homology. C. gigantea (?) presumably has two pairs of nucleolus chromosomes. Two bivalents were often seen in association with the nucleolus at diakinesis and one pollen mother cell was observed which had two distinct nucleoli each associated with a bivalent. But the prominent satellites characteristic of Coix were only seen to be displayed by one of the bivalents (pl. 12, fig. 6). Both Polytoca macrophylla and Chionachne Koenigii behave normally at diakinesis, regularly forming 20 and 10 bivalents respectively (pl. 13, fig. 4; pl. 14, fig. 3). The former has satellites to its two nucleolus chromosomes but none were observed in Chionachne. Formation of metaphase plates and separation at anaphase I take place in the normal fashion. There is an orderly migration to the poles; and no formation of [Vor. 42 120 ANNALS OF THE MISSOURI BOTANICAL GARDEN bridges, lagging chromosomes, nor any other irregularities were observed in any of the five plants (pl. 9, fig. 6; pl. 11, fig. 3; pl. 12, fig. 7; and pl. 14, fig. 4). At telophase I the chromosomes are seen to have formed two compact groups. By the time they emerge from telophase I and enter prophase II, the arms of the component chromatids of each chromosome are seen to diverge from the centro- mere with the result that the chromosomes appear cruciform. A nucleolus has been re-formed in each group. This behavior is similar in all five taxa under dis- cussion (pl. 12, fig. 3; pl. 11, fig. 6; pl. 13, fig. 1; pl. 14, fig. 1). At this stage, the second meiotic prophase, a curious body was seen to occur regularly in each of the two groups of chromosomes in C. gigantea (?). It ap- peared to be composed of a few darkly staining granules embedded in a lightly staining matrix (pl. 14, fig. 1). It varied in shape, being sometimes small and rounded, sometimes larger and lens-shaped. The number of granules also varied. Ап attempt to investigate the nature of this body was made by halving an anther containing pollen mother cells in second prophase and staining one half with aceto- carmine and the other with aceto-orcein. With aceto-carmine both nucleolus and the body showed up clearly, while the orcein-stained preparation stained neither the nucleolus nor the matrix. The granules showed extremely faintly. It would seem from this evidence that the body was in some way connected with the nuc- leolus. It was not seen in contact with any of the chromosomes or with the nucleolus, and so far as was observed was not seen to be in later stages. The remaining stages of the second meiotic division leading to tetrad formation are completed in normal fashion. In C. gigantea? and Ch. Koenigii at anaphase II separation of chromatids and migration along the spindle may not take place simultaneously in both cells (pl. 13, fig. 2; pl. 14, fig. 6 Relationship between Chromosome Number and Gross Morpbology.— Coix Poilanei?, with 10 somatic chromosomes, is essentially a small plant and even when grown outdoors did not exceed a height of three feet. The varieties of C. Lacryma-Jobi, 2n = 20, of which several forms of var. “Ma-yuen” and var. typica from different regions were raised, showed a wide variation in size, some attaining a height of 3 feet, others growing 9 feet tall. Most of them averaged 7—8 feet. Coix gigantea?, with the highest chromosome number (2n — 40), exhibits characteristics typical of polyploids. Roxburgh (1832) describes it as attaining a height of 8-15 feet. Our plants of C. gigantea? were only between 5 and 6 feet high, but the leaves were coarser than in the two other species and the plants came into flower much later. Relationship between Chromosome Number and Geographical Distribution.— Species of Согх seem to have reached an optimum in the somatic number of 20. While the 10- and 40-chromosome kinds are more or less restricted in their distri- bution, C. Lacryma Jobi and its varieties have spread over the world. 19551 NIRODI—ASIATIC RELATIVES OF MAIZE 121 Discussion On finding that accurate chromosome counts could be made from tapetal cells, a study of their chromosome morphology was undertaken. While tapetal chromo- somes at the right stage for study were being sought it gradually became apparent that they undergo what is essentially a process of endomitosis. Endomitosis in tapetal cells was first reported Бу Witkus (1945) in Spinacia. Brown (1949) showed that tapetum endomitosis in tomato differed in some respects from the process in Spinacia. In both cases the process ends in the formation of polyploid nuclei but whereas in Spinacia the nuclear membrane remains intact throughout and the chromosomes are irregularly distributed, in tomato there is a breakdown of the nuclear membrane and the chromosomes form clumps or are arranged in plates at endometaphase. In spite of the breakdown of the nuclear membrane, Brown retains the term “endomitosis,” explaining that “lack of movement of the chromosomes on a spindle rather than persistence of a nuclear membrane would seem to be the better criterion of endomitosis.” In the present study the behavior of tapetal cells follows a somewhat similar pattern to that described in tomato. The cells are uninucleate to start with but soon become binucleate following a normal mitosis, i. e., with spindle formation but without the formation of a cell wall. The chromosomes become visible as sep- arate entities at endoprophase and contraction commences. At endometaphase the chromosomes are very much contracted and the beginning of a longitudinal split, prior to anaphase separation, is distinctly noticeable (pl. 8, fig. 9; pl. 10, fig. 4). However, the cruciform configurations at late metaphase described in C. poilanei? and Polytoca and presumably present in the other three taxa are reported by Brown to be absent in tomato. At endoanaphase the chromosomes form dense clumps. The fact that the chromosomes fall apart at this stage was only gathered from the small size of stray chromosomes at the periphery of the clumps. That the completion of endomitosis results in tetraploid nuclei was concluded from the larger size of many resting nuclei and from instances like the one por- trayed in pl. 10, fig. 2, where a tetraploid nucleus of С. poilanei? is shown with its chromosomes in endoprophase of a second endomitosis. Whether this is con- cluded with the formation of an octoploid nucleus is not known. Except in C. poilanei? no tetraploid nuclei were observed in a state in which the chromosomes could be counted. For this to be possible the chromosomes would have to emerge from the resting stage and start on a second endoprophase, and this does not seem to happen in the other four taxa, and was only rarely observed in C. poilanei? Presumably then, endomitotic activity stops after one division and the formation of tetraploid nuclei, i. e., two tetraploid nuclei per tapetal cell. Brown reports three endomitotic cycles as normal to tapetal cells in tomato. Sectioned material at endoprophase showed the nuclear membrane intact and fig. 9 of pl. 8 shows an endometaphase in var. "Ма-ушеп (1)" with the membrane [Vor. 42 122 ANNALS OF THE MISSOURI BOTANICAL GARDEN still intact. Whether it disappears at a later stage is not definitely known but it seems unlikely. Assuming that it does not, the endomitosis in Coix, Poly£oca and Chionachne resembles Spinacia regarding this point. According to Brown, in tomato the membrane disappears toward late endoprophase. From what study has been made on Coix chromosomes it would appear that the various species and varieties dealt with in this paper form a fairly homogeneous group, with 5 as the basic chromosome number. С. poilanei? would then be con- sidered a diploid species, C. Lacryma Jobi and its varieties tetraploids, and C. gigantea? an octoploid. Morphologically, the octoploid is similar to the diploid in an outstanding characteristic, the presence of crateriform glands bearing hairs. (Watt considers this characteristic as one of the more important ones that sep- arate what he terms "the gigantea-aquatica series" from the Lacryma-Jobi series). In its chromosomes, however, the octoploid is more like the tetraploids. Since no crossing experiments were undertaken for this study any hypothesis regarding the possible origin of the tetraploid and octoploid should be made with the necessary caution. It could be surmised that the tetraploids arose by a crossing of two similar diploids and a doubling of the chromosome number of the hybrid. They function like true diploids forming no multivalents. In the octoploid two nucleoli have been observed, and there are two sets of nucleolus chromosomes. Multivalent configurations are found. It probably originated from the tetraploid as an autopolyploid. Since its origin, its genomes have undergone changes so that it has come to function essentially as a diploid but the occasional formation of tetravalents indicates its autopolyploid beginnings. On the basis of its pachytene chromosomes, Сойх as a genus seems closer to the New World Maydeae than to Polytoca and Chionachne. In staining reactions, as previously described, Polytoca and Chionachne behave differently from Coix. Morphologically, they are similar in having the walls of their fruit-cases formed from the hardened first glume, whereas Coix has hardened, modified leaf- sheaths. The probability must not be overlooked that investigation of species of Polytoca and Chionachne other than those treated here might disclose the presence of pachytene knobs. The resemblance of endometaphase chromosomes of Polytoca to those of artificially shortened somatic chromosomes of corn has been mentioned. On the basis of the present evidence, it would seem that Polytoca and Chionachne bear a closer affinity to each other than to either Coix or to the New World May BIBLIOGRAPHY же + Р. (1931). Кагуо ose а deen der Familie Graminéen. Bull. Appl. Bot., Genet. & Pl. Breed. (Leningrad), Sup Backer, c. ГА: (1927). вод voor de Fiora van a Ra 2:32-33. Bailey, F. M. (1902). Queensland Flora 6:1 Balansa, B. Seen Catalogue des Graminées ge lIndo-Chine Francaise. Jour. de Bot. 4:76-84. Bennett, J. J., Т. Horsfield, and Robert Brown (1838). Plantae Javanicae Жана. рр. 15—19. 19551 NIRODI—-ASIATIC RELATIVES OF MAIZE 123 Bentham, е (1878). Flora im pei А. ід --, m . Notes on Gram Jou Вог. 19:14—134. ————, ay р. о (1883). баню утри y 1112-1115. Berger, с. ге and E. Witkus (1943). A Airaa study of c-mitosis in the plant Spinacia oleracea with с ari ative observations on Allium сера. Bull. Torr. Bot. Club 70:457—467. Bor, N. L. (1940). Flora y^ Assam 5:453—459. um of tomato. Amer. Jour. Bot. аот Brown, pud W. (1949). Endomitosis in the tapet Burkill, I. H. (1935). A dictionary of economic darn of the Malay Peninsula 1:629—6 Conger, Ala ucile M. ы ld (1953). A quick-freeze method for making smear ғ де n n D., and L Pon: ee echnol. 281—283. Cooper, D. C. (1933). Nuclear. divisions in the tapetal cells of certain angiosperms. Amer. Jour. Bot. йі 364. Darlington, C. D., and Е. К. Janaki Ammal (1945). Chromosome atlas of cultivated plants. 397 pp. London. Duthie, J. F. (1888). The fodder gr — M northern India. 90 pp. Roorkee. Gamble, J. S. (1934). Flora of Madra Hackel, E. eet Descriptiones Gramin um novorum. Oest. Bot. gro 41: ). Notes оп "Philip e Gramineae. II. Phil ilipp. Jour. Sci. "dig A :263—269. 31). А Бабине тона to the knowledge of the Indian Mi due Meded. Rijks — 06 dec J ~ (19 Lei w Chionachne from Queensland. Blumea 3:238—242. ————, (1 210). On a new species of У оне from Java. Ibid. 238—2 1^ Heyne, K. (1927). Die nuttige planten van патике ч io 1:150- Holland, T. H. (1926). Adlay—its uses and р . Agr ae Taito 1926:60—61. Hooker, J. D. (1896). eun E Briti sh India 7:10 Jacquin, J. Е. von (1820). Eclogae Graminum. р. а. Koorders s Hi (1918). uer али einer von Dr. Ouwehand im Toba-See, in Sumatra, entdeckten Art von Coix. ее Sul ПІ, 1:190-191. Jar Камай, Y. ec Ueber de оленя zahl von Zea Mays L. = Mag. Tokyo 29:83—89. Lamarc ene y Encyclopédie Méthodique. 3:4 Levan, Аве ( b. effect of colc agr Ae m itoses in Айн». Hereditas 24:471—486. С. (1753). "ЫШ peana pP- ine vi, fili us 780; Suppl. ‚ Vere Gen "BL, et Sp. РІ. Lnd: A. = МЕ 39). Кпођ 86 са. on corn osomes. Jour. Agr. PA 59:475—4 = 41). и а morphology in maize and Из relatives. Вог. Rev. 7: oes 218. е у. ds dq: Miri Coc Verein 2:550—551. : : — , P. С. and В. С. Reeves (1939). The origin of Indian corn and its relatives. Texas Agr. Éxp. Sta, Ball. pe 1-315. Merrill, E. E (1906). An ее of Philippine Gramineae with keys to genera and species. hilipp. Jour. Sci. Suppl. 1:320-3 Mimeur, Geneve ve (1 ge Systématique ылга ifique du genre Coix et systématique varietale de yma-Jobi. Morp wc е? tte petite céréale et étude de за plantule. Rev. Internat. Coix Bot КО A Agric. Trop. 31:197- x Muel - ; Pieris, Н. А. (1936). M Trop. Agr. (Ceylon) 86:217-2 3 deri 5. í Бан Gramineae. Ш. Unterfamilie Panicoideae. ‘Engler & Plantl's Nat. Pflanzenfam. Rape Filed (1921). A handbook of some South Indian grasses. pp. Reeves, R. G., and P. C. i pm (1935). Chromosome numbers in Mure of Zea Mays L. Amer. Nat. 69: 633-6 idle . N. (1906). Malay drugs. Agr. — Straits & Fed. Malay States 5:193-206. Roxburgh, W. (1832). Flora Indica 3:568—5 tege А. Van es d = Leydensis. P. 5 Salisbury, R. A. (1796). romus stirpium in Horto ad Chapel Allerton vigentium. Smith, E Н. (1933). Nucliar divisions in the tapetal cells of Galtonia candicans. утте е ud Bot. 20:341—347. Sprengel, race (1825). и vegetabilium 1:238—239. - Stapf, O. ( ee Tis (Coix Тасфтута L. var. stenocarpa). Kew Bull Misc. Inf: ы Е, d 1 Gramin 1:405. e у ум eros ў er distinguishing characters in plants. Tas. William Randolph (1925). Chromosome constrictions as er. Jour. Bot. 12:238—244. [Vor. 42, 1955] 124 ANNALS OF THE MISSOURI BOTANICAL GARDEN un C. P. (1784). Flora Japonica. Thwaites, G. H. K., and J. D. Hooker Tea Enumeratio на NDA Ey 357. Valleri, 'G. os Le “Coix Lacryma «эб ". Bull. Agric. Congo Belge 39:2 Warburg, О. von (1891). Beitrage zur Kenn nis de г papvanischen d Bot. le 13:230-272. Watt, dum ane. А dictionary of the economic produc of India. 2:492—500. — (1904). Coix spp. ог Job’s Tears. A review of NT Рис С Iulormitien, Agr. Ledg. Н 189-22 9. horses P. (1926). Comparative morphology of the Oriental Maydeae. Indiana Univ. Studies Weser, P. T. 920). i on Adlay. Philipp. — ie 13:217-222. Willdeno йе L (1805). Species Plantarum 4:20 Witkus, E. E (1945). й tapetal cell а in Spinacia. Amer. Jour. Bot. 32:326-330. EXPLANATION OF PLATE PLATE 8 Figs. 1-4. Coix poilanei Mimeur. Tapetal chromosomes in endoprophase. Fig. 5. Coix poilanei. Tapetal chromosomes in endometaphase. Fig. 6. Coix poilanei. Tapetal chromosomes in late endometaphase. ig. 7. “Blue Adlay" (C. Lacryma Jobi var. typica Watt), from Trinidad. Tapetal chromosomes in endoprophase. . 8. Var. "Ma-yuen (1)” (С. Lacryma Jobi L. var. “Ма-уцеп” Stapf). Tapetal dés in endoprophase. Fig 9. Var. "Mayuen (1)". Tapetal chromosomes in endometaphase. Ann. Мо. Вот. GARD., Vor. 42, 1955 NIRODI—ASIATIC RELATIVES OF MAIZE 125 Fig. 1. Fig. 2. Fig. 3. prophase. Fig. 4. Fig. 5. Fig. 6. [Vor. 42, 1955] NIRODI—ASIATIC RELATIVES OF MAIZE EXPLANATION ОБ PLATE PLATE 9 Var. “Ма-учеп (2)" (С. Lacryma-Jobi L. var. “Ма-уџеп” Stapf). Tapetal chromosomes in endoprophase. Coix gigantea Koen. ex Roxb. Tapetal chromosomes in endoprophase. Chionachne Koenigii (Spreng.) Thwaites. Coix poilanei. Mimeur. Diplotene. Coix poilanei. Diakinesis. Coix poilanei. Anaphase I. Tapetal chromosomes in endo- ANN. Мо. Bor. Слвр., VoL. 42, 1955 PLATE 9 NIRODI—ASIATIC RELATIVES OF MAIZE 126 Fig. 1. Fig. 2. [Vor. 42, 1955] ANNALS OF THE MISSOURI BOTANICAL GARDEN EXPLANATION ОБ PLATE PLATE 10 Coix poilanei Mimeur. Showing twelve tapetal chromosomes in endoprophase. Coix poilanei. Tapetal cell with tetraploid number of chromosomes under- going a second endomitotic cycle. Fig. 3. Fig. 4. Fig. 5. Fig. 6. Fig. 7. Polytoca macrophylla Benth. Tapetal chromosomes in endoprophase. Х 1350. Polytoca macrophylla. Tapetal chromosomes in endometaphase. X 1350. Polytoca macrophylla. Tapetal chromosomes in late endometaphase. "Blue Adlay." Pachytene. X 1350. Coix Lacryma Jobi var. typica, from Coimbatore. Pachytene. Х 1350. ANN. Mo. Вот. Ganp., Vor. 42, 1955 РТАГЕ 10 NIRODI—ASIATIC RELATIVES OF MAIZE PLATE 11 ANN. Мо. Bor. Gard., Vor. 42, 1955 ам, NIRODI—ASIATIC RELATIVES OF MAIZE | VoL. 42, 1955] зәм во» ю з NIRODI—ASIATIC RELATIVES ОЕ MAIZE EXPLANATION OF PLATE PLATE 11 Coix poilanei Mimeur. Telophase I of a 12-chromosome cell. "Blue Adlay." Diakinesis. "Blue Adlay." Anaphase 1. "Blue Adlay." Telophase I. "Blue Adlay." Late Te!ophase I. "Blue Adlay." Prophase Il. "Blue Adlay." Metaphase Il. 127 |Хог. 42, 1955] 128 ANNALS OF THE MISSOURI BOTANICAL GARDEN ExPLANATION OF PLATE PLATE 12 Fig. 1. Coix gigantea Koen. ех Roxb. Pachytene. X 1350. Fig. 2. Coix poilanei Mimeur. Telophase I of normal 10-chromosome cell. Fig. 3. Coix poilanei. Prophase Il. X 1150. Fig. 4. "Blue Adlay." Telophase II. Fig. 5. Coix gigantea. Diakinesis showing ring formation. X 1350. Fig. 6. Coix gigantea. Diakincs's showing satellite bivalent. X 1800. Fig. 7. Coix gigantea. Anaphase І. X 1350. ANN. Mo. Bor. Garp., VoL. 42, 1955 PLaTE 12 # NIRODI—ASIATIC RELATIVES ОЕ MAIZE ANN. Mo. Вот. Сакр., Vor. 42, 1955 PLATE 13 NIRODI—ASIATIC RELATIVES OF MAIZE [Vor. 42, 1955] NIRODI—ASIATIC RELATIVES ОЕ MAIZE 129 ExPLANATION OF PLATE PLATE 13 Fig. l. Coix gigantea Koen. ex Roxb. Prophase II showing extra bodies. Х 990. Fig. 2. Coix gigantea. Anaphase II. X 1170. Fig. 3. Polytoca macrophylla Benth. Diplotene. X 875. Fig. 4. Polytoca macrofbylla. Diakiness. X 1350. Polytoca macrophylla. Telophase I. Х 1100. a 99 мл ANN. Mo. Вот. Сакр., Vor. 42, 1955 PLATE 14 NIRODI—ASIATIC RELATIVES OF MAIZE 130 Nays [Vor. 42, 1955 ANNALS OF THE MISSOURI BOTANICAL GARDEN EXPLANATION OF PLATE Polytoca macrophylla Benth. Prophase Il. Х 1170. Polytoca macrophylla. Chionachne Koenigii (Spreng.) Thwaites. Diakinesis. Chionachne Кости. Chionachne Koenigii. Chionachne Koenigii. Chionachne Koenigii. PLATE 14 Telophase II. X 800. Anaphase I. Telophase I. Anaphase П. Х 1550. Telophase П. X 1350. ша LICHENOLOGICAL NOTES ON THE FLORA OF THE ANTARCTIC CONTINENT AND THE SUBANTARCTIC ISLANDS. I-IV CARROLL W. DODGE AND EMANUEL D. RUDOLPH* In 1947-1948, the Australian National Antarctic Research Expedition estab- lished weather stations at Heard Island and Macquarie Island and somewhat later at Mawson in MacRobertson Land, on the Antarctic Continent. Through the kindness of Dr. Н. W. Jessep of the National Herbarium, Melbourne Botanic Garden, and Dr. A. M. Gwynn, Medical Officer and Biologist of the Australian National Antarctic Research Expedition ( A.N.A.R.E.), we have had the privilege of studying the lichen coilections. I. ADDITIONS To THE LICHEN FLORA OF HEARD ISLAND Previous collections were reported by Dodge (1948) based on the British Australian New Zealand Antarctic Research Expedition (B. A.N.Z.A.R.E.), No- vember 27-December 2, 1929, all between Atlas Cove and Corinthian Bay. The weather station was established at Atlas Cove in December 1947 (Scholes, 1951) and closed in 1955. The earlier collections received in 1949 were by D. Alan Gilchrist, Medical Officer; the collector of the later specimens was not recorded on the herbarium labels and are cited: A.N.A.R.E. The island was more accurately mapped in 1948 (A.N.A.R.E. 1949). Most of the southern part of the island is covered by glaciers and ice fields, so that the lichen collections have come from the northwestern part, especially the Cape Laurens peninsula on the northwest corner. Thirty-two species are represented, of which three are new and fifteen have not been previously reported although known from Kerguelen Island to the northwest, making a total of 52 species known from Heard Island. THELIDIUM HEARDENSE Dodge, B.A.N.Z.A.R.E. Rept. В. 7:44. 1948. The thallus is lighter (vinaceous buff) than the type, but it agrees micro- scopically. North of Cape Laurens on volcanic rock, A.N АКОН, 75. THELIDIUM PRAEVALESCENS (Му!) Zahlbr., Deutsche Südpolar Exp. 8:51. 1906. Verrucaria praevalescens Nyl., in Crombie, Jour. Linn. Soc. Bot. 15:192. 1876. As in most previous collections, all of our material is sterile although the thallus has a characteristic appearance. West Bay, A.N.A.R.E. 743; north of Cape Laurens, on broken lava, in cave, A.N.A.R.E. 67, 72. * Graduate Student, Henry Shaw School of Botany of Washington University. Issued June 23, 1955. (131) [Vor. 42 132 ANNALS OF THE MISSOURI BOTANICAL GARDEN MICROGLAENA KERGUELENA (Му) Zahlbr., Deutsche Südpolar Exp. 8:51. 1906. Verrucaria kerguelena Nyl., in Crombie, Jour. Bot. Brit. For. 14:22. 1876. D. Alan Gilchrist 5, growing with Steinera Werthii Zahlbr. New to Heard Island. XANTHOPYRENIA heardensis Dodge & Rudolph, n. sp. Type: Heard Island, north of Cape Laurens at base of black cliffs, A.N. A.R.E. 76. Thallus areolatus, areolis 0.3—0.5 mm. diametro subconvexis substipitatisque, marginibus liberis crenulatis, humectatis obscure flavo-viridibus, siccatis flavo- citrinis, homoeomericus; algae Xanthocapsa, coloniis sphaericis aut oblongis, vaginis flavo-brunneis, cellulis 7-8 и diametro; hyphae 2-3 и diametro inter colonias algarum. Perithecia immersa, 1-3 in quaque areola, subsphaerica, 200-250 p pre ostiolo minuto; parathecium obscure brunneum, 15-20 y crassitudine, cel polyhedricis; asci 12-15 Х 115 р, leptodermei; ascosporae octonae, gg monostichae, late fusiformes, 19-24 X 8-9 д, tenui cum halone dum in asco, dein liberae 24 X 10-11 р, biloculares, cellula superiori majori, hyalinae, septo con- strictae. Thallus areolate, areoles 0.3—0.5 mm. in diameter, slightly convex above, sub- stipitate below, margins free, crenulate, dark yellow-green when moist, old gold to buffy citrine when dry, homoeomerous; algae Xanthocapsa, colonies rounded to oblong, of 4—16 cells with a thick yellow-brown sheath at first, becoming densely packed in a homogeneous gel with abundant hyphae and cells more rounded, 7-8 p in diameter, each with its own sheath about 2 p thick; hyphae 2-3 in diameter, filling most of the interstices between the algal cells and colonies. Perithecia immersed or nearly so, 1—3 per areole, showing as minute dark brown to black points; subspherical, about 200—250 y in diameter, ostiole small; wall dark brown, 15-20 д thick, of polyhedral cells; asci 115 X 12-15 y, thin-walled, 8- spored; ascospores imbricately monostichous, broad fusiform, 19-24 X 8-9 m, with а thin halo while still in the ascus, 24 X 10-12 д when free, bilocular, the upper cell larger, hyaline, slightly constricted at the septum. STEINERA GLAUCELLA (Tuck.) Dodge, B.A.N.Z.A.R.E. Rept. B. 7:66. 1948. Pannaria glaucella Tuck., Bull. Torrey Bot. Club 6:57. 1875. Growing with Placopsis bicolor (Tuck.) B. de Lesd., on broken lava, north of Cape Laurens, A.N.A.R.E. 80. New to Heard Island. STEINERA WrnTHm Zahlbr., Deutsche Südpolar Exp. 8:43. 1906. D. Alan Gilchrist 5 and unnumbered specimen LICHINA ANTARCTICA Crombie, Jour. Bot. Brit. For. 14:21. 1876. On crystalline rock (sandstone?), A.N.A.R.E. New to Heard Island. 19551 DODGE & RUDOLPH—LICHENOLOGICAL NOTES, I-IV 133 ? SIPHULASTRUM CLADINOIDES Dodge, B.A.N.Z.A.R.E. Rept. B. 7:69. 1948. We have doubtfully referred our material to this species. It forms compact hemispheric cushions up to 2.5 cm. in diameter. The habit resembles young dense tufts of Spbaeropborus fragilis Pers. from the Arctic and Subarctic but it is com- pletely different in structure. The thalli are stiffer than in the type. Although the tips of the ultimate branches become very dark brown, we have been unable to find any reproductive structures. It may represent а new species. North of Cape Laurens on broken lava, A.N.A.R.E. 78. New to Heard Island. CorrEMopsimiUM pyrenuloides Dodge & Rudolph, n. sp. PL 15, не. 3. Type: Heard Island, north of Cape Laurens, on humus in cave, A.N. A.R.E. 65. Thallus gelifactus, flavidus, siccitate fragillimus, membrano-foliosus, semi- pellucidus, lobis irregularibus sinibus rotundatis, erectis, subrugosis; homoeomericus; algae Xanthocapsoideae, cellulis 5—6 р diametro; hyphae sparsae, tenues. 0.4—0.5 mm. diametro; amphithecium non bene Apothecia lecanorina immersa, his periclinalibus intertextis; evolutum; hypothecium са. 25 y crassitudine, hyp thecium 100 y altitudine; paraphyses 3-4 p diametro, pachydermeae; asci clavati, juventute apice incrassati; ascosporae octonae, brunneae, late fusiformes, biloculares, septo constrictae, pachydermeae, 25-30 Х 10-14 p. Thallus a yellowish gel, very fragile when dry, yellow, foliose-membranous, semipellucid, lobes very irregular with somewhat rounded sinuses, erect, subrugose; homoeomerous; algae Xanthocapsoid, cells mostly singly dispersed in the gel with- th, 5—6 и in diameter, occasionally in small subspherical colonies out evident shea d a somewhat thicker up to 40 u in diameter with thin sheaths about each cell an sheath surrounding the colonies, best seen in the amphithecium; hyphae very slen- der, not abundant; some bacterial colonies imbedded in the gel as well as a few moss (?) fragments. Apothecia lecanorine, immersed or nearly so, amphithecium not clearly differentiated, a zone of Xanthocapsoid colonies with more abundant subvertical hyphae; hypothecium about 25 р thick, of interwoven periclinal hyphae; paraphyses 3—4 p in diameter with thick walls, occasionally branched; thecium 100 р tall; asci clavate, thickened at the tip when young, dif- fluent, 8-spored; ascospores brown, broad, fusiform, sometimes flattened on one side, bilocular, constricted at the septum, wall thick, protoplasts rounded, rarely ell, resembling the spores of Pyrenula, 25-30 X about 0.4-0.5 mm. in diameter; 2 distinct protoplasts in each c 12—14 р. The thallus suggests the Pyrenopsidaceae, but the algal cells mostly occur singly without an evident sheath and the gel is much softer when moist so that we have been unable to secure good sections. It is possible that the fungus is a parasitic Tichothecium, but the mycelium below the hypothecium is continuous with that of the thallus and it lacks a parathecium. Although C. pyrenuloides is foliose, it seems more closely related to Collemopsidium than to any other genus of the Pyrenopsidaceae. [Vor. 42 134 ANNALS OF THE MISSOURI BOTANICAL GARDEN North of Cape Laurens, on humus in cave, A.N.A.R.E. 64, type, 67 sterile. PANNARIA DICHROA (Hook. f. & Tayl.) Crombie, Jour. Linn. Soc. Bot. 16:220. 1876. Lecanora dichroa Hook. f. & Tayl., London Jour. Bot. 3:643. 1844. Probably owing to the very rough surface of the lava, the lobes are somewhat ascendant and imbricate. The thallus is darker, light brownish olive and not stained with iron. On another very dense rock without locality, nearly covered by Blastenia keroplasta Zahlbr., the thallus is quite typical. D. Alan Gilchrist 3; north of Cape Laurens on broken lava, A.N.A.R.E. 74. Lecwea AUBERTI B. de Lesd., Ann. Crypt. Exot. 4:99. 1931. D. Alan Gilchrist 8. LECIDEA ASSENTIENS Nyl., in Crombie, Jour. Bot. Brit. For. 13:334. 1875. North of Cape Laurens, on volcanic rock, A.N.A.R.E. 75. New to Heard Island. LECIDEA SUBASSENTIENS Nyl., in Crombie, Jour. Bot. Brit. For. 14:21. 1876. D. Alan Gilchrist 4. RHIZOCARPON KERGUELENSE Dodge, B.A.N.Z.A.R.E. Rept. B. 7:116. 1948. Јаска Valley, on cliffs, A.N.A.R.E. 33. RHIZOCARPON Mawson! Dodge, B.A.N. Z.A.R.E. Rept. В. 7:115. 1948. D. Alan Gilchrist 7. New to Heard Island. RHIZOCARPON JOHNSTONI Dodge, B.A.N.Z.A.R.E. Rept. В. 7:118. 1948. On moraine of Schmidt Glacier, A.N.A.R.E. 242. New to Heard Island. CLADONIA PHYLLOPHORA (Tayl) Dodge, B.A.N.Z.A.R.E. Rept. B. 7:132. 1948. Cenomyce phyllophora Tayl. in Hook. f. & Tayl, London Jour. Bot. 3:652. 1844. : North of Cape Laurens, A.N.A.R.E. 70. CLADONIA JOHNSTONI Dodge, B.A.N.Z.A.R.E. Rept. B. 7:135. 1948. Some podetia are nearly clothed with coarse granules, rarely almost isidioid, thus somewhat resembling C. pyxidata (L.) Fr. Among mosses on broken lava, north of Cape Laurens, A.N.A.R.E. 69. New to Heard Island. ARGOPSIS CYMOSA (Crombie) Stzbgr., Ber. Thiatigk. St. Gall. Naturw. Ges. 1889—90:231. 1891 Stereocaulon cymosum Crombie, Jour. Linn. Soc. Bot. 15:182. 1876. On broken lava, north of Cape Laurens, A.N.A.R.E. 71. New to Heard Island. URECEOLINA KERGUELIENSIS Tuck., Bull. Torrey Bot. Club 6:58. 1875. Jacka Valley, 600 ft., on cliff, A.N.A.R.E. 34. New to Heard Island. 19551 DODGE & RUDOLPH—LICHENOLOGICAL NOTES, I-IV 135 ASPICILIA LYGOMMA (Nyl.) Dodge, B.A.N.Z.A.R.E. Rept. B. 7:164. 1948. Lecidea lygomma Nyl. in Crombie, Jour. Bot. Brit. For. 13:334. 1875. On boulders, 20 ft., West Вау, A.N.A.R.E. 743, growing with Thelidium praevalescens (Nyl.) Zahlbr. and Kuttlingeria crozetica (Zahlbr.) Dodge. New to Heard Island. ASPICILIA DISJUNGUENDA (Nyl) Dodge, B.A.N.Z.A.R.E. Rept. B. 7:167. 1948. Lecanora disjunguenda Nyl. in Crombie, Jour. Bot. Brit. For. 15:105. 1877. D. Alan Gilchrist 6, 7, 9. ASPICILIOPSIS MACROPHTHALMA (Hook. f. & Tayl) Dodge, B.A.N.Z.A.R.E. Rept. B. 7:175. 1948. Urceolaria macrophthalma Hook. f. & Tayl., London Jour. Bot. 3:640. 1844. D. Alan Gilchrist ro. Pracopsis BICOLOR (Tuck.) B. de Геза. Ann. Crypt. Exot. 4:100. 1931. Placodium bicolor Tuck., Bull. Torrey Bot. Club 6:57. 1875. The specimens are much paler than usual, probably from less iron in the rocks, and cephalodia are very rare. North of Cape Laurens, on broken lava, A.N.A.R.E. 80, growing with Steinera glaucella (Tuck.) Dodge; on cliff, Јаска Valley, 600 ft., A.N.A.R.E. 31. UsNEA Taytort Hook. f., London Jour. Bot. 3:657. 1844. Moraine of Baudessen Glacier, 800 ft., А.М.А.В.Е. 250. New to Heard Island. UsNEA INSULARIS (Lamb) Dodge, B.A.N.Z.A.R.E. Rept. B. 7:211. 1948. Neuropogon insularis Lamb, Jour. Linn. Soc. Bot. 52:215. pl. 8, fig. 17. 1939. Only a few plants, beginning to form cupulate apothecia. Mt. Aubert de la Rüe, 300 ft., A.N.A.R.E. 4. UsNEA TRACHYCARPA (Stirton) Müll. Arg. Nuovo Giorn. Bot. Ital. 21:37. 1889. Neuropogon trachycarpus Stirton, Scottish Nat. 6:105. 1881. D. Alan Gilchrist 2. New to Heard Island. BLASTENIA KEROPLASTA Zahlbr., Deutsche Südpolar Exp. 8:28. 1906. Two specimens without locality, A.N.A.R.E. New to Heard Island. KUTTLINGERIA CROZETICA (Zahlbr.) Dodge, B.A.N.Z.A.R.E. Rept. B. 7:226. 1948. Caloplaca crozetica Zahlbr., Deutsche Südpolar Exp. 8:29. 1906. Atlas Cove, on rocks above high water, A.N.A.R.E. 225; West Bay, 20 ft., A.N.A.R.E. 743; on cliff in Јаска Valley, 600 ft., A.N.A.R.E. 32; D. Alan Gil- christ т. New to Heard Island. BurrLiA sUBPLICATA (Nyl) Müll. Arg., Bot. Jahrb. [Engler] 5:158. 1884. Lecidea subplicata Nyl. in Crombie, Jour. Bot. Brit. For. 13:334. 1875. On cliff, Jacka Valley, A.N.A.R.E. 33. [Vor. 42 136 ANNALS ОЕ THE MISSOURI BOTANICAL GARDEN RINODINA ASPICILINA Zahlbr., Deutsche Südpolar Exp. 8:50. 1906. Two small specimens without locality, A.N.A.R.E. DEUTEROLICHENES (LICHENES IMPERFECTI) Occasionally lichenologists have encountered conidial fructifications on lichen thalli, sometimes associated with apothecia, sometimes not. Müller-Argau (1881) described an otiform conidial structure which he named a campylidium. Vainio (1890) found the same structure on a thallus of Lopadium јетра да (Nyl.) Zahlbr. (Lecidea perpallida Nyl.) and referred it to Cyphella aeruginascens Karst. Spegazzini (1909) named it as a genus of lichenes imperfecti, Chlorocyphella, based on C. subtropica Speg. Keissler (1927) transferred Cyphella aeruginascens Karst. and described several new varieties. Mameli Calvino (1930) proposed the name Deuterolichenes to include Chlorocyphella, and Cengia Sambo (1937, 1941) and Rizzini (1952) have reported species of Chlorocyphella not associated with apothecia. Malme (1935) considered C. aeruginascens (Karst.) Keissl. to be a conidial stage of Lopadium perpallidum (Nyl) Zahlbr., and Dodge (1953) de- scribed а campylidium as a conidial state of Г. Deightoni Dodge. Campylidia have also been observed on Sporopodium sp. Müller-Argau (1890) described another type of fructification, the orthidium, which resembles an apothecium except that the thecium is replaced by conidiophores, and the senior author has seen a similar fructification on a foliicolous thallus from Jamaica. Our material contains an orthidium-bearing lichen from Heard Island, resembling Ephelis Fr. of the Excipulaceae, which has unicellular spores, while those of our lichen become septate as do those of Ephelis trinitensis Cooke & Massee, the imperfect state of Balansia trinitensis Cooke & Massee on Panicum palmifolium in Trinidad, B.W.I. EPHELIDIUM Dodge & Rudolph, п. gen. Type: Ephelidium heardense Dodge & Rudolph. Thallus crustosus, indeterminatus, sorediosus, ecorticatus, heteromericus; algae protococcoideae. Orthidium sessile, concavum, margine persistente; conidiophorae simplices; conidia singulatim disposita, terminalia, acicularia, septata. Thallus crustose, indeterminate, sorediose, ecorticate, heteromerous; algae protococcoid. Orthidium sessile, concave with a persistent margin, resembling a lecanorine (or biatorine as the algae die above) apothecium; conidiophores un- branched; conidia single, terminal, acicular, long remaining unicellular but finally multiseptate. EPHELIDIUM heardense Dodge & Rudolph, n. sp. Pl. 15, fig. 1. Type: Heard Island, Atlas cove at foot of Poa mound, A.N.A.R.E. 147. Thallus crustosus, indeterminatus, 0.7-1.2 mm. crassitudine, citrinus, sore- diosus; ecorticatus; stratum algarum ca. 280 р crassitudine, cellulis protococcoideis, 8.7-12.2 р diametro; medulla crassa, hyphis 1 м diametro dense intertextis, nubilatis. 1955] DODGE & RUDOLPH—LICHENOLOGICAL NOTES, I—IV 137 Orthidium sessile, basi constrictum, orbiculare, 0.6—1.5 mm. diametro, sub- ochraceum, margine 250—280 м crassitudine; conidiophorae tenues, unicellulares, 13—17 y longitudine; conidia singulatim disposita, acicularia, hyalina, 30-44 X 2 p, primum unicellulares, dende ad 7-septata, recta aut subcurvata. allus crustose, indeterminate, 0.7—1.2 mm. thick, buffy citrine, sorediose, К orange brown, C—; ecorticate; algal layer about 280 и thick, cells protococcoid, spherical to somewhat polyhedral from mutual pressure, 8.7—12.2 и in diameter; medulla thick, of closely woven hyphae about 1 м in diameter, somewhat nubilated with granules and including pieces of roots etc. from the substrate. Orthidium sessile, constricted at the base, circular, 0.6—1.5 mm. in diameter, ochraceous buff, disc very concave; margin 250—280 и thick, finally undulate; algae in a discontinuous layer on the outside of the layer of conidiophores, tending to die out above and forming a continuous layer below; the medulla around and between the algal colonies is formed of compactly woven hyphae about 1 p in diameter; conidiophores arising from the medullary hyphae, forming a continuous layer 13-17 р thick; conidia borne singly, acicular, 30—44 X 2 р, hyaline, non- septate until late, finally up to 7-septate, straight or slightly curved, slightly taper- ing at the ends. П. ADDITIONS TO THE LICHEN FLORA ОЕ MACQUARIE IsLAND Previous collections were reported by Dodge (1948). Most of the present collections are from the northern half of the island, collected mostly by Norman R. Laird and by N. M. Haysom of the A.N.A.R.E. Taylor (1954) has discussed the problem of distribution of the flowering plants which apparently have much wider ranges than the lichens. Twenty-seven species are reported from the present collection, three of which are new and three not previously reported bringing the total species of lichens to forty-four. Several other possible new species will be reported in a later number of these Lichenological Notes. In the following list MI/49/ has been omitted from N. M. Haysom’s numbers of collections. MICROTHELIA MACQUARIENSIS Dodge, B.A.N.Z.A.R.E. Rept. B. 7:48. 1948. Plateau, 800 ft., between Mt. Elder and Sandy Bay, N. M. Haysom 28. CoENOGONIUM SUBTORULOSUM Müll. Arg., Jour. Linn. Soc. Bot. 32:207. 1896. Orange red when fresh, drying olive buff. The Trentepoblia filaments have very few corticating hyphae. From glacial moraine above Sandy Bay, 500 ft., N. M. Haysom Z2. РБОКОМА VERSICOLOR (Hook. f. & Tayl.) Müll. Arg., Flora 71:538. 1888. Lecanora versicolor Hook. #. & Tayl., London Jour. Bot. 3:642. 1844, поп Ach. The squamules are less well developed than in previous collections but the apothecia agree microscopically. Norman R. Laird 2, 2a, 2c; Featherbed Terrace, growing over decaying hepatics and other vegetable debris, A.N.A.R.E. [Vor. 42 138 ANNALS OF THE MISSOURI BOTANICAL GARDEN PsEUDOCYPHELLARIA GLABRA (Hook. f. & Tayl.) Dodge, B. A.N.Z.A.R.E. Rept. B. 7:79. 1948 Stricta glabra Hook. f. & Tayl., London Jour. Bot. 3:647. 1844. South end of Plateau, on moss, N. M. Haysom 2114; north end of Plateau, on soil and over decaying mosses, Norman К. Laird; Featherbed Terrace, A.N. A.R.E.; Wireless НШ, М. M. Haysom Zr; north of Lusitania Вау, М. M. Haysom 7134. РЕТЛІСЕКА Гата! Dodge & Rudolph, п. sp. Type: Macquarie Island, growing on soil over decaying grasses, mosses and hepatics, Norman R. Laird 4, A.N.A.R.E Thallus foliosus, in herbario isabellinus, ad 10 cm. diametro, 400 y crassitudine, lobis sterilibus ca. 1.5 cm. latitudine, 2 cm. longitudine; tomentum crassum, verru- cosum in partibus junioribus; superficies inferior arachnoideus inter venas; venae sparsae, elevatae, arachnoideo-tomentosae, rhizinis simplicibus vel semel dichotome ramosis, са. 5 mm. longitudine, concoloribus; cortex ca. 90 и crassitudine, pseudo- parenchymaticus, cellulis 12-15 и diametro leptodermeis; stratum algarum 50-55 p crassitudine, cellulis nostocaceis, 5 и diametro; medulla 270 м crassitudine, hyphis periclinalibus pachydermeis, 4 и diametro, parte superiore compacta, inferiore laxi- ore, hyphis ad 8 и diametro, lumine 5 м Apothecia in lobis erectis non revolutis, 15 mm. altitudine, 5-6 mm. latitudine, sita; discus nigro-brunneus, 4 mm. diametro; sine amphithecio nec parathecio, stratum algarum sub hypothecio crassiore, ad 120 y crassitudine; hypothecium 30 м crassitudine, brunneum, superne obscurius, pseudoparenchymaticum, hyphis periclinalibus; thecium 120—130 y altitudine; paraphyses simplices, са. 2 м diametro; asci 100—105 Х 11-12 р, juventute apice incrassati; ascosporae octonae, hyalinae, aciculares, multiseptatae, ca. 56 X 3 y. Thallus foliose, becoming isabella color in the herbarium, up to 10 cm. in diameter, 400 » thick, margin lobed, sterile lobes about 1.5 cm. wide and 2 cm. long, somewhat lacerate on drying, ends of lobes with thick, verrucose tomentum above, becoming smooth but surface dull in the older portions as the tomentum weathers away; underside pinkish buff, arachnoid between the sparse veins; veins elevated, arachnoid-tomentose; rhizinae not abundant, simple or once-dichotomous near the tips, about 5 mm. long, concolorous; cortex about 90 м thick, pseudo- parenchymatous, cells rather thin-walled, 12—15 м in diameter; algal layer 50-55 p thick, of Nostoc colonies, cells about 5 и in diameter in an inconspicuous gel; medulla 270 & thick, of closely packed periclinal hyphae, thick-walled, about 4 p in diameter, less compact and more irregularly arranged below, forming the arach- noid underside with hyphae up to 8 џ in diameter, lumen 5 y. Apothecia on erect lobes, not revolute, 15 mm. tall, 5-6 mm. wide, disc very dark brown, 4 mm. in diameter; no amphithecium nor parathecium differentiated; algal layer thicker below the hypothecium, up to 120 м thick; hypothecium 30 р. thick, lower half pale brown, upper half very dark brown, pseudoparenchymatous from periclinal hyphae; thecium 120—130 м tall; paraphyses about 2p in diameter, unbranched, tips ending in the dark brown epithecial gel; asci 100-105 X 11-12 p, 19551 DODGE к RUDOLPH—LICHENOLOGICAL NOTES, I-IV 139 tips slightly thickened when young, 8-spored; ascospores hyaline, acicular, 56 Ж 3 m, thin-walled, many septate. This species somewhat resembles Ре рета dilacerata Gyelnik from Auckland, New Zealand, which has longer and narrower lobes with dilacerate margins and is about half as thick. LECIDEA SUBGLOBULATA Knight, Trans. Proc. М. Zeal. Inst. 8:314. 1875 [1876]. South slopes of Wireless Hill, A.N.A.R.E. New to Macquarie Island. CATILLARIA (ЕОСАТПЛ.АКІа) Rudolphi Dodge, п. sp. Type: Macquarie Island, without locality, but the rock types closely resemble those from the south slopes of Wireless Hill, A.N.A.R.E. Thallus crustosus, albidus, continuus aut ad centrum areolatus crassior, mar- gine irregulariter lobatus; cortex fastigiatus subdecompositus, erosus, ad 30 м crassitudine; stratum algarum 65 и crassitudine, cellulis protococcoideis 6-8 р diametro; medulla 250—300 p crassitudine, hyphis tenuibus, granulis brunneis nubilatis, dense contexta. Apothecia subimmersa aut sessilia, ad 2.5 mm. diametro, orbicularia, aut mutua pressione angularia, margine elevato, disco plano aut subconvexo nigro; para- thecium carbonaceum, in margine 50 м crassitudine ad 125-160 р sub thecio; hypothecium non bene evolutum; thecium 90—120 p altitudine; paraphyses tenues, septatae, super ascos ramosae, apicibus clavatis brunneis, ca. 1.5 и diametro; asci 50-60 X 7-8.5 и, cylindrico-clavati, leptodermei; ascosporae octonae, hyalinae, biloculares, uniseriales, ellipsoideae, 13-15 Х 6-7 allus whitish, sometimes stained ferruginous from iron in the underlying rock, thin, continuous at the margin, thicker toward the areolate center, K yellow then orange, margin irregularly lobate, distinct; cortex fastigiate, somewhat де- composed and eroded, up to about 30 y thick; algal layer 65 р thick, cells proto- coccoid, 6-8 и in diameter in a continuous dense layer; medulla 250-300 p thick, of densely woven slender hyphae, heavily nubilated with dark brown granules. Apothecia slightly immersed to sessile, up to 2.5 mm. in diameter, abundant, circular or angular from mutual pressure, margins elevated, disc plane to slightly convex, black; parathecium carbonaceous, 50 м thick at the margin, 125-160 p thick below the thecium; hypothecium not clearly differentiated; thecium 90-120 p tall; paraphyses slender, septate, branched at the level of the tips of the asci, tips slender, clavate, brownish, about 1.5 и in diameter, asci 50-60 X 7-8.5 p, cylindric- clavate, thin-walled, 8-spored; ascospores hyaline, bilocular, mostly uniseriate, ellipsoidal, 13-15 Х 6-7 p. The germinating ascospore produces an extensive black hypothallus. As con- tact is made with algal cells, thin assimilative areoles develop and finally fuse, covering the hypothallus as a continuous thallus which gradually thickens with age. Without locality but probably from south slopes of Wireless Hill, A.N.A.R.E., type; west side of Wireless Hill, N. M. Haysom Z152; south end of Plateau, N. M. [Vor. 42 140 ANNALS OF THE MISSOURI BOTANICAL GARDEN Haysom 2137; Plateau, 900 ft., young, A.N.A.R.E.; north end of Plateau, N. M. Haysom 286; without locality, Norman К. Гата 8. СТАРТА AGGREGATA (Sw.) Nyl., Bull. Soc. Linn. Normand. П, 4:167. 1870. Licben aggregatus Sw., Nova Gen. Sp. Pl. Ind. Occ. 147. 1788. West coast, fluviatile swamp, 50 ft., A.N.A.R.E. CLADONIA FLoRIFORMIS Dodge, B.A.N.Z.A.R.E. Rept. B. 7:134. 1948. Wireless Hill, 300 ft., Norman В. Гата 1d; Nuggets Creek, Norman R. Laird Ia, Ib, young; without locality, Norman R. Laird 1; only a few young plants among mosses, A.N.A.R.E. 5c Ciaponia Mawson! Dodge, B. A.N.Z.A.R.E. Rept. B. 7:128. 1948. Nuggets Valley, 100 ft., A.N.A.R.E.; ? primary thallus only, mouth of cave, south end of isthmus, A.N.A.R.E. i CLADONIA SARMENTOSA (Tayl.) Dodge, B.A.N.Z.A.R.E. Rept. B. 7:129. 1948. Cenomyce sarmentosa Tayl. in Hook. f. & Tayl., London Jour. Bot. 3:651. 1844. Interior of cave, 35 ft., A.N.A.R.E.; near Brothert Point, 250 ft., A.N.A.R.E.; near Nuggets Creek, Norman R. Laird Ic, Id, 2c; Lusitania Valley, east coast, 150 ft., A.N.A.R.E. STEREOCAULON ARGODES Nyl, Compt. Rend. Acad. Sci. [Paris] 83:88. 1876. Stereocaulon Argus Th. Fr., Nova Acta R. Soc. Sci. Upsal. III, 21:333. 1858; Hook. f. & Tayl., London Jour. Bot. 3:653. 1844, pro parte min. For discussion of nomenclature, see Dodge (1948, pp. 142—144) sub Argopsis megalospora. In the present collection the dimensions of the ascospores are some- what greater, 45-53 Ж5.5-7 р, than those given by Th. M. Fries and Nylander. On coastal vegetation north of Lusitania Bay, N. M. Haysom 2,128; Plateau, М. M. Haysom 71 STEREOCAULON PULVINARE Dodge, B.A.N.Z.A.R.E. Rept. B. 7:139. 1948. Near Nuggets Creek, Norman R. Laird. STEREOCAULON SUBMOLLESCENS Nyl., Compt. Rend. Acad. Sci. [Paris] 83:88. 76. Featherbed Terrace, A.N.A.R.E.; north end of Plateau, N. M. Haysom 285. PERTUSARIA TYLOPLACA Nyl. Compt. Rend. Acad. Sci. [Paris] 83:90. 1876. South end of Plateau, М. М. Haysom 2,122; Plateau, 900 ft., A.N.A.R.E.; north end of Plateau, М. M. Haysom 250, 762; growing over Menegazzia circum- sorediata Santesson on radio mast erected by A.A.E. in 1911 on Wireless Hill, A.N.A.R.E. PLacopsis PERRUGOSA (Nyl) Nyl., Lich. Nov. Zelandiae, 57. 1888. Lecanora perrugosa Nyl., Flora 48:338. 1865. For complete description, see I. M. Lamb, Lilloa 13:268—272. 1947. 19551 DODGE & RUDOLPH—LICHENOLOGICAL NOTES, I-IV 141 On rocky banks of Lusitania Creek, М. М. Haysom 2123. New to Macquarie Island. THAMNOLECANIA macquariensis Dodge & Rudolph, n. sp. Type: Macquarie Island, Wireless Hill, on cliff subject to water seepage, A.N.A.R.E. Thallus fruticosus erectus aut subdecumbens, pulvinos hemisphaericos ad 3 cm. diametro formans, di- aut tri-chotome ramosus, torulosus, cinnamomeo-alutaceus, inferne ad 0.8 mm. diametro, superne tenuior, ad 1 cm. altitudine; ecorticatus; algae flavo-virides, cellulis sphaericis vel mutua pressione polyhedricis, 8-11 р diametro; hyphae medullares са. 3 diametro, conglutinatae, inter cellulas algarum. Apothecia ad 1 mm. diametro in lateribus ramorum sessilia, plana dein convexa emarginataque, са. 250 y altitudine, disco nigro; amphithecium 250 p crassitudine; parathecium non bene distinctum, margine ad 40 p crassitudine; hypothecium obconicum, са. 190 p altitudine, hyphis verticalibus nigro-brunneis; thecium 135 р altitudine; paraphyses 2.5-3 р diametro, septatae, cellula terminali clavata 4 Х 5.5 и, pachydermea nigro-brunnea; asci ca. 65 Х 12 p, clavati, juventute apice incrassati; ascosporae octonae, subdistichae іп ascis 16 X 8 p, biloculares, brunneae, septo tenui constrictae, liberatae, 12 7X 6 р, nigro-brunneae. - Thallus fruticose, erect or subdecumbent, forming dense, depressed hemispheric cushions up to 3 cm. in diameter, branching di- or trichotomous, torulose, pinkis buff to cinnamon buff, up to 0.8 mm. in diameter below, more slender above, circular in cross-section, about 1 cm. tall; ecorticate; algae yellow-green, spherical or somewhat polyhedric from mutual pressure in the outer portion, 8—11 p in diameter with a thick sheath of periclinal medullary hyphae, more scattered throughout the medulla; hyphae about 3 in diameter, compact and conglutinate between the algal cells. Apothecia up to 1 mm. in diameter, sessile on the sides of the branches, not on the ultimate branches, flat at first becoming convex and emarginate, about 250 д tall; disc black; amphithecium about 250 y thick, similar in structure to that of the thallus but the medullary hyphae more vertical; parathecium not well differ- entiated, about 40 и thick at the margin, scarcely differentiated from the para- physes, continuous below with the dark brown obconic hypothecium about 190 р tall at the center, of vertical dark brown hyphae, not sharply differentiated from the thecium above; thecium 135 p tall; paraphyses 2.5-3 p in diameter, septate, terminal cell broadly clavate, about 4 Х 5.5 р, thick-walled, dark brown above; thin septum, shrinking to 12 X 6 p when free from the ascus and very dark brown. gest that the base may expand into a circular holdfast. The systematic position of this species is not clear. The bilocular brown spores with a thin septum, the para- physes and lecanorine apothecium suggest Rinodina sect. Beltraminia; the very dark hypothecium suggests Melanaspicilia Vainio. From both of these it differs in [Vor. 42 142 ANNALS OF THE MISSOURI BOTANICAL GARDEN its fruticose thallus. The structures of the thallus and of the apothecia are wholly different from those of the fruticose species of Anaptycbia, but do resemble those of Tbamnolecania from which it differs in a very dark brown hypothecium and broader brown bilocular spores. Its yellowish green algae suggest a relationship to Thelidea Hue, which is foliose with biatorine apothecia and hyaline ascospores. It is not a parasite as the medullary hyphae are continuous with those at the base of the hypothecium and the amphithecium is well developed; hence it cannot be considered a Karschia sp. on a Sipbula. PARMELIA TURGIDULA Bitter, Hedwigia 40:246. 1901. Featherbed Terrace, over mosses. A single sterile plant. New to Macquarie Island but previously known from New Zealand. | PARMELIA SUBLUGUBRIS Dodge, B.A.N.Z.A.R.E. Rept. B. 7:188. 1948. Featherbed Terrace, A.N.A.R.E.; Norman R. Laird 3; north end of Plateau, №. M. Haysom Zór. PARMELIA TENUIRIMA Hook. f. & Tayl, London Jour. Bot. 3:645. 1844. Featherbed fluviatile terrace on northwest slopes, growing over mosses, Norman R. Laird; from rock face, Half Moon Bay, west coast, A.N.A.R.E. MENEGAZZIA CIRCUMSOREDIATA Santesson, Ark. f. Bot. 30:11:14. 1942. Thallus 9 cm. in diameter. Wireless Hill, on radio mast erected by the A.A.E. in 1911, A.N.A.R.E.; coastal rocks north of Lusitania Вау, М. М. Haysom 2127. USNEA ARIDA У. MUsCICOLA Dodge, ВАМА КЕ. Rept. B. 7:207. 1948. Wireless Hill, on planks of A.A.E. radio hut, N. M. Haysom Z53. UsNEA СОМТЕХТА Motyka, Lich. Gen. Usnea Stud. Monog. 2:436. 1937. Featherbed Terrace, A.N.A.R.E.; north end of Plateau, М. М. Haysom 783. UsNEA ToruLosa (Müll Arg.) Zahlbr., Cat. Lich. Univ. 6:594. 1930. Usnea dasypogoides f. torulosa Müll. Arg., Flora 66:19. 1883. Our plants are only 3 cm. tall with shorter lower internodes. North of Lusi- tania Вау, М. M. Haysom 7130. UsNEA ХАМТНОРОСА Nyl., Compt. Rend. Acad. Sci. [Paris] 83:89. 1876. Among mosses, Norman R, Laird 5a, 5b, 5c, 54, бе; Featherbed Terrace, very young, A.N.A.R.E. RAMALINA GENICULATA Hook. f. & Tayl., London Jour. Bot. 3:665. 1844. Turf from rock surface, west coast, 50 ft., A.N.A.R.E.; Featherbed Terrace, A.N.A.R.E. GASPARRINIA MACQUARIENSIS Dodge, B.A.N.Z.A.R.E. Rept. B. 7:234. 1948. One plant has much larger apothecia, up to 4 mm. in diameter, with flat discs and nearly excluded margins but agrees microscopically. Hasselborough Bay, isthmus beach, Norman R. Laird 8a; beach of Garden Cove, М. M. Haysom 711, Z223; Green Gorge, N. M. Haysom 2102. 19551 DODGE & RUDOLPH——LICHENOLOGICAL NOTES, I-IV 143 ВоЕмлА Mawson Dodge, B.A.N.Z.A.R.E. Rept. B. 7:243. 1948. South end of Plateau, N. M. Haysom 2,137. Ш. ЕТОКА or ADELIE LAND А small collection was received from Cape Margerie, 66° 50' S., 141? 20' E., collected on November 4, 1950, the first specimens taken from Adélie Land. Since the collection is small and Adélie Land lies between King George V Land and Queen Mary Land, it is not surprising that all these species have been reported before. UMBILICARIA HUNTERI Dodge, B. A.N.Z.A.R.E. Rept. В. 7:148. 1948. A.N.A.R.E. 4. Previously known from King George Land. ALECTORIA CONGESTA (Zahlbr.) Dodge, B. A.N.Z.A.R.E. Rept. B. 7:195. 1948. Parmelia pubescens v. congesta Zahlbr., Deutsche Südpolar Exp. 8:52. 1906. A.N.A.R.E. 3. Previously known from King George V Land to MacRobertson Land. USNEA ANTARCTICA DuRietz, Svensk Bot. Tidskr. 20:90, 93. 1926, A.N.A.R.E. 6. Previously known from Marie Byrd Land to King George V Land. UsNEA PUSTULATA Dodge, B.A.N.Z.A.R.E. Rept. B. 7:203. 1948. A.N.A.R.E. 5а. Previously known from King George V Land and Queen Mary Land. USNEA SCABRIDULA (Lamb) Dodge, B.A.N.Z.A.R.E. Rept. B. 7:204. 1948. Neuropogon acromelanus у. inactivus f. scabridulus Lamb, Jour. Linn. Soc. Bot. 52:2202 3939. A.N.A.R.E. 5. Previously known from South Victoria Land to King George V Land. XawTHORIA Mawson! Dodge, B.A.N.Z.A.R.E. Rept. В. 7:236. 1948. A.N.A.R.E. т. Previously known from King George V Land to MacRobertson Land. RiNoDINA FRIGIDA (Darb.) Dodge, B.A.N.Z.A.R.E. Rept. B. 7:259. 1948. Buellia frigida Darb., Brit. Nat. Antarct. "Discovery" Exp. Nat. Hist. 5:Lich.:7. 1910. A.N.A.R.E. 3. Previously known from Marie Byrd Land to MacRobertson Land. IV. AppITIONs то THE FLorA OF МАСКовектзом LAND Previous collections from Cape Bruce were reported by Dodge (1948). The present collections are from the A.N.A.R.E. weather station at Mawson, about 67? S. and 50? E., collected by the Medical Officer, Dr. A. M. Gwynn. Fourteen species are represented, one new species, six others new to MacRobertson Land, [Уог. 42 144 ANNALS OF THE MISSOURI BOTANICAL GARDEN making twenty-four species so far known. The Verrucariaceae are still absent from collections. НЕРРІА ANTARCTICA Dodge, B.A.N.Z.A.R.E. Rept. В. 7:71. 1948. Small sterile thalli along with several genera of Myxophyceae on a weathered crystalline rock with Parmelia Johnstoni Dodge, A. M. Gwynn 1:23. ГЕСШЕА Harrissoni Dodge, B.A.N.Z.A.R.E. Rept. B. 7:101. 1948. On rock with Alectoria congesta (Zahlbr.) Dodge, A. M. Gwynn 1220. UMBILICARIA SPONGIOSA v. SUBVIRGINIS (Frey & Lamb) Dodge, B.A.N.Z.A.R.E. Rept. B. 7:148. 1948. Umbilicaria antarctica v. subvirginis Frey & Lamb, Trans. Brit. Myc. Soc. 252475. 1939, Thallus about 7 cm. in diameter, upper surface pale smoke-gray or lighter, very minutely rimulose. А. M. Gwynn 1230. New to MacRobertson Land. UMBILICARIA SUBCEREBRIFORMIS Dodge, B.A.N.Z.A.R.E. Rept. B. 7:149. 1948. A. M. Gwynn Li30. New to MacRobertson Land. ACAROSPORA (РАСНМОГЕРА) Gwynni Dodge & Rudolph, n. sp. PL 15, fig, 2: Type: MacRobertson Land, Mawson, on weathered reddish granite? A. M. Gwynn 1227. Thallus cerebriformis subfruticulosus, juventute citrinus dein magis olivaceo- viridis, ad 3 mm. diametro, substipitatus, marginibus liberis, са. 500 м crassitudine, inferne alutaceus; cortex superne 20-25 y inferne ad 30 y crassitudine, fastigiatus, pseudoparenchymaticus, cellulis pachydermeis, luminibus 2.5 и diametro, strato amorpho gelifacto 15 и crassitudine tectus; algae protococcoideae, cellulis 13-14 p diametro; medulla non distincte evoluta, hyphis 2 м diametro. pothecia ca. 0.5 mm. diametro, orbicularia vel mutua pressione elliptica, im- mersa, disco obscuro; parathecium hyalinum, 15 y crassitudine in lateribus thecii, subtus 25-30 y crassitudine, hyphis periclinalibus; hypothecium 15 y crassitudine, hyphis tenuibus grumosis dense contextum; thecium 230 y altitudine; paraphyses cohaerentes, distinctae, pachydermeae, 2 р diametro; asci 100 X 20 p, juventute apice incrassati; ascosporae multae, ellipsoideae, ca. 3.5 X 1.8 и (liberae non visae). Spermogonia immersa, 200 и diametro; perifulerum hyalinum, 15 p crassi- tudine, hyphis pachydermeis, periclinalibus conglutinatis; fulcrum in cavitate in- vaginans; spermatiophorae 12-15 р longitudine, subramosae (modo Cladoniae) ; spermatia hyalina, anguste ellipsoideae, 3 X 1 y. Thallus cerebriform, subfruticulose, lemon-yellow in younger portions to warbler green when the apothecia are well developed, covering areas up to 2 cm. in diameter, individual thalli up to 3 mm. in diameter, attached at the center, margins free, about 500 и thick, upper surface smooth to deeply furrowed and cerebriform, under surface warm-buff, K—; cortex 20-25 и above to 30 y thick below, completely surrounding the thallus, fastigiate, of thick-walled pseudo- 1955] DODGE & RUDOLPH—LICHENOLOGICAL NOTES, I-IV 145 parenchyma, cell lumina 2.5 шіп diameter, densely nubilated with minute yellowish crystals, especially above, covered by an amorphous layer of gel about 15 y thick, with scattered granules; algal layer filling the thallus between the cortices, cells protococcoid, 13-14 р in diameter, often polyhedral from mutual pressure; medulla not differentiated, hyphae about 2 м in diameter between the algal cells Apothecia very numerous in the central thalli, about 0.5 mm. in diameter, circular to somewhat elliptic from mutual pressure, immersed, disc greenish black; amphithecium not differentiated from the thallus, but often with a circular furrow about 0.2 mm. from the parathecium; parathecium hyaline, of periclinal hyphae about 15 р thick at the sides of the thecium to 25-30 y thick below; hypothecium about 15 д thick, of grumose, densely woven, slender hyphae; thecium 230 y tall, the upper 25 д brownish and covered by a hyaline amorphous gel about 13 и thick; paraphyses coherent, distinct, thick-walled, about 2 м in diameter; asci about 100 Х 20 р, tips thickened when young; ascospores numerous, ellipsoid, about 3.5 X 1.8 в (not seen free from ascus). S onia immersed, about 200 д in diameter, wall hyaline, 15 м thick, of већа, conglutinate, periclinal hyphae; layer of spermatiophores invaginated, forming cerebriform cavities, spermatiophores 12-15 р long, somewhat branched (Cladonia туре); spermatia hyaline, narrow ellipsoid, about 3 X 1 p On weathered reddish granite, А. M. Gwynn Li2r, 1427. LECANORA GRISEOMARGINATA Dodge & Baker, Ann. Mo. Bot. Gard. 25:572. 1958. Growing on moss cushion, А. М. Gwynn Li20. New to MacRobertson Land. LECANORA EXSULANS (Th. Fr.) Dodge & Baker, Ann. Mo. Bot. Gard. 25:570. 1938. Lecanora chrysoleuca v. melanophthalma £. exsulans Th. Fr., Nyt Mag. Natur- vidensk. 40:208. 1902. А. М. Gwynn Li21, Li24, 1238, 1431 (a single young plant not sectioned). PARMELIA JOHNSTONI Dodge, B.A.N.Z.A.R.E. Rept. B. 7:191. 1948. Growing over weathered rocks and mosses, А. М. Gwynn 1221, Li22, 1223, 1225, Li28, 1237. ALECTORIA CONGESTA (2аћ Бг.) Dodge, B.A.N.Z.A.R.E. Rept. В. 7:195. 1948. Parmelia pubescens v. о Zahlbr., Deutsche Südpolar Exp. 8:52. 1906. On rock, A. M. Gwynn PROTOPLASTENIA CITRINA Dodge, B.A.N.Z.A.R.E. Rept. B. 7:222. 1948. The color is more orange than in the type and all the material is sterile, although the thallus agrees microscopically. Growing over moss cushions, А. M. Gwynn 1220, Li21, Li25. [Vor. 42 146 ANNALS OF THE MISSOURI BOTANICAL GARDEN GasPARRINIA Harrissoni Dodge, B.A.N.Z.A.R.E. Rept. B. 7:235. 1948. Thallus about 5 cm. long and 2 cm. wide, apparently starting near the edge of the rock fragment, so probably capable of forming a circular thallus about 5 cm. in diameter when the substrate permits. А. M. Gwynn Li26, [431, a smaller thallus. XANTHORIA MawsoNi Dodge, B.A.N.Z.A.R.E. Rept. В. 7:236. 1948. А. M. Gwynn Li20, Li28, Li32. BUELLIA JOHNSTONI Dodge, B. A.N.Z.A.R.E. Rept. B. 7:249. 1948. A. M. Gwynn 1428, associated with Lecanora exsulans (Th.Fr.) Dodge & Baker, as in the type. New to MacRobertson Land. Rinopina FRIGIDA (Darb.) Dodge, B.A.N.Z.A.R.E. Rept. B. 7:259. 1948. Buellia Hue Darb., Brit. Nat. Antarct. "Discovery" Exp., Nat. Hist. 5:Lich.:7. 1 . M. Leod 1421, with lecanorine apothecia nearly immersed; Li27, 1433, bs older thallus up to 6 cm. in diameter. Мовтос sp. Several thalli growing on a moss cushion, akinetes abundant. A few young perithecia or cleistothecia were seen which may be parasites, or perhaps the thalli are those of а Pyrenidiaceous lichen. Very few fungus hyphae were seen in the gel between the algal filaments but in one or two places there is a faint suggestion of а cortex. Some thalli contained a few bacterial colonies. We hope it will be foun again in a more mature condition, if it be a lichen. А. M. Gwynn Li20, along with Lecanora griseomarginata Dodge & Baker and sterile Protoblastenia citrina Dodge. BIBLIOGR APHY Australian National Antarctic Research Expedition eed [Map of] Mee Island from survey by Aust desi ре Antarctic Research Expedition, 1948, scale 1:8 Cengia ыы М (193 m мемен Africae Oran. ations: Moore: ва. Bot. Ital. N.S. HAM г; з —— ——, (19 Became пе Ш. Ibid. 48:495—516. [see p. 500—501]. pee Carrol Y. (1948). Lichens and s мае areas Australian New Zealand Antarctic Sas (B.A.N.Z.A.R.E.) R В. 7:1-27 x su J: e lichens of й PUES ке Мо. Вог. Gard. 40:271—401. [see Mae Keissler, к. v (927). аат Untersuchungen über die Flechtenparasiten und lichenoide Pilze. V. ‚ Mus. Wien 41:157-170. [see р. 157-163 меде, с. и. Das sa pylidium, ein verkanntes Fortpliuixongsorgat der Flechten. Svensk Tidskr. 29:3 мама Calvino, Eva (1930). асы su una forma singolare di Deuterolicheni: Cblorocypbella Speg. Nuovo Giorn. Bot. Ital. N.S. 37:369—379. pi. Maller “Argh, J. (1881). ма чин га Beitráge. XII (Schluss). Ms 64:100—112. [see p. 111- — — ——, (1890). Lichenologische Beiträge. XXXIII. Ibid. 73:187-202. [see р. 202]. Rizzini с. Toledo (1952). Lichenes in Horto Botanico Fluminis Januarii crescentes. Г. Ага. Jard. ot. Rio Scholes, Arthur (1951). Fourteen men, м» rà of the Australian Antarctic Expedition to Heard Island. London кы 5 р. i. 273 pp I0 N. 1955] DODGE & RUDOLPH—LICHENOLOGICAL NOTES, I-IV 147 Spegazzini, С. (1909). Mycetes Argentinenses. Series ТУ. Ап. Mus. Nac. Buenos Aires 19:257—458. g. [see 279 Taylor, B. W. (1954). An xample of long distance icc Ecology 35:569— Vainio, E. A. (1890). Bendi sur la рен naturelle et la morphologie Ра cre du Brésil. [Thesis Univ. Helsingfors]. Acta Soc. Fau a Fl. Pei. fa 1-247; Pars II, 1-256. [see part 2 271. 148 EXPLANATION OF PLATE PLATE 15 Fig. 1. Epbelidium heardensis Dodge & Rudolph. A. Cross-section of thallus. B. Cross-section of orthidium. C. Conidiophores. D. Conidiospores. Fig. 2. Acrospora Gwynni een & ure A. Cross-section of a B. ad -section of ен shown C. Asc D. Specritiophores, g spermogonium. Fig. 3. Collemopsidium pyrenuloides Dodge & Rudolph. А. Ascospores. B. Algal colonies. ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vor. 42, 1955] ANN. Мо. Bor. Gard., Vor. 42, 1955 PLATE 15 DODGE & RUDOLPH—LICHENOLOGICAL NOTES, I-IV ANN. Мо. Вот. Garb., Vor. 42, 1955 PLATE 16 2 DODGE & RUDOLPH—LICHENOLOGICAL NOTES, I-IV (Vor. 42, 1955] DODGE & RUDOLPH—LICHENOLOGICAL NOTES, I-IV 149 PLATE 16 Fig. 1. Heard Island, Laurens Peninsula. Fig. 2. Heard Island, Coast of Atlas Cove. (Photographs taken by the Australian National Antarctic Expedition.) THREE NEW АММОМАСЕАЕ FROM PANAMA! ROBERT E. FRIES ANAXAGOREA Allenii R. E. Fr. n. sp.—Arbor сис. 6 m. alta; ramuli novelli brevissime rufo-tomentelli. Foliorum petiolus 10-12 mm. longus, crassus; lamina papyracea, (exsiccata) flavo-viridis, supra glaberrima laevisque, subtus demum glaber et pallidior, oblongo-lanceolata, basi rotundata, apicem versus sensim long- eque angustata, 30—35 cm. longa et 8-10 cm. lata; costa supra subplana, subtus valde prominens, teres, rugosus; nervi laterales I. utrinque circ. 15, sub angulo circ. 80? exeuntes, deinde sursum curvati, sicut venulae subtus paulo elevatae. Pedicelli florum breves (3—5 mm. longi), crassi, sicut sepala et petala rufo-tomentelli, juxta calycem bractea vaginaeformi instructi. Sepala ovato-orbicularia, apice rotundata, non recurvata, 7—10 mm. longa. Petala crassa, exteriora oblonga, apicem rotund- atum versus haud angustata intus apicem versus carinata, ad 12 mm. longa et 5 mm. lata; petala interiora lanceolata, acutissima, ad mediam partem pilosa et supra medium intus carinata. Stamina linearia, 3 mm. longa; appendix connectivi tenuis, circ. 0, 7 mm. longa, apice rotundato-truncata. (Fructus ignotus). CANAL ZONE: Quebrada López, alt. 30 m., in flower, Feb. 11, 1940, P. H. Allen 2143 (Type in Herb. Missouri Bot. Garden). This species is very like the Anaxagorea crassipetala Hemsl., found in Guate- mala and Nicaragua, but differs from this especially in its leaves being larger, rounded at the base, and with much-elongated, almost thread-like apices; in its pedicels being short and thick, in its sepals being longer and of rounded oval, not triangular-shape, and its outer petals longer and not tapering toward their apices. it is also similar to Anaxagorea dolichocarpa Sprague and Sandwith, from which it differs by sepals rounded at the tips and not retroflexed, short inner petals, shorter stamina, etc. Unonopsis panamensis В. E. Fr. n. sp.—Ramuli novelli pilis brevibus fer- rugineis erecto-adpressis vestiti. Foliorum petiolus teres vel supra planus, tomen- tellus, 4—5 mm. longus; lamina chartacea, exsiccata flavo-virens, supra laevis et primo in costa minute hirsuta, demum omnino glabra, subtus in costa hirsuta glabrescens, oblongo-oblanceolata, supra medium latissima, basi breviter acuta, apice abrupte contracta et 1-1.5 cm. longe cuspidata, 15—20 cm. longa et 4—5 cm. lata; costa utrinque elevata; nervi validiores utrinque circ. 15, arcuatim adscen- dentes et 1—2 mm. a margine conjuncti, tenues et supra elevati; reticulum venularum in foliis adultis haud conspicuum. Inflorescentiae ех axillis foliorum delapsorum evolutae; pedicelli florum crassi, ferrugineo-sericei, breves (7 mm. longi vel ultra?), paulo supra basin articulati. Sepala inter se libera, rotundato- ovata, acutiuscula, circ. 2 mm. longa. Petala rigida, exsiccata nigrescentia, ex- teriora rotundata et extra ferrugineo-sericea, circ. 7 mm. in diam., interiora minora, secundum nervum medium ferrugineo-tomentosa. Stamina 1.8—2 mm. longa, 1Jssued June 23, 1955. (151) [Vor. 42 152 ANNALS OF THE MISSOURI BOTANICAL GARDEN apice truncata. Ovaria 1, 5 mm. longa, prismatica et breviter strigosa, stigmate globoso-ovoideo coronata; ovula 3—4, lateralia. (Fructus ignotus). PANAMÁ: Cerro Campana, trail from Campana to Chica, alt. 600-800 m., with flowers Aug. 10, 1941, Allen 2647 (Туре in Herb. Missouri Bot. Garden). The material is unfortunately scanty, with no fruits at all. The number and lateral positions of the ovules, however, prove it to be well differentiated from the Central American species U. РИНет Saff. and Schippii R.E.Fr. Flowers of the third and last Central American species, costaricensis R.E.Fr. (Acta Horti Bergiani 13:105. 1941) are unknown so far, but its fruit structure with one single oblique to horizontal seed favours the view that, structurally, the flower agrees with that of Pittieri-Schippii, and accordingly differs from panamensis. Unonopsis panamensis also differs from costaricensis in having much larger and differently shaped leaves. In my survey of the species of this genus (Acta Horti Bergiani 12:237) panamensis should be ranged among the guatteriodes, obovata, and allied species, but cannot be joined to any of these. GUATTERIA INUNCTA var. caudata В.Е. Fr. nov. var.—Differt a typo foliis pro rata angustioribus, basi longe cuneato-angustatis, apice acumine 2-3.5 cm. longo caudiformi acutissimo ornatis, pedicellis florum tenuioribus. Bocas DEL Toro, vicinity of Chiriqui Lagoon, Old Bank Island, von Wedel 2108 (Type in Herb. Missouri Bot. Garden). THE BOTANICAL CATALOGUES OF AUGUSTE DE ST. HILAIRE JOHN D. DWYER* The recent essay of Jenkins! serving as an introduction to Auguste de St. Hilaire's Esquisse de mes voyages au Brésil et Paraguay focused considerable atten- tion on this French naturalist who collected plants and animals in southeastern Brazil during 1816—1822. Few of the 30,000 plant specimens collected by St. Hilaire? have found their way into other botanical institutions from the herbaria of the Muséum National d'Histoire Naturelle in Paris or the Université de Mont- pellier”. This has contributed to the failure of many modern monographers of Brazilian plants to cite St. Hilaire material. While studying at the Muséum National d'Histoire Naturelle in Paris I came Ue the ten handwritten books constituting the catalogues of plants collected by t. Hilaire*. It is my hope to demonstrate the importance of these catalogues and E to rectify some errors made in the handling of the St. Hilaire collections. I wish to express my thanks to Prof. H. Humbert, Directeur Laboratoire de Phanérogamie, Muséum National d'Histoire Naturelle, Paris, France, for permission to examine and to have microfilmed the catalogues of Auguste de St. Hilaire, as well as to use the facilities of the Herbarium. I owe especial gratitude to Monsieur J. Leandri, Sous-Directeur of the same institution, for his substantial assistance. То Mr. George Swett, Examiner of Questioned Documents, Clayton, Mo., I extend my heartfelt appreciation for his comparison of handwritings found in and related to the catalogues. ST. HILAIRE IN BRAZIL St. Hilaire set foot on Brazilian soil at Rio de Janeiro on June 1, 1816. Не was destined to be the first French naturalist to penetrate into the interior of south- eastern Brazil. Rio de Janeiro, the infant capital of Portuguese Brazil, was from the time of Vellozo in 1780 the hub of botanical activity in South America. Three years before the arrival of St. Hilaire, Georg von Langsdorff, а German, found the capital an ideal center for his botanical work. Two English collectors who botanized in Rio de Janeiro and in Sáo Paulo, Brazil, James Bowie and Alan у Јев ins, А. Е Chron. Bot. 101:5-21. Rapport sur le idi e de M. Auguste ме узи Hilaire -— le Brésil et les Missions du Para, гаку, lu à l'Académie Royale des Sciences, die Jussieu Rapport p. Paris, 1823. This rare reprint may be found in the library of the Missouri Botanica Garden, go Г ouis, Missouri, as well as the Oliveira Lima Library, C atholic University, > n America, for вічну TA 5 а “ұны than 100 sheets of St. Hilaire material deposited in the ый ем Е п (U.S. ten Henn Washington, D. C., and less than 40 sheets in the Miss кучаву, Gen St. Lou 41 wi tude to she е Science — Washington, D. C., for the e VIIIth Botanical Congress held cata t gues, despite the fact that it may be more accurate to designate the ten books collectively as а cat ime; e. * Missouri Botanical Garden and St. Louis University. (153) [Vor. 42 154 ANNALS OF THE MISSOURI BOTANICAL GARDEN unningham?, in their unpublished (Brazilian) diary speak of visiting "Mr. Langs- ped garden” on January 10, 1815. Four fellow-countrymen of Langsdorff had already penetrated into the interior of southeastern Brazil: Sellow, von Chamisso, Eschscholtz, and Maximilian, Prinz zu Wied-Neuwied. The most illus- trious German systematist of this era, von Martius, was soon to arrive (July, 1817). Urban’s account of St. Hilaire’s five "voyages" informs us that St. Hilaire met two of these botanists, Langsdorff and Sellow®. One of the most concise and accurate summaries of the travels and activities of St. Hilaire during his five journeys into the interior from Rio de Janeiro in search of plants, animals, minerals, etc. is to be found in the report of St. Hilaire's travels presented to the French Academy by Baron Cuvier and five other members. This represents the acceptance by the Academy of St. Hilaire's Aperçu d'un voyage dans l'intérieur du Brésil of the previous year’. I have taken the liberty of trans- lating a section of this "summary" from the French. Wherever possible, the spelling of geographical names has been changed to that of modern usage. fter speaking of St. Hilaire's arrival in Brazil we read: In the following December he pr көне northwards, іп the province of Minas [Gerais], a territory as vast as France, where he sojourned for 15 months in order to explore its various arts. At first he headed іп = бика n of Villa Rica, penetrating into old and very ive covering in t country w ifferen i i n tribes. He noted i e “district” of Minas Novas a change in he flora, with woods composed of only small shrubs. ` are зр Botocudos, warlike Indians who live in the forests, naked and epi definite ае иу з я о: е 1 banks and entered the District of the Diam ds [Diama ұры; and visited the areas where he had previously noted their methods of Mica ng gold from the mountains and rivers. Returning to Villa Rica by a long detour, = again headed ot - Rio (de) Janeiro arriving in March of 1818, whence he sent to the Muséum d'Histoire Naturelle a слом of 200 birds, some quadrupeds, 800 species of insects, (C two hundred packets P as well as two botanical papers which have been printed in the publication of the Muséu A second voyage of several months duration brought him ope reds districts, to the north of Rio de Janeiro, as far as the Rio Doce, an yiel we im about the same materials adj edly е ones found on the sandy ерни ses. о, de town of Ж “Salvador pos, the entire province of Espi о a ere unhealthy banks of the Rio Doce, a ео embarked at Villa da Мема, he | returned by sea to Rio de Janeiro, whence he а а new shipment of animals and of "e o the — His third => the longest and most importan t, begun i in January of 1819, was at first toward the northwest in the eastern erii of the province of Minas [Gerais] where he saw a beautiful waterf i i Araxá 520 the whose waters cattle drink unhesitatingly. То arrive at = he followed а barren plateau 5Bowie and Cunningham were to depart from € de Janeiro in October of 1816, the former t collect at the Cape of Good Hi Hore. the latter in Australia, Their unpublished Brazilian journal is i the R oyal Botanic | Gardens, Kew, England. Ud. in Martius' Flora Brasiliensis 11:93—98. M Hilaire, / Auguste de, in Mém. Mus. d'Hist. Nat. > на 1822. 19551 DWYER—BOTANICAL CATALOGUES OF ST. HILAIRE 155 whose opposite sides form the sources of the Sao Francisco river and the Dos Tocantins both duraring the sea from different directions. The v vegetation str ruck him as being little different at seen on the га ъз of the Sio Francisco. The pastures were at times entirely к rren, at times covered w — deg trees. M. de St-Hilaire after having visited Villa Bóa, capital of the pos of G pent a few шу among the Coyapos Indians, then went to the се f Matto Сон: from i ен he t outheastern route to Si0-Paul. Не arrived there cep rete 1819, afte long and ar rin s journey marked by great heat and dryness, and here he deposited bison various collections, —— with кергі to permit him uthw adi. bh iid the western ом of the vast — of mountains which separate the the interior б the country from eashore, and saw in succession the following villages Sorocaba, Itu, Pórto Feliz and their на Һе Рыа the boundaries of the most impo tent colonial industries. Then he entered Campos Gerais, such delightful country because of its diversified countryside. e made a goodly collection of plan ich d a He ma temperate climate. There he observed the erroneously named herbe du Paraguay; this he co нач си аѕ а ин of the genus Ilex. Не also identified the local quinquina which «з. over а chain of mountains at a difficult spot called Serra die балак, Һе found но» e sea coast at а much lower altitude and hotter than the interior plateau big crops ані coin more those of the tropics even as far south as 27? latitude. He was able to visit the isla St. Francis and of Catherine and their whaling-posts. ii ihe ground beyond is a sandy and dry stretch enlivened only by the presence of a prodigious number of aquatic birds. e province of Rio Grande, which he soon entered, is very — nap a more vi igorous population. Не observed that the culture of sugar-cane stops a ut while cotton- growing extends to 51, palms to 347, just at noted in New-Hol us [Australia]; He spent the winter, already severe in these areas, at Pôrto Alegre and at Rio е, rting in October 1820, that is in the springtime of this country. The plants which flourish at this ime belong to the same genera as those constituting the spring ens i in Europe, such as the Е eeds enaria е up to а degree north of Pórto is a ей ose their gras at Rio Grande, and at degrees fertnr dri scarcely a tenth have retained their leaves. t. Hilaire then = the Spanis nish possessions, p Lodi through several towns and ul estates іе the banks of the Вю de la Plata, pes at Montevideo where the i of Euro ith zzara. m ficult between Belem and the provin the Missions, and he spent thirteen days in uninhabited desert, populated “n =н еб Гена), сей [deer], and jaguars, against which he had to defend his own life, and especially that А horses; утур ве y | tuted in e past of Paragua | : м г еді be groan at finding only ruins = ead sight of such misery in a fertile land whose former splendor our writers have not еха Then re-entering Brazil proper and traveling in ihe Па deni n, he frequently ran n pod of losing his notes and collections; he returned at the end of a "year by a detour to Pórto Alegre, embarked on Lake Patos for Rio Grande; returned by sea to Rio Janeiro, where he cid toward the end of 1821, after an absence of nearly iie years St. Hilaire has elaborated on his travels in his own Esquisse (loc. cit.) and his Voyages. These are in need of careful study by contemporary scholars. The recent remarks of the Brazilian P. R. Reitz (Anais Botanicos do Herb. Barbosa Rodrigues 1:86. 1949) on St. Hilaire reflect the warm affection of the Brazilians for the French naturalist. I have taken the liberty of translating freely the same from the Portuguese: [Vor. 42 156 ANNALS OF THE MISSOURI BOTANICAL GARDEN He had a kind heart, adapting himself to the environment in which he lived and like- wise to the а people with whom he often tr: pre раю with the manners ої a gentleman those who helped him, ME: amiably th aka treated him badly. He described ener the customs of our lands. St. " Hilaire stands in marked contrast to those pseudo-investigators who, as навече describe our country as a grand carnival from the seats of gambling hall ‹ or even from Corcovado? where they sit to write their oes ations, their memoirs of the га вечи of the Brazilian way of life and customs Only . Hilaire, а ME or a Bonpland who razil with great sacrifice, who traveled poem inhospitable ке к sense те » m of ps Brazilian чар ра ње in the cities but in the back c and small is ot could чач in their habitats great diversity of plants, animals, ирк pe minerals. O these or others of their alibre could de: Sake. in the role of тен алы, our natural caches and our p ople. There is an excellent map of the itinerary of St. Hilaire's five "voyages", pre- pared by J. Santos for Chronica Botanica (loc. cit. p. 12), from St. Hilaire’s original in the back of his Voyage à Rio Grande do Sul. А large map found in the first yolume of Martius’ Flora Brasiliensis labelled “Itinera Botanicum" with inked- in itineraries of the majority of the principal botanists who collected in Brazil from 1780 to 1852 is unfortunately inaccurate, at least with respect to St. Hilaire. For example, the delineator indicates that St. Hilaire collected as far north on the eastern coast of Brazil as Villa de Porto Seguro (Bahia); this is almost 225 miles - north of the Rio Doce, the most northeastern limits of St. Hilaire's collections. THE CATALOGUES OF ST. HILAIRE The catalogues are made up of separate books in St. Hilaire's own hand- writing. Approximately 8900 collections are entered in the 1575 pages?. My count from the catalogues stands in marked contrast to Urban's statement (Мағ из” Flora Brasiliensis 11:92—98. 1906) of the number deposited in three herbaria, that: "Collectio princeps (cr. 7600 numeri) in herbario musei historiae naturalis Parisiensis. Dupla in Montpellier, parca in museo Berolinensi". Urban’s estimate may have been the result of adding up the collection numbers in the catalogues without considering that often single collection numbers were subdivided one to several times, or from St. Hilaire's own statement (Esquisse, p. 61) that: "Le nombre des plantes que j'ai recueillies s'éléve à environ sept mille. . ..". Int catalogues there are approximately 835 collection numbers which are subdivided а la “Б”, “ter”, "4", etc. In one instance (in series Сә) the collection num- ber 2701 is subdivided thirteen times and applied to eleven different families of angiosperms. The failure on the part of monographers or copyists to add "bis", “ter”, etc, to a specific collection number has caused considerable confusion in the citing of specimens. Following each collection number is (usually) the specific plant family (written in French) to which the collection belongs. Occasionally only the generic name is given. For approximately 550 collection numbers St. Hilaire fails to supply any taxa although usually one finds an "analysis" or a few words. In more than one ЗА high peak overlooking Rio de те easily reached in our times Ах street-car. 9In a letter ги received, M. Leandri writes: "Our catalogue of entries bears actually 7692 gatherings [collections] from St. ud maybe the others have Бега ees or dede: by A. de Jussieu and Cambessedes and others) ." 1955) DWYER—BOTANICAL CATALOGUES OF ST. HILAIRE 157 third of the entries a detailed description ("analysis") of the plant is given in French. In some families, for example in the Orchidaceae, these descriptions may take up three full pages. St. Hilaire paid particular attention to listing the com- mon names of the plants as well as their economic uses. The place of collection is usually found at the end of the "analysis".!? The "analyses" are models of pre- cision and critical observation. St. Hilaire himself states (Esquisse, p. 61) «..... је les ai toutes analysées sur les lieux-mémes, et me suis principalement attaché à la dissection des parties dont la connoissance répand le plus de lumiéres sur les rapports naturels". Dreuzy, a great-nephew of St. Hilaire, in an appendix to the Voyage à Rio Grande do Sul, refers to the difficulties encountered by Saint Hilaire in preparing his "journal botanique", the name by which Dreuzy apparently designates St. Hilaire's diary and his catalogues collectively: donnons ci-dessous quelques extraits de ce Hager сати iv pour faire comprende au Nue eur quel travail s'imposait Phair au soir de chacune de ses fatigantes journées de voyage, assis sur ses malles sous ncho, éclairé par я Metis бота а d'une chandelle fumeuse, dévoré par les с - attirait la lumiér Dreuzy then supplies two samples of St. Hilaire's "analyses" from catalogues C2, Section 2, one collection number 2682, and the other 2787”. То my knowl- edge these are the only "analyses" which have been published in toto. In fact, I have found no reference of any length in the literature concerning: St. Hilaire’s catalogues. We may use Dreuzy’s sample 2682 as typical of an “analysis” of St. Hilaire: EsTANCIA DE SANTO-REI, 1%" mars, n° 2682. Graminée. те chaque dent де Гері son 2 m гаш sessile, l'autre ређе. Épillets sessiles, 1 fl. her арт ipa sous- сот glumes 2-valves, à valves ext. hérisée, lancéolée, tantót e ac ta = courtement è e Pau plus _ Glu elles 2-valv., à va у. s ees sous-violettes, sous-pubescentes => 4 как ем: uscules, 2 paléoles tronquée tyles sétacés, 2 stig., — continus, plumeux, "us glabre shee Epillets pédicelle е Сите 2 valv. à valve ext. pubescente, lancéolée-linéaire- , plane, à valve int. lancéolée-aigué, presque edo. : зң той Кораб» 2 ét. glumelles 2 уају. à valves lancéo lées, linéaires, aigués, presque égales, е membraneuses, un peu violettes. Toute la plapte ed une forte odeur de citron. Nom. vulg. cepim limao. Босанатын St. Hilaire made no great effort to list binomials іп the cata- logues. In the more than 8900 collection numbers there are approximately 160 binomials, the majority in St. Hilaire’s handwriting and a few in Spach’s. Most of 10 The catalogues oe not list as many specific localities as does Urban in his elaborate t of th itineraries of id RÀ a in ve sic — inpet ми (loc. age! Here more than 800 loc ка ities are given a in sequen according to the years of St. Ийме "voyages". It is у probable shee ure еы this detailed irs a from St. Hilaire's own pub- lished reports of his т in 2682 Dreuzy terminates this "analysis" of a Gr aminée with the binomial Echites Loa e de S.-H. This binomial, despite the fact that it appears in the — to be w E “analysis” = the upper right-hand corner, obviously refers to collec : number pret "26H Apocyna To make matters worse, the second net. шун й is labelled "Composée К т? КК the catalogues the collection number is simply 278 [Vor. 42 158 ANNALS OF THE MISSOURI BOTANICAL GARDEN the binomials are addenda to the "analyses"; those written by St. Hilaire belong as a rule to families on which he published extensively: e. g., Violaceae (including cer- tain genera now assigned to the Ochnaceae), Droseraceae, etc. For some reason "analyses" of certain families, e. g., Malvaceae, Malpighiaceae, etc. are often cross- hatched in ink. In the case of the Rutaceae the numbers of the family were designated by a scrawled name which looks like "Oranger" (pl. 17, under 270112). Figures 1 and 2 show the number of collections made in specific families (with the exception of the ferns) according to the catalogues of St. Hilaire. Figure 1 lists the families represented by more than 50 collection numbers; fig. 2 lists those represented by less than 50 and more than 20 collection numbers. The family Leguminosae is considered in sensu lato to include the Mimosaceae and Саеза!- piniaceae. The Ochnaceae includes certain genera which St. Hilaire related to the Violaceae, e. g., Sauvagesia, Lavradia. To avoid confusion I have combined the Liliaceae with the Amaryllidaceae. NUMERICAL SEQUENCES OF ST. HILAIRE COLLECTIONS In preparing a collection of plants for herbarium deposit the most convenient sequence of collection numbers is a continuous linear one which is correlated with the date of collection and the itinerary of the collector. In examining the more than 8900 collections of St. Hilaire we find the following numerical sequences!? in the ten books Book 1— 1- 758 Book 6— 86- 96 464— 650 % 4 xi ot Book 2— | 1B- 1328 | 7—} з 60 (not in St. Hilaire's handwriting) Book 3— I 63, 103-2085 gg. | 1176-1899, 2000-2797 Book 4— 1- 112 | 1-219 жый s pande Book 9— 692- 709, 1858-1869, 2004-2120 1- 408 Book 10— 1- $1 Despite the patch-work pattern of these sequences of numbers definite cor- relations exist between sets of numbers and the itineraries of St. Hilaire's five voyages. It is not impossible that St. Hilaire prepared the catalogues following his return to Paris in 1822. I have seen certain individual "analyses" attached to herbarium sheets in the Paris Herbarium, e. g. to one of several of St. Hilaire's un- numbered collections of Lavradia capillaris (Ochnaceae). It may be that the "analyses" were written on separate sheets of paper in the field, inserted separately with the unmounted specimens and then removed in Paris to be gathered together and recopied into a catalogue. The mounting of his 30,000 specimens in Paris may have resulted in some confusion in the rc-organization of the "analyses" in the catalogues. Occasionally, one finds the date 1822 at the end of an analysis, sug- gesting that an addendum was made to the original more-abbreviated "analysis". "analyses", too, are generally written in a neat (but often difficult to decipher) 1218 a given run of a number in the сайл it is not uncommon to find nal n missing or out of place. For the sake of brevity I am not specifying ds; sheng а the absentees have been taken into account in computing the sum-total of collection in the 1955] DWYER—BOTANICAL CATALOGUES OF ST. HILAIRE 159 COMPOSITAE МЕНВЕНАСЕА ONVOLVULACEAC | ____ |] MALPIGMIACEAE AMAR AN’ E TMACEAE — EGG — 100 200 300 400 500 600 700 #0 900 1000 100 Fig Approximate numbers of collections listed bars желе families of Angiospermae (ош a Ferns) in the Maris sime of Aug . Hilaire. Only those families with show collections in excess of 50 a GUTTIFERAE ONAGRACEAE OCHMACEAE ERICACEAE СЕНА ськ a пада A RR RE МА LO ЛАН SEE ce re a] URTICACEAE СЕНЕ ЯСИН зая URN ee LOBELIACEAE р RE SENT DNS а Ко | SAPINDACEAE RE AT Б PR EA TR С=с Пт ы з SE TURNERACEAE EES IRIDACEAE DIOSCOREACEAE _____---- 20 25 30 35 40 45 50 2. bers of collections listed under specific families of Angiospe ae in the posters ы. el dx ge su 8. Hilaire. Only those families with at least 20 if ar more than 49 collections are shown. [Vor. 42 160 ANNALS OF THE MISSOURI BOTANICAL GARDEN hand. Perhaps legibility is too much to expect of a scientist working in the evening after an exhaustive day in the fie In Book 5 Weddell writes that certain notes for collection numbers 2085—2101 were not sent to the Museum for insertion. Several points argue against this possibility, however. In the first place, it is more logical, if not more safe, to insert the "analyses" initially into a field book; again we note, especially in the early books, that St. Hilaire often fails to complete an "analysis" before beginning the next collection number but occasionally completes the "analysis" on another page and consequent to a different collection number. It is difficult to believe that he would do this if he were preparing the catalogues in Paris. Тһе fact that volume 6 has only 11 collection numbers sandwiched in with what is obviously a segment of his diary written in the field is another substantial point of evidence. St. Hilaire and his contemporaries used these numbers in the catalogues of col- lections and for herbarium sheets, and not for purposes of citation in the literature. St. Hilaire and his co-workers in the Flora Brasiliae Meridionalis, Jacques Cam- bessédes and Adrien de Jussieu, never cite collections by number. No doubt, they did not consider individual collections to be as important as do our botanists of today who adhere to the principles of the type concept. Obviously, if a collection of St. Hilaire had been cited, e.g.,No. 26, the reader would wonder to which sequence to assign it unless the locality were given. Wed- де! 3, no doubt stimulated by the pioneer efforts of certain authors in Martius’ Flora Brasiliensis in citing collectors and collection numbers, attempted (probably in the 1860's) to organize the catalogues in a manner more suitable for шш and for labelling of specimens. Unfortunately, Weddell did not publish his “К to the catalogues!*. The facts that Spach, who was curator of the Кыш, аї the Muséum d'Histoire Naturelle, employed Weddell’s system on printed labels and that certain monographers have used the Weddellian system give some status to the key. \Уе shall discuss the popularity of Weddell's key in following para- graphs. BH. A. bier d 877), an "aide-naturaliste" to Adrien de Jussieu, шеси extensively in Ir in — е was a distinguished taxonomist at the herbarium of the Muséum d'Histoire ec spat. typewritten copy of Weddell’s key. Weddell’s ы "аи may be found scattered in the ilai с 18). № he initi о: talog . Sw aminer of Questioned Documents, Clayton, Missouri, expresses the probabilicy vm eg was ем Бу а person (or persons) other than St. Hilaire, Weddell, or the author or authors of the page " 1. 19, i g a 5 ertain. М. Leandr бобів Че OB. mde D'Histoire Naturelle, Paris, writes me that the typewritten key was "copied from notes in the books [St. еч xatslo gues] partly of the handwriting of Weddell”. While Weddell may Б based his insertions on the чолі ized summary found іп k 8, it seems safer to suppose that the latter was gister from his n 19551 DWYER—BOTANICAL CATALOGUES OF ST. HILAIRE 161 WEDDELL'S KEY TO THE CATALOGUES OF ST. HILAIRE Weddell divided the books!” constituting St. Hilaire's catalogues into four series: А (books 1, 2, 3, and 4); B (books 5 and 6); C (books 7, 8, and 9) ; and D (book 10). Series A is divided into two volumes, B, into two volumes, C, into three volumes, and D consists of one volume. The following revised outline is based on a recent typewritten summary of Weddell's key found in the herbarium at Paris: SERIES A (TWO VOLUMES) А1, comprising numbers $.— Plants collected du uring the year HA some in the vicinity of Rio de Janeiro (1—469 and 651— 758), = ers in the vicinity of Uba, the Rio Paranhyba about 25 leagues from Rio de Janeiro, and in t in me province Туы у. A^, іп m леве" Section 1, comprising numbers 464—650, which are duplications of the same numbers in Volu. 1 | } Section 2, comprising numbers 18–1308 (ог 1328). Plants from the environs of Rio de Janeiro ы іп 1818. SERIES B (TWO VOLUMES) omprising numbers 1-2085 (December 1816 to March 1818).— 62; nts —Ü in the е of Minas Gerais, with the exception of 1—25, which are from the province o de е В", т two s ction ^s “comprising the numbers 2101"'" to 2493; these follow in order the series of plants of ише. Gerais which constitutes the three previous volu n 2, "comprising the е 1-408. Plants collected during a "voyage" on the of Mad vii o de Janeiro аң 2” Doce іп the provinces of Rio de Janeiro and Espirito Santo, from 2 to ара эщ SERIES C (THREE VOLUMES) C!, com g numbers I prisin EIS Plants collected i in € ; following districts: 10—Province of Rio de Janeiro. 1819. es [» is. 5 —Province of 626— 920—Province of Goi 9 to May, 1820. 986-1175--Ргоуіпсе of São Қы” dre ES to Mer. 1820. ca, mun iets two sections.— omprising numbers 117 6—2797. us erat in the following regions: aulo. 2569—2797—Province of Rio Grande до Sul. 1820. Нани ме San Pedro de Rio Grande to Montevideo. 1820. 2261-2568--Егот > и bp to the Brazilian frontier on the banks of the Uruguay. 1820. Section 2, comprising nap numbers 1— Han. tlle ected in the environs of раң > Janeiro, in July, 1821. Comprising three small series of numbers (692-709, Чет ea to and eher чи, to be in- tercalated in the large series of numbers contained in volumes C 15 Weddell referred to these as volumes. [Vor. 42 162 ANNALS OF THE MISSOURI BOTANICAL GARDEN SERIES D (ONE VOLUME) D, comprising the numbers Plants — м 1821 а а 1822 during = last voyage of St. Hilaire, in the following places: —Province of Rio de Jan as —Pro of Minas Can 588- in отне of 520 Paulo. Certain elements of confusion exist in the interpretation of these numbers. In Book 2 the numbers 464—650 are simply duplicates ("double emploi") of the same numbers іп the 1—758 sequence of Book 1. The numbers 1-112 in Book 4 are duplicates of those found in Book 3 except for 64—102 which are eg jr in Boo The numbers 86-96 in Book 6 (if one should refer to this as Book 6 asmuch as it is apparently one of St. Hilaire's diaries) belong to the sequence “of numbers in Book 5 from which they have been omitted. In Book 7 the initial sequence of numbers 1—60 are not in St. Hilaire's handwriting'®, In Book 8 num- bers 1900-1999 are missing, apparently due to carelessness. Weddell (?) used paper labels for all the specimens of the collections except numbers 1-758 in catalogue A! (Rio de Janeiro), 1-25 in В! (Rio de Janeiro), and 25-2085 in В1 (Minas Gerais), for which parchment labels were used. АП paper labels, except for numbers 2101""" to 2469 of B? section 1 (Minas Gerais), and 2470 to 2493 B? section 1 (Rio de Janeiro) bear, or should bear, symbols. This is summarized as follows: | ТАВІЕ І Labels Numbers Catalogues . Localities Paper à 1– 240 B? Sect. 2 Rio de Janeiro 241— 408 B? Sect. 2 Espirítu Santo Paper B 1 A? Sect. 2 Rio de Janeiro Paper C 1– 219 C? Sect. 2 Rio de Ja Paper D 1- 3 D Rio de Janeiro — 587 D Mi Ge 588— 818 D Sio Paulo Paper O - 10 cet Rio de Janeiro 11- 625 с: 626— 691 c Goiaz 692- 709 c Goiaz 710- 920 ce Goiaz 9 985 с: Minas Gerais 986-1175 с São Paulo 1176-1704 C? Sect. 2 520 Paulo 1705-1763 С? Sta. Catarína 1764-1775 e ~ Peo 1776-1806 c 1807-1857 с Rio бы, do Sul 1858-1870 с Rio Grande do Sul 1871-1874 с Rio Gran: 1875-2003 e Banda Oriental del Uruguay 2004—2020 с” Вапда Оп 2021-2568 (7 Banda Oriental 2569-2797 С Rio Grande do Sul S'These 60 numbers have neither families nor “analyses” appended, ае the localities of the collections are given. Some of the localities (e. 8», Barbacena Rio Ouro Branco) suggest the begin- ning of St. Hilaire's second voyage to Minas Gerais in 1816. Оп woke voyage he was рана n b es He patie алі Рыб dudes itv абы of one or both of th 1955) DWYER—BOTANICAL CATALOGUES OF ST. HILAIRE 163 THE ROLE OF THE CATALOGUES OF ST. HILAIRE IN TAXONOMIC MONOGRAPHS Although only a few botanists such as Weddell, Spach, Dreuzy, etc., are known to have consulted the catalogues of St. Hilaire, the various check-marks and binomials inserted suggest that they may have been studied by others. Prior to an arbitrary date of 1860 monographers in citing St. Hilaire collec- tions, in imitation of St. Hilaire, Cambessédes, and de Jussieu failed (usually) to supply collection numbers but gave clear-cut geographical data. Later mono- graphers often used St. Hilaire collection numbers, and, almost without exception, they furnished only a fraction of collections as listed in the catalogues. Mueller, for example, in his classic work on the Apocynaceae (Mart. Fl. Bras. 61:1-195 1860) gave collection numbers for only 23 of the 56 collections of St. Hilaire which he cites. These 56 collections represent about one-third of this family as listed in St. Hilaire's catalogues, all of which have specific collection numbers. In general, in Martius’ Flora Brasiliensis it is the exception to find collection numbers when St. Hilaire material is cited; collecton numbers are missing, for example, in Radlkofer's monograph of the Sapindaceae (1895), Bennett's work on the Poly- galaceae (1874), Koehne's work on the Lythraceae (1887), etc. (see Table II). No doubt much of the confusion and inconsistency encountered in the citation of St. Hilaire's collections, especially by the 19th century monographers, arise from several circumstances: (a) localization of the bulk of the St. Hilaire collections in the Paris herbarium; (b) the fact that the duplicate collections often have labels bearing only a binomial and the geographical data (often undecipherable)'*; (c) the lack of interest in collection numbers by the contemporaries of St. Hilaire; (d) awkward and overlapping sequences of numbers prior to Weddell's groupings; (£) the subsequent failure to apply the Weddellian system to many of St. Hilaire's collections. Table II, based on an examination of twenty monographs (1832 to the present) which include St. Hilaire's collections, reveals many important points!. In the Anacardiaceae (D.C. Monograph. Phanerogam. 4. 1883) Engler cites eleven col- lection numbers of St. Hilaire, none of which has catalogue numbers; in attempt- ing to check them against the catalogues it appears from the blank spaces in the latter that St. Hilaire may not have been certain of the characters of the Anacardiaceae and thus never entered the family name. In Cogniaux's Cucur- bitaceae (D.C. Monograph. Phanerogam. 3. 1881) the number of St. Hilaire's collections cited exceeds the total number estimated to be entered in the catalogues under this family. This disparity is a reflection of the number of blanks left by Y Of the a sheets of St. Hilaire reported by Dr. Lyman B. Smith, of the Smithsonian енн (Washington C.), to be in еі herbarium fre 34 bear collection numbers; of these only have the *Weddelian series appended to the number. I have been unable to locate ten of the num- 18 When знов я numbers а ot Hated i in a monograph the M желек „o> be including more than one collection (of St. Hilaire) for the species in question the same clear This makes an accurate statistical analysis of the маай ie collection cited impossible. [Vor. 42 164 ANNALS OF THE MISSOURI BOTANICAL GARDEN TABLE II SOME T 5 | 35 | =. | 55 ла = ТЕ c T. Ы Lg о 8 б "TM es $ MON a 58 ЛЕ, е 5% 5 Se i soak] 5» as Specific Monograph Date ч В T 78 т“ ^e 2j m £ g” © == = ge ENT o iis: | 45 [433| 35 | 3$ 84 | ve | за | e23 за | За же Duk Рош које Ep eda Bod uem Eg BE E ЕЗ Еф зб 26 2a 5 З. 3 8 zB Zi Za 2.8 крен 1832 64 64 0 ? 97 А. де Jussieu Мештошасеае 1849—1852 51 51 0 0 ? 378 Naudin Apocynaceae 1860 58 35 23 0 16 147 Polygalaceae 1874 30 30 0 0 ? 122 Erythroxylaceae 1878 8 8 0 0 ? 0 Peyritsc Alismaceae 1881 7 0 7 6 4 6 Micheli Commelinaceae 1881 11 0 11 0 6 18 Clarke Cucurbitaceae 1881 36 0 36 35 31 35 gniaux Turneraceae 1883 15 1 14 9 10 28 Urban Anacardiaceae 1883 11 0 11 0 0 0 Engler Lythraceae 1887 24 24 0 0 ? 62 оећпе Rubiaceae Tribes VII-XIX |1889—1889 8 2 1 0 1 386 Schumann Guttiferae 1893 4 0 4 1 4 46 Vesque Sapindaceae (Serjania) 1896 19 19 0 "|. tf 35 Radlkofer Bromeliaceae 1896 19 0 19 10 7? 14 Bignoniaceae 1897 2 1 1 0 1 21 ureau Ochnaceae ее 1902 7 1 6 0 2 41 ospermées) Van Tieghem Passifloraceae 1938 17 0 17 0 11 22 1 Vochysiaceae (Salvertia & |1948 8 1 7 E ^ 15 ocbysia) Stafleu Labiatae (Hyptis) 1949 100 8 92 2 79 244 Epling 497 250 249 67 176 1878 1955] DWYER—BOTANICAL CATALOGUES OF ST. HILAIRE 165 St. Hilaire after collection numbers in the catalogues, some of which are un- doubtedly of the Cucurbitaceae. In Naudin's monograph of the Melastomaceae!?, we find very few of St. Hilaire's collections cited despite the fact that Naudin must have been well acquainted with them??. What is even more remarkable is the complete absence of St. Hilaire material in the “second half" of the monograph (i. e., vol. 16, after p. 87. 1851), even though such common Brazilian genera as Clidemia, Miconia, etc. are included?!, In the several sample monographs from Martius’ Flora Brasil- iensis we note that few collections of St. Hilaire are cited as compared with the total number of collections of the specific families estimated to be in the catalogues. Killip, in his recent treatment of the Passifloraceae (Field Mus. Nat. Hist. Bot. 19:1—613. 1938), cites six numbers which do not check with the corresponding ones in the catalogues. Of these, three (717, 751, and 970) were probably erroneously copied from labels bearing the numbers, 719, 755, and 976 respectively, all of which are entered in the catalogues under Passifloraceae; a fig. 9 may be easily mistaken for a 7, a 1 for a 3, etc. Epling, in his monograph of Hyptis (Rev. Mus. de La Plata, n. s. Secc. Bot. 30:153—497. 1949) cites more St. Hilaire material than any contemporary mono- grapher??, Unfortunately, only two of the hundred collections cited bear collec- tion numbers with Weddell's series added. This suggests, of course, that the Weddellian system has not been applied in any great degree to the labelling of specimens. Despite the quantity of St. Hilaire material cited by Epling, approxi- mately 65 collection numbers of Hyptis were not included in his work. In checking this genus in the Paris Herbarium 13 of the 92 collection numbers (St. Hilaire) cited by Epling, were not located; but eight of these thirteen can be located readily when "bis", "ter", etc. are appended. (Some copyist of course may have failed to insert "bis", "ter", etc. on the labels.) Of the remaining numbers, prob- ably 578 was misread for 518, and 1155 for 1133. In those monographs using the Weddellian system the collection numbers are usually matched with ease in the proper volume in the catalogue. 19 Ann. Sci. Nat. Bot. Ser. ПІ, 12:196—284. 1849; 13:25-39, 126-159, 273-303, 347-362. 1850; 14:53-76, 118-165. 1850; 15:43-79, 276-345. 1851; 16:83-246. 1851; 17:305-382. 1852; 18:85-154, 258-294. 1852 206; Hilaire and Naudin were co-authors of a paper entitled "Revue de la Flore du Brésil Meridionale” which appeared in three sections (Ann. Sci. Nat. Bot. Ser. II, 18:24—54, 209-213. 1842). In addition they were co-authors of a new genus, Augustines, of the Melastomaceae (Ann. Sci. Nat. Bot. Ser. III, 2:145. 1844) 211 hope to discuss in another paper the stormy history of the Melastomaceae Hilaire. 22 Hyptis is one of the few cases where St. Hilaire wrote i his catalogues in his "analyses" instead of the amily name "Labiée". However, in c Epling's monograph against the catalogues, it is obvious that St. Hilaire did on occasion taxon of ily i ptis. family in place of Нур: collected by St. a genus name with fair consistency into hecking use th [Vor. 42 166 ANNALS ОЕ THE MISSOURI BOTANICAL GARDEN A PARTIAL LIST OF THE GENERA COLLECTED BY ST. HILAIRE In the herbarium of the Muséum d'Histoire Naturelle, Paris, there exists an unpublished list of about 2000 entries of St. Hilaire's collection numbers; these are arranged according to genera (occasionally binomials are listed) and cover eighteen families or tribes. The author of this list is unknown??. Fortunately, the collection numbers of St. Hilaire are entered according to the Weddellian system. The presence of geographical data increases its value considerably (cf. pl. 19). In the list the entries from numbers 1 to 364 are of the Graminée only, divided according to 13 tribes. Among the Dicotyledoneae we find the following families and tribes listed in the order given (entry numbers 1–1589): Chloranthaceae (entries 1-2) Nyctagineae (entries 150—182^!*) Piperaceae (entries 3—44) Chenopodeae entries 183—192) Celtideae (entries 45—53) Amarantacea entries 193—3 07) tries 54— Aristolochieae (entries 308—322P^!*) Artocarpiées (entries 85^!*—95) Proteaceae ies 3 8) Urticacées (entries 96-100) Monimieae entries 329—337) Salicineae (entries 101—103) Lau 338-390"'*) Lacistemeae (entries 103°'*—1035) Compositae entries 391-1572) Polygoneae (entries 104—149) Campanulacées (entries 1573-1589) It is obvious that such a list may be of great value to specialists of the above groups working on material from southeastern Brazil. As an example of its use- fulness, I was able to ascertain in the model analysis (cf. p. 157) that Elionurus is the genus of the grass family described by St. Hilaire. SUMMARY Monographers have used only sporadically the Weddellian system in checking the collection numbers of St. Hilaire’s material with the specific catalogue series and section, probably due to the absence of these numbers from the majority of the labels of the specimens. The Weddellian system answers the problem of the possible overlapping of collection numbers, and monographers should be encouraged to adopt it. Being aware of the existence of the catalogues of St. Hilaire, they are now able to check collection numbers against St. Hilaire's "analyses"?*, The cata- logues themselves represent an original source of reference for the 8900 collections of Aug. St. Hilaire in southeastern Brazil. The discovery of a list (author unknown) of some of the genera collected by St. Hilaire, with some 2000 collection numbers entered according to the Weddellian system, should be of great assistance to monographers. rs of which are obviously in Seach laneis. This permits one to give ње date of : the list ?4^'The author, through the Missouri Botanical агау will be happy to supply sets of collection numbers of St. Hilaire for specific families 19551 DWYER—BOTANICAL САТАГОСОЕ$ OF ST. HILAIRE 167 APPENDIX Through a letter received during the proofing of this paper I learned that the Bibliothéque Centrale (Mus. Nat. Hist. Naturelle) has no manuscripts of Aug. de St. Hilaire except several fragments of the Flore du Brésil Meridionale and four of his letters. Apparently there are no St. Hilaire manuscripts in the Library of the Division de Phanérogamie of the Museum except the catalogues herein described, and a single letter. In addition, I learned that the editor of St. Hilaire's Voyage à Rio Grande do Sul is Charles Roland de Dreuzy (1837-1899), a native of Nivernais. He was the grand-nephew of St. Hilaire, heir to his estate at Turpiniére, and grandson of a sister of St. Hilaire. At his own expense he published the Voyage à Rio Grande do Sul, in accordance with the last wishes of his "parent". The chateau at Tur- римеге is located near Ferté Saint-Aubin, Sennely (Loiret). The present owner is M. Robert de Dreuzy, son of the editor of Voyage à Rio Grande do Sul. The grass described in the model "analysis" (cf. p. 157 of text) was determined by Miss Agnes Chase as Elionurus candidus var. bisetosus Hack. and Linden. [Vor. 42, 1955] 168 ANNALS OF THE MISSOURI BOTANICAL GARDEN EXPLANATION OF PLATE : PLATE 17 A portion of one page of Catalogue C? of Auguste de St. Hilaire. This is in his hand- writing and covers numbers "2791" Hypoxis, 279119, Labiée, 2791", Solanée, 2791", Oranger", and "Suite à 2791".” The last is a continuation of the "analysis" of Hypoxis at the top of the page. PLATE 17 - ANN. Mo. Вот. Garp., Vor. 42, 1955 рвы ее Soret Ж. Lom afe run Port nie ~- вн df =. ручак ond 5” QU - deu сёе sata oriad Ж. ANE A фа Е. Sd do o — 1 „= у Ж oblong sme prm 34. m REA C. а аа сув, i i pation ы) dre ас и Жу гео wey any © әкей ae нее ера ас rn << EL 4 =; ! a er ње саро ооо actin а 244. Е ЛИВ. | Loa E ЖЕ “2 22 о) Зая 22241. “+ Ё parse У диван —24< "e с вад “= eS Dre cer! „учет ш. 4 Р еп авва аА | я. Pi СЫН co мети в В yer? bpn р» ДЕР ЭШ... а.с. aget ат IO sr, ке г o шо emmamma Om Leta | ^ c А losak 25 лаа АВ Poit flam 22.2 ва ее. | на M 2 а TENS НИ PAT Fm adis barenga i 2 MIB a а = ЈУРИ + EE 2 Aa Dns Prem остао с Су p —€— Км “о “= 2-22 ern ry im metet ә ма 4. T. e a. А. EY ен и ж Ф; andit p зерно о pe re ntn ж по О: ст HH rm rt o e , | : ‚ » 7 amant se D I tm B arn ба ei DWYER—BOTANICAL CATALOGUES OF ST. HILAIRE Ann. Mo. Bor. Сакр., Vor. 42, 1955 PLATE 18 hy 9. ТЗ а ооч "M. . 016 - 985 o o кб = ie dA | n g6- И. m : | | (+: с ME ies 9% d iJ e. ve B Fut Ace J йл 4220. ете 7 2 UN mm о e 2... DWYER—BOTANICAL CATALOGUES OF ST. HILAIRE (Мог. 42, 1955] DWYER—BOTANICAL CATALOGUES OF ST. HILAIRE 169 ExPLANATION OF PLATE PLATE 18 A portion of the "key" to the botanical catalogues of Auguste de St. Hilaire. This is in Weddell's handwriting and is from Series C, catalogue СІ. The handwriting at the top and bottom of the page is that of the librarian. [Vor. 42, 1955] 170 ANNALS ОЕ THE MISSOURI BOTANICAL GARDEN EXPLANATION OF PLATE PLATE 19 A page from a list of approx mately 2000 collection numbers of Auguste de St. Hilaire. Note that the Grasses are entered according to tribes and genera. The fourth collection entered under the genus Elionurus (C? No. 2682) is that upon which the “model analysis" (see p. 157) is based. Thelist is deposited in the Herbarium of the Muséum National d'Histoire Naturelle, Paris. ANN. Мо. Вот. Сакр., Vor. 42, 1955 PLATE 19 [gum : | Origine, Donateur, ж, 2-2 1 eder me тайот of À а Nome vulgaires Judwatous T фузсват | : Sigues à Е.ни “4 Catalogue (m yet јеле : т.317 Жоғ айыға. жа PED d ие 7204 he 7 ed т.%% “A, ‚тм. ot we 2247 |% я. з я. Kes pac 24 Се A ed к^ я. м. 330 2. Bada 7“ 47 5 Ға” F M %. Н A м XN. ; 4 + я d са m Li 27 ‚ ATP > e - баба Хат. DW YER—BOTANICAL | с“ тела; A nt rn A! nepe 6/4” 21 dus E: d'un tta gt m^ өзе mtt ase 27.54 и 1370 tmo три CATALOGUES OF ST. HILAIRE Ww * Y MSN AS SM $ COMPARISON OF JUNIPER POPULATIONS ON AN OZARK GLADE AND OLD FIELDS MARION TRUFANT HALL* INTRODUCTION Using comparative morphology as criteria, I presented evidence in an earlier paper (Hall, 1952) that Juniperus virginiana L. has differentiated into five geo- graphic races, and two of these races, called Typica and Ozark, were described in detail. Typica is known to the horticulturist as the Eastern form and may be characterized as the very tall, narrowly pyramidal tree found from Virginia to southern Illinois. Within its present distribution Typica is concordantly variable, and at least so homogeneous that it seems to be carrying no introgressant genes from other species. The influence of Typica in the juniper populations of the Ozark Plateau is not evident even though isolated plantings of this race are to be found on the western banks of the Mississippi River from St. Louis southward, usually about old farmsteads. The only other race of red cedar which occurs occasionally in the St. Louis area is the Northern, which is characterized by high frequencies of the spire-like form known as var. crebra, by crooked fruiting peduncles, and a preference for well-drained, sandy soils with a slightly acid reac- tion. I postulated that this Northern race may be the consequence of a few genes from J. horizontalis becoming stabilized in red cedar. The Northern race, with its tendency toward suppressed lateral branching, follows the glacial drift down the Mississippi River, and where the periglacial loess occurs these junipers are abundant. South of the loess deposits, particularly in the vicinity of St. Louis, juniper popula- tions closely resembling the Northern race occur sporadically. The Ozark race is indeed the most abundant and important in the St. Louis area and represents red cedar throughout the Southwest. The Ozark race, which is the result of introgression by genes of J. Asbei Buchholz into J. virginiana L., includes a highly variable assemblage of junipers distributed from central Texas across eastern Oklahoma and the Boston Mountains of Arkansas through the Mis- souri Ozarks to the Mississippi and Missouri rivers. Introgression by genes of J. Ashei into red cedar has probably occurred repeatedly, at least during the Pleisto- cene, and current mixing is evident in restricted areas where extensive hybrid swarms exist. e purpose of this investigation was to determine the differences between the habitats in relation to the differences in the variability within and between popula- tions of these Northern and Ozark races of junipers. The habitats studied are close together, less than a mile apart, and are surrounded by many hundreds of popula- tions of junipers both large and small. It was thought that these habitats may select from this large gene pool adaptive genotypes which may vary from popula- tion to population proportionately to the habitat differences. * Cranbrook Institute of Science, Bloomfield Hills, Michigan. 171 [Vor. 42 172 ANNALS OF THE MISSOURI BOTANICAL GARDEN ACKNOWLEDGMENTS In the spring of 1952 I was granted a leave from the Cranbrook Institute of Science to be a guest lecturer in Genetics and Evolution at the Henry Shaw School of Botany, Washington University, St. Louis, Missouri, and to take over Dr. Edgar Anderson's class during his absence. This study was part of a project undertaken by Dr. Anderson's class under my direction. I am most grateful to Dean Henry N. Andrews for his advice and cooperation and to Dr. Anderson for his enthusiasm in regard to field studies as well as his helpful criticism. The study was carried out in the native tracts of the Missouri Botanical Garden Arboretum, Gray Summit, Missouri, which is approximately forty miles southwest of St. Louis on US Highway 66. I am especially grateful to Mr. August P. Beilmann, Superintendent of the Arboretum, for his cooperation and detailed knowledge of land use in the Ozark region. GENERAL DESCRIPTION OF THE AREAS The areas studied are within the Missouri Botanical Garden Arboretum, which is situated 40 miles southwest of St. Louis, Missouri, just south of the junction of Highway 66 and Highway 50 near Gray Summit, a hilly region principally of weathered dolomitic limestones in the northern portion of the Ozark highland. The climate is designated by Thornthwaite as BB'r (humid, mesothermal, with precipitation adequate at all seasons). Records in St. Louis for the past hundred years show such variations in precipitation that no general trends are evident, so that it is difficult to judge whether climatic amelioration or deterioration is in process. The Ozark Plateau is a part of the Oak-Hickory Forest Region described by E. Lucy Braun (1950) as the Interior Highlands, Southern Division, which lies essentially south of the glacial border. Braun wrote: "The Southern Division is characterized by the prominence of southern species of oaks and hickories, as Quercus stellata, О. marilandi Carya Buckleyi (var. arkansana and var. villosa), and in the bottomlands, О. nigra, О. lyrata, Carya Pecan, С. myristicaeformis, and C. aquatica . . . . Prairie openings, limestone glades, and balds locally interrupt the forest cover." Those who know the Ozark landscape consider these local interruptions of forest cover as the particular mark of beauty of the uplands, and the “openness” was a major feature in selection for settlement in pioneer days. Visitors to the Ozarks from foreign countries usually are most impressed by the red cedar glades which are island playgrounds in the present-day dense oak woods. In the proper season these open areas serve as edge for much wildlife activity, and perhaps the most characteristic features in this wise are the summer call of the chuck-will's- widow and the "booming" of the nighthawk that so often chooses a glade for his target. The soils indicate a long history of open conditions where prairie vegetation, probably maintained by fire and grazing, played a very important role in the development of soil horizons. Soils in the area are characteristic of prairie-forest 19551 HALL—OZARK JUNIPER POPULATIONS H3 transition regions and are much like the nut-structured prairie-forest Grood soils of continental sub-humid steppes, being intermediate between chernozems and podsols. Regardless of its history of prairie development, the Ozark region has a forest climate and probably has had since the end of the xerothermic period. Thus the climax plant formation, which is of course subject to continual change, may be the oak-hickory forest which has been spreading extensively throughout Missouri roughly in the last hundred years. Beilmann and Brenner (1951) and Etter (1953) report this trend in their studies of Ozark vegetation. It is not likely that any region has ever had a vegetational mantle entirely in equilibrium with climate alone, for other factors affecting balance are legion—e. g. nature of geological substrate, grazing and browsing, fire, and man. Rather, as Sauer (1950) has pointed out, "Plant associations are contemporary expressions of historical events and processes, involving changes in environment and biota over a large span of geologic time." However, Braun (1950) stated that knowledge of these historical events and processes is very scant even though general trends seem to be quite clear. The history of land usage in the Gray Summit area is not known in detail, but titles to land were granted beginning about 1850, nearly 50 years after the first settlers, mostly French, began to locate along the major waterways in what is now Franklin County, and over 80 years after St. Louis was founded, also by the French. Evidence of timber cutting, heavy grazing and burning variously from the time of settlement until the Missouri Botanical Garden acquired the property in 1925 is clear and common. Apparently, in the days before the settlement of the region most of the timber was found on the headwaters of the creeks and was scattered, giving a prairie- or park-like aspect presumably as a result of a somewhat drier climate than today and the effects of grazing, fire, etc. The timber, where dense enough to be called woodland, was usually situated in the rough lands or dissected headwaters of the streams where it was largely safe from fire. Today the Ozarks are well covered with close forests, and the once open areas show signs of rapid forest encroachment; old fields develop rapidly to forest, but the glades show little invasion by woody species even when protected from fire or grazing. Steyermark (1940) made a vegetational survey of Missouri which cov- ered fifteen years of extensive observation, and even though the survey was strictly qualitative it was detailed enough to permit rough comparisons with present-day conditions. He stressed the role of rock strata and soil, i. e., edaphic factors, in determining the distribution of plant associations. Не described six plant associa- tions and many more associes based on differences in soil, slope, and water avail- ability; in his conclusions he favors the polyclimax concept subscribed to by Domin, Gams, Gleason, Tansley, and Du Rietz. He indicated that the glade flora was the result of edaphic factors, and in most sites would eventually support a maple-white oak association which in the Ozarks seems to be restricted to slopes and bluffs with neutral to alkaline soils underlain by limestone. The Glade.—The Glade, which is one of several within the boundaries of the Missouri Botanical Garden Arboretum, may be characterized as a small open space [Vor. 42 17 4 ANNALS OF THE MISSOURI BOTANICAL GARDEN (open glade), approximately 60,000 square feet in extent, covered by grasses and junipers surrounded by a closed area (closed glade), or buffer, approximately fifty feet in width, chiefly composed of junipers and chinquapin oaks and situated on a west-facing slope (upward ENE, 10°20’). The area studied constituted a rectangle 400 Х 150 feet, approximately the total extent of the open glade. In this study the term Glade refers to the open glade within the enclosure studied. This same glade was described by Erickson, Brenner, and Wraight (1942), who demonstrated a direct correlation between the structure and composition of the glade flora and the occurrence of a dolomite stratum at or near the surface. The glade soil is very shallow (averaging about 4 inches), organically rich and derived from a dolomitic rock which under the open glade is a thin, porous dolomite locally known as “cotton-rock,” an upper stratum of the Cotter Formation, and lies un- conformably on the Jefferson City Formation, both in the Canadian Series of the lower Ordovician. Under the closed glade a less porous dolomite with lower magnesium content is covered by a fairly thin mantle of Union Silt Loam. The dense oak-hickory upland forest on the Union Silt Loam gives way to red cedar- chinquapin oak forest at the periphery of the glade, but wherever the "cotton- rock" occurs it drains the area sufficiently that the red cedar-chinquapin oak forest is fairly well stabilized. In winter and spring, as a consequence of greater precip- itation and less evaporation, the open glade (“cotton-rock” stratum) is wet, often with standing water, while in summer and fall, when precipitation effectiveness is least, the porosity of the rock permits rapid drainage which results in rather extreme dryness in the upper few feet of the stratum. I have been frequently observing several of these glades since 1947, and, besides the rather obvious seasonal cycle of wetness and dryness, there are fluctuations every few years which markedly affect the population density of the species and the cover. In order for the peren- nials, in particular, to be successful in the glade community, they must be able to withstand fairly great fluctuations of the environmental conditions. The Old Field.—The Old Field is situated on a gently sloping knoll below the orchid greenhouse. The upward slope (5°10’) is NNW facing SSE. The knoll is covered by Union Silt Loam and probably will eventually support an oak-hickory forest. On this knoll a plot 60,000 square feet in extent was laid out as a rectangle 300 Х 200 feet. Prior to 1925 the area was under cultivation in the traditional “Ozark fashion.” A few acres of beans, peas, and corn were planted, rows running uphill, and tilled by means of the mouldboard plow and mule driven by a tenant farmer. In 1925, when the property was.acquired by the Arboretum, this field was badly eroded and almost bare of vegetation, but since 1925 it has not been grazed or burned, and the revegetation has been accomplished in the last 28 years. At present, the arborescent growth on the Old Field consists of scattered juni- pers, black oak, white ash, and thickets of red elm. Wisteria is an abundant shrub 1955] HALL—OZARK JUNIPER POPULATIONS 175 having invaded the field from plantings nearby. The upper edge of the sloping field supports a growth of sugar maple, white ash, black oak, shingle oak, post oak, black cherry, smooth sumac, buckbrush, and poison ivy. The middle portion of the field consists of thickets or clumps of red elm saplings with a maximum height of about ten feet, and a thick growth of smooth sumac and wisteria. On the lower slope which drains into a tiny creek are growing black oak, white ash, northern red oak, persimmon, rough-leaved dogwood, winged sumac, poison ivy, woodbine, and summer grape. The oldest arborescent growth is made up of persimmon, post oak, black oak, and shingle oak, which apparently invaded the field approximately five years after its abandonment. Then northern red oak, sugar maple, white ash and red cedar came in six to eight years after abandonment. At any rate, the older specimens of red cedar were twenty years old, eight years having elapsed since abandonment of the field. The rate of invasion by red cedar depends on the balance of several variables such as the number and proximity of seed-producing plants, condition of the invaded area, amount of cover, slope, pH, etc. Drew (1942), in his studies of abandoned crop-land in the Cedar Creek area, Boone and Callaway counties, Mis- souri, found that five years after abandonment sassafras, persimmon, shagbark hickory, post oak, and shingle oak invaded; after six to seven years, American elm; after eight to ten, northern red oak, sugar maple, white ash, honey locust, and black walnut; after eleven to twelve years, red cedar and black oak. In the herb layer the general cover consisted of fairly evenly distributed clumps of broom-sedge with Kentucky blue-grass in swales and level places bordered by clumps of little blue-stem. Оп the up-slope, broom-sedge, tall red-top, Canada goldenrod, and old field goldenrod were conspicuous. From midway to the bottom of the slope were these same species plus dewberry (Rubus flagellaris) and Aster pilosus. Неге and there were almost bare sandy slopes, a few feet in extent, con- sisting of stands of triple awn-grass with a few lichens and mosses. Directly beneath the crowns of most of the larger trees were areas covered by lichens and mosses. Cedar Hill.—Another old field, here called Cedar Hill, is ап exclosure with a gentle slope from west to east mowed annually with a sickle-bar and covered with well-spaced red cedar, black oak, black cherry, honey locust, and red elm. The shrub layer consists of dewberry (Rubus flagellaris), smooth sumac, trumpet-vine, buckthorn, aromatic sumac, large clumps of buckbrush, partridge pea, and seedlings of sassafras, shingle oak, and persimmon. The herb layer consists principally of Canada blue-grass, triple awn-grass, tall red-top, Panicum capillare, and frequent clumps of broom-sedge, scattered clumps of little blue-stem, and occasional clumps of mountain mint and ironweed. Dewberry (Rubus flagellaris) was scattered fairly evenly throughout the field. The junipers average about 30 feet in height and thirty years of age, somewhat younger than those of the Glade. The juniper reproduction in the exclosure was [Vor. 42 176 ANNALS OF THE MISSOURI BOTANICAL GARDEN negligible. Although there were abundant seedlings, particularly at the edge of the crowns of those mature junipers bearing berry-cones, all were below sickle-bar height and usually under three years of age. Just outside the exclosure and in the general area juniper reproduction was very good, and ages up to thirty-five years were represented. Species composition was not analyzed because of the artificial situation. In the meadow the juniper population in an area of 60,000 square feet was sampled. In general features this meadow is closely similar to the Old Field. The soil is Union Silt Loam but much deeper since Cedar Hill lies above an old river channel. The Cedar Hill population was included in this study because of its intermediate position morphologically between the Glade junipers and those of the Old Field METHODS OF STUDY Several factors were studied for the Glade and Old Field in order to make com- parisons. Species composition, distribution of juniper seedlings, and the population structure of junipers were studied, and the data appear in Tables I-II, figs. 1-3, and pls. 20—21. Frequency was used as a quick method of measuring species com- position. The Glade and the Old Field were divided into four plots by staking out the diagonals. Along the diagonals alternating from one side to the other, 20 quadrats (1 meter in size) were laid out, equally spaced from the center. Prelim- inary work was done in late spring of 1952, but other surveys were made in late spring and late summer of 1953. The presence of species in each quadrat was tabulated and the results for the Glade and Old Field appear in Table I. The areas studied seemed sufficiently homogeneous for the use of 1-meter quadrat size, specif- ically to determine the species composition of the herb layer. А sample area about 0.5 per cent of the total was used. А compromise in favor of a small sample area was made because of the limited time the class could be in the field. Variations of the juniper populations were measured by pictorialized scatter diagrams as demonstrated in Juniperus in a previous paper (Hall, 1952). Five characters were scored: gland length-width ratio; length of whip (longshoot) leaf; length of terminal whip; length of lateral whip; on secondary shoots, the per cent of tertiary branches which have the decussate leaf arrangement. Two of these characters were also analyzed by use of the parameters of the log-normal curve which the data fit. These data, when plotted on logarithmic p paper, result in a straight line. In an attempt to get some estimate of seed-dispersal pattern, the distribution of seedlings was studied. Observations were made to determine the role that birds, mammals, and gravity may play in seed dispersal. SPECIES COMPOSITION Rough comparisons were made of species composition of vascular plants for the Glade and the Old Field. Only one quadrat size was used and no attempt was made 1o obtain quantitative data on density or cover. For convenience a frequence index after the method of Raunkiaer (1918) was made up for the Glade and the 1955] HALL—OZARK JUNIPER POPULATIONS 177 Old Field. The percentage of frequence for a given species is the percentage ratio the plots on which the species occur bear to the whole number of plots sampled. Raunkiaer found that the greatest number of species had the least frequence; and that as the frequence increased the number declined steadily, until at the highest frequence the number increased slightly. Raunkiaer used five frequence groups, A, B, C, D, and E, designating the species of frequence 1—20 per cent, 21—40 per cent, 41—60 per cent, 61—80 per cent, 81—100 per cent, respectively. His law of dis- tribution, determined from quadrat studies in Europe was A > В ^C = р <Е; the actual percentages were approximately 53, 14, 9, 8, 16. In eastern America, Kenoyer (1927), found that 69, 12, 6, 4, 9 per cent were more characteristic. In this study the Glade had a frequence index of 28, 8, 2, 2, 2 or A>B>C=D=E. Thi index is similar to that found by Kenoyer in his studies for America and indicates a fairly high proportion of sporadic species and a very few acting in a "dominant" role. The Glade has uniformly distributed patches of Sporobolus neglectus, and the size of each patch fluctuates somewhat from year to year. When the patches are smaller, more Bouteloua curtipendula and forbs are in evidence. In good years the Glade has a good cover contributed principally by a very small number of species even though there is an appreciable number of sporadic species between the clumps of the cover plants. The Old Field index was 22, 3, 8, 5, 0 ог A>BD>E. This somewhat irregular result is a product of the heterogeneous nature of the Old Field flora. There are a large number of sporadic species, a fairly large number of species inter- mediate in frequence, and no "dominants." The floral composition and vegeta- tional structure are markedly different from those of the Glade. Factors of competition and selection are less delicately balanced in the Old Field, and these factors undoubtedly cause the population structure of the junipers there to be different from that of the Glade. Table I lists the species and their frequencies found in sampling 0.5 per cent of the area studied. Only the under-story is represented. Woody plants in the list were seedlings or sprouts in size class 1 (0—0.9 feet) of Weaver and Clements (1938). The Old Field had seven species of woody plants represented in this size class for the area surveyed, which compares with nine species of similar size found by Drew (1942) to be common in old fields twenty-five years after abandonment. In the Old Field all sizes of woody plants were abundant up to size class 4 (3.6-9.5 inches D.B.H.), but on the Glade very few specimens larger than size class 1 were seen, suggesting a high seedling mortality. The frequency data in Table I show that the Glade and Old Field are quite different in species composition and suggest a difference in the pattern of succession. The Glade is a “prairie” association with Andropogon scoparius and Rudbeckia missouriensis contributing most to its aspect and with Andropogon scoparius and Sporobolus neglectus contributing most to the cover. Even though there is some fluctuation in abundance from year to year. there are definite dominants, a good variety to the flora, and a stability in its over- all composition which suggests that succession is very slow and that the Glade [Vor. 42 178 ANNALS OF THE MISSOURI BOTANICAL GARDEN TABLE I SPECIES LIST WITH FREQUENCIES FOR UNDER STORY OF GLADE AND OLD FIELD* Glade Frequencies Old Field 1. Sporobolus neglectus 1.00 2. Carex Crawei .90 3. Andropogon scoparius .80 Poa prate 4. Houstonia longifolia .75 Rubus Први 5. 70 Andropogon virginicus 6. .70 Aster pilosus 2% .65 Desmodium glabellum 8. Eupborbia corollata .60 9. Rudbeckia missouriensis 255 Китех тілден 10. 285 Oxalis s 11. .50 pencil prie HE .50 Lichens 13. Sisyrinchium campestre .45 > rigeron strigosus 14. .45 Lom finer 15. 45 eracium Gronovii 16. 45 е radicans 17. Hypericum prolificum 40 Triodia flava 18. Gaillardia pulcbella .40 19. Echinacea pallida .40 20. Ruellia bumilis :35 21; ther ouriensis 539 22. Heliotropium tenellum 35 23. 30 д кые cte 24. Viola pedata :25 artbenocissus quinquefolia 25. Selaginella rupestris 25 = 26. Juniperus virginiana PL 27. Viola papilionacea .20 rubra 28. .20 Celtis оета: e cp 29. Coreopsis lanceolata var. villosa .15 Hypericu ye 30. Ulmus rubra T Prunus s 31. 115 Cercis Padre: 5 32. 13 Rhus aromatica 33. Mosses .10 Juniperus virginiana 34. Eleocharis compressa 10 Asplenium platyneu 35. Aster elon pilates 10 Ат iens не о 36. Zizea а 110 Allium vineale 37. Mona rda ли .10 ronments carolinense 38. Cercis с жиын ‚10 атех епі 39 Қаған с jana var. mollis 10 Eragrostis spectabilis var. sparsibirsuta 40. Comandra Rich rdsi .10 41. Lotus ameri .10 42. Orobancbe uniflora .10 43. Acalypha graciliens 110 44. Рапісит virgatum .10 45. Triodia fla .05 onica arvensis 46. Lic hia .05 ~ syria 47. Bumcls ко ли уаг. .05 Acer ры iin ongifolia 48. қөмек americana .05 Pbysalis pruinosa 49. Monarda Dante .05 Pentstemon palli 50. Acer saccharu .05 Sporobolus neglectus 51. Anemonella thalictroides .05 Daucus pusill 52. Bouteloua curtip la .05 Antennaria fallax 53. Silphium laciniat .05 Potentilla тесі 54, Psoralea esculenta .05 Сайит ар 55. Quercus Mueblenbergii .05 Poa compressa 19551 HALL—OZARK JUNIPER POPULATIONS 179 Glade Frequencies Old Field 56. Pellaea atropurpurea .05 Rhus Toxicodendron 57. Fimbristylis Drummondii .05 58. Petalostemu .05 Sympboricarpos orbiculatus .05 60. Pentstemon pallidus .05 * Nomenclature after Gray's Manual of Botany, 8th edition, 1950. community is one of real integrity. Because of the uniformity in distribution of Andropogon, Sporobolus, and Bouteloua the quality of the cover on the Glade can be determined fairly well with frequency data. In a good year the clumps of Andropogon are larger, denser, and more frequent. The Sporobolus effectively fills in between the clumps. In a dry or unseasonable year the Andropogon may lose ground, the Sporobolus to а lesser extent, but the Bouteloua may increase con- siderably. Thus, one can tell when the Glade is in good condition, so far as climatic fluctuations are concerned, by the relative frequencies (or preferably cover) of Andropogon scoparius and Bouteloua curtipendula. 1 think the Glade’s general aspect and its severe selection against trees, shrubs, and woodland herbs, particularly as evidenced by selection against all but very Ashei-like junipers, support the con- clusions of Erickson, Brenner, and Wraight (1942) that these glades are edaphic sub-climaxes. The frequency data show the Old Field to be fairly typical for the Ozarks. There are several abundant species, but none of these contribute decisively to the field’s aspect or cover. Poa pratensis, Andropogon virginicus, Panicum virgatum, Triodia flava, and Rubus flagellaris are conspicuous. While the number of species is much the same for the two areas, the Old Field has fewer species with high frequencies or very low frequencies, and more with average frequencies. The O Field is obviously in a rapid stage of succession toward woodland. Several other species were not found in the quadrats sampled. Mostly, these are not common plants in these areas but in some cases may occur with frequencies of 5 or 10 per cent in localized areas of the Glade or Old Field. The following species occurred on the Glade but did not occur in the quadrats: Ulmus rubra, Celtis occidentalis, Physocarpus opulifolius, Amelanchier arborea, Rosa carolina, Prunus serotina, Rhus glabra, R. aromatica, Ilex decidua, Celastrus scandens, Acer saccharum, Cornus Drummondi, Diospyros virginiana var. pubescens, Fraxinus quadrangulata, Viburnum rufidulum, Sporobolus heterolepis, Elymus virginicus forma hirsutiglumis, Allium stellatum, Menispermum canadense, Psoralea psora- lioides var. eglandulosa, Linum sulcatum, Petalostemum purpureum, Croton monanthogynus, C. capitatus, Asclepias tuberosa ssp. interior, Triosteum perfoli- atum, Vernonia crinita, Kubnia eupatorioides var. corymbulosa, Hypericum рет- foratum, Amsonia illustris, Isantbus brachiatus, Ruellia humilis, Solidago rugosa var. aspera, S. rugosa var. celtidifolia, S. nemoralis, and Partbenium bispidum. А [Vor. 42 180 ANNALS OF THE MISSOURI BOTANICAL GARDEN total of 80 species of vascular plants was collected on the Glade (45 from the quadrats; 35 by systematic search). Eighteen other species, mostly seedlings of the woody species already mentioned, were found in the Old Field but not in the quadrats. According to experience and estimates, it seems likely that slightly over 100 species may be found in similar areas the size of the Glade and Old Field. Since these studies did not run continuously through the seasons involved, a considerable number of species was probably missed. It is also obvious from the data that a sampling area 0.5 per cent of the total is too small for best results; an area between 1 and 2 per cent would have been effective. The closed glade is a tension zone between the open glade and the oak-hickory woodland. It is a mixture of the open glade and oak-hickory woodland floras, but most of the cover in the arborescent layer is from juniper and chestnut oak. The transition from one community to the other seems abrupt since it occurs over a short distance, usually from 50 to 100 feet. From transect studies in the closed glade I listed 75 species, 39 of which were frequent species in the oak-hickory woodland and 43 were frequent in the open glade (Glade). Plate 20 shows the aspect of the Glade with close-ups of two quadrats (1 meter in size) on the up-slope and two on the down-slope. Plate 21 shows the Old Field aspect with two quadrats up-slope and two down-slope. DISTRIBUTION OF JUNIPER SEEDLINGS The diagonals and sides of the Old Field and the Glade were used as transects for seedling counts. All juniper seedlings were counted within one-half meter on either side of the diagonals and one meter to the inside of the area boundaries. Seedlings were divided into three classes as follows: Class 1, 18 inches or less; Class 2, 18.1—36 inches; Class 3, 36.1—72 inches. Total number of seedlings re- ported is based on counts in approximately one twentieth of the total area. In the Old Field, a total of 621 seedlings was estimated, of which 276 were in Class 1, 207 in Class 2, 138 in Class 3. The actual number is probably somewhat higher since the smaller seedlings are rather easily overlooked. The seedlings were not distributed at random, but were either clumped about the mature (fruiting) female trees (with over 90 per cent of them on the down-slope side), or were more or less aggregated at the bottom of the slopes or in other areas where materials carried by run-off water are deposited. Birds play an important role in the distri- bution of seedlings in the Old Field. Many young seedlings were clustered below the branches of shrubs and trees other than junipers. A very few were found on isolated knolls and in the open where they may have been deposited by birds in flight or by ground-feeding birds. The Glade supported a great many more seedlings under four years of age than did the Old Field. The estimate was: Class 1, 6,672; Class 2, 10; Class 3, 12. The total estimate was 6,764 seedlings for the 60,000 square feet of glade exclosure. These were distributed much as in the Old Field, but even more uniformly because of the greater uniformity of the 10° slope characterizing the Glade. The seedlings 19551 HALL—OZARK JUNIPER POPULATIONS 181 were either clumped on the down-slope side of the mature (fruiting) female trees or were aggregated at the bottom of the Glade slope where the Sporobolus neglectus, because of its increasing cover toward the lower edge, acted as a brake to tumbling berry-cones carried by run-off water. There were numerous seedlings in the open, probably from seeds deposited by birds. Great flocks of robins and cedar waxwings are a common sight on the glades, particularly in the fall after the crop of berry- cones has ripened. Fortuitous fluctuations in climate probably add to the seedling mortality. In very wet seasons the seedlings with more genes of red cedar are probably favored only to be wiped out in a later year by drought. Likewise, seedlings possessing a strong J. Asbei component are favored in drought seasons but may suffer defeat in wetter periods. POPULATION STRUCTURE OF JUNIPERUS The variations in the junipers (Juniperus virginiana) of the Glade, Old Field, and Cedar Hill were studied, and five morphological characters were measured and plotted as pictorialized scatter diagrams. Two of the five characters, measured, gland length-width ratio and lateral whip length, were used on the ordinate and abscissa of logarithmic paper (1 X 1 cycle) and the other three as rays on the scatter diagrams. The data for the three characters plotted as rays were grouped as follows: (1) whip-leaf length: long-bar == 8 mm. or more, half-bar = 5-7 mm., no bar == 2—4 mm.; (2) per cent decussate: long-bar = 25—100 per cent, half-bar = 6—24 per cent, по bar = 0—5 per cent; (3) terminal whip length: long-bar == 80 mm. or more, half-bar = 31-79 mm., по bar = 0-30 mm. The scatter diagrams for the Glade, Cedar Hill, and the Old Field populations, as well as for populations typical for the species, are shown in fig. 1. Population means and ranges were obtained from the grouped data (pictorialized scatter diagrams) in the manner described in an earlier paper (Hall, 1952, p. 53). The numbers given have no absolute value but represent a graded scale or index of characteristics. The scale is made up in such a way that low scores belong to J. Ashei, high scores to J. virginiana, and intermediate scores to morphological intermediates or introgressants of the two species. From these scatter diagrams the following population means on this arbitrary scale were obtained: Glade—4.35; Cedar Hill—5.48; Old Field—10.0. Typical J. Ashei has a population mean of 2.0-2.3; typical J. virginiana, Eastern race, has a mean of 9.0-10.0. The two characters plotted on the abscissa and the ordinate for the pictorialized scatter diagrams were also analyzed by using ordered values. The data were ex- pected to fit a log-normal distribution which was confirmed, for the data produced a straight-line function when the ordered values (from smallest to largest) were plotted on logarithmic probability paper. I followed the practical short-cut recom- mended by George W. Thomson of the Ethyl Corporation, which used the range as a measure of dispersion instead of the standard deviation. Statistical techniques using the range are particularly efficient with small samples. This is an excellent method of checking the confidence of data where only an average and the range are available. The confidence limits a, b = х = суду, [Vor. 42 182 ANNALS OF THE MISSOURI BOTANICAL GARDEN OLD FIELD "m оғ LIMITS OF VARIATION IN Juniperus virginiona TYPICAL RACES OF THE SPECIES № “ „Е lu ұ ұ T т Y зуу Ы 2 [ Juniperus чуј i ұ Y T " "ДУ ET + E F E T 5 Е % - 5 [122 T 4 3 D к зо wo в „с сш $00 600 ҡо 960900 è 008.888 wo = а Bi 500 600 100 800 4 TERMINAL WHIP LENGTH \№ xoccussare ceaves СЕОАЯ HILL POPULATION OF GLADE POPULATION OF Juniperus virginiano L. Juniperus virginiona L. і у у i Log é WwW і v че уч М у : i vs | ^L Y ся ' : № LU у ba à sk ә 5 і vi jv v димних V 3. з 9» t i ы қ * = | Wo v Hos Ф о = ийе eem = еее è PNE. с = па = = ы ~ бирои елес й синю Fig. 1. Scatter diagrams showing the variations in five characters d a typical race of mesa о virginiana from southwestern Vee я a Маине гасе а A Asbei from the Edwards P f 'Texas, and of three populations of Lone t Gra mit, Mo. "T e Рави at roue доза left for the typical races shows the means s (large d ots x Eun extremes (small dots), the limits of variation (the curved lines). (The symbol in the space at the upper right of this diagram — have had three long "ic In the diagrams for the three Gray omit populations each dot resents a single individua 1955] HALL—OZARK JUNIPER POPULATIONS 183 where x is the mean, t is Student's t, s is the square root of the estimated popu- lation variance, and n is the number in the sample. Or, by using the short-cut the 95 per cent confidence limits are equal to the range multiplied by the proper value! (0.1064 for n — 25, 0.0720 for n — 42) plus or minus the mean. Table II lists the population means, confidence limits, and standard deviations (log values) for two characters of each of the three populations and typical popu- lations of both J. Ashei and J. virginiana. Figure 2 represents the ordered values plotted on logarithmic probability paper. The slopes of the lines depend on the amount of variation within each group; the greater slopes indicate the more heterogeneous populations. The point of intersection of each line with the 50 per cent value is the estimated mean for each population. The points of intersection of the 25 per cent and 75 per cent values with each line delimit the range where half the values for that character will probably lie. In fig. 3 rectangles are made about the means to correspond to the 50 per cent probability level, and the farther two rectangles are separated the greater the probability that the difference between the corresponding means is not due to chance. The dotted lines indicate the 95 per cent confidence limits of the means from Table II. These five populations were compared to determine the probability that they represent the same population with equal means by means of the t-test. If one assumes the means of two populations to be equal, the sample difference is tested to see whether it is no more than sampling variation from the hypothetical dif- ference, zero. Each of these comparisons led to a large t value, indicating a low probability that they represent samples from a single population. Welch’s modifi- cation of the t-test was used for the comparison because the variances were not homogeneous (Comparison of р-р», regardless of o,°/c,”). TABLE II MEANS AND THEIR 95 то CENT CONFIDENCE LIMITS FOR TWO CHARACTERS IN VE POPULATIONS OF JUNIPERUS A: Number Whip leaf length (mm.) | Gland length/width Populations Class us Means | Conf. limits | Stand.* | Means | Conf. limits | Stand.* sample d d ev. ev. J. Asbei 30 yrs. 25 4.8 (4.0-5.7) | 0.179 1.0 | (0.95-1.05) | 0.045 Glade 35 yrs. 42 2.6 (2.3-3.0) | 0.500 2.2 (2.0-2.5) | 0.145 Cedar Hill 30 yrs. 25 6.8 (5.98.0) | 0.522 3.9 (3.44.5) | 0.187 Old ыы. 20 угз. 25 23 (6.3-8.3) | 0.140 8.1 (7.6-8.8) | 0.093 J. vir 30 yrs. 25 8.8 (8.1-9.6) | 0.088 6.4 (6.0—6.9) 100 *Log values. 1See the Appendix to this paper. The values for these re ЖЕМЕ m worked out for sample numbers from 2 to 1000 by George Thomson and a in Table [Vor. 42 184 ANNALS OF THE MISSOURI BOTANICAL GARDEN - GLAND LENGTH -WIDTH RATIO "ES OLD FIELD з L 1 diss: | | J. Ashei E d cp | | 1 І 5 10 20 30 40 50 то во 90 95 98 PERCENT з cE WHIP LEAF LENGTH E T ш » * he чор = L zio ш | > . LJ о b J. virginiana i cL e е OLD FIELD ж l < ВАН НЕ d б x T o p. ы - LJ Е з ZJ. Ashei s x а Е D | | | | | | | | | 1 1 " Џ = 5 10 20 30 40 50 60 70 во 90 95 98 PERCENT Fig. 2. Ordered values, кы on logarithmic probability paper, for two сћаг- acters s typical M ени s virginiana, J. Asbei, and че n popula- tions of J. virginiana from mmit. Length of he lines delimit the range where vn > cent of the values ја ы two characters will probably lie. 1955] HALL—OZARK JUNIPER POPULATIONS 185 | GLADE -Whip CEDAR HILL -Whip иа em OLD FIELD - Whip | GLADE - Gland CEDAR HILL- Gland Е a OLD FIELD - Gland J Variation in whip leaf length and gland сир eg ratio in the three populations = Ји uniperus vir к from Gray Summit, Mo. The estimated means vertica the means b vertical Ба the dinde delimit the ranges where 50 per cent of the values will probably lie; the length of the horizontal lines delimit the ranges where 95 per cent of the values will probably lie. DISCUSSION In the Missouri Botanical Garden Arboretum at Gray Summit, Missouri, these three populations of Juniperus virginiana called here Glade, Cedar Hill, and Old Field, are each distinct, having considerably different variances and significantly different means. Each is also distinct from either species, J. Ashei or J. Уз вуче which contribute to the variability of these Ozark populations. Yet, these Ozark populations exist within less than a mile of one another, but in bild which, at least in part, may be noted for their differences more than their similar- ities. In the sums of several morphological characters these populations differ from one another in nearly the same degree, so that a crude ranking using five characters [Мог. 42 186 ANNALS OF THE MISSOURI BOTANICAL GARDEN gives arbitrary values of 1.0 for J. Ashei, 2.1 for Glade, 3.0 for Cedar Hill, 4.0 for Old Field, 4.9 for J. virginiana. Where each character is compared separately there is greater apparent difference between populations. Methods utilizing several characters therefore give a more accurate picture of the population. With several characters plotted, scatter dia- grams give a good impression or "bird's eye view" of the variation within a popu- lation even though statistical inferences may not be drawn from them alone. When the populations are compared character for character, one finds that the degree of difference between characters is not necessarily similar for each character. The Glade population may closely resemble Cedar Hill in one character and more closely approach J. Ashei in another. Unequal variability, character for character, is characteristic of species hybrids and introgressant populations. Even when con- sidering variation within the individual, most naturalists know that hybrids, par- ticularly in wide crosses, have more intra-individual variability than do more typical plants of the species. The Glade and Cedar Hill populations, both of the Ozark race, i.e. intro- gressants of J. virginiana by J. Asbei, are the most variable. The Old Field popu- lation is a mixture of the Ozark race and the Northern race (including var. crebra type) of red cedar; but as far southwestward as Gray Summit, Missouri, it occurs sporadically in disturbed habitats where the soil is sandy, light, and well drained. I have not yet seen this northern invader on a glade even though it could con- ceivably occur there within a favorable micro-habitat. The Old Field and Cedar Hill habitats are very similar in every aspect but history of use by man. The Old Field was allowed to erode severely, and conse- quently it has less topsoil and organic matter than the Cedar Hill. The Old Field soil is more acid, sandy and coarser than Cedar Hill soil. Both Old Field and Cedar Hill soils are quite different from the very thin, organically rich, spring-wet, summer-dry soil of the Glade The Northern race of red cedar (including var. crebra) tends to be arenicolous, and even though it develops well on sandy areas overlying limestone or on the strands of marly lakes, it does best on slightly acid, well-drained strand or dune areas where competition with other plants is at а minimum. The Northern race becomes more arenicolous as it becomes more columnar or spire-like. Because of the ease with which junipers may be distributed over great distances, it is not sur- prising to find an occasional mixture of the Northern race of red cedar in the northeastern Ozarks. However, its occurrence there is a result of man inad- vertently preparing a suitable habitat, an eroded, acidic, sandy field with poor cover. Except for the minor role the Northern race plays in introgression with the Ozark race, its effect on the Ozark populations is transient and dependent on the perpetuation of poor land-use practices. Ozark fields which have had fair treatment, particularly where cover has been protected, support stands of the Ozark race alone. In part, the Ozark race is also dependent on man since a large number of its populations occur on cleared or grazed land. 19551 HALL—OZARK JUNIPER POPULATIONS 187 Several other factors affect the distribution and structure of juniper popula- tions. Some of these factors are intrinsic and some extrinsic, but it is impossible to consider them independently. Rate of reproduction, per cent viability, presence of introgressive genes, character of the climatic extremes (whether unfavorable for natural seed stratification, which seems to be the situation at the northern limits of red cedar), amount of bare area in the habitat, slope, soil conditions, presence and abundance of bird and mammal species which aid in seed dissemination, role of run-off water and gravity in seed dissemination, and several other factors occur in different combinations and degrees to produce populations ranging from uniform (as in a cedar-brake) to less uniform (apparently random) to clumped. In gen- eral, when a large enough area is examined there may be found "parent" popula- tions consisting of cores of ancient specimens which supply or supplied the seed stock for the general area. In southwestern Missouri these mother-plants may be found on the bluffs of the White River where they have fruited for centuries safe from fire and other hazards. In northern Missouri the bluffs of the Missouri and Meramec Rivers support populations of ancients. These bluff habitats of the major rivers were probably the original environment for Juniperus long before the Ozark glades were first colonized by junipers. On most of the glades in the eastern Ozarks the oldest junipers are over 100 years of age, probably between 125 and 135 years. Since all the older junipers from which increment borings were taken were hollow at the center, it was not possible to get precise ring counts, but by applying a factor determined by the rate of decrease in years per inch from the center to the outside of an increment, a good estimate of age may be made. Much older trees (judged by circumference) have been found occasionally on bluffs throughout the Ozarks, but those from which increments were taken were hollow and usually with more or less eccentric rings. Such old junipers are not found in Ozark fields. In the early years of the 19th century and previously, written accounts of travels and casual reports by geologists indicate that junipers were to be found only along bluffs of the major rivers, but reports in the last half of that century presented a picture of juniper distribution much as it is today. Wherever the rocky limestone glades are burned each year, they are devoid of junipers and shrubs, giving an open grassland aspect with an abrupt transition to forest at the edges. Also, the junipers do not easily invade a glade that has been protected from grazing and supports a good stand of grass. The treeless glades, which are not uncommon, always have good cover and protection from man’s grazing animals. Where glades show signs of erosion junipers are localized. In general, juniper density is proportional to the degree of land abuse, so that glades may support junipers distributed as dense "brakes", open stands with evenly but widely spaced individuals, or scattered, clumped colonies. Because of the nature of the glades slight grazing could easily weaken the cover to permit invasion of junipers where fire-sere grassland would not be so easily weak- ened. Before white man came to the Ozarks, wild herds probably initiated the [Vor. 42 188 ANNALS OF THE MISSOURI BOTANICAL GARDEN invasion of the glades by junipers; white man tremendously accelerated the process through his domestic animals. In an earlier paper (Hall, 1952) I suggested that introgression between ]. Asbei and J. virginiana may have begun as early as the late glacial period, followed by an expansion during the xerothermic period. Throughout this time, before settlement of the Ozarks, the most continuous habitats for junipers were the bluffs, knobs, rough glades, and youthful streams or rivers with flood-plains in an early stage of succession. Probably continuously the bluffs, knobs, and glades (edaphic sub- climaxes) were inhabited by Ashei-like junipers while the streams and river banks supported scattered stands of virginiana-like plants. e distribution pattern of junipers in the Ozarks has probably changed radically since settlement. It is clear that the oldest junipers are situated on bluffs or glades, while old fields in the St. Louis area may support mature colonies (oldest plants up to 70 years) of the Ozark race or occasionally more youthful colonies (40 years or less) of the Northern race mixed with Ozark. From this study and cursory examination of other fields, it is evident that the condition of the land in large part determines the kind of junipers selected from the local gene pool. Literally, explosive distribution of juniper has occurred in the last hundred years, progressively increasing in amplitude as abused or worn-out land was abandoned to old-field succession. The fact that these neighbor populations, particularly Glade and Old Field, are distinct even though intermediate does not mean that hybridization between them is limited, but that each habitat is selecting the best- adapted genotypes. Natural selection is strong enough to limit the portion of the available gene pool which may be realized. Неге are two powerful forces working against one another. Strong natural selection restricts the gene pool; hybridization amplifies it. With species like those of Juniperus, which hybridize freely and are readily dispersed, I suggest that continuously variable habitats result in continu- ously variable populations, and the simplest situation is when the hybridization is between two allopatric species. Anderson (1948) stated that hybrids between two species differing in habitat requirements are expected to occupy intermediate habitats. Тһе Е; should occupy a habitat more or less intermediate between the two species; the Ро should occupy more variable habitats, but the range should lie between the habitats of the species; backcrosses or introgressants should occupy habitats closest to the backcross parent. Two species may be quite interfertile, exist side by side, yet show little evidence of hybridization between them because the habitats select the parent-type combina- tions. If the habitat is disturbed, these available mongrels may become established. The homes of these variants are the seres either initiated by natural processes or man. While hybrid swarms and hybrid habitats are transient, they have their in- fluence toward increasing variability both through cryptic processes, e. g. structural differentiation, and toward accelerating regional differentiation. If, for the sake of discussion, man as a producer of hybrid habitats is ignored, it will be seen that processes which initiate succession produce hybrid habitats, and indeed the edges 1955] HALL—OZARK JUNIPER POPULATIONS 189 of ecologic systems, physiographic provinces, and climatic regions are hybrid habitats. Ecologic tension zones are hybrid habitats where species which meet may develop discordant variability. When man enters the scene, the pattern becomes more complex. Hybridization may do its cryptic work in tension zones through all levels of ecologic structure. The obvious evidence of hybridization, the hybrid swarms, are transient; and they may occur whenever the right hybrid habitat is available, but in most situations the habitats select gene combinations more like the parents and tend to keep the populations distinct. SUMMARY In the northeastern Ozarks Juniperus virginiana is represented by the Ozark race (introgressants from J. Asbei) on bluffs, glades, and most old fields. Oc- casionally, the Northern race is found locally with a little admixture of the Ozark гасе. Three populations of J. virginiana in the northeastern Ozarks were studied in detail in order to compare their variation patterns and habitats. Variation patterns were compared by means of scatter diagrams for five characters and ordered values, means, 95 per cent confidence limits, and standard deviations for two characters. The t-test modified by Welch was used to compare the means of the populations. Habitats were compared in terms of their histories, slope, soil profile, flora, frequency of species, rough estimates of cover, seedling distribution. The evidence suggests that the three populations—Glade, Cedar Hill, and Old Field—are distinct and differ more or less proportionately to the differences in their habitats. The Glade (Ozark race) is the most southwestern in affinity, more closely resembling a population and habitat of J. Ashei. Cedar Hill (Ozark race) is intermediate between Glade and typical red cedar and occurs on old fields which are in good condition or, farther southwestward from Gray Summit, Missouri, on more worn-out lands. The Old Field (Northern race with a little mixing from the Ozark race) occurs on worn-out acidic and sandy lands in the vicinity of St. Louis and northeastward. Extremely high seedling mortality (much higher on the Glade than Old Field) suggests that strong natural selection restricts the field for variability even in the face of considerable hybridization. The present balance between these two evolu- tionary forces depends on man making and keeping available a variety of suitable habitats. A shift in habitat qualities will result in a shift in the variation pattern of the junipers. Distribution of age classes in the junipers indicates that the bluffs, knobs, and glades have been colonized longest by junipers, followed by the old fields supporting the Ozark race, and last the worn-out sandy, acidic fields supporting youthful colonies of the Northern race. An explosive а of juniper colonization resulting from man's activities seems to have occurred within the last hundred years, growing progressively as land has been worn out and abandoned. It is proposed that man is primarily responsible for the present explosive evolution in Juniperus in eastern North America and that if his influence were removed from [Vor. 42, 1955] 190 ANNALS OF THE MISSOURI BOTANICAL GARDEN the scene, the field of variability would decline as the area of occupation contracts. Without man's help junipers would be more or less restricted to the areas in a natural state of arrested development such as the bluffs, knobs, and glades. BIBLIOGRAPHY Anderson, Edgar (1948). e st the habitat. Evolution 2:1-9. Bits. August P., and Louis G. vois r (1951). The recent ins of forests in the Ozarks. . Mo. Bot. Ga га, 38: 261-28 —, == (1951). The changing forest flora of the Ozarks. Ibid. 38:283—291. Braun, Е. Lucy (1 950). Deciduous for eastern North America. Philadelphia, Dre м рит В. (1942). Тће revegetation o ая —— crop-land in the Cedar Creek cem e and Callaway counties, Missouri. Uni l. Agr., Agr. Ex xp. Sta. Res. Bu S Ralph Oa ong G. еле es Joseph Wie и. Dolomitic glades of S cag Ar Missouri. Anh . Bot. Gard. Etter, Mu G. (1953). us ood A 3 st sng in historical ecology. Ibid. 40:227—257. Fernald, M. L, (1950). Gray’ nual of Botany. Fig th nesters оо ork. Ha il ae T (1952). „Аб шс ‘and hybridization in Juni M Bot. Gard. 39:1—64. Kenoyer, Leslie A. (1927). A study of Raunkaier's т эры 2 pes of frequence. Ecology 1—3 Жаша; © У (19 18). пее statistiques sur les formations végétales. Kgl. Danske Vidensk. Selskab, ра Mo su Carl O. (1950). bs climax, fire, and man. Jour. Range Management 3:16—2 Steyermark, ‘abies A. (1940). Studies of - vegetation of Missouri. y “ tural plant associations и eon in the Ozarks of Misso Field Mus. Nat. Hist, . Ser, 9:349—475. Personal аво (Also, see Appendix 'Thom George W. on. x). маан Е ХУ. (1931). е of North America according to a new classification. Geographical Review 21:633—6 Weaver, John E., and Frederic C. а. (1938). Plant ecology. New York. ExPLANATION or РТАТЕ PLATE 20 Aspect of Glade Fig. 1. Open glade looking south. Fig. 2. Open glade looking north. Figs. 3 and 4. Two Жааны ле in open-glade up-slope. Figs. 5 and 6. Two quadrats in open-glade down-slope. ANN. Mo. Вот. Сакр., Vor. 42, 1955 PLATE TUA AA ^r ek 4 HALL—OZARK JUNIPER POPULATIC ANN. Мо. Bor. Garp., Vor. 42, 1955 РТАТЕ 21 HALL—OZARK JUNIPER POPULATIONS [Мог.. 42, 1955] HALL—OZARK JUNIPER POPULATIONS EXPLANATION OF PLATE PLATE 21 Aspect of Old Field Fig. 1. Old Field begins beyond mowed area. Fig. 2. Old Field, interior. Figs. 3 and 4. Two quadrats, up-slope. Figs. 5 and 6. Two quadrats, down-slope. 191 [Vor. 42 192 ANNALS OF THE MISSOURI BOTANICAL GARDEN APPENDIX ESTIMATION OF CONFIDENCE LIMITS OF POPULATION MEAN FROM RANGE I am very grateful to George W. Thomson for the method by which 95 per cent confidence limits of the mean may be rapidly estimated by multiplying the sample range by a factor. Lord (1947) prepared a table of factors for sample sizes from two to twenty for six classes of probability (90—99.9 per cent). Thom- son prepared a table of factors including greater sample sizes for 95 per cent probability. Thomson’s table and brief discussion are published here for the benefit of other biologists who may wish to use this short method of getting 95 per cent confidence limits of the mean. The usual method of estimating the variability of a population from a sample is by the use of the usual standard deviation estimator, s. However, recent research has shown that much more convenient estimates can be obtained from the sample range, which is defined to be the difference between the smallest and the largest values in the sample. These estimates are not as efficient in the statistical sense, but the loss is not important from a practical point of view. А useful by-product of this research is the rapid estimation of confidence limits of the mean of a normal population by the multiplication of the sample range by a factor. These results are associated with a smaller number of equivalent degrees of freedom than con- fidence limits based on the usual s estimates of the population standard deviation. A close approximation for the multiplying factors was found by Patnaik (1950) who approximated the distribution of the range in normal samples by the х— distribution. If w is the range іп a sample of size n, then w/o = c х/ Му where с is a scale factor and v is an equivalent number of degrees of freedom for x. The following table extends previous calculations (Thomson, 1953) to sample sizes as large as 1000. TABLE Ш SCALE FACTORS AND ia Oe DEGREES OF FREEDOM. FOR Х—АРРЕОХТМАТТОМ RANGE IN NORMAL SAMPLES umb. in Degrees of Scale Equivalent : sample freedom factor two-sided Confidence limits** n » c 5% t + t/(cVn) 2 1.0000 1.41421 12.7062 6.3531 3 1.9846 1.91155 4.3349 1.3093 E 2.9291 2.23887 3.2265 206 5 3.8267 2.48124 2.8267 5095 6 4.6772 253 2.6249 4010 7 5.4841 2.82981 2.5038 3344 8 6.2512 2.96288 2.4233 2892 9 6.9818 3.07794 2.3658 .2562 10 7.6798 3.17905 2.3228 .2311 11 8.3485 3.26909 2.2894 0.2112 19551 HALL—OZARK JUNIPER POPULATIONS 193 TABLE III (Continued) Numb. in Degrees of Scale таом A sample freedom factor o-sided Confidence limits** n » c due + У (cVn) 12 8.9896 3.35016 2.2626 .1950 13 9.6057 3.42378 2.2406 .1815 14 10.1997 3.49116 2.2249 .1703 15 10.7717 3.55323 2.2067 .1604 16 11.3249 3.61071 2.1933 0.1519 17 11.8602 3.66422 2 141» .1444 18 12.3789 3.71424 2.1715 .1378 19 12.8823 3.76118 2.1624 .1319 20 13,3723 3.80537 2.1543 .1266 21 13.85 3.847 2.147 0.1218 22 14.31 3.887 2.141 174 23 14.76 3.924 2.135 1134 24 15.20 3.960 2.129 1098 25 15.61 3.994 2.124 1064 26 16.05 4.027 2.120 0.1032 27 16.44 4.058 2.113 .1003 28 16.84 4.088 2.111 .0976 29 17.23 4.116 2.108 .0951 50 17.63 4.144 2.104 0927 35 19.38 4.268 2.090 0,0828 40 21.04 4.373 2.079 .0752 45 22.61 4.465 2.071 0691 50 24.04 4.545 2.064 0642 55 25.37 4.617 2.058 0601 60 26.59 4.682 2.053 0.0566 65 27.80 4.742 2.049 ‚0536 70 28.91 4.796 2.046 .0510 75 29,03 4.846 2.042 .0487 80 30.98 4.893 2.040 .0466 85 31.96 4.936 2.037 0448 90 32.82 4.977 2.035 0431 ЗЕ 5.015 2.033 .0416 100 34.61 5:032, 2.031 .0402 200 47.3 4521 2.011 0.0258 500 67.4 6.096 1.996 0.0146 1000 85.3 6.502 1.988 0.0097 * All etri are believed to be correct to within 3 units in the last place shown. Values for а = 2(1)10 are from Thomson 241953). ** Multipliers ње range to get 95% confidence limits of normal population mean. The following example demonstrates the use of the method. А sample size of twenty-five log values of whip leaf length from the Cedar Hill population of the [Vor. 42, 1955] 194 ANNALS OF THE MISSOURI BOTANICAL GARDEN Ozark race of Juniperus virginiana had a range of 0.637 and a mean of 0.842. It is estimated with 95 per cent confidence that the population mean lies within 0.842 + 0.067 or 0.775 to 0.909 (5.96 mm. to 8.11 mm.). APPENDIX BIBLIOGRAPHY > Е. (1947). The use of range f standard deviation in the t- test. Biometrika 34:41—67. naik, P. B. (1950). The use of mean range as an estimator of variance in statistical tests. Ibid. 37:78—87. Thomson, еа, W. (1953). Scale factors = = of freedom for small sample sizes for x-approximation to the range. lbid. 40:449 MISSOURI BOTANICAL GARDEN STAFF Emeritus Director OORE CARROLL W. DopcE, | | з Свовсв В. МАМ SCHAACK, Mycologist A Honorary Curato Ковевт E. Woopson, JR. Curator of the Herbarium Henry М. ANDREWS, ndrolc Paleobotanist | = Arme Е. Тихому ої RorrA М. Tryon, до. ; 220222 Research Associate Assistant Curator of the ` ен - Jonn D. Dwyer, - Huc С. CUTLER, | pe С Hone . Pu Curator Museum of : орел | Librarian and Editor — Volume XLII ed Annals of the Missouri Botanical Garden Tassel Modifications i in , Zea m сб л, S "Norton H. Nickerson and Ernest Е Баје 195-212 | А New Species of Doryopteris from Surinam Karl U. Kram | Annals of the Missouri Botanical Garden A Quarterly Journal containing Scientific Contributions from the Missouri Botanical Garden and the Henry Shaw School of Botany of Washington University in affiliation with the Missouri Botanical Garden. Information The ANNALS OF THE Missoumr BOTANICAL GARDEN appears four times ж the calendar Шыр Са February, Мау, September, and November. Four umbers tea v bscription Price __ Su $10.00 per volume Single Numbers Contents of previous issues of the ANNALS OF THE MissoURI BOTANICAL GARDEN are listed in the Agricultural Index, published by the Н. W. Wilson Company. Аппа[$ of the Missouri Botanical Garden Vol. 42 SEPTEMBER, 1955 No. 3 TASSEL MODIFICATIONS IN ZEA MAYS! NORTON H. NIC Department of Botany, feas of jui Amberst and Kerckboff Laboratories ё Biology, California Institute of Technology, Pasadena AND ERNEST E. DALE? Emeritus Professor of Biology, Union College, Schenectady, New York Zea Mays L., in addition to its great economic importance, is preeminent among plants as a tool in the study of heredity. However, even though our knowledge of maize genetics has made rapid and often spectacular advances, certain problems involving basic morphology of the plant have been left unanswered. Foremost among these is that of the pistillate inflorescence, or ear, the structure of which has been discussed elsewhere (Nickerson, 1954). Another equally basic problem involves the staminate inflorescence, or tassel. Anderson (1951) briefly sum- marized botanical knowledge dealing with this inflorescence, and Alava (1952) showed how certain maize races could be characterized on the basis of information obtained from their tassels. Preliminary studies on tassel morphology by the senior author indicate that the somewhat stereotyped and simple construction of tassels is misleading; certain parts apparently have been subjected to reductions even greater than those affect- ing corresponding ear parts. Morphological analysis of the tassel as it occurs atop a typical maize plant is even more difficult than morphological analysis of the ear. One source of information on ear morphology was through studies of certain genetic forms in which particular parts were present in an accentuated form. It was felt that the same technique might profitably be applied to the tassel, since variation in maize tassels is, as in other parts of this specialized grass, variation on a theme. Specimen tassels from the collecton of sixteen tasse] mutants assembled by the junior author were therefore examined. They contained a wealth of material; This work, which was per- in part 1 Massachusetts Agricultur al Experiment Station Contribution 994. i ћ tomic Епег 5 ће а freely shared experience in carefu observation and for his cial ‘reading of = — In addition, they wish to thank Dt. Edgar Anderson for his cooperation in ны projec (195) [Vor. 42 196 ANNALS OF THE MISSOURI BOTANICAL GARDEN a search of the literature revealed that, in addition to the fact that several of them had not been described, many of the known forms had very meager descriptions. Since knowledge of di tassel mutants in themselves will serve both as а prelim- inary step in analysis of tassel morphology and as a basis for their employment in studies involving gene interaction, fuller descriptions of these anomalous forms are given. We have attempted to give below a careful description of how each mutant type affects structures present in tassels and ears of North American corn-belt maize. Plants of each mutant have been bred to various combinations of two widely known corn-belt inbreds, CC5 and 1.317, sufficiently often to allow mean- ingful comparisons to be made between them. It must be remembered, however, that the standard stock is largely responsible for these particular manifestations, and that the background here employed was one found to be well adapted to grow- ing conditions in southern California. Certain forms which might logically be included in this discussion, such as Corn-grass, Teopod, Silky-1 and Silky-2, were not studied. Silkless, a well-known ear anomaly, also was not included, because in our stocks Silkless tassels are indistinguishable from those of Standard plants. It is possible to construct a key by which these tassel mutants may be sep- arated. However, since even in relatively homogeneous stocks, an example of one mutant may sometimes match the description of another, such an artificial device cannot be too heavily relied upon. А key is here intended only to show certain general trends among the sixteen forms; it may or may not indicate underlying genetic and physiological similarities. The several descriptions given later are arranged in the numerical order indicated in the key. KEY TO CERTAIN TASSEL MUTANTS A. Tassel of normal (Standard) or nearly normal proportions B. Tassel with silks C. Pollen shed; fruits formed D. Fruits on proxi mal екы areas from imperfect florets Tss — 6 D. All fruits from perfect jh — $ illea ids no fruits за на == 9 В. = without silks Glumes undeveloped Үс — 13 с Chemis well develope D. Glumes long (average 11.5 mm.) Ти — 14 D. Glumes near normal length (average 8-10 mm.) E. No awa ж та — 10 E. Pollen s F. Up to % or A of branch tips sterile; pedicellate spikelets often with — eem о — 11 пећ e never sterile; pedicellate spikelets not branched A. Tassel 4 [че ү. portions B. 24 with silks Half or more of tassels e те staminate spikelets о to middle areas or outer halves of branches; spikelets with no more than two florets each D. Branches indet terminate e, mostly ending in immature саа. на Т роза Та — 3 D. Branches othe E. Spikelets loosely spaced on iquam fsa m | E. әліде crowded on bra — C. Tassels either peu э. p or mixed; if mixed, staminate spikelets M with more than two flor 1955 NICKERSON & DALE—TASSEL MODIFICATIONS IN Zea Mays 197 D. е бе fen ужа on both sessile and pedicellate spikelets; average д — 7 D. 5 dores AN on pedicellate spikelets only; average peduncle length more E с set very sparsely if at all; average glume length 6 mm hi e 4 set ей noticeable quantities; average glume ае 8.5 тт в — 5 ks . Seed B. тої without G nches n or slightly lax fioi reference to central culm ra, — 15 С iure upright, close to central cu таз — 16 Certain measurements have been made on at least fifteen and often twenty-five plants of each mutant form, and averages of these data are given in Table I. Peduncle lengtb is the distance from the uppermost leaf-bearing node to the node at which the lowermost tassel branch arises. Tassel length refers to distance from the node at which the lowermost tassel branch arises to the tip of the central culm. Branching area is that distance along the central culm from the lowermost branch node below the central spike; it is included within the tassel length. Primary branches are axes of the second order, when the main culm of the plant is con- sidered to be an axis of the first order. The percentage values for peduncles and branching areas were obtained by dividing these respective lengths by the sum of peduncle and tassel lengths. These same data are presented graphically in fig. 1 by means of a pictorialized scatter diagram (Anderson, 1949). The diagram is meant only to indicate the average extent to which each of the mutants departs from average measurements of Standards. Limits were chosen so that Standard averages would always be char- TABLE I AVERAGE MEASUREMENTS OF CERTAIN MORPHOLOGICAL FEATURES IN TASSELS OF NORMAL AND MUTANT MAIZE FORMS E 2 o ~ 4 -- [is «| |57 81% eg |95 | 38 | 8 Еогт 3.8 9 85 мы» 58 9.9 Е ti = d 5 я 8 = "2 o “= 3% % Б FC DER T Ба ДР Se а 2% Er DM TIS. я B.S аз Za | SE |556 Standard 17.1 35.0 10.4 14 3 20 (1) Tassel-seed 1 (#51) ; 23.4 i 21 29 (2) Tassel-seed 2 (#52) э 26.4 9.7 21 16 31 (3) Tassel-seed 3 (Tss) 22 22.5 9.0 26 15 34 (4) Tassel-seed 4 (#54) 16.2 25.2 19.1 34 39 46 (5) Tassel-seed 4* (Ts4^) 14.4 24.8 16.3 43 37 42 (6) Tassel-seed 5 (Т) 13.0 33.5 12.4 24 28 27 (7) Tassel-seed 6 (Тз) 10.0 18.5 12.0 36 35 42 (8) Tassel-seed 7 (із) 17.9 33.1 15.0 23 35 29 (9) Tassel-seed 8 (1) 18.0 31.9 8.8 14 36 18 (10) Male-sterile 1 (та) 19.0 33.3 10.8 18 36 21 (11) ммен (bd) 15.1 34.4 10.7 18 31 22 (12) Club 13.9 30.3 10.8 20 31 24 (13) Vestigial бій (Ух) 14.9 35.3 117 20 30 23 (14) Tunicate (Tu) 13.7 33.3 12.1 22 29 26 (15) Ramosa 1 (ға!) 14.6 33.8 8.0 72 30 58 (16) goin 2 (таг) 12.0 27.7 21.7. 67 30 55 [Vor. 42 198 ANNALS OF THE MISSOURI BOTANICAL GARDEN acterized by long flags. Thus, similarity to Standard in both graph position and flag lengths means that a particular mutant departs relatively slightly from a normal form. The converse is also true; the farther away in position and the greater the flag deviations from Standard, the more a given mutant differs in form. eo 7. Ts, \ ; STANDARD TS, 15Г 4 8D Ya 13,8 ^ ы CLUB "Y > Ts; о ЛА? є P» T 5% f= o z ш «І ш TS, = Sr Y 152 = ~ а Ts, ш n. ———————— do eo es 30 35 TASSEL LENGTH IN CM. PERCENT PEDUNCLE е s f£ NO. PRIMARY BRANCHES Nite PERCENT BRANCHING AREA Ж xm BRANCHING AREA IN CM. р. 1. Fig. 1. Pictorialized scatter diagram of information contained in Table I, showing relationships between ко aa features in tassels of normal and mutant forms of maize e text. explanation in t 1955] : NICKERSON & DALE—TASSEL MODIFICATIONS IN Zea Mays 199 DESCRIPTIONS Standards, 1317 and СС5 (pl. 22, fig. 1).— Various mixtures of these two corn-belt inbreds were used. The history of Lancaster Surecropper, from which M. T. Jenkins developed L317, has been re- viewed by Anderson (1944b). Wisconsin Inbred CC5 has an ancestry in the open-pollinated maize which was common to that section. Anderson and Brown (1952a, 1952b) showed that these corn-belt forms were various mixtures of Northern Flint and Southern Dent races. Nickerson further pointed out (1953) that these two races had previously been mixed in some degree, and that the excellent hybrid vigor manifest in present-day crosses, such as those between these two standards, is based on rather small differences. Ears are 14- to 18-rowed, 8 to 10 inches long, and bear bright yellow, dented kernels. 1. Tassel-seed 1 (15) (pl. 23, fig. 1).— First described as "tassel-ear" by Emerson (1920), this mutant was classified as Tassel-seed 1 by Emerson e£ a] (1935), and assigned by them to Chromosome 2. Emerson originally described the inflorescence as being completely pistillate, com- pact and distinctly ear-like, with “glumes and palae short, broad and rounded, in all respects much like those of true ears." He noted that these characters could be observed in both mature and immature tassels as well as in poorly pollinated mature ones. It may be added that each pair of pistillate spikelets is associated with a well-developed cupule (Nickerson, 1954). The branches are thickened so that they appear as wide green bands in adaxial view. In our cultures, as in Emer- son’s, seed set was good and very often both florets of a spikelet develop a caryopsis. This development obscures any evidence of regular rowing in the central spike. Emerson noted the general plant weakness in comparison to sibs, profuse silk production in the tassel, a general lack of elongation in upper internodes, and early assumption of a pendant position by the tassel. In our stock, this pendant position is due to bending of the culm; in Tassel-seed 2, likewise a pendant form, it is attributable mostly to bending of the tassel proper. These differences are mani- festations of the degree of condensation (Anderson, 1944a), at least in part. Spikelet pairs of Tassel-seed 1 are generally closely spaced, resulting in rigid branches, while those of Tassel-seed 2 are more widely spaced, resulting in lax branches. Emerson mentioned these relative spikelet-pair densities, but he did not call attention to their differing effects on the area of bending. However, his figures 10 and 11 show each form quite well. In our stocks, branch tips most commonly end in sterile zones characterized by closely overlapping glumes. Occasionally at the very tip, these glumes simulate those of male spikelets; they sometimes even contain stamens, but no pollen shedding has been observed. [Vor. 42 200 ANNALS OF THE MISSOURI BOTANICAL GARDEN The ear of Tassel-seed 1 is slow to develop, and, as in Emerson's strains, fails to set fruit unless the tassel is either removed or poorly pollinated. It is normal except that, just as in the tassel, both florets of each spikelet often develop, obscuring evidence of regular rowing. 2. Tassel-seed 2 (ts,) (pl. 23, fig. 2).— Emerson (1920) originally described this mutant, located on Chromosome 1 (Emerson e£ al, 1935), as "tassel-seed," and noted that the tassels were loose, like those of normal plants, with individual spikelets more or less separated. Our material never showed individual spikelets, but always paired spikelets. In Emer- son's specimens, spikelet pairs sometimes occurred more densely, "but not ear-like in any way." Although it rarely occurred, he noted that staminate flowers might develop with pistillate ones throughout the entire tassel, but he did not determine whether these male florets were functional Не also observed that glumes and "palae" (paleas and lemmas) of such male spikelets were long, narrow, and pointed, as in normal tassels, while in female spikelets these parts were shorter, broader, and more rounded. In our material, spikelets were either female or perfect, no male spikelets having been observed. Оп perfect spikelets, glumes were elongate and near normal. Also, each spikelet pair was associated with a cupule which was often elongated above spikelet pairs located on central spike tips. Kernels developed on nearly all spike- lets; in many, as with Tassel-seed 1, both florets were functional and two kernels were formed. Each tassel branch is thin, with spikelet pairs more widely spaced in its distal portion, but the base may be as thickly set as any found in Tassel-seed 1. Tassels of this mutant are pendant, but differ from those of Tassel-seed 1 plants in that the bending is mostly accomplished in the tassel rather than in the supporting culm. This point is discussed further with regard to Tassel-seed 1 above. Ears of Tassel-seed 2 develop to a degree depending on the fate of the tassel. If little ог no successful pollination takes place, ears are developed rapidly, but always later than on normal sibs. If tassels are removed soon after their appear- ance, ears are formed about the same time as in normals. These results are in accord with the findings of Emerson. Like most other tassel-seeded forms, second florets in many spikelets set fruit, after the manner characteristic of Country Gentleman sweet corn (Weatherwax, 1916), so that regular rowing may be obscured. When ears develop, they аге of normal size. 3. Tassel-seed 3 (Ts,) (pl. 24, fig. 1).— The first published mention of this dominant mutant was made by Phipps (1928), who stated that “а third type of tassel seed designated as Tassel-seed 3 has been studied by Emerson, but the data have not as yet been published." Emerson e£ al (1935) also listed Tassel-seed 3 as being an unpublished discovery of Emerson’s. They noted that this form was similar to Tassel-seeds 1 and 2, except that the inflorescence was generally mixed pistillate and staminate, and that usually pollen could be obtained. The gene is located on Chromosome 1. 1955] NICKERSON & DALE—TASSEL MODIFICATIONS IN Zea Mays 201 Tassel-seed 3 tassels are easily recognizable, but, like other forms, they are variable in appearance. Branches bearing distichously arranged pistillate spikelets resemble flat green ribbons on their adaxial surfaces; both branches and central spikes most often end in tapering sterile rudimentary pistillate spikelets. Often, these sterile spikelets have an appearance of being whorled rather than distichously arranged. About three-fourths of our specimens produced some spikelets from which the florets shed pollen. These staminate spikelets are imperfect, most com- monly produced on proximal portions of branches and central spikes and sometimes separated by sterile zones from pistillate regions. An occasional branch is entirely staminate and ends with a zig-zag axis, but most branches which bear staminate spikelets end in the characteristic pistillate tips described above. ‘Tassels also are lacking in stiffness, thus becoming pendant as soon as they are exserted. The upper three to five internodes of Tassel-seed 3 plants do not elongate to the degree common in Standards, and as a result a rather tight collar of leaf sheaths is formed covering the lower sixth of each tassel. One tassel in our culture was infected with smut (Ustilago zeae), a probable reflection of the extreme growth and lasting succulence of tassel parts. Emerson (1920) observed, and Dr. E. G. Anderson agreed, that Tassel-seeds 1 and 2 were highly susceptible to smut, but this year we noted no infections in any of our stocks except Tassel-seed 3, There were often instances in which a spikelet pair consisted of one staminate and one pistillate spikelet; in these mixed pairs, the staminate one is always pedi- cellate. No perfect florets were observed, and only one floret in any particular pistillate spikelet would form a caryopsis. Cupules are well-developed adaxial to both pistillate and mixed spikelet pairs. The ear of Tassel-seed 3 shows very few secondary florets, although Emerson et al (1935) mentioned that such florets do ет In other respects, the ear resembles that of a Standard plant. 4. Tassel-seed 4 (ts,) (pl. 24, fig. 2).— First reported by Phipps (1928), and also as Sorghum Tassel (Hayes and Brewbaker (1928), /s, was assigned by Emerson e£ a] (1935) to Chromosome 3. hipps described the tassel as being tassel-like in structure but predominantly pistillate-flowered, and stated that when it emerged it was a mass of silks. Our material showed tassels much shorter than normal, with an abnormal number of weak branches held upright by a tangled growth of silks. In Phipps' stock, mature tassels were studded with kernels, usually densely packed together, especially on the central axis of the tassel, and not arranged in regular rows. Our specimens varied from setting fruit to this extent to setting no fruit at all. Phipps interpreted this irregular rowing as a result of development of a second floret in each spikelet; with this finding we are in agreement. In both his material and in ours, tassel branches were short and had distichously arranged spikelets, the spikelets on the proximal branch and the lower central spike producing stamens which contained functional pollen. Tassel glumes were short and papery, but we cannot agree with (Мог. 42 202 ANNALS OF THE MISSOURI BOTANICAL GARDEN Phipps that they were similar to glumes of a normal ear; we observed no glume induration. We noted also that sessile spikelets rarely branched, while pedicellate ones often formed short branches bearing two to six spikelets. Our standard stock has an average condensation index of 1.3; spikelets may occur at a given node in pairs, in 4’s, in 3’s and 5's and occasionally in 6's. As in Phipps’ stock, both florets of each spikelet always developed, and thus each node had at least four and oc- casionally as many as twelve florets in addition to possible extra ones formed on branches of proliferated pedicellate spikelets. Phipps mentioned that careful study of his material showed “а few multi-flowered spikelets.” Silks were produced only from pistils, and such florets were always imperfect. No glumes were modified into silks, as Phipps reported. It is doubtful if his ob- servation is accurate, because he noted “the basal parts of such glumes were normal.” The silk-like structures were most likely awns, and if so, they were probably produced not on glumes but upon lemmas. We observed no such struc- tures in this mutant, but in a mutant combination not reported on here, awns are developed on the lemmas. In some tassels, spikelets were sterile, producing neither silks nor stamens. These sterile spikelets were most common on branches between the proximally located obviously staminate areas and the distally located obviously pistillate areas. Characteristic of our Tassel-seed 4 stock were both its generally light set of fruit in the tassel and production of staminate and pistillate florets in widely varying amounts. No cupules were developed adaxial to any spikelet pairs of the tassel. Ears of Tassel-seed 4 were often shorter than normal because of failure of the tip to mature; they developed better if tassels were removed early. In our stocks, no excessive development of silks was noted, but Phipps reported that some glumes were modified into silks. Development of second florets, referred to by Phipps as common, was apparently much less common in our stock, and regular rowing was not obscured. Glumes were usually less indurated than in normal ears. 5. Tassel-seed 42 (#5,7).— This allele of Tassel-seed 4 was found by E. G. Anderson in a background stock of CC5 X L317 which had been exposed to radiation at Eniwetok. Crosses made by the junior author between this mutant and Tassel-seed 4 showed the two types to be allelic. The tassel of Tassel-seed 42 resembles that of Tassel-seed 4, but the quantity of seed set is much greater. Pollen is much more freely produced than in Tassel- seed 4. Pistillate florets which set fruit are always borne on pedicellate spikelets and are generally imperfect, but some perfect florets have been found which always appeared to set fruit. In general, pistillate florets are confined to the outer third of branches. The central spike is sometimes all staminate, with several extra florets produced on short branches which are proliferations of pedicellate spikelets. Cupules are sometimes weakly developed on the culm of the central spike, but their presence is not universal. Average lengths of glume are about half again as great as those of Tassel-seed 4. 19551 NICKERSON & DALE—TASSEL MODIFICATIONS IN Zea Mays 203 In Tassel-seed 4* a few pedicellate spikelets have a tendency to proliferate slightly in the upper (distal) third of the ear. After kernels are removed, the right- angled insertion and excessive lengths of the glumes, paleas and lemmas of this allele give the cob a markedly different appearance from that of Tassel-seed 4. 6. Tassel-seed 5 (Ts;) (pl. 25, fig. 1).— Although this mutant is attributed to Emerson (1932), the only mention of Tassel-seed 5 in his paper was its inclusion on Chromosome 4 of a linkage map (see his fig. 1, p. 145). Emerson e£ al (1935), in a brief description, noted that it contained both silks and anthers but was not compacted as in Tassel-seed 4. In our material, tassels were very close to normal proportions, but they were never exserted as far as in normal sibs. Apparently this condition was due to lack of elongation of the internode directly below each tassel, as the plants appeared to be normal in height (Table I and fig. 1). Silks are much shorter and more scattered than in previously discovered tassel-seed forms. Their occurrence varies from very few being located in small branch areas (either basal or terminal), to being uniform over the entire tassel. Spikelets may bear either staminate, pistillate, or perfect florets, but these types occur in specific places. Imperfect pistillate spikelets are usually located in proximal parts of lower branches, and when they occur on central spikes, it is only in their lower regions. Cupules are well developed adaxial to such spikelets, which most often occur as sessile members of spikelet pairs; the pedicellate ones are always both imperfect and staminate. The imperfect pistillate florets form most of the caryopses. Perfect spikelets are never associated with well-developed cupules, and seldom set fruit unless there are few or no imperfect pistillate spikelets present. Even here, however, fruits are formed predominantly on sessile spikelets. If a fruit is set in such a floret, the stamens are generally not exserted, but examination shows their anthers to be full of pollen and occasionally dehisced inside the palea and lemma. Imperfect staminate spikelets occur wherever the other two forms are absent; they never have any trace of silks, and their stamens shed pollen copiously. Silks not pollinated quickly withered. The ear of Tassel-seed 5 in our strain developed very few second florets, but Emerson e£ al noted that "secondary florets develop in ears." Otherwise, the ear closely resembles that of our Standard stock. 7. Tassel-seed 6 (Т) (рі. 25, fig. 2).— The first apparent reference to this dominant mutant was made by Emerson (page 14 of the Maize Genetics Cooperation Newsletter for January, 1953). In subsequent ‘Newsletters’, he established that this mutant is located on Chromosome 1, but no description of its morphology has been published. Plants of Tassel-seed 6 average about one foot shorter than normal sibs. Tassels appear at the same time, but they are only half as long and are borne on peduncles whose average length is only three-fifths as great as those of Standards. In general [Vor. 42 204 ANNALS OF THE MISSOURI BOTANICAL GARDEN appearance, the tassels resemble those of Tassel-seed 4, but they usually set a considerably greater number of fruits. Branches are numerous, short, thin, and lax, with no cupules developed adaxial to any spikelet pairs. They end in many small florets which rarely form fruits and which may be borne on whorled instead of distichous spikelets. Spikelet pairs are closely set on both branches and central spike. Scarcely any distinction exists between pedicellate and sessile individuals, as each one forms several closely packed florets. Short branches with 4 to 12 spike- lets are developed on the central culm, giving it a thickened appearance. Glumes are hyaline, without chlorophyll, and extremely short, averaging about 5 mm. in length. Tassels of this mutant have two general forms. Approximately half of them are entirely pistillate; the others produce staminate spikelets on proximal branch areas and on the lower third of their central spikes. Tassels of this latter sort may have their remaining spikelets constructed in one of three different orders which occurred in our samples with equal frequency: (1) florets bearing silks might also bear stamens and thus be perfect; (2) they may bear no stamens and be imperfect, or (3) a zone of perfect florets of variable length may grade off into a tip which is entirely imperfect. Stamens in perfect florets often neither extruded their anthers nor shed pollen, but those in staminate florets invariably shed good quantities of pollen. Silks are produced only from pistils, but an occasional lemma near a branch tip may develop a short awn. Ears of Tassel-seed 6 are not well developed unless the tassel is removed early. There is the same characteristic production of second florets found in other tassel- seeded forms, so that regular rowing is often obscured. In other respects, ears are comparable with those of Standard plants. : 8. Tassel-seed 7 (tsı) (pl. 26, fig. 1).— This mutant was found by Е. С. Anderson at Pasadena in material which had been exposed to radiation at Bikini. (See Anderson её al, 1949, for a further dis- cussion of radiation effects.) It is not known to which chromosome it should be assigned. Tassels have proportions and dimensions of normal ones, and, like Tassel- seed 5, have few functional pistillate florets. Silks are generally short and evenly distributed; their occurrence varies from almost none to profuse. Each silk arises from a pistil in a perfect floret; florets which do not produce silks are imperfect and staminate, producing pollen freely. Only two florets are borne on each spike- let. No cupules are developed in the tassel. The number of fruits seldom exceeds twelve per tassel, and on about half the tassels they are not formed at all. The ear of Tassel-seed 7 commonly bears up to six short branches at its base. These branches are produced from pistillate spikelets, and an adaxial cupule is present above each one. The sessile spikelet of such a pair is generally not fully developed. These basal branches do not form any viable kernels. Silks form only slightly if at all, even though the branches remain meristematic after cessation of growth elsewhere throughout the plant. Occasionally, two or three small 1955] NICKERSON & DALE—TASSEL MODIFICATIONS IN Zea Mays 205 branches are found on the shank below an ear. These sub-ear structures have always been sterile, and none of them developed beyond the point at which pistillate spikelets could be easily recognized. 9. Tassel-seed 8 (153) (pl. 26, fig. 2).— This mutant, found by E. G. Anderson at Pasadena, was segregating in a cul- ture of chromosome translocations. It is known to be linked to white endosperm (э). Both #5; and ms, tassels are male-sterile, in chromosome 6 and linked to Yı» but the two forms are not allelic. Further, since T's, sets no fruit but produces silks and since Emerson e£ al (1935) mentioned no male-sterile forms which so behaved, the material is here described as an eighth tassel-seed form. In over-all appearance, plants resemble normal sibs. Tassels are of standard proportions and size, but glumes are occasionally shorter than normal and the spikelets never swell as time for anthesis approaches. Silk production varies from none to a profusion as great as may occur in Tassel-seed 7. Four-fifths of our population of this mutant had tassels in which no stamens were formed; in the remaining one-fifth, they were formed but their anthers were empty. No stamens are ever exserted. Silks arise from abortive pistils. Three florets per spikelet may send out silks, but no fruits are ever set regardless of ample exposure to pollen. Branch tips are often undeveloped. The ear of Tassel-seed 8 produces a great profusion of silks; 4 to 6 are formed by each spikelet. The first silk formed (from the lowest flower) is of greatest diameter, and its pistil forms a caryopsis. Each of the others sends out a silk which generally is exserted, but no fruits are formed. Kernels are arranged in rows, and, except for the fact that many extra silks protrude between the tightly packed kernels, a mature ear resembles that of a Standard plant. 10. Male-sterile (ms,) (рі. 27, ће. 1).— This mutant was reported by Singleton and Jones (1930) and listed by Emer- son e£ al (1935) as occurring on Chromosome 6. Singleton and Jones noted that _ tassels of male-sterile plants were more slender than normals, that anthers failed to extrude, and that no pollen was shed. They also reported that Beadle found meiosis essentially normal but that the haploid nucleus did not go through the first mitotic division, and by the time pollen should be shed it was difficult to find even the remnants of spores. Emerson e al noted further that shriveled anthers were exserted much later than in normal sibs. In our material, no stamens were exserted. In other respects, tassels of ms, resembled those of our Standards. There was a tendency for the tassel to emerge sooner than with normals, and this tendency was reflected in slightly longer pedicels (Table I and fig. 1) since elongation ceased at the same time as in normal 51 [Vor. 42 206 ANNALS OF THE MISSOURI BOTANICAL GARDEN The ear is normal. Singleton and Jones reported that ms, is closely linked to the white endosperm locus (y,) with about 5 per cent recombination. Our material behaved in the same fashion; white kernels almost invariably produced male-sterile plants. 11. Brancbed-silkless (bd) (pl. 28, fig. 2.)— First described by Kempton (1934), this mutant is listed by Emerson e£ al (1935) as occurring on Chromosome 7. Kempton noted that the division of the tassels into a central spike and branches is as definite as in normals, but that the tassels have a thickened appearance suggesting those of Tunicate plants. Не at- tributed this thickening to development of short branches in place of paired spikelets, a condition which also occurred in our material. He also mentioned the fact that this branching was most common on central spikes, and that on the branches there was more tendency toward retaining a pedicellate-sessile form. Kempton made no mention of sterility, nor did his illustrations show any, but in our material sterile areas occurred at branch tips and sometimes at the tip of the central spike. Not uncommonly lowermost branches were sterile throughout. In sterile areas, spikelets were less developed as distance from the central culm in- creased. Kempton noted that many plants shed pollen in good quantity, but stamens often were not fully exserted. ars of Branched-silkless are of two forms, a fact which Kempton likewise noted; his figures show each type clearly. In one type, each spikelet primordium develops into a short branch bearing female spikelet rudiments which never develop beyond the earliest stages. In the other type, such branch development is con- fined to the basal quarter of the ear, the upper three-fourths being composed of sterile spikelets with elongated and sometimes thickened glumes. That these branches develop from spikelet primordia on an otherwise normal cob can be easily shown by removal of the branches; underlying each pair adaxially is а well- developed cupule. It is this latter type which may occasionally produce a scatter- ing of silks too short to become exserted. Kempton stated that his material was wholly without silks, but Emerson е/ al mentioned that the ear occurs "often without silks.” 12. Club (club) (pl. 27, fig. 2).— This mutant was first noted by E. G. Anderson and segregated from non- irradiated genetic stocks at the Caltech farm in Arcadia. Its chromosome length is own, The tassel is shorter than normal, with branches held at angles approximating 45°. Branch ends do not droop and their thickly set spikelet pairs are fertile throughout. The central spike is exceedingly thick, and it is on account of this character that the mutant was named. This thickening is brought about by a shortening of internodes, so that the spikelets stand nearly at right angles to the central culm. Further, these closely spaced spikelets occur singly or in pairs; there is no pedicellate branch development such as is found responsible for creation of thickened portions of the central spike of Branched-silkless. 19551 NICKERSON & DALE—TASSEL MODIFICATIONS IN Zea Mays 207 The ear of Club generally matures a week later than the tassel. Its silks are profuse, and there are 4—6 silks per spikelet, each from a rudimentary pistil—a condition also found in ears of Tassel-seed 8. However, Club ears are fasciated, the branching strongly resembling a type listed by Kempton (1923) as Bearsfoot (see his fig. 1 for a clear illustration). It is apparently caused by one or more incomplete divisions of the growing point when the ear is partly formed, followed by simultaneous development of each new point into a more or less independent ear tip. The upper half of such an ear is generally hollow, and no spikelets are formed on the inside walls. 13. Vestigial glume (Vg) (pl. 28, fig. 1).— First reported by Sprague (1939), Vestigial glume was found to be a dominant mutation on Chromosome 1. It is easily noted in the tassel, where the hyaline outer glumes range from awl-shaped vestiges to nearly one-third normal length. Sprague pointed out that “flowering glumes” (lemmas) and "palets" (paleas) were also reduced in size, with the result that the stamens were nearly completely exposed. In our material, as in his, anthers generally dried up before dehiscence, but occasionally a plant shed viable pollen. Many specimens in our stock had sterile areas (i.e., no stamens were formed) at tips of branches and central spikes, in some cases involving the outer sixth of all branches. Generally, over-all tassel size was smaller and tassel construction, as exemplified by thickness of central culm and branches, was lighter than in normal sibs. The ear of Vg is easily recognized, as Sprague pointed out, after removal of kernels. There are no chaffy upper and lower glumes, paleas or lemmas present, only short pedestals, each with a low, hard ridge on both upper and lower sides. Above each spikelet pair, in an adaxial position, is a cupule with somewhat reduced rachis-flaps (Lenz, 1948). 14. Tunicate (Ти) (pl. 22, fig. 2) — According to Collins (1917a), this mutant has been reported from several sources and its origin is not known. Recently, it has been cited in connection with theories of maize origin (Mangelsdorf, 1948; Mangelsdorf and Smith, 1949). Cutler (1944) suggested that its widespread occurrence among Indian tribes was because of mystical significance attached to it. Emerson et al (1935) listed the character as occurring on Chromosome 4. Tassels of heterozygous Tunicate plants have normal measurements in all visible parts except the glumes. The glumes are one-fifth to one-fourth longer than those on Standard tassels and enclose paleas and lemmas similarly elongated. The stamen length being normal, the anthers, especially in the central spike, are not fully exserted and therefore are not pendant at anthesis. Pollen is shed abundantly. Ears of heterozygous Tunicate plants are easily classified. Collins (1917a, b) stated that "the glumes of the female inflorescence, or ear, are developed so that each seed [fruit] is entirely enclosed.” Emerson et al (1935) likewise attributed [Vor. 42 208 ANNALS OF THE MISSOURI BOTANICAL GARDEN these covers to glumes. In our stock, paleas and lemmas were also elongated, and in many basal spikelets, paleas and lemmas of both upper and lower florets were noted. No fruit occurred in these lower florets, however, and regular rowing was externally apparent. The rachis of a Tunicate cob is rather flexible. Cupule development adaxial to each spikelet pair was not excessive in comparison with Standards; but in the Tunicate ear of Guarani maize sent to the senior author by Dr. P. C. Mangelsdorf, the cupules are much more apparent than is normally the case. Homozygous Tunicate plants, according to Emerson e£ al (1935), are usually female-sterile. Мо homozygous plants were available in our collection. An allele of Tunicate, tu*, was reported by Mangelsdorf (1948) as being present іп Maiz chapolote. How this allele would behave with our Standards as background was not determined. 15. Ramosa 1 (ra,) (pl. 29, fig. 1).— This form was originally described by Gernert (1912) under the name of Zea ramosa, and was listed by Emerson e£ al (1935) as Ramosa-ear 1, located on Chromosome 7. Gernert's description of the tassel was brief. In his stocks, tassels were slightly smaller than normal, invariably much branched and cone-shaped. Collins (1917a, b) noted that branches were much more numerous than normal. Kempton (1921) reported 400 as an extreme number; the branches gradually de- creased in size upward, the transition from branches to pairs of spikelets being imperceptible. A short central spike was characteristic in our specimens—a fact which Kempton had likewise noted in his plants but which Collins did not mention. It was not possible, according to Collins, to distinguish between plants heterozygous for Ramosa 1 and normals. However, in our material, it was quite simple to sep- arate them; in five F, families of Ramosa 1 Х Standard, four with 10 plants and one with 9 plants, separation was easily made into 24 normal and 25 heterozygous individuals. Resemblance of tassels of homozygous Ramosa 1 plants to grass panicles is striking (Kempton's plate 13 is typical). Tassels of plants heterozygous for this character were intermediate between a pyramidal paniculate tassel and a normal one, identical to one figured by Kempton in his plate 14. The ear of Ramosa 1 was described by Gernert as being much branched, with- out male florets, covered with husks, and composed of a mass of kernels borne on numerous irregular branches. His description applies to our specimens as well, except for two other points: (1) branches on a Ramosa 1 ear are about as num- erous and have much the same irregular whorled arrangement as do those of the tassel; (2) there are no adaxial cupules at the junctions of ear branches and central culm. Cupules were present, however, adaxial to each spikelet pair of the branches. No other ear among the mutants in this collection is so organized. It was also a common occurrence for tips of those branches originating near the upper part of the ear to be sterile. 1955] NICKERSON & DALE—TASSEL MODIFICATIONS IN Zea Mays 209 16. Ramosa 2 (ra,) (pl. 29, fig. 2).— The only published reference to this mutant is found in Emerson ef al (1935), which simply listed Ramosa 2 as being located on Chromosome 3 and credited its discovery to Brink. Ramosa 2 has a tassel characterized by stiff upright branches which remain closely appressed to the central culm. The central spike region does not bear short- pedicelled spikelet pairs; instead the spikes are borne on branches which decrease in length from base to apex. There is a pronounced transition from well-developed tassel branches bearing many pairs of spikelets to smaller branches bearing fewer spikelets, some of which appear to occur singly. As one proceeds acropetally, these multi-spikeleted branches are replaced by stalks each bearing only one pair of spikelets which are themselves borne on pedicels longer than normal In the adaxial area of each of these branches and stalks, a small cupule-like depression in the central culm can be observed which is generally of a different color from the surrounding surface. Although these depressions and short stalks are not confined to Ramosa 2, they are here most common and best developed. The spike- let pairs in proximal areas of well-developed lower branches are likewise stalked and also have abnormally long pedicels, but there is no adaxial depression. Pollen is shed freely, but much of it remains within the confines of the tassel because close proximity of branches does not allow unrestricted air passage. Each spikelet has two male florets, and each of these has three functional stamens. Tassel-seed 42 might possibly be confused with Ramosa 2, as its tassels sometimes show the same tendency to have numerous stiff, erect branches and stalked spikelet pairs in the central culm region, with an abundance of pollen produced. It can be separated from Ramosa 2 by the lack of stalked spikelet pairs on lower branches and the fact that each of its branches end in a few sterile undeveloped spikelets. The ear of Ramosa 2 sometimes ends in a staminate structure looking like a normal central spike. Most commonly, it produces scattered branches on the upper (distal) half of the cob, which in turn bear female spikelets in pairs. These branches are nearly always found in younger ontogenetic stages than the rest of the plant and consequently set seed only rarely. А branch originates by growth of what would normally be the pedicellate spikelet of a pair of kernel-bearing spikelets. No instances were observed in which the sessile spikelet of such a pair formed anything but a normal caryopsis. The spikelet pairs themselves are borne on short stalks; the cupule is bent at nearly right angles, and is adherent half to the stalk and half to the cob. In addition, these stalks are spaced rather widely apart, so that a cob of Ramosa 2 from which the chaff has been removed resembles a similarly treated cob of Coroico maize (Cutler, 1946). SUMMARY Morphological studies have been made on tassels and ears of sixteen mutant forms of maize. Four of these forms are reported for the first time, and the others, for which descriptions were either inadequate or non-existent, are here [Vor. 42 210 ANNALS OF THE MISSOURI BOTANICAL GARDEN described. Each genetic form was introduced into a standard background derived from the inbreds CC5 and L317. The mutants studied included nine tassel-seeded forms: #51, fs), Тез, #54, 45,2, Ту», Тз, із; and ts,. Others studied were Male- sterile (ms,), Branched-silkless (bd), "Club (club), Vestigial glume (Vg), Tuni- cate (Ти), Ramosa-1 (ra,) and Ramosa-2 (ra,). LITERATURE CITED Alava, Reino O. (1952). Spikelet variation in Zea Mays L. Ann. Mo. Bot. бы 39:65—96. Апдегзоп, died (194423). Homologies of the ear а tassel in Zea Mays. . Мо. Bot. Gard. 31:325— S piel The sources of effective "re in hybrid maize. Ibid. 355—361. ‚ (1949). mom aa New у ми ) аге red plume. 24 pp. Publ. Puer Hi-Bred Co Des ————, L кока ( 1952а ). The history of the common maize variation me the United States corn belt. Agric. History 2 ———— 009 (195525). Origin die corn belt maize "n its genetic significance. In: Heterosis. 124—148. Edited by ues W. Gowen. State Colleg FEE E UG та; Є корі Ыы, ane К, Г. на (1949). Hereditary effects pro- duced in maize фі radiations s from the Bikini atomic bomb. I. Studies on seedlings and pollen of e ов generation. pru s 34:639—646. Collins, G. N. (1917a). Hy brids ds г? eo ramosa and Zea tunicata. Jour. Agric. Res. 9:383—395. —————, (1917b). Hybrids of Zea tunicata and Zea ramosa. National Pee Sci. Proc. 3:345—349. Sarien. H А part; хисар men and the preservation of а relict gene in maize. Jour. Hered. ————, 15 46). Races об maize in South America. Нагу Mus. ГеаЙ. 12:257-292. Emerson, R. А. (1920). Heritable Paises of maize. II. PD Re maize uas Jour. Hered. 11:65-76. , (1932). The present status of maize genetics. Sixth Internat. Cong. Genetics Proc. 1:141—152 ‚ С. W. Beadle, and A. C. org (1935). A summary of linkage studies in maize. Cor nell Univ. Agric. Exp. Sta. Mem. 180:1 я Gernert, У. В. (1912). А new subspecies of Zea Amer. Nat. 46:616—622. Hayes, H. K., and H. E. Uti тыз (1928). s characters in maize. ХХХШ. Sorghum tassel. Jour. Негев; 19:561—567. а n, LH. (1921): Inheritance of ramose inflorescence in maize. U. S. Dept. Agr. Bull. 971:1—20 ‚ (1925); паса зе aracters of maize. XIV. LIA ears. Jour. Hered. a 2” ------- (1934). ritable characters in maize. XLVII nched sillón. Ibid. —32. Lenz, L. Y. (1948). а arative ная of the female D eed of Zea Mays L yes Mo. Bo ard. 35:353-376. Mangelsdorf, Р.С Men The role of pod corn in м origin and evolution of maize. Ann. Мо. 35:377-406 СЕ . Smith, Б се New archaeological evidence on evolution in maize. Harv. Uni ONE ; Mus Leafl. 13:213-24 Nickerson, N. H. ee Wide i in cob morphology ер certain archaeological and ethno- logical ee of m Ann. Mo. Bot. Gard. 40:79-11 ------, 54). Morphólogical analysis of ak maize ear. Am. Jou Phipps, I. F зр Heritable characters іп m XXXI. Tas ES d ү Tout, Hered. 19: ice Singleton, W. R., and D. F. Jones (1930). Heritable characters in maize. XXXV. Male sterile. 268. ur. Sprague, G. F. (1939). Heritable characters in maize. L. Ves tigal glume. Jour. Hered. 30:143—145. Wether, Paul (1916). Morphology of the flower of Zea Mays. Bull. Torrey Bot. Club 43:127- 1955] NICKERSON & DALE—TASSEL MODIFICATIONS IN Zea Mays 211 rawings to same scale of spikelet pin found on маније primary tassel branches ie 4. h of certain mutants showing characteristic = #51, adaxial v of thick, ribbon-like branch showing cupules and a developed second flor t (silks removed). Тв, » рейесіме eee and sessile (below) spikelets with hyaline awnless sonde pe ис spikelet for ous naked small pistils. bd, both spikelet orm extra spik Tdo, an extra internode 2 м чак tween ead У Бо rimary axis and point of depa e left reni ts4, sessile обода left has one male and t (right) f numerous spikelets, with many of the thin glumes endin awn-like tips. Тз, и а and pedicellate male іп one solide pair; сд ахіѕ ж кор го vun of cupule for part of its length (silks removed). B E б 55. et > о an о Фо й о - p. ВЕ © et с < os PE E 23 2 e © B ot two оту es ÉL pum кати -lik PLATE 22 ANN. Mo. Bor. Garp., Vor. 42, 1955 SAVW VAZ NI SNOLLVOHIGON 1455У1--З1У А INV NOSWAMOIN (mp) озвогипі (со ‘ЗЫ ртеритзс ‘1 ‘Зы PLATE 23 ANN. Мо. Bor. Garp., Vor. 42, 1955 Fie t: Tassel-seed 1 (/s,) Fig. 2. Tassel-seed 2 (6,) NICKERSON лмо DALE—TASSEL MODIFICATIONS IN ZEA MAYS 4 > PLATE VoL. GARD., Mo. Bor. ANN. SAVW УЯ7 NI SNOLLVOLIIGOW м pees-[osse] ‘2 '814 IHSSVL—HIVQ( аму МО5ЧЧЯЯОІМ [ur рәәв-|әвее Т. SAVW VHZ NI SNOLLVOLIHIIGONW TASSV.L—aTVG INV NOSWIWXOIN "Su iL (5) с poos-[ssse]. "с ‘ЗЫ (954) 9 poos-[osse] a ANN. Mo. Bor. Garb., Vor. 42, 1955 PLATE 25 PLATE 26 ANN. Мо. Bor. Garp., Vor. 42, 1955 (557) 8 poos-[osse] 4 SAVW VIZ NI SNOLLVOISIGOW TASSVL—AIVG INV NOSYAMOIN ie Зы Си) 2 pees-pessep СІ '814 PLATE 27 ANN. Мо. Вот. Garb., Vor. 42, 1955 m Fig. Т; Male-sterile 1 (75,) ^ Fig. 2. Club (club) NICKERSON лмо DALE—TASSEL MODIFICATIONS IN ZEA MAYS SAVW V3Z М SNOLLVOIJIGOW 71455У1-ЯЛУАа INV NOSWIMOIN (pq) ssepqurs-poqouezq ‘с ‘ЗЫ (ЗА рел PLATE 28 ANN. Mo. Вот. Garp., Vor. 42, 1955 SAVW V3Z NI SNOLLVOIJIGOW T43SSV.L—31VG ANV МО5МАМОМ "St tp (ва) т еѕошеу с ‘Зы "(Єра) z vsowry ANN. Мо. Вот. Garb., Vor. 42, 1955 - PLATE 29 А NEW SPECIES OF DORYOPTERIS FROM SURINAM* KARL U. KRAMER лмо ROLLA М. TRYON, JR. In the course of examination of specimens in the fern collection of the Botanisch Museum en Herbarium, Utrecht, the senior author discovered a speci- men of Doryopteris that seemed to be different from any species treated in the junior author's revision of the genus!. be the case and it is here described as new. ПОКХОРТЕКЛ conformis, spec. nov. Rhizoma modice crassum breviter repens, squamis elongatis angustissimis dense vestitum, partium hyalinarum cellulis latitudine maxime partem quintum longi- tudinis aequantibus; stipes obscurus laevis vel leviter rugosiusculus, fasciculis vascularibus duobus; lamina fertilis sterili similis coriacea, venatio libera; sterilis suborbicularis—quinquangularis, profunde bipinnatifida; fertilis conformis, pro- funde bi- vel tripinnatifida; receptaculum intramarginale plus minusve continuum; sporangia breviter pedicellata. Typus: Gonggrijp 9 Stabel (В. W.) 5609; Surinam: Mt. Hendriktop, alt. 1080 m.; moist, sunny rocks; in Herb. Utrecht. Rhizome moderately stout, short-creeping; scales of the apex of the rhizome very long and narrow, the cells of the hyaline portions at least five times as long as broad; stipe dark purple to black, naked or slightly scaly at the base, glabrous, smooth or minutely and irregularly roughened, with two vascular bundles at the base, terete; fertile and sterile blades similar, without proliferous buds, coriaceous; sterile leaf about 25 cm. long; blade 7 cm. long, suborbicular-pentagonal, deeply bipinnatifid with about 12 oblong, broadly rounded, entire or partially crenulate ultimate lobes; margin with a pale brown cartilaginous border; venation free; hydathodes prominent on the upper surface; fertile leaf about 25 cm. long; blade about 8 cm. long, suborbicular-pentagonal, deeply bi- to tripinnatifid, with num- erous ultimate segments; primary segments broadly decurrent and surcurrent, the bases forming wings along the rachis with symmetrically concave sides; ultimate segments oblong-lanceolate, narrowly rounded, entire; soral lines continuous around the sinuses; sporangia short-stalked, і.е. the stalk somewhat shorter than capsule (which is 340—360 р long), borne оп a more or less continuous vascular commis- sure; spores subglobose, triplanate, slightly rugose, pale yellow-brown, about 55 p. Doryopteris conformis is most closely related to D. lomariacea from which it differs in the non-dimorphic fertile and sterile leaves, the segments of the fertile being much broader than in D. lomariacea, in the short-stalked sporangia and coriaceous texture of the blades; in D. lomariacea, the sporangia are long-stalked and the leaf-tissue is herbaceous. This species is notable in that it is the only local endemic in the genus outside of southeastern Brazil. However, the Guiana Highlands represent a weak sec- Further examination has proved this to 1Tryon, Rolla M. A revision of the genus Doryopteris. Contr. Gray Herb. 143:1-80. 1942. *Issued November 17, 1955. (21 [Vor. 42, 1955] 214 ANNALS OF THE MISSOURI BOTANICAL GARDEN ondary center for the genus in South America. This is the sixth species known from there, not including the widespread and somewhat doubtfully allied D. concolor; a concentration of species exceeded only in the southeastern Brazilian Highlands. It is also significant that all the other five species, D. lomariacea, sagittifolia, collina, varians, and pedata var. multipartita are variously disjunct tween southeastern Brazil and the Guiana Highlands. ether D. conformis represents a local offshoot of D. lomariacea or whether it was evolved, as all other Guiana species evidently were, in southeastern Brazil and migrated via the Andes to the Guianas, can only be determined if it is eventually discovered in Brazil. EXPLANATION OF PLATE PLATE 30 Fertile and sterile leaf of Doryopteris conformis, from type. X 24. ANN. Мо. Вот. Сакр., Vor. 42, 1955 PLATE 30 KRAMER AND TRYON—DORYOPTERIS CONFORMIS А REVISION OF THE GENUS CELASTRUS* DING HOU** INTRODUCTION There are about 500 species which have been referred in the past to the genus Celastrus; of these half are African plants. In 1942 Loesener!, in a review of the genus for Engler & Prantl’s ‘Pflanzenfamilien’, transferred some of the African species to the genus Gymnosporia and some to the allied genus Maytenus. In indicating the geographical distribution of Celastrus only in Asia, America, Australia, and Madagascar, he seems to have been aware of the fact that the so- called species of Celastrus of Africa proper did not truly belong to that genus. Loesener has expressed the opinion that the confusion regarding species limits in Celastrus and the synonymy indicate the need for a more precise study. I have attempted in this treatment to define the generic limits of Celastrus and its rela- tionship with other closely related genera, to review and to check all the published binomials of Celastrus, and to clarify the complicated synonymy. I have also made a study of the morphological characters and geographical distributions of the dif- ferent species. I have used the data to separate species or groups of species and to show their inter-relationships (fig. 1). All the “Celastrus” species from Africa proper have been excluded in this treatment, as they are referable either to Сутто- | sporia or Maytenus. Y have tried to delimit definitely the three genera, Celastrus, Gymnosporia, and Maytenus, which have been very much confused. It is my intention in the near future to make a detailed study of these last two genera. This treatment of Celastrus includes two subgenera with thirty-one species and five subspecies. The generic name comes originally from Theophrastos, who, however, desig- nated with the Greek word xeAaorpos (Kelastros) an evergreen tree (Phillyrea), that has nothing to do with our genus”. It is feminine in gender as used by Theo- phrastos, but Linnaeus in adopting it, made it masculine. It has already been pointed out by Airy ae that under the International Rules the masculine gender must be retained. Hence, in this treatment, Celastrus is treated as masculine and I have standardized all di epithets accordingly. lLoesener, Th., in Engler, A. & Prantl, К. Die Natürlichen Pflanzenfamilien 2 Aufl. 20b:131, 124. 1942 Loes. loc; cit. 2 Aufl. 20B:132. rem Shaw, H. K. in Curtis's hot. Ми 158:4. 0304. 1935. Botany of Washington peter of An investigation carried out in the Henry Shaw School of апі rang as a thesis in partial fulfillment for the requirements for the degree of Doc Philoso ** Arnold Arboretum of Harvard University, Cambridge, Mass. Issued November 17, 1955. (215) [Vor. 42 216 ANNALS OF THE MISSOURI BOTANICAL GARDEN НЕТОКУ The genus Celastrus was founded and described by Linnaeus* in his ‘Genera Plantarum’ (1737). Later, in his ‘Species Plantarum” (1753), he described three American and two South African species. Only one of these, Celastrus scandens, is now retained in the genus Celastrus. In 1824, Kunth® pointed out that some of the Celastrus species, for example, Celastrus buxifolia L., C. montana Roxb., C. trigynus Lam., etc., have peculiar characters and might well constitute a new genus. He gave a detailed description for that group of species, but provided no name for it. In 1834, Wight and Arnott’ studied the Indian species of Celastrus, and, chiefly on the basis of ovule or seed characters, divided the genus into two sections, EUCELAsTRUS and GYMNO- SPORIA. They gave each section a very concise and clear description. In the sec- tion EUCELASTRUS, the ovary is free from the disc, the ovules have a cup-shaped aril at their base, and the seeds are surrounded by an entire fleshy aril. In the section GYMNOSPORIA, which corresponds to Kunth's unnamed genus, the ovary is half- immersed in the disc, the ovules are naked at the base, and the seeds are apparently without an aril or with a very short, imperfect, and membranaceous one at the hilum. In 1862, Bentham and Hooker? elevated the section СУММОЗРОЕЈА to the rank of genus. At present the limits of the genus Celastrus are identical with those of section EUCELAsTRUS of Wight & Arnott. GENERIC RELATIONSHIPS Gymnosporia and Maytenus are closely related to one another and to Celastrus, They often have been confused with Celastrus both in the literature and in the herbaria. Exell? has pointed out that Loesener's!? separation of Gymnosporia from Maytenus, on the presence in the former of either thorns or inflorescences borne on short shoots, appears artificial Не says: "There seems little point in keeping the two genera distinct." After examining all available specimens of Celastrus, Gymnosporia, and May- tenus, Y have concluded that they are distinct though very closely related genera. The following table for distinguishing the three genera is based on a consideration of several characters taken together and not any one single character taken by itself (pl 31) енен С. Сепега Plantarum. ed. 1. "Wight, R. & Arnott, G. A. W. Prodr. F FI. I bt Ori rient. 152. 1834. ЗВепећат, С. & Hooker, J. D. Genera ать М 1:364. 1862—67. „o Exell, А. W., in Kew Bull. 1953:103. Loes. loc. cit. 2 Aufl. 20b:109. 1942. ANN. Mo. Вот. Garb., Vor. 42, 1955 PLaTE 31 CELASTRUS MAYTENUS GYMNOSPORIA HABITS TWINERS SHRUBS, SMALL TREES OR TREES SHRUBS OR TREES PHYLLOTAXY ALTERNATE ALTERNATE RARELY OPPOSITE ALTERNATE LENTICELS DISTINCT OBSCURE OBSCURE DISTINCT DICHASIA OR RARELY FASCICLED OR RARELY SOLITARY DISTINCT DICHASIA OR FASCICLED SOLITARY INFLORESCENCES ~ $r f. AXILLARY BUD AND BUD SCALES DISTINCT Ж OBSCURE 2 OBSCURE þe FLOWERS POLYGAMOUS RARELY BISEXUAL BISEXUAL BISEXUAL ARTICULATION ON THE PEDICEL DISTINCT а NONE y DISTINCT 2 OVARY AND DISC OVARY PES FROM THE DISC, 3- CELLED; DISC USUALLY CUP-SHAPED, RARELY MERELY CONCAVE, MEMBRAN - OUS, RARELY SUBFLESHY v 4 OVARY SEMI-IMMERSED IN THE "DISC, USUALLY 2-CELLED, RARELY 3-CELLED; DISC FLESHY FLAT, RARELY CONCAVE OVARY SEMI-IMMERSED IN THE DISC, USUALLY 3- CELLED, RARELY 2-CELLED; DISC FLESHY FLAT, RARELY CUP-SHAPED y Y USUALLY 2. (1 IN CENTRAL USUALLY 2, RARELY 1 ALWAYS 2 NO. OF OVULES AMERICAN SPECIES) is ® © © Ф. SUBGLOBOSE , e ELLIPSOID OR те ЗНАРЕ FURROWED OBOVOID , SMOOTH Q 9 Ф 3-VALVED FRUITS NO. OF VALVES NO. OF SEEDS USUALLY LI SEEDED, RARELY 1-SEED! USUALLY 2-VALVED, RARELY 3-VALVED USUALLY 1-SEEDED, RARELY 2-4-SEEDED. USUALLY 3-VALVED, RARELY VED e USUALLY 3-6-SEEDED USUALLY OVOID OR ELLIPSOID; ROUND IN CROSS-SECTION USUALLY OVATE; OVAT PLANO-CONVEX IN о SECTION USUALLY OVOID OR ELLIPSOID ROUND IN CROSS-SECTION SHAPE z c с> С) ш з COMPLETE COMPLETE OR INCOMPLETE AT THE BASE OR INCOMPLETE ARIL ^ T: CUP-SHAPED ARIL NONE OR SLIGHTLY AT THE ВАЗЕ NONE ARIL OF THE OVULE 0 DISTRIBUTIONS CHIEFLY IN ASIA, SOME IN AMERICA 8 MADAGASCAR c. W IN AUSTRALIA AND AFRICA CHIEFLY IN S. & AMERICA FEW CHIEFLY IN S. AFRICA AND ASIA FEW IN PACIFIC ILS. Celastrus Maytenus, and Gymnosporia contrasted. [ +. 42 218 ANNALS ОЕ THE MISSOURI BOTANICAL GARDEN MORPHOLOGY HABIT:—Plants of Celastrus are scandent, from 1 to 50 meters in height. The branches are hardy, rapid-growing and vigorous. There are very few available records of the diameter of older branches for all the species, but two records of Celastrus monospermus show this diameter to be about 10 cm. The branches and branchlets usually are terete, except in Celastrus angulatus, and are glabrous except on young branchlets, for example, of Celastrus hirsutus. Most older stems have orbicular to ovate lenticels. In Celastrus hindsii, lenticels | are usually absent on the current year’s growth. They are fine and dense in Celastrus angulatus and C. lenticellatus, but in C. hirsutus and C. caseariifolius they are large, elevated, ovate, or orbicular, and are found on the peduncles and pedicels. This feature may well serve as a supporting character to assist in identification of some species. VEGETATIVE BUDS:—Celastrus gemmatus bears axillary buds which are conical and characteristically large. All other species bear small, depressed and ovoid ones. The inflorescences of the species of series AXILLARES (except Celastrus monospermus, C. monospermoides, and C. bindsii) and one species of the subgenus КАСЕМО- CELASTRUS, C. panamensis, are associated with accompanying vegetative buds, while the inflorescences of other species are not associated with vegetative buds. The outermost bud scales are usually deciduous, occasionally persistent, falcate, and spiny. These persistent bud scales are called “stipules” by some authors. Usually in Celastrus aculeatus and in some specimens of other species, there аге four or more outermost bud scales as opposed to the usual two. These are nearly triangular, spiny, and semi-persistent or persistent. LEAVES:—The leaves are extremely variable, in the same species, and even on the same plant, the result of environmental conditions and the age of the plant. Previous authors mainly have used leaf characters for distinguishing species and varieties in the genus, but such characters are very unreliable. Most of the species are deciduous, but there are evergreen ones, for example, Celastrus monospermus, C. monos permoides, and C. bindsii, 'The leaf shapes vary from elliptic to oblong, or from broadly ovate to orbicular. The apex and the base vary from acute to obtuse or rotund. The margins usually are serrate, serrulate or subentire; in Celastrus flagellaris they appear finely ciliate. In size, the leaves range from 2.0 cm. long and 0.8 cm. wide in Celastrus punctatus to 16 cm. long and 16 cm. wide in C. angulatus. Texture usually varies from delicately to firmly membranous. In general, the texture of the leaves in flowering and fruiting stages of the plant is different: for example, in Celastrus angulatus the leaves of flowering specimens are delicately membranous, while those on the fruiting specimens are firmly membranous (so- called “coriaceous” of some authors). The leaves are glabrous or pubescent on the veins below. Sometimes they are pubescent in the juvenile stage and later become glabrous. In Celastrus birsutus there is a dense brownish pubescence on both surfaces. 1955] HOU—REVISION OF THE GENUS CELASTRUS 219 The leaves of all the species have pinnate and netted veins. The midrib usually is elevated below, and distinctly or slightly elevated above. The primary lateral veins are four to nine pairs on each side of the midrib, arcuate towards the apex. The veinlets though usually visible, are obscure on both surfaces in Celastrus aculeatus, and while usually loosely reticulated are densely reticulated in C. hindsii and C. gemmatus. The phyllotaxy always is alternate. The petioles are canaliculate, from 0.5 to 3.0 cm. long. In Celastrus flagellaris the petiole is half as long as the blade while in the other species it is less than half as lon FLORIFEROUS BRANCHES AND INFLORESCENCES:—Diagrams of the floriferous branches of the current season’s or year’s growth of most of the species of Celastrus have been drawn from herbarium specimens (pls. 32 and 33). These diagrams are arranged according to the sequence of the species in the key, and are not in- tended to show any evolutionary paths. They will make it convenient for a com- parison of the relationships among the species. They also reveal that the male and female plants of the same species have different inflorescence patterns. A “flower-cluster” has been thought of as а whole inflorescence or as a part of the inflorescence. Parkin’! has commented: “It seems futile to quibble over the question whether the inflorescence means the mode of floral branching or the flower-group itself. Custom has sanctioned the latter meaning.” I have adopted the customary approach of considering а "flower-group" as an inflorescence only as a matter of convenience in this treatment of the genus Celastrus. Dichasium and Aggregate Dichasium:—The fundamental type of inflorescence in the genus Celastrus is an axillary, solitary dichasium. A dichasium in its simplest form consists of three flowers—one terminal and two lateral, the terminal flower blooming first—and is organized as follows: First, there is the primary peduncle terminating in three secondary peduncles to each of which the pedicels of the flow- ers are articulated. What appears at first to be the entire flower stalk therefore is really a part of the peduncle and the whole pedicel. The dichasium usually is com- pound, and sometimes the branches are much multiplied to form a terminal panic- uliform aggregate dichasium. The aggregate dichasium differs from the thyrse in all axes being determinate and lacking any association with vegetative buds. In the literature the paniculiform floriferous branches are described usually as “panicles.” Each dichasium may or may not be subtended by a foliage leaf or bract, but in the series AXILLARES it is always associated with a dormant bud in the axil of the subtending appendage (pls. 31 and 32). АП the species of subgenus RACEMO- CELASTRUS have axillary dichasia with or without distinct primary peduncles. The Latin American species of RACEMOCELASTRUS, in addition, usually have dichasia arranged to form a racemiform floriferous branch on the current year's growth (pl 33). In this case the racemiform branch may be compounded and be located terminally on the branchlets to form a pseudopaniculiform aggregate dichasium 11Parkin, J. The evolution of the inflorescence. Jour. Linn. Soc. Bot. 42:512-563. 1914. ANN. Mo. Вот. GARD., Vor. 42, 1955 PLATE 32 EY P ‚Уу oT Hy, УР ке Ма У о лье m E * . maisade 79 if) Көзі 7 (Fr.) Clemens 5152 (Young Fr.) "ыа ы chow зәт (Pre) С. PANICULATUS C. NOVOGUINEENSIS C. ANGULATUS / M^ 7 Й а | ~ Жон 4 a A 3 P NZ ке => 4 2 МИ 4 aei ~ 5 Ske A дз J bot аф TEN 4 am. 2 а га - > 2 те зі - ~ 2 a P = Я ad Lo > а [m P ~ m | i Pia lat мо E Caley (ПВ 1276013) (тт.) Dag b Ред. зе че E ша williams /fr.) "aang 80000 (27) Levime 401 12) Wilaec ( Pr.) C. AUSTRALIS C. RICHII C. SCANDENS C. HINDSII 7 zA E —- ә. Y ге | 5 xm iae зет, те ЦУ Forrest 26407 И) 8.р.ке 50008 (ут,) & › 6. SPERMUS с. GLAUCOPRYLLUS f от C. HYPOLEUCUS ANN. Mo. Вот. Сакр., Vor. 42, 1955 PLATE 33 ТЕР; Zn Cult. fn X. лот (Ф) Wilson 7412 tfr.) Rosthorn 197; wilson 1173 (Fr.) Е рейін с. зані CULATUS c. . "ROSTHORNI ANUS Fr.) Огавеїла 1414) Hu Pételot өз 0) > Henry 5569 (0) 14 1181 (Rm) Б ES о US C. KUSANOI C. HIRSUTUS C.STYLOSUS SSP.GLABER SUBGENUS RACEMOCELASTRUS bor, Sad 1 AD q Tsang 20092 12) Tamm 724 (Fre) ‘Wilson 8450 ($) Tee 21823 (Fr) penne z [^ ACULEATUS С. FLAGELLARIS Viene C.RACE- C.VULCANI- — C.MERI- - axillary bud - floral bud = blooming flower - fruit - bract - leaf ~ -0 а о Plates 32 and 33. Diagrams of the floriferous branches of current season's or year's growth of Celastrus species. [Vor. 42 220 ANNALS OF THE MISSOURI BOTANICAL GARDEN (pl. 33). This type of dichasium differs from the paniculiform aggregate dichasium in series PANICULATI in having a vegetative bud in the axil of each dichasium. FLOWERS:—TIhe flowers of this genus are small and inconspicuous, usually about 3—5 mm. long. According to the available records the flowers are yellowish-green or white. They are generally dioecious in the subgenus CELAsTRUs, and usually bisexual in the subgenus RACEMOCELASTRUS. Calyx:—The calyx is campanulate and persistent, and sometimes accrescent as in Celastrus aculeatus. It always has five short and equal, deltoid, ovate, or oblong lobes which are usually imbricate or in some species rarely open in aestivation. The lobes have simple venation and are glandular-ciliate, slightly erose, or entire; they are glabrous or slightly pubescent on the outer surface, and reflexed or spreading in the fruit. Petals;— The flowers have five petals which usually are oblong or elliptic, glandular-ciliate to slightly erose, or entire. In Celastrus paniculatus the petals have parallel veins while in other species each of the petals has а “costa” and some lateral veins. They are generally glabrous, but in Celastrus kusanoi are slightly puberulent on the inner surface. The petals usually are reflexed. Floral Tube and Disc:—In the flower, there is a cup-shaped or plate-like struc- ture which is usually referred to as the disc. Berkeley!?, who made a morphological. and anatomical study of the flower of two species of Celastrus, C. scandens and C. orbiculatus, on the basis of the behavior of the vascular supply to the flower, interpreted the cup-shaped structure as the floral tube consisting of fused stamen, petal, and sepal bases. Не considered the inner layer of the floral tube to be com- posed of fused basal portions of existing stamens and of stamens no longer present; between the existing байок there are small projections of tissue of reduced stamens constituting the "disc" in his sense. The cup-shaped and реална "floral tube" is common in most of the species in this genus. It surrounds the pistil but is free from it; on the rim of the tube are attached the free floral parts. А plate-shaped, flat and fleshy "floral tube" is usually present, chiefly in all the species of subgenus RACEMOCELASTRUS an several other species of the subgenus cELAsTRUs. The ovary is situated at the cen- ter of this type of "tube" and free from it, while all the floral parts are attached on the outer edge of the tube. The two species that Berkeley had studied lacked the flat floral tube, and it is difficult to interpret its origin without morphological and anatomical evidence. The term disc connotes the “Нога! tube" of кар and disc-lobe is equivalent to the "disc" in Berkeley’s sense in the present treatment of Celastrus; this is just өлі convenience. The disc-lobe may be distinct or obscure, deltoid, oblong ог iform. Sometimes these are used as supporting characters to identify species. Berkeley, E. Morphological studies in the Celastraceae. Jour. Elisha Mitchell Scientif. Soc. 70:185-206. 1954. 1955] HOU—REVISION OF THE GENUS CELASTRUS 221 Fertile and Sterile Stamens:—There are five stamens in the staminate flowers. The stamens depart from the rim of the cup-shaped disc, or are attached immedi- ately beneath the edge of the flat disc. The filaments generally are filiform, fleshy or thin. They are glabrous in all of the species except Celastrus stylosus ssp. stylosus, C. aculeatus, C. hirsutus, and С. kusanoi, where they аге papillose-tuberculate. The filaments usually are longer than the anther, but in Celastrus novoguineensis they are shorter. The anthers in the staminate flowers are generally oblong-ovoid, obtuse or cordate, but in some species they are distinctly apiculate, i.e., Celastrus pringlei. The sterile stamens in the pistillate flowers are small, short, and about 1 mm. long; the anthers of these are triangular and cordate, and sometimes they may contain sterile pollen grains. Fertile and Sterile Pistils:—The fertile pistil usually is superior and free from the disc. The ovary usually is subglobose or ovoid, and narrowed into a columnar style. The ovary has three cells. Each cell has two ovules in the species of sub- genus CELASTRUS while there is only one ovule per cell in the species of subgenus RACEMOCELASTRUS. The stigma is trilobed or the lobes are obscure. In some species the lobes are further divided in half, for example, Celastrus orbiculatus. The lobes usually are filiform and reflexed, but they are tangentially flattened in Celastrus angulatus. The sterile pistil in the staminate flowers usually is small and columnar. Ovules:—The ovules are anatropous and attached at the base of the ovary by a distinct funiculus. The raphe is clearly visible on the seed. Ina free-hand sec- tion of the ovule the micropyle is sometimes visible towards the base. The ovule has the characteristic cup-shaped aril towards its base, which is a definitive char- acter of this genus. The embryo is erect, thin-foliate and broadly spatulate!®, The radicle is cylindric and stalk-like. | FRUITS:—Fruits аге subglobose, ovoid, obovoid or cylindric, and are smooth, or rarely angular as in Celastrus monospermoides. The bases of the fruits are obtuse trus monospermus, where the base is contracted into a The fruits generally have three cells, but two- or four- The fruits are three- to or rotund, except in Celas column and is stalk-like. celled fruits are occasionally found on some plants. six-seeded in most of the species. Celastrus monospermus, C. hindsii, and C. monospermoides have only one developed seed with five undeveloped ovules, while all the species of the subgenus RACEMOCELASTRUS have one developed seed with two undeveloped ovules. The fruit is three-valved, dehiscence being loculicidal. The valves are bright yellow or dark brown, slightly transversely wrinkled on the out- side, and spreading or reflexed in dehiscence. 13 Martin, А. C. The comparative internal morphology of seeds. Amer. Midl. Nat. 36:520, fig. 2, Ы. 55. 1946. [Vor. 42 222 ANNALS OF THE MISSOURI BOTANICAL GARDEN HYPOLEUCUS ORBICULATUS MEE GEMMATUS GLAUCOPHYLLUS PUNCTATUS SUBGENUS ROSTHORN | ANUS CEL ASTRUS HIRSUTU KUSANOI ACULEATUS TYLOSUS HINDSII MONOSPERMUS MONOSPERMOIDES “SERIES KXFLTARES ANGULATUS PANICULATUS RICHI I NOYOGU INEENS 15 SUBSP ICATUS MADAGASCARIENSIS 1 OVULE ІМ EACH CELL OF THE OVARY 2 OVULES IN EACH CELL OF THE OVARY | че ee | wi узу КЛ № ТЕ ШЕ аы] Fig. 1. Chart showing interrelationship of species of Celastrus and their geographical distribution. 1955] HOU—REVISION OF THE GENUS CELASTRUS 223 Harris! studied the relationship between the number of flowers formed and the number of fruits maturing per inflorescence in Celastrus scandens. Не also studied the size of an inflorescence as measured by the number of flowers it produces and the number of seeds per fruit. His conclusion is that there is no correlation between these pairs of characters. GEOGRAPHICAL DisTRIBUTION Celastrus is widely distributed in eastern Asia, Oceania, both Americas and Madagascar, between the latitudes of about 40° S. and 47° N. Most of the species are found in the subtropical and tropical zones. From the distribution patterns of the species, it appears that there are two centers of dispersal: Yunnan-Burma- eastern India, and Central America (fig. 2). In eastern Asia, the species radiate from the Yunnan-Burma-eastern India region. Celastrus paniculatus, one of the two most widely distributed species, radiates from that region eastward through Kwangsi, Kwangtung, and Hainan, eastward to Formosa; southward through the Malay Peninsula to Indonesia and then northward to the Philippines, and westward to India and Ceylon. Celastrus hindsii, the other widespread species, extends from the center westward to Sikkim; northeastward through Yunnan and Kweichow to Szechuan and Hupeh; eastward to Kwangtung, Fukien, Formosa and Bonin Islands, Japan; and southward through the Malay peninsula to Indonesia. The other species of that area have much more restricted distributions. From the Asiatic center of dispersal, the number of species gradually decreases both northward and southward, as is shown on the distribution map (fig. 2). In North America there is only one species, Celastrus scandens. It is widely distributed forming an ellipse from Quebec southward to Virginia, then westward to Texas and eastern Wyoming. It is closely related to the Asiatic species of the series PANICULATI. The disjunct distribution and floristic relationship of plants of eastern Asia and eastern North America recently have been interpreted by 1115, Не says that this familiar disjunct distribution appears to be the remnants of great mesophytic forests that extended over all the northern hemisphere and reached the arctic regions in the Tertiary period. Geological changes have destroyed and changed the floras of many lands so that this mesophytic forest of the Tertiary times in the northern hemisphere survives mainly in eastern Asia and eastern North America. Berry!9 states: "The genus Celastrus Linn. is the largest fossil genus of the family. Though its present center of distribution lies in the upper lands of south- 14 Harris, J. A. Correlation in the inflorescence of Celastrus scandens. Mo. Bot. Gard. Ann. Rept. 20:116-122. 1909. ; Ж 151; Hui-lin. Floristic relationships between eastern Asia and eastern North America. Trans. mor. Philos. бос. 42:371-429. 1952. ; MEA sect RS рату, E. W. The Lower Eocene flora сі southeastern Norta America. U. 2. бе», 5urv. rre. Paper 91:105. 1919. [Vor. 42 224 ANNALS OF THE MISSOURI BOTANICAL GARDEN eastern Asia and East Indies, its history shows that the ancestral stock was cosmo- politan and very abundant in the Tertiary of America and Europe. It is highly probable that it originated in America at the dawn of the Upper Cretaceous or somewhat earlier.” Croizat!" says that this view cannot be accepted as it is in conflict with the distribution data, but he cites no alternative center. АП that can be said from present distribution patterns is that Celas£rus scandens is related to the eastern Asiatic center of dispersion rather than to the one in Central America. It is rather difficult to interpret the relationships of the Central and South American species. Lcesener!? has pointed out that the Latin American species show relationship with tropical Asiatic species. "They closely resemble the Asiatic species Celastrus monospermus, C. monospermoides, and C. hindsii, especially in their in- florescence patterns, discs, and one-seeded fruits; but differ in the number of ovules i : fe j | | f, а је E LOT ! i г E EN uet NI \ 11 o Пи і 55 реБ Е - N aL. E I ae ~ Қз E E 1 Бы УЛЫ % анна BN PUR ХИ У ГЕТ БҮ SE M a Ly pope = 2 ue rancho la d. | јој. Пасау Fig. 2. Distribution patterns of Celastrus All the maps used in this paper are from Goode’s Series о: эк ане: of Chicago Press, except qp map of rigs United Sat. which was kindly с бекет c Dr. Woodson in each cell of their ovaries and in their sexuality. In the Central and South American species there is only one ovule in each ovary cell, while in all the other species the ovary cells have two ovules; the Asiatic and North American species are dioecious while the Latin American species are bisexual. Croizat!?, in his studie of the intercontinental distribution of plants, mentions that some Asiatic genera “run a transpacific channel [sic!]” from the Far East to 1 Croizat, Leon. Manual of ve cian de р. 302. 1952. 18T oes, loc. cit. 20b:103. 19 19 Croizat, Leon. loc. cit. pp. 54, 305, and fig. 16. 1952. 1955] HOU—REVISION OF THE GENUS CELASTRUS 225 the New World. Не holds the celastraceous genus Perrottetia as a perfect example. There are about fifteen species of this genus distributed in Queensland, New Guinea, Moluccas, Celebes, Philippines, northern Borneo, Formosa, central China, Hawaii, Mexico, Colombia, and Venezuela, and every essential station in the “traject [sic!]” is exemplified by one or more species. An alternative hypothesis is suggested by the fact that the Latin American species of Celastrus constitute a distinct subgenus without representatives in the Eastern Hemisphere. These species appear to be closely related to the neighboring genus Maytenus, which includes species of both Latin America and Africa, a dis- tributional phenomenon which might be explained upon the hypothesis of trans- Atlantic continental drift. From a rather dogmatic view of phylogeny, one might assume the hermaphroditic flowers of Maytenus (as well as of Gymnosporia) to be "primitive" from which the bisexual condition general for the Latin American Celastrus might be derived. The North American Celastrus scandens is closely related to the species of the Eastern Hemisphere rather than to those of the Western Hemisphere, thus suggesting that the North American and Latin American Celastrus might have different origins. The two subgenera perhaps therefore can be said to have independent origins. HYBRIDIZATION White and Bowden?? have reported identical chromosome numbers (n — 23) for the American Celastrus scandens L. and the Asiatic C. orbiculatus Thunb. They were able to hybridize these species with interesting results. The hybrid is less vigorous than the parents and sparingly fertile; the fruits are smaller than those of either parent, and the seeds are not uniform in size. When they back-crossed the hybrid to both parents both the Fs's had leaves closely resembling those of Celas- trus scandens. The plants resulting from the back-cross to Celastrus orbiculatus yielded one-seeded fruits, with only two exceptions among the 475 fruits pro- duced, while the fruits of both parent species contained three to six seeds. Of the plants resulting from those back-crosses, there is no further record available regard- ing the inflorescences, the sexuality of the flowers, the ovule number in each cell of the ovary, and the sizes and shapes of the seeds. All the Latin American Celastrus species have one-seeded fruits. It would be interesting to learn the chromosome number and morphology in these species. Since the Latin American species are more or less an uniform and isolated group, cyto- logical data might help in relating them to other groups of the genus Celastrus. In Celastrus, there appears to be a strong possibility of hybridization taking place in nature in certain localities. For example, Celastrus vanioti looks like a hybrid between Celastrus angulatus and С. bypoleucus. A specimen collected by Wang 76880 from Yunnan, China, seems a hybrid between Celastrus hindsii and C. monospermus. - 39 White, О. E., and Bowden, W. М. Oriental and American bittersweet hybrids. Jour. Hered. 38:125-127. 1947. ГУ ог. 42 226 ANNALS OF THE MISSOURI BOTANICAL GARDEN Economic Џ5Е5 There are very few available records of economic uses of Celastrus. Two species, Celastrus scandens and C. orbiculatus, are widely cultivated commercially for their beautiful yellowish fruits which open at maturity and disclose the crimson arillate seeds. Balfour?! states that the bark of Celastrus scandens has emetic and purgative properties. The seeds of this species are said?? to be useful pharmaceutically. An oil is extracted from the seeds of Celastrus paniculatus, which is described as having a very hot and bitter taste. It (oleum migrum) is useful for burning?? and has been employed in India successfully in treating beriberi disease?*, Burkill? reports that in Malaya the seeds may be used, externally, in poultices, or in the Philippines, internally, as an anti-rheumatic and as a treatment of paralysis. In Java, the leaves are used in dysentery. In the Philippines, the sap is given as an antidote in opium- poisoning. | Нетпзјеу2в, quoting from A. Henry's record, states that the leaves and roots of Celastrus latifolius Hemsl. (— C. angulatus Maxim.), when dried in the sun and pounded to a powder, are efficient in killing insects infesting turnips and other vegetables in China. ACKNOWLEDGMENTS This work was completed under the direction and supervision of Dr. Robert E. Woodson, Jr. I should like to express my deep gratitude for his kind interest, and more particularly for his extraordinary patience and understanding in helping me with the problem of my languages. Also, I cannot be sufficiently grateful to him for the financial assistance he has obtained for me during the course of my study. My thanks are due to Dr. Edgar Anderson, Director of the Missouri Botanical Garden, for facilities extended to me and for useful suggestions. I also acknowledge my gratitude to the Committee on Awards of Fellowships, China Institute in America, for the award of the Frank M. Shu Scientific Fellowship for the years 1952-53-54, enabling me to undertake this study. My sincere appreciation is due to Dr. Rolla M. Tryon, Jr., Dr. Frederick С. Meyer, and Dr. Hui-lin Li for suggestions and criticisms during the course of the investigation. I am also in- debted to Miss Nell C. Horner for much constructive criticism. I am happy to acknowledge my indebtedness to the many herbaria and their curators for the loans and facilities extended. Finally, I should like to express my thanks to many other friends who have directly or indirectly aided and encouraged me in the course of this study. ~ 21 Balfou 22 Gray, Asa. The Genera of the Plants of the United States. 2:185. 1849. awson, G. in Trans. Bot. Soc. Edinburgh 6:364. 1860. igt, J. O. Hortus suburbanus Calcuttensis. A Catalogue of the Plants which have been E: ec by the Hon. East ndia Company's botanical garden, Calcutta, etc. p. 5 . 166. 5 ll, 1. Н. A Dictionary of the Economic Products of the Malay Peninsula. 1:505. 1935. 26 Hemsley, W. B. in Jour. Linn. Soc. Bot. 23:124. 1886—88. 1955] HOU—REVISION OF THE GENUS CELASTRUS TE. ABBREVIATIONS Herbarium specimens used in this study include those of the following herbaria or institutions. The abbreviations for them are taken from Lanjouw & Stafleu's "Index Herbariorum', part I (Regnum Vegetabile, vol. 2, 2nd. ed., 1954). nold Arboretum, Harvard University, puit тен BLAT вача Herbarium, St. Xavier's College, Bombay, In —Herbarium of the Broo x Botanic Garden, Heck. New York. CAL Indian Botanic Garden, Howrah, Calcutta, India. —Chicago Natural History Muse eum, Chicago, Illinois. ime Gray Herbarium of Harvard University, Cambridge, Massachusetts. HK—Herbarium, Gardens Pee Hongkong, China K—The Herbarium, Royal Botanic era ка England. L—Ri iks he barum нса Netherla MICH.—Universi ity Herbarium, ро lee ман. Ann Arbor, Michigan. MO—Missouri Botanical Gate: St. Louis, Missour МУ— Мех York Botanical Garden, New York, М. Y P—Museum National d'Histoire Naturelle, Тае де Phanérogamie, Paris, Fra PENN-— Herbarium of the University of Pennsylvania, Philadelphia, Pennsylvania. РН—Асадету of Natural Sciences, Philadelphia, Pennsylvania. S—Naturhistoriska Riksmuseum, Botani cal Department, Stockholm, Sweden. ; Та ; U—Botanical Museum and Herbarium, Utrecht, Netherlands. UC—Herbarium of the University of California, Berkeley, California. UPS—Institute of Systematic Botany, pem Garden and Botanical Мок of the University of Uppsala, Uppsala, en. US—National Museum, Smithsonian аны Washington, D. C. TAXONOMY CELASTRUS L. Gen. Pl. ed. 1. 59. 1737; Sp. Pl. ed. 1. 196. 1753; Gray, Gen. РІ. U. S. 2:185, pl. 170. 1849; Benth. & Hook. Gen. Pl. 1:364. 1862; Loes. in Engl. & Prantl, Nat. Pflanzenfam. 2 Aufl. 20b:131. 1942. (Type species: C. scandens L.). Euonymoides Isnard, ex Medicus, Philos. Bot. 1:173. 1789 (pro к, Celastrus $ Eucelastrus Wight & Arn. Prodr. Fl. Ind. 2 158. 18 Semarilla Raf. Sylva Tellur. 146. 1838. (T.: S. bicolor Raf.). Scandent shrubs, up to 50 m. tall; branches usually terete, rarely the branchlets angular, usually glabrous, rarely pubescent, scarcely to densely lenticellate. Leaves extremely variable, elliptic to orbicular, serrate or subentire, rarely finely ciliate- serrulate, deciduous, rarely evergreen, alternate, petiolate; stipules small, usually laciniate and deciduous. Inflorescences dichasia, paniculiform to racemiform, soli- tary, sometimes branched, axillary or terminal, pedunculate or sessile, few- to many-flowered. Flowers small, usually unisexual and dioecious, rarely bisexual, pedicellate and articulate. Calyx campanulate, persistent on the fruits, 5-lobed; petals oblong, obovate-oblong or ovate, glandular-ciliate to erose, or entire, inserted under the disc, alternate with the calyx lobes; disc usually membranous and cup- [Vor. 42 228 ANNALS OF THE MISSOURI BOTANICAL GARDEN shaped, or fleshy and flat, entire or 5-lobed, the lobes alternate with the stamens; stamens 5, the filaments glabrous or papillose-tuberculate, arising from the margin of the disc proper or attached immediately beneath it, the anthers ovoid or oblong- ellipsoid, obtuse or apiculate, dehiscing laterally or extrorsely, dorsifixed, versatile, the thecae separated at the lower half or third; ovary superior, subglobose or ovoid, free from the disc or rarely slightly confluent with it, 3-celled, each 2-ovuled (sub- gen. CELASTRUS) or l-ovuled (subgen. RACEMOCELASTRUS), the ovules arising from the base of the ovary, axile, anatropous, with a cup-shaped aril towards its base, sessile or with a short funicle, the style usually columnar, the stigma usually 3- lobed or the lobes obscure, rarely each bifid. Fruit a capsule, usually subglobose, rarely cylindric, tipped by the persistent style, 3-celled, dehiscence loculicidal, 3- valved, the valves terminating in part of the style, 1- to 6-seeded; seeds enclosed in a fleshy crimson aril, the areolae distinct or obscure, the albumen copious, the embryo erect, thin-foliate and broadly spatulate. KEY TO THE SUBGENERA AND SERIES A. duh apis —— and dioecious; ovary 3-celled, each cell 2-ovuled; fruits 3- to 6- seeded, or in some — ies 1-seeded with 5 ченні» undeveloped ovules. Айы Осе Месне and North America gen. I. Салатын В. Inflorescences ната M ке ог, if in the axils of the uppermost н as ut accompanying vegetative buds er. 1. PANICULATI BB. Infotéscences both terminal pas axillary, or axillary only, but the axillary alw with acco mpanying vegetative buds (except in C. monospermus, C. mono. ospermoides 5; 2. а c . bindsii AXILLARES AA. Flowers bisexual; ovary 3-celled, each cell 1-ovuled; fruits pe with 2 p ctly undeveloped ovules. Central and South America en. П. RACEMOCELASTRUS Subgenus I. CELAsTRUS Series 1. PANICULATI Rehd. & Wils. in Sarg. Pl. Wils. 2:354. 1915. KEY TO THE SPECIES A. Calyx lobes imbricate, broader than long. B. eee usually at least thrice compound; petioles usually more than 10 mm. С дн of the male flowers truncate at the € = distinct filaments about 1 . long; pistil of the female flower about 2.5 long, the style half as long as a a ovary. India, Burma, Siam, foie CR Chins, Malaya, Indonesia, and the Philippines 1. C. paniculatus CC. Anthers of the male flowers apiculate, sessile or subsessile; pistil of the female flower about 1.5 mm. long, the style about one-fourth as long as me vary. New C. novoguineensis BB. €— once or twice compound; petiole less than 10 mm. long. . Anthers of the male flowers not apicu ulate or obscurely so; stigmata distinctly 4. ed; disc-lobes truncate; orescences usually 3-9 cm. long. E. Fruiting pedicels 1-3 mm ма leaves firmly membranous. Eastern Australia, Papua, and New Caledonia 3. C. subspicatus EE. Fruiting pedicels 4— m. long; leaves thinly membranous. Еїўї.............. Е С DD. Anthers of the male ‘oars, apiculate: stigmata discoid c or —, 5-lobed; disc- cuspidate; орн 1-3 cm. long. Майар; „нб, С wade gascariensis AA. Calyx m open, longer than broa . Young br: ж асаку leaves orbicular or broadly iig w petioles of nepr} ‚Зв usually more than 2 cm. long; елеп more than 10 . long, dense, кушы ді disc fleshy, flat; т subsessile. 6. C. angulatus 1955] HOU—REVISION OF THE GENUS CELASTRUS 229 FF. Young branchlets bw звање etie or broadly elliptic, ar petioles of upper leaves usually 1.6 с ong; initam Ж usually less pan 2, lax, s pod пее ys mem veux "о ed fruits и њу he pedicels 2-5 mm. long. North Атегїса...................... . C. scandens 1. CELASTRUS PANICULATUS Willd. Sp. Pl. 1:1125. 1797 (as dup weh. Lawson in Hook. Fl. Brit. Ind. 1:617. 1875, ex char Scandent shrubs up to 10 m. tall; branches terete, glabrous, brown, the young branchlets terete, usually pubescent, densely lenticellate, the lenticels elliptic, some- times elevated; axillary buds deltoid to orbicular, about 1.5 mm. long. Leaves elliptic, obovate, suborbicular, broadly ovate, ovate-oblong to oblong, the apex abruptly acute, obtuse or rarely emarginate, the base cuneate, broadly acute to obtuse, the margins serrate or remotely crenate, 5—15 cm. long, 2.5—6.0 cm. wide, firmly membranous, glabrous, sometimes pubescent on the veins below, the primary lateral veins usually 5—7 pairs, elevated on both surfaces, curved toward the apex, the veinlets distinct to slightly elevated below, immersed but visible above; stipules laciniate, about 1 mm. long; petioles 0.5—1.5 cm. long. Inflorescences terminal, paniculiform, thrice- to multi-compound, spreading, usually 5—10 cm. long, rarely up to 20 cm. long, pedunculate, the peduncles glabrous or puberulous, the primary peduncles usually 6—10 mm. long; flowers dioecious, pale or yellowish green, the pedicels about 4—6 mm. long, the articulation at the basal part of the stalk. Male flowers: calyx lobes semi-orbicular, imbricate, ciliate, about 0.7 mm. long and 1.5 mm. wide; petals oblong or obovate-oblong, obtuse, entire, about 3 mm. long and 1.5 mm. wide, with parallel veins; disc cup-shaped, the lobes inconspicuous, slightly triangular; stamens arising from the margin of the disc proper, 3 mm. long, the filaments subulate, flat, glabrous, the anthers ovoid, obtuse, cordate; sterile pistil columnar, about 1.3 mm. long. Female flowers: calyx lobes, petals, and disc as in the male; sterile stamens 1.3 mm. long; pistil 2.5 mm. long, the ovary globose, the style columnar, the stigmata 3-lobed, each deeply bifid, slender. Fruits subglobose, the valves broadly elliptic, about 6–9 mm. long and 5-8 mm. wide, 3- to 6-seeded; seeds ellipsoid, 3.5—5.0 mm. long and 2—5 mm. wide, yellowish-brown, smooth. Chiefly in thickets, at altitudes from 200 to 1,800 m.; widely distributed in India, Burma, Siam, Indo-China, China, Malaya, Indonesia, and the Philippines; flowering from April to June. KEY TO THE SUBSPECIES A. Leaves orbicular to broadly o B. Leaves orbicular, sd sic ай or obovate. India, Ceylon, Burma, oan taped and southwestern Chin ssp. paniculatus a ssp. serratus 1 BB. Leaves broadly ovate. Philippines and Indonesia AA. Leaves elliptic to ине ороо. India, Burma, Siam, Malay, Indonesia, Palawan 141. (Philippines), and С 1c. ssp. multiflorus la. CELASTRUS PANICULATUS ssp. PANICULATUS. Celastrus mutans Roxb. Hort. Beng. 18. 1814, nom. nud.; Fl Ind., ed. Carey & Wall. 2:390. 1824, ex Celastrus Ren Roem. & Schult. Syst, 5:423. 1819; DC. Prod. 2:8. 1825, ex char. (T.: Heyne s. n.). [Vor. 42 230 ANNALS OF THE MISSOURI BOTANICAL GARDEN Ceanotbus paniculatus Heyne ex Roth, Nov. Pl. Sp. 154. Toni ex char. Celastrus metzianus Turcz. in Bull. Soc. Nat. Moit. 31 :448. 1858. (T.: Metz 1549, GH!). Celastrus pesce var. pubescens Kurz ex Prain, in Jour. Asiat. Soc. Bengal 73:196. 1904, nom Celastrus pid enit var. andamanica Kurz ex Prain, loc. cit. 73:196. 1904, nom nud. ANUS. = Fig. 3. Celastrus paniculatus ssp. paniculatus BurMa: Fort Stedman, Abdul Khalil s. и: 1894 (CAL); Dr. King's collector 454 (CAL), 479 (L); Cokteih Shaub, Meebold 8055 (S); Maymyo Hill, Badal Kban 36, 259 CAL); Maymyo Plateau, Lace 3130, 3280 (e AL); P egu, Eyre s. n., 1887 (CAL), Kurz 1926 (CAL); sHAN HILLs: Rh «СН ird SABE Pwehla, Collett 672 (CAL); Shwebo, Abdul Huk 62 (CAL); souTHER E: Laikow, Abdul Khalil s.n., 1893 (CAL), s.n., 1894 (CAL), Ташу, Abdul Khalil 5. n., 1893 (CAL), s. n., 1894 Г, Г, UPS); Yeu, Cole s. п., 1890 is gio Rewamb, Watson I 76 САТУ, ; without precise locality, Thwaites 1232 CHINA: YUNNAN: Che-li, Wang 75551, 76335 (А); Fo-hai Hsien, Wang 74570, 74650 (A); Nan-chiao, Wang 75024 (А); Shung-kiang Hsien, Wang 72006 (A); Lang-tsang P »- 76644 (A); Szemeo, Henry 11 em US). DIA: ASSAM: Khasia Hills, 1,000—3,000 p Hooker. t & Thompson s.n. (CAL, L, 5, U), Duphla Hills, Radhu, Pokri, Lister s. п. (CAL); oo s, Dr. King’s collector s.n., Aug. 2, 1891 (CAL), Dr. Prain’s collector 70, 78 (СА M mon Lower Bengal, Kurz s.n. (CAL); Darjelling, Reyang, Osmaston s. n., 1903 ((CAL), Manbhum, Campbell 5. п. Е Maymbhing, Жармы, Hooper 388o (CAL), Orissa, Ball s.n., ALi: STATE: 8780 (CAL); а Hazaribagh, Chota Nag ain s.n., Nov. 29, 1891 (CAL); Parasnath, Kurz s.s. (CAL); Кайдай! | Hill а pant peines Kurz 5.п., May 1867 (CAL); терені. Көзі 65 (CAL). MBAY STATE: vic. Concan , Stocks. з. я; (CAL, І, S. U, UC, UPS); Khandala, Элка n J. 12707 (MO); Poona, Deccan, Woodrow 1955] HOU—REVISION OF THE GENUS CELASTRUS 231 5. п., 1882 (CAL); Dharasu, Dudgeon & Kenoyer 237 (MO); Bhawani, in montibus Nilagiricis, one 1549 (type of Celastrus metzianus Turcz., UPS). коді зго Кит » Str 273 (CAL); нь ed (CA но MADRAS STATE: и E Chilka Lake, Hooper 30537 (CAL); "Godala. Chodavarum, Ramaswami 1490 (CAL); Naikeneri, onn ner-chittur, Fischer 4730 (CAL); рія, Narayanaswami 105 (CAL), ADHYA PRADESH: Gwalior, Maries 22 (CAL); Khandwa, Duthie 8211 (CAL). MYSORE: Somanathapur, Barber 6846 (CAL) ; Shimoga, Barber 7018 (CAL) ; Bettiah, near the Nepal frontier, ан onymus 443 (CAL). HYDERABAD: Jalna, Peninsular India, Rodriguez 2025 CAL). PUNJAB: Bhadwar, Kangra, Koelz 4266 (MICH, US), 4377 (MICH, UC, US); Gurdaspur pu. Bis Ram 300 (S); Karnal Distr., Drummond 21651, 21691 (UC); Aryan Karn, Calder 9 Ramaswami 654 (CAL); Nokara, Rama Rao 1588 (CAL). яккм: Garubathen, Dr. Prain’s collector s. n., 1,500 ft. alt., 1900 (CAL); Rungit, Clarke 26330 (САТ); M gru precise locality, alt. 2,000—4,000 ft., Hooker s.n. (CAL, L). RAJASTHAN Abu, King s. ». (CAL), Stocks 231 (CAL). UTTAR PRADESH: Dehra Dun, Pradhan 34 (S); Poir. & а Dun, Anrudh Singh 25 (S), Vicary (2) 175 (CAL); Upper Gangetic plain, Harsukh 21400 (UC), Dehra Dun, Mackinnon s. n. (CAL); Singh 43 (UC). Імро-Снтма: Cambodia, Pierre 805 (L). 1b. CELASTRUS PANICULATUS ssp. serratus (Blanco) Ding Hou, stat. nov. Diosma serrata Blanco, Fl. Filip. 168. 1837; ed. 2. 119. 1845, ex char Celastrus polybotrys Turcz. in Bull. Soc . Nat. Mosc. 31:449. 1858. (та Cuming 1321). NDONESIA: CELEBEs: Prov. Minahassa, Koorders 196348 (1); 2” dap iie а ($); Paloe, Curran 3415 (A). JAVA: Prov. Bawean, Coert 3758 (L) ; Moenoeng, de Voo, 682 (А, L); Poeger, Besochi, Koorders 301788 (L, UC); ers Mousset irf (іл. iin: Nonboumn, T'eysmann 8731 (1); P: de voer 23 38 (А, І. : Ргоу. Вас p INES: drm Ramos s ‚ 1923 (A, UC) Lamao, Cuiran 7499 (SING); Mt. Marive 5 Whi itford ФІ, 414 (Е, US): Prov. eade, Taal Volcano, Gates 7468 I is 1 Mira precise сит, Ramos 1813 (L, МО, SING); Prov. Bulacan, Ramos 346 (A, Е O, US); Cagayan, Cuming 1324 (L, NY, UPS); Prov. Cavite, Ram eroy s к: кетте 1915 (L, ; Prov. Nueva Viscaya, Мс- Gregor s arch-April, 1912 (MO, S); Prov. dy ws Otanes s. n., on a 1914 (A, SING, US); Rizal, Ramos 9 (US); prd Карони ‚ Магсћ US); Prov. Tayabas, Alcasid & Едапо 4669 (А precise locality, Cuming 1209 (L, MO, 5 NAO Cotabat Қ дек $$ и 1910 (Е); Pagpawan Sitio, Davao Ргоу., Edaño s. n. jo ые а, Merrill 044 US). Ic. CELASTRUS PANICULATUS ssp. multiflorus (Roxb.) Ding Hou, stat. nov. Celastrus grey ntn Roxb. Hort. Beng. 18. 1814, nom. nud.; Fl. Ind. % Carey & Wall. :389, ; Fl. Ind. ed. саз геу 1:622. 1832, ех сћаг., поп Гатаг Celastrus pA s Wall. in Roxb. d Ind. ed. Carey & Wall. 2: 389. от. де P sei Батый ы алдыдан 1. Jungh. e 1881-85. (T.: Junghubn s.n., 11). . Gén. 11 1912. (T.: Pierre 2794, P! us euphlebiph- lus ыш Icon. РЇ, ылы 5:15. 1915, ex char. (Т.: Fujii n. Celastrus euphlebiphyllus (Hay.) Makino & Nemoto, Fl. Jap. ed. 2. 597. 19 Celastrus euphlebiphyllus (Hay.) Kanehira, Formos. Trees, ed. 2. 383, f. 340. ке, пот. illegit. 1824, in obs. [Vor. 42 232 ANNALS ОЕ THE MISSOURI BOTANICAL GARDEN Burma: Chin Hills, Prazer 236 (CAL); Chindwin, near Tummoc, Prazer 167 (CAL); Kachin Hill, Sbaik Mokim s. n., 1897 (CAL); Koni, Prazer s. п., May 1888 (CAL); Luay Wyo, Rodger 198 (CAL); Maymyo Hill, Badal Khan 11 (CAL); Myitkyina, Rogers 864 (CAL); Pinmona, Abdul Huk s. n., Aug. 26, 1890 (CAL); southern Shan State, Taungyi, Abdul Khalil s. n., 1894 (CAL); Tavoy, Oomaigoi, Shaik Mokim 266 (CAL, 1, U); Younghue, Carter 466 (CAL); without precise Unde Prazer 54 (CAL), 236 (L). : HAINAN: Ka La, McClure ara (A, MO); Kan-en Hsien, Lau 3658 (A, S); Ngai Hsien, Lau 283 (A, MICH, MO, UC, US); Yaichow, Не 103742 (А x How АСТЫ А, US). TAW nn-yeh Hsien, Wang 80588, 80194 (A); Shunning, Y: GR 202 (A); слаба о, "s Pozo (MO), pate | 12122А (US), 125728 (А), 119724 (US), 11993 (A, МО), 11993“ (US); Gadus Shan (Chang-kiang Dist.), Law 3191 (S) Inpa: Khasia, 1,000—3,000 ft. alt., Hooker & Thompson s. n. (S). Ixpo-CHriNA: Prov. Sonla, 4049 (A, UC); Bao-chiang, Prov. Bien-hoa, Pierre 2794 (lectotype of Celastrus laotica Pitard, Р); Phuoc-than, Thorel 5. п., 1862-66 (paratype of C. laotica Pitard, P) ; Laos, ў юре thane, Spire 289 (paratype of C. otica, Pitard, P). INDONESIA: CELEBES , Lam 2 гаа & у. Hasselt 18 (L); Madiun, есесі. быа pe B (1); Parve- ань Backer r 37252 (L ); Pehalongan, Koordert FEE (L); Soerabaja, Dorgelo 2290 (L). RA: without precise locality, Junghu . (type of Ае ы = а. Ly. ALAY га Lud: Pahang: Kwala бае. Mat s. n., 1893 (SING); Sea Pardens, Perak, Ridley s. n., 1892 Gr Рнплррімеѕ: Palawan ғап Merrill 9311 (А, Е, L, MO, SING, US) Sram: Bangtapham, Keitb 2 3 (SING); Bitsenuloke, Groff 6132 Aie). Kin Sayo, about 120 km. northwest of eut Kostermans 1027 ( у; М апека, Nei River basin, Kostermans 441 (L, SING); WITHOUT PRECISE LOCALITY: Kasin 347 (L), Sowett 752 (L). The wide distribution and the variable leaf forms of this species have been the cause of the many synonyms. Because of the inadequacy of flowering material, I selected the middle leaf of a branch of current season’s or year’s growth for com- parative statistical study, following Woodson’s?” method of measuring leaf char- acters in Asclepias. A study of the herbarium material according to the geo- graphical distribution and pedem differences among the populations has demon- strated that this species includes three distinct populations, which I am treating as three subspecies (pl. 34, upper diagram). Celastrus paniculatus ssp. paniculatus has leaves which are usually suborbicular or broadly obovate. It is chiefly distributed in India, but it extends to Burma, Siam, and Yunnan, China. The leaves of ssp. multiflorus are oblong, and its dis- tribution is chiefly in Burma and Yunnan, where it overlaps ssp. paniculatus. It is also distributed in Hainan, Formosa, Siam, Indo-China, Malaya, Indonesia, and Palawan Isl., in the westernmost Philippines. The leaves of ssp. serratus are usually broadly ovate. This subspecies is chiefly distributed in the Philippines, and is less common in Java where it overlaps ssp. multiflorus. Woodson, R. E., Jr. Some dynamics of leaf variation in Asclepias tuberosa. Ann. Mo. Bot. sid 34:353-432. 1947. ANN. Mo. Bor. Сакр., Vor. 42, 1955 PLATE 34 C. PANICU LATUS О SSP. MULTIFLORUS @ SSP. PANICULATUS ZA 105° 100° 95° о о 90° о о о 85" 80" о е е 75° . . Lt 42 LE ТЕТЕ 18 17 [6 ТЕ 20 ТТ 2.22.3 2.4 2.5 я б t 45 ae 3.0 LEAF LENGTH/WIOTH RATIO Frequency distribution ee subspecies of Celastrus paniculatus: — dots, ssp. paniculatus; small dots, ssp. serratus; pei, sed multiflorus. Explanation in the tex (Vor. 42 234 ANNALS OF THE MISSOURI BOTANICAL GARDEN I have not seen any specimen of C. paniculata collected in Borneo, but I think that it may be found there from the distribution pattern. It is interesting to note from the geographical distribution of these three subspecies that the specimens collected in Palawan Island and Formosa belong to the ssp. multiflorus instead of the Philippine ssp. serratus. Geologically, Palawan Island and Formosa belong to the Malaysian continent. The leaves borne on a branch of current season's or year's growth are slightly variable. If a leaf picked at random from a plant of one subspecies is compared with that of another subspecies the same size and form can be found. Following Anderson’s?® method, I have traced the leaf forms on a current season's or year's branch for each subspecies, and the differences in leaf patterns were easily observed. 2. CELASTRUS NOVOGUINEENSIS Merr. & Perry, in Jour. Arn. Arb. 22:260. 1941. (T.: Clemens 5152, A!). Scandent shrubs; branchlets terete, glabrous, brown to reddish-brown; the lenticels numerous, elliptic, elevated, white; axillary buds conoid, obtuse, about 2 mm. long. Leaves oblong, ovate to elliptic, the apex acute or obtuse, the base obtuse or rotund, the margins minutely and remotely serrulate, 12—16 cm. long, 5.0—7.5 cm. wide, firmly membranous, glabrous, the primary veins 6 pairs, elevated below, distinct, plane or slightly elevated above, the veinlets obscure below, plainly visible above; stipules linear, about 2.5 mm. long; petioles 1.0—1.5 cm. long. Inflorescences terminal, paniculiform, usually thrice compound, about 15—20 cm. long, the floriferous branchlets divaricate, the peduncles glabrous, the primary peduncles about 6—10 mm. long; flowers green, dioecious, the pedicels about 1 mm. long, the articulation varying from the lower to upper half of the stalk. Male flowers: calyx lobes imbricate, subreniform, slightly erose, about 1 mm. long; petals oblong, rotund, subentire, scarious-marginate, about 2.5 mm. long and 1.5 mm. wide, subfleshy, brownish-punctate; disc cup-shaped, membranous, the lobes obscure, truncate or mucronate; stamens subsessile or with very short flat filaments, the anthers narrow-ovoid, obtuse, apiculate, brownish-dotted; sterile pistil 1 mm. long. Female flowers: calyx lobes, petals, and disc as in the male; sterile stamens 1 mm. long; pistil about 1.4 mm. long, the ovary subglobose, wider than long, the style obsolete, the stigmata 3-lobed, each bifid. Fruits subglobose, the valves broadly ovate, about 9—12 mm. long and 7-10 mm. wide, 3- to 6-seeded; seeds ellipsoid, 5—8 mm. long and 2.5—3.5 mm. wide, reddish-brown, smooth. In thickets, at altitudes 1,220-1,830 m.; New Guinea, flowering from February to September. New GUINEA: NORTHEAST NEW GUINEA: Morobe-Dist.: Oguawuang (Ogeramnang), J. 8 M. S. Clemens 5152 (А, (pe 5394: (A); Wantoat (Wantot), к 11003 (А); Yoangen (Yunzaing), J. 9 М. S. Clemens 3523, 6606, 7230 (А). Таја River, Ca 15608 (А). beim NEW GUINEA: Meervlikte; ни ы. v. Leeuwen 1 Bh 28 Anderson, я Concordant versus discordant variation in relation to introgression. Evolution 5:133—141. 195 19551 HOU—REVISION OF THE GENUS НЕН; Е LEES ii 7, E TN X \ ый Wow --./ CELASTRUS 235 а А | SP. MI ares Sp. MU dicic р = i DE : EI 58: o= z Е a SERRATI їв DE Cg ба * ТН СЕ LE NOCT АО ; 2 : CY vir x 1 1 r к, xd e 7. C. SCANDENS ~ " Map 1. Distribution of species of Celastrus, Subgenus CELASTRUS, Series I. PANICULATI. [Vor. 42 236 ANNALS OF THE MISSOURI BOTANICAL GARDEN ' This is an endemic species of New Guinea. It is distinctive and can be dis- tinguished from other species by the thrice compound aggregate dichasia and main floriferous axis with branchlets extending out at nearly right angles. The stamens in the male flowers usually are sessile or with very short, flat filaments. The pistil in the female flower has a very short style. 3. CELAsTRUs sUBsPICATUS Hook. Icon. Pl. 5:#. 462. 1842. (Т.: Collector un- known, s. n., cultivated at Kew, K!). Celastrus australis Harv. & Muell. in Trans. Phil. Soc. Vict. 1:41. 1855; Benth. Fl. Austral 1:398. 1863, ex char. (T.: Mueller s. Celastrus papuana Warb. in Engl. Bot. Jahrb. 15:366. 71891, ex char. (Т.: Warburg s. A Celastrus "— Willd. var. Balansae Loes. in Engl. Bot. Jahrb. 39:160. 1906, char. ( Balansa Scandent shrubs; branchlets terete, brownish-pubescent, the lenticels elliptic, distinct, and white; axillary buds deltoid, spiniform, about 1-2 mm. long, the outermost scales acuminate. Leaves elliptic, ovate, suborbiculate, or ovate- lanceolate, the apex acuminate, the base obtuse to acute, the margins subentire to remotely serrulate, 4-12 cm. long, 1-6 cm. wide, membranous, glabrous, the primary lateral veins 5—8 pairs, the veins and veinlets slightly elevated on bot surfaces; stipules linear, about 2.5 mm. long; petioles usually 4—7 mm. long. In- florescences terminal, once compound, rarely twice compound in the male, 2-9 cm. long, occasional in the axils of the uppermost leaves, the peduncles usually brownish-puberulous; flowers white, the pedicels about 3 mm. long, the articula- tion toward the base of the stalk. Male flowers (young): anthers ovoid, obtuse. Female flowers: calyx lobes imbricate, orbicular, broad, ciliate, about 1 mm. long; petals ovate-oblong, rotund, slightly erose, about 2.5 mm. long and 1.0-1.5 mm. wide; disc cup-shaped, the lobes obscure, truncate; sterile stamens 0.5 mm. long, the filaments very short; pistil ovoid, 2 mm. long, the style short, the stigmata 3-lobed, spread. Fruits subglobose, the valves broadly elliptic, about 5—10 long and 5—8 mm. wide, 3- to 6-seeded; seeds ellipsoid, 5 mm. long and 2.5 mm. wide, reddish-brown, the areolae obscure. In the кык Australia, New Caledonia, and Papua. Aus : QUEENSLAND: Cedar Beenleigh, White s. n., Мау 1920 (SING); head of Burnett Ет i Miller s $. п. (GH); Benarkin, Cameron s. n., April 1924 (A); Mt. Glorious, say, Wbite 12 517 € US); Yarraman d Forest Sack Clemens s. n., Aug. ба vf (UC); without definite locality: nieces! 1124 (GH), Clemens 43083 (A). ws SOUTH WALES: Suc ‚ Caley s.n., 1807 (A, UC, US); Tweed River, Moore end wet а x ke. collector unknown s.n. (holotype of Celastrus пашу ие id CALEDONIA: Balansa 1870 (A); ve нууч a (А). New Gone v PAPUA: Kanosia, Carr 11260 (A, L). Celastrus subspicatus was published by W. L Hooker, based on a plant long cultivated in the Royal Botanic Gardens at Kew. Its history and habitat are unknown. In the type specimen, borrowed from Kew Gardens, the leaf forms, texture, short petioles, distinct and elevated lenticels, acuminate outermost 1955] HOU-—REVISION OF THE GENUS CELASTRUS 237. bud scales, and inflorescence patterns are similar to Celastrus australis. ‘This simi- larity leads me to believe that Celastrus subspicatus of Kew Gardens was grown from seeds obtained from Australia. Specimens collected from New Caledonia and lowland Papua, New Guinea, are morphologically similar and geographically related to the Australian ones except in the leaf shape. Meanwhile, not enough specimens are available for a population study to find out whether they really differ from the Australian species and whether they should be put into subspecific or varietal category, so they are treated tenta- tively as one species. 4. CELASTRUs RICHU А. Gray, in Wilkes U. S. Expl. Exped. (Bot. Phanerogam., pt. 1):289. 1854. (T.: Wilkes Exped. s. n., GH!). Scandent shrubs; branchlets terete, glabrous, light to dark brown, the lenticels distinct and numerous, white and elliptic, rarely lacking on the young branchlets; axillary buds small, conic, about 1 mm. long. Leaves elliptic, the apex and the base acute, the margins subentire or remotely serrulate, thinly membranous, shining above in dry condition, glabrous, the primary veins 5—7 pairs, elevated on both surfaces, the veinlets distinct on both surfaces, densely reticulate; stipules laciniate, filiform, about 1 mm. long; petioles 2-9 mm. long. Inflorescences terminal, once or twice compound, 2-7 cm. long, the peduncles glabrous, the primary peduncles 3.5—5.5 mm. long; flowers white, the pedicels 1.0—2.5 mm. long, accrescent, up to 15 mm. long on the fruits, the articulation toward the base of the stalk. Male flowers: calyx lobes imbricate, subreniform or semi-orbicular, about 1 mm. long, ciliate; petals oblong, entire, 3 mm. long and 2 mm. wide; disc concave, mem- branous, the lobes obscure, mucronate, the margins confluent with the base of the filaments; stamens about 2 mm. long, the filaments short, flat, broad at the base, gradually attenuated toward the top, glabrous, the anthers ovoid, slightly cordate and apiculate; sterile pistil conoid, 1.5 mm. long. Female flowers: calyx lobes and petals as in the male; disc cup-shaped, membranous, the lobes subreniform; sterile stamens 0.8 mm. long; pistil 2 mm. long, the ovary subglobose, the style short, the stigmata 3-lobed, each bifid. Fruits subglobose, the valves suborbicular or obovate, about 9 mm. long and 7—8 mm. wide; seeds ellipsoid, about 4.5 mm. long and 3 mm. wide, reddish-brown, distinctly areolate. In sunny forest; at altitudes 30-900 m.; Fiji Isl.; flowering from December to May. NGA: Smith 1122 (GH, МУ, 5, UC, US). міті Levu: Lautoka, ома 900 pi pos Smith 4301 (A, US), 5827, 6085 (A); Tholo, Degener 14798 (М Y, US), 14940, 14973 (A, NY, US), 14978 (A, MICH, NY), 15050, 15315 (A, MICH, NY, 5 US) ; Gillespie 4182 (GH, NY, UC); -— Pots 153 > (А, Б, $, US) ; Ra, Degener 15352 (А, NY), 1 (A, MICH, NY, h . WITHOUT PRECISE E Hed ; M rA e Mrs. iden. 3 (GHD; Wilkes ad, 5. n., 1838-42 (GH, holotype). [Vor. 42 238 ANNALS OF THE MISSOURI BOTANICAL GARDEN This is an endemic species of Fiji. Its geographical distribution and accrescent pedicels make this species easily separable from other species. А specimen collected by Degener (15330) in an "arid patch of forest along coast" near Nandi has obovoid fruits while all others have subgloboid ones. 5. CELASTRUS MADAGASCARIENSIS Loes. in Notiz. Bot. Gart. Berlin 13:215. 1936, ex char. (T.: Perrier de la Batbie 2067). Scandent shrubs; branches terete, glabrescent, brown, the lenticels elliptic or orbicular, slightly elevated; axillary buds globose to deltoid, about 1 mm. long. Leaves usually elliptic to ovate-oblong, the apex acute to shortly acuminate, the base cuneate or obtuse, the margins crenate-serrulate, 4—7 cm. long, 2.5—3.5 cm. wide, chartaceous to firmly membranous, glabrous, the primary lateral veins 5-7 pairs, elevated on both surfaces, curved toward the apex, the veinlets distinct and slightly elevated on both surfaces; stipules laciniate, about 2 mm. long; petioles 5-11 mm. long. Inflorescences terminal, paniculiform, once compound, 1—3 cm. long, the peduncle puberulous, the primary peduncles 2-3 mm. long; flowers dioecious, the pedicels about 2.5 mm. long, the articulation toward the base of the stalk. Male flowers: calyx lobes imbricate, ovate, obtuse, slightly erose, about 1 mm. long; petals elliptic to ovate, obtuse and subentire, about 3 mm. long and 2 mm. wide; disc membranous, the lobes inconspicuous, cuspidate; stamens arising from the margin of the disc, about 3 mm. long, the filaments shorter than the anthers, glabrous, the anthers ellipsoid, distinctly apiculate; sterile pistil columnar, about 1 mm. long. Female flowers: calyx lobes, petals and disc as in the male; sterile stamens about 1 mm. long; pistil about 2.5 mm. long, the ovary globose, the style shortly columnar, the stigma discoid. Fruit unknown. In forests; at altitudes 100—1,500 m.; endemic in Madagascar; flowering from October to November. ADAGASCAR: Valley of the Mandraré, d’Alleizette 1288 (L); Basaltic near Betafo, FA 5. *t., Sept. 1905 (L). This species quite clearly belongs to the series PANICULATI by its distinct terminal paniculiform inflorescences. It is closely related to the Australian Celastrus subspicatus, rather than to the Asiatic C. paniculatus, by its short and once compound inflorescences. It can be distinguished from those two species by its apiculate anthers, cuspidate disc-lobes, and discoid stigma. Loesener has pointed out that the Celastrus species of Madagascar enumerated by Drake?) are Сутпо- sporia. The excellent illustrations reveal that Loesener's view is quite right. Thus Celastrus madagascariensis is the only remaining Celastrus species described from Madagascar. 29 Drake del Castillo, M. E., F Do Hist., Phys., Nat. et Pol. Madag. 36 (Hist. Nat. Pl. Atlas 3):pls. 280, 2802, 2800, 1955] HOU-——REVISION OF THE GENUS CELASTRUS 239 TOES К. | LNCS 3% 9 AWRY раја б Ме Fig. 4. Celastrus angulatus Maxim. 6. CELASTRUS ANGULATUS Maxim. in Bull. Acad. Sci. St. Pétersb. III, 27:455. 1881 (as angulata); in Mél. Biol. Acad. St. Pétersb. 11:199. 1881, ex char. (T.: Piaseski s. n.). Celastrus latifolius Hemsl. in Jour. Linn. Soc. 23:123. 1886; Oliver, in Hook. Icon. Pl. і. 2206. 1894, ex char. & Ш. (T.: Henry s. n.). Scandent shrubs up to 10 m. tall; branches subterete, the branchlets usually angular, glabrous, both reddish to dark brown, shining, densely lenticellate, the lenticels small, orbicular or ovate, white; axillary buds short-ovoid, about 2-5 mm. long. Leaves broadly elliptic, broadly ovate to suborbicular, the apex abruptly acute, the base obtuse, the margins crenate-serrate, usually 9-16 cm. long, 7-15 cm. wide, firmly membranous, glabrous or pubescent on the veins below, the primary lateral veins usually 6—7 pairs, prominently elevated below, plane or slightly elevated above, the secondary lateral veins slightly elevated below, plane and visible above, nearly parallel to one another; stipules filiform, tufted; petioles 0.6—3.0 cm. long. Inflorescences terminal, densely paniculiform, up to 20 cm. long, the peduncles brownish-puberulous, the primary peduncles 5-8 mm. long; flowers small, greenish-white or yellowish-green, the pedicels obscure, the articula- [Vor. 42 240 ANNALS OF THE MISSOURI BOTANICAL GARDEN tion just beneath the flower. Male flowers: calyx lobes open, ovate or oblong-ovate, obtuse, subentire, about 1.5 mm. long; petals elliptic-oblong, obtuse, ciliate to slightly erose, reflexed, about 2.5 mm. long and 1.2 mm. wide; disc subfleshy, nearly flat, plate-like, the lobes usually depressed-subquadrate; stamens attached just under the margin of the disc, about 3 mm. long, the filaments filiform, glabrous, the anthers ovoid, obtuse, apiculate; sterile pistil ovoid, about 1 mm. long. Female flowers: calyx lobes, petals and disc as in the male; sterile stamens about 1 mm. long; pistil about 3-4 mm. long, the ovary subglobose, the style distinctly columnar, the stigmata 3-lobed, each bifid, recurved. Fruits subglobose, the valves broadly elliptic to suborbicular, about 6—9 mm. long and 5-9 mm. wide, 3- to 6- seeded; seeds ellipsoid, about 3—5 mm. long and 1.5—3.0 mm. wide, brownish-red, shining, the areolae usually obscure, rarely distinct. In open places, light woods and thickets; at altitudes of 400—3,600 m.; China; flowering from May to INA: HONAN: Hia Hsien, idis 1831 (A); Hwei wem Hers 772 (A); Lushih, Hiung-eul Shan, Hers 861 (A); g Hsien, Hers 1288, 1314 (A); Teng-feng Hsien, Hers 269 (А); Tsi-yuan Hsien, 2. de (A); Yungning, йт 417 (А). HuPEI: Chang- yang, Wilson 753A (A); Enshih Hsien, Ho-ch’ang Chow 2005 (A, NY); Fang Hsien, Mes 364", 753 (А, Е, GH, МО, US); Ho-long-teng, Silvertri 5099 (A); Ichang, Chun Chien : 6050 (UC), Henry 2084, 3405, 3883 (GH), 5925 (A, GH, NY), Wilson 364 М Е, MO, US); Ma-pan-scian, Silvestri 1336 (А); Nanto, Wilson 783 (A); Paokan, Wilson po (A, NY); Patung Hsien, Ho-cb'ang Chow 495, 597 (A, NY); Sin-yeh-sie, Chun 3724 (А), 4066 (US); теси у вей, Ho-ch’ang Chow 405 (А, NY). SU: ienschui, Fenzel 2817 (A). KWANGsI: Ch'uan Hsien, Tsang 27666 (US). KWANGTUNG: oh-ch'ang, MO, 1 G, J: Kiangkou, Steward et al 541 (А , F, NY, S, US); Kweiyang, Tieng 8653 (A, NY); 5 у 07 (А; Е, МУ; 5, U А Tuhshan, Тане 6098 (А, МУ); риса, Tsiang 4945, 4970 (А, NY). sHENsI: -Shan, Hers 3097 (А); Shih Tan, Hers 2442 (A); Tai-pei Shan, Purdom 946 (A,US), у Роа 738 (А); Tsi-yuan Hsien, Hers 2670 (A) ; Tsinling Shan, Fenzel 678 (A), Hers bé ig: 3010 (A); Tze-wu Hsien, Meyer 1722 d SZECHUAN: Li-fan 70 а ара sien, Wang 2 A); Kwang-yun Н and vicinity, Wang 22633 (A); inter Nitou et linpin, Smitb 13 UPS); -chuan, i 571 (A), = (А, 5245 еп- chuan, Wilson 364" (А); without precise locality 8 v. Rosthorn 700 (А). YUNNAN: Eche-hay, Maire 385 (A); Hsinlung, Handel-Maxxetii 5692 (A); rone а Maire 22, 393 (А), 6224 (UC). This species has the largest leaves of any species of Celastrus. They reach as much as 16 cm. in length and 15 cm. in width. The characteristic leaves and angular branchlets with dense, small and whitish lenticels make it easily dis- tinguishable from allied species. I saw a living plant of this species cultivated at the Brooklyn Botanic Garden, New York, which has long and dense, paniculiform fruiting branches. During winter and early spring the dehiscing fruits with their crimson arillate seeds present a beautiful and eye-catching sight. This species would make a valuable ornamental plant although it is not now used extensively. 7. CELASTRUS SCANDENS Г. Sp. Pl. 196. 1753; Gray, Gen. Pl. U. S. 2:186, Ё. 170. 1849, ex char. Celastrus bullatus L. Sp. Pl. 196. 1753, ex char. Euonymoides scandens (L.) Moench, Meth. Pl. 70. 1794. 19551 HOU—REVISION OF THE GENUS CELASTRUS 241 Scandent shrubs up to 20 m. tall; branches terete, glabrous, castaneous, the lenticels scattered, orbicular or elliptic, the current year's growth yellowish-green, with lenticels lacking or inconspicuous; axillary buds subglobose, about 1.5 mm. long. Leaves elliptic, ovate-elliptic to obovate, the apex acute to short-acuminate, the base rotund to cuneate, the margins crenate-serrulate, incurved, glandular- mucronate, 4-12 cm. long, 2-6 cm. wide, membranous, glabrous; the primary lateral veins 4—6 pairs, distinct and slightly elevated on both surfaces, the veinlets visible on both surfaces; stipules filiform, tufted, about 1.5 mm. long; petioles usually 1.0-2.5 cm. long. Inflorescences terminal, usually once compound, 2-8 cm. long, the peduncles glabrous, the primary peduncles obscure to 5 mm. long; flowers dioecious, green, the pedicels obscure to 3.5 mm. long, the articulation at the upper half of the stalk. Male flowers: calyx lobes open, oblong, rotund, slightly erose, accrescent, about 1 mm. long; petals thin, more or less transparent, rotund, slightly erose, about 4 mm. long, 1.5 mm. wide; disc cup-shaped, the lobes trun- cate; stamens about 2.5 mm. long, the filaments slender, glabrous, the anthers ovoid, obtuse, apiculate; sterile pistil 1.2 mm. long. Female flowers: calyx lobes, petals, and disc as in the male; sterile stamens about 1 mm. long; pistil flask-shaped, about 3 mm. long, the ovary ellipsoid, the style cylindric, the stigmata 3-lobed, patent. Fruiting inflorescences usually racemiform; fruits su globose, the valves sub- orbicular or broadly elliptic, about 8—12 mm. long and 7-10 mm. wide; seeds ellipsoid or slightly plano-convex, obtuse at the base, attenuate toward the apex, about 4.5 mm. long and 2 mm. wide, brown, the areolae obscure. Roadsides and thickets; lowlands up to 600 m. alt.; North America; flowering from May to June. CANADA: ONTARIO: Algoma, Bruce, Essex, Frontenac, Glengarry, Leeds, Renfrew, Waterloo, Welland, and York counties. QUEBEC: Brome, Chambly, Chateauguay, Mis- sisquoi, and Pontiac counties. UNITED STATES: ARKANSAS: Benton County. CONNECTICUT: Litchfield and New sin counties. DELAWARE: Newcastle County. ILLINOIS: A a son counties. KEN : Boyd, Fayette, Greenup, Rockcastle, an ren counties. MAINE: Cumberland, Kennebec, Penobscot, and York counties. MARYLAND: Caroline, Cecil, Harford, Montgomery, an albot counties. MASSACHUSETTS: Berkshire, Dukes, Essex, Franklin, Hampden, Hampshire, Middlesex, Norfolk, and Suffolk counties. MICHI- сам: Cheboygan, Emmet, Keweenaw, Menominee, Midland, and Oceana counties. E sora: Clearwater, Goodhue, Hennepin, Houston, Kandiyohi, Nicollet, and Otter Tail counties. MISSOURI: Iron, Jackson, Jasper, Jefferson, Johnson, Lafayette, Lawrence, Oregon, zark, Polk, Ralls, Reynolds, St. Clair, St. Louis, Stoddard, Stone, Taney, Warren, and Wright counties. NEB i i EW RASKA: Antelope, Cass, Franklin, Lincoln, Sioux, Thomas, and York counties. HAMPSHIRE: Cheshire, Grafton, Hillsboro, Merrimack, and Rockingham counties. JERSEY: Atlantic, Bergen, Burlington, Camden, Cape May, Cumberland, Gloucester, Hun- terdon, Mercer, Middlesex, Monmouth, Ocean, Salem, Somerset, Sussex, and Warren counties. NEW YORK: Albany, Cayuga, Columbia, Dutchess, Orange, Saratoga, Suffolk, St. Lawrence, Tompkins, Ulster, Warren, and W ington counties. NORTH CAROLINA: Buncombe, Iredell, and Mitchell counties. NORTH DAKOTA: Pembina, Ramsey, and Stark [Vor. 42 242 ANNALS OF THE MISSOURI BOTANICAL GARDEN counties. OHIO: Clinton, Cuyahoga, Gallia, peony Harrison, Jackson, Lawrence, Meigs, Richland, Sciota, and Wood counties. OKLAHOMA: Blaine, Cleveland, Johnston, Kay, Murray, Payne, and Woo gae counties. PENNSYLVANIA: Adams, Allegheny, Beaver, Bedford, Berks, Blair, Bucks, Cambria, Carbon, Center, Chester, Clarion, Clearfield, tin Jefferson, Juniata, vr ior Lebanon, Lehigh, Luzerne, Lyco: ‚ Montgom Montour, Northam ‚ Northumberland, Peery, Pali Bic, ‘Schuylkill, eed Somerset, Sullivan, Susquehanna, Tioga, Union, Venan oreland, and York counties. RHODE ISL : Prov idence Cou м с океанами Више, Grant, Lawrence, and Белдік UR counties. T ча EE: Crockett, Dyer, Gibson, Hamb- len, Obion, and White Counter TEXA s: Armstrong, Brewster, Culberson, Jeff Davis, and sle of Wig t c WEST VIRGINIA: Cabell, "rei Summers, and counties. WIS lo, WYOMING: roo This is the ae ibm distributed Celastrus species in North America. It is also commonly cultivated as an ornamental plant in the United States. This species is easily recognized by its paniculiform inflorescences, open calyx lobes, cup- shaped disc, and especially by the crimson arillate seeds of the opening fruits. Since it is widely distributed and the only commonly known species of Celastrus in North America, it will not be confused in the field or in the herbarium. I have seen many specimens of this species all of which are identified correctly. In the present treatment, I have not cited individual collections, but just listed the counties where they are found. All of the collectors and their field numbers can be found in the Index to Exsiccatae at the end of this paper. The variations of morphological characters and geographical distribution of this species have been discussed under the section on general morphology. Series 2. AXILLARES Rehd. & Wils. in Sarg. Pl. Wils. 2:355. 1915. Sempervirentes Maxim. ex Rehd. & Wils. in Sarg. Pl. Wils. 2:357. 1915. KEY TO THE SPECIES interc B. Flowering оком terminal both in male and in female plants; inflorescences both lateral and terminal. б. Evergreen Зери ‚ the current season's branches usually subtended by foliage lea inflor orescences usually without accompanying vegetative buds (except Ж. Flowering branches terminal or rarely in some species those of the female plants alar elastrus bindsi р 4 D. Young branches with distinct lenticels; flowers numerous in dense umbelli- form cymes; discs subfleshy or fleshy, flat; — not ie €— E. Fruit conspicuously stipitate; > about 15 wide. India, Burma, Siam, Indo-China, лыс: and nn doses. 8 . monospermus EE. Fruit not stipitate; seeds about 9 mm. long, 5 mm. wide. Мајауа, Indo- nesia, Philippines and New Guinea . monospermoides DD. EN omes rarely with lenticels; flowers few in simple lax cymes, or кн m cup-shaped; пе densely reticulate; seeds about x s =: m. wide. Southeastern Asi то. C. bindsii СС. савані стій tec current season's branc М not seston а е leaves; inflorescences with a ing vegetative buds; frui F. a fasciculate « ог пара еј ithout distinct наа ог ій very short n the terminal Re idis calyx lobes imbricate; petals thin and 1955] HOU—REVISION OF THE GENUS CELASTRUS 243 with veri venation, 2.5—3.0 mm. long, 1.0-1.5 mm. wide; anthers of the male flowers 0.7—1.0 mm. long; fruiting inflorescence erect, the stalks of fruits not accrescent after anthesis, about 1.0—1.5 cm. long; seeds ellipsoid у ог оуо с. Terminal aes pe usually 2—6 cm. ME сар inflorescences usually veral- to many-flowered, ат frequen H. M me ts ar leaves usually immersed ior иза elevated beneath; seeds г кокс? of male flowers not apiculate; disc lobes subquadrate to ob- long, obtuse or sli а ет sion leaves usually paler beneath Chi I glencopbyllus Southwestern and cen a China and Indo-China II. ко б Е 3 owers apiculate; disc lobes subreniform, mucronate; a, Burma, and China 12. C. hookeri I lea color: HH. Veinlets Of ine deo. od both "Pirfsces; seeds highly jure ndia. . membranifolius GG. Terminal жененин usually 6—18 cm. long, axillary ires few- ppressed, fruiting infrequently. Southwestern Chi 14. C. vanioti FF. Flowers in ук ре ‘wale unculate dichasia on the lower two-thi ~ of the ; relatively thicker and nd terminal orescences; calyx lobes Bee Haag ДЕМ without distinct venation, abot A= ong a wide; anthers of the meg flowers 1.5 m ong; iting пе при "pendulous the m of th ari эссесе ты anthesis, about 2—4 . long; seeds LT x western Chi bypoleucus p. only; de both шелі с, o BB. Flowering | ни nal in male pla termina male ыд арі only i in female plan usly pedonculite; pedicels of fruits green. RE arcs uri dichasial, conspicuous K. as d buds relatively large. í ‚ 5-11 mm. long; leaves y ovate, rarely broadly elliptic, the secondary veinlets densely ulate. China south of t gtze 16. C. gemmatus Tain buds relatively gris conoid, obtuse, 1-3 mm. long; leaves usually broadly ellipt x or broadly obovate to nearly rbieni, the secondary vein lets sli dr рона. d elevated; anthers not apiculate. China = of the Yangtze R . С. orbiculatus JJ. Inflorescences ic бесіне ог КЕ usually sessile, iy very shortly pedunculate; cels of fruits bro L. Disc lobes broader th feed about пез t one-fifth as long as the disc proper, the cedes nid at the ау аре ог lower half of the жай. С. rostbornianus Бгоа тайсам d elevated; anthers apic Riv _ ngtze n broad, ple half to one-third as long as the pred e articulation usua = at the upper third or upper hal stalk. Sod iter China, Formosa, southern Japan, and Ryukyu sands . C. punctatus China uk of the Yan i nge AA. точ branches usually distinctly intercalary. . Flowe n pedunculate cymes, axillary or ИРИ leaf margins ѕег- rs usually in rate, pon petioles less than half as gue as the N. Seeds slightly lunate, үнө, Мен => usually erose or decns elevated on both s econdary veinlets of lea paid usually аты увео pn goer й disc subfleshy, у ог r slightly Р. “Plants ogy св я сау pubescent on young flowering branchlets and t eaves es usually suborbicular or gorge бымен lobes mdi hes go Кук, Kwangtung, Hainan, and F ое 20. С. kusanoi РР. — deu yellowish brown-pubescent, especially on both surfaces of es; leaves usually b bend ovate or broadly obovate; disc lobes rig aei еі rcuate. Burma, China, and ІпФо-СВіпа......................... ... . Leaves indir с ог Якобс -oblong; disc membranous, i n Eastern India, чайн и 22. с. stylosus NN. Seeds semi-annular, жы at both ends; sepals десен т уе! — of leaves obscure on both surfaces. a [OTORGA MM. Flowers usually solitary and axillary, or rarely fasciculate; a rgins ~ саге метке the petioles nearly half as bu as the blades. Northeastern and rn China 4. C. flagellaris , Korea, and Japan . С. hirsutus [Vor. 42 244 ANNALS OF THE MISSOURI BOTANICAL GARDEN Fig. 5. Celastrus monospermus Roxb. 8. CELASTRUS MONOSPERMUS Roxb. Hort. Beng. 18. 1814, пот. nud.; Fl. Ind. ed. Carey & Wall. 2:394. 1824, ed. 2, 1:625. 1832 (as monosperma). (Т.: Desiloa s. n. NY!). Catha iig Gardn. & Champ. in Hook. Kew Jour. 1:310. 1849, ex char. (Т.: bampion s. n.). Celastrus chaidii Benth. in Hook. loc. cit. 3:334. 1851, ex char, (based on Catha benthamii Gardn. & Champ. Celastrus benthamii (Gard. & Champ.) Rehd. & Wils. in Sarg. Pl. Wils. 2:358. 1916 Celastrus annamensis Tardieu, in Fl. Gen. indo Ching § 5. 1:803, fig. 96, t. 7. 1948, har. (T.: Poilane 24300). Scandent shrubs up to 10 m. tall; branches terete, smooth, glabrous, reddish or dark brown, the lenticels ovate to orbicular, sparse to dense, sometimes obsolete; axillary buds conoid, about 2 mm. long. Leaves variable, elliptic to oblong or broadly ovate, the apex acute to obtuse, the base cuneate to obtuse, the margins serrate, 7-17 cm. long, 3—9 cm. wide, firmly membranous, glabrous on both sur- faces, the primary lateral veins 5—8 pairs, curved toward the apex, elevated below, immersed above, the veinlets distinct, slightly elevated below, immersed and o lete above; stipules laciniate, about 1 mm. long; petioles 0.5—1.5 cm. long. In- 19551 HOU—REVISION OF THE GENUS CELASTRUS 245 florescences axillary, simply dichotomous or paniculiform, few and laxly branched, up to 20 cm. long, sometimes longer than the subtending leaf, distinctly peduncu- late, the peduncles usually glabrous, rarely pubescent, the primary peduncles 0.5—3.0 cm. long; flowers dioecious, small, greenish-yellow or white, the pedicels obsolete to 4 mm. long, the articulation toward the base of the stalk. Male flowers: calyx lobes imbricate, semi-orbicular, slightly glandular-ciliate, about 1 mm. long; petals shortly oblong, about 2.5 mm. long and 2 mm. wide; disc fleshy, annular, slightly lobed, the lobes entire, subquadrate, broader than long; stamens attached just under the margin of the disc, about 2.7 mm. long, the filaments filiform, the anthers slightly ovoid, pink-punctate, slightly apiculate; sterile pistil 1.3 mm. long, ovoid, slightly immersed in the disc. Female flowers: calyx lobes, petals, and disc as in the male, usually the petals smaller than in the male; sterile stamens about 1 mm. long; pistil about 3 mm. long, the ovary subglobose, narrowed into the style, the stigmata 3-lobed, reflexed. Fruits cylindric, stipitate, 1-seeded, the valves ovate- oblong, about 2 cm. long and 1 ст. wide; seeds cylindric, about 1.5 cm. long and 0.8 cm. wide, smooth, pinkish brown. Chiefly in thickets, at altitudes from 246 to 1,880 m.; commonly distributed in India, Pakistan, Burma, Indo-China, and China; flowering from March to June. BurMa: Pang Hoi Phi and Peng Sai, Rock 2237 (A, US); without precise locality, Prazer s.m. С). CHINA: HAINAN: Bak Sa, Lau 2628 5 26624 (A); Fan Yah, Chun 8 Tso 4416 (А, locality, Liang 63419 (А, L, NY, S, US), 65253 (А, МҮ, S, US), Wang 35222 (МҮ), 35803, de (А, МУ), 36607 (МУ). но Bako Shan, western Розећ, Ching 7525 (A, UC, US); Me-kon, southern Nanning, Ching 8433 (А, NY, UC, US); Ping Nam Hsien, Wang od; 5 (А); я taai Shan, Shang-sze Hsien, Tsang 22070 vd S), 24106, 24166, ir (А, MO, NY); Yeo-mar Shan, northern Hin-yen, Ching 7120 (NY). KWANGTUNG: Kowloon, Wang 3056 (NY); Lofou Shan, Ho 60086 (NY), Ko pos (NY), Tsiang 1705 (A, МУ, UC); Lung-tau Shan, near Iu, Kan-peng To et al 473 (ОС); Sin-fung Hsien, Taam 720 (A); muse Ко 51463 (A); Tai-o, Chun 310 еф ~ Тари T ; Wung-yuen Hsien, Lau 800 (А, МУ, SING), 2 о (А); Tsang 8 Wong 2681 (UC), Tsui 348 (MO, NY). EASTERN TIBET А green WESTERN CHINA: without precise locality, Forrest 26407 (А). YUNNAN: Che-li, Wang 75920, 78458, 79015, 70450, 70776 (A); Chen Kang, Yu 17405 (A); Fo-hai, Wang 73570, 73590, 73825, 73839, 74430, 74907, 77202" (А); Jenn-yeh Hsien, Wang 80312, 80781 (A); Lan-tsang Hsien, Wang 73415 (A); Lung-ling, Tsai 55679 (A); Lu-se, Tsai ry 10446 NY, US); 8 ЮМ ~ 5 3 да У 24 “~ 77 Un $3 rr T 2 99 T x i pt a em = = LR ©з 8 5 = 9- wn between Palut and Nam-dip, on the trail from Raheng to Mesawt, Ping-pien Hsien, Tsai 55076, 55130, 60152, 60450, 61680, 62295 (А); xdg рум bles 11874 (UC); Si-chour Hsien, Feng 11445, 11530, 12469 (A); Sunning, Hila, У зоб МАЈ Szemao, Henry 11972 (А, ME 119728 (А, A, MO, NY), Rock 270, 9 2823 А, Wen-sha ity, iu 8201 tay HONGKONG сзи п 401 ў усн a, Chun 4831 MO. pre (A, UC); LE I (NY); Таат 1792 (А, US), 2087 (US); Lanto, Tsieng 720 (UC); Wrigbt г. ‚ 92 (СН). Е do Darrang, Upendranath Kanjilal 3801 (CAL); Jawai, Dr. King’s col- lector s. п. (L); Khasi Hills, Kurz 182 (CAL); Kohima, Dr. King’s collector 178 (U, [Vor. 42 246 ANNALS OF THE MISSOURI BOTANICAL GARDEN UPS); Коровы Dr. King's collector 233 | (A); Namchung, Luckimpore, сени 379498 (CAL); Naga Hills, Jagarmani 499 (А, US); Dr. King's collector 795 (A, CAL), 1 Сани Silhet (as Sillet), Say 5. 9». (NY, type of Celastrus Honoran} а у mson s.m. (AS СН, Г, $); Khasia, Liam, Clarke 451528 (CAL); without precise locality, Griffith 666 ( CAL). BENGAL: зас, Biswas 2033 (А, NY); Darjeeling, Lace 2229 (CAL) ; Darjeeling, Anderson 107 (CAL); bé en Kurz s. n. (CAL) ; Runquo Valley, Anderson 108, 1 m (CAL); Sureil, Агаш I7 (CAL). INDO-CHINA: Chapa Tonkin, Pételot has d hes 5931 (A), 6368 (A, NY, US). PAKISTAN: Griffith 1992 (CAL, GH, Sy. This species is distinguished mb к stipitate fruit bases, the one-seeded capsules, the large cylindric seeds, the characteristic fleshy annular disc, and the attachment of the stamens immediately beneath the margin of the disc. he synonymy of this species is due to the failure, first, of Bentham to appre- ciate the true nature of Roxburgh's Celastrus monospermus, and later, of Gardner and Champion, of Bentham, and still later of Rehder and Wilson to appreciate the variability of that species. My present interpretation is that of Lawson (in Hook. Fl. Brit. Ind. 1:618. 1875) Generally speaking, plants of Yunnan, Burma, and India have oblong or elliptic- oblong leaves, while those of Kwangtung and Kwangsi have broadly elliptic or broadly ovate leaves. At first I attempted either to maintain these plants as two species or to unite them into one species with two subspecies having differing leaf characters. Since then, I have measured the length and width and angles A (apical) and B (basal)?? from the different areas, and found that the frequency curves of the ratio of length and width, and the angles A and B overlap. Since leaf shape is an extremely variable character and the floral structures of these plants are so similar, I feel justified in considering them as representing a single species. 9. CELASTRUS MONOSPERMOIDES Loes. in Lorentz, Nova Guinea 8:280. 1910. (T.: Versteegb 1643, L!). Celastrus championii sensu King in Jour. Asiat. Bes Beng. 65?:639. (Mat. Fl. Malay Penin. No. 3). 1896, as to spec. cited, non Benth. Celastrus ‘aboensis Elmer, Leafl. Philipp. Bot. 7: 12579. 1915. (T.: Elmer 11411, A!). Celastrus malayensis Ridl. in Jour. Roy. Asiat. Soc., Str. Branch 75:18. 1917. (T.: Ridley 13536, SING Scandent shrubs up to 17 m. tall; branches terete, smooth to striate, glabrous, black or reddish-brown; lenticels lacking or obscure; axillary buds conoid, about 2 mm. long. Leaves ovate to elliptic, the apex acute to acuminate, the base shortly cuneate, obtuse to rounded, the margins serrate to nearly entire, 5-16 cm. long, 2.5-7.5 cm. wide, firmly membranous, lucid and usually blackish-brown on both sides or yellowish-green below in dry condition, the primary lateral veins 4-7 pairs, curved toward the apex, conspicuous and elevated below, plane above, the veinlets distinct and slightly elevated below, immersed and obsolete above; stipules laciniate, about 1 mm. long; petioles 6-13 mm. long. Inflorescences axillary, soli- 30 Woodson, R. E., Jr. Some dynamics of leaf variation in Asclepias tuberosa. Ann. Мо. Bot. Gard. 34:353—432. 1947. 1955] HOU—REVISION OF THE GENUS CELASTRUS 247 Fig. 6. Celastrus monospermoides Loes. tary or in clusters of three, aggregate dichasia, frequently fasciculate-racemiform, once to thrice compound, up to 10 cm. long, the peduncles glabrous, the primary peduncles obscure to 16 mm. long; flowers dioecious, white, the pedicels about 4 mm. long, the articulation at the basal part of the stalk. Маје flowers: calyx lobes imbricate, suborbicular, entire, about 0.7 mm. long; petals oblong, the apex rounded, about 2 mm. long and 1 mm. wide, entire, pink-punctate; disc fleshy, plane, the lobes obscure, truncate; stamens attached slightly under the margin of the disc, about 2.5 mm. long, the filaments filiform, glabrous, the anthers ovoid and obtuse; sterile pistil conoid, 2 mm. Jong. Female flowers: calyx lobes, petals, and disc as in the male, sterile stamens about 0.7 mm. long; pistil flask-shaped, 1.2 mm. long, the ovary globose, the style slender, distinct, the stigmata discoid. Fruits angular-ovoid, obovoid to subglobose, the valves ovate or obovate to broadly elliptic, about 12-17 mm. long and 7-11 mm. wide; seeds ovoid, about 7-12 mm. long and 5—10 mm. wide, blackish-brown, smooth. In dense woods, at altitudes 100—3,300 m.; Borneo, Indonesia, Malaya, New Guinea, and Philippines; flowering from January to September. Borneo: western Koetai, Endert 3846 (L). BRITISH NORTH BORNEO: Kini Taki River, J. Є М. S. Clemens 31827 (А, L), 32052 (L); Мага Parai, J. 8 M. S. Clemens 32860 (L), 33182 (L, UC); Penibukan, J. . S. Clemens 32120 (L), 50253 (1, UC); Upper Kinabalu, J. 9 М. S. Clemens 27870, 20019 (L), 50641 (A), 51205 (А, L, UC). JL. 42 248 ANNALS ОЕ THE MISSOURI BOTANICAL GARDEN SUMATRA: Саји & Alas Land, van Steenis 8717 (L), 90651 (А, L); old jungle near the Ack Kanopan, Loendoet — Koeale, Bartlett 6864 (F, MICH, NY, US); Laboe wipes Batoe, Kota Pinang, Tor 000 (A T e et ), А: SELANGOR: ме 5228 ASIN G). NEGRI SEM utu, Alims 1605, 2103 (SING). ENANG: Government Hill, Curtis 404 (SING), "Riley 404 (SING, para- type of спис malayensis) ; тайна, Ridley 13538 (SING, paratype of С. malayensis) ; 2 3 5690 (L, SING, UC, UPS), 6028 (L, US); Мате, fel Curtis 2005 (SING), Ridley s.n., 1890 (A); Тара, Wray 175 (L). SINGAPORE: Sungei Loyang, Mat s.n., 1894 New Сім UTCH NEW GUINEA: Bali, Stresemann s. n., May-Aug. 1911 (L): s. w. of Be eode des Idenburg (pues pes 12000 (A, L); Dalman, 45 km. inward of Nabire, Kanebira З Hatusima 12102 ( A m southern a lees via E = то oe. Mts., Verstee gb 1643 (1 type of Celastru ospermoides). NORTHEAST Ogeram a "Мамі e, J. З M. S. Closed. ait, bg: 6253 (A). PAPUA: ‚ боа. poma pluie e MINDANAO: Davao, Mt. Apo, Elmer MV (А, type of Celastrus apoensis, СН, L, MO, МҮ, US); Mt. McKinley, Е4а%о 000 (А, L, SING, UC); Bu і Du snalucpthan: Ramos & Едапо 38961 (А, L, US); Mt. “Katanglad, Sulit 3188, 3427, 0017, 10086 (Т.). The wide distribution, within geographically isolated areas, and the тогрћо- logical variations of this species are the causes of several names. Specimens col- lected in the Malay Peninsula from lowlands up to 1,330 m. altitude have leaf forms varying from ovate to elliptic. They have thick and large leaves and long in- florescences. A specimen (van Steenis 0651, A, L) collected at altitude 3,000 m. from northern Sumatra has small and thin leaves and short inflorescences—about 1 cm. long. The type specimen of Celastrus apoensis Elmer from Mt. Apo (alt. 2,000 m.), Mindanao, Philippines, has unusually small leaves. In addition, two other speci- mens (Sulit 9917 and 10086, L) collected at altitudes 1,700—1,800 m. from Mt. Katanglad, Mindanao, have larger leaves. The shapes and textures of leaves and inflorescences are similar to those from Malaya, Borneo, and New Guinea in similar environments. There are several collections from Mt. Kinabalu, North Borneo, at altitude about 3,000 m. The leaf form is ovate with rounded base which is very distinctive and might be thought as a form of this species. In addition, there is a specimen (Endert 3846, L) collected at altitude 1,200 m. from east-central Borneo having elliptic leaves, acute at both ends. I think it may be possible, in addition to those mentioned above, to find more specimens with various leaf forms from different environments. The leaves of the New Guinea specimens are usually elliptic, acute at both ends. They are quite similar to those from other localities in similar environments. I have made some studies and measurements of leaf characters. Except those from Mt. Kinabalu, Borneo, they can hardly be separated into groups. The shapes and sizes of fruits are variable also. As far as the shapes of fruits are concerned, those of the Malayan specimens seem to be broadly elliptic while those of the Philippine specimens are obovate. However, they display intermediate forms as well. 19551 HOU—REVISION OF THE GENUS CELASTRUS 249 з $. ~ 7 a t т м 4 & i * Я N чан e З шұ ~, | ў з e t б ; [o . ху 1 Р “А, 1 4 й “ | ДЕ ~ -, РУ 4 Ў 5 A ; i ° U “ ШЕ Let [А] dt c. marec пр | P^ У; ш e 10. C. HINDSII pn == 1 7 Y А 2 қ d joe v. N^ ~ АЛА | • s m . (/ * Р 7 Lr a >, -. V %/ i е Tr. еле $ rM / ur d н 19 р 5 E ds я! EN Мы Ж i ди] A еі” De "t паре A WU 5 к - м ub. 5» | г urs св 4 S ulla + | кій d" MES i € 2 ed \ ~ 4 r А => | 5 у“ e t MEWBRANIFOLIUS e 8. С. MONOSPERMUS, | $i po n gu p. Хәтер < | ) 2 ^ SEL. qd wd. ва en "Yu y dele д. => eut Гү . Р Lt DE M p fg. Pf 2. я И = D EAD = 4 ^ r 2 ж ж. ft й “ W А 0 а P № у “:-. "d $ 413 ~ г. e * 5 • Raye st IN “a А ке 4 N LA Lr x \ ma зи z E 3 алии ~e JE: tdg Y » ^ d ы , ё Ss a Lic Л. ; у: 4 ~~ / S UC ТЫ г men pls • La с = vede p 7 т Кес Жа ~ e rote , d За RE % M e — о =>) кої = е а >= : === 611. C. GLAUCOPHYLLUS e p cay te П Cae geist звання сло З #12. С. HOOKERI у Я AXILLARES. Map 2. Distribution of seven species of Celastrus, Subgenus CELASTRUS, Series II 10. CELASTRUS HINDSII Benth. in Hook. Jour. Bot. 3:334. 1851, ex char. (T.: Hinds s.n.). Catha ма и tes Benth: in Hook. Lond. Jour. Bot. 1:483. ‚ non сани а Wall. а d irri C emulosus Hassk. Ног Flueggea serrata Miq. Fl. ines E н 356. elastrus топозретта sensu Benth. Fl. Hongk. 63. Celastrus cantonensis Hance, in Jour. Bot. 23:325. 1885. (Tit aes 22101, Бептуі Loes. іп Engl. Bot. Jahrb. 29:444. 1901; 30:467. тас sensu . (Т.: Henry 3495, At). 1842, quoad descr. 1831, nom. nud. or. Descr. 1: 1915. 1858, ex char. 1859. (1: Junghubn 5. в., ІМ). 1861, поп R А, а photo [Vor. 42 250 ANNALS OF THE MISSOURI BOTANICAL GARDEN Celastrus oblongifolia Hay. Icon. Pl. Formos. 3:58. 1913. (T.: Hayata s. n., A!). Gees ves 5 marianensis Koidz. in Bot. Mag. Tokyo 30:400. 1915, ex char. (T.: Koidzumi dehet axillaris Ridley, in Jour. Roy. Asia . Soc. Malay. Branch, 1:56. 1923, ex char. Celastrus approximata Craib, in Kew Bull. Misc. TR 1926:349. 1926. (Т.: Kerr 9941, ! Celastrus пи Peas in Bull. Soc. Bot. France 95:180. 1948; in Uh Gén. l'Indo- Chine Suppl. 1:805, fig. 98, t. 1, 2 & 3. 1948. (T.: Poilane 28714, P Fig. 7. Celastrus hindsii Benth. Scandent shrubs up to 19 m. tall; branches terete, glabrous, brown, rarely lenticelled, the branchlets smooth and green, the lenticels usually lacking on the current season's growth; axillary buds deltoid, about 1 mm. long. Leaves elliptic- oblong, obovate-oblong, elliptic or rarely broadly elliptic, the apex abruptly acute to acute, the base cuneate to obtuse, the margins remotely and obscurely serrate to serrate, 4.0-14.5 cm. long, 1.5-6.0 cm. wide, coriaceous, shining, glabrous, the lateral veins 6—9 pairs, the veins and veinlets usually distinctly eleva and densely reticulated on both surfaces, rarely slightly elevated or obsolete above; ерин laciniate, about 1 mm. long; petioles about 0.5—1.0 cm. long. Inflorescences y axillary as well as terminal, sometimes axillary only, solitary, usually 3- to зіниці. ог sometimes racemiform and longer than the subtending leaf, the 1955] HOU—REVISION OF THE GENUS CELASTRUS 231 peduncles glabrous, the primary peduncles almost obsolete to 5 mm. long, rarely up to 25 mm. long; flowers dioecious, white to pale yellow, the pedicels almost obsolete to " mm. long, the articulation on the upper half of the stalk. Male flowers: calyx lobes semi-orbiculate, imbricate, obtuse, ciliate, about 1.5 mm. long; petals oblong to obovate, obtuse, glandular-ciliate, 2—5 mm. long and 1.3-2.5 mm. wide; disc cup-shaped, the lobes usually deltoid, rarely slightly oblong, shortly acute, rounded, or slightly dentate; stamens arising from the margin of the disc proper, about 2.3 mm. long, the filaments subulate, glabrous, the anthers ovoid, obtuse, cordate; sterile pistil ovoid, about 1.5 mm. long. Female flowers: calyx lobes, petals, and disc as in the male, but the petals smaller; sterile stamens about 1.5 mm. long, the ovary subglobose, the style columnar, the stigmata 3-lobed. Fruits ovoid to subglobose, the valves broadly ovate to suborbicular, about 8—11 mm. long and 7-9 mm. wide, usually 1-seeded; seeds cylindric, 5-8 mm. long and 5 mm. wide, reddish-pink, smooth. Chiefly in thickets, at altitudes from 232—1,800 m.; Bonin Islands, Burma, China, India, Indonesia, Indo-China, and Siam; flowering from January to October. d NortH Borneo: Penibukan, J. 9 M. S. Clemens 30493 (А, L, UC). ВМА: Kachin Hills, Shaik Mokim s. п. (CAL); Kaitao, Торрт 4292 (CAL); Kajen Lap., "Tobbin 6154 per » CHINA: FUKIEN: Amoy, Chung 335, ы 1771 AE eee 4599, 4783, 4851, 5601 (A); Vimus Chung 1174, 1 UC); Chuanchow, Chung 1084 (UC); Foochow and vicinity, Sin Ging Tang 7030 igh 7175 со. a Кыш а уісіпісу, Sin Ging Tang а (МО, UC). HAINAN: Bak Sa, Lau 26227 (A); Ching-mai Hsien, Lei 129 (L, NY, SING, UC, US), 440 (NY, SING, UC, US); Hung-mo Shan and vicinity, Lai area, Tsang 8 Fung 241 (A, US); Модоа, McClure s. n. ЖА O); Yaichow, How 70 368 (A, F, NY); without precise locality, Levine s.n. As MO, US), 484 (А, Е, GH, MO). норе: Chang-lo Hsien, Wilson 561^ (A); n Hsien, Ho-Ch’ang Chow 1944 (A, NY); Ichang, T 7 (А, US); Patung, Ho-Ch’ang Chow 548 (A), 705 (A, NY), Wilson 450 (A, NY); wout precise locality, Henry 3241 (NY, type of Celastrus hindsii var. henryi), 3495 (A, GH, US), 3856, 3495^ (са), v. Rosthorn 9 (А). KWANGTUNG: Au-tsai, McClure 3544 (NY); Canton and vicinity, Hance 22191 (A, photo z ge of Celastrus cantonensis, with a тек of leaf), Levine s. n. (C.C.C. пов. 180--А, UC; 441--А, Е, GH, ‚ US; 1236—A, GH, MO, US), Merrill 10070 (UC), Keng- d To 1868 (UC); near Fung- wang, Kang-ping To et al 889 (US); Kochow, Tsiang 1002 (UC), 2300 (NY); Mei Hsien, Tsang 21422 (A); Pok-Lor, Hom Fung A-511 (NY); Sin-fung Hsien, Taam 587, 759 (A); Sun-wui Hsien, Tso & Tsiang 2024 (A, F); р Wang 3196 (NY); Tseng- 2 shing, Tsang 20. NY); g-hwa Hsien, Tsang (MO, NY), 25019 (A); Wung-yuen, Lew 693 (A, NY, SING), 2672 (A); Ying-tak Hsien, Tsui 391 (A, MO, NY, UC, US), Tsiang 1919 (SING, UC), Tsang & Kam-chow Wong 2661 (UC); Wung- yuen, 74 693 (A, S Ў дылы ном: without precise — Cavalerie 3976 (A). SZECHUAN: Kiating Fu, Wis 3 4 (A); Ота Shan, Chiao & Fan 458, 8639, 11712, 14081, 15221 (А); Ни 7380 (A). s Lee 4351 (A), Wilson ge ae US), Wang (A). AIWAN: ru, Heito, DH 11162 (ТАГ); Arisan, Faurie 1376 (A, TAI), Hayata s. n. A, photo of type of Celastrus oblongifolius). YUNNAN: ен Hsien, Wang 75846, pu, 79470 (А); Chen-pien Hsien, Ko 56089 (A); Jenn-yeh Hsien, Wang 80136, 80181, 2 (A); Lu-se, Tsai 56342 (A); Luh-shuen Hsien, Wang 81177 (A); ii ы Feng 1 13068, 13210, 13 387, 13454, 13464, 1 вый (А); Mengtze, Henry 1055 (А, МО, NY, US), Ји 2071 (А); geom : "eus peni 11085, ag (A); Зашто, Неп M 370 P SE re n. 2147 (US), s ae (A, US); “Tsang 77 (UC); ubi I (СН), біт (GH, NY). (Мог. 42 252 ANNALS OF THE MISSOURI BOTANICAL GARDEN ASSAM: Darrang, Upendranath Kanjilal 4976 (CAL); Kegwima, Naga Hills, P. жопе (CAL); Khasia, Tserra, Clarke 5015 (CAL). ѕккім: Duphla Hills, Lister 68 (CAL). Імро-Сніма: Cho Gank, Pételot 863 (A, UC), 991 (UC); Tonkin, Pételot 340 pii dence ТАУ ae Arsin 19568 (L); Tjibodas, Hallier 462 (A), Valeton 102 (A), Koorders Е (L), Valeton s. п. (L, UC); Tjidadap, Tjibeber, Winckel 928, а (1), 16338 s UC), 16828 (L, SING); Mt. Waripin, Premyer, Forbes 961 (CAL), ерінін 27, 10р (1), s. п. (type of PME serrata Miq., L), Kortbals 747 (A). MOLUCCAS: Am a, de Vrise 8 Teijsm ). SUMATRA: Kabajakan to Tretel, Walter & "ві %о (А, МҮ); Waldregion, Janghabn s. n. (L). ейі BONIN ISLANDs: Hah-jima, Wilson 8300 (A). : Kao Lem, Kerr 9941 (8059, type of Celastrus approximata) ; Mt. Doi Chang, Ше Косе. 1744 (А, ОС, 05). Bentham in 1842 determined a Hongkong plant, collected by Hinds, as Catha monosperma (Roxb.) Benth., based on Celastrus monosperma Roxb. Subsequently in 1851, he described the same specimen as representing a distinct species, Celastrus hindsii Benth. He commented in the discussion: "The form, size, and structure of the capsule are like those of C. [Celastrus] paniculata, without any narrowing of the base, as in С. monosperma and C. Championi, although it is always mono- spermous (by abortion) as in the two latter, not three-seeded as it usually is in C. paniculata. All these species belong to the true Celastri, not to Catha, to which I had erroneously referred them in my former paper”. So he reduced Catha mono- sperma (Roxb.) Benth. (excl. syn. Roxb.) into synonymy under Celastrus hindsii Benth.; but in 1861 he again adopted Celastrus monospermus Roxb. for the Hong- kong plant and reduced his own Celastrus hindsii to synonymy. Actually, Celastrus bindsii Benth. is readily distinguished from Celastrus monospermus Roxb. in both vegetative and reproductive structures, as Bentham has pointed out in the discus- sion under Celastrus bindsii, and the two species should be maintained as distinct. e ly membranous leaves usually are elliptic-oblong, shining, and densely reticulate; and the capsules usually one-seeded, which make this species very easily distinguished from the others. I have seen a photograph of the type specimen and fragment of a leaf from the type of Celastrus cantonensis Hance. This plant has leaves, venation, and inflorescences which match the description of Celastrus hindsii. The original description indicates that Celastrus bindsii var. henryi Loes. differs only from the typical variety by the larger and thicker leaves, and also the nearly immersed obsolete veins. Later (1916) Rehder and Wilson stated: "in addition to the characters given by Loesener the variety differs from the type species in having subsessile and very shortly peduncled cymes". In all the speci- mens seen, I have not been able to distinguish with certainty var. henryi from the typical variety. Also, this variety does not differ at all from the typical form of the species in geographical range. Thus, this variety is reduced to synonymy here. A specimen collected from Yunnan by Wang (no. 76880) has leaves like Celastrus hindsii whereas the fruits are similar to C. monospermus. This may be a hybrid between the two, if indeed they are distinct. 1955] HOU—REVISION OF THE GENUS CELASTRUS 253 Fig. 8. Celastrus glaucophyllus Rehd. & Wils. 11. CELASTRUS GLAUCOPHYLLUS Rehd. & Wils. in Sarg. Pl. Wils. 2:347. 1915. (Т.: Wilson 952, At). Celastrus rugosus Rehd. & Wils. loc. cit. 2:349. 1915. (Т.: Wilson 1106, А!). Celastrus orbiculatus sensu Tardieu, in Fl. Gén. l'Indo-Chine, Suppl. 1:806. 1948, поп unb., as to spec. cited. Scandent shrubs up to 2-5 m. tall; branches terete, glabrous, castaneous, the lenticels scattered, elliptic; branchlets glabrous, yellowish-brown, the lenticels lacking or obscure, rarely distinct; axillary buds ovoid, obtuse, about 3 mm. long. Leaves elliptic, obovate or ovate, the apex acute to shortly acuminate, the base rotund to broadly cuneate, the margins crenate-serrate, 4—13 cm. long, 2.5-8.0 cm. wide, firmly membranous, glabrous, or pubescent on the veins below, rarely bullate, usually glaucous or pale beneath, the primary lateral veins 5—7 pairs, ele- vated below, immersed to slightly elevated above, the veinlets distinct, slightly ele- vated below, immersed and obscure above; stipules laciniate, filiform, about 1.5 mm. long; petioles 5-15 mm. long. Inflorescences axillary as well as ахі і 3- to 7-flowered, the terminal ones racemiform, usually 2-6 [Vor. 42 254 ANNALS OF THE MISSOURI BOTANICAL GARDEN cm. long, the peduncles suppressed or up to 2 mm. long, glabrous, each associated with an axillary bud except the uppermost one of the terminal inflorescences; flowers dioecious, green, the pedicels about 2 mm. long, the articulation at the middle or upper half of the stalk. Male flowers: calyx lobes valvate, ovate, rotund, glandular-ciliate, about 1.5 mm. long; petals oblong, rotund, slightly erose, 3—4 mm. long and 1.0-1.5 mm. wide; disc membranous, the lobes ovate, obtuse; stamens 2.5—4.5 mm. long, the filaments filiform, the anthers ellipsoid; sterile pistil 1 mm. long. Female flowers: calyx lobes, petals, and disc as in the male; pistil flask- shaped, 5 mm. long, the ovary globose, the style distinct, slender, the stigmata 3- lobed, the lobes patent. Fruits subglobose, the pedicels accrescent, about 3—5 mm. long, the valves suborbicular, 7-10 mm. long and 8—9 mm. wide, 3- to 6-seeded. Seeds ellipsoid to slightly plano-convex, about 3—4 mm. long and 2 mm. wide, black, the areolae distinct. Chiefly in mixed thickets, at altitudes from 700 to 3,300 m.; China and Indo- China; flowering from March to June. | ANHWEI: Chu-hwa Shan, Ching 2700 (UC), Sun 1189 (NY). норе: sien, Wilson 357 (А, MO, US), 072 (A). ( i Shan, C 6448, 6468, 7597, 8028 (A), Fang 3089 (A, NY), 12883, 13051 (A, US), 16624, 17317 (A), Hu 7380 (А), Гее 3561 (А), Та 204, 1071 (A), Wilson 4782 (А); O-pien Hsien, 52^ (A), 1106 y Celastrus rugosus; ); Wen-chuan, Wang 21124 (A); Yen-yuan Hsien, Schneider 4112 (A); Wa-shan, Wilson 9 (A); without precise locality, Fang 1202 » Schneider 646, 64 YUNNAN: Che-tse-lo, Tsai 58555 Henry 9679 (NY, US); Pe-yen-tsin, Ten 40 (A), 550 (A, US); Ping-pien Hsien, Tsai 61004 (A); south of Red River, Manmei, Henry 96794 (A, MO, NY); Ta-pin-tze, Ten 358 (A, US); Tengchuan, Schneider 2874 (A, GH, US); Likiang, Ching 20489, 20638, 21618 (A), Feng 3047 (A), Rock 3556 (A, UC, US), 3031 (A, US), 4032 (A, UC, US), 8312 (A), 8540 (A, NY, UC, US), Schneider 1984, 2839 (A, GH); Wei-se Hsien, Tsai 57899, 59824 (A). Іхоо-Сніма: Chapa, Tonkin, Pételot 5941 (A, NY). 12. CELASTRUS HOOKERI Prain, in Jour. Asiat. Soc. Bengal 73:179. 1904; Novic. Ind. 418. 1905. (T.: Hooker s. n., CAL!). Celastrus stylosa sensu Lawson, in Hook. Pl. Brit. Ind. 1:618. 1875 (pro parte, non Wall.). Celastrus stylosa Wall. var. Loeseneri (Rehd. & Wils.) Tardieu, in Fl. Gén. l'Indo-Chine, Suppl. 1:806. 1948, sensu Tardieu, non Rehd. & Wils. Scandent shrubs up to 10 m. tall; branches terete to slightly striate, glabrous, brown to reddish-brown, the lenticels sparse or lacking on the current year's growth, suborbiculate to elliptic; axillary buds conoid, 2-4 mm. long, the bud 19551 HOU—REVISION OF THE GENUS CELASTRUS 255 scales accrescent, persistent. Leaves broadly elliptic to ovate, the apex acute, the base obtuse, the margins serrate, 6-12 cm. long, 4-7 cm. wide, thin-membranous on flowering branches, glabrous, the primary lateral veins 5—6 pairs, slightly ele- vated below, immersed and distinct above, the veinlets visible below, obscure above, relatively thick-membranous on fruiting branches; stipules laciniate, about 1.5 mm. long; petioles 8—15 mm. long. Inflorescences axillary as well as terminal, the axillary ones 3- to 5-flowered, the primary peduncles glabrous, about 1.5—3.5 mm. long, the terminal inflorescences very short-pedunculate dichasia, fasciculate or solitary, about 4 cm. long, the axils of the peduncles associated with vegetative buds (except the uppermost) ; flowers dioecious, green, the pedicels 1.5—2.4 mm. long, the artic- ulation on the upper half of the stalk. Маје flowers: calyx lobes deltoid, imbricate, ciliate, about 1.5 mm. long; petals oblong to obovate, slightly glandular on the margins, about 3.5 mm. long and 1.5 mm. wide; disc cup-shaped, the lobes sub- reniform; stamens arising from the margin of the disc proper, about 2 mm. long, the filaments linear, glabrous, the anthers apiculate; sterile pistil 1.5 mm. long. Female flowers: calyx lobes, petals, and disc as in the male, the sterile stamens 1.5 mm. long, the pistil 4 mm. long, the ovary globose, narrowed into a distinct style, the stigmata 3-lobed, each lobe bifid, linear. Fruits subglobose, the valves broadly elliptic, about 10 mm. long and 7 mm. wide, 3- to 6-seeded; seeds ellipsoid or ovoid, about 4 mm. long and 2 mm. wide, black, the areolae distinct. Chiefly in thickets, at altitudes from 1,525 to 3,050 m.; Burma, India, China, and porte flowering from March to May. a: Adung Valley, Kingdon Ward 9455 (A); d Lapyeka to Sinlum Kaba, iem 2 (САТ); without precise locality, Prazer s. n., 1890 (C AL), Lace 5753 сира HINA: YUNNAN: between Chienchuan Plain and cs Mekong drainage basin Rock 8615 (A, UC, US); Wei-se Filek, Tsai 59901 (A); without к ). chining, locality, Yu 8491 (A NDI a, Griffith 605 (CAL). BENGAL: Darjeeling, Clarke 682", Yr CAL); Бони i tera 5. п. (CAL); note Be Dr. Prain’s collector 289 (CAL); rupe сн 46569 (CAL). WITHOUT PRECISE LOCALITY: temp. region, 8,000—10,000 - r s. n. (type of Celastrus bookeri, CAL), Kurz s. n. (CAL). Ain Griffith 1993 ( , GH). Rehder and Wilson 31 have cited two Indian specimens, Hooker & Thomson, “Khasia, alt. 4-5,000 ped." and С. С. Rogers, “Sikkim, January 1900”, as Celastrus bookeri Prain. Actually, they are Celastrus stylosus Wall. As it was stated by Prain, "The two can be at once distinguished by the stamens which have hirsute filaments in Celastrus stylosus, but glabrous ones in Hooker’s species”. In addition, the intercalary inflorescences and slightly lunate seeds of Celastrus stylosus are easily distinguished from terminal inflorescences and ovoid seeds of C. hookeri. 13. CELASTRUS MEMBRANIFOLIUS Prain, in Jour. Asiat. Soc. Bengal 73:197. 1904; Novic. Ind. 418. 1905. (T.: Mann s. n., CAL!). Branches terete, the branchlets slightly angular on the dry specimens, glabrous, 31 Кећдег, A., and Wilson, E. H., in Sarg. Pl. Wils. 2:352. 1915. [Vor. 42 256 ANNALS OF THE MISSOURI BOTANICAL GARDEN the lenticels scattered, ovate or orbiculate; axillary buds conoid, acute, about 1.5 mm. long. Leaves ovate-lanceolate to elliptic, the apex acuminate, the base cuneate, the margins serrate, 7.5—14.0 cm. long, 3.5—5.5 cm. wide, membranous, pale green in dry condition, glabrous, the primary veins 4—5 pairs, elevated below, immersed or slightly elevated above, the veinlets prominent on both surfaces; petioles 1.0—1.5 cm. long. Flowers unseen. Fruiting inflorescences axillary as well as terminal, short-pedunculate or fasciculate, associated with axillary buds, the stalks 7-15 mm. long, the articulation on the upper half of the stalk. Fruits subglobose, the valves broadly elliptic, about 7 mm. long and 5 mm. wide, 3- to 6-seeded; seeds elliptic, rotund at both ends, wrinkled (immature?), black, about 2-3 mm. long and 1-2 mm. wide. Innra: А55АМ: Khasia and Junteah Hills, Mann s.n., 1877 (CAL, type; MO); Khasia Hills and Brahmaputra plains, Kurz s. п. (CAL). The few specimens available are fruiting specimens in which the seeds are highly wrinkled, indicating that they might be immature. However, I have seen no immature seeds with such wrinkling in any other species of Celastrus. 14. CrLasTRUs vANIOTI (Lévl.) Rehd. in Jour. Arn. Arb. 14:249. 1933. Saurauja vaniota Lévl. Fl. Kouy-Tchéou, 415. 1915. (T.: Bodinier s. n., А!). Celastrus spiciformis Rehd. & Wils. in Sarg. Pl. Wils. 2:348. 1915. (T.: Wilson 2312, ! Celastrus spiciformis var. laevis loc. cit. 2:349. 1915. (T.: Wilson 1176, А!). Scandent shrubs up to 10 m. tall; branches terete ог slightly striate, glabrous, light brown to reddish or fuscous, the lenticels scattered, orbicular or elliptic, some- times lacking on the current year's growth, slightly elevated on the older branches; axillary buds globose, about 2 mm. long. Leaves ovate or elliptic, the apex acute to shortly acuminate, the base rotund to shortly cuneate, the margins crenate- serrate, the teeth incurved, glandular-mucronate, 5-13 cm. long, 3.5—7.5 cm. wide, membranous, glabrous or slightly puberulous on the veins below, the primary lateral veins 6—7 pairs, slightly elevated on both surfaces, the secondary lateral veins parallel; stipules laciniate, filiform, about 1.5 mm. long; petioles 1-2 cm. long. orescences axillary as well as terminal, shortly pedunculate or fasciculate, the peduncles almost obsolete to 5 mm. long, the axillary inflorescences dichasia or very short-racemiform, sometimes lacking in the axils on the current season's growth, the terminal inflorescences fasciculate-racemiform, 6-18 cm. long, associated with axillary buds; flowers dioecious, greenish-white, the pedicels about 2 mm. long, the articulation at the middle or on the lower half of the stalk. Male flowers: calyx lobes deltoid, imbricate, obtuse, glandular-ciliate, about 1 mm. long; petals oblong, rotund, slightly erose, about 3 mm. long and 1.5 mm. wide; disc cup-shaped, membranous, the lobes acute; stamens arising from the margin of the disc proper, about 2.5 mm. long, the filaments linear, glabrous, the anthers ovoid, obtuse, slightly apiculate; sterile pistil conic, about 1 mm. long. Female flowers: calyx 1955] HOU—REVISION OF THE GENUS CELASTRUS 257 lobes, petals, and disc as in the male; sterile stamens 1 mm. long; pistil 3 mm. long, the ovary ellipsoid, gradually attenuate into the style, the stigmata 3-lobed and spreading. Fruits subglobose, the valves broadly ovate, about 9 mm. long and 8 mm. wide, 3- to 6-seeded; seeds ellipsoid, obtuse at both ends, areolae distinct, about 4 mm. long and 2 mm. wide, black. In mixed woods, at altitudes from 500 to 2,000 m.; China; flowering from e to July. N: Wukang, Te-bui = 113 (А). норк:: Fang Hsien, Wilson 2215 а NY) Heng iet Hsien, Wilson 2312 (A, type of Celastrus spiciformis; GH, MO); without туе осу, Непту 5035 А. СН, US), Wilson 2215" (NY). KWEICHOW: Bin Lon u Shan, Luchen, poe 6057 (NY); Tuhshan, Tsiang 6655, 6003 Qut Ан и locality, Bodinier 2287 (A, type of — vanioli). SZECHUAN: Loo shan Hsien, Wang 23582 (A); Ma-pien Hsien, Wang 23101 (A); Omei Shan, Chow 7004 5 у Wilson 1176 (А, holotype of Celastrus spiciformis var. Aou I 383 (А, 5925 YUNNAN: (A, NY): Та Няеп, Feng 11165 (A); Yi-liang Нее, Tsai 52000 "uk Celastrus vanioti, as pointed out by Rehder and Wilson, "is a puzzling plant somewhat intermediate in character between С. angulata Maximowicz and C. bypoleuca Warburg". It appears to be a hybrid between those two species. Super- ficially, the type of C. spiciformis is similar to C. angulatus, except for the in- florescences which are both terminal and axillary. Оп examination, the floral structures were found to be similar to those of Celastrus hypoleucus. 15. CELASTRUS HYPoLEUCUS (Oliv.) Warb. apud Loes. in Engl. Bot. Jahrb. 29:445. 1900 (as bypoleuca). Erythrospermum hypoleucum Oliv. in Hook. Icon. Ш, 9:/. 1899. 1889, ex char. & ill. (T.: Henry 5877). Celastrus hypoglaucus Hemsl. in Ann. Bot. 9:150. 1895. (T.: Henry 2837, GH!). Celastrus bypoleucus forma a genuina pnm in Diels, in Engl. Bot. Jahrb. 29:445. 1900 (based on Erythrospermum bypoleucu iv. Celastrus bypoleucus forma B argutior Lar loc. cit. 1900. (T.: Henry 6771, A). Scandent shrubs 3—5 m. tall; branches slightly striate, glabrous, darkish brown, the lenticels sparse to dense, rarely lacking on the current year's growth; axillary buds orbicular, about 2 mm. in diameter. Leaves elliptic to elliptic-oblong, the apex acute, the base obtuse, the margins remotely serrate to serrulate, 3.0—9.5 cm long, 2.0-5.5 cm. wide, usually delicately membranaceous, glabrous, sometimes puberulous on the veins below, glaucous below, pallid above, the primary lateral veins 5—6 pairs, slightly elevated below, immersed above, the veinlets prominent below, obscure above; stipules filiform, about 2 mm. long; petioles 1-2 cm. long. Inflorescences axillary as well as terminal, the terminal portion usually up to 10 cm. long, usually short-pedunculate, the peduncles glabrous, accrescent, the rimary peduncles usually 2—8 mm. long; flowers dioecious, light green or yellow, the pedicels accrescent, about 2-8 mm. long, the articulation at the upper half of the stalk, after anthesis usually only the flowers of the terminal inflorescences maturing into fruits, and the stalk accrescent up to 4 cm. long, spreading and [Vor. 42 258 ANNALS OF THE MISSOURI BOTANICAL GARDEN pendulous. Male flowers: calyx lobes valvate, deltoid to oblong, subentire to slightly erose, about 1.5 mm. long; petals oblong or rarely elliptic, usually entire, rarely obscurely erose, about 4.3 mm. long and 2 mm. wide; disc cup-shaped, the lobes obscure, truncate to mucronate; stamens arising from the margin of the disc proper, about 4 mm. long, the filaments filiform, glabrous, about 3 mm. long, the anthers ovoid, slightly apiculate, about 1.5 mm. long; sterile pistil about 2 mm. long. Female flowers: calyx lobes, petals, and disc as in the male; sterile stamens about 1.5 mm. long; pistil flask-shaped, about 3 mm. long, the ovary ellipsoid, the style cylindric, distinct, the stigmata 3-lobed, flat, spreading. Fruits subglobose, the valves broadly elliptic or suborbicular, about 8-12 mm. long and 8—10 mm. wide, irregularly pink-punctate within, 3- to 6-seeded; seeds more or less plano- convex to slightly lunate, attenuate at both ends, about 6 mm. long and 2 mm. wide, black-brown, the areolae distinct. Chiefly in thickets, at altitudes from 1,700 to 2,745 m.; China; flowering from June to July. CHINA: ANHWEI: Wang Shan, Ching 3065 (UC). nuper: Fang Hsien, Wilson 362 (A, Е, СН, MO, 05); Hsun-tien-tsze, Chun 4427 (A); Ichang, Wilson 1063 (A, NY); without precise locality, Henry 2837 (GH, paratype of Celastrus hypoglaucus), 5887 (GH, NY, US), 6811 (GH), 6771 (А, type of Celastrus hypoleucus forma B ch еді: ). SHENSI: Kin-qua Shan, Giraldi s. n., July 10, 1897 (А); Qua-in Shan, Giraldi s. July 16, 1897 (A); Tai-pei Shan, Giraldi s.n., Ju Му 15-20, 1897 (А); Tsinling, met 027 (A). SZECHUAN: Kuan Hsien, Wilson 2306 (A, US). This species is easily recognized by its a2 terminal pendulous racemiform dichasia of long accrescent stalked fruits and by the slightly lunate seeds. The leaves are usually glaucous below, but this character is extremely variable. 16. CELASTRUS GEMMATUs Loes. in Engl. Bot. Jahrb. 30:468. 1901 (as gemmata). (T.: Henry 98724, МО!). Embelia esquirolii eis in Fedde, Rep. Sp. Nov. 10:374. 1912. (T.: Esquirol t А!). Celastrus lokcbongensis Masamune, in Trans. Nat. Hist. Soc. Formosa 25:15. 1935. (Т.: Tsiang 1346, А ў Scandent shrubs, 3—7 m. tall; branches terete ог slightly striate, light to darkish brown, the lenticels scattered, suborbicular or oval, usually elevated, white; axillary buds conoid, acuminate, about 4-11 mm. long. Leaves usually broadly elliptic to elliptic-ovate, the apex gradually acuminate to acute, the base obtuse, rotund or truncate, the margins serrate, 5—15 cm. long, 2-8 cm. wide, firmly membranous, usually glabrous, rarely yellowish-puberulous on the veins below, the primary lateral veins 5—7 pairs, curved toward the apex, the veinlets densely retic- ulate, the veins and veinlets elevated on both surfaces; stipules filiform, tufted, about 1.5 mm. long; petioles 1.0—2.5 cm. long. Inflorescences axillary, usually 3- to 7-flowered, the peduncles glabrous, the primary peduncles 4—7 mm. long; flowers dioecious, white or yellowish-green, the pedicels 3-8 mm. long, the articu- lation at the lower half of the stalk. Маје flowers: calyx lobes valvate, ovate- deltoid, obtuse, glandular-ciliate, about 1 mm. long; petals subequal, obtuse, 1955) HOU—REVISION OF THE GENUS CELASTRUS 259 Аз у Fig. 9. Celastrus gemmatus Loes. slightly erose, about 3.5—4.5 mm. long and 1.5-2.0 mm. wide; stamens inserted between the disc lobes, about 3 mm. long, the filaments more or less complanate, glabrous, gradually tapered toward the apex, the anthers oblong-ellipsoid, obtuse, usually apiculate, pink-punctate; disc thin, cup-shaped, the lobes acute; sterile pistil columnar, 1.5-2.5 mm. long. Female flowers: calyx lobes and petals as in the male; sterile stamens about 1 mm. long; disc thin, cup-shaped, the lobes in- conspicuous; pistil flask-shaped, about 4 mm. long, the ovary ovoid, the style columnar, the stigmata 3-lobed, each bifid, reflexed. Fruits usually 1—3 on each dichasium, the pedicels green, the fruits subglobose, the valves broadly oval or suborbicular, 7-15 mm. long and 7-15 mm. wide, 3- to 6-seeded; seeds ellipsoid or ovoid, about 4.5 mm. long and 2.5 mm. wide, reddish-brown, smooth. on in thickets, valleys or open slopes, at altitudes from 400 to 3,000 m.; widely distributed in China: south of the Yangtze River; flowering from April to October. ANHWEI: Chu-hwa Shan, Ching 2638 (А, UC), € ASIE Wang Shan, rad 15: [490 t (UC) ; Wu-yuen, Ling 1349 "cdi CHEKIA Tien-mu, Ни 5 м: Dingschou ‚ Sone schan ang 399 “U пуан Fang Hsien [Vor. 42 260 ANNALS OF THE MISSOURI BOTANICAL GARDEN СН, MO, US); Patung Hsien, Ho-Ch’ang Chow gor (A, NY); Wan-tsao Shan, Chun 3884 (A), 4229 (US); without precise locality, Henry 7614 (GH). кілхсві: Hsin-feng Hsien, Ни 1119 (A); Lu-shan, Chung & Sun 502 (А EO Steward 1234 (A), 5470 (A, UC, US), Wilson Тыр (А, МО, US). KIANGsI ET FUKIEN: in monte Dunghwa-schan inter Schitscheng ес Ninghwa, Wang 200 (A). KWANGSI: ићи. Shan, Thwang, T'siang 10047 (NY); Tzu-yuen Hsien, Chung 83605 (A); Huangtung, Yao-shan, Sim 11835 : Jen-hwa A Tsang 208 55 (А, L, MO, NY, SING, UC, US), Tsiang 1346 (А, isotype of C. lokcbong- ensis, UC), 1410 (A, UC); Tsungfa, Tsang 25189 (A); Yang-shan Hsien, Tsui 763 MO, МУ). kwricHOw: Nan-kan, Cheng-feng, Tsiang 4661 (А, NY); Pinfa, Kweiting, Tsiang 5360 (A, NY); Расы "PH 002 5 (A); Tungtze, Chun 4013 (A, 8, UC), Tsiang 5013 (NY); Yao-ren Shan, Sanhoa, Tsiang 6450 ( ); without precise locality, Esquirol 4 (A, type of one’s P. SIKANG: нае (Tachien-lu), Smith 12808 (S), Wilson се. Le MO, US). UAN: Kuan Hsien, Fang 2124 (A, ); Mo-tien-ling, Wang 22451 (А); о боже Wilson 2305“ (А); Mow Hsien, Fang 5529 (A); Nan-c а. “sien, Fang 503 (A); Omei Shan, Chow í Fang & Lee 3575 (A, US), Yu-shih Liu 1147, 1423 (А), Tai 1113, 1120 (А); O-pien "pepe a 781 (A); Pao-Hsing Hsien, Chu 3761 (A); Wen-chuan Hsien, Wilson 2305 (A, US); Yun-ching Hsien, Sun 12 А]. YUNNAN: Chien-chuan-Mekong Divide, coe Y» 22311 (A, US); Chungtien, Feng 2868 (A), Rock 24670 (A, MO, NY, UC, US); in the mountain above Dashao, Handel-Mazzetti 13074 (A); between Kambaiti and Tengyueh, via Kuyung, Rock 7568 (A, UC, US); Kunming, Wang 62000 (A); ( gemmatus; MO, NY, US), 11471 (A, paratype of C. gemmatus; MO, NY); Muli Wachin, Yu 14375 (A); Ping-pien Hsien, Tsai 61004 (A); south of Red River from Manmei, Henry 0670% (A); between Shweli and Yengyueh valleys, Forrest 8704 (А, $), 11173, 12050 (A); Suen-oui, Maire 14 (A); Tali, Rock 6809 (A, US); Tai-pon, Maire 102 (A); Wei-si Hsien, Tsai 50523, 50799, 63116 (A), Wang 63709 (A); without precise locality, Ducloux 132 (NY, UC), Forrest 16241, 14510 (A), Yu 5562, 5771 (А). Celastrus gemmatus often has been confused with C. orbiculatus Thunb., since the two are closely related. After a careful examination of specimens of the two species, it was found that although their floral morphology is similar, C. gemmatus has larger conic axillary buds, firmly membranaceous and densely reticulated leaves, and, in addition, a distinct geographical distribution which separates it from Celastrus orbiculatus and other related species. Celastrus gemmatus is chiefly distributed in southwestern China and is rarely found in southeastern China; some plants are recorded as far north as Hupei which may represent its northern limit. I have seen the types of Embelia esquirolii Lévl. and С. lokchongensis Masamune, They both bear large conical axillary buds and densely reticulated leaves, and are within the geographical range of Celastrus gemmatus. Непсе, I consider these species as synonyms of Celastrus gemmatus. 17. CELASTRUS ORBICULATUS Thunb. Fl. Јар. pp. xlii, 97 (errore "'articulatus"). 1784; Gmel. Syst. Veg. 406. 1796; Lam. Tab. Encycl. et Meth. Bot. 2:94. 1797; Airy Shaw, in Curtis's Bot. Mag. 158:/ab. 0304. 1935. (Т.: collector unknown s. n., A, photo!). Celastrus асве Thunb. ЕІ. e 97. Шы. (sub. Celastrus og quoad nomen ч Tu in Fl. Jap. p. xlii); Maxim. in Bull. Acad. Sci. St.-Petersb., ПІ, 27:456. z^ ol. Acad. St.-Peténb. 11:200. бі нн азл" Rupr. in Bull. Acad. Sci. St.-Petersb., IL, 40:357. 1857, ex char. 1955) HOU—REVISION OF THE GENUS CELASTRUS 261 Celastrus punctatus sensu Regel, in Gartenfl. 9:407, tab. 312, f. 6. 1860, non Thunb. Celastrus articulatus Thunb. var. pubescens Makino, in Bot. Mag. Tokyo 7:102. 1893, ex har. : Makino s. n Celastrus orbiculata Thunb. forma В. microphylla Loes. in Engl. Bot. Jahrb. 30:469. 1901, ex char. (T.: Henry 3827). Celastrus orbiculata Thunb. forma y. maior Loes. loc. cit. 30:469. 1901, ex char. (Т.: Giraldi 237). Celastrus articulata 'Thunb. var. cuneata Rehd. & Wils. in Sarg. Pl. Wils. 2:350. 1915. T.: Wilson 2308, A!). Celastrus lancifolia Nakai, in Bot. Mag. Tokyo 37:3. 1923, ex char. (T.: Nakai s. n.). Celastrus insularis Koidz. in Bot. Mag. Tokyo 39:22. 1925, ех char. (T.: Koidzumi s. n.). Celastrus strigillosus Nakai, loc. cit, 40:492. 1926, ex char. (T.: Nakai s. n.). Celastrus articulatus Thunb. var. stephanotiifolius Makino, in Jour. Јар. Bot. 3:24. 1926, ex char. (T.: Макто s. n.). Celastrus stephanotiifolius (Makino) Makino, loc. cit. 3:46. 1926. Celastrus orbiculata var. aureo-arillata Honda, in Bot. Mag. Tokyo 45:422. 1931, ex char. (T.: D. Shimisu 50). Celastrus jebolensis Nakai, apud Nakai & Kitagawa, in Rept. Ist. Scientif. Exped. Man- choukuo, Sect. IV, 1:6, pl. Г. 1934, ex char. (T.: Nakai et al. s. n.). Celastrus articulatus var. papillosus Nakai ex Hara, in Jour. Jap. Bot. 10:84. 1934, ex char. .: H. Hara s. n.). Celastrus articulata var. orbiculata (Thunb.) Wang, in Chin. Jour. Bot. 1:62. 1936. Celastrus orbiculatus var. papillosus (Nakai) Ohwi, Fl. Jap. 736. 1953. Celastrus versicolor Nakai, in Bull. Sci. Mus. Tokyo 33:16. 1955, ex char. (Т.: Nakai 8 Maruyama s. n.). Celastrus orbiculatus var. pilosus Nakai, loc. cit. 33:16. 1953, ex char. (T.: Hozawa- Sigeo s. n.). Celastrus orbiculatus £. papillosus Nakai ex Hara, Enum. Spermat. Japon. 80. 1954. Scandent shrubs up to 10 or 12 m. tall; branches terete to slightly striate, glabrous, light to darkish brown, the lenticels inconspicuous or sparse; axillary buds small, depressed, ovoid to subglobose, 1-3 mm. long, sometimes the outermost scales becoming deltoid, sharp-spinose, 1-2 mm. long. Leaves extremely variable in size and shape, usually obovate to suborbicular, ovate, or oval-oblong, the apex rounded, shortly cuspidate to acute, the base cuneate to obtuse, the margins crenate- serrate, 2-12 cm. long, 1.5—8.0 cm. wide, delicately membranaceous, glabrous above, rarely puberulent below, the primary lateral veins 3—6 pairs, distinctly elevated below, slightly elevated above, the veinlets slightly reticulated, prominent and slightly elevated below, immersed and obscure above; stipules filiform, tufted, about 1-2 mm. long; petioles 1-3 cm. long. Inflorescences axillary, rarely also terminal in the male plant, usually 3- to 7-flowered, occasionally solitary, often rather congested along the shoots of the current year, the subtending leaf some- times not developed, distinctly pedunculate, the peduncles subequal, glabrous, the primary peduncles 3-8 mm. long; flowers yellowish-green, regularly dioecious, but occasionally monoecious in late growth, the pedicels about 2-3 mm. long, accres- cent, the articulation usually at the base or lower third of the stalk. Male flowers: calyx lobes open, ovate-deltoid, subacute to obtuse, glandular-ciliolate, about 1.5 mm. long; petals narrowly oblong to oblanceolate, obtuse, subentire to slightly ciliolate or erose, usually 3—5 mm. long and 1-2 mm. wide; disc thin, cup-shaped, the lobes usually acute; stamens arising from the margin of the disc proper, about [Vor. 42 262 ANNALS OF THE MISSOURI BOTANICAL GARDEN 3 mm. long, the filaments filiform, glabrous, the anthers oblong-ellipsoid, obtuse, pink-punctate; sterile pistil columnar, about 2 mm. long. Female flowers: calyx, petals, and disc as in the male, sometimes the petals smaller; sterile stamens about 1.3 mm. long; pistil flask-shaped, about 4 mm. long, the ovary ovoid, nar- rowed into a columnar style, the stigmata 3-lobed, the lobes flat, recurved, some- times each slightly bifid. Fruits subglobose, bright yellow, the valves broadly elliptic to suborbicular, about 6-8 mm. long and 5-7 mm. wide, 3- to 6- seeded, the pedicels usually green in dry condition. Seeds usually ovoid or ellipsoid, about 5 mm. long and 3.5 mm. wide, minutely areolate, pinkish-brown. Common in lowland slopes or in thickets, at altitudes from 100 to 1 ,400 m.; widely distributed in northern and central Japan, Korea, and China; flowering from April to June. CHINA: CHEKIANG: Hai-wei, mode I-shing, Ching 9 Tso 580 (A). HARBIN: (Pinkiang), along a river, Skvortzov s.n., Dec. 7, 1929 (A). HONAN: Siashih, Hers 345 A); Sunghsien, Shih-tze-miao, Hers ino (A); den feng Hsien, Yu-tai Shan, Hers 2791 (A); Tsi-yuan Hsien, Hers 1758, 2705 (A); Yungning, Yu-t e-ping, Her. шай m without precise locality, Hers 2 HOPEI (cHILI): Ch'o-K'ou-tien, Peking, Boblin 104 (S, UPS); Chanel: Dorsett 8 Morse 7152 (US); western hills, Dorsett & Morse 7206 (UC, US); Prince Park, Wang 1 ; mont. á l'ouest Peking han, an-tso-ling, Liu 1176 (UC) ; Ming Tombs, Nankou, Chiao 2126 (US) ; Tsing-ling-chiao, Chiao 2116 (NY), Dorsett б Morse 70 4 (US); near eastern Tombs, [40и 385 (NY); western Tombs, Liu 308 (UC); without precise locality, Li 11160 (NY). норе: Ichang, vr 3505 (A), Wilson 2308 (А, holotype of Celastrus Pes d var. cuneata Rehd. & ils.) ; Nok: chen Hsien, Chow 369 (NY); Lo-kia Shan, Wuchang, Sun 108 (NY); western Hupei, without preci locality, Wilson 181 (СН, US). JEHOL (GEHOL): David 1783 (СН, US). kxaNcsu: Mien Shan, Lin-shih Hsien, Tang 889 (A); Tienschui, Fenzel 2827 (A). она Liaoyang, Mukden, Li бо (UC). port двтнов: Wilson 8801 (А). NANKING: Le 1402 (UC). SHANSI: Chioh-hsin Dist., Smith 5745 ; Fu-pin зегі "A 348 ( UPS); Бела і "та a-wu-tai Shan, Janet A340 ); Hweihsien, Shansi border, Hers ^y o. PO Red уні A79 (UPS); Dao-hui-gou, Wenshui, 4 ( ; Yuan-chu Dist., Yang-shu-ling, Smith 6154 ). SHANGTUNG: Meng Shan, Fei Нама, Cheo & Yen "103 ( Tsing-tao, First Park, Chiao 2305 (A, NY, US), 2612 ); (А, Е, МУ, ОС AR Tsing-tao, Lioent 13262 (A); 6005 Ching-lung Shan, Chiao 3016 (A, NY, UC, US). ѕнем1: Lao-y-san, Giraldi s . June 5, 1897 (A, UC); Tai- Sieg рий 044 (А, 05), 045 (А); Tingling. ДЫ centr. inter Mei ес Шира, Fenzel 502 (А JAPAN: HONSHU: Chiba Pref., ме шде 5717 (US); Kanagawa, Hakone т Sagami, Obwi 9 Okamot 504, 506 (s US); Hakone, Sawada 2161 (S); Mt. Tanzawa, Suzuki 479013 (UC); Mino Ела P 5222 3905, 0643, 0766 (А); Ргоу. Musashi, Нитизаша 10650 (MO), Yamazala 44 (US); Nagano, Kimura 307 (US), Suzuki 455009 (UC); Niigata, Sado, ar 2502 ШО. Nikako; Shimotsuke, Wilson 58» 1914 (А); Otake-gawa, Shinano Prov., Wilson s. n., 1914 4 (A); Suruga Prov., Wilson 00 4 ; е 0134 (А), стене 44-А amu Kamohura, Sagami Prov., Wilson 6627 (A, US); Kumashii, Tomari, Toyama, Bergman 439 (S); Prov. Hone Mei 1920 (MO); Mitsutooge, bieres Pref, Suzuki MM (UC); Mt. Hoo amanashi, Suzuki 493027 (UC). KKAIDO: Prov. Iburi, Zezo, Takenauchi s. n., Des b. 1916 (S); Sapporo, Faurie 6133 (A), jake. % жа LAE 22, 1905 (A, GH), Miyabe 5. п., June 1880 (А, GH), Wilson 7412 (A); vicinity of Obihiro, Tokachi, Dorsett & Morse 1070 (A, US); southern Hokkaido, 1955] HOU—REVISION OF THE GENUS CELASTRUS 263 without precise locality, Brooks 44 (А, UC), 544 (UPS). ковки ILANDs: Shikotan, Ohwi 436 (UPS). WITHOUT PRECISE LOCALITY: Thunberg s.n. (A, type photo of C. orbiculatus, pl. 961). OREA: Port Chusan, Keisho Nan, Wilford s. n., 1859 (GH, S); Port Hamilton, Wil- ford s. п., 1859 (СН, 5); Keijo (Seoul), Mills s. п., Oct. 2, 1915 (UC); Kongo-san, Prov. Kogen, Wilson 10461 (US); Pyongang, Wilson 0203 (US); Taiyudo Prov., Hsian Hoku, Wilson 8611 (US); without precise locality, Gilbert 6 (UC). Thunberg describes this species (Fl. Jap. p. 97, 1784) as Celastrus articulatus. However, in his index, p. xlii, articulatus is omitted and orbiculatus is listed. In 1796, Gmelin?? used the name orbiculatus and referred to Thunberg's Fl. Jap. p. 97, but he did not mention articulatus. Later, іп 1881, Maximowicz pointed out?? that the name articulatus was due to a printer’s error, having been changed from C. orbiculatus, as shown in the ‘Flora Japonica’ and the figures of Thunberg's manuscript. However, Maximowicz chose Celastrus articulatus as the species name because it was universally accepted at that time. Airy Shaw?* states in 1935 that “The form articulatus is a ‘vain tradition’, and the sooner it is dropped the better—but tradition dies hard!” I agree that it is best to use the originally correct name for this species, even more so because "orbiculatus" refers to the leaf shape. This is a very widely distributed eastern Asiatic species. It is found in northern and central Japan, and provinces north of the Yangtze River in China. е Yangtze River seems to be the southern limit of its distribution, except for a few specimens that have been collected south of the river (northern Chekiang). It is the commonly cultivated and nearly naturalized exotic Celastrus of the eastern United States where it was introduced at the New York Botanical Garden in 1891 from seeds secured from the Royal Botanic Gardens, Kew?*. The fruit- clusters are axillary and remain attractive through the winter after the leaves have fallen; the fruiting branches are as showy as those of the paniculiform Celastrus scandens. Birds are fond of the пре seeds?? and help to propagate them. I have observed two living plants of this species cultivated in the Missouri Botanical Garden. In the spring both have male flowers; all the floriferous branches bear both axillary and terminal inflorescences. In June, the male flowers fall, and the terminal inflorescences fall away from the plant; some of the axillary buds or buds that are near the top of the branchlets sometimes now develop into shoots which bear only female flowers. Some of the female flowers develop into fruits which in these plants are scarce. The leaves are variable. Due to the variation caused by these characteristics, there are many synonyms for this species. S2Syst. Veg. p. 406. 1796. 33M]. Biol. Acad. St. Pétersb. 11:201. 1881. 34Сиге. Bot. Mag. 158:tab. 0304. 1935. 35Nash, C. У. Celastrus articulatus. Addisonia 4:9-10, pL 135. 1919. i articulatus). Gard. Chron. ge, F. HL 23:28, Б n. 18 [Vor. 42 ANNALS OF THE MISSOURI BOTANICAL GARDEN кы = = PROJECT'ON TUS CULATUS TATUS [EE ^ AEN ЕЗ PUNC томата YPOLEUCUS OSTHORNIANUS P m ATUS ELLARI 1$ Мар 3. Distribution of ten species of Celastrus, Subgenus cELastrus, Series II. 149 AXILLARES. 19551 HOU—REVISION OF THE GENUS CELASTRUS 265 18. CELASTRUS ROSTHORNIANUS Loes. in Engl. Bot. Jahrb. 29:445. tab. V. F-H. 1900. (T.: Rosthorn 1572, A, photo!). Celastrus stylosa sensu Lévl. Fl. Kouy-Tchéou, p. 69. 1914, = Wall. Celastrus loeseneri Rehd. & Wils. in Sarg. Pl. Wilson. 2:350. 19 (T.: Wilson 357", A!). quem ton He Lévl. in Monde des Pl. П, 18:31. 1916. (T.: BEI 496, A, photo); n Lévl. 1914 Scandent shrubs up to 7 m. tall; branches terete, occasionally striate, glabrous, gray-brown to brown-red, densely to sparsely lenticelled, the lenticels elevated, ovate; axillary buds ovoid, about 3 mm. long. Leaves elliptic, oval, or obovate- oblong, the apex acute, the base cuneate to obtuse, the margins remotely serrulate to serrate, usually 4—11 cm. long, 2—6 cm. wide, thinly membranaceous on the flowering specimens, firmly membranaceous on the fruiting specimens, glabrous, the primary lateral veins 4—7 pairs, prominent, slightly elevated on both surfaces, the veinlets obscure on both surfaces; stipules filiform, tufted; petioles 0.5—1.5 cm. ong. Inflorescences axillary as well as terminal in the male plants, axillary only in the female, the terminal portion sometimes up to 5 cm. long, usually fascicular, rarely solitary; peduncles glabrous, the primary peduncles very short to nearly obsolete in the fascicular inflorescences, 2-5 mm. long at the solitary ones; flowers dioecious, yellowish-green, the pedicels 2—5 mm. long, the articulation at the middle or lower half of the stalk. Маје flowers: calyx lobes valvate, ovate to oblong, obtuse, subentire or glandular-ciliate, about 3 mm. long and 1 mm. wide; stamens arising from the margin of the disc proper, 2.5 mm. long, the filaments filiform, glabrous, the anthers ovoid, obtuse, slightly brownish-punctate; sterile pistil subulate columnar, about 1.5 mm. long. Female flowers: calyx lobes, petals, and disc as in the male; sterile stamens about 1.3 mm. long; pistil about 3.5 mm. long, the ovary subglobose, the style about 1.8 mm. long, the stigmata 3-lobed, each deeply bifid, filiform. Fruits subglobose, the valves broadly elliptic, about 7 mm. long and 6 mm. wide, 3- to 6-seeded; seeds ovoid or ellipsoid, about 4 mm. long and 2 mm. wide, the areolae distinct, yellowish-brown. Chiefly in thickets, at altitudes from 400 to 3,400 m.; China; flowering from April to May. CHINA: НОРЕГ: Bo-moh-ping, Chun 3570 (А); Changyang, Wilson 688 (A), 706 precise locality, Henry gis (А, GH), 5909 (GH, NY, OS). HUNAN: Hbri Te-Hui Wang KWANGSI T -yuin Hsien, Stewar. Cheo (A, NY); Ling ý 28207, 28763 ( Hsien, Lau А). KWANGTUNG у, Wang 32148 (А) -shan Hsien, Tsui 542 (А, MO, UC, US). xwricHow: Anlung, Tsiang 9354 ( ; (A, МҮ); Kweiting, Tsiang 5443 (A, МУ), 5614 (А); Lang-tai, Y. Yeh-tin, Tsiang 9513 (A, NY); Wai-ho, near Sio-chang, T 5601 (A); without precise locality, Cavalerie 790. (суре о 720 (А cavaleriei, А). SIKANG: near San-tao-chiao, 40 li from I-tung, EE NY 1720 (A); Konting lachen: Wilson 4187 eb rede. Fang 3424 MAY 2 SZECHUAN: Chao -hua Hsien, Hopkinson 297 (5); С tu, Chien 5920 (A), Feng БИ; (А); € g Shan, Wilson 2311 ЖҰ жы; Е * rig ef i Schneider 1375 (А); Mian-nin [Vor. 42 266 ANNALS OF THE MISSOURI BOTANICAL GARDEN Hsien, Yu 1756 (А); Mow Hsien, d ‚5589 (А, МУ); Nan-chuan, Rosthorn 1572 (type, photo of Celas£rus rostbornianus, A), 1753 (А, n of C. rostbornianus) ; Ning-yuen, Handel-Mazzetti 130 5 (A); Telipu, Scheer З I (A); Tien- chuan, Yu-shib Liu 1319 М : 8 (U i ow 12013 (A), Chu 3701 (A), Fang 16410, 1779 ыы Lee 2805 (А), 3069 (A, US), Wil- son 4781 (A); without precise locality, Faber 227 (NY), Henry 5640 (A, GH, МО), 5734 (GH, NY). YUNNAN: A-tun-tze, Wang 69361 (A), Yu 7078 (A); Chung- tien, 2428, 2479, 3262 (A); La Kou, Maire 3633 ser. B (A, NY, UC); Lapping Hsien, Tsai 56096 (A); Liang-shan, Tsai 51340, 51343 (А); иа Ching 22119 ; Muli, near rre Yu 14848 (A); Pan-pien-kai, Maire 7256 (UC); Pin-chuan idu Tsai 5200 I (A); Ping-pien Hsien, Tsai 60781 (A); Si-chour Hsien, Feng 11720, 12084 (A); Shun-ning, Wang 71910 (A); Suen-oui, Maire 81 (A); between Sung-kweh and и, бебнеійсі 268 3 (А, СН, МУ); Tso-si, Майе 254 (A); Wei-se Hsien, Tsai 57974, 59548, 61753, 63050, 64147, 68203 (A); without precise locality, Forrest Pies 26 (A); Tsai 57253, 57271, 57343, 57510, 57769, 60933 (A); Хи 5185, 7204, 5945 This is a common species chiefly distributed in southwestern and central China. Since the leaf characters are quite variable, this species has sometimes been mis- identifed. However, the characters of the flower, fruit, and seed are constant. The racemiform inflorescences bear clustered, usually sessile flowers; the disc-lobes are subquadrate, broader than long, and mucronate; and the seeds are ovoid or ellipsoid and usually yellowish-brown. These characters can distinguish Celastrus rostbornianus from related species. Léveillé first identified a specimen collected by Cavalerie (no. 496) from Kweichow as Celastrus stylosa Wall, and later described it as a new species C. cavaleriei (non Léveillé 1914) although he had previously published the same name based on another specimen. The earlier name subsequently was shown by Rehder?" to be a Myrsine. I have seen the type photo of the Cavalerie 496 specimen, and it compares closely with others collected in Kweichow. They are all fruiting speci- mens, and clearly belong to Celastrus rostbornianus. These Kweichow plants have grayish-brown, striate-fissured branches, and remarkably uniform, elliptic, firmly membranous leaves. This uniform population perhaps could be considered as an ecotype. 19. CELASTRUS PUNCTATUS Thunb. Fl. Jap. 97. 1784. (T.: Collector unknown, 5. п. А, photo!). Celastrus scandens sensu Thunb. in Trans. Linn. Soc. 2:332. 1794, non L. Celastrus punctulatus in Abh. Bayer. Akad. Muench. II, 4:150. 1845 учакта ап еггог Celastrus crispulus Regel, in Gartenfl. 9:407, tab. 312, f. 1-5. 1860, ex char. & ill. Celastrus striatus Miq. in Ann. Mus. Bot. Lugd.-Bat. 2:210. 1865-6 6, non Thunb., ex € Celastrus kiusianus Franch. & Sav. Enum. РІ. Jap. 2:314. 1879, ex char. (T.: Sav 35: Celastrus orbiculeius var. punctatus (Thunb.) Rehd. in Bailey, Cycl. Am. Hort. 1:267. 1900. Celastrus articulatus Thunb. var. punctatus (Thunb.) Makino, in Bot. Mag. Tokyo 21:138. 1907. 37 Your. Arn. Arb. 15:292. 1934. 1955] HOU—REVISION OF THE GENUS CELASTRUS 267 Celastrus gracillimus Hn Icon. Pl. Formosa. 5:24. 1915. (T.: Hayata s. n. A, photo!). Celastrus leiocarpus Hay. loc. cit. 5:22. 1915. (T.: Mori s.n., A, phot o!). Celastrus longe-racemosus Hay. loc. cit. 5:23, pl. * 1913. (ТА His $. no ТАП). Celastrus geminiflorus Hay. boe. cit. 5:25, f. 9. 1915. (T.: Nagasama s. п., А, photo! ). Celastrus elevativenus Hay. loc. cit. 6:14. 1916. (T.: Faurie s. „и А; Celastrus punctatus Thunb. var. тон Li & Hou ex Hou, in Тине 1:172. 1950. (Т.: Yamamoto 802, ТАП). Scandent shrubs, about 2—3 m. tall; branches terete, glabrous, hazel-brown, the lenticels rounded or oval, scattered and elevated; axillary buds deltoid, about 2 mm. long, the outermost bud scales sometimes sharply spinose. Leaves usually "e the apex acute, the base usually cuneate, the margins remotely serrate, 2—7 cm long, 0.8—3.0 cm. wide, membranous, glabrous, the primary lateral veins 4—5 pairs, slightly elevated below, immersed above, the veinlets obscure to distinct; stipules filiform, tufted, about 2 mm. long; petioles about 0.5—1.0 cm. long. Inflorescences axillary as well as terminal in the male plant, axillary only in the female, fascicu- late, racemiform, or solitary, the peduncles obsolete, glabrous, the primary peduncles suppressed to about 5 mm. long; flowers dioecious, white or pale green, the pedicels 1.0—1.5 mm. long, the articulation at the middle or upper third of the stalk. Male flowers: calyx lobes open, deltoid to oblong, obtuse, glandular-ciliate, about 1 mm. long; petals usually oblong to slightly obovate-oblong, ciliate to slightly erose, 2.5-4.5 mm. long and 1.3-1.5 mm. wide; disc cupuliform, the lobes erect, oblong, about half to one-third as long as the disc proper, obtuse; stamens arising from the margin of the disc proper, the filaments filiform, glabrous, the anthers broadly ovoid, obtuse, the base divided to above the middle, pink-punctate; sterile pistil subconical-columniform, slightly 3-lobed, about 1.5 mm. long. Female flower unseen. Fruits globose or subglobose, usually solitary, fasciculate, sessile or oc- casionally short-pedunculate, pale yellow when dried, the valves suborbicular, about 6 mm. in diameter, 3- to 6-seeded; seeds broadly ellipsoid, about 2 mm. long, smooth, pinkish-brown. Hillsides and in thickets, at altitudes from 100 to 2350 m.; southeastern China, Riukiu Islands, and southern Japan; flowering from March to August. A: ANHWEI: Chu-hwa Shan, Ching 2661 (UC). снЕекллмс: Changhua, - stet (А, LUC); south t Siachu, Ching 1677 (MO, a 296 US). FUKIEN: Amoy 493 (SING), Hun. 2447, 1491 (А, wee 1515 (UC), 4604 (A, NY), 5957 (А); Diongloh, Риз Еп п 2447, 2701 (UC), Siu Ging ap 13657 (UC); how, Siu Ging Tang jor A y (A, UC); елей Dunn 44 (А); Mt. Useke, 0 Eder s.n. (A, photo of Рио of C. geminiflorus). ТАГУАМ: Arisan, Gressitt 141 (А, L, МУ, S, U), 185 (A, L, NY, S, U), Hayata s. n., "April 26, 1914 (ТА photo of type of C. longe- racemosus; А), s.n. (А, photo of type of С. gracillimus); Nanko-taisan, Sasaki s. n., July 22, 1922 (A, photo); Kwarenko, Faurie s.n. (A, photo of type of C. elevativenus), Nakamura Krat (TAI); Shinchow, Prov. Karenko, Wilson 11096 (A, US); Taito, Mori . п. (A, photo of уре of C. leiocarpus) ; Mt. Taito, Yamamoto 802 (type of C. punctatus var. полови из, ТАТ). AN: KYUSHU: Mt. Kirishima, Tashiro s.n., May 5, 1917 (A); Kudsi Kadsura, Tsuro Ome (?) 500 (A); Машан. "Oldham IÓI (GH, L, S), 162 (GH), Wilson 630 ^: (A); Satuma, uk s.n., Aug. 13, 1930 (TAI); евини, Kita-tane, Mori s.n Aug. 13, 1934 (TAI); Tanaga-shima, Wilson 6121 (A). u IstaNps: Wright 54 (GH). [Vor. 42 268 ANNALS OF THE MISSOURI BOTANICAL GARDEN The type of Celastrus longe-racemosus is a flowering specimen whereas the type of C. punctatus var. micropbyllus is а fruiting specimen. Until now no fruiting or flowering specimens, respectively, for these two species have been found. On further examination of additional specimens, it was observed that the fruiting branchlets of the type of C. punctatus var. micropbyllus are the same as the so- called long-racemose inflorescences of C. longe-racemosus; consequently, they appear to belong to the present species because they cannot be distinguished from it. On comparing material from Fukien and southern Japan, this interpretation is confirmed by both morphological characters and geographical distribution. In general appearance, the present species differs conspicuously from Celastrus orbiculatus and more closely resembles C. rosthornianus. However, the deeply cleft, oblong disc lobes, the articulations located at the upper third or half of the stalk, as well as the distinct geographical distribution distinguish this species from related ones. 20. CELASTRUS KUSANOI Hayata, in Jour. Coll. Sci. Imp. Univ. Tokyo 30:60. 1911; Icon. Pl. Formos. 1:137. 1911; 5:20. А. 8. 1915, ex char. & ill. (T.: Kusano s. n.). Scandent shrubs up to 18 m. tall; branches terete, sometimes striate, brown, the branchlets glabrous to brownish-pubescent, both sparsely or rarely densely lenticelled, the lenticels orbicular or ovate; axillary buds deltoid, about 2.5 mm. ong. Leaves broadly elliptic to nearly orbicular, rarely elliptic, the apex rounded to shortly cuspidate, the base broadly truncate, rarely obtuse to cordate, 5.0—10.5 cm. long, 5-11 cm. wide, chartaceous, glabrous or pubescent on the veins below, the primary lateral veins 5—7 pairs, slightly elevated on both surfaces, the veinlets distinct below, obsolete above; petioles usually 1.5-2.5 cm. long, rarely up to 5 cm. long; inflorescences axillary and cauline at the basal position of the current year's growth, rarely also terminal in the male plant, usually 3- to 7-flowered, the peduncles usually pubescent, sometimes yellow, the pedicels about 3—5 mm. long, usually pubescent, the articulation nearly toward the base of the stalk. Male flowers: calyx lobes deltoid, obtuse, entire, about 1 mm. long; petals obovate to oblong, rounded, about 4 mm. long and 1.5 mm. wide, ciliate, sometimes pubescent wi on the lower portion; disc subfleshy, flat, the lobes obscure, truncate; stamens arising from the margin of the disc, about 3 mm. long, the filaments fili- form, papillose-tuberculate, the anthers ovoid, subcordate; sterile pistil conical, about 1 mm. long. Female flowers: calyx, petals, and disc as in the male; sterile stamens about 1.5 mm. long; pistil about 3.3 mm. long, the ovary subglobose, the style columnar, the stigmata 3-lobed, reflexed. Fruits globose or subglobose, the valves suborbicular or broadly elliptic, about 8—10 mm. long and 5-7 mm. wide, 3- to 6-seeded; seeds slightly lunate, minutely areolate, dark brown. Chiefly in thickets, at altitudes from 100 to 1,000 m.; southeastern China; pes. from February to April. HAINAN: Bak Sa, Lau 26184, 26606 (А); TUS Ды, Hsien, Lei A, NY, SING, U UC, US); Fan Yah, Chun & Tso 44028 (A, Е, NY, US). NM. 19551 HOU—REVISION OF THE GENUS CELASTRUS 269 = — — = L g » Fig. 10. Celastrus kusanoi Hayata Lokchong, Tsiang 1393 (A, UC), Tso eios A^ 40045 (NY); Tsinleong Shan, Mei Hsien, Gressitt 1240 (А, MO); Lung-t'au village, Kang-peng To 131 (UC), 56, A MO, UC); Nam Shan , Ho-yuen d ылы 22% G (A); Ying-tak, Kang-peng То 8 Кат-с bow Wong 2784, 2928 (UC). TAIWAN: Arisan, Faurie 1374 (А); Козуип, Kudo & Suzuki 16048 (TAI), mper 6195 (TAI); Kuaru, айбу 1215 (TAI); t. Naier, Bansyoryo, Kawakami & Mori 3145 (TAI); Nanto, Wilson 9981 (A); Taipei, Wilson 11214 (A); Takou, Apes Hill, Hee ZR (A, NY. , TAI). Henry?! first regarded the Formosan specimen, collected by himself (по. 1803), as a variety of Celastrus articulatus Thunb., but did not give it a name. Later Hayata described a specimen collected by Kusano as Celastrus kusanoi Hayata, but he did not cite Henry's specimen. Не stated that Celastrus kusanoi is near С. articulatus (C. orbiculatus) but differs from it in having more rounded leaves and 38A List of Plants from Formosa. р. 27. 1895. [Vor. 42 270 ANNALS OF THE MISSOURI BOTANICAL GARDEN carpels transversely wrinkled in the dried condition. His original description was based on a fruiting specimen; later?? he obtained a flowering specimen, and then amplified his description. Celastrus kusanoi bears pubescent filaments, lunate seeds and usually nearly rounded leaves, hence it is easily distinguished. Specimens collected from central and southern Formosa have nearly rounded and sometimes even cordate leaves, whereas those from Hainan and Kwangtung usually bear suborbicular or broadly obovate leaves without cordate bases. Merrill and Chun“? have said that Kwangtung records of Celastrus articulatus Thunb. (C. orbiculatus Thunb.) actually belong to C. kusanoi Hay. and that Thunberg's species does not extend as far as Kwangtung. Of the many Kwang- tung specimens labelled Celastrus articulatus which I have examined, none of them actually is that species; they are either Celastrus Визапої Hay. or C. gemmatus Loes. 21. CELAsTRUS HIRSUTUS Comber, in Notes Roy. Bot. Gard. Edinb. 18:233. 1954, ex char. (T.: Forrest 17963). Scandent shrubs up to 12 m. tall; branches terete or slightly compressed, some- times more or less sulcate, densely brownish-pubescent or glabrescent, the lenticels elevated, large, ovate, elliptic or orbicular, rarely lacking on the current year's growth; axillary buds ovoid, usually about 3 mm., rarely 5 mm., long, the bud scales accrescent, sometimes persistent. Leaves broadly ovate to obovate, the apex round, shortly cuspidate, the base rotund or broadly cuneate, the margins crenate- serrate, 7—14 cm. long, 4—10 cm. wide, membranous, densely pubescent on both surfaces especially so on the veins when young; primary lateral veins 6-7 pairs prominently elevated below, slightly elevated above, the veinlets distinct below, immersed to obscure above; stipules laciniate, about 1 mm. long; petioles pubescent, about 1.5—3.0 cm. long. Inflorescences axillary and cauline at the basal portion of the flowering branch, usually 7- to 14-flowered, the peduncles hirsute, the primary peduncles 5-15 mm. long; flowers dioecious, creamy-white to greenish-yellow, the pedicels 3—5 mm. long, the articulation toward the base of the stalk. Male flowers: calyx lobes valvate, deltoid, obtuse, sparsely ciliate and pubescent without, about 1 mm. long; petals obovate-oblong, puberulous at the basal portions of both sur- faces, about 3.5 mm. long and 2 mm. wide, obtuse, slightly erose; disc membranous, slightly concave, the lobes arcuate, distinct; stamens arising between the disc obes, about 2.5—3.5 mm. long, the filaments subulate, papillose-tuberculate, the anthers ovoid, obtuse, cordate; sterile pistil small, ovoid, about 1.3 mm. long. Female flowers: calyx lobes, petals, and disc as in the male; sterile stamens about 1 mm. long; pistil 3—5 mm. long, the style columnar, the stigmata conspicuously 3-lobed. Fruits subglobose, the valves broadly elliptic, about 8—10 mm. long and 6–8 mm. wide, 3- to 6-seeded; seeds slightly lunate, attenuate toward both ends, about 5 mm. long and 1.5 mm. wide, blackish-brown, the areolae distinct. 39 соп. Pl. Formos. 5:20. 1915. 4011 Sunyatsenia 5:111. 1940. 1955] Fig. 11. Celastrus hirsutus Comber In thickets, at altitudes from 1,400—2,500 m.; China, Burma, and Indo-China; flowering from March to A Burma: Adung valley, 27°30’ to 28°30’ lat., and 97730" to 98730" long., Kingdon Ward 9362 (A). CHINA: YUNNAN: Kengma, Уи 17294 (A); Liukiang, Muchietu, Уи 21012 (A); Ma- ine Hsien, Tsai “1045 (A); Mar-li-po, Sze-tai-po, Feng 13765 (A); Mienning, Уи (A); Ping-pien Hsien, Tsai 55421, 60261, 60499, 62396 (A); Shang-pa Hsien, Tu ‚54/00 M: ; without precise e locality, Forrest er (A). a: Chapa, Tonkin, Pételot 5936 (A The particularly dense pubescence on both ы of the leaf makes Celastrus hirsutus easily recognized and distinguished from other species of the genus. On the other hand, as Comber previously has pointed out, its floral characters are quite similar to Celastrus kusanoi Hay. Specimens of these two entities indicate that although their floral characters are similar, their geographical distributions and [Vor. 42 272 ANNALS ОЕ THE MISSOURI BOTANICAL GARDEN leaf and lenticel characters are distinct. For these reasons I am considering them as two separate species. 22. CELASTRUS STYLOSUS Wall. in Roxb. Fl. Ind. ed. Carey 2:401. 1824; Wall. Cat. 4313. 1831; Lawson, in Hook. Fl. Brit. Ind. 1:618. 1875 (pro parte) ; Prain, in Jour. Asiat, Soc. Bengal 73:198. 1904. (T.: Wallich 4313, МО, photo!). Scandent shrubs, 3—4 m. tall; branches terete, glabrous or puberulent on the branchlets, light brown to dark brown, sparsely lenticelled, the lenticels small, elliptic or ovate; axillary buds globose or ovoid, about 2 mm. long. Leaves usually elliptic-oblong, ovate to obovate, the apex acute, the base acute or obtuse, the margins serrate, 6—15 cm. long, 3—9 cm. wide, membranous to firmly membranous, glabrous or rarely pubescent on the veins below, the primary lateral veins usually 5-7 pairs, distinctly elevated below, plane or slightly elevated above, the veinlets prominent below, obscure to visible above; stipules filiform, about 1 mm. long; petioles usually 1-2 cm. long. Inflorescences axillary and cauline, usually 3- to 7-flowered, at the basal part of the current year's growth (rarely also terminal in the male plant of Celastrus stylosus ssp. stylosus), distinctly pedunculate; the peduncles puberulous to glabrous, the primary peduncles 5-11 mm. long; flowers dioecious, green or pale green, the pedicels 2—5 mm. long, the articulation toward the base of the stalk. Male flowers: calyx lobes imbricate, oval to oblong, obtuse, slightly erose to entire, about 1.5 mm. long; petals obovate, obtuse, slightly erose, about 2—4 mm. long and 1.0—1.5 mm. wide; stamens arising between the disc-lobes, about 2.5 mm. long, the filaments filiform, fleshy, glabrous to papillose-tuberculate, the anthers ovoid, cordate; disc membranous, cup-shaped, the lobes distinctly arcuate or depressed-quadrate; sterile pistil about 1.5 mm. long. Female flowers: calyx lobes, petals, and disc as in the male; sterile stamens about 1 mm. long; pistil flask-shaped, about 3 mm. long, the ovary subglobose, the style distinctly columnar, the stigmata 3-lobed, each lobe bifid, flat, reflexed. Fruits subglobose, the valves broadly elliptic, about 7-12 mm. long and 5-10 mm. wide, 3- to 6-seeded; seeds more or less plano-convex to slightly lunate, attenuate at both ends, reddish- to blackish-brown, about 4—6 mm. long and 1-2 mm. wide, the areolae distinct. KEY TO THE SUBSPECIES A. Staminal filaments always densely papillose-tuberculate; the disc-lobes depressed- quadrate; leaves elliptic, membranous. India АА. Stami 22а. CELASTRUS STYLOSUS ssp. STYLOSUS. Celastrus neglecta Wall. Cat. 4341. 1831, nom. nud. Gymnosporia neglecta Wall. ex Lawson, in Hook. Fl. Brit. Ind. 1:619. 1875; Prain, in Jour. Asiatic Soc. Bengal 73:198. 1904, Novic. Ind. 419. 1905. (based on Celastrus neglecta Wall., Wallich 4341). In thickets, at altitudes from 1,000 to 2,745 m.; India; flowering from March to July. | 1955] HOU—REVISION OF THE GENUS CELASTRUS 273 Fig. 12. Celastrus stylosus ssp. stylosus дім; 2; (CAL), vay 267018 CAL), Kurz s. n. (CAL) SIKKIM Doll Hills Lister 157 ach ith 9 Cave 064 5. п., March 1878 CAL); Punkabari, Lister s. n., April 1878 (CAL); Singolila forest, Rogers s. n., Jan. 0 190 (CAL); Sureil, Dr. Prain’s collector 439 (CAL). WITHOUT PRECISE LOCALITY: Anderson 104 (CAL); Meebold 15012 (S) 22b. CrrasrRus sTyLosus ssp. glaber Ding Hou, stat. et nom. nov. Celastrus hypoleucus forma y. puberula Loes. in Engl. Bot. Jahrb. 29:445. 1900. (T.: Rosthorn 1556", A, pho to!). kar o етем ені; Желе. in Chin. irn Bot. 1:62. 1936; in Contr. Bot. Surv. N. W. China 1:62. арго тю at — in thickets, at altitudes _ 800 to 2,000 m.; China and Indo-China; flowering in CHINA: ANHWEI: Kimen, Ip 41 (UC); E 3165 поља Wong Shan, Ling 1104 (UC). куламеч: Chuen Yuan, {чооң 4 2041 (А); Kwei-lin Hsien, Tsang 38365 (US); Ling-yuin Hsien, Steward & Cheo 87 (А, NY, S, SING), 306 (МУ, 5); Nanning, Seh-feng-dar Shan, Ching 8205 (NY, UC, US); Shang-sze дяк Tsang 241 31 (А, MO); Pin-lam, Ko 55544 (A); Tsin-hung Shió, Hin Yen, Ching 7 Mapo, Pingchow, — 6832 (А, NY). ѕкамс: Tien-chuan diim, (A). szEcHUAN: Nan-chuan, v. Rosthorn 15569 (A, photo of туре of C. byboleuca forma y. puberula); Omei Shan, Fang 15211, 15280 (A), Lee 3168 (A), Yu-shib Liu if | : + &— Q Prj ~ precise locality, Henry 5550 (А, СН). YUNNAN: Chungtien, Feng 3365 (А); Hokin, ii 747 is үз rid Hsien, Tsai 54563, 54578 (A); Mar-li-po, Chung-dzia, Feng 7 (A); e, Henry 10522, 11267 (А, MO); ig веле we Tsai 60420, 60010, АТА dró (A); Maas Hsien, Tsai 54020 (А); Shunning, Wumulune, Yu 16502 ; Si-chour Hsien, Feng 12260, 12283 (A); without овим locality, Forrest 0306, 15900 ХА), Li 1181 (A), Tsai 62841 4 Cura: Chapa, Tonkin, Pételot 5829, 5046 (A, NY, US). These two subspecies are superficially similar except for the staminal filaments, the leaf shapes, and the geographical distributions as shown in [Vor. 42 274 ANNALS OF THE MISSOURI BOTANICAL GARDEN Subspecies glaber is widely distributed in China; its extra-axillary inflorescences and firmly membranous elliptic-oblong leaves make it easy to separate from other related species. The staminal filaments are usually glabrous; however, a few speci- mens collected from Kwangsi and Yunnan are slightly papillose-tuberculate. Because of the intermediate leaf forms, staminal filaments, and distinctly geograph- ical distributions, I consider the Chinese population as a subspecies of C. stylosus. Under the provisions of the International Rules, priority of publication operates only within individual taxa. It would therefore be rather inappropriate to employ Loesener’s epithet “puberula” in this instance, since Loesener chose the epithet with reference to the puberulence of the leaves of Rosthorn 1556", which actually is an abnormal condition for the other known specimens of Celastrus stylosus ssp. glaber. Thus, I have chosen the epithet "glaber" with reference to the more significant character of the staminal filaments. 23. CELASTRUS ACULEATUS Merr. in Lingn. Sci. Jour. 13:37. 1934. (T.: Tsang 20092, NY!). Celastrus bookeri sensu Rehd. & Wils. in Sarg. Pl. Wils. 2:352. 1915, as to specimens cited, in part, non Prain. Celastrus oblanceifolia Wang & Tsoong, in Chin. Jour. Bot. 1:65. 1936, ex char. (T.: 443). Scandent shrubs up to 10 m. tall; branches terete, glabrous, the young branch- lets sometimes brownish-pubescent, brownish-red to dark brown, both branches and branchlets lenticellate, the lenticels orbicular, sparse to dense. АхШагу buds ovoid, about 2.5 mm. long, the outermost scales persistent, usually deltoid and spiny, acute to acuminate, accrescent, up to 5 mm. long. Leaves elliptic to ob- lanceolate, the apex acute, the base cuneate to obtuse, the margins remotely serrulate, 3—10 cm. long, 1.5—6.0 cm. wide, membranous, usually glabrous, rarely pubescent on the veins below, the primary lateral veins 4—5 pairs, arcuate toward the apex, immersed, distinct to slightly elevated below, obscure above, the veinlets usually obsolete on both surfaces; stipules laciniate, filiform, about 1 mm. long; petioles 7-12 mm. long. Inflorescences axillary and cauline at the lower part of the flowering branch, shortly pedunculate, the primary peduncles almost obsolete to 3 mm. long, usually 3-flowered, puberulous; flowers dioecious, greenish-yellow, subsessile, the articulation just below the flower. Male flowers: calyx lobes im- bricate, ovate-deltoid to oblong, obtuse, entire, subglabrous without, accrescent, about 1.0-1.5 mm. long; petals oblong to oblanceolate, obtuse, slightly undulate, puberulent at the basal parts on both surfaces, about 3.5—4.5 mm. long and 1 mm. wide; disc fleshy, annular, entire; stamens arising from the margin of the disc proper, about 5 mm. long, the filaments filiform, densely papillose-tuberculate; sterile pistil ovoid, about 2 mm. long. Female flowers: calyx, petals, and disc as in the male; sterile stamens 1.5 mm. long, papillose; pistil subglobose, the style columnar, distinct, the stigmata discoid or slightly 3-lobed. Fruits subglobose, the valves broadly elliptic, about 8 mm. long and 7 mm. wide, 3- to 6-seeded; seeds 1955] HOU— REVISION OF THE GENUS CELASTRUS 275 D Fig. 13. Celastrus aculeatus Merr. arcuate ог semi-annular, about 4.5 mm. long and 1.5 mm. wide, brown, slightly wrinkled, distinctly areolate. In open fields or in thickets, from lowland up to about 900 m. elevation; south- eastern China; FOVERE from March to April. ed Pose 285 (A); Mar "Hsien 73 , Chung 14 T , UC), 1693 (A, SING, US), 1773 (US), 4774 (А), 4777 (A), Po 6235 (A); Baek- liang and vicinity, Siu-Ging Tang UG); chow, С UC); Diongloh and vicinity, Ku Та; Lin 11677 (UC); Foochow, Norton 1346 (UC), Chung-Chang Tang & Shan En Ma 2936 (US); Кобан, Chung 7264 (А, Е); Kushan and vicinity, Chea 1573 (UC); Kutien, Chung 40 4036 (A); Minhow Hsien, Chung 2065 (UC); Pu- cheng, Ching 2508 (A, UC, US); Tsze-chook-Hang, central Fukien, Dunn s.n. (HK, A); Yenping, Kuang-Han Chou 8215, 8460 (UC); Yuen-fu gorges, Dunn s.n. (HK, A). HUNAN: Yi-chang Hsien, теті 23574 (А, US). KIANGsI: Kaoan, Tsiang 10460 (NY); iennan Hsien, Lau 4367 (A, $, US); Lungnan Hsien, Lau 4801 (А, $, US); Lu pare Chiao 18768 (US); Swe-chuen, Hu 88 9 (A); ле; Ни 784 (A); without prec locality, Ни s.n., 1920 (UC). куамстиме: Ho-yuen, Tsang 28843 (A); бета Wang 3060 (NY, SING); Mt. Lung-t'au, near Iu, Kang-peng To et al 631 (US); Lung- tung, ч 21640 (NY); Ма Hsien, Tsang 21416 (А, NY, S); Mui-nen Hsien, McClure & Shang 4 (UC); Naam-kwan-shan, Lung-men, Tieng 25268 (A); Sin-fung Hsien, Taam 2 (Мог. 42 276 ANNALS OF THE MISSOURI BOTANICAL GARDEN 724, 1041 (A); Kakchieh, Swatow, Gressitt 1767 (МО); Ta-ching, Chun 5530 (A); Тари, Tsang 21189 (А, NY, S); Tsengshing Hsien, Tsang 20092 (МҮ, holotype of С. aculeatus; A, isotype) ; Wai-yeung, Tsui 154 (A, MO, NY, UC, US); Yao-shan, Sin 11889 (NY). This species is well characterized by the arcuate seeds, the papillose-tuberculate filaments, the fleshy and annular disc, and the obsolete veinlets of the leaves. Celastrus aculeatus is a very distinctive species of southeastern China. Most of the specimens which are cited here have been identified as Celastrus bookeri Prain. Celastrus hookeri, however, is easy to separate from Celastrus aculeatus by its glabrous filaments, its cup-shaped disc, and its ovoid seeds. A specimen collected by Ching (2508) from northern Fukien has puberulent veins and slightly oblanceolate leaves. It matches the description of Celastrus oblanceifolia Wang & Tsoong from southern Anhwei which also has oblanceolate eaves. The veinlets described for Celastrus oblanceifolia are reported to be obsolete here and the seeds are said to be curved, both typical characters of Celastrus aculeatus. Because of these morphological similarities and the identical geographical distribution, Celastrus oblanceifolia is placed in synonymy. 24. CELASTRUS FLAGELLARIS Rupr. in Bull. Acad. Sci. St. Pétersb. II, 15:357. 1857, ex char. Celastrus ciliidens Miq. in Ann. Mus. Bot. Lugd.-Bat. 2:85. 1865—66, ex char. Celastrus clemacanthus Lévl, in Fedde, Rep. Spec. Nov. 8:284. 1910. (Т.: Taquet 632, A1). Scandent shrubs; branches and branchlets terete to slightly striate, glabrous, brown to dark brown, the lenticels small, oval, sparse, the young branchlets cling- ing by a line of filiform, branched aerial roots; axillary buds small, depressed-ovoid, usually protected by two prominent, broadly falcate, spiny scales, acuminate, about 3 mm. long. Leaves broadly elliptic to suborbicular, the apex obtuse to shortly acuminate, the base cuneate to obtuse, the margins finely ciliate-serrate, 3.0—5.5 cm. long, 2—5 cm. wide, delicately membranous, glabrous or puberulous on the veins below, the primary lateral veins 4—6 pairs, distinctly elevated below, slightly elevated above, the veinlets distinct, sometimes elevated below, obscure above; stipules laciniate, filiform, about 5 mm. long; petioles 1.5—2.5 cm. long. In- florescences axillary, solitary or fascicled, or clustered on the short young branchlets, the peduncles about 2—5 mm. long; flowers dioecious, white or yellowish-green, the pedicels 1—4 mm. long, the location of articulation varying from the lower to upper half of the stalk. Маје flowers: calyx lobes imbricate, oblong, obtuse, ciliate, about 2 mm. long; petals obovate-oblong to elliptic-oblong, ciliate to slightly erose, about 4 mm. long and 1.3 mm. wide; disc cup-shaped, the lobes incon- spicuous; stamens arising from the margin of the disc, about 4 mm. long, the fila- ments fleshy, filiform, glabrous, the anthers ovoid, obtuse, apiculate, cordate; sterile pistil columnar, about 1 mm. long. Female flowers: calyx lobes, petals, and disc as in the male but smaller; sterile stamens about 0.5 mm. long, subsessile, tri- angular-cordate; pistil about 2.8 mm. long, the ovary subglobose, the style 1955] HOU—REVISION OF THE GENUS CELASTRUS 277 columnar, distinct, the stigmata trilobed, each lobe deeply bifid, reflexed. Fruits globose, the valves broadly elliptic to suborbicular, 6–7 mm. long and 5-7 mm. wide, 3- to 6-seeded; seeds shortly ellipsoid, about 3.5 mm. long and 2 mm. wide, brown, the areolae obscure. Lowland thickets, at altitudes up to 1,000 m.; China, Korea, and Japan; flower- ing from May to August. CHINA: Seas ы Yu-tsien, Ни 1631 (А, UC). NORTHEASTERN CHINA: ad fl. Amur, Maack s.n., 1855 (СН); sinus Possict, Maximowicz s.n., 1860 (GH) Kore a: Chulla, Mrs. ^ - Smitb s. п. (А); Gyouhfeng, Taquet 2723 (A т Quelpaert, аг in muris agror «C «ооо 632 (А, туре 2” {наган ; Hongo-san, Prov. Но apes. а ie 5 (A); Ping Yang, lack » Sept. 18, 1908 (А ); Рак Han, Seoul, Jack s.n., Sept. us pe (A); Мат-зап, жый (Нато), и Heiki, Wilson 8450 (A,US). ЈАРАМ: Kawagishi, Naganoken, Uno 21823 (A, NY); Musatre, Mitake, Hayakawa 5. n., Мау 20, 1910 (S); Thinano, Suwa, Sakurai s.n., Мау 21, 1913 (A); Thinano, Nagano, Sakurai s. n., June 19, 1913 (A); near Lake Yamanaka, Dorsett & Morse 610 A, US). This species is easily recognized by its characteristic ciliate-serrate leaves and the two persistent spiny outermost bud scales. Its branches sometimes produce aerial roots which function as attachment organs on smooth surfaces. The flowers are usually clustered on young shoots which occasionally elongate during the fruiting stage. Subgenus П. RAcEMOocrELasTRUS Ding Hou, subgen. nov. Frutices scandentes. Flores hermaphroditi; ovario triloculare, in quoque loculo ovulo singulo. Capsula semine singulo, ovulis manifeste abortivis binis. America centralis et australis. Type species: Celastrus racemosus (Reiss.) Loes. KEY TO THE SPECIES A. a paniculiform, obviously compound. бы ade or multi- E: up to 14 cm. long; кенесі. d se. Mexico: Chia . lenticellatus i эйс "чаї once or twice compound, up to 6 cm. long; Лайн 15 авг е. тылай сы pud reos in I^ leaf axils; peduncles not associated w зані buds; leaves elliptic. М leges vati Gain Rica, British Мын , Guatem Km El Salvador г, „а ас Мехі 26. C. racemosus CC. Inflorescences solitary = leaf axils; per Рите бно with — ; leaves ovate. an . panamensis AA. ви racemiform, decer compound. D. Pedicels of the flowers 1-2 mm. long. rupe distinctly apiculate; fruit valves about 12-14 mm. long, the "e m. wide; leaves closely serrate. Central Мехісо................-------------- . C. pringlei EE. Aner ‘obscurely apiculate; fruit valves about 14-18 mm. long, the Жы ; leaves remotely serrate to entire. Chiapas, — пин е- ог 9. С. vulcanicolus der тиг ie “ы Salva DD. Pedicels 2 the flowers obsolete than m. lon F. Leaves — te-oblong; ris y large, Ve and Avid; anthers apiculate 30. C. caseariifolins FF. Leaves teas elliptic or rn lenticels small, sparse and obscure; pu not apiculate. Colombia enezuela ‚ C. meridensis [Vor. 42 278 ANNALS OF THE MISSOURI BOTANICAL GARDEN T. ытта й ZZ ESL X. m = KLL KER: aX Tod PAS Map 4. Distribution of seven species of Celastrus Subgenus RACEMOCELASTRUS. 25. CELASTRUS LENTICELLATUs Lundell, in Bull. Torr. Bot. Club 67:616. 1940. (T.: Purpus 7370, US!). Scandent shrubs; branches glabrous, black-brown, the lenticels small, dense, white, slightly elevated; axillary buds conoid, acute, about 2 mm. long. Leaves broadly elliptic, the apex abruptly short-acuminate, the base cuneate to rotund, the margins remotely serrulate, 10—20 cm. long, 5.0-7.8 cm. wide, chartaceous, glabrous, the primary lateral veins distinctly elevated below, plane or slightly im- pressed above, the veinlets prominent below, obscure to visible above. Stipules filiform, about 1 mm. long; petioles rather stout, 6-12 mm. long. Inflorescences axillary as well as terminal, solitary or fasciculate, thrice or multicompound, paniculiform, up to 14 cm. long, much branched at the base, the peduncles glabrous, the primary peduncles 1—5 mm. long; flowers bisexual, the pedicels about 1 mm long, the articulation at the lower part of the stalk. Calyx lobes valvate, ovate, rotund, minutely erose, 0.5—0.8 mm. long; petals oblong, rotund, entire, brown- punctate, 1.8 mm. long and 1 mm. wide; disc fleshy, flat, about 1.5 mm. in 19551 HOU—REVISION OF THE GENUS CELASTRUS 279 diameter, the lobes depressed-rectangular; stamens attached slightly under the margin of the disc, about 2 mm. long, the filaments filiform, glabrous, the anthers ovoid, obtuse, slightly apiculate; pistil short, ovoid, about 1 mm. long, the style stout, the stigmata obscure. Fruit unknown. Mexico: Chiapas; flowering in June. Mexico: Chiapas, Finca San Cristobal, Purpus 7370 (F, type; MICH; US, holotype). The paniculiform inflorescences and the densely lenticellate branches make this species easily separated from other Latin American Celastrus species. 26. CELASTRUS RACEMOSUs (Reiss.) Loes. in Engl. Bot. Jahrb. 24:199. 1898, (as racemosa). Maytenus racemosus Reiss. in Mart. Fl. Bras, 11:30, pl. 4, f. 15. 1861. (T.: Riedel s.n., Сене liebmannii Standl. in Publ. Field Mus. Bot. 8:316. 1931. (T.: Liebmann 14871, ! Celastrus pachyrachis Lundell, in Lilloa 4:382. 1939. (T.: Jahn 476, US!). Celastrus mainsiana Lundell, in Lloydia 2:99. 1939. (T.: Lundell Ag MICH!). Scandent shrubs up to 50 m. tall; branches terete, glabrous, blackish-brown, the lenticels scattered to dense, white, elliptic to orbicular; axillary buds conoid, acute, about 1 mm. long. Leaves elliptic to ovate, the apex acute, the base cuneate to rotund, the margins crenulate-serrulate, 5-12 cm. long, 3—5 cm. wide, firmly membranous, glabrous, the primary lateral veins 7—9 pairs, slightly elevated below, plane and visible above; stipules subulate, erose, about 1 mm. long; petioles 3—10 mm. long. Inflorescences axillary, 1- to 3-branched, once compound, up to 4 ст. long, the primary peduncle glabrous, about 3 mm. long; flowers bisexual, greenish- white, the pedicels 2—3 mm. long, accrescent, the articulation toward the base of the stalk. Calyx lobes imbricate, deltoid, obtuse, subentire, brownish-punctate, about 1 mm. long; petals arie obtuse, subentire, about 1.5 mm. long and 1 mm. wide; disc fleshy, flat, about 1.5 mm. in diameter, the lobes subreniform; stamens attached slightly beneath the disc margin, about 1.5 mm. long, the filaments linear, glabrous, the anthers ovoid, obtuse, rarely apiculate; pistil ovoid, about 1 mm. long, the style columnar and blunt. Fruits ellipsoid, the pedicels 2.5-5.0 mm. long, the valves broadly elliptic, 14—21 mm. long and 8—11 mm. wide, the septa 2.5—6.0 mm. wide, 1-seeded; seeds cylindric, about 10-13 mm. long and 7-9 mm. wide, pinkish-brown, areolae obscure. In wet forests or thickets, at altitudes 1,400—2,400 m.; Mexico, British Hon- duras, Guatemala, Costa Rica, Venezuela, and Brazil; flowering from March to August. Вкл?п.: Rio de Janeiro, Riedel s. п. (US, type of Maytenus racemosus). British Honpuras: El = ayo Dist., Lundell 6307 Ye holotype of C. mainsiana). Costa Rica: к=: andley 8 Valerio 52024 (US). GUATEMALA: Tac iy Stole 713 MEXICO: VERA CR r Jalapa, Pringle 8133 (K, MO, NY, S, UC, US); Mirador, Liebmann 14871 T. ту. Че с liebmannii; MO), 14872, 14573, 14874 4, 148758, ien (F), 14875 (GH, Uc. US), Purpus 8026 (MO. NY, UC, US); Zacuapan, Purpus 7 (Е, MO, NY, UC), 8080 (MO, NY, UC, (Мог. 42 280 ANNALS OF THE MISSOURI BOTANICAL GARDEN NEZUELA: east of El Junquito, Steyermark 57015 (Е, NY, US); Масагао, Jabn 476 (Us. ré не of C. pacbyracbis). This is a widely and disjunctively distributed species of the Latin American Celastrus. It is found in the rain forests or thickets from Vera Cruz, Mexico, southward to Rio de Janeiro, Brazil. Lundell*! has stated, "from examination of the type photograph of C. racemosus (Reiss.) Loes., the similarity of that species and the type of C. liebmannii is striking." I have seen the types of Celastrus racemosus and C. liebmannii; in addition, I have examined specimens collected from Vera Cruz, British Honduras, Guatemala, Costa Rica, and Venezuela. They are all similar and excessively difficult to separate into two species. This species is closely related to Celastrus pringlei Rose and C. vulcanicolus Donn. It can be distinguished from them by the distinctly once compound, aggregate dichasia and longer pedicels. The other two species bear racemiforin inflorescences and have very short pedicels. 27. CELASTRUS PANAMENSIS Lundell, in Contr. Univ. Mich. Herb. 6:40. 1941. (T.: Allen 319, М Scandent shrubs; branches terete, glabrous, smooth, shining, reddish-brown, the lenticels small, obscure, elliptic; axillary buds conoid, acute, about 1 mm. long. Leaves ovate, the apex obtuse, the base rotund, the margins shallowly crenate, 9-14 cm. long, 5.0-7.5 cm. wide, membranous, glabrous, the primary lateral veins 7—8 pairs, elevated below, plane and distinct above, the veinlets slightly elevated below, plane and visible above; stipules deltoid, erose, about 1 mm. long; petioles 11-13 mm. long. Inflorescences axillary as well as terminal, up to 6 cm. long, usually twice compound, the primary peduncles glabrous, about 1.2-2.5 cm. long, associated with a vegetative bud in the axil; flowers bisexual, white, the pedicels about 1 mm. long, the articulation toward the base of the stalk. Calyx lobes im- bricate, obtuse, minutely erose, about 1.2 mm. long, brownish-punctate; petals oblong, rotund, more or less entire, about 2 mm. long and 1.2 mm. wide; disc fleshy, flat, about 2 mm, in diameter, the lobes subreniform; stamens attached slightly beneath the margin of the disc, about 2 mm. long, the filaments filiform, glabrous, the anthers ovoid, slightly apiculate; pistil short, conoid, about 1.5 mm. long, the style columnar and blunt. Fruit unknown. At altitudes 1,400-2,300 m.; Panama; flowering in April. PANAMA: CHRIQUI: Allen 319 (MO, type; P). The floriferous branch of this species is very characteristic and different from other Latin American Celastrus species in having the peduncle with an associated axillary vegetative bud. I have seen only the type collection. It is a young flowering branch. Чо Lilloa 4:380. 1939. 19551 HOU— REVISION OF THE GENUS CELASTRUS 281 Fig. 14. Celastrus Pringlei Rose 28. CELASTRUS PRINGLEI Rose, in Contr. U. S. Nat. Herb. 5:195. 1899. (T.: Pringle 6842, MO!). Celastrus longipes Lundell, in Lilloa 4:381. 1939. (T.: Palmer 106, MO!). Scandent shrubs up to 5 m. high; branches terete, glabrous, brown or reddish- brown, densely lenticelled, the lenticels elliptic or ovate, slightly elevated, white; xillary buds suborbicular, about 1 mm. long. Leaves narrowly elliptic, the apex acute to acuminate, the base attenuate, the margins serrulate, 6–10 cm. long, 2-4 cm. wide, membranous, glabrous, the primary lateral veins 7 pairs, elevated below, plane and distinct above, the veinlets slightly elevated beneath, visible above; stipules filiform, tufted, about 1 mm. long; petioles 5-12 mm. long. Inflorescences axillary, 1- to 4-branched, racemiform, usually 2.0-3.5 cm. long, the pri peduncles glabrous, obscure to about 5 mm. long, the secondary daiis about 3-8 mm. long, with 2—3 prophylls; flowers bisexual, white, the pedicels about 1.2 mm. long, the articulation usually at the upper half of the stalk. Calyx lobes imbricate, ovate, marginate, scarious and slightly ciliate, about 1.5 mm. long; petals obovate, rotund, slightly erose, 2.5 mm. long and 1.2 mm. wide; disc fleshy, flat, the lobes depressed-subquadrate; stamens attached just beneath the disc margin, about 2.5 mm. long, the filaments linear, glabrous, the anthers suborbicular, apiculate; pistil flask-shaped, about 2 mm. long, the ovary ovoid, the style cylindric and blunt. Fruits ovoid, the valves 12-14 mm. long and 7.5-9.5 mm. wide, the (Мог. 42 282 ANNALS OF THE MISSOURI BOTANICAL GARDEN septa 1.0—2.5 mm. wide, 1-seeded; seeds broadly ellipsoid, 10-12 mm. long and 7 mm. wide, pinkish-brown, shining, the areolae distinct. In forests, at altitudes 790-2,300 m.; Mexico; flowering from March to June. Mexico: canyons of mountains above Cuernavaca, Pringle Е (MO, type of С. pringlei; СН, 1, S, UC); Temascaltepec, Hinton 290 (Е), 3506 ча 22 3574 (NY, US), 3717, 6076, 7203 (СН, US), 7380 (US), 9020 (СН, US). RANGO: San Ramón, Palmer 106 (MO, type of C. longipes; US). JALISCO: McVaugh 10246, 10308, 13906 (MICH). MICHOACAN: Tancitaro, Leavenworth % Hoogstraal 1023 (Е, GH, MICH, MO, NY). Moretos: Sierra de Tepoxtlan, Pringle 6008 (GH, MICH, MO, NY; S DC). The type specimen of Celastrus longipes seems to have larger leaves and longer peduncles than the type of C. pringlei. These characters are variable in all the specimens I have examined, and I think that they might be due to the differences in altitude. I have found some floriferous branches which are subtended by foliage leaves. This subtending foliage leaf may be from the previous year’s growth. For example, a specimen (McVaugh 10308), collected at altitudes 2,400—2,600 m. in Jalisco, Mexico, on April 14, has a main branch bearing three floriferous branches. Each of the branches is subtended by a foliage leaf. When I made a cross-section of the main branch two porous rings of spring wood showed clearly. From this I assumed it might be an evergreen species, but, Dr. McVaugh told me he has col- lected another specimen in fruit with all of the leaves fallen off. 29. CELASTRUS VULCANICOLUs Donn. Smith, in Bot. Gaz. 61:373. 1916. (Т.: Donn. Smith 2549, US!). Celastrus chiapensis е in Lilloa 4:380. 1939. (Т.: Matuda 2080, n Celastrus siltepecanus Lundell, in Wrightia 1:155. 1946. (T.: Matuda 5192, MO!). Maytenus williamsii A. Molina R., in Cei ba 1:258. 1951. (T.: Merrill et al Le F!). Scandent shrubs up to 7 m. tall; branches terete, glabrous, gray to reddish- brown, the lenticels obscure on the gray branches while distinct on the reddish- brown ones, orbicular or ovate; axillary buds globose, about 1 mm. in diameter. Leaves elliptic, the apex acuminate or acute, the base cuneate or obtuse, the margins entire, repand, or slightly serrulate, 6-12 cm. long, 2.5—5.0 cm. wide, firmly mem- branous, glabrous, the primary lateral veins 7—9 pairs, curved toward the apex, elevated below, plane and distinct above, the veinlets slightly elevated below, ob- scure to visible above; stipules filiform, about 1 mm. long; petioles 4-12 mm. long. Inflorescences axillary, racemiform, 1- to 5-branched, up to 3.5 cm. long, few- flowered, the primary peduncles glabrous, obsolete to about 5 mm. long, the secondary peduncles about 3 mm. long, usually with 2 prophylls; flowers bisexual, pale green, the pedicels about 2 mm. long, the articulation usually at the middle or lower half of the stalk. Calyx lobes ovate or deltoid, thick, obtuse, marginate, scarious, slightly erose, about 1 mm. long; petals oblong-elliptic, obtuse, minutely erose, about 2.0-2.5 mm. long and 1 mm. wide; disc fleshy, flat, about 2 mm. іп diameter, the lobes depressed-rectangular; stamens attached beneath the disc mar- gin, about 2 mm. long, the filaments filiform, glabrous, the anthers subglobose, slightly apiculate; pistil 1.5 mm. long, the style slender and blunt. Fruits ellipsoid, the valves broadly elliptic, about 14-18 mm. long and 7-10 mm. wide, 1-seeded; 1955] HOU—REVISION OF THE GENUS CELASTRUS 283 seeds ellipsoid, 12-14 mm. long and 6-9 mm. wide, reddish- brown, shining, the areolae obscure. In forests, at altitudes 1,300—2,400 m.; Mexico, Guatemala, Honduras, and El Salvador; flowering from December to January. Er ЗАГУАРОК: Santa Ana, north of Metapán, Carlson F). URAS: Dept. Morazan, Мер et al. 15640 (Е, type of Maytenus williamsii; US); San Juancito, Williams & Molin (F). UATEMALA: Dept. Alta Vocus Standley 71 349, P Dept. Chiquimula, Steyer- mark 31480 (Е); Dept. Quezaltenango, Steyermark 33635, 33602, 33775 (F), Standley 65402 (F), 65411, 86072 (F, MICH); Dept. Sacatepequer, Standley 63667 (F), Donn. Smith 2549 (US, суре of С. vulcanicolus). EXICO: CHIAPAS: Fraylesca, Siltepec, Matuda 5102 (MO, type of C. siltepecanus) ; Cascada, Siltepec, Matuda 5145 (Е); Mt. Ovando, Matuda 2080 (MICH, holotype of C. chiapensis), 3044 (Е, MICH, MO, NY, US), 16396 (UC). нірлісо: Molango, Moore 2698 (G oaxaca: Cumbre de Talea, Reko 4019 (US). This species is closely related to Celastrus pringlei Rose, but can be distinguished from it by the entire or remotely serrulate leaf margins, the obscure connectives, the larger fruits with wider septa, and especially the geographical distribution. The type specimens of Celastrus siltepecanus Lundell is similar to the present species except for the distinct and slightly elevated veins on both surfaces of the leaves, and the small distinct lenticels borne on the branches. These characters might be caused by the environment and are within the range of variation of the species. The specimens assigned to Maytenus williamsii A. Molina В. are Celastrus and are congruent with the present species. The leaves and fruits are smaller, which might be due to the high altitude of the habitat. 30. CELASTRUs CASEARUFOLIUS Lundell, in Lilloa 4:379. 1939. (Т.: Lehmann 5. n., F1). Branchlets terete, glabrous, reddish-brown, densely lenticellate, the lenticels oval, elevated; axillary buds conoid, about 1.5 mm. long. Leaves oblanceolate- oblong or elliptic-oblong, the apex acute to shortly acuminate, the base obtuse, the margins remotely serrulate, 4.5—10.5 cm. long, 1.4—4.0 cm. wide, chartaceous, glabrous, the primary lateral veins 7-9 pairs, veins and veinlets slightly elevated below, obscure above; petioles 3—5 mm. long. Inflorescences axillary, racemiform, 1- or 2-branched, up to 2.5 cm. long, the primary peduncles obscure, the secondary peduncles about 1.5 mm. long. Flowers (young) bisexual, the pedicels obsolete to about 0.8 mm. long, the articulation at the upper half of the stalk. Calyx lobes imbricate, ovate or ovate-deltoid, erose-ciliolate; petals ovate-oblong, obtuse, sub- entire; disc fleshy, flat, the lobes obscure, the stamens arising from the margin of the disc, the filaments glabrous, the anthers suborbicular, apiculate; pistil ovoid, the style short and columnar, the stigmata obscure. Fruits cylindric, the valves oblong, about 16 mm. long and 7 mm. wide, 1-seeded; seeds cylindric, obtuse at both ends, 15 mm. long and 6 mm. wide, black and smooth. In the forests of highlands, at altitudes 1,600—2,000 m.; Colombia; flowering from May to June. [Уот.. 42 284 ANNALS OF THE MISSOURI BOTANICAL GARDEN LoMBIA: Dept. de Antioquia, Daniel 3205 (Е); күйе of Popayan, Lebmann s. n., 1,600-2,000 m. elev., Мау 1889 (Е, holotype), 399 (L, NY). Celastrus caseariifolius is characterized by the innumerable, elevated and crowded lenticels, very much resembling a crowded colony of plant lice (aphids). Further, its chartaceous and oblanceolate-oblong leaves are distinctive. 31. CELASTRUS MERIDENSIS Pittier, in Bol. Soc. Venez. Cienc. Nat. 3:423. 1927. (T.: Gebringer 298, US!). Maytenus meridensis (Pittier) Cuatr. in Fieldiana, Bot. 27?:82. 1951. Scandent shrubs; branches slightly striate, glabrous, brownish, the lenticels sparse, elliptic, slightly elevated, white, obscure on the young branchlets; axillary ovoid, about 2 mm. long. Leaves ovate, ovate-oblong, or obovate, the apex rotund to abruptly acute, the base cuneate to rotund, the margins remotely crenate- serrate, 4.5—7.0 cm. long, 2—5 cm. wide, firmly chartaceous, glabrous, the primary lateral veins 5—7 pairs, elevated below, plane and distinct above, the veinlets visible below, obscure above; stipules filiform, about 1 mm. long; petioles 3-7 mm. long. Inflorescences axillary, simple, racemiform, up to 5 cm. long, the primary peduncles obscure, glabrous, the secondary peduncles about 1 mm. long, with 2 small pro- phylls; flowers bisexual, white, the pedicels obscure, accrescent, up to 3 mm. long on the fruiting specimens. Calyx lobes,imbricate, ovate, rotund, slightly erose, thick, about 1.2 mm. long; petals oblong, obtuse, slightly erose, about 2.6 mm. long and 1.3 mm. wide; disc fleshy, flat, about 1.6 mm. in diameter, the lobes depressed, subreniform; stamens attached just beneath the disc margin, about 2 mm. long, the filaments linear, glabrous, the anthers ovoid, obtuse; pistil pear- shaped, about 1.5 mm. long, the style short, columnar and blunt. Fruits ovoid, the valves broadly ovate to suborbicular, about 11 mm. long and 7.5-9.0 mm. wide, the septa 2—3 mm. wide, 1-seeded; seeds ellipsoid, 7 mm. long and 5 mm. wide, reddish-brown, shining, the areolae obscure. In thickets, at altitudes 2,490—2,700 m.; Colombia and Venezuela; flowering in July. CoLoMBiA: Cordillera Oriental, Dept. Boyacá, Me gro 1813, 1831 (F, US). Ченкн: Маны, Gebriger 208 (US, суре; Е, NY). This species distinctly belongs to ж The morphological characters match the generic characters very well The type specimen is a scandent, flower- ing plant. In addition to the type, I have several fruiting specimens at hand, which con this view. Cuatrecasas/? transferred this species to Maytenus based on his own collections (1813 and 1831, Е, US). Не says, “Му specimens were obtained from trees, justifying their inclusion in the genus Mayfenus". Оп examining the specimens, especially Cua£recasas 1813 US, one can see easily the twisted branches; it seems a scandent plant. Unfortunately, there is no habit data on the specimens. Since all the Celastrus species are scandent shrubs, I assume that Dr. Cuatrecasas' mom "obtained from trees" were in reality scandent. 42In Fieldiana, Bot. 27?:82. 1951. 1955] 4 HOU—REVISION OF THE GENUS CELASTRUS 285 DOUBTFUL SPECIES he type specimens or representative specimens of the following species are not available, while their original descriptions alone are not sufficient to place them. CELASTRUS DISCOLOR Lévl. in Bull. Géogr. Bot. 24:142. 1914 (T.: Cavalerie 3019). China. CELASTRUS GRENADENSIS Urb. Symb. Antill. 5:51. 1904 (T.: Eggers 6222). Ind. Occ. CELASTRUS MicrocarPus D. Don, Prod. Fl. Мер. 191. 1825 (Т.: Kamroop s. n.). Reg. Himal. CELASTRUS RACEMOSUS var. TRINITENSIS Urb. Symb. Antill 5:52. 1904 (T.: Baptiste 5857). Trinidad. CELASTRUS REPANDUS Bl. Bijdr. Fl. Ned. Ind. 1145. 1825 (Т.: none). Java. CELASTRUS RETICULATUS Wang in Chin. Jour. Bot. 2:68. 1937 (T.: Leu 233). China EXCLUDED SPECIES From this list have been omitted all indigenous species of Africa proper, most of which are species of Gymnosporia or Maytenus and require special study of precise disposition. Celastrus acuminatus Wall. Cat. по. 4342. 1831 == CHAILLETIA GELONIOIDES Hook, f. Fl. Brit. Ind. 1:570. 1875 Celastrus adenophylla Мід. Ann. Mus. Bot. Lugd. Bat. 2:85. 1865—1ILEX CRENATA Thunb. Fl. Jap. 78. 1784. Celastrus alatus Thunb. Fl. Јар. 98. 1784 == EUONYMUS THUNBERGIANUs Bl. Bijdr. Fl. Ned. Ind. 1147. 1825. Celastrus alpestris Bl. Bijdr. Fl. Ned. Ind. 1145. 1825 — PERROTTETIA ALPESTRIS Loes. in Engl. & Prantl, Nat. Pflanzenfam. III, 5:220. 1892. Celastrus apbyllus Schlecht. in Linnaea 15:458. 1841 == ACANTHOTHAMNUS APHYLLUS Standl. in Contr. U. S. Nat. Herb. 23:684. 1923. Celastrus aquifolius Regel, Ind. Sem. Hort. Pétrop. 36. 1856, nom. nud. Celastrus attenuatus Wall. Cat. no. 4319. 1831 = GYMNOSPORIA NEGLECTA Lawson in Hook. Fl. Brit. Ind. 1:619. 1875. Celastrus bilocularis Е. Muell. in Trans. Phil. Inst. Vict. 3:31. 1859 == MAYTENUS BILOCULARIS (F. Muell.) Loes. in Engl. & Prantl, Nat. Pflanzenfam. 2 Abt. 20b:135. 1942. Celastrus iiis Jack, in Malay. Misc. 1:19. 1820 == MicnornopPis BIVALVis Wall. Cat. no. 4340. 1831. кос паз borda Baill. Hist. Pl. 6:26. 1877, in text, nom. nud. == MAYTENUS BOARIA Molina, Saggio Stor. Nat. Chile, ed. 1, 177. 1778. Celastrus bodinieri Lévl. in Fedde, Rep. Sp. Nov. 13:263. 1914 == ILEX PURPUREA Hassk. Cat. Pl. Bogor. 230. 1844; Rehd. in Jour. Arnold Arb. 14:239. 1933. Celastrus buxifolia Wall. in Roxb. Hort. Beng. 18. 1814, nom. nud.; Fl. Ind. ed. [Vor. 42 286 ANNALS OF THE MISSOURI BOTANICAL GARDEN Carey & Wall. 2:396. 1824, in syn.: Celastrus rigidus Wall. = GyMNOSPORIA WALLICHIANA Laws. in Hook. Fl. Brit. Ind. 1:621. 1875. Celastrus cavaleriei Lévl. in Fedde, Rep. Sp. Nov. 13:262. 1914 == Мувяме SEMI- SERRATA Wall. in Roxb. Fl. Ind. ed. Carey & Wall. 2:293. 1824; Rehd. in Jour. Arnold Arb. 15:292. 1934. Celastrus cbungii Merr. in Sunyatsenia 3:253. 1937 == Ткіртекүсіум WILFORDII Hook. in Benth. & Hook. Gen. Pl. 1:368. 1862-67. Celastrus circumcissus Pavon, mss.; Briq. in Ann. Conserv. & Jard. Bot. Genéve 20:253. 1919 == MayTENus ORBICULARIS (Willd.) Loes. in Engl. Bot. Jahrb. 50 (Beibl 111):10. 1913. Celastrus colombianus Cuatrec. in Fieldiana Вог. 272:81. 1951 == ILEX $САМРЕМ$ Cuatrec. in Lloydia 11:207. 1949. Celastrus confertus Ruiz & Pavon, Fl. Peruv. 3:7. 1802 == MAYTENUS CONFERTUS (Ruiz & Pavon) Loes. in Engl. & Prantl, Nat. Pflanzenfam. 2 Abt. 20b:146. 1942. Celastrus crenatus Forst. f. Prod. 19. 1786 == бүммовромА СКЕМАТА (Forst.) Seem. Fl. Vit. 1:40. 1865. Celastrus crenatus sensu Е. Brown in Bull. Bishop Mus. Honolulu no. 130:158. 1935, non Forst. == GYMNOSPORIA sp., ex char. Celastrus crenatus sensu Hook. & Arn. Bot. Beechey Voy. 61. 1841 == Суммо- SPORIA VITIENSIS (А. Gray) Seem. Fl. Vit. 1:41. 1865 с crenatus Roth, Nov. Pl. Sp. 156. 1821 == GYMNOSPORIA MONTANA . Fl. Austral. 1:400. 1863. | зі crenulatus Wall. Cat. по. 4323. 1831, nom. nud. Celastrus cuneifolius (Wr. ex A. Gray) Gomez de la Maza in Ann. Inst. Segunda Enseñanza 2:172. 1895 == EUONYMUS CUNEIFOLIUs C. Wright ex A. Gray in Mem. Am. Acad. n. s. 8:171. 1861— not Celastrus (capsules 2-valved). Celastrus cunninghamii Е. Muell. in Trans. Phil. Inst. Vict. 3:30. 1859 == MAY- TENUS CUNNINGHAMII (F. Muell.) Loes. in Engl. & Prantl, Nat. Pflanzenfam. 2 Abt. 20b:136. Celastrus cunninghamii var. parviflora Е. M. Bailey in Queensland Agr. Jour. 29:178, pl. 22. 1912 == MAYTENUs CUNNINGHAMII (Е. Muell.) Loes. in Engl. & Prantl, loc. cit. 1942. Celastrus dilatatus Thunb. in Trans. Linn. Soc. 2:332. 1794 == OrIxa JAPONICA Thunb. Nov. Gen. Pl. 3:57. 1783. Celastrus disperma F. Muell. in Trans. Phil. Inst. Vict. 3:31. 1859 — MAYTENUS DISPERMUS (F. Muell) Loes. in Engl. & Prantl, Nat. Pflanzenfam. Abt. 2. 20b:135. 1942. Celastrus diversifolia Hemsl. in Jour. Linn. Soc. 23:123. 1886 == GYMNOSPORIA DIVERSIFOLIA Maxim. in Bull. Acad. Sci. St. Pétersb. III, 27:459. 1881. Celastrus dubia Spreng. Syst. Veg. 1:774. 1825— not Celastrus (capsules 2-valved). Celastrus emarginatus Ruiz & Pavon, Fl. Peruv. 3:6, t. 229, f. a. 1802 — Mar- TENUS RETUSA Briq. in Ann. Conserv. & Jard. Bot. Genéve 20:351. 1919. Celastrus emarginatus Willd. Sp. Pl. 12:1128. 1798 == GyMNOSPORIA EMARGINATA Thw. Enum. Pl. Zeyl. 409. 1864. Celastrus esquirolianus Lévl. Fl. Kouy-Tchéou, 69. 1914 — RHAMNUS CRENATUS Sieb. & Zucc. in Abh. Bayer. Akad. Muench. II, 4:146. (Fl. Jap. Fam. Маг. 1:38). 1845; Rehd. in Jour. Arnold Arb. 15:13. 1934. 1955] HOU—REVISION OF THE GENUS CELASTRUS 287 Celastrus esquirolii Lévl. in Fedde, Rep. Sp. Nov. 13:262. 1914 — SABIA PARVI- FLORA Wall. var. морима Lévl. Fl. Kouy-Tchéou, 379. 1915; Rehd. in Jour. Arnold Arb. 15:10. 1934. Celastrus euonymoidea Lévl. Fl. Kouy-Tchéou, 419. 1915 == GREWIA FEDDEI (Lévl.) Burret in Notizbl. Bot. Gart. Berlin 9:678. 1926. Celastrus fasciculatus Drake in Grandidier, Hist. Phys., Nat., e Pol. de Madag. 35 (Hist. Nat. Pl. Atlas 3:рі. 280). 1896 == Gymnosporia sp., ex ill. Celastrus feddei Lévl. in Fedde, Rep. Sp. Nov. 13:263. 1914, quoad specim. Esquirol 3189 == Grewia HENRY! Burret in Notizbl. Bot. Gart. Berlin 9:674. 1926; Rehd. in Jour. Arnold Arb. 18:221. 1937. Celastrus feddei Lévl. in Fedde, Rep. Sp. Nov. 13:263. 1914, excl. Esquirol 3189 == GREWIA FEDDEI (Lévl.) Burret, loc. cit. 678. 1926. Celastrus finlaysonianus Wall. Cat. no. 4324. 1831, nom. nud. Celastrus floribundus Span. in Linnaea 15:186. 1841, nom. nud. Celastrus fournieri Panch & Sebert, Not. Bois Nouv. Caléd. 234. 1874 == May- TENUS FOURNIERI (Panch. & Sebert) Loes. in Engl. & Prantl Nat. Pflanzenfam. 2, Abt. 20b:138. 1942. Celastrus glaucus Vahl, Symb. Bot. 2:42. 1791 == ELAEODENDRON GLAUCUM Pers. Syn. Pl. 1:241. 1805. Celastrus baenkea Spreng. Syst. Veg. 42:88. 1827 == SCHOEPFIA FLEXUOSA Roem. & Schult. Syst. Veg. 5:160. 1819. Celastrus bamelii Spreng. Syst. Veg. 1:774. 1825 == RHAMNUS RACEMOSUS Duham. Traité Arb. et Arbust. ed. nov. 3:48. 1806. Celastrus heterophyllus Savi in Mem. Accad. Sci. Torino 38:163 (Mem. de С. С. Savi, р. 11, tab. П, f. 2.). 1835—not Celastrus (fruit triangular). Celastrus beyneana Roth in Roem. & Schult. Syst. Veg. 5:421. 1819, p. p. — GYMNOSPORIA HEYNEANA Laws. in Hook. Fl. Brit. 1:620. 1875. Celastrus ilicifolius Schrad. in Goett. Gel. Anz. 1:716. 1821 == MayTENus TRUNCATUS Reiss. in Mart. Fl. Bras. 111:5. 1861. Celastrus japonicus (Thunb.) Koch, Dendrol. 1:625. 1869 == ORIXA JAPONICA Thunb. Nov. Gen. Pl. 3:57. 1783. Celastrus jodinii Steud. ex Gópp. іп Gartenflora 3:312. 1854, in obs.: Ilex cunei- folia Hook. f. == Trica уогими (Steud. ex Gópp.) Briq. in Candollea 6:21. 1935. Celastrus kouytcbensis Lévl. in Fedde, Rep. Sp. Nov. 13:264. 1914 = RHAMNUS CRENATUS Sieb. & Zucc. in Abh. Bayer. Akad. Müench. II, 4:146. (Fl. Jap. . Nat. 1:38.) 1845. слон leptopus Drake in Grandidier, Hist. ies 35. (Hist. Nat. Pl. Atlas А). 1896 == GYMNOSPORIA sp., ex i Седаны linearis var. madagascariensis Drake іп НМР loc. cit., pl. 2808. 1896 — GYMNOSPORIA sp., ex ill Celastrus lineatus (Wr.) Gomez de la Maza in Ann. Soc. Езрай. Hist. Nat. 19:239. 1890 == MAYTENUS LINEATUS С. Wr. in Griseb. Cat. Pl. Cub. 54. 1866. [Vor. 42 288 ANNALS OF THE MISSOURI BOTANICAL GARDEN Celastrus listeri Prain in Jour. Asiat. Soc. Beng. 73:197. 1904 — GYMNOSPORIA sp. Celastrus lucida Wall. in Roxb. Fl. Ind. ed. Carey & Wall. 2:400. 1824—not Celastrus (ovary many-celled). Celastrus lycioides Bross. ex Willd. in | Roem. & Schult. Syst. Veg. 5:427. 1819— not Celas£rus (branches spiny). Celastrus lyi Lévl. in Fedde, Rep. Sp. Nov. 13:264. 1914 — RHAMNUS ESQUIROLII Lévl. in Fedde, Rep. Sp. Nov. 10:473. Celastrus macrocarpus Ruiz & Pavon, Fl. Peruv. 3:8, #. 230, f. b. 1802 == May- TENUS MACROCARPUS (Ruiz & Pav.) Briq. in Ann. Conserv. & Jard. Bot. Genéve 20:561. 1919. Celastrus ? magellanicus DC. Prod. 2:8. 1825 — MAYTENUS MAGELLANICANUS . f. Fl. Antarct. 254. 1847. Celastrus mairei Lévl. in Fedde, Rep. Spec. Nov. 13:264. 1934 — SABIA YUNNAN- ENSIS Franch. in Bull. Soc. Bot. Fr. 33:465. 1886; Loes. in Ber. Deut. Bot. Ges. 32:543. 1914. Celastrus mauritiana Stadtm. ex Willem. in Ust. Ann. Bot. 18:22. 1798; Roem. & Schult. Syst. Veg. 5:428. 1819— not Celastrus (leaves ternate). Celastrus maytenus Willd. Sp. Pl. 12:1127. 1798 == Maytenus волміл Molina, Saggio Stor. Nat. Chile ed. 1. 177. 1778. Celastrus mexicanus Moc. & Sessé ex DC. Prod. 2:8. 1825 — WIMMERIA MEXI- CANA (Moc. & Sessé) Lundell in Bull. Torr. Bot. Club 67:618. 1940. Celastrus micrantba Roxb. Hort. Beng. 86. 1814, nom. nud.; Fl. Ind. ed. Carey 2:393. 1824—not Celastrus (leaves pinnate). Celastrus mollis Decne. in Rev. Hort. II, 4:425. 1845—46—пог Celastrus (leaves posite) Celastrus montana Roth ex Roem. & Schult. Syst. Veg. 5:427. 1819; Roth, Nov. Pl. Sp. 154. 1821==GyMNosroria MONTANA Benth. Fl. Austral. 1:400. 1863. Celastrus moya О. Kuntze, Rev. Gen. 32:37. 1898 == Moya sposa Griseb. РІ. Lorentz. 63, pl. І, f. 3. 1874. Celastrus muelleri Benth. Fl. Austral. 1:399. 1863 — MAYTENUS sp. Celastrus myrtifolius Linn. Sp. Pl. 196. 1753 == Prunus МУКПЕОЏА (Linn.) . Urban, Syn. Ant. 5:93. 1904 Celastrus neglecta Wall. Cat. по. 4341. 1831 == GYMNOSPORIA NEGLECTA Laws. in Hook. Fl. Brit. Ind. 1:619. 1875 Celastrus nepalensis Steud. Nom. ed. 2, 1:315. 1840 == PrrTOSPORUM FLORI- BUNDUM Wight & Arn. Prod. Ind. Orient. 154. 1834. Celastrus obtusatus Presl, Bot. Bemerk. 34. 1844 — SIMMONDSIA CALIFORNICA Nutt. in Hook. Lond. Jour. Bot. 3:400, 7. 16. 1844. Celastrus obtusifolia Roxb. Hort. Beng. 86. 1814, nom. nud.; Fl. Ind. ed. Carey & Wall. 2:393. 1824 = Gymnosporia TRIGYNA Baker, Fl. Maurit. 50. 1877. Celastrus octogonus L'Hér. Sért. 7. 1788 — MAYTENUs ORBICULARIS ( Willd.) Loes. in Engl. Bot. Jahrb. 50 (Beibl. 111):10. 1913. 1955] HOU—REVISION OF THE GENUS CELASTRUS 289 Celastrus opposita Wall. in Roxb. Fl. Ind. ed. Carey & Wall. 2:398. 1824 — PLEUROSTYLIA COCHINCHINENSIs Pierre, Fl. For. Cochinch. Fasc. 20, sub. і. 305, in text. 1894. Celastrus orbicularis Willd. mss. ex НВК. Nov. Gen. et Sp. Pl. 7:65. 1825 = MAYTENUs orBICULARIS ( Willd.) Loes. in Engl. Bot. Jahrb. 50 (Beibl. 111):10. 1913. Celastrus orixa Sieb. & Zucc. in Abh. Bayer. Akad. Muench. II, 4:150. 1845 — Orxa JAPONICA Thunb. Nov. Gen, Pl. 3:57. 1783. Celastrus ovalifolius Steud. Nom. Bot. ed. 2, 1:315. 1840 == СУММО5РОКЈА OVATA Laws. in Hook. Fl. Brit. Ind. 1:619. 1875. Celastrus ovatus Hill. Veg. Syst. 13:62, f. 12. 1824 — COLUBRINA FERRUGINOSA Brongn. in Ann. Sci. Nat. I, 10:369. 1827. Celastrus ovatus Wall. Cat. no. 4308. 1831 == MayTENus ovata (Wall.) Loes. in Engl. & Prantl, Nat. Pflanzenfam. 2 Aufl. 20b:140. 1942. Celastrus oxyphyllus Wall. Cat. no. 4312. 1831 = GYMNOSPORIA ACUMINATA Hook. Fl. Brit. Ind. 1:619. 1875. - Celastrus pallidus Wall. Cat. по. 4307. 1831 == СУММОЗРОКЈА MONTANA Benth. Fl. Austral. 1:400. 1863. Celastrus parviflorus Vahl, Symb. Bot. 1:21. 1790 = GYMNOSPORIA sp. Celastrus parvifolius А. Rich. Ess. Fl. Cub. 349. 1845— not Celastrus (fruits 2- valved and 1- to 2-seeded) . Celastrus pauciflora Wall. in Roxb. Fl. Ind. ed. Carey & Wall. 2:400. 1824—not Celastrus (ovary 1-celled). Celastrus pentagyna Zipp. ex Span. in Linnaea 15:186. 1841, nom. nud. Celastrus quadrangulatus Schrad. in Goett. Gel. Anz. 1:716. 1821 == MayTENus QUADRANGULATUS (Schrad.) Loes. in Engl. & Prantl, Nat. Pflanzenfam. 2 Aufl. 20b:142. 1942 Celastrus retusa Poir. in Lam. Encycl. Méth. Suppl. 2:146. 1811 == MayTENus RETUSA Briq. in Ann. Conserv. & Jard. Bot. Genéve 20:351. 1919. Celastrus rbombifolius Hook. & Arn. in Hook. Bot. Misc. 3:170. 1833 == IoDINA RHOMBIFOLIA Hook. & Arn. ex Reissek, in Mart. Fl. Bras, 11:78. 1861. Celastrus richardi Gomez de la Maza, Dicc. Bot. Nom. Vulg. 25. 1889 == May- NUS BUXIFOLIUS Griseb. Cat. Pl. Cub. 53. 1866. Celastrus ricbardi y. latifolius Gomez de la Maza in Ann. Soc. Езрай. Hist. Nat. 19:239. 1890, nom. nud. Celastrus richardi 8. cocblearifolius Gomez de la Maza, loc. cit. 239. 1890 == MAYTENUS COCHLEARIFOLIUS Griseb. Cat. Pl. Cub. 53. 1866. Celastrus richardi є. elaeodendroides Gomez де la Maza, loc. cit. 239. 1890 = MAYTENUS ELAEODENDROIDES Griseb. loc. cit. 54. 1866. Celastrus rigida Wall. in Roxb. Fl. Ind. ed. Carey & Wall. 2:396. 1824 == Суммо- SPORIA WALLICHIANA Laws. in Hook. Fl. Brit. Ind. 1:621. 1875. Celastrus robustus Roxb. Hort. Beng. 18. 1814, nom. nud.; Fl. Ind. ed. Carey & Wall. 2:395. 1824 == Kurrimia sp. [Vor. 42 290 ANNALS OF THE MISSOURI BOTANICAL GARDEN Celastrus rotbianus Wight & Arn. Prod. Fl. Ind. Orient. 1:159. 1834 = Сумко- SPORIA ROTHIANA Laws. in Hook. Fl. Brit. Ind. 1:620. 1875. Celastrus royleanus Wall. Cat. no. 4317. 1831 == GvMNosPonIA ROYLEANA Laws. loc. cit. 620. 1875. Celastrus rufa Wall. in Roxb. Fl. Ind. s ord & Wall. 2:397. 1824 — GvMNo- SPORIA RUFA Laws. loc. cit. 620. 18 Celastrus salicifolia Lévl. in Fedde, ee и Nov. 13:263. 1914 == ILEX МАСКО- CARPA Oliv. in Hook. Icon. Pl. 18:#. 1787. 1888; Rehd. in Jour. Arnold Arb. 14:242. 1955; Celastrus seguini Lévl. in Fedde, Rep. Spec. Nov. 13:262. 1914 == MYRSINE SEMI- SERRATA Wall. in Roxb. Fl. Ind. ed. Carey & Wall. 2:293. 1824. Celastrus semiarillata Turcz in Bull. Soc. Nat. Мосс. 361:599. 1863— not Celastrus (branches spiny). Celastrus sepiarius Dennst. Schlüs. Hort. Ind. Malab. 31. 1818, nom. nud. Celastrus serrulatus Roth, Nov. РІ. Sp. 155. 1821, p. р. = GYMNOSPORIA ROTHIANA Laws. in Hook. Fl. Brit. Ind. 1:620. 1875. Celastrus spicatus Vell. Fl. Flum. 92, f. 138. 1827 = QGOUANIA CORYLIFOLIA addi, in Mem. Soc. Ital. (Modena) 18, f. 304. 1820. Dno: spinifolius Larrañaga, Escritos D. А, Larrafiaga 2:96. 1923—not Celastrus (capsules bivalved). Celastrus spinosus Royle, Ill. Bot. Himal. 157. 1835 — GYMNOSPORIA ROYLEANA Laws. in Hook. Fl. Brit. Ind. 1:620. 1875 Celastrus striatus Thunb. Fl. Jap. 98. 1784 == EUONYMUS ALTUS (Thunb.) Sieb. Syn. Pl. Oecon. Jap. 49. 1830. Celastrus suaveolens Lévl. in Fedde, Rep. Sp. Nov. 13:263. 1914 — НЕХ sUAVE- OLENS (Lévl.) Loes. in Ber. Deut. Bot. Ges. 32:541. 1914. Celastrus tetramerus Standley in Contr. U. S. Nat. Herb. 23:679. 1923 == PHYLLANTHUS sp. Celastrus tristis Lévl. in Fedde, Rep. Sp. Nov. 13:263. 1914 == RHAMNUS NAPAL- ENSIS (Wall.) Laws. in Hook. Fl. Brit. Ind. 1:640. 1875. Celastrus uncinatus Ruiz & Pavon, Fl. Peruv. 3: 7, 2. 230, f. а. 1802 == MAYTENUS BOARIA Molina, Saggio Stor. Nat. Chile, ed. 1, 177. Celastrus variabilis Hemsl. in Jour. Linn. Soc. 23:124. 1886 — GYMNOSPORIA VARIABILIS Loes. in Engl. Bot. Jahrb. 29:446. 1900. Celastrus verticillata Roxb. Hort. Beng. 18. 1814, nom. nud.; Fl. Ind. ed. Carey 321. 1824 == PrrrosPORUM FLoRIBUNDUM Wight & Arn. Prod. Fl. Ind. Orient. 154. 1834. Celastrus verticillatus Ruiz & Pavon, Fl. Peruv. 3:6, #. 220, f. b. 1802 == May- TENUS VERTICILLATUS DC. Prod. 2:10. 1825 1955] HOU— REVISION OF THE GENUS CELASTRUS 291 Celastrus vitiensis (A. Gray) [incorrectly ascribed to Benth. & Hook. by] Drake, . Polyn. France, 30. 1893 == САТНА vitiensis А. Gray, Bot. Рћапегор., Wilkes U. S. Expl. Exped. 287, І. 23. 1854 == GYMNOSPORIA VITIENSIS (A. Gray) Seem. Fl. Vit. 40. 1865. Celastrus wallichianus Spreng. Syst. Veg. 5, Index 150. 1828 == GyMNospPorRIA WALLICHIANA Laws. in Hook. Fl. Brit. Ind. 1:621. 1875. Celastrus wallichianus Wall. in Roxb, Fl. Ind. ed. Carey & Wall. 2:400. 1824 == GYMNOSPORIA WALLICHIANA Laws. in Hook. loc. cit. 621. 1875. Celastrus wallichii G. Don, Gen. Syst. 2:8. 1832 == CELastrus LUCIDA Wall., поп Linn.—not Celastrus. Celastrus wightianus Wall. Cat. no, 4332. 1831 == PLEUROSTYLIA COCHIN- CHINENSIS Pierre, Fl. For. Cochinch, Fasc. 20, sub. Ё. 305 in text. 1894. Celastrus yunnanensis Lévl. Cat. Pl. Yun-Nan, 32. 1915 == PREMA PARVILIMBA P’ei in Mem. Sci. Soc. China 1:62. 1932; Rehd. in Jour. Arnold Arb, 15:324. 1934. Celastrus zeylanica Roth ех Roem. & Schult. Syst. Veg. 5:427. 1819 == ScuTia COMMERSONII Brongn. in Ann. Sci. Nat. I, 10:363. 1827 ENUMERATION OF THE SPECIES SUBGENUS Ll CELAsTRUs 14. vanioti (Lévl. tian З SERIES 1. PANICULATI d за = E (Oliver) War т. г рі Ша. 17. orbiculatus Thunb. p. pan niculetu. 18. rostbornianus Loes. js ње serratus (Blanco) 19. punctatus Thunb. Ding Hou 20. kusanoi Hayata It. хе ‘multiflorus (Roxb.) 21. hirsutus Comber 22. stylosus Wal 2. я ime on маг & Реггу 22a. ssp. stylosus 3. subspicatus Hook. f. 22b. ssp. glaber Ding Hou 4. пери A. Gray 23. aculeatus Merr. % таана Loes. 24. flagellaris Rupr. 6. angulatus Maxi 7. scandens. Linn. SuBGENUs Il. RACEMOCELASTRUS SERIES 2. AXILLARES 25. lenticellatus Lundell 8. monospermus Roxb. 26. racemosus (Reiss.) Loes. Ж личи Гоез. 27. panamensis Lundell ae # Benth. 28. pringlei Rose Ns собуй Rehd. & Wils. 29. vulcanicolus Donn. Smith 30. caseariifolius Lundell 13. ы, Ргаш 31. meridensis Pittier 292 [Vor. 42 ANNALS OF THE MISSOURI BOTANICAL GARDEN INDEX то EXsICCATAE talicized numerals refer to collectors’ number. . (sine numero) to unnumbered е parenthetical numerals refer to the EE of the species conserved in this revision Adi. J. Ж. е с 2865, а (7); Adams, J. W. & 1420, 21 d Adams, de Ws « E = Adams r4 ктг Adams, 1 Y. & R. Tash we p. . Wherry 2300 (7). 2889 (Bureau of Sci. ams, J. W. & E. Mak collector Alcasid, —, & С. Е. Едаћо 4609 (1Ъ). 9). Alims, —. 7): Апдегзоп, Т. 104 (22а); 107, 108, 109 Andrews, BD Ww uU ru CA 37252 (1c). ‚Куз. жж (7). m (7 9, 1842, 2223, 5600, Bennett, W. 5720, 6308, бие, 7023, 8682 (7). сага. Н. Е. 2101, 2630, 2855, 3120 (7 ns Bergman, S. 439 (17). эры сагабы р. 4756 (7). Bicknell, E. P. 3836 (7). Biswas, K. 2033 Bock, C. . v. Rosthorn 6 о зак 700 (6). Boswastore, —. 73 (22a) randis, 32 (1a) Brass, L. T 12999 (9). Breitung, A. J. да (7). Brooks, ор: 44, 544 (17). Визћ, В. Е. 118, 732 (7). Bush, Y. 00 (7). Calder, C. C. & M. S. Ramaswami 654 la). о С.Е. 11260 (3); 13036 (9); 15608 А М. С. ооо (29). Carter, Н. С. 466 (1 с). r oon 2976 (10). азе, У. Н. 255, 853, de 16652 (7). 9) Chen, P. E. 2447, 270. Chen, Y. 1000 (18). Cheng, C. ue 1573 (Dy. Cheo, T. Y. & Г. Yen 193 (17 Chiao, C. Y. 1720 (18); 1721 (11); 2116, 2126, 2395, 2612, 3016 (17); 18641 (22b); ай >). ћао, С. У. & C. S. Fan 458 (10); 767 11 Chien, $$ те о (18 hing, В. 1336, 1664 (23); 1677 (19); An (23); 2638 (16); 266I (19); 2700 (11); 2845 (22b); 3065 (15); 3165, 5461 (22b); 6057 (14); 7080 (22b); M ); 7566 (11); 22119 ts Ching, — & — ^ 580 (п). Chou, К.Н. rs 17 , 8460 (23 ее І. 5707 (18); 6448, 6468 (11). ow, Н. C. 11712 (10); 11742, 12013 (18). Chu, К. L. 3701 (18); 3761 (16). Chus, N. K. 40143, 40276 (8); 42950 16). Chun, М. К. & C. E. Too 44028 (20); 44165, 2. ие Chun, W. Y. 7 (8); 3503 d 3570 (18); 3724 (sy; 3884 (16); 4066 (6); 4427 (15); 4220 (16); 4831 (8); 4840 mih. 3 4043 I = (16); 5539 (23). 5. Chie 8050 ( eu H. a мт: (10); 493 4” 954, І 206 (10); 1437, 144 Ж 55 Өзі 1514 (23); 1815 ani И Clarke, C. В. 682с 2); 5615 (10); 12046 (222); 18678С (222); 23681В, 24857, 26339 (1а); 26701В ( 27040 (12); 27683 (8); 35758Е (12); 37040В (Di 41911C (10); 45152 (8); » J. & M. S. Clemens 3523 (2); 4581, "5084 (9); 5152, 5304 (2); 6253 9); 6606, 7230 (2); 27870 (9); 28701, 28701A (10); 20010 (9); 30493 (10); 1955] HOU— REVISION OF THE GENUS CELASTRUS 31827, 32052, 32120, 32860, 33182, 50253, 20065; se (2). Clemens, M. S 073 (2); 11680 (7); 43683 (3). Clokey, 1. W. 2470 (7) uming, H. Curran, H. Curtis, C. 404, 2005 (9). Daniel, H. 3295 (31). Darsteri, A. C. 1 David, A. 1783 (17). Davis, John. 100, 716, 720, 3709, 6368, 7344 (7) Day, M. A. 43,96 (7). Deam, Chas 15456, 18042, 18240, е А . 18331, 18810, 19320, 35238, 37014, 38141 (7). Degener, Otto 14798, 14940, 14073, 14078, 3 50, 15315, 15330, 15352, 15424, 4). Demaree, D. 0740, 6055, 11340, 11572 7 Deplanch, — 92 (3) Desiloa, F. . (8) Dickason, Е. С. 5845 (1a) Dicke. 5. 5. 27 (7); Dorgelo, J. D. 2200 (Іс). Dorsett, P. H. . J. Morse 610 (24); 1070, 7074, 7152, 7206 (17). Dougboo, —. 86225 (12). Drummond cr ВК. 21515, 21516, 21651, 21001 ia 132 (16). Dudgeon, У. & L. A. Kenoyer 237 (1a). unn, —. 44 (19); 2479, 24798, 3420 (23). Duthie, І. Е. 8211 (Па). Ессјег, К. С. 20 (7). Едаћо, С.Е. 000 (9); 2 (15). Edmondson, T. У. 3426, 3876 (7). Eggleston, W. У. & К. М. Wiegand 22503 Elizabeth, Е-Е 978 3903 (7). Elmer, A. D. E. II4II в, М. Кайа, Е.Н. PE (9). Enequist, J. 435 (7). 493 Engelmann, G. 1107 (7). Esquirol, 4 (16 Evans, I. H. N. 13247 (1c). Evers, R. r 143 ( Evers, В. A., С. N ds & F. F. Jones 541 Eyles, D .E. & M. S. Eyles $505 (7). Eyles, M. S. 8380 (7). P g 227 (18). SEX Xo i» "m. 205; (22Ь). Fang, У. P. ; 593 (16); 1202a (11); 1362 (2; M ien 3089 (11); 190 (22b); 210, I (10); 15211 (22b); 15221 (10); 15280 (225); 16419 О; 16624, 17317 (11); P 4 (18 Fang, W. P C Lu 3575 5 (16). Faurie, U. 4 (20); 1376 (10); 2502, баз, 2 “Gn: 8015 (19). Fay 08 (7 i T ч 43, "556, 615, 616, 863, 1158, 1320, 1411, oo (7). Fendler, A. 104 Feng, . 747 (225); 787, 1226 (11); 2425, 2470 (18) ; 2808 (1 6); 3047 (11); (18); (11); 3365 (22b); ере (10); 13360 (18); 13387, 13454, 4 (10); 13472, 13740 (8); 13765 (21); 13847 el, С. 502 (17); 2 738 (6); 927 (15); 2817 (6); 2827 (17). Fernald, M. L. й 2709 73; Еегпа!4, М. І. 0858, 0850, “ В. Гоп 12716, 13660, 13973, 14031 (7). Fester, С. $. 1288 (7). Ае а б, n. e 4739 (1a). Fogg, J. 5807, 8605, 7410, 8213, 8554 sos 12203, 12527, 13032, 14678, ie (7). Forbes Н. О. ger, 961, 080, 980a (10); 3820 (224). Forrest, G. 7812 (18); 8704 (16); 9306 (22b); ‚Доо? 10142; 11152 (11); 11173 (16); І (16); 15080 PAS r622 (18); (16); 16538. (11); I 21); pd (18); 21464 (16); 22235 (11); 22311 (16); Ге» 23230 Gus 26407 (8). Fosberg, F 15146 (7 Fox, R. B. yy (1b). 294 Friener, К. С. 11777 (2). Fung, Н. 00161 б А-551 (10). Galen, J. 480 (7). Gamble, J. S. 8780 (1a). с во: М. oe (7 2973 (7 2250, 3218 (7 (7). Greenman, J. М. 783, 837, 2409, 2510 Greenman; J. M., О. Е. Lansing, Jr. & В. A. Dixon 44 (7). Greenwood, W. 900 fex G . 605 (12); 666, 1992 (8). Groff, G. У. 6132 (Іс). Grood, J. H. 799 (7). Hallier, —. 462 (10). Hance, Н. Е. 370, 22191 (10). H^ e ноу Н. 132 (6); 583, 810, 305, 2008 (18); 5692 (6); 1307 (16). Haneef, Md. & Md. Jan 2446 (9). Harbison, T. с ud (7 arsukh, —. 27. (1a Hayden, A. 0500, 10070 (7) Hayward, H. E. 616 (7) Heiser, Charles B., Jr. 1801 (7). Heller, A. A. & E. G. Halbach 522 (7). vue 9. II 3: . ба 11); 1893 (20); 2084 (6): 2837 (15); 3115 (18); 3241 (10); 3405 (6); 3405, 3405А, 3856 (10); 3883 (6); 5550 (22b); 5040, 5734 (18); 5887 15); (6); comi 2 ib); [Vor. 42 ANNALS OF THE MISSOURI BOTANICAL GARDEN Hermann, F. J. aes Него, J. 26 (17 442, 2670 (6); 2791, 2795 (17); 2932, 3010, 3097 CRANE P. 443 (1a). Hinckley, L. С. 1108 (7). He G. B. 290 , 3506, 3574, 3717, 76, 7230, Er 9020 (28). Hitchcock, A. 5. бо (7). Ho, H. T. 60086 (8). Hooper, Р. 38801, 39537 (1a). а 297 (1 8). Hopkinso. Horne, + 5 (9). Horr, W. H. ERL McGregor (7). E493 Horsey, R. E. 290, 1283, HS 1545, 1581, 1748, 1754, гері, 2380 ow, Е. ‘wire 70573 RIPE) У К. Сћип 70273 (8). Hsiung, У. К. 5516 (23); 5655, 5701 (16); n ad Hu, H. H. 784, 88 (23); 1110, 1585 наб E ow 380 ‘UD. Ha z3 2 Hunnewell, Е. W. ` 568 + (7). Hurusawa, L 10650 (17). Ip, N. K. 41 (22b). Jack, J. G. 3857 (7 Janet, —. A79, д, 4349, 1639 (17). Jenkins, P. Н. & С. Stebbins, Jr. 170 1867 (7). қ 11458, ой, 16251 (7). Junghuhn, —. 27, ros (10). ева К. & S. Hatusima 12102 (9). . 4 Kanjilal, Upendranath 3801 (8); 4076 n зе Каз 347 (1с). Kawakami, T. & U. Mori 3145 (20). Keith, —. 213 ign 1а). КҺап, В. I (еј 36, 250 (1а). Kiener, W. 17023 (7). T 1955] HOU—REVISION OF THE GENUS CELASTRUS Kimura, Y. 307 + 34 (8). Dr. Rink’ collector s.n. (1а); 178, 233 (8); 454 479 "eH 5690, 6928 (9). 30 ат (9); 2907 (1b). 0089, 51463 (8); 55544 (225); EIE Koelz, W. ent 4377 ( IL la). Борга, 3, 196328 (1b); 273728, 038 (1c); 301788 (1b); 424308 Ki ЗХ Jera Е: Korthals, —. 747 (10). Kostermans, A. 441, 1027 (1с). 7567. 9 қ (1с). Kurz, S. 65 (1a); 171 (22a); 182 (8); 1926 ба 1027 (1c). се, J. Н. 2220 (8); 3130, 3280 (Па); ). 26227 (10); 2 (8); (2 0); 26624 = aon, 28 558, E. 28763 SP. ; Гау WA. 4 (7). Leavenworth, W. С. & Н. Hoogstraal 1023 5, 3060 (18); 3168 (225); are (1); Bom (10 eeuwen, W. ыр А у. 11008 (2). Lehmann, — 30 La, С. I 129 pe dos (20); 440 (10). Levine, C. О. 180, 240, 441, 484, 1236 (CCC nos., ag ICT HI Li МК: ré (2293: 2807, 2947 (1c). ТЕТЕ à 173. Liang, H. Y. l Sas (8); 63042 (1c); 63419 (8); өй», (1с); 65203 (8). Licent, E. 2 (17). Liebmann, —. 14871, 14872, 14873, ue 14875, 5, 148750 198756 (26). in, 3 Lig 5” 95,40. (16); 1104 (225); 1246, 40 (16); 5 1607, 1664, 2197 (17). 85 (17). пам 3. L. 68 би. T57 (223); 295 Lay 9 204 1176 (17). Liu, У. & С. №. Wang 82047 (8). Lu r Xm Bi (16); 1184 (22b); 1319 (18); 1423 (16); 1714 (225); 2231 11). dei С. С. 480 ( (7). Long, B. 3408, 5700, 8273, 9708, 13088, 8, 412, 15016, 16886, 352, 402 3, 66044, 67273 (7). и с. L. 6307 (26). MacElwee, AC 2072 (ту: Mackenzie, K. K. 4764 (7). acoun, —. E d. p E. 14 (16); 22 (6); 81 (18); I 16); 254 (18); 385, 303 (6); о ced (18); 3014, 6224 (6); 7256 Маше, мане Victorin 16280, 18534 (7). Marie-Victorin & Rolland-Germain 45524, же 14 rie-Victorin, Rolland-Germain & М. Ray- . 0529, 92 16369 ( McClur re, F. А. 3544, 7938 do: no (1c McClure, F. A. & Chan Kwai Shang 4 805 (7). McCree, H., P» ЯТІ. (7). McGregor, В. С ue (1b) McGregor, R. L. 645 (7). McVaugh, R. 220, 605, 759 (7); 10246, 10308, 13 Meebold, А. i (Та); oo (22a). Merrill, (1b); 9311 (1с); 10070, der (0) 3; Merrill, E. D. et al. 15640 (29) 1722 (6). W.H. 120, 300, 3500 (7). 296 Mizushima, M. 1029 db c Moyle, J. Muenscher, Лу. C. x А. R Bechtel 264 7. Muenacher, У. C. & R. T. Clausen 4680 Mukherjee. Revoti Mohon 32 (1a). Mumbauer, J. В. 350 (7). Nakamura, Т. 3618 (19). Narayanaswami, У. 105 (1а). Nath, Amar 24 (1a). Norton, J. В. бо? (7); 1346 (20). Ohwi, J. 436 (17). Ohwi, J. & Okamoto, K. 504, 506 (17). ee | к. IÓI (19); 162 (17). Ome, Т а 500 (19). Otanes, E 17985 (1b). Ownbey, M. 1570 (7). Painter, J. H. 1194 (7). Palmer, E. 106 (28). Palmer, E. |. 712, 772, 13903, 16145, 18923, 21238, 22124, 26057, 27583, 28527, 30917, 933, бо 15237, 31134, 32 ~ Uv S A ~ 4170 53, 36000, 37131, 41926-A, р,45399, 45570, 45645, 49016 (7) mel, pa 26 икону ~ a B A. Bane & 5. В. Fitz 1647 (7). Pammel, L. H., Н. E. Pammel & P. $. Mc- Ал rs.) 3 (4). 14, ио, "2671, 11963 (7). 7). PW. fna How 16487 (7). Peters, J. E '2 Phelps, О. P. 647 tae (7). MN. L: 505 Io 2794 (1c). Poilane, —. 28714 (10). Poole, 5 Pradham, — der (и); Dr. Prain's collector (12); dieu (22а); Жар Pratt, 5 (7). 0, ded (1а); 289 2 (8). [Vor. 42 ANNALS OF THE MISSOURI BOTANICAL GARDEN Prazer, J.C. 54, 167 (1а); 236 (1с); 273 Pretz, H. W. 11403 (7) Pringle, C. Ga 6842, 6008 (28); 8133 Punt: W. 944, 945 (17); 946, 946° вы А. 7094 (26); 7370 (25); oed 5026 (26). Кат, В. 300 (1а). Ramaswami, M. S. 1400 (1a). Ramos, M. & Deroy, B. 22511 (1b). Ramos, x: & С. Едаћо 3 9). 9, 346, 1813, 42711 (15). Ramos, К. Randolph, L.F.& Е. В. Randolph 35(7). ге К. 1588 (1а). Капор, Н. М. 7426, 1 s e Redfield, L Н. тобі (7). Reko, B. P 19 (29) Richards, C. D. 2453 (7). Ridley, 404, 5237, 7127, 6226, 13538 (9) Robinson, B. L 550, 798, еді (7). 11617 25 b). Robbins, G. T. 22 R 202 024 "бу; 1744 (10); Rodger, А. 1 Rodriguez, G. 202 5 а а). Rogers, C. Gilbert 804 (1c). Rose, D. K. планине C. О, 315 (7). — A * з ne 15565 (22b); ). E ( die 5 1453 (7). Rydberg, P. —. Ralphlmer 1237 (7). Ruth, А. 42, 356, 360 (7). Samnasena, М. S. 34 (1a). Santapau, H. 87.6, 47-7, 87.8, 87.0, 515, 570, 603, 1229, 1235, 1967, 1968, 1975, 2100, 4102, 8710, 12707, 12708, 14258 а). Sawada, T. 2101 (1 d Schaeffer, R. L., Jr. 3, 5861, x 8770, 8776, m Жа. 9979, 14728, 15100, 20521, 25023, 20418, Baer. DN 32768, 33851, 37752, 40474 Schneider, C. 644, 646, 647, 1027 (11); 1131, 1375 (18); 1084, 2839 (11); 2683 (18); 2874, 4112, 4130 (11). Seymour, Е. С. 1268 (7). 19551 HOU—REVISION OF THE GENUS CELASTRUS Shanks, В. E., A. J. Sharp & E. Clebach 50. vae J, L I34 (7). Sharp, A. ]., E. Clebach & А. Clebach 6309 (7 бай А. £ S. Fairchild & E. Clebach 7) 1604 (7) Shiota, . 3899, #52 0643, 9766 (17). Silvestri, = 133 6 Sin, S. 5. 11835 ee 11889 (23). a ]. 39284 (9). Singh, А. 25 (1a). Singh, U. 43 wok mith, — & — 964 (22a). Smith, A. C. p F 5827, 6085 (4). Smith, H. 959, 5745, 6154, 7697 (17); 72808 (16); 13326 (11); 13448 (6). Зтић, Н. Н. 7142, 7785 (7). Smith, U. C. 2 S Standley, P. C. 40524 (7); 71349 (26). Standley, P. C. & J. Valerio 52024 (26). Steenis, С. С. G. bw Stevens, G. W Steward, A. N. 1234, 5470 (16). teward, A. N. . Cheo 87, 306 22b); 429 (18) Steward, A. М. e£ al. 97, 541 (6) Stewart, C. C Ж.Т Stewart, R. R We p Steyermark, J. 7572, 12131, 12875, 13583, E 20825, 22055, 23533, 24382, 24480, 24563, 24670, 24874, 25065, 26398 (7); 57015 (26). Stocks, J. Е. 231 (1a). Stone, W. gar, 3921, 10200, 10682, 10847, 21%, 13375 (7 Strachey, К. & J. ‚ Winterbostom 2 (1a). Strathdickie, L. 1124 (3). tresemann, E. 9o (9). Sulit, M. D. 3188, 3421, 244) 10086 (9). Sun, С.І. 781, 1247 (16 Sdn: $. С. 1180 (22b); 108 (17); 1180 (1 1). poer S. 11162 (10); 444020, 455000, 9013, 486024, 493027 (17). о Т. 6195 (20). Swartley, J.C. 425 (7). с т. М; 8 (10); 720 (8); 724 аз; 759 (16); 1041 (23); 1477 (10); 1792 (8); 1798, 1951 (10); 2087 (8); K (10 Та Р.С & с. M. Feng 5238 (22b). 297 Tai, Е. C. е? al. 5142 (11). Tai, Г. Y. 204, 1071 (11); 1113, 1120 16). ( Jaw С. CE > > we 2936 (23). Tang, S.G. 1677 (19); m I OP: 7085, 13657 (19); 15609 Tang, T (17). Tanger, —. 4674 (7). Taquet, T. 632, 2723 (24). Tatnall, R. R. 2098, 3633, 4770, 4967, gU 499. . Taylor, Т. M. С. e£ al. 1239 (7). ТЕУ. To K.P. & К. С. ione 2784 (20). о, К. P. её al. 131, 442 (20); 473 (8); 631 (23); 880 (10). Tolstead, W. І. 41887 (7). Toppin, S. M. 4292, 6154 (10). Toroes, R. S. (9). (16); 5082 8); 21); 60420 (22b); 60450 (8); 60409 (21); бо781 (18); боді b); о (8); 61753 (18); 62295 pe SEA в (225); 63050 de 6311 16566 SAM 20002 (23 404 (1 ); 27666 (6 ); (225); 28843 (23); Ж (20); ЭС n У. T. & —. Fung 241 (10). [Vor. 42 298 ANNALS OF THE MISSOURI BOTANICAL GARDEN Tsang, W . C. Wong 2061 (10); 2081 B Se (20). Teiang, Y. 720 (8); 1002 (10); f (16); 1393 (20); 1419 (16); 1044, 6832 (22b) ; 7499 (18); 3053 (6); 9354, = SENT 20788. (16); 10460 (23). ‚С 8, 20796, 21026 (20); PT 5. s (23). Teo, СКОК У. re 2040 (10). Tsui, T. M. a» 348 (8); 391 (10); 542 (18); d io > Turner, —. Uno, K. 21823 (24). Valeton, Т. 102 (10). Versteegh, —. 1643 (9). N. H. de 682, 2338 (15). ХУ admond, S. С. 2. ox Walter, М. & С. М. В қанық 870 (10). Wang, С. SEM Or bee Eg 35222, 35893, 36064, 30007, 40415 (8). 8 9 (11); 62990, 63709 (16); 64147, 63203 (18); 68718 (1a); 857 (8); 72996 5 оё; 4 o 8). vor Ж Е "Dg (11); 21670 (6); 22451 (16); Peg (11); 22633 (6); 22643 (11 .М. 3101 (10). Wang, Т. Н. 114 (18); 299, 399 (16). Wang, T. P. 176 (17). Wang, У. H. 8096 (10). о“ К. 3056 (8); 3060 (23); 3196 i Е.К. 9362 (21); 9455 (12). т. ЈЕ ІРЕК 3030 -Т?У, Watkins, Il. 51 (7). 76 (1а). White, C. Т. 1043, 12517 (2). Whitford, Н. М. 87, 414 (1b). Whitney, Е. С. 2311 (7). VR K. M. ауе 6794, aee! (7). ‚©: Molin Wilson, Е, Н. 181 с ЕР au; od (18); 362 (15); 363 (16) ; 364, 364a, b (15); 1106, 1148 (11); 1175 (18); 1176 (14); 1184 (22b); 1289 (15); 1302, 1519, 2305, 2305A (16); 2 (15); 2307 (10); 2308 (17); 2309 (18); 2310 (11); 2311 (18); 2312 (14) ; 22154 (16); 3324 (10); 3325, 4117, 4122, 4157 (11); ps 4195 (18); 4317 (11); 4781 (18); 4782 (11); 6121, 6304 (19); 6627, 6007, Рат) 4500 (10); 8450 (24); 8611, 8801, 9203 ; (20); 10425 (24); 10461 в: ум M ed. II2I4 (20 зн nckel, W. Е. 928, 088, 16338, 16828 Wood, C. E Je pu 338 3784 (7). Wray, Б, Jr 1751 Wright, с: 9 (8); rd (19); 9I (10); 0). 92 (8); бт (1 amada, K. 1215 (20). een moto, У. 602 (19) Yamazala, Т » 44-A (17) York, H. H. 155 (7). Yu, Т.Т. 445 (22b); 750, 940, 951 (11); 1756 (18); 5522 (11); e rA не: 5845, 7204, 7978, 8185 (18); (1c); 17204 (21); 17405 (8); 17803 (18): Мрз, pu (213. ee T. с . H. Welch 10669 Zinck, N. H. 740 (7). 1955] HOU—REVISION OF THE GENUS CELASTRUS 299 INDEX TO NAMES Roman type nies accepted, ратын names; italics indicates synonyms; bold face indicates new names or combinatio: Acanthothamnus aphyllus _________ 285 cunningbamii 286 Alsodeia glabra 231 V регеа — s 7 "ВБ Сара ЖЫРЫ -- 244 dependens 231 mon a 249 dilatatus - 286 vitiensis ____ NT discolor im MES Ceanothus paniculatus |... 230 disperma ____ 286 Celas 227 diversifolia ___ 286 aculeatus ___ 274 dubia 286 acuminatus _ 285 elevativenus 267 pa opbylla _ 235 emarginatus _ 286 Ине. 285 esquirolianus _ 286 4! upra 285 esquirolii = AME angulatus 239 euonymoidea = AM annamensis 244 euphlebiphyllus 231 apbyllus ___ 285 fasciculatus 287 и 246 feddei 287 approximata ____ 250 ТОНИНИ —— — X 287 ifolius __ 285 flagellaris ___ 276 articulata floribundus __ 287 var wo 261 fournieri 287 ar. orbiculata 261 geminiflorus 267 aridum . — 260 gemmata ___ 258 r. pubescens 261 [4 258 var. р сирі : 261 еол —— 253 var. 266 glau -- ЖЕ аг. ste ето ee ЫНЫ 261 gracillimus 267 attenuatus ________ 285 grenadensis 285 НИНА — 236 aenkea 287 о Е 250 bamelii - M7 benthamii |... 244 beteropbyllus ____ 287 bilocularis __ 285 beyneana 287 bivalvis ___ 285 hindsii 249 и 285 var. henryi 249 bodinieri __. 285 270 bullatus ____ 240 hookeri 254 buxifolia ____ 285 boo 274 cantonensis ___ 249 hy poglaucus 257 caseariifolius ___ 282 hypoleucus 257 о — 265, 286 f. В. argutior 257 cbampionii — 244, 246 f. a. genuina 257 cbiapensis 28 berula 273 о 286 ilicifolius 287 ------.-.......... 276 І, 2 6 1 circumcissus ______ 286 зазорі о зат. 280 14 айақ pa clemacanthus 276 jebolensis ее ест 25) іш __ 286 jodimii ан crassifolia — 273 Riusianus 266 tus 286 kouytcbensis 287 crenulatus __ 286 usanoi 268 crispulus 266 lancifolia 261 pu cuneifolius 286 || laotica ids 300 latifolius leiocarpus enticellatus eptopus iebmanni inearis var. madagascariensis- Longipes TER ——27%, sis. novoguineensis vide Amo сонын — — — 2 22 ошон . . — 1 octogonus . 5 лы E ПОРОМ о f.m f. persis ы. vat. ewreo-érilleja 0 02 atus orbi [Vor. 42 ANNALS OF THE MISSOURI BOTANICAL GARDEN var. punctatus стан un ene eer ха вирі муу. var Вийти - === — LS o ssp. пене )Ding Hou мрорядісційния о ssp. serratus з (Blanco) Ding Hou tifolius _____ 8. сос жаг аа E elaeodendroides 19551 HOU—REVISION OF THE GENUS CELASTRUS sepiarius serrulatus siltebecanus spicatus ersicolor _ verticillatus з ze BN. si ed gelonioides Colubrina ferruginosa Diosma s serrata emarginata heyneana __ 301 montana 286, 288, 289 neglecta 285, 288 neglecta 272 ata 289 rothiana 290 royleana 290 fa 290 igyna 288 variabilis 290 itiensis 286, 291 wallichiana 286, 289, 291 lex crenat 285 cuneifolia 287 acrocarpa 290 purpurea 285 scand 286 su 290 Iodina лавре 289 Kurrim 289 May ан" сезне 285 агіа 285, 288, 290 uxifolius 289 cochlearifolius 289 confertus 286 unninghamii 286 dispermus 286 elaeodendroides 289 fournieri 287 ineatus __ 287 macrocarpus 288 magellanicus 288 den 284 orbicularis 286, 288, 289 ovata -- 289 quadrangulatus . 289 асетоѕи 279 286, 289 truncatus 287 verticillatus ___ 290 williamsii 4 Microtropis bivalvis 285 inosa 288 Myrsine semiserrata 286, 290 rixa japonica ________ _ 286, 287, 289 Perrottetia д бес 285 Phyllanthus sp. 290 Pittosporum floribu 22. 288, 290 Pleurostylia cochinchinensis _ рді Prema parvilimba __ 291 Prunus myrtifolia — 288 Rhamnu crena 00 286, 287 esquirolii 288 napalensis 55870 racemosus 287 Sabi parviflora var. nitidissima —— 287 [Vor. 42, 1955] 302 ANNALS OF THE MISSOURI BOTANICAL GARDEN yunnanensis 288 Sempervirentes 242 Saurauja vaniota 256 ondsia californica DE 27525575. 288 ee ee 287 Trichilia ЖЫ... 287 Scuti mmerso: 291 Tripterygium wilfordii и 286 Sentia 1 g сіна m 230 Wimmeria mexicana 22288 Semar 227 MISSOURI BOTANICAL GARDEN Cannons W. Donor, | ce Еовкат Е Woonsox, Tee г Senior Taxonomist Volume XLII Number 4 Annals of the Missouri Botanical Garden NOVEMBER, 1955 The Food of a Hiudu Village of North India . (Reprinted from Bulletin No. 2, Bureau Ри _ ава Pranoni Res , United Provin * Charlotte. Vial Wiser, PhB, M.Sc. 303-412 is ori тезда | PUBLISHED сбанткіл а AT GALESBURG. | : вт THE BOARD inu ех ТНЕ MISSOURI BOTANICAL CARDEN — зт. | Annals of the Missouri Botanical Garden А Quarterly Journal containing Scientific Contributions from the Missouri Botanical Garden and the Henry Shaw School of Botany of Washington University in affiliation with the Missouri Botancal Garden. Information The ANNALS OF THE MISSOURI BOTANICAL- GARDEN appears four times during the calendar year: =» May, September, and November. Four numbers constitute a volum Subscription Price ______ $10.00 per volume Single Numbers — __ 2.50 each Contents of previous issues of the ANNALS OF THE MISSOURI BOTANICAL GARDEN are listed in the Agricultural Index, published by the H. W. Wilson Company. Annals of the Missouri Botanical Garden Vol. 42 NOVEMBER, 1955 No. 4 "THE FOODS OF AN INDIAN VILLAGE OF NORTH INDIA" By CHARLOTTE ViALL WISER Men Wiser's monograph originally appeared as Bulletin No. 2 б. T Bureau r Ж апа nomic Research of the z. ес. d inces, paliar in Allahaba d, U. P., India, 917 Е was s n data wires by Mrs. and her hus ze а den серой. socio ан: pes en they lived in es villa age of Karimpur at iong internals аи 1931. She eventually worked it us as a master's thesis in the рю nt of Foods and Маша of Cornell ређа Ithaca New York. Since copies of it are now dicm Мене та obtain, either in Indi the United елен we have obtained Permission 1 to reprint it from е Wiser, from the аи У к. апі from Cornell University. A single copy ati it ыер to the I of the Missouri Botanical Garden has been "feda on loan duri a good part of = ince it was received. an accurate, detailed, and pe — account of what ми nts w Е жайда for which purposes іп а S Indian village, it is ge its way to becoming a tinis in ы field of Ethno-botany.—E. Anderson.] ote aid to bibliographers the oigo, title-pages are here pud d in i Throughout the he ined effort has been made to reproduce the text exactly, e o the variant spelling Nothing has been changed from de нее P except the жешсе te to seed numbers The half-title is as follows: BULLETIN NO. 2 THE Foops oF А HINDU VILLAGE OF NORTH INDIA The title-page bears the following: Bureau of Statistics and Economic Research, United Provinces THE Foops oF А HiNDU VILLAGE OF NORTH INDIA CHARLOTTE VIALL WISER, РЋВ., M.Sc., July, 1936 ALLAHABAD: Superintendent, Printing and Stationery, United Provinces 1937 TABLE OF CONTENTS INTRODUCTION os ге r4 I—DESCRIPTION OF THE VILLAGE OF KARIMPUR AND ITS PEOPLE II—Foop SourcEs .. i E A—VARIETY OF FOODS MADE POSSIBLE BY CLIMATE AND SOIL . B—THE crops (1) CEREALS (2) PULSES ОК LEGUMES (3) VEGETABLES .. 59 E (4) Fruits ex i c C—ANIMAL PRODUCTS D—Srices —Foop COSTS IN кла. Ш—ЕКОМ FIELD TO SERVING TRAY i ARE OF FOODS AT HARVEST TIME B— STORAGE AND PRESERVATION OF FOODS C—EQUIPMENT D—DIsHEs PREPARED FROM (1) СЕКЕА15 Es ча (2) PULSES чом еі (3) VEGETABLES .. a (4) “GREENS” се 52% vx (5) Екџтт5 қ 92% 74 (6) ANIMAL PRODUCTS С. E—FEsTIVAL DISHES NER IV—Foop PRACTICES —Tasoos .. B—NUTRITION PRACTICES V— VALUE OF FOODS NOW IN USE : A— JAIL DIETARIES OF THE Раса B—THE work or McCarrison C—Foops AVAILABLE ІМ KARIMPUR .. D—A STUDY OF DIETARIES OF FOUR KARIMPUR FAMILIES VI—CONCLUSION .. А-- кийн. is В—Тне PRoBLEM (SUMMARY) + + . о I—INTRODUCTION We moved to the village of Karimpur in the autumn of 1925. Му husband and I had found during our years in India that if we were to understand the people of the villages, we must not be content to visit them, but we must live among them. After a study of the villages in the district, we chose Karimpur for our village home. Uninvited, we moved our tents and belongings into a mango grove on the edge of the village. In our ignorance we thought that our village neigh- bours would be glad to see us. But they showed no inclination to welcome white strangers, and were decidedly aloof until we had proven ourselves harmless. When they did finally accept us, they took our residence among them seriously, and saw no reason why we should ever leave. We thought, in the beginning, that a few months would give us what we ought to know of custom, activity, and thought. But when the hot season of 1926 arrived and we were obliged to retreat to a bungalow in the town of Mainpuri, five miles away, we had just begun to appreciate our ignorance. We commuted to the village from the town during the summer, and returned to live in it as soon as the heat was bearable in tents. We repeated this migration annually until the spring of 1931, when we left for America. My intention was that I should act as the good will of the study. With one son five, and another not yet two, I felt that I had all that I could undertake. But the village women, confined to their courtyards, could not come to me. They kept calling me to them. And as I saw more of them and their share in family activities, I realized that to understand and interpret the life of the village, we must know more about the women. They are of much more importance, eco- пописа у, than most students of Indian villages have given them credit for being. Observations in this field led me into further interest in their share in the care and preparation of foods Meanwhile, the increasing number of calls on us for medical help impressed on us the futility of trying to cure ills while the causes of those ills went unques- tioned. This led me into a study of village food from the point of view of health. I knew very little about food. But I had to acknowledge that if I wanted to understand and help Indian village women I must learn more about food, par- ticularly their food. And out of this need grew the present study. I learned all that I could about the foods procurable and the foods used, of food preparation and food customs. This material I carried to America, and with the help of specialists in the fields of nutrition and biological chemistry, 1 attempted to evaluate the foods eaten in our village, and to study the effect of customary methods of preparation on food values. The work was supervised by Miss Helen Monsch, head of the Department of Foods and Nutrition at Cornell University. (303) [Vor. 42 304 ANNALS OF THE MISSOURI BOTANICAL GARDEN Miss Monsch and other members of the staff, especially Miss Mary Henry and Miss Marion Pfund, helped in the study of the chemistry of foods, and in securing standard tables of food values. Others who made the work possible were Dr. J. B. Sumner, Professor of Biological Chemistry, and Dr. L. A. Maynard and Dr. C. M. McCay, both of whom have much to bring to the field of human nutrition from their researches in animal nutrition. I—A DESCRIPTION OF KARIMPUR AND ITS PEOPLE Karimpur, the village of our study, is in the northern part of India, in the much fought over Ganges plain. This section of the country is known, under British rule, as the United Provinces of Agra and Oudh. The latitude of Karimpur is 27" 9" North, about that of Palm Beach, Florida. The climate, however, is more continental than that of Palm Beach, and might better be compared with the same latitude in Mexico. The winters are delightful. In October, the heat relaxes, and the weather is progressively cooler until January when the mean temperature is 48.7° Е. For us these are the months of refreshment. But for village families with lives regulated for hot seasons and no provision for cold, January and Feb- ruary are months of misery. In March the hot winds begin to blow, and from then until June the temperature rises steadily to а mean maximum of 112? Е. The rainy season begins in July: over 60 per cent of the annual rainfall comes in July and August. The temperature drops only slightly but the rain brings relief to men and fields. September, at the end of the rains, with its heat and humidity, unrelieved by rain, is the most trying month for all of us. It is made endurable by the thought that the siege of heat cannot last much longer. arimpur is in an area which has played an important part in the history of India. Not far away are Капай}, capital of the Harsha dynasty (a.D. 604), and Kampiyala, the stronghold of the Panchalas, heroes of the Indian epic, Mahabharata (before 400 B.c.). Sixty-five miles to the south-west is Agra, with its Taj Mahal, and great fort of the Moghuls. Agra is the nearest city. Cawnpore, next nearest, is 110 miles south. As yet, neither city has made its influence felt in Karimpur. There are two market towns which link the village with the outside world. One is Kuraoli, 716 miles to the north, with a population of 4,717. The other is Mainpuri, 544 miles south. The population of Mainpuri is 15,599. Farmers of Karimpur take their surplus produce to either of these market towns to sell, and while on such an expedition, they may stop to buy clothing or food not available in the village. But they fear the clever merchants of the town, and avoid dealing with them as far as possible. They prefer to confine their activities to the village, with the minimum of outside con- tacts. Mainpuri is the official headquarters of the district. This necessitates trips to the Mainpuri courts to settle major disputes, chiefly those over land. But such trips are costly and confusing to farmers. And they prefer, if possible, to settle their disagreements at home with the advice of their own village council, 19551 WISER—FOODS OF A NORTH INDIA VILLAGE 305 This aloofness has protected the Hindu culture of Karimpur from inroads of changing dynasties—an important reason why it is a good field for our study of Hindu foods. It shares this advantage with other villages of the district. But it has additional points in its favour. Mainpuri District covers 1,675 square miles and includes 1,388 inhabited towns and villages. Two hundred and five of these have a populaton of over 1,000, and are too large to be representative. Of the 1,183 under 1,000, many are very small and confined to few castes, and often to only one or two industries. Karimpur, with its population of 754, presents a variety of castes (seventeen), as well as two groups or castes of outcastes, four groups of Muhammadans, and several families of Christians. It is dominated by Brahmans, which strengthens its barriers against foreign ideas. Like most of North India, farming is its outstanding occupation, with the subsidiary industries needed in a self-supporting community. The dusty, mango-shaded highway from Mainpuri to Karimpur, winds north- ward through a patchwork of small, oddly shaped farm plots separated from one another by low mounds of earth or narrow irrigation ditches. Sometimes there is an open stretch of uncultivated land which is too alkaline for crops, but which puts forth a stubby growth sufficient for meagre grazing. The animals must keep moving and searching if they are to keep alive. One is impressed by the leanness of the cows, and the thinness of the small boys who herd them. There are no houses on the farm plots. But there are hamlets, not much more than a mile apart, where the houses of farmers are huddled together. They are at a distance from the highway and easily passed unnoticed, with their monotonous gray walls. If one were to follow this highway a few miles further north, he would come to the old Grand Trunk Road travelled by Kipling's Kim. Five and a half miles out from Mainpuri is the village of Karimpur. There are two small sections of the village directly on the highway, while the village proper stands back 200 yards to the east. The first of the roadside sections is Chamar Nagariya, home of leather-workers. The leather-workers are counted as outcastes, but not as untouchables. All of the leather-workers' families save one, live in this section. ЈЕ consists of just two enclosures side by side, with a high mud wall extending across the front of both of them. Along two sides of each enclosure are small rooms, one room to a family. The men who live here wear scant loin wrappings, and no shirt except in cold weather. They are dark-skinned. Their legs are long and spindling, and their knees and ankles bulge conspicuously. Their ribs and shoulder blades stand out, and their arms seem disproportionately long, like their legs. Their faces are deeply lined. Their teeth are ground down, and dark. The women's bodies are covered by long skirts, badly torn, but still so full that they are a protection. Over their heads they wear stained, grayish scarves, and some have short vests. One or two have smoother lines than the men, but the others have the same long thin arms and hard, bony hands. The boys and girls are dressed like their elders, and look like them except for their fuller faces, fewer wrinkles and better teeth. The babies аге [Vor. 42 306 ANNALS OF THE MISSOURI BOTANICAL GARDEN plump and laughing. They go unclad when it is warm, and wear short vests when it is cold. The other section by the roadside is Kachhi ka Nagla, hamlet of vegetable growers. They are true farmers by right of caste, along with the family of rice- growers which lives in the section with them. Although poor and in debt to Brahmans, they are independent, which may account for their living apart from the rest of the village. Among them we have the nearest approach to the joint- family that we find in the village. High mud walls enclose each joint-family, with its animals, its implements and its stores of grain—all of its possessions except land. The men, dressed in loin cloths, are short and slight but not as gaunt as the leather- workers. Their dark skin fits more smoothly over bone and muscle. One of their elders might be mistaken for Gandhi, especially since a cataract operation has made it necessary for him to wear spectacles. The women are small and thin, and their babies are plump and placid. The older children are copies of their parents. As one turns east and bumps over the remnants of irrigation ditches which cut across the dirt road, to the village proper, he sees a long line of almost unbroken mud walls. He can guess where one man’s wall ends and his neighbor’s begins, only by the differences in height and smoothness of different portions of the wall. At the north end and at the south of this long wall there is a cluster of small, low- walled enclosures, each cut off from contact with the main wall by a lane. These are occupied by two castes of outcastes. Those living on the southern end are mat-makers, who make their living as farmhands. Their wives serve as midwives. The men are small and most of them are as thin as the leather-workers. But the arms and shoulders of the young women curve smoothly. And their skirts do not hang from protruding hip bones, like the leather-worker women, but from plump hips and abdomens. The outcastes at the northern end are the village scavengers, and therefore untouchables. They have recently been baptized as Christians, but are still regarded as untouchables by the villagers. Here too, the men are thin, with three exceptions. The younger women are attractively smooth-skinned, while the middle-aged and older women look tired, and their dark skin is drawn tightly over broad cheek bones, and hangs loose and dry between sharp elbow and wrist. In this group of untouchables one finds the darkest skins, the flattest noses, and the broadest faces in the village—evidences of Dravidian heritage. Between the two outcaste communities there extends the long line of walls of caste enclosures. And beyond these enclosures are more like them, with the same high walls and small doorways. Аз far as possible they are built side to side and back to back, the only break being an occasional narrow lane or footpath leading to enclosures deeper in the village. If one is a stranger, he may follow any of the footpaths through the village and see only high gray barriers and blank doorways leading to stables. The stable is the entry-way to every enclosure. But if he is a friend, he sees the family enclosures overflowing into the paths. Children, puppies and goats tumble under his feet. Carpenters and potters carry on their work in the lanes before their doorways. And serving women salam him as they pass by 1955] WISER—FOODS OF A NORTH INDIA VILLAGE 307 with head scarves drawn down over their faces. Мо matter how good a friend he may be, he never enters an enclosure other than his own. If it is necessary for а carpenter to repair a door frame or a beam, a man of the house accompanies him inside and remains until the work is finished. Women of serving castes are obliged to leave their own homes to work, but those of bigher castes are jealously hidden within high-walled courtyards behind high-walled stables. The families of each caste live close together, preferably in adjoining en- closures. The Brahmans (188) occupy the whole eastern section of the village. They are of priest-class by birth, but in practice they are farmers. Below them in the caste scale, in the order of their rank, are: bards (15), accountants (6), gold- smiths (10), flower growers (17), vegetable growers (152), rice growers (6), carpenters (42), barbers (2), water bearers (83), shepherds (26), grain parchers (10), seamsters (21), potters (9), tradesmen (14), oil pressers (10). In addition there are the outastes—washermen (6), mat-makers (28), and leather-workers (22). There are also the Christian sweepers (35); and four groups of M madans—faquirs or beggars (23), bangle sellers (10), cotton carders (9), and dancing girls (3). The numbers given include all members of each caste, not just the members who are actually engaged in the caste industry. All of these, caste and outcaste, serve the Brahman farmers, and at the same time are able to exchange services with each other. It is comparatively easy to recognize a Brahman. Even the older men among them who are bowed and withered, retain the bearing and features which mark them as Brahmans. There are exceptions, but one is fairly safe in picking the tallest, straightest, most firm-limbed young men as Brahmans. Their skin is fairer and their features finer than those of lower castes. Their noses are high and паг- row. The Brahman women, always kept safely in the courtyards of their family " enclosures, work as hard as other women. But they too have the fair skin, the narrow noses and faces, the firm, smooth bodies which set them apart from others of the village. None of them is fat. We have known only one fat person in the village, a woman of carpenter caste. Her fat was diagnosed as a symptom of her diseased condition. Brahman babies seem no plumper than others, and when weaned, they pass through the same period of loss in weight and resistance as do low caste babies. But if they survive this precarious stage, they surpass other children in weight, height, and strength. Smallpox scars and inflamed eyes are less common among Brahmans than among other castes. Passing down from the Brahmans, the next three castes compare with them favourably. The goldsmith brothers are the village wrestlers, and are fed and massaged with this in view. Below them there is a decline in the evidences of well- being. From appearances it would be impossible to ascertain to what caste a man, woman or child of the humbler levels belongs. Sometimes an individual family gives its children a little better care than others, but such families are not confined to any one caste. A few Muhammadans may be recognized because of their more hawk-like noses. And the Christians are set apart by the consciousness of their [Vor. 42 308 ANNALS OF THE MISSOURI BOTANICAL GARDEN untouchability and their broader Dravidian features. Aside from these differences, all are very much alike. One becomes so accustomed to signs of insufficient food and care that his attitude toward standards of living is apt to shift. Most of the men and women of the lower castes are angular, with hollows where flesh should be smooth. Bodies go unbathed, in contrast to higher castes where pour-over baths are supposed to be a part of the day's routine. They sleep and work in the same garments, and change them once in two or three weeks, whenever the washerman collects them to wash in the village pond. Не uses cold water, beats the clothes on a rock, and spreads them on the dusty ground to dry. After a few washings a garment takes on the shade of gray of the earth and mud walls. The men's hair is cut short, with one long lock at the top. The women and girls do not use combs, and their hair tangles and mats, and offers refuge to multitudes of lice. Smallpox scars and eye infections or disfigurement are so common as to excite no comment. Both men and women work to the limit of their strength, and are thankful if they can earn enough food to make it possible for them to perform the tasks required of them. They do not hope for more, except on festival days when Brahman housewives hand out bread or cakes to the women and children of families performing regular services for them. The boys who run about the village lanes, or herd cows or sheep are small, lithe, and active, ready for any sort of fun. The little girls are the same. But before they are ten, they retire demurely to the courtyards where they care for their baby brothers or cousins, and imitate their mothers. The life cycle of the women is briefer than ours. e girls mature earlier—at ten or eleven—, are married at twelve ог thirteen, and go to live in their husband's homes when fourteen or fifteen. From this time on, they are treated as grown women, capable of carrying a full share of the heavy work of farm households. They begin bearing children in the next two or three years, and although they do not have many babies, there is a constant sex- strain on their bodies. All of their work is done sitting or squatting on the ground, which is less wearying than standing. In spite of this, almost every mature woman of the village complains of the dragging down of her body. While we, in our early thirties, are facing the choice between a career and rearing a family, they are ready to sit back and tell their daughters-in-law how to feed their babies. The Indian Year Book for 1931 gives the mean age of males as 24.8 and of females as 24.7, according to the 1921 census. II—FOOD SOURCES The families of Karimpur retain their customary aloofness in meeting the prob- lem of food supply. They prefer to accept the limitations of their own immediate environment, rather than venture into strange markets. Salt and spices are brought in from the outside, and sold by the small tradesmen. And an occasional purchase is made on a grain selling trip, or on a visit to a religious fair. These purchases are confined to familiar products which are considered a treat because of their limited 19551 WISER— FOODS OF A NORTH INDIA VILLAGE 309 season or supply within the village. Unfamiliar products have no charm. About 40 miles from us is a district reputed for its oranges. These oranges are sold cheaply in the Mainpuri market during two winter months. But they are not brought home to the village. On the other hand, bananas are familiar, because a few are grown in the village. And bananas are brought from Mainpuri, when the price is sufficiently low. Each year tomatoes are sold in increasing quantities in the Mainpuri bazaar, but in the village the tomato is still a strange fruit, of an un- desirable colour, and the villagers regard it with suspicion. Their diet is thus largely determined by the boundaries of their own fields. And it rises and falls with good and lean years. If the rains are favourable, and there have been no hail storms or destructive winds or pests, everyone has enough to eat. The villagers have learned to count on one or two such years out of every five. If the rains fail, or come too late, the whole village cuts rations. If one particular crop is damaged, they get along without it until it is harvested the following year. Their exclusiveness does not limit their diet as much as one might expect. There may not be a variation in foods in one single day or in a week. But over a period of months there is considerable variety. The villager accepts his changes in food with the changing seasons. Не is thus better prepared to welcome each new crop as it comes than is the town man who has greater variety each day. Still further, although the town man seems to have a greater choice of foods in the bazaar than the villager has in his fields, the villager has discovered many edible things which vary his diet and which rarely find their way to the town. A—VARIETY OF FOODS MADE POSSIBLE BY CLIMATE AND SOIL One reason why a greater variety of foods can be produced from a limited area, in the United Provinces than in Bengal and further south, is the greater distinction between the seasons. We have our hot dry summers when land rests, followed by warm rains which make crops of warm climates, like rice and sugarcane, possible. The rains are followed by a long dry season, a part of which is much cooler than Bengal, and which permits the growing of cold weather crops, such as wheat. One can better appreciate the variety of foods in Karimpur, by comparing them with the diet of less favoured districts. In his study of the use of food-grains in India, Church (1) found that "There are many districts in India where rice forms not merely the chief food-stuff but three-fourths or even four-fifths of its total amount. In some places it even rises to seven-eighths or to fifteen-sixteenths of the whole quantity, as in Burdwan, Dinajpur, Maldah, Kuch Behar, Manbhum, and Darrang; other districts might be named in which it constitutes the only food staple.” Another factor which contributes to the variety of the foods of our village is the variety of soils. The soils of Karimpur are typical of those of village com- munities of the Ganges plain. The following division of the revenue village is given in the latest Settlement report. These reports are made every twenty or thirty years by government settlement officers. (1) See References. [Vor. 42 310 ANNALS OF THE MISSOURI BOTANICAL GARDEN Revenue village ој Karimpur Acres Ы ње sites қ E jä => 2% егей with майт. = e S с щі 202 ise barren 1,609 Culturable land 5 2,103 Miscellaneous land .. 4,093 The 2,103 culturable acres are distributed as follows: Type of land Acres | n acs i 39 Manjha wet 204 Manjha dry 18 Barha Dumat wet 554 arha Dumat dry 242 Barha Bhur wet 518 Barha Bhur і 160 Maiyar 68 arai 2 300 2,103 "Gauhan" is the land immediately encircling the village. It is the best manured, and is rented at the highest rate. It may be any kind of soil, usually a loam, improved by constant manuring. Farmers prefer to grow maize, sugarcane, peas and potatoes on this soil, or on the next grade, which is listed as wet and. dry "Manjha." “Мапа” is the same type of land as “gauhan,” but it is the strip just outside the "gauhan," and therefore less manured and less desirable. Most of the vegetables are grown on one of these two soils, partly because of the improvement of the soil, and partly because vegetables require constant care and constant water- ing except during the rainy season. The outlying land is known as “barha.” "Barha dumat” is a mixture of clay and sand in almost equal proportions. It is “generally of a rich brownish colour, adhesive without tenacity, briable [friable] without looseness, slippery and greasy when wet and with a soapy feeling when dry, and cutting like a cheese when ploughed wet." “Вага bhur” is outlying land which is “loose and sandy, and quite incapable of retaining moisture." (2) The “dumat” will bear a large variety of crops, including wheat, barley, several varieties of millet, and pulses. Whereas the "bhur" is so sandy that crops which survive with the least irrigation are reserved for it. It is usually limited to spiked millet and coarse pulses. “Tarai” is the low lying land, which "includes all classes of natural soils from very heavy clays to loose sand." (3) In our area it is chiefly heavy clay. On the “tarai,” rice is usually grown, and sometimes water melons. In ordi- nary years it is swamped during the rainy season. “Мауаг” is “a stiff, unyielding clay of a dark colour, shrinking and cracking in dry weather into a network of (2), (3) See References. 19551 WISER——FOODS ОЕ A NORTH INDIA VILLAGE 311 fissures, but expanding when moistened into a sticky clayey mass." (4). It can be made to produce some of the less desirable pulses, and millets, and perhaps rice. Each farmer wants a share of the better land, and not too much of the poorer land. This has led to the partitioning of land into small, queerly shaped plots. А man may have as many as sixty or seventy plots scattered over the different kinds of soil in every direction from the village. В. Mukerjee writes of these scattered plots, “Various аге the causes which tend to render a peasant's fields so widely scattered. Іп the village communities there was a deliberate attempt to distribute plots among the settlers in different soil areas so as to allot plots of different degrees of fertility to each . . . The Indian cultivator’s farm tends to fly into fragments and grows steadily smaller and less regular. As the population increases, the holdings on account of the law of succession come to be unduly fragmented." And further on he refers to "the pepper pot distribution of holdings." (5) The following table shows the proportions of each type of land included in the scattered plots of J, one of the largest land-holders of the village, and compares the proportions of his holdings with those of the same kinds of land of the whole village: Percentage of сасђ Ета of land in Karimpur compared witb percentage of same kind of land beld by J, a Brabman Percentage Percentage Percentage Kind of soil in revenue in J's in J’ village hereditary total holding holding Gauhan 1.9 Es 0.9 Manjha X са ~ 10.6 15,3 15.1 Barha Dumat се 2% 37.8 47.9 46.2 Barha Bhur .. t» фі 323 22:3 21.0 Tarai : зві У 14.3 13.3 16.8 Maiyar 3.2 5 100.0 100.0 100.0 "The hereditary holdings show the closest correlation because they conform more closely to the original division of land.” (6) With such an arrangement, each long-established family has a chance to raise a variety of crops, even though its acres are few. Poorer farmers, and those whose forefathers came in after the original division was made, may have little or none of the best land, and a larger share of undesirable land. But even they manage to get a variety of crops, by hard work and a resigned acceptance of inferior products as part of their inferior lot. У. Н. Moreland of the Civil Service discusses the management of farm land composed of scattered holdings, in his Agriculture of the United Provinces: (4), (5), (6) See References. [Vor. 42 312 ANNALS OF THE MISSOURI BOTANICAL GARDEN "Now › considering "e er - soils and of possible crops, t чуб x ooks like a very complicated prohi ма it is fortunate the individual cultivator has to face it without some guidan The гаки ion i ds fran ані y many generations of Азія ors and their accumu- lated | experience is to be found in the custom Be the country which guides the individual on such pase s is xi Ж means s$ 22. cds manure over the holding, in what order should crops be grown, and so on. ‘This custom of the country is not an ие uide to the most protttable utiliza- tion of the any em the best course ma penu ried or may have been discarded through some иза ну ог све nges in m p v = ог as in the = and for particular products may make some changes in the custom desirab. le; but it is usually a safe guide to making a living, and the ordinary cultivator is well perg in samedis it itus than applying his limited intellect to working a solution afresh And сагаа оп, ће xus the arrangement of crops as actually practised: “To see how ы? custom works out in practice, we may take the case 248; holding of the ordinary эш, say 5 or 6 acres, containing different soils. There may be a field of - rend land; xe be сой and it dries so quickly that it would be very risky to s crops on i s land will usually have a Ee season crop. But even in the rains it X T pn poti in it will suffer from either or drought: not very much will be spent on its m really fitted, while if possible some cheap pulse "n be sown ғә ough tillage after Le rice has been gathered. The rest of the holding consists, we will su spa irrigable loam e portion of it, however, is so low lying that it is tears a more or dae сы in the rains: bes portion will probably be sown regularly with a winter crop, which may sometimes be wheat and some- times a mi e о at i — ог реаз. season and will be divided between the tw iem way that Rara. can be effected in the available. Possibly the cultivator will do dis in the following way: taking the bow алі "is house he wi y his manure to it and sow maize, which will be ir и by some winter crop: another field he will put under great milles, arhar and urd (two pulses), with some sesamum to supply his house with oil, a ede е гдег оЁ ћетр She give fibre for well-ropes, etc.: the remaining land will be left for the winter and ds ter manuring, sown with wheat along with a border of lins о: ines of гар re oil. i early supply of f from his maize, say about the end of August or September: rice in е ber or October; juar, bajra, urd and moth (two millets biu two pulses) in November: peas СТ ин abies winter e he had sown in March and April, when he чі wire have his wheat ready to sell for the “This is the holdin sie a cultivator ot no special skill: a better man would y in a све за sugarcane or opium, ses ge to save more manure and take two crops larger area, but the general principles will be the same; to secure sufficient food, wi га Prefebl an early supply in a rains; to have something to sell; to hav a good v riety of cro arrange so that there TN time to till for all with the ile pair of еі which р api Mac a holding of this size us support." (7) B—THE CROPS One of the farming customs of the Ganges plain is the mixing of crops. Products intended for sale are kept unmixed. At least the farmer tries to keep them unmixed. But ће is so accustomed to mixtures, that if barley or pigeon peas get into the wheat he sees nothing objectionable in the mixture. It is when his product reaches the city market, particularly where foreigners buy, that he is accused of adulteration. In his study of "Food-Grains of India," Church remarks, “The value of Indian wheats in European markets is often much lowered by pre- ventable impurities. Very frequently they contain other cereal grains, especially barley; gram and linseed some times occur in them." (8) And McCay in his study of jail dietaries in the United Provinces, writes, "The chemical analyses of wheat in use in the different jails show greater variation than any other of the (7), (8) See References. 19551 WISER— FOODS OF A NORTH INDIA VILLAGE 313 food materials. This is practically entirely due to the different degrees of con- tamination with other grains, most of the samples in ordinary use were of second class quality and contained a large proportion of foreign grains." (9) Wheat grown for home use is almost always mixed with the poorer quality cereals. Its presence makes bread made of barley or millet more palatable. And as long as the grains will be cooked together, they may as well be grown together. The more common mixtures have names, according to the grains they contain, an the proportions of each. “Соја is a mixture of wheat and barley. “Тита” is wheat, barley and gram in equal parts. “Bejhar” may be equal parts of barley and peas, or equal parts of barley, gram and peas. Within the village, the selling or loaning of mixed grains can be carried on satisfactorily with the help of these names. There are several advantages to the Karimpur farmer in raising mixed crops. A mixed crop may serve as insurance. А crop, such as rice, which flourishes in heavy rains, may be planted with a coarse millet, which flourishes in dry weather. Then, if the ground is flooded there will be rice, and if there is very little rain the rice may be lost, but there will be a millet crop. Sometimes two crops are grown together to save land. Millet and the pigeon pea are such a combination. The millt grows up more rapidly and the pigeon pea grows more slowly, until the millet is cut. As soon as the millet is cut the pigeon pea grows more rapidly, and when harvested yields about two-thirds of what it would without the millet. And the farmer is one millet crop ahead. In some cases a creeping pulse is grown with taller crops, to retain as much of the moisture in the ground as possible. Or a deep-rooted plant may be grown with one with shallow spreading roots. The most common crop mixture is that of at least one pulse with a cereal. The farmer does not know about nitrogen but he and his forefathers have found that the soil is benefited by the presence of a pulse. They have also discovered that pulse alone is not a satisfactory food. The mixture of pulse and cereal is the happy solution in the field and on the brass eating tray. The following record of his own crops, given us by a village cultivator, reveals the extent of crop mixing, particularly with food crops: The grain crops of the village are familiar to us of the West except in the use for which some of them are intended. We in the United States regard certain millets, such as sorghum vulgare, as fodder crops, but in Karimpur they are grown as a complete family food—first for the human members of the family, and secondly for the animal members. Moreland, in his Agriculture of the United Provinces, remarks: "in many parts the cultivator cannot spare any part of his land solely for fodder crops, but must grow crops which will feed himself and his family as well as his cattle." To a Westerner, the pulses are less familiar than the cereals. The “arhar,” or pigeon pea, is the most common. Its yellow flowers colour the village fields during the cold season. It is one of the hardiest of the field crops. The peas, dried and (9) See References. [Vor. 42 314 ANNALS OF THE MISSOURI BOTANICAL GARDEN Crops grown by a Karimpur farmer Crop Grown with Reason for mixing Wheat x 1. Barley, or 1. Mixed food. 2. Barley and gram 2. Mixed food. Barley R 1. Gram, or. . 1. Good for soil, and 2. Wheat 2. Mixed food. Field peas Alone, or with wheat Mixed food, and good for Sarson (mustard) Gram Duan (mustard) Potatoes Sugarcane 27% Water melons . . Little millet .. Great millet Spiked millet . Maize Rs Rice Pigeon pea .. Urd (pulse) .. Mung (pulse) . . Moth (pulse) . Ronsa (pulse) . . ~ Cluster bean .. Sesamum і . 1. Wheat,or.. 22 Barley с. l Wheat ... 2. Wheat and barley 1. Wheat,or. 2. Barley a А]опе. 1. Hemp, or. . а 2. Castor plant on border. Alone. Cotton E M e 1. Pigeon pea, mung, or urd. № Mustard .. ME 1. Moth (pulse) or pigeon pea. 2. Mustard .. га Alone. Alone 1. Millet _ 2. Cotton 3 1. Millet Spiked miller . . 1. Great millet 2. Spiked millet 1. Great millet 22% 2. Spiked millet € Cotton > soil. Both save ground. Both for mixed food. Both save ground. Both, 1 and 2, save ground. Saves ground. 1. Mixed food. 2. Saves ground. 1. Mixed food. 2. Saves ground. 1. Mixed food. 2. Saves ground. 1. Mixed food. 2. Saves ground. Mixed food. Mixed food. Both, mixed food. Both, mixed food. Saves ground. 1955] WISER— FOODS OF A NORTH INDIA VILLAGE 315 split, are used by everyone. "Urd" and "mung" are much alike and are usually classed together. They are the best liked of all the pulses. “Moth” is the least used. Some do not care for it, while others keep a small store of it on hand as a special food for invalids. These three, “шга,” "mung" and “moth,” are dried and split, like the pigeon pea. Bengal gram is one of the most welcome of the products of the spring harvest. In the fields, around the home fire-place, and at the grain- parchers, it supersedes all other parchable foods. Travellers carry it parched, often their sole food. In the Indian army it has been adopted as emergency rations. A few pulses, like the field pea and the cluster bean, can be used green for a short period. But as soon as the peas are ripe they are treated like the other pulses—split and dried. When split, this whole group of pulses is known as "dal," an important division of any family's diet. Vegetable crops are traditionally in the hands of vegetable growers, Kachhis. The task of growing them is not so much an exclusive privilege as a laborious caste duty. We have been interested in the slight increase in vegetables grown by ordinary farmers during our years in the village. They have discovered that enough water spills over the edge of the big leather water bags to keep vegetables alive on a small patch of ground surrounding an irrigation well Beyond this they have neither the patience nor the interest to struggle with vegetables, especially in the hot weather when it seems as though nothing green can possibly survive. They leave vegetable growing to the Kachhis. The vine crop is the largest vegetable crop in the village. It includes several kinds of cucumbers and melons which were new to us. Like other foreigners in India, we found the hot season made almost unbearable by the absence of fresh vegetables, until our Indian friends introduced us to several unattractive forms of cucumber, melon and gourd, and showed us how appetizing they could be made with the help of green mangoes and spices. Tubers and roots are depended upon for winter vegetables. Potatoes are dug and used when very small, chiefly because tbe crop is in danger of being stolen, or of being destroyed by porcupines. Radishes and carrots, on the other hand, are allowed to grow very large. In his Agriculture of the United Provinces (p. 222), У. Н. Moreland reports “Radishes are grown for their bulk, and may be anything up toa foot in length." Of carrots he writes, "The Indian carrot is different from that which is familiar to Europeans as a vegetable; the root is dark coloured and is coarse and flavourless.” Beets and tomatoes are grown on a small scale in the vicinity of Mainpuri, but Karimpur conservatism bars them from its fields. The prejudice is chiefly against their colour which is that of blood. Onions are limited in their growing and use to the lower castes. In the Laws of Manu, V. 19, it is written “А twice-born man who knowingly eats murshrooms, a village pig, garlic, a village cock, onions or leeks, will become an outcaste." There appear, in the village list of vegetables, foods which we overlook. Buds of certain trees are treated as vegetables. And a whole series of leaves, some wild and others incidental to the growing of some field crop, are utilized. Among the former is the white goose foot, and among the latter are radish tops, the leaves of . [Vor. 42 316 ANNALS OF THE MISSOURI BOTANICAL GARDEN the Egyptian arum, and the tender tips of pulse and mustard plants. All of these leafy foods are set apart from other vegetable foods, and cooked in a special way. In the village they are called "sag," or greens. The village looks to its trees as a special source of variety in the diet. And the trees have something to offer during most of the year. Some of the fruits, like the neem and peepal berries, we would never consider as human food, and in govern- ment publications they are listed as famine foods. But the children eat them in large quantities and farmers munch them on their way to or from the fields. The children also gather and eat the blossoms of the mahwa tree, and forage for wild plums. Their enjoyment of plums, as well as of other fruits, when hard and green, is amazing. Anyone who has tried to eat a tamarind pod knows how pain- fully sour it 15. We use it to make a hot season drink. But the village youngsters chew one after another without a blink. The tamarind is valued in Sanskrit medi- cine for its antiscorbutic effect. Certain fruits are so rare that only a few can have them, even for a limited season. Among these are the pomegranate, custard apple, guava, date and lime. Bananas are somewhat less rare, but are treated as a luxury. They are sliced green, skin and all, and cooked with vegetables. Musk- melons and water-melons appear when most welcome, in the hot, dry weather. Everyone manages to get a share of The most popular and the most abundant of all the fruits is the mango. In the hot months, when the ground is baked hard and few growing things survive, the mango trees yield their crop. Mangoes seldom ripen. Wind storms bring down some of the fruit. And the rest comes down, with the help of sticks and stones and secret trips up the trees, before we think it ready to eat. It is the right of everyone in the village to have a share of the mangoes. This is in accord with the Laws of Manu (VIII, 339). “The taking of roots and of fruit from trees. . Manu has declared (to be) no theft.” The owners of groves do not object to the stripping of their trees, as long as the trees are not damaged, and as long as the fruit is reasonably distributed within the village without being sold. If mangoes were picked and sold, even by the owner of a grove, the village would deal with the offender by its own unique method, that of non-co-operation. We have known progressive landlords who have raised mangoes for sale, but they have been power- ful enough to assert their rights. In our village, the old custom prevails to the extent that owners of groves entertain no thought of financial profit. One man, of farmer caste, owns an unfenced grove of 25 old trees, bearing in a good season over 50,000 mangoes Our office clerk lived in the home of this farmer, and was in a position to observe closely. No mangoes were brought into the owner’s house from his own grove, during the clerk’s two year’s stay with him. Fruit from near- by trees, belonging to others, met the needs of his family. He was following the practice of his forefathers in letting others enjoy the fruit from his trees. And he is assured that as they were blessed by the religious merit thus acquired, so will he be blessed, especially since a number of those who eat his mangoes are twice-born. The Government encourages the preservation of groves by not taxing them as 1955) WISER—FOODS OF А NORTH INDIA VILLAGE 317 long as they remain groves. Rule 4, Chapter I of the Wajib-ul-arz (Customary Law) states that, "If the trees of a garden are cut down and new trees are not replaced, the Government will have a right to assess land revenue according to circle rates, if that grove has been exempted from the assessment of land revenue at the time of the current settlement." This has helped retain the groves which existed before the last settlement. But there has been almost no effort to plant new groves. The custom of free distribution of fruit fitted into the old regime of simple, personal exchange. Under the comparatively recent system of money exchange, men want some assurance of a money return, before taking the trouble to plant and foster young trees. The following table gives a picture of the sources of the village supply of cultivated fruits. Number end hinds of trees found in groves of вија Мит Name of t р г mature oe ом 159 187 Tamarind E d RUP TER 1 4 Neem => S 422 ES 219 1,051 Peepul 3% vss Vx і 12 8 Gular (wild fig) з ba “а 5 37 Bel Ue . 5 s 36 66 -apple . і 5 27 Indan jujube 5 55% ka 46 132 i. (kind of plum um) .. rs с 45 г (kind of date) .. I er 9 76 Khan labhera iem) > VÀ 5% 24 - ч Jack 2% т У: 9 ы: (kind of date) . 4% «Зі 55% 9 ceram ЕИ Ж 5 2 220 ote es oe 50 25 Plantain as i 2% n 18 4 3 5 Mahua Tap 27% 12 3 n bers “(kind of plum) sa н d ; Number and kind s trees on p village site Name of tree Number Mango age 3 22 1 жағасы = і > 10 Реер А š 3 Ber ndun за . . . 4 Khara isbhers (plum) : сс . 2 г (pomegranate) P5 i4 $i 1 A (guava) as и» e 2 Banyan >% Ko» 3 Bel i : The ponds near the ы contribute a share T the did supply. Vegetable growers and water-bearers collect generous quantities of water chestnuts and seeds of phapola, a water vine, from nearby ponds, and sell or exchange them at such a low rate that everyone can buy. The phapola, puffed, or popped, was one of our happiest discoveries among village foods. We like it better than popped corn. [У от. 42 318 ANNALS OF THE MISSOURI BOTANICAL GARDEN At the festival known as Deotan, in November, everyone in the village is allowed a taste of sugarcane, even though it is not ready for cutting. From this time on, one meets children and men chewing and sucking stalks of cane. A stalk may be broken off as one passes by a field, without offence. Not much fresh cane reaches the women shut in the courtyards. But they enjoy whatever the children remember to bring home to them. The fresh juice adds variety to a number of dishes during the pressing season. And the raw sugar, made by boiling down the juice and pressing it down into large cakes, supplies the village with sweetening for the year. C—ANIMAL PRODUCTS Animals are a minor source of food supply in a vegetarian community like Karimpur. Cows are kept and milked, but the milk is seldom used as such. Boiled, churned into butter, and clarified, it supplies the pure fat for deep fried foods. Beef is never eaten. Goat meat is eaten about twice a month by the goldsmiths. And Muhammadans eat it oftener, if they can afford it. The goats kept by the village goatherds are supposedly kept for milk, which is sold in Mainpuri. But occasionally a goat, no longer useful, is sold quietly to a Muhammadan butcher. Any goat meat beyond this comes from Muhammadan communities outside Karim- ur. Fish is not eaten by members of the higher castes. But members of lower castes have a feast of fish, during the few days before the ponds dry up for the hot season. The men who are draining the pond with their reed water-lifting baskets, let it be known when the water is getting too low for further usefulness for irriga- tion. The word passes swiftly, and on the following morning a crowd of low caste men and children gathers at the edge of the pond. Each one has a big round basket without a bottom. The instant the irrigators announce that they are through, the fishers wade into the shallow water and push their baskets down into the mud bottom. They scoop out any fish they have thus captured, and move forward a few steps where they try again. The fish are small, but they are an annual treat. The few chickens in the village belong to the sweepers, lowest of the outcastes. No one else would eat their eggs or their flesh. The sweepers, along with the Dhanuks, another group of outcastes, own the swine, which act as village sewage system, and which are the only animals raised with a view to butchering. There are many foods denied these outcaste families, but when a swine grows old or ill and is butchered, the feast is theirs. No caste person would come near enough to interfere. If a family is small, and can afford to keep a buffalo, there will be "ghi" or clarified butter, for all of the cooking. But in most families the "ghi"* must be supplemented by mustard oil, or sesamum oil. There are several varieties of mustard and rape grown in the fields of Karimpur, usually mixed with grain crops, but harvested separately. Each farmer tries to have his own supply, which he *Pronounced “ghee”. 19551 WISER—FOODS OF A NORTH INDIA VILLAGE 319 stores, and has pressed month by month, by the oil presser. The better grades are very satisfactory for deep-fat frying. The poorer grades are bitter. The women who are obliged to use the poorer oil, try to relieve it by adding some of the better. When mustard oil becomes scarce, the sesamum harvest is ready with its fresh oil supply. The sesamum is valued not only for its oil but for its flavour. The seeds, with crude sugar, are made into balls. These are among the most popular sweets in the autumn D—SPICES If one tries eating Indian vegetables with only salt added, and then visits an Indian home where these same vegetables are prepared with spices, he has a new appreciation of the art of spicing. Ап untempting vegetable can be spiced and cooked in several ways, and scarcely be recognized as coming from the same plant. This ability to use spices properly is even more important where there is no meat or meat stock to supply flavour. The variety of spices used in a village home is limited by purchasing power. Everyone tries to have at least red and black pepper, coriander, asafoetida, and turmeric, along with salt. Other spices are necessary for certain dishes, but may be omitted from ordinary daily coo ist of spices sold by a village tradesman follows. Two other tradesmen have smaller stocks of the same things. e list does not include peppermint, which is grown in some of the larger courtyards, or coriander, which is grown in fields with grain crops. Condiments sold by Moban Singb, Karimpur Local name English name Botanical name* . . jAsafoetida . . . | Ferula alliacea Lal oid .. |Red pepper : эре frutescens used both green ripe. Kali mirch. . .. |Black pepper Piper nigrum. From unsorted berries, with some husks. Safed mirch . |White pepper m Piper nigrum. From sorted berries Saunf У ied .. . |Pimpinella anis ал. .. |Small fennel . | Nigella sativa Safed jira . . c3 raway .. . |Corum carui. Idi Ы .. Turmeric . Сигс uma lon onga. ung Ls .. |Clove ~ E кр caryophyllata. Chhoti elachi .. [Small cardamom . | Ellettaria cardamom, true cardamom. Bari elachi. . .. |Greater cardamom Amom sub Nariyal . |Cocoanut .. Cocos nucif r Ajwan .. і 21% Са сора Sonth " .. [Green ginger root ngiber officinale ismi i ene 5 Vintis vinifera Chiraunji . s ra Buchanania latifolia. Namak ‚> 16016 .. |Common salt with sodium c arbonate, in sort of fedes. Chhoti pipar and bari Long as cloves, but thin. Not quite as ap rest = .. [As thick as cardamom, and a half inch long. Supposed to aid digestion. *As in the rest of the paper, the spelling is reproduced here exactly as in the original copy. [Vor. 42 320 ANNALS OF THE MISSOURI BOTANICAL GARDEN E—FOOD COSTS IN KARIMPUR Food prices given in the accompanying table are rarely paid. Farmers would not have money to buy food at these rates, low as they seem to us. They produce most of their own food, or exchange food which they have stored, for food or services which others have to offer. When the exchange is made, there is a great deal of arguing, and he who is the best barterer gets the best bargain. Either the seller or the buyer produces balance scales. One article to be exchanged, such as cucumbers, is put into one pan, and barley, corn, rice or almost any food is placed in the other. The exchangers agree on how the scales should hang, according to their estimate of the value of each article. "The pan of cucumbers should swing far down if wheat is in the other pan. If there are sweet potatoes in the other pan, they should almost balance the cucumbers. With this in mind, one realizes the futility of measuring a farmer's food expenditure in terms of money. At best, it can only give a general idea of comparative values. The figures here given we believe to be as accurate as can be found. They were recorded by our clerk, in the household of farmer caste in which Һе lived. If we were to list the prices which we were actually paying for the same articles at the same time, our figures would be much higher—twice as high in some cases. A white face always sends prices up. Prices of foods in Karimpur in 1928 Number of pounds Article purchased for 1 rupee Wheat SH 1 5 e 1 Barley a а = es 24 i a ay 20 Rice ‚> $ Р» 14 Spiked millet : i А == 16 Great millet Ta ee ili 25; 17 Стат En s 30 Urd 2% ey ia = 16 Mung os Е КЎ А гі 16 Moth ahs s ix i s 10 Pigeon pea Р a СА > 32 Milk $3 2% і й : 12 Ghi ix PX ate 1.25 Mustard oil oe : : 3 Potatoes .. bg RT 50 Lauka (kind of pumpkin) 50 Lauki (kind of gourd) . . 5% ix s 53 Turayan (kind of gourd) 252 es © 43 kra 45% A і 27 Cauliflower - 43 Egg plant 3 66 reen fi 5 * 33 Bitter melon 27 String У mi E ©, 66 Radishes, with tops “ү 7 i it 66 Carrots Jà wa З 66 Евургіап агит Ps 43 Khata labhera (blossoms) b. rz ie 25 Khata labhera (green fruit) A 2% кк» 43 Cluster bean, with pods x P3 T3 65 Cow pea .. i s Ра 55% ey 65 Field pea . . 4» e. фо s 33 1955] WISER— FOODS OF A Article Onions . rem се “з 5: $ = ria сом melon) € Сану 7 ee Mus - Mus d. 5 тенор ig пе Ea 5 White goose foot 2 Gram tops. . s Spinach a bd dill 2.4 аен чи pot-herb) .. Опі “4 Раш La v (eia фону орз us aunrai (pot herb) Leaves of Egyptian arum Mangoes i Guavas Jaman (kind of Иса. ‚ Garlic nS du е Fenugreek (seeds) 24 .. Caraway .. F 4% NORTH INDIA VILLAGE 321 Number of pounds purchased for 1 rupee o N со со ка ка N V) \д N на UW Price of meat А two-year-old pig, weighing about 75 pounds, was butchered on 10th Feb- ruary, 1928. It was divided into 12 shares plus the head. Each share weighed about 5 pounds, and cost about $.16. most desirable part. Cost $.20. The head weighed more, and was the [Vor. 42 322 ANNALS OF THE MISSOURI BOTANICAL GARDEN III—FROM FIELD TO SERVING TRAY A—THE CARE OF FOODS AT HARVEST TIME At harvest time, all of the able bodied men and boys of the village and some of the low caste women go out to the fields. Some cut their own crops, and some are employed to cut those of others. The men squat in a row at one end of a field and move slowly toward the opposite end, cutting the grain handful by handful with their sickles as they hunch along. They tie the cut stalks into bundles, bind- ing them with tough grass. When а field is finished, they carry the bundles, usually on their heads, to one of the groves beside the village. We camped in one of these groves and used to watch the men coming in, balancing their loads and dropping them on the large pile belonging to the master of the field. The millets are treated differently. Women of farmer and water-carrier castes walk slowly through the field breaking off the heads of grain and dropping them into large flat baskets which they carry. Later the men cut the stalks. This job can be postponed until more important crops are harvested. Rice is cut in the usual way, but is not carried to a grove. It is threshed close beside the field where it has grown. Most of the grains and pulses are threshed by the treading of bullocks on a threshing floor. The threshing floor is simply a circle of ground from which grass and weeds have been scraped. The threshing-floor for rice is prepared beside the rice field. Other threshing-floors are in the groves where the cereal and pulses have been stacked. In our grove there were five and sometimes six threshing-floors being used. On each of them from two to six bullocks were driven round and round day after day. Small boys usually do the driving. A stack of grain about three feet high, covering а threshing-floor with a radius of ten feet, takes three days to be threshed by four bullocks. This varies of course with the activity of the driver. People without bullocks must leave their grain stacked in the grove until some one else’s animals are available. In the threshing, the husks are loosened, and many, but not all, are separated from the kernel. The husks of rice and barley are not loosened, by the treading of ocks. With these two the threshing simply breaks the kernels from the stalk. The actual husking has to be done later by the women in the courtyards. Close after the threshing comes the winnowing. The men hold shallow baskets of grain high above their heads, and tilt and jerk them in such a way that the kernels fall to the ground. The chaff is carried away by the breeze. If there is no air stirring, two men stand behind the winnowers and flap a long sheet, thus making a breeze. Grain to be sold is weighed out and poured into gunny bags, beside the thresh- ing-floor. The two men who keep large store-houses of grain to loan or sell later in the season to village neighbours, measure their grain into sacks and have their hired helpers carry it to their store-houses. A few others keep their surplus in 19551 WISER—FOODS OF А NORTH INDIA VILLAGE 323 bags. But most of the products of any harvest—grains, pulses, vegetables, or fruits—are carried home from threshing-floor, grove or field, and poured out on the floor of the courtyard. Whatever remains to be done, is the work of the women. B—STORAGE AND PRESERVATION OF FOODS The women shape tall jars of clay which remind one of the tale of Ali Baba. They must have them made and well dried before the harvest. As rapidly as the grain is brought in, they spread it on sheets, on their roped cots, and keep moving the cots so that they may get all of the sun possible within the courtyard. When the grain is thus thoroughly dried, they clean it basket by basket full, and fill the jars with it. Near the base of each jar is a small opening, plugged with rags. · Through this, а day's supply of grain can be drawn out. Pulses to be stored require additional treatment. The women must remove any pods not removed in threshing. Then they soak the pulse over night in water and oil. The quantity of water varies. To five pounds of pigeon peas they add ten pounds of water and three and a half ounces of oil. To five pounds of "mung", "urd", or "moth" they add only four ounces of water and an ounce of oil. After the soaking they dry the pulse in the sun for a day or a day and a half. For this they requisition every cot in the house and every sunny inch of the courtyard. When the pulse is thoroughly dried, they split it in the light stone hand mill described. If a woman does not own such a mill, she tries to borrow one from a neighbour or employer. ТЕ she cannot borrow, she pries apart the stones of her flour mill, with a wad of cotton on the top of the pivot, and splits the pulse in it. She may split her whole season's supply immediately after the harvest, or she may take care of enough for three or four months at a time. During the splitting process many husks are loosened. She gets rid of them by winnowing the pulse and sifting it, before storing it in clay jars. There is heavy work left for the women, in the care of rice and barley. Both of these have been partially threshed by the treading of bullocks. Bullock's hoofs аге able to separate the husks from the kernels of some grains, but not those of rice or barley. This task is left for the women. With an iron-bound wooden pestle, a woman husks her rice or her barley in a stone mortar set in her floor. It is tiring, strenuous work, and she prefers to do a small portion at a time, as it is needed. Where there is a supply of a certain grain to be stored, too small to require a tall jar, it is stored in smaller jars, made and baked by the village potter. These are often found in rope baskets, suspended from beams of the family's store-room, 9ff of the courtyard. Grain kept for seed may be stored in the same way, or hung in the courtyard on a beam protruding beyond the store-room wall, protected by eaves. The village housewife wages continuous battle with the rats and insects. Which covet her stores, and with the crows which settle on the courtyard wall waiting their chance to swoop down on drying grain. [Vor. 42 324 ANNALS OF THE MISSOURI BOTANICAL GARDEN Drying is not limited to cereals and pulses. Greens, such as the tips of peas and mustard plants, are chopped, and dried on the cots in the sun, and kept in baked clay jars, or in bags improvised from old skirts and shirts. “Mahua” flowers, if not all eaten fresh, are similarly dried. Coriander, the most used of spices, is gathered by the children and dried by their mothers. Mangoes, peeled and dried, are kept in jars to be used later to flavour insipid dishes. Mustard seed is dried and stored, ready to be doled out to the oil-presser when a fresh supply of oil is needed. The only other forms of preservation within the reach of a village housewife аге spicing, preserving in oil, and preserving in salt. Vinegar is not used. Тһе containers in each case are clay jars of various sizes, made by the village potter. Carrots and radishes will keep about a week, if well mixed with spices and covered with oil. The "rai labhera" (small plums) keep longer. And horse-radish, limes and mangoes keep for months. The "rai labhera" and limes are scarce, and hence available to only а few. They require a little oil, and spice. The horse-radish and mango are within every one's reach. Mangoes lend themselves to the greatest variety of preserves. In salt they are acceptable, in spices and oil they are a happy addition to a noon day meal, and in raw sugar they are a special treat. With clay jars or old cloths as her only containers, and a stifling room as her only storing space, there it little more that the village woman can aspire to, in carrying foods beyond their season. C—EQUIPMENT Before considering the further processes through which the food materials must go in the home, it will be necessary to study the equipment with which a village woman works. Her work-room is the courtyard in which all family activities are carried оп. The earth is her floor, and the sky is her ceiling. From the courtyard she sees only these and the walls of mud around her. The walls which surround her are made of mud dug from the bottom of a near-by pond. She cannot see over the tops of the walls and, more important still, no one can look over the walls at her. If she is а Brahman, her courtyard will be large enough for her to work in, and to spread out her grain comfortably, with plenty of sunshine most of the year, and at least a patch of it crossing the floor in mid-winter. If she is an outcaste, her courtyard will be small, originally high, but her walls will be eroded until low enough to admit abundant sun and air. The women who suffer are those whose courtyards are small, and whose husbands insist upon keeping them in strict seclusion. Such a one is the oil-presser’s wife, whose courtyard is cramped, and high-walled. During the rains, she works in puddles, and during the winter she shivers between damp walls which shut out the sun. She and her children are among the most wretched in the village. The carpenters are as near average as any group in the village; and a diagram of a carpenter’s house is shown on opposite [page]. The enclosure is a typical in that it has two rooms off the courtyard. Usually only Brahmans have two, whereas those GROUND PLAN ОЕ А KARIMPUR ENCLOSURE 8 Feet = i Inch — — | | і | | | 2” | | d е Height | x о С or | % т COURT YARD "Р * ~ |- 7 2 g З =. СО ній Ж ! o Ж for бог тд mt у | і 0 | | T | | i | | Bathing | ace - | Clay тегатр l | Jar. for EMT | М ” | | 2-9 | | Height of room | | 8-2 | } еа EAST ROOM ^ um | ES | | еи => Wem ! | k 3 то Height of wall i | Si 5 H gs J à $ у Y [Vor. 42 326 ANNALS OF THE MISSOURI BOTANICAL GARDEN less than Brahmans should be content with one. In this particular case the second room was added to accommodate extra brothers. It is now falling into disrepair. The doors of the courtyard lead to store-room or cattle-room, not to the out- side world. On rare occasions when a village woman is going to her father's home for a visit or to the wedding of a relative, she swathes herself in a coarse sheet and goes out through the cattle-room. Her only other means of exit is by ladder. If her courtyard has a ladder, she climbs to the top of her wall and looks over at her neighbour, and if no men are at home in her own or her neighbour's house, she drops down into her neighbour's courtyard for a hasty visit. Aside from these rare breaks, she spends her life in her own courtyard, with food-preparation as her chief task. Her children wear few garments, and these are made by the village seamster, and washed by the village washerman. She has no furniture to care for beyon the unvarnished roped cots, used as beds by night and as drying-racks by day. During the winter weeks, coverings of home spun sheets or padded quilts are used. These she throws over wooden pegs projecting from the courtyard wall. She has no hangings or rugs, or silver or china to care for. If food processes were as simplified for her as they are for us, her life would be easy. But as it is, her work is heavy and slow. Her activities centre around her "chulha" (fireplace). Generations ago a great-grandmother-in-law made it in a secluded corner of the courtyard and each daughter-in-law since has kept it in repair. There is a supplementary "chulha" in the store-room to be used in stormy weather, but the one in the courtyard is the true family hearth. It is made of mud piled up and shaped so as to protect a small flame and to support a cooking utensil. (The daily care which the housewife gives her "chulha" will be discussed in connexion with the ceremonies of cooking and eating.) Her fuel is the twigs which the children gather from groves, or the dung cakes which she and other women of the household have shaped. Almost as necessary as her fireplace is her stone flour-mill. All the grains and pulses which are to be made into bread must be ground in the home mill. mil is made of two stones, cut like discs about two inches thick and from one and one- half to two and one-half feet in diameter. The housewife has embedded the lower stone in a base of clay, which lifts it to a convenient height from the floor. The upper stone is free to turn on the lower one, being held in place by a peg in the centre of the lower stone which passes through a hole in the centre of the upper stone. The clay base has a diameter several inches greater than that of the stones, so that it forms a border around the outside of the mill. This is trenched to loo like a moat surrounding a low stone castle. As the woman turns the upper stone round and round with a wooden peg fitted into a slot near the rim, she pours her grain into the hole at the centre of the stone. As the grain is crushed and ground between the stones it gradually spills out around the edges into the encircling moat. When the grinding is finished, the housewife brushes the flour out into a winnow- ing basket, and separates chaff and clay-dust from the ground grain. She returns some of the bran to the flour, unless the bread is intended for a special feast. The grinding is heavy, tedious work, and if possible two women share it, sitting facing 1955] WISER— FOODS OF А NORTH INDIA VILLAGE 327 each other across the mill, each with a hand on the turning peg. One of them pours the grain into the centre hole, without breaking the rhythm of the turning. The steady hum of the grinding stone is one of the earliest morning sounds of the village. In а few homes опе finds а "darenti", a mill patterned after the one de- scribed, but smaller and lighter. It is used in splitting the pulses. Аз the supply for several months can be food. For the winnowing and sifting there are baskets and sieves. Cooking utensils are never borrowed, but the "darenti" is regarded differently, although it touches the food. efore rice or barley can be ground they must be husked. For this the house- wife has а pestle and mortar. The pestle is a thick pole about four feet long, bound with iron rings, one of which forms a rim at the bottom. The mortar is a stone cup set into the earth floor of the courtyard. The husking is another strenuous task usually shared by two women. They sit on the floor one on either side of the mortar. Each woman holds a pestle, and turn by turn, each lifts her pestle above her head and brings it down heavily into the rice, some of which is struck, but much of which flies out of the mortar. One of the women keeps brushing the scattered grains back into the mortar, while managing to keep her hand out of the way of the pestles, and not missing her turn. The grain thus treated is not only husked but chipped and broken. However, skilful winnowing saves every scrap that might [be] food. For the winnowing and sifting there are baskets and sieves of several sizes and shapes. The winnowing baskets are of reeds. The sieves may be strung with coarse threads or reeds. The utensils which the village housewife uses are made of brass, iron, stone, clay or wood. When she makes her daily bread (she actually does make it every day), she may mix and knead her dough in a bow] of brass or wood, or baked clay. She pats out the flat round cakes with her hands if the flour comes from mixtures of corn, barley, millets or pulses. If the cakes are of wheat, and especially if they are intended for guests, she rolls them out thin on her small round breadboard, which is of stone rather than wood if she can afford it. To bake her bread, she may rest over her fire a convex iron plate which she brushes with clay, to keep the cakes from sticking. Or she may fry the cakes in deep fat in an iron vessel shaped very much like those which we use for deep-fat frying. She lifts the cakes out of the fat with a long-handled, flat-bowled iron spoon, which may or may not be perforated. And when her bread is ready and her husband calls for it she serves it on a brass tray, always brightly polished. For cleaning her brass she rubs loose a bit of her earth floor, applies it with a knot of grass and much rubbing, and rinses it off. en she prepares vegetables she rarely removes the skins, but cuts up the whole vegetable with a small sickle. She braces the handle under her toes so that the blade stands upright. Against this she pushes the vegetable. Greens are held fin bundles and cut through in the same way. To fry her vegetable, she uses the same iron pan for deep-frying. To boil her "greens" (leaves and herbs), she uses a brass kettle with straight sides and no handles. Both “greens” and vegetables 328 ANNALS OF THE MISSOURI BOTANICAL GARDEN she serves on the brass tray with the bread. The bread is broken off bit by bit and used to dip up the other food. ТЕ she is serving guests, she will place bread and vegetables and any other food on clay saucers or dishes made of large leaves bound together with twigs. These can be thrown away when the guests depart. In no case should household utensils or dishes be touched by others than members of the amily. When I admire a piece of brass, I always admire from afar. Every day the housewife grinds the spices to be used in the food. For this she has a stone about an inch thick, about six inches wide and a foot or more long, curving gracefully to a point at one end. On its corrugated surface she lays her spices, sprinkles them with water, then rolls a small, rough stone roller back and forth over them until they form a paste. Every morning and evening she puts fresh milk on to simmer. For this she uses a baked clay jar, set in a smouldering fire of dung cakes. A separate fire- arrangement is required for this long period cooking. It consists of a large, shallow basin with a thick base which she has made of clay. In it she places a burning dung cake, with others arranged over and around it, like a nest for the jar of milk. After hours of simmering she pours the scalded milk from this clay jar into another kept for the purpose. And the next morning she pours the clabbered milk into still another clay jar which serves as a churn. Each of these jars has a distinct Shape and is called by a different name. The churning paddle is а stick of woo with cross pieces at the bottom. The woman who is to churn loops a rope around the leg of a cot and then twists it at the middle around the paddle, and holds the two ends in her two hands. By pulling one end of the rope and then the other, she rotates the paddle horizontally and keeps it rotating until butter forms. e churned butter is placed in a special jar with other butter and accumulated for a week, at which time it is clarified. The utensils mentioned, along with the large clay jars which are filled each day with water, our housewife considers necessary to the feeding of her family. With increasing means and increasing family, she may add to the number of these same articles, or she may replace clay with iron, or iron and wood with brass. Or she may add brass cooking utensils and serving dishes of varied sizes and shapes. She may aspire to a serving dish or cup of bell metal, which is more expensive than brass. Beyond this, her desires do not go. The following list of utensils indicates the variety used, and the relative im- portance of the different articles, as shown by their frequency. The list includes forty-eight households which represent all castes and all degrees of prosperity. I have divided them roughly into what our village friends call “rich”, “moderate”, and “poor”. They would add to the list "very poor", but these I have included under “poor”. There is no way of estimating the actual wealth of a family. Those who appear poor are very apt to have money and jewellery hidden away in a wall or the ground. Indebtedness is no register, as most of the village is in debt: 1955] WISER—FOODS OF A NORTH INDIA VILLAGE 329 List of articles, with percentage of rich, moderate, and poor homes, in which each occurs Classification of homes Local name Description 5 rich 22 mode- 21. poor te Per cent. Per cent. Per cent. ha қ Grinding s i 100 100 86 Darenti .. | Light mill foc splitting omn ag. 60 24 5 usa ы Tuo рес for huskin is 100 100 100 Dahkli 4 Жы ог ri 248 : 100 100 100 Charpai t т со which foods аге 100 100 100 rie Chalni . . | Round 500008 sieves strung with 100 100 95 th Sup 2% eed winnowing es 54% 100 100 100 Parat .. | Brass mixing bow А 80 50 14 Tasla .. | Iron mixing bo 3 80 32 .. Kanthani . . | Wood mixing bowl 100 64 62 Kathel а 9а wood mixing bowl 100 73 67 Кипга .. | Сау mixing bowl. . 80 5 .. Kunri .. |Smaller mixing bowl 100 55 24 Pata (w) .. | Wooden rolling ка % 100 100 95 Pata (s) .. | Stone rolling boa 80 41 33 Belan .. | Wooden rolli g se 4 100 85 95 Тама .. | Сопуех ir riddle 100 100 100 Karahiya .. | Шоп pan for deep f g 100 91 86 ansiy .. | Small sickle for cutting vegetables 100 95 86 Sil batta .. |Stone on which spices аге ne 100 100 100 orhwa .. |Roller for grinding spic 100 100 100 Batua e esee іні for "bons е 100 77 52 atili 5 sae ooking .. 80 45 24 Mathni m $ 80 77 38 Chamch iv pe a ані t iron spoon : 100 100 90 Kalcha . . | Different type of iron spoon 100 100 100 Chamcha (wood) Wooden sp v4 ar oa .. 19 Chimta .. | шоп fire ton 45 is 100 86 57 Chimti .. | Smaller iron 20 9 vs Sansi .. | Tongs аг, for lifting kettles . 80 9 “> Тћага .. | Clay water - 4 100 100 100 Kalsa .. | Brass water Aot x s 100 59 .. Chilamchi .. |Brass wash basin 2 20 .. . Silafchi .. | Brass basin 2 "s 20 ко .. ali .. | Brass serving tray . . it 100 95 86 Phul Thali .. | Tray of bell metal . . es 80 45 5 Rakabi 5, рги ate 4% 80 18 "s ota .. s drinking bow 4 ... 100 100 62 Phul lota .. Drinking bow of bell pya ira 40 .. 27% la Eye which food may be 100 100 86 a Phul bella .. | Bell metal cup. © 60 27 2 Gilas .. | Brass glass .. 100 86 er Ghanti .. | Small ен water vessel рі 80 68 81 Tarazu .. | Balance 100 95 76 Pandan +. | Box contsining betel син 60 2 5 Clay saucers .. | Cups and jars z 100 100 100 [Vor. 42 330 ANNALS OF THE MISSOURI BOTANICAL GARDEN D—DISHES PREPARED FROM— Cereals Every day the village woman makes fresh bread for her family. Аз has been mentioned, she makes it of wheat, barley, or mixtures of these with pulses, from the spring harvest until the end of the rainy season. From the end of the rains until spring, she makes it of maize, different millets, and perhaps rice, with a wheat mixture occasionally as a treat. Whatever the grain, she suns it well the day before she uses it. Early on the day when she is to bake it into bread, she, with the help of another woman of the family, grinds and winnows it. She mixes the flour with water in her mixing bowl and kneads it by alternately working it vigorously with both hands, and pounding it with her knuckles. When she considers it “soft enough," she takes up a ball of the dough and works it further in her hands, before patting and rotating it into a round, flat cake. While she turns the cake in her hands, she pinches the edges. She brushes a daub of wet clay over her thin, convex griddle, to keep the cakes from sticking. As each cake is shaped, she flops it on to the griddle. There is only room for one cake at a time. She turns it over with her long, flat iron spoon. While the first cake cooks, she shapes the second. Аз soon as the first is done well on one side, and slightly on the other, she stands it upright inside the fireplace under the griddle, to toast, while the second cake cooks above, and she shapes a third. Meanwhile, she must keep her fire alive by blowing it and adding fresh | fuel. Her fuel may be small twigs, gathered by the children, or dung which she has gathered in her stable, or which the children have collected in fields or on roads, and which she has mixed with chaff and water and shaped into cakes of con- venient size for her fireplace. The bread which toasts close besides the fire, must be watched lest the burning fuel touch it, or lest it topple over into the flame. If its edges have been pressed together properly, the cake swells out like a balloon, as the moisture inside turns to steam. At this point, the cook lifts it out with her iron fire tongs, and bangs it against anything convenient to drive out the expanded air. The steps in bread making are listed on page 335. _ The women prefer wheat, or mixtures in which wheat appears, not only because of the better flavour, but because wheat flour is more satisfactory to work with. When making wheaten cakes, the housewife can roll them out with her small | roling-pin, after patting them between her hands. Cakes of other grains fall apart if she tries to make them equally thin. Corn cakes are a cross between corn- meal mush, and corn bread. Cakes of the little millet look like clay, before they are baked. They аге mealy and have а raw taste. Cakes of sorgum are brown with a purplish tint, and are slightly bitter, and mealy. Cakes of Italian and of Poorman’s millet are still less desirable. Wheat is always used for bread which is to be fried in deep fat. When fried, the bread is known as “puri” as distinguished from the toasted bread which is "chapati". Both "puris" and "chapatis" are 1955] WISER— FOODS ОЕ A NORTH INDIA VILLAGE 331 covered by the general term “roti”. The occasions on which “puris” are required are discussed under the ceremonial significance of food Parched grains provide a useful variation in village fare. The grain-parcher is kept busy most of the day when a fresh crop of grain is cut. At other times he keeps his furnace going late each afternoon. The farmer or one of his children brings the grain to be parched. The parcher has a clay furnace in which he burns leaves and twigs. In the furnace he heats clay jars of sand. He mixes the grain with the hot sand and keeps mixing them until the grain has popped or puffed, or browned. Then he sifts out the sand, and gives the grain back to the farmer, after taking out a portion as his pay. The grain thus roasted may be eaten as it is. Or the farmer's wife may grind it coarsely for his morning meal. If there is butter- milk, it is added to the ground, parched grain. If not, water is added, and a little raw sugar, if there is any. If not, there is always salt. The daily ceremonies which must precede any cooking make it impossible to prepare an early cooked breakfast. This ground, parched grain, called “satua,” solves the breakfast problem for the farmer, and perhaps for the children. The women and children usually eat cold food left from the day before, but the men scorn this. If the grain to be parched or roasted for “satua” is fresh, it goes straight to ‘the parcher. But if it is dry—as it is most of the year—the farmer’s wife must ‘soak it for twelve hours, and sun it for six hours before her husband carries it to the parcher. Corn, barley, or a combination of barley and gram or of barley and peas, are the most popular. A quantity is sent to the parcher that will supply the eee! with “satua” for several weeks, or perhaps, months. A prosperous farmer y carry a maund and a half (about one hundred and twenty pounds) to the жаалын s, while his poorer neighbour carries about ten pounds. And each hopes to have “satua” enough to feed himself for the same period. When the parched grain is brought home, the women pound it in the stone mortar set in the court- yard floor. This loosens the outer skins. Then they clean it, and grind a week's supply at a time. At harvest time there is a feast of parched foods in which the parcher has no share. While men rest from harvesting in the fields, or while they sit with their brotherhood after a strenuous day's work, they gather a pile of leaves or stalks and scatter the heads of green grain or the pods of legumes within the pile. They light it from a scrap of smouldering cow-dung, always on hand for lighting pipes. When the blaze dies down, they scatter the charred remains and pick out the smoking heads of grain. These they rub between their palms until the kernels are freed. Some are still green; others are charred black. But the combination is very agree- able. The women and children manage to get a share of the heads for roasting, from the stalks which the men bring home at the close of each day's work in the harvest fields. Peas, wheat, and barley are thus treated and eaten in March and early April. During the harvest at the end of the rains, the heads of both large and small millets are laid on the fire, and the green ears of corn are husked and turned in a blaze or in the embers of a fire of twigs. The corn is field and not [Vor. 42 332 ANNALS OF THE MISSOURI BOTANICAL GARDEN sweet corn, but it is good roasted. These popular products are available to everyone in the village, for at these seasons every able man has a chance to help some farmer. And those who cannot, plan to be present when generous neighbours indulge. The amount of parching done in the fields depends on the good will of the owner of the field. If he is miserly, or work is pressing, he may allow little time for rest | and feasting. But ordinarily, it is an accepted part of the day's routine. Other dishes in which cereals appear, are prepared for certain festivals. These come so seldom that the foods are not of great significance from the point of view of nutrition. In the eyes of the village housewife they are tremendously important and consume much of her conversation and time. They are at the end of this chapter, along with other festive dishes. Rice is treated differently from the usual cereals. For bread it must be husked with mortar and pestle, then ground. It makes a very soft dough. The cook keeps her hands moist while handling it. It is toasted like any other bread. Rice is used for bread only when there are no other grains on hand. - It is more often used in “khichri.” Water is boiled in a brass kettle, and rice and a split pulse are dropped in together and cooked. They may be in equal pro- portions, or there may be less pulse. A variation of this with a decidedly different flavour and "feel" is “да! bhat.” The rice and split pulse are cooked separately and combined when almost dry. The women of the family drink the rice-water. Sometimes rice is soaked and later fried in "ghi" (clarified butter) and salted. Oil may not be used for this. “Khir” is a dish which is limited to households where there is enough milk so that all is not needed for "ghi." Rice is boiled in milk and sweetened (one-half pound rice to about three quarts milk). Villagers say that there should be as much raw sugar added as there is rice. Just before serving, a little "ghi" should be added. A variation of this dish, which is available to a much larger number, is "Raskhir." The juice of the sugarcane is boiled, and rice cooked in it. It has a flavour which we think disagreeably strong, but village folk eat quantities of it with pleasure. А cross between these two, is a dish called “таһегі,” in which buttermilk and raw sugar are added to cooked rice. We can get puffed rice in the market town during most of the year. But in the village it only appears at the time of the festival of lights, in the autumn. Then the grain parcher buys the better variety of rice necessary and puffs it in his hot sand. Everyone in the village comes to him to get some. Other occasional uses of rice will be described under special dishes. 1955] WISER—FOODS ОЕ А NORTH INDIA VILLAGE 333 STEPS IN MAKING TOASTED BREAD Maize Process oa d Weight before | Weight after Bringing dung from drying pile and bring- | 20 minutes. ing water. Pounding with pestle 3 minutes .. : а. алы то аы ей husks, etc. | 2 minutes .. «= 72-25 13 05 Grindin à 1 hr. 5 min. Ib 13 oz. 2]b. 9 oz. Siftin E ОИ c ib. 9 oz. 2 Ib. 8 oz. Making into cakes 1. Мау ада gram. In this case, is only eaten by i 2. May be added to чит Same as ordinary, combined with potatoes. When cooked with ан we frying гаа а 15 — per and xed wich vegetables а nay Other at end. Same as ordinary, combined with potatoes. May add gram to filling. СРЕ [Vor. 42 344 ANNALS OF THE MISSOURI BOTANICAL GARDEN cooked, іп combination with vegetables. “Mahua” blossoms are eaten raw by the children. In case they are unusually plentiful, some are dried and cooked with vegetables later. "Bel" fruit and the wood-apple are usually roasted over a home fire until they burst. Pomegranates are rare. When they do appear, they are eaten raw. Guavas are eaten raw, usually when green. Wood-apples are also eaten raw. The sour fibre of tamarind pods is chewed by the children. Grown-ups soak the pods and make a sweet, cool drink from the liquid. Lemons are used in "achar"—pickle. The lemons are cut up, and ground spices are rubbed into the exposed inner surfaces. The spices used are cardamom, black pepper, fenugreek, and coriander. The lemons must remain covered with the spices for four or five days before being eaten. If prepared carefully, they will keep for a year. Plums are used to make a preserve. They are cooked, dried, and put into sugarcane juice. Mangoes are the utility fruit of the village. They are eaten raw, cooked with vegetables, and made into a variety of pickles and preserves. An oil pickle is made by cutting the mango almost in half and removing the stone. Aniseed, red pepper, and salt are ground and added to the fruit, and oil is added to coat the surface. A salty pickle is prepared by cutting the mangoes into four sections and removing the stones. They are put into a clay jar and sprinkled well with salt. Another form of mango preserve is made with sugarcane juice. Green mangoes are soaked and dried, or ripe mangoes are used, without cooking. They are dropped into a jar of strained, fermenting sugarcane juice, and left there for months or years. They are not ready for use until after remaining in the juice two months. Mangoes may be peeled, split, and dried after the stones have been removed. In this form, they may be added to almost anything which lacks flavour. A mango preserve is made by peeling the mangoes and rubbing them to a pulp with the spice-stone and roller. Salt, pepper, coriander and peppermint are added. This “chatni” (chutney) spoils quickly. Even the kernels of the stones removed from mangoes are used. The stones are split, the kernels taken out, boiled, dried and eaten. They may be roasted, if preferred, Animal products Milk and milk products are the chief forms of animal food in the village. Fresh milk is seldom used, unless there is an unusually large supply. When a cow first gives milk, the fresh milk is boiled, and clabbers almost immediately. This is eaten, with or without raw sugar. It is eaten during the next four or five days, as long as the milk is drunk after boiling. After this it is made into 7 ghi". In most homes all of the milk is reserved for "ghi". The milk is heated slowly in a clay jar, over a nest of smouldering dung-cakes. А little buttermilk is added, and it is set aside all night. By morning it has clabbered. At this stage it is called “dahi”, and may be eaten with sugar. In city homes it is often served. But in the village, it seldom stops here, but goes on to butter. 19551 WISER— FOODS ОЕ A NORTH INDIA VILLAGE 345 The “dahi” is put into the churn, water is added, and it is churned. When butter forms, it is taken out and put into another jar. The butter is never used, but is collected for eight days. The sour buttermilk which remains in the churn has a variety of uses. The men of the family like it with their "satua"—parched, ground cereal—in the morning. It may be used with vegetables in “reota”, de- scribed among vegetable dishes. Or it may be cooked with pulses in the dishes which we treat as meat substitutes. If there is not time to use it in any of these combinations, the men drink the buttermilk. Sometimes the children have a drink. The women do not drink it. The accumulated butter is heated in a deep kettle over the smouldering dung- cakes, or in the kettle for deep-fat frying over the fire-place. It may be heated from one-half to four hours, depending on the fire. Foreign matter which collects on the surface is skimmed off. en the heated butter is removed from the fire it is strained through a cloth. The clear fat which goes through is "ghi". It is regarded as the ideal medium in which to fry foods. Frying is an important process in village cooking, as it must be used for all foods served to guests and for food which members of the family must eat away from home, even though they eat it in their own fields. In the absence of ovens, many things are fried in "ghi" which we would bake. “Ghi” is also used in the preparation of a number of sweet- meats. In prosperous homes, it is an important ingredient in the special dishes given a woman two months before and a month after child birth. The Brahmans of our village eat no meat. They do not take advantage of the opportunity offered them by the Sacred Law, of eating meat as sacrificial food. Manu provides that “Не who eats meat, when he honours the gods and manes, commits no sin, whether he has bought it, or himself has killed (the animal) or has received it as a present from others.” V. 31. It may be that with the giving up of priestly duties, for farming, they have given up priestly privileges as well. If they were to eat meat, without sacrificial intention, they would be guilty of sin. “There is no greater sinner than that (man) who, though not worshipping the gods and manes, seeks to increase (the bulk of) his own flesh by the flesh of other (beings).” V. 52. The attitude of our Brahmans and of the orthodox Hindus who try to follow them is expressed in the Law. “Meat can never be obtained without injury to living creatures, and injury to sentient beings is detrimental to (the attainment of) heavenly bliss; let him therefore shun (the use of) meat.” У. 48.... "Having well considered the (disgusting) origin of flesh and the cruelty of fettering and slaying corporeal beings, let him entirely abstain from eating meat.” 9. It is this attitude which keeps meat out of sight in the village. The Hindus who eat it, do their buying in the Muhammadan quarters or in their own shadow stables, as quietly as the excitement of bartering allows. The goldsmiths show themselves unorthodox in many ways, one of which is meat-eating. They do not talk about it but every one knows that they eat flesh. One or two carpenter families eat meat, as well as several families below them in the caste scale. They [Vor. 42 346 ANNALS OF THE MISSOURI BOTANICAL GARDEN all limit themselves to goat meat. The cow is revered too highly to be thought of as a source of food. Muhammadans eat as much goat meat as they can afford. Living in a Hindu community they have much the same attitude as their Hindu neighbours toward beef-eating. Hindus and Muhammadans prepare meat in the same way. They cut it up into small pieces, sear it in fat, add water and spices and cook it. They eat it with rice or bread. Meat dishes were the only ones denied me by my little Hindu cooking-teacher. She warned me that she would never touch our experimental fire-place or utensils if I were to defile them with meat. She knew that we ate meat, but that was not her responsibility, as long as it was kept out of her field of activity. Sweepers and Dhanuks, two castes of outcastes, are the only ones who will touch pork. After watching the scavenging activities of village swine one is not surprised that their flesh is repulsive to most villagers. Only those who are brought up with the consciousness that they are themselves unclean, could eat it. The men do the butchering and divide the pig into shares. The women cut it up still further, into small pieces which can be fried and spiced, and eaten with bread or rice. The head is the most desirable share. Fish is not classed with meat. Many eat it who abhor meat. We were sur- prised at the number of castes represented on the banks of drying ponds, waiting to start their "basket-fishing." On the occasional fishing days in the spring, the whole village smells of frying fish. Only Brahmans and those others who pride themselves on keeping the letter of the Law, abstain. Eggs are prohibited almost as much as pork. The fact that eggs hold potential life makes it impossible for orthodox Hindus to eat them. In so doing they would be destroying life. A further objection to the eggs is the objection to the chickens that lay them. Chickens are kept by outcastes. No one feeds them and they wander about the lanes near the outcaste section as scavengers. They are scrawny and offensive. The outcaste families who own them eat the few eggs which they lay. E—FESTIVAL DISHES Special dishes are reserved for days of religious festivals, and weddings. From the passing of one festival to the coming of the next housewives and children, and often the men, recount the special dishes they have had or are going to have. A meal of toasted millet bread and stewed pulse tastes better if one thinks while eating it, of sweetmeats fried in "ghi". The pleasure of anticipation is granted to all, as those with much share with their dependents on these occasions. "Puris"—bread of finely ground wheat flour, fried in "ghi"—appear on all feast days. The other articles vary. “Holi” is the great festival of the spring. With its coming there appear "gojha", “уаѕѕе”, "chandia", “puri”, *kachauri" and "halwa". For "gojha", wheat flour is ground fine, sifted through a coarse cloth, kneaded and shaped into cakes as for bread. The cakes are dried for several hours, then fried in deep fat. When fried, they are pounded into small pieces and dried 1955] WISER— FOODS OF A NORTH INDIA VILLAGE 347 again. When well dried, they are ground in the stone flour mill, and put through the thread-strung sieve. The sifted material is combined with black pepper and crude sugar and again sifted. Raisins and "chiraunji" (small nuts) and cut up fresh cocoanut are then mixed with it. This is the filling. The crust is made of fine wheat flour, kneaded and rolled out as for small "puris". Each cake is flattened out in the hand and filled with the mixture. Then one side of the cake is folded over and the edges pressed together, very much like small turnovers. As rapidly as the turnovers are prepared they are fried in deep fat. ““Yassee” is made of rice. The rice is husked, soaked, and pounded fine. It is then beaten up with a little warm "ghi" or oil, and raw sugar is added while beating. "Chandia" is made of split “urd.” The “urd” is soaked and its husks removed. Then it is rubbed to a thick paste, together with asafoetida and salt. This paste is spread on a wet cloth in a small circle (about four inches in diameter). A hole is cut in the centre, and it is lifted from the cloth, fried in deep fat and then added to buttermilk. "Puri" and "kachauri" have been described under cereal foods. “Halwa” is prepared by making "puris" and grinding them up as for “gojha”. One pound of this is browned well in one pound of "ghi", and combined with one pound raw sugar. A little water may be added if necessary. It is like a stiff mush, patted out and cut into cubes. Raisins may be added. In town there are many variations of “halwa”, such as adding grated carrot, nuts and cocoanut. But in the village, the simple form is adhered to. In the spring, just before the last field of the winter crops is cut, the farmer's wife makes “sira”. For this she browns wheat flour in "ghi". She melts raw sugar in a small quantity of water and adds hot water to it. Then she adds the browned flour and raisins and cocoanut and serves it. The wedding season comes in May and June. Wedding parties go from our village to be entertained in other villages and wedding parties from other villages аге feasted by us. There are "puris", "kachauris," “halwa”, and fried vegetables and pulse dishes, all of which have been described. А special dish is called “tikian”. t is made of gram flour, mixed with enough water to make a stiff dough. In a deep kettle, water is heated—about three quarts water to one pound flour. The stiff dough is made into little flat round cakes, a little larger than the palm of the hand. These cakes are dropped into the boiling water and boiled about twenty minutes, until tough. They are then lifted out and spread on a tray to cool. Each cake is cut in half, and then into cross-wise strips. The gram flour inside looks packed and uncooked. Salt, red pepper and turmeric are ground and mixed through the strips. The sticky surfaces hold the spices. Two ounces of "ghi" and a pinch of asafoetida are heated in the deep-fat frying kettle and the spiced strips are fried in it. When the strips are well browned, water is added almost to the top of the kettle. It is cooked until most of the water boils away, when salt, cloves, carda- mom, and black pepper are added. Те looks like noodles in thick brown gravy and tastes better than it looks. [Vor. 42 348 ANNALS OF THE MISSOURI BOTANICAL GARDEN Two festival dishes go through the same steps, but one is sweetened and one is made salty. The sweet cakes are "gul-gula" and the salty ones are "pakauri" To a pound of wheat flour a pound of melted raw sugar is added, also a small ginger root and a large cardamom. Water is added until the batter is like that for ginger bread. The frying kettle is half filled with "ghi" or oil, and heated. A handful of batter is dropped into the fat. The cakes thus dropped puff out and turn a rich brown. They are delicious, like doughnuts, with a strong molasses flavour. For the salty cakes, salt and asafoetida are added to the flour and the same procedure is followed as above. We prefer the salty cakes. At wedding time mangoes are at their height, and appear in many forms. In addition to the preserves described, a special mango dish is prepared, called “Атіуап chinghora”. Turmeric, pepper, salt and coriander are ground together. Oil is put into the frying kettle and the spices are heated in it. Gram flour and mangoes have previously been peeled and stoned. When it is ready to remove from the fire, heated asafoetida is stirred into it. In August, during the rains, there is a festival which is a happy one for the girls. Most brides, and often women long-married, go to their father’s homes for this festival, “Raksha Bandan”. The dish most definitely associated with this festival is “samai”. Every woman makes it. She makes a stiff dough of the finest wheat flour she can grind. She inverts a large clay jar, and on its bottom she rolls the dough into a long thin string like vermicelli. The dried branch of a tree is stood on a roped cot, and the vermicelli is draped over it. When it is dry, the whole family enjoys it, mixing it with melted raw sugar. In the Autumn comes the Feast of Lights which is another occasion worthy of feasts. Puffed rice, plain or sweetened is the special dish for this festival. And there is "lapasi". A thin sweetened flour batter is poured into hot fat and stirred until it is a smooth thick brown sauce. This is poured out on a tray where it becomes firm almost immediately. It is patted out and cut into squares for serving. "Lapasi" may be made of wheat, maize, or barley, and may be sweetened or salted. Another dish associated with the Feast of Lights is "pitaua". Spiked millet or great millet is ground, and combined with raw sugar and sesame seeds. It is worked into a dough, and shaped piece by piece into round flat cakes, about six inches in diameter and three-eighths inches thick. These are fried. Maize may be used in place of the millet, in which case salt is added instead of sesame. “Halwa”, described under the Spring Festival, also appears on the Feast of Lights. t any time during the year which is proclaimed auspicious, a man may give what is called “Katha”. Не calls in a priest to read from the sacred books, and invites all relatives within reach to come and listen, and feast. By giving a ‘Katha”, a man gains much merit. At the feast he serves the guests mango pre- serve, with “Channa mirt". “Channa mirt” is a mixture of cow's milk, leaves of the holy basil tree, Ganges water, sugar, curds, and gram. It is served in little clay cups, or poured into the hands of guests. At the same time, two wheat flour dishes are served. One of these is simply wheat flour mixed with hot "ghi" and 19551 WISER—FOODS ОЕ A NORTH INDIA VILLAGE 349 then with raw sugar. For the other, wheat flour is made into a stiff dough with water and melted raw sugar. The dough is shaped into a long slender roll from which small slices are cut and fried in deep fat. Among the special dishes of the village must be included those prepared for women before and after their babies are born. One such dish is “mewa kilaurna", given to a woman during the early nursing period. Cocoanut, almonds, pistachies, raisins, dried dates, are all chopped together and heated in "ghi". Gum from the gum acacia tree is added. Raw sugar and water are boiled together until it forms a thread. It is stirred into the prepared condiments and the whole is beaten and shaped into balls. It resembles Persian sweets. This is only available in the better omes. Another special food given a woman after child-birth is “harira”. Aniseed, coriander, turmeric, fennel and *ajwain" (caryota copticum) are ground together. Equal weights of these and “ghi” are combined and added to their total weight of raw sugar. А little water is added. Both of these sweets are supposed to add strength, and increase the flow of milk. IV—FOOD PRACTICES Certain practices in regard to food have become customary in Karimpur. They are the product of the experiences of generations. Some of them are common to all Hindu communities, others are limited to the Ganges area and still others are found only in our district. There are two sets of these practices. One set is asso- ciated with the good, and the other with the harm which may come to a person through food. Both have the support of custom, but those associated with the good which food may do, are such a simple part of daily living that one is hardly aware of them. They encourage the distribution and use of foods shown by ex- perience to be beneficial. Those practices associated with the harm which may come through food receive much more emphasis. They take the form of food taboos, supported by religious law and ceremony. A—FOOD TABOOS Blunt, in his discussion of the caste system of Northern India, lists seven kinds of food taboos. (10). “The food taboos of Hindu life complicate it to an almost incredible degree: and as has many a time been pointed out, prevent ‘the growth of the good fellowship which we are wont to cement at the dinner table.’ “Hindu food taboos are of several kinds: 1. The commensal taboo—which lays down the persons in whose com- pany a man may eat food. 2. Тһе cooking taboo—which lays down the persons who may cook the food that a man eats. (10) See References. [Vor. 42 350 ANNALS OF THE MISSOURI BOTANICAL GARDEN 3. The food taboo—which lays down what kinds of food a man may eat. 4. The eating taboo—which lays down proper ritual at а meal. 5. The drinking taboo—which lays down the persons from whom a man may take water. The smoking taboo—which lays down the persons whose pipe a man may smoke, and in whose company he may smoke. 7. The vessels taboo—which lays down the nature of the vessels that a man may use for eating, drinking, and cooking. To a new-comer, these and related taboos seem an absurd waste of time. But as we observed the measures which we took to protect ourselves from disease in community with no sewers, no screens and no quarantine rules, we admired the skill with which high caste Hindus had formulated rules with all the prestige of ceremonialism, which served to protect them and their progeny from the dangers that apparently lurk in food. We tried to explain our reluctance to eat in village homes in terms of flies, and small-pox and cholera germs. But our village friends pleasantly ignored our terminology and explained our practices to each other in their own familiar terms of caste rules. They were relieved to find that although we mingled recklessly with all castes, we still maintained certain prohibitions. This discovery made us more comprehensible to them. The Brahmans at the top of the caste scale, are meticulous in their observances of the religious law. This includes bodily cleanliness. Even our busy Brahman farmers try to bathe every day. А man stands beside а well and pours water, freshly drawn, over his body, while he repeats prayers. Не wears his loin cloth while bathing and changes it for a dry one if he has it, after the ceremony is com- pleted. After this bath he is ready to eat. Мо one is supposed to touch him until he has finished his food. Bose, in describing Hindu customs in Bengal, writes, “When Hindus sit together to eat, their seats are generally wooden planks ог ‘Kusa’ grass seats placed on the floor. These seats must not touch one another and no Hindu will touch another person who is also eating by his side. Among Brahmins this rule is very strictly observed. Мо one except the mother and the wife can touch a Brahmin while he is eating; if he does, he will stop eating and get up and wash his hands and mouth. This is a most barbarous custom. Some Brahmins do not allow their wives to touch them while they are eating. Аз the mother is re- garded as a goddess by all Hindus they make an exception always in her case." (11) The Brahman wife, who serves her husband, has also bathed or at least washed her hands before preparing his food. S. N. Jafri, Deputy Director of Public Information of the Government of India, gives an interesting picture of eating in а Hindu home of the United Provinces. (11) See References, 1955] WISER— FOODS ОЕ A NORTH INDIA VILLAGE 351 ""The system of taking food by the Hindu pepin or for the matter of that all Hindus except non "0 о аге influenced Ьу modernism, is и “The in a part of the floor of the а erc is called ‘chauka’ and is meant as a place cmd e 22 food. Before Ming ара take аа all their clothes except the bine or the rouser сї, о ес ey g che caste. In very cold weather, while no they жұға d a blanket over them The p on *ch i i d is i auka' at once, but only the ig ages quantity. The woman of the family who is the principal cook at the time, or officiates in the kitchen, serves the food for the eater or eaters. She sits near the fireplace, where the * in is situated. The man squats, in which case he has mooth boar int a kind 6 the bis manage to carry it to "hd poet d diia they can eat very фондю ith one han “А Hindu at a time " eating must not be touched by one of inferior caste or by a non Hindu for, if he were, he would immediately rise and not take еее "e у if he had to go тшен food the whole day; he would also throw out Pec e й thoi in his mouth. He would never eat food prepared by a non-Hindu evi indo an inferior caste; ET to some the shadow of a low caste man falling on ах Е pi s unfit to eat. А Hindu's food is of two kinds, 'kachcha € and 'pa ana.’ All the ее tioned ceremonies are required ~ ‘kachcha khan No нина жк картая attaches t dry things, such as flour and grain, and none also me > (сона. Such things a man of higher caste can receive from a man of lower gie 5112) Before one can attempt to understand the intricacies of the food "Thou shalt nots" of a Hindu community, he must know the distinction between “kache and “pakka” food. Ghurye in his presentation of Caste and Race in India, gives one of the clearest statements of this distinction that I have found. Я дня; on feeding and social intercourse. There are minute rules as to what sort of food or can be accepted by a person an nd from what castes. But there is very grea зані in this matter. The practices in the matter of food and social intercourse divide India nto two broad belts. eign os proper (our section of Indis) кес сап ivided into ћу a rule a man food unless it is pre by a ro casteman, which in actual Mio spia means a а mem | 4 own endogamous group, ум к» caste or sub-caste, or else by his Brahmin ael or spiritual guide. Las in practice t castes seem to take rom has it three Ба require по ре than thirteen rs ris ок i fe the ‘Pakka’ food, it may be taken by a Brahmin at the bands of some of the only. A man of се caste cann : ot accept "Kachcha' food from one of the raving знай pe latter may Е imself i similar food offered by a member of one of the castes accepted to be higher than his own. 13). (12), (13) See References. [Vor. 42 352 ANNALS OF THE MISSOURI BOTANICAL GARDEN Blunt gives the explanation of the distinction between the two kinds of food. “The Hindu draws a distinction between ‘Kachcha’ food, which i ked in water, and ‘Pakka’ food, which is cooked with ‘ghi’ (clarified p This ditas cios depends o on the principle that ‘gh?’ like all the — of the sacred cow, protects from impurity: and since such protect ion is the object of a aboos, this convenient Жент? enables the Hindu to be less particular in the case of vies сэй of ‘kachcha’ food, and to relax his restrictions accordingly." (14). In our village *kachcha" food is an exclusive, family-only product. But оп certain definitely fixed occasions a Hindu may eat “pakka” food, prepared by others than members of his own family. The persons who may cook this food are all listed in the rules of his particular caste. In this way, although he ventures to dine abroad he is still somewhat protected. The persons who may cook the feast of “pakka” food, which he shares must be his equals or superiors in the caste scale. This guarantees that they will be at least as clean as he is, if not cleaner, in the care of their bodies and their food. If he is a Brahman, he is supposed to feast only at the house of a Brahman, with the understanding that the food has been prepared by members of the Brahman household, Brahman servants, or servants of water- carrier caste. He takes for granted that any sweets too elaborate to be prepared at home, have come from a bona fide "halwai," maker of sweets. If doubt arises about the source of any article of food, the Brahman guest does not hesitate to make pointed inquiries of his host about it, before touching it The following is just a short section from pages and pages of laws stating from whom a Brahman shall not accept food: “The food of a king impairs his а Brahman’s vigour, the food of a Sudra his excellence in sacred зване. the food of а Жө a "his ТІ ches of a leather-cutter his fame. The food of an artisan destroys his offspring, that of a washerman his (bodily) strength.” Code of Manu. Iv. 21 The Brahmans of Karimpur treat their goldsmith neighbours as being worthy of becoming their hosts on special occasions. At the lower end of the caste scale are the Sudras. They can eat "pak food or drink water from members of the long line of castes above them. But better than any other food for the Sudra is that which a Brahman has touched. Sudras who live according to the law, shall each month shave (their heads); their ә = purification (hy (shall о the same as that of Vaisyas, and their food the fragments of an Ary: 140. Below the Sudras are the outcastes. They haunt the weddings and other feasts of any caste, ready to hurry in ahead of the dogs and crows, to gather up the scraps left by the guests on their leaf plates. These left-overs they carry home in baskets or ends of scarves or loin cloths. They eat what they can at the time, and dry out the rest to be eaten later. No one worries about protecting the outcastes from possible contamination. (14) See References. 1955] WISER— FOODS OF А NORTH INDIA VILLAGE 353 Our years in India convinced us that experience had taught the law-makers well—at least for their own benefit—in the matter of "kachcha" and “pakka” food. We found ourselves unconsciously following their practice. We learned that we could safely go to a wedding feast in a village home, where we would hesi- tate to eat the daily food. This was because the wedding foods were all “раККа”— fried in hot "ghi" and served at once. It was not the sacred source of the "ghi" which influenced us, but the fact that it was sizzling hot. The weighty distinction between "kachcha" and “pakka”? food complicates food preparation for the village housewife. When her husband and sons eat in the family courtyard, everything may be "kachcha." She makes the ordinary cereal cakes and toasts them, and she may cook “вар” (boiled green vegetables) or pulses. If one of the men wants his food sent to the fields where he is working, she must prepare “pakka” food for him. She cannot send boiled food to the fields, as it would be “Касћећа," and “kachcha” food may not be eaten in the fields where there is a possibility that the food or the eater might be touched. So she must fry the vegetables, and she is supposed to fry the cereal cakes in deep fat. These fried cakes are made only of pure wheat flour, which she cannot afford to use often. Neither can she afford large amounts of “ghi” or oil for deep fat frying. So she toasts cakes of mixed flour in the usual way, and sprinkles a little "ghi" or oil over those which are to be carried to the worker in the fields, and pronounces them "pakka". 'Thus she keeps the law with a few drops of "ghi". But her problem does not end here. If she should prepare her “Касһсһа” food such as toasted cakes and pulse first, then any food which she prepares after it becomes “kachcha”, even though it would seem to be “pakka,” by being fried in "ghi". It can then be eaten only by members of the family within the courtyard. If she has "pakka" food to prepare she must finish it first, and set it aside in a prescribed place, and after that, cook the "kachcha" food. She finds it easier whenever possible, to have everything fried in "ghi"—"pakka", or everything "kachcha". "Kachcha" food must be eaten immediately. If a woman discovers that she has prepared more of some “kachcha” food, such as boiled vegetables or pulse, than will be used at noon, she can mud plaster a corner near her fire and set some of the food there to be kept for the evening meal of the same day. But she must do this before any one has been served. If any "kachcha" food is left after a meal has been served, the women and children may eat it later on in the day or even the next morning. The men will not touch it. One appreciates this precaution after observing what might walk over food standing on the floor, even though it be in a special corner, where no human treads. Drinking water is protected from contamination by a number of laws. Only three Brahman courtyards in Karimpur are large enough for wells. The other wells—19 in all—are along the lanes and paths. Each well may be used by certain families, and must not be touched by others. Brahman women and other women kept in seclusion cannot leave their courtyards. Women of water-carrier caste draw their water for them. Whoever uses the well lets his or her jar down into [Vor. 42 354 ANNALS ОЕ THE MISSOURI BOTANICAL GARDEN the water. This makes it important that unclean castes and strangers be prohibited from using a high caste well. We were not allowed to get our water from any of the village wells, as the boy who carried our water was a leather-worker, and the well of the leather-workers was too far away to be used. So our water came from - an irrigation well in our grove and sometimes from one in the fields. We were not as particular as our village friends about who shared our well. In place of caste prohibitions, we resorted to thorough boiling. When cholera broke out in the village one summer, the district sanitary department sent out men armed with potassium permanganate. They did not limit themselves to the wells used by contaminated households but treated all wells, for safety. Someone told our people that the permanganate was poison. Аз we went the rounds to re-assure them we found men laboriously trying to empty the wells of the red water, a jarful of water at a time. If the permanganate had been somehow related to ceremonial purification or if waterboiling could have been substituted and given some spirit- appeasing significance, it would have been accepted readily. But being introduced officiously, without explanation, it was upsetting. Dishes in which food is prepared or served are carefully guarded. They are used for the immediate family only. Red clay dishes and clay cups without handles are used for guests and thrown away. They are never used a second time. This is as satisfactory as our rinsing with boiling water and it is much easier. At least so our young sons thought when we washed dishes after our own guests had de- parted. If clay dishes are too expensive, a host may serve his guests on plates of large leaves bound together with twigs. At a feast, nothing is ever passed around. The host or his sons serve the men, and his womenfolk serve the women, after the men have finished. As a server moves from guest to guest he is careful not to touch anyone with his garments and he is still more careful not to touch his hands, or the big iron serving-spoon against the plate of a guest. This would render the spoon, himself, and the serving ish in his hand unclean Food dropped on the floor must not be eaten. The advantages of this are obvious, especially where the floors are the earth, and children and animals relieve themselves anywhere. The villager has his own explanation, expressed in an ol saying, "If edibles at the time of eating drop down they are taken by the shadowy spirits and should not be picked up and eaten. If you do eat, you are sure to be possessed by the spirits who will harass you much.” (15) There are other food regulations, the burden of which rests on the women. These have to do with the ceremonies of food preparation. In the morning as soon as her grinding is finished, a woman plasters her little fireplace and the floor all around it with fresh clay. The plastered floor space extends about three feet out (15) See References. 1955] WISER— FOODS OF A NORTH INDIA VILLAGE 355 from the fireplace. The clay she uses is not ordinary clay. It is collected from the edge of a particular pond, by the children of the family if she is too high in caste to get it herself. She mixes the clay with a little water and a bit of cow- dung in her mixing bowl, and plasters it on the surface of the fireplace and floor with a rag kept for the purpose. When the plastering has been done, the fireplace and the plastered area around it are "purified". No one must go near them. Even the children of the family know that they must not approach it. It was not until I had my own fireplace that I appreciated the importance of this. In the absence of a kitchen table, it is necessary to have some place where one can put eatables and utensils and know that they will not be walked over. My fireplace was under a tree in our grove, with children playing all around, and farmers stopping to add their advice to that of my instructor. And only once did anyone approach this sanctum. That one was a little toddler of washerman caste. My teacher rushed at the child in a fury, and a torrent of reminders of her unworthy birth. The child fled, and we had no lesson that day. The fireplace had to go through a fresh mud- plastering before it could be used again. While the fireplace is drying, a woman may perform any of the tasks which do not demand strict purity. She may churn or make her fuel cakes of dung or chop vegetables. When everything is collected for the meal—flour, water, pulse or vegetable, and utensils and spices—she bathes. This she does by standing in a sunny part of the courtyard and pouring water over herself and her clothes. Then she changes to a “‘dhoti”, a scant garment made of one piece of cloth draped around the body and over the head. She never allows the village washerman to touch this garment, but rinses it out at home herself. The washerman’s touch would in- directly defile her, and the food she touches. Women lower than the goldsmiths in caste, let handwashing take the place of the bathing ceremony. When bathed, or washed, the woman enters the plastered area, by her fireplace, taking all of her materials and utensils with her. From this moment until after the men have been served, she must not leave it. The other women help with the preliminaries, but only the one who is ceremonially and physically clean, may attend to the final cooking. She alone may serve the men their food as they come and sit near her. In Karimpur the family never eats together, and even the men seldom eat together. Each one comes in when he is free, eats his meal in silence, and leaves the courtyard at once. The cook must wait until every man has been served before she is allowed to leave her post. Thus does a woman do her share in protecting the food which she gives her menfolk. For her own food, there are many less rules. She and the other women eat what the men leave. While they eat they are free to sit anywhere within the courtyard. Women are of little im- portance in the village interpretation of the Hindu code. [Vor. 42 356 ANNALS OF THE MISSOURI BOTANICAL GARDEN B—NUTRITION PRACTICES The second set of food practices, mentioned at the beginning of this chapter, are much less formal and less obvious than the ones just discussed. They have risen from experience which has demonstrated the good effects which apparently result from the eating of certain foods. Most of these practices are simply an ex- pression of physical desire for certain foods. I have called them "nutrition prac- tices," as they express the village way of meeting body needs. First there is the custom of sharing raw products. If there were barbed wire fences around orchards and fields our poorer families would suffer. As it is, each family of low caste with low food supply is dependant upon, and attached by custom to one or more high caste, prosperous families whom they call their patrons. When leguminous plants and mustard plants are young and their tops are tender, children of poor homes are allowed to pick them in the fields of their patrons. They gather enough so that their families have what they want to eat fresh, with some left over to dry and store for later use. The children also go into the fields of patrons and dig up certain of the green leaves which grow without tending, close to the ground. There are a number of those listed in the food table (p. 372). With a diet based on cereals and legumes, supplemented with vegetables which are chiefly of the pumpkin and tuber varieties, villagefolk need the minerals and the vitamins which these green leaves offer. And nature and custom have com- bined to make the leaves available, not just to the prosperous but to all. The same is true of sugarcane. From the time it is ripe until it is cut and pressed, one sees children and men of all castes sucking and chewing at the stalks. In so doing they are supplying their bodies with the calcium of which they get little in their low-milk diet. In the cane juice there are also the calories which they need. At harvest time sharing reaches its high point. Every man and boy in the village has a chance to work in the harvest fields. He may be a carpenter, a potter, or a leather-worker the rest of the year, but at this time he turns farm-hand. The untouchables are the only men excluded. We asked why? One of the farmers explained, "When they work just a little while, they perspire and then they com- plain because they are tired." We thought at first that this was an excuse for keeping the untouchables out of the fields. One expects to have real reasons veiled. But after seeing the untouchables at lighter field work, and their quick tiring, we agreed that the farmer had expressed his actual thought. А Multan proverb says, “Не that eats a ‘ser’ (from 2 to 3 pounds) works like a lion, but he that eats only a quarter of a ‘ser’ works like wood-ashes.” And we wondered if this might not help explain the seeming laziness of the untouchables who have no fields of their own and who get only a meagre share from others. The men of all castes above the untouchables, and some of the low caste women, work daily while the harvest lasts. And every night each one carries 1955] WISER— FOODS OF A NORTH INDIA VILLAGE 357 home a headload of grain from the field just cut. This gives them varieties of grain and pulse which they would not otherwise have. At harvest time, women who cannot work all day, follow the harvesters, gleaning. A woman may glean in any of the fields belonging to her husband's patrons. In a few hours she can gather up several pounds of heads of grain— grain which would otherwise feed the crows. At harvest time representatives of families of craftsmen and serving castes appear at the fields of patrons at the proper time, to claim their shares. Even the outcastes have rights at this time. Мо questions are asked. Everyone knows whose duty it is to give, and whose right it is to receive, and the proper amount. The ethics of this system of payment may be questioned, but it assures the low caste and outcaste families of a variety of cereals and pulses which they would not otherwise get, either from their own fields or from the bazaar. In the fruit crops, we find a high degree of sharing. As has been noted, mangoes may be gathered by any member of the village, in any grove of the village. And mangoes are apparently their chief source of Vitamin C. The tamarind pods are available to all. And long ago Sanskrit medicine found in the tamarind, a quality now recognized as anti-scorbutic. The children are the ones who let their physical needs direct their foraging. None of the edible fruits miss them. Тһе wild ones which they eat, we know little about. But we feel sure that they must supply some Vitamin C, and calories, and perhaps some of the iron which they need. The gum from the gum acacia tree of which they are particularly fond, must add calories, if nothing more. We tried most of the strange fruits which they brought us to sample. A few were delicious, but we would not recommend many of them for flavour. The only offering which we absolutely declined was something which they scraped from the trees of our grove. It was apparently the deposit of mud made by white ants to serve as their passageway up the side of the tree. What the children got from it which was satisfying, I cannot imagine. When the outcastes butcher a pig they treat the head as the choicest portion. It reminds us of the boar’s head, praised by our own forefathers. The outcastes have not been told why the pig’s head appeals to them especially. They only know that they like it. Here, as with the children and the fruits, nature guides them. The people of Karimpur have learned to save and eat many things which we, with our plenty, have wasted. It is comparatively recently that we have been taught that the water in which potaoes and certain other vegetables have been boiled is valuable. We try to save some of the water for soups. In Karimpur the water which is added to vegetables is so small in quantity that little or none re- mains when the vegetable is cooked. What does remain is eaten along with the vegetable. My village friends were horrified when they heard that in some places this water is thrown away. [Vor. 42 358 ANNALS OF THE MISSOURI BOTANICAL GARDEN They hate to throw anything away. When they cook vegetables, they never remove the skins, unless too tough for chewing. When their grain is threshed some of the finer husks remain. These come loose during the grinding of the grain into flour and are separated in the sifting. But the women are careful not to lose them. They put some back into the flour and save the rest to be mixed in with other flour which has less roughage. When I tell them that in America we зер- arate bran, and sell it in packages apart from the flour, at a price which seems fabulous to them, they think that we must be very stupid, or at least shockingly extravagant. en a woman prepares a batter of any of the buttermilk or pulse mixtures in her mixing basin she does the stirring with her right hand, and scrapes the basin clean with the same hand. But she is not satisfied with this. She pours water into the mixing basin, washes the basin and her hand and pours the washed product into the frying pan or jar or other container along with the original preparation. As nothing is measured exactly, this additional liquid makes little difference. And the cook prefers to have it in the soup, rather than in her open mud drain which is already messy and alive with flies. As I walk through the village lanes on a festival day, I meet the women of serving castes, and outcaste women, going from door to door among their patrons. At each door they stand until one of the women of the patron's household tosses out several "puris" and the particular cakes or sweets associated with the occasion. If the serving woman thinks that she has not received her due, she says so, and stands talking about it until the women of the house give her something more, to get rid of her. ТЕ they think that she wants more than she should have, they shout at her until she is convinced that waiting will not profit her, and leaves. This sort of tumult alarmed me, until I learned that it was a perfectly friendly way of settling the amount of payment due. А definite payment in cash would seem more satisfactory to us. But cash, in the village is not as useful as food, especially the kind of food which is distributed on festival days. In the homes rom which these low caste and outcaste women come, there is no wheat except that distributed at harvest time. They own no cows or buffaloes, and therefore have no "ghi". Meanwhile, there is plenty of wheat and "ghi" in the prosperous homes. And this method of handing them out in the form of festival cakes, as- sures the donors of religious merit, while giving the recipients and their children something which their bodies need. There are sometimes other rare things used, to fill the cakes, such as cocoanut and dried dates. "These help still further to meet neglected needs. It seems like an unpleasant welfare method to из. But both parties regard it as an exciting and profitable procedure. Life would be dull without the anticipation of festivals. And as yet the West has not discovered a method of caring for the unfed, success- ful enough to be offered as an improvement. 19551 WISER—FOODS OF А NORTH INDIA VILLAGE 359 Among villagefolk there is a conviction that certain foods have special qualities. Some foods are "heating", while others are "cooling." We noticed that patients almost always asked what food they should eat. Knowing little about their foods, we hesitated to advise. But there was always someone around ready to speak with authority. If the patient suffered from toothache or earache, or a cold, the term "heating food" almost always occurred somewhere in the lengthy advice. If the patient had fever, "cooling food" always occurred. The reasons also seemed important. In cold ані when а person had н difficulties, "heating foods" were advised. In similar cases in hot weather, "cooling foods" were recom- mended. We collected a list of the foods which we heard mentioned in these two classes. “Cooling foods" include: buttermilk; sugarcane juice; bread of barley flour; tips of gram plants; "raskhir"; a combination of rice and sugarcane juice; radishes; carrots; Egyptian arum, or dasheen; and water-chestnuts. "Heating foods" include: bread of spiked millet; "dal" of pigeon pea; raw sugar; milk; potatoes; corn and sesamum oil. As I recall some of these foods, in the setting of an Indian hot season, I have reactions towards them, similar to those suggested by the villagers. But I have no adequate explanations of them as two distinct classes. There is no one quality running through either class which might explain its desig- nation. Further analysis of Indian foods may disclose a reason. At present, I simply offer the lists as demonstrating the effort of village folk to express the possible effects of food. Another class of foods accepted in the village is that of "strengthening" foods. These are supposed to give superior strength to the already strong, or to restore strength to the weak. The goldsmith brothers who act as village wrestlers, and are the acknowledged "strong men", eat almonds and pistachies, both of which are luxuries from the Mainpuri bazaar. While in training, they give up meat, and add milk to their usual diet. Strength-giving foods are considered important for expectant and nursing mothers. Two dishes reserved for them, “mewa kilaurna" and “Harira”, are described in the section on festival dishes. These are too expensive for most homes, Where simpler festival dishes of wheat or gram flour, rich in "ghi" and raw sugar, are substituted. Cocoanut and "phaphola" (nymphaea lotus) are included if possible, even though they must be begged. or anyone who has become very weak, without particular reason, mixtures like the two following are recommended by the local practitioners. Three drams of roots of the okra, and twelve black pepper corns, ground together. This is one dose. Or, one green seed pod of the gum acacia tree and ten black pepper corns. Grind and make into sherbet, with one € of water. Sweeten to taste. Drink this every morning until strength return: Sunshine is an acknowledged ising in the village. Every morning good Brahman farmers chant hymns in its praise. We agreed with them, that but for the sunshine we would all have died. On several occasions, we found boys with smallpox scabs still peeling, riding with others on our boy’s wagon or tricycle. We [Vor. 42 360 ANNALS OF THE MISSOURI BOTANICAL GARDEN washed the playthings with disinfectant immediately. But we had to trust to the sun to save us from the germs which had had time to attach themselves before the discovery was made. In the rainy season one appreciates the sun, by its absence. After a few days of heavy clouds and rain, the village lanes are like pools, fed by the streams which trickle from open house-drains and privies. The smells which rise up to greet one are nauseating. А few days of bright sunshine clear the pathway and at- mosphere. The streams still trickle but they dry up before they become objection- able. In this discussion, however, we are more interested in the sun's light as a source, or producer of, Vitamin D. As far as we know, it is the villagers sole source of Vitamin D, except for a small amount which he may get from green leaves. I have heard Westerners inquire, “If sunshine is beneficial, then why aren't Indian children large and stronger than they are?" This is demanding too much of sun- shine. It cannot be expected to correct all diet defects. I have heard American es reply, that in rural India sunlight does well the work which they look to o do. If our village children depended upon diet alone for adequate supplies of ао and phosphorus, and for proper mobilization of the supplies, cases of rickets would be much more general and extreme than they are. I have found no definite records, beyond that of the work of Dr. Hutchinson and Dr. Shah. “А recent study of rickets in India by Hutchinson and Shah is clearly illustrative of the preventive role of sunlight. Surely there should be no rickets in India. Yet the infants of the top caste Hindus almost always develop extremely severe rickets. Owing to the religious custom of ‘purdah,’ which compels their mothers to live beyond the public gaze, these infants are practically never taken out of dusky rooms. In contrast the infants of Hindu laborers who cannot afford too expensive a religious detail practically never develop rickets. The women work with the men in the fields and leave their infants along-side in the sunlight. The rickets pre- ventive action of sunlight is here particularly clear cut, since in both instances the dietaries of the mothers almost entirely lack the food-stuffs which contain the substance D.” (16). In Karimpur, custom co-operates with nature, in supplying sunshine to all castes. The women live and work in roofless courts. A few of these courtyard admit little sun in the winter months, but even they are better than stuffy rooms. Babies of all castes creep around naked during nine months of the year, and during the remaining three months they wear abbreviated cotton skirts. The children work in the fields or play in the lanes in scanty clothes which grow scantier with increasing heat. No one wears stockings, and few wear shoes. Craftsmen, instead of working in shops or work-rooms, carry on their crafts in village lanes or in open courtyards. The washerman, assisted by his whole family works beside a pond near our grove, not in a dark basement. Every one lives in the sunlight. And the sunshine blesses them. We have no cases of severe rickets in the village. (16) See References. 19551 WISER— FOODS OF А NORTH INDIA VILLAGE 361 Some of the attitudes of our people toward food are expressed graphically in sayings and proverbs. These are an expression of their experience with food. Some of them аге revealing enough to be included in this study of their practices. The rhythm of the lines is lost in translation. “Juar (great millet) is my mother and makes my cheeks swell like raised sweet-cakes. Bajra (spiked millet) is my brother and restores my wasted form." “Whoever eats ‘mung’ pulse daily, becomes flatulent and drowsy.” "Don't call me ‘arhar’ (pigeon pea). My name is dear maiden. When all other grains are gone you come fumbling after me. Don’t call me 'arhar'! My name is dear maiden. Two cakes of me are as filling as sixteen of other grains." “Arhar with dried mango and an ounce of ‘ghi’ makes a dish for the king." "Rice is good, but split peas are my life." "Everything is false in the world, save pulse and bread." “Eat bitter melons in September, radishes in August, and raw sugar in April— this is the way to spend your money and buy illness.” “Two things agree with a man—his own wife, and plain toasted cakes and pulse.” “Two things bring pain— puris' (cakes fried in deep fat) and strange women." [Vor. 42 362 ANNALS OF THE MISSOURI BOTANICAL GARDEN V—VALUE OF FOODS NOW IN USE The scientific study of food in India seems to have originated in the planning of diets for prisoners. In 1880, Surgeon-Major T. R. Lewis in his annual report points out a phase of the vegetarian diet of our people, which more recent students of nutrition have emphasized: “That ipe Mags: however оо can only afford such information as will enable a proximate estim о be formed of the nutritive value of any food seeing that it is not only what nutriment a De cular Ag tuff contains that is of moment, but also what portion of it can med be dig ini and зені by the body. Іп а diet composed entirely of vegetable subst the quality of the cooking is of much more importance than it is in animal duin a don that a large proportion of осе guis in cereals and pulses is enclosed in extremely utri indigestible ср з ich if n Soaps они disposed of by proper cooking defeat all attempts on the part of the dei organs to profit by the food ...... In some cases, however, the e excess of nit rogenous elements is given Ба 5%. рани ої рагс AS or Және зе ope tly cooked grain, so that it is probable that a large proportion > the d. ned nu t will not be Mise ин On several grounds, therefore the addition of Prud proportion a са ка especially ill-cooked pulses—is a doubtful advantage, and may be even injurious." (17). The most outstanding contributions in the study of jail dietaries were those made by Major D. McCay, of the Medical College, Calcutta. His first investiga- tions were in Bengal, but it is his investigations into the jail dietaries of the United Provinces in which we are particularly interested. His statement of the jail dietaries in use in 1910 is useful in a consideration of diets of North India. “The diet scales z yey native ges on hard labour consist of 1 lb. 12 oz. of a cereal pulse combination, 2 oz. of ‘dal’, 6 of vegetables with salt, a little oil м chilli. The cereal pulse сона 15 ern = up according to the busta able Cub Principal Adjuvant nation З diet Gin дивну Саб € Total A 5 сас à 115. 6 oz. . . |Barley 6 oz 28 oz. B · + | Spiked millet 1 Ib. 8 oz. Pulse 4 oz 28 oz. C 5 aize ; 1 lb. 8 oz. . |Pulse 2 4 от 28 oz. р ..| Millet 2 115. 7 oz. . . |Pulse і 5 oz 28 oz. Е eat FAB. 7:02: Gra 5 5 oz 28 oz. Е . . | Great millet А 1 Ib. 6 oz. . . |Pulse š 6 oz 28 oz. G Barley 115. 6 oz. se eee ш 28 oz. H . | Rice à (HT Aw. . Wia ..] 77 o | зок "Any two of the above may be combined sd taking half of each 5 and half of es adjuvant; thus "57 will give spiked millet 12 oz. + maize 12 oz. + pulse 4 oz. = 28, Are will give wheat 11 oz. + barley 14 oz. = pulse 3 oz. — 28. These are са official diet scales as —— in the Jail Code, but in practice they are атон. Па ‚ diet comm of the above table; in fact giv Ron RE vince. During the c weather months great millet, spiked millet, and barley are sometimes made use of, but hardly 5 down ifferent superintendents give their ow inations dam stuffs with wheat and pulse, which are usually very different from E pou officially (17) See References. 1955] WISER— FOODS ОЕ А NORTH INDIA VILLAGE 363 "Rice has been very rarely used in the United Provinces. As it is pue grown to a limited extent, Ж ries | is prohibitive. It was, however some years ago used in two st pil close o the borders = Bengal, but an со of Вегі-Вегі in ex of these caused i e dis- continued and its use has never been rev * *Marua' (miller) is s never given at “и in n fact, we failed to obtain samples even for analysis from € T vae pini 222. were c carried o “Ма seldom, if ever used: why this is so we do dil edicti as it is a very nutritious Го material E соу superior to som yr ose in use “The бог of in the United Provinces are e principally * arhar dal’ (pigeon рег) and ‘ur oda vt: appear from the above table gram 'dal' is not classed as a pulse, but кезі is оба as dissi *dal' belongs to the pulses "For at least nine months in the year Diet E, i.e.— pias 2244 = S. VY пуття р Ста si Vd. Ке T. Th. dal (pigeon реа) d e И D MA Vegetable i с 3 о. | Salt s 150 grains. Oil n 9.08 grams. Condiments OZ. is given to the exclusion of all E ^na The diet now in use in our Mainpuri District Jail is practically what McCay recommended as a result of his investigatons. In jail, the prisoners have a cereal, a pulse, and a vegetable daily. Our village people are glad if they can have a pulse or a vegetable with their cereal bread. They rarely have both on the same day. And sometimes they have neither, and eat salt with their bread. But the villager spends his time in the fields where he may snip off the green tips of plants or roast a handful of grain. The prisoner's fare is exactly that which is recorded. The farmer's is very likely not exactly what we have recorded. McCay has some interesting comments on the contents of jail diets as compared with diets of the province outside the jail. “We ma ^ Басы eue ae —€— г ied pulses of these diet scales, by suns that they esent in quan uch f what is required, P3 ar grea iced than e people of da лайка сап а а ed. or wish to have. Two | or five ounces is the maximum amount of pulse that е enter into any "diet, and, if gram is to cooked, not made of as parched gram, and it should never be the "pir pulse in the piece of the diet. "The superiority of 'arhar dal' (pigeon te form f the Unit that it is the favourite Ғ of pulse with natives of t ed Provinces should ensure m this pulse will be вий in all diets to as great xte ible. ‘Ч d dal’ being more suitable than other forms in preparation o of ‘bajra’ өр millet) and 'juar (great millet) for consumption, this ‘dal’ а ould be used in diets into t position of whic those cereals enter. One chittack (2 ounces) of ‘urd dal’ will be found ‘amply sufficient in the cooking of the quantities of ‘juar’ Се Se recommended in new dietaries.” (19). “Diet C in which the principal is ‘makka’ or maize, we have not investigated. It is never used, so far as we could gather, in the dietaries of the prisoners. Maize we found in Bengal was a very good food-material although not very palatable as prepared in Bengal Jails. It is not suitable alone for baking into bread or ‘chapatti,’ but is very good when mixed with wheat or prepared as porridge. We do not think it suitable as a principal, as in Diet C, but see no reason why maize should (18), (19) See References. 18 Т 28 1 23 6 18 8 20 8 т euoN | ‘20 | ў '20 g || 20 05 | °° 8ашподнј-пом esr "3 1 "3 1 28 6 48 8 *zo 8 "40 91 "zo v чор | ў '20 $ |,70% | '' ** Заштодет бод | рез чим | ЧИМ о р ww | и MET pecie ратовао Smd зда) teq | рд — 364 8661 чо s4nduvo ит oy) ил poen oq 1955] WISER—FOODS ОЕ А NORTH INDIA VILLAGE 365 пос be given as a substitute for part of the wheat in dietaries. Why it should never be given in the jails of the United Provinces is difficult to understand, as, next to wheat, maize is the best cereal on the list. It is used during its season fairly largely by the inhabitants of the province, and the average crop being up to one million tons, there cannot be any great scarcity of this food to enhance its price. We are strongly of the opinion that maize has been neglected unjustly, and that, if given as we suggest, mixed with wheat, it would be found quite satis- factory." (20) "Judging from inquiries we have made, wheat is the food of the higher and middle classes and a luxury to the poorer. In times of scarcity or famine, when wheat, gram and 'juar' are practically the same price, wheat is eaten largely by the people, but in ordinary times ‘bajra’ and ‘juar’ are the principal food-materials of the poorer classes and small cultivator. “МаККа or maize is also largely used as a food-material, barley and ‘marua’ not to so great an extent as the other cereals. ‘Arhar dal’ is the favourite form of pulse, and a small amount is partaken of daily. “It is, therefore, only right and fair to the prisoners that these secondary food- materials, ‘bajra,’ ‘juar, ‘makka,’ barley and ‘marua,’ should not be denied them. While they are not suitable to be used alone, when given along with wheat, in the manner we have indicated in the diets suggested, they are exceedingly useful in varying the monotony of the wheat diet, and in providing food-materials to which the great mass of the prisoners are accustomed. (21). *... . from a physiological point of view it must be admitted that the dietaries at present in force, or any framed in future to maintain the same level are dis- tinctly superior to those available for the great mass of the population, and that this will tend to place a premium on crime in periods of scarcity. This, how- ever, is a problem with which we have no concern; we simply point out its bearing in passing and leave the present policy of the maintenance of the superiority of jail dietaries, over those possible to the same classes outside the walls of the jails, to be determined by the government of the country." (2 ) We hope that our people will find an adequate diet without being obliged to go to jail to find it. McCay's investigations led him to the conclusion that the protein factor determines the place of Indian races in the scale of physical efficiency. Не traced all differences in physique to the different levels of nitrogenous interchange possible in the diet of wheat-eaters and the diet of rice-eaters. If he had made his investi- gations a few years later he would have modified his conclusions to include minerals and vitamins. Since the studies of McCay there has been an increasing interest in the study of foods in India as elsewhere. But very little scientific work has been done. Colonel R. McCarrison has contributed more than anyone else through his experi- ments and through his efforts to use the results of his experiments in educating Indians in the possibilities of indigenous foods. He was a member of the Indian Medical Service until 1919 when he became Director of the Deficiency Diseases Inquiry, in the Pasteur Institute, Coonoor. (20), (21), (22) See References. [Vor. 42 366 ANNALS OF THE MISSOURI BOTANICAL GARDEN One experiment which he carried on with two colonies of rats (20 in each colony) contrasted what he calls a “good diet" with a "bad diet." “Опе colony received a ‘good diet,’ designed to resemble that eaten by the Sikhs. It consisted of ‘chapatties’ made of whole wheat flour; uncooked, green vegetables, fresh fruits (tomatoes, etc.) ; sprouted gram (legumes); butter; fresh whole milk; and fresh meat occasionally.” (23) Liberal quantities of each article were supplied so that the animals could select for themselves the amounts of each they cared to eat. is diet is of interest to us, because all of the foods in it are available in our part of the country. Our village folk grow wheat, but they sell much of it. They do not grow as many green vegetables as they might, but at the same time they make use of every edible leaf within reach. They have plenty of gram, but they are accustomed to parching or cooking it. It would be easy for them to sprout it. They could grow tomatoes and oranges and lemons. Prejudice keeps them from growing tomatoes, and they have not tried growing citrus fruits, other than a few lime trees. Oranges are now grown in our area of the Province, and can be grown in Karimpur. They would be in season during the months when there are no mangoes. The village folk might have butter, but they clarify it all to make "ghi. e same is true of milk. There is milk in almost every home and some families drink it. But in most households the milk is heated and churned to be made into "ghi." As for meat, a few eat it now. The meat at present sold is very inferior, coming from old, decrepit or diseased animals. It is a poor recommenda- tion for meat-eating. Colonel McCarrison’s "poor diet” consisted of foods eaten by any Western people of the poorer classes—white bread, cooked vegetables, a cocoanut oil mar- garine, canned meat and jam, with preservatives, and tea, with milk and sugar. The experiment lasted six months, during which time all of the rats were given equally good care. They all had a two-hour sun bath daily. During the six months, three rats out of the 20 on the “good diet” died, one of them from injury. Nine of the 20 rats on the “poor diet” died. Three of these were killed and eaten by their fellows. After this, the rats of the ill-fed colony were separated every night and a small quantity of fresh vegetables was given three times a week. The other six died from broncho-pneumonia. There were 20 litters of rats born in the well-fed colony, and all of the 134 born were reared to maturity. In the ill-fed colony, two litters were born, one of seven and one of four rats. Two of these eleven died and the others were eaten before they could be removed. The aggregate body weight in each colony was 2,540 grams at the start. At the end of the experiment, the aggregate weight of the “well-fed” colony (17 rats) was 3,170 grams, whereas that of the eleven survivors of the ill-fed colony was 1,300 grams. Colonel McCarrison’s conclusions are: “Not only does the former (good diet) promote physical efficiency and health but the latter (poor diet) gives rise to stunting of growth, to physical inefficiency, and often-times to di v Saes of ға” bad diet is so apt to lay the foundation аге lung-diseases and gastro- (23), (24) See References. 1955] WISER— FOODS OF A NORTH INDIA VILLAGE 367 This last statement is based on his post mortem examinations. Our people have the symptoms of the bad diet, while the good diet is within their reach. Experience has taught them much. But Science can add what ex- perience has evidently missed. Few of them can read and those few see no news- papers or journals. The contributions of Science have no way of reaching them as yet. With its help they will be enabled to build up stronger bodies, without the need of importing new foods. With the raw materials already available, they have a better selection of foods than people in the West who are not as poor as they. An article by Colonel McCarrison which appeared in "The Practitioner" on “The Relation of Diet to the Physical Efficiency of Indian Races” (25) is of value in our study. It is worth quoting at length. who have travelled far in India can have failed to notice the remarkable differences i in phy sus efficiency of different Indian races. So great is the contrast between ce төмен and resolute people of the North and certain кош developed and toneless inhabitants of the [s tts East, tat the qu geben arises: Why should there be these great differences between one rac ӨРМЕ ? Іп attempting to answer it, I shall put aside such factors as climate, де. e al о the actinic rays of the sun, peculiar religious customs and endemic diseases. For while it d be doubted that these play t their r part in determining the position of different Indian races in the scale of physical efficiency diet is the most important of the factors concerned. .... “The food materials available to the people of India . . . are mostly derived fom the vegetable kingdom. They include rice, wheat, barley, 2 millet, legumes, vegetables a nd fruit. For which locality, p age on рео] пед, 25 certain races n Ra pianis the P ak "fontem India. Su ipea taple of the masses thus c of dee ger m рази eun ue the cae of t tissues’ (Mc allen. ~ viz., en 5 and are d in at least three dietary factors. All contai дар d ty; al contain xe md f certain mineral elements, — calcium, sodium, and chlorine and all are deficient in ов cn vitamin A. It is iet i i etabl its, b i h 'd plexes necessary for the construction of living јин ктк т). а lar і io way in кесе the defects of what d D 2 1 Бг фер ы of other foods c table protein, more of А essential mineral elemen ts, and more vitamins, dec peri S. dba and шен. ої the different races of India depend. all the diets in common use in India is that consisting of rice, ‘dal’ кл. ва and condiments. This diet is used by millions of people in Bengal, еее and elsewhere; and it is these who exhibit the пази grade of physical efficiency and health We need not follow in detail, as it does not apply to our section of the country. "If now it be found that the diet of other races is such as corrects some or all of the five f egumes, then, pace! things being equal, the better diet ought to be associated with a higher grade of physica 1 efficiency in the races using it; this is i we ters of Lower Bengal to the wh t anl eaters of the United Provinces—cli and other factors being. practically ities—we see that the latter are енің вири у“ physically, more capable of jr aa irda and more alive th е average resident of Bengal of the same ( у). easons for are that wheat pro a зё Бемег quality than rice; tion diet a ts of the absorption of a larger amount of of - ment in short of a higher level of nitr interc ed И there is not the same (25) See References. [Vor. 42 368 ANNALS ОЕ THE MISSOURI BOTANICAL GARDEN of mineral elements i is not so great, and the diet is, in — — in vitamin ever, faulty in regard to the ау of its protein and to its content of m vitamin À; and since its users do not, as a general зад асыла tit with a sufficiency of other foods, rich in those elements and complexes in | whi ch it is poor, they lag behind, in physical кушык and well-being those races өлме о. This describes the people of Karimpur. They are better developed than those of the south, but they have not reached the degree of physical efficiency attained by the races further north, whom McCarrison proceeds to describe. “The high cns of physical efficiency of the Sikhs is attained on a diet of wholemeal bread, legume seeds, tubers and roots, vegetables and clarified butter, supplemented = libe на amounts of cow's, а ог pais milk (whole milk, buttermilk and curds). Beef is never them, but goat's flesh, n and game are eaten occasionally, although not in large quantities. In Sikh regiments, for instance, a pound of meat is eaten by each man two or thr month, while 16 o s of milk are taken daily. In their own "уље m = ее; milk are used ‘har ia is sepes in duin siii ife. It is to be noted that milk and milk products are constant esee of r food, other animal foods being regarded more as a luxury than as a necessity. Their c Seis en are breast- fed for two and a half years, the mother’s mil supplemented with diluted cow’s or buffalo’ ч ог oe s milk during the later ter they are i amounts and are gradually Креја on to the iil in use by their elders. The Sikhs use veia: E but they do t fruit to the extent as certain hill people whose climate is more suited to its ней e f of the Pathans is much the same as die Sikhs, but they © eat more meat end less gum They € 2 advantage of a better climate and more fruit and nuts are available for db use—mul walnuts, etc. “The people of Honi mes dn use of meat than either the Sikhs or the Pathan Their Bed is much си ѕате аѕ 2s that of the Sikhs i iall but le: great а especially of apricots and mulberries. They use ry t heir r t or use «Қаны che citas pes Winter months. Dried mulberries i е а ся аге тіх wholemeal flour and ече into cakes which f their stanle FON of diet. Mea ux used only on special occasions. ese реорје are unsurpassed bv Indian race in lends. of physique: they are long lived, „Vigorous in youth and age, ане a great endurance and enjoy a remarkable freedom from disease in general. To see a man of this race throw off his s garments and plunge into a T aedes river in the middle of winter is to realize that enforced restriction to the чае е дарр 004$ of nature—provided these be of the right kind—is no bar to the ot eee perfecion of physique. The of these е races—t as Sikhs, ~ Pathans and the people of Hunza—provides = interesting сыйы: in m the accuracy of the conclusions drawn by McCollum asa result o is extensive inquiry ith de kinds of diet: tag succeed best in nutrition of a decies orous type of di which satisfactory nutrition can be mai jateined “ther are two o ch he has found to be successful; ы 1) the seed, rab r, root and muscle type i i ent ith liberal amounts не oe and or dra d (2) the type of diet derived from cereals, legume seeds, cer , fles € wit сады without meat, i ilk— eing a con good all the The diet of the ы 15 речей ehe with the la type; that of the Pathans with the former; while th the people of Hunza is a happy combination of 'the storage parts of plants, leafy exactor s fruit and milk, which gives, per- haps, the = result of all. "So m then, for the RE of diet to the physical To of the Indian races. There remains now to co: r, very briefly, the extent to which their diets predispose them to di eir freed m i have seen that the eos. mass of the people of Bengal od i on a rice-and-legume diet having at least five faults. Associated with these dise entry -beri, malnutritional oedemas, epidemic dropsy, and xerophalmia, which diet о! f such о epidemiological inquiries have ae kee pace with those in the laboratory. But in this connexion it may be remarked that к pian higher анна of leprosy in the South West, and 1955] WISER— FOODS ОЕ A NORTH INDIA VILLAGE 369 East of India than in the North has reps a nutritional basis. The high incidence of dysentery in the jails of Bengal as compared with its low incidence in those of the Punjab is border exam nple of t e sam fing ee ontrast now with these morbid effects of food having the faults € —À to—faults present in some degree in many dietaries at Home as well as in India—the state of health of the people of Munza who combine their foodstuffs so happily. With them пена to infection is remarkable: anthrax, for example, does no more than produce a Duet pustule, while epticaemia is its common consequences in those less resistant to й tion. Gastro-intestinal neve мбар i ts the rest F Asia has kept them free of many diseases carried by infected , but there can be no doubt that iren freedom from disease is largely due to the food ые eat and js health-giving life they lead in a bracing climat After this clear picture of MacCarrison's of the possible range of diets from foods grown in India, it will be of interest to observe the foods available in our particular area. The following table is a summary of the Food Sources discussed in Chapter I References used in identifying the products were: (а) “The Commercial Products of India” by Sir George Watt (Published under the authority of His Majesty's Secretary of State for India in Council). John алар, Гопдоп. (b) "Food Grains of India", by A. H. Church. Chapman and Hall, London. | (c) "Flora of the Upper Gangetic Plain", by J. Е. Duthie. When names were too local to be found in any of these, specimens were sent to Winfield Dudgeon, рн.р., Botany, of Ewing Christian College, Allahabad, who identified them for us. euno оца ur se e par рээпроз4эл st Зи 24$ әз zoded oui P 3591 942 ut sy ота Зав! уво вот. звису vod ^o) onu seprorpe1osq sedouso Tirereq свод 194800 өЗетріл JO sprer.r umure 1901) зиизчо | ** (0913) ved yoryy, оде твои 831484 JO веря | wz£xo jo fend on Кеј 19893 pu РИМ хр Їх x ота ere3[nA отерлон овр Корза x x ота orema umuZiog теп goru 78919 х onu urmnuo[oo urnorueq 30189 өш в,ивш 1004 x ота шторгоца А umjosruuoq ve1fegr gomu poxrdg х|х J* Due '* обета Jo SPINT өтөш umorueq тязом зет err зер Ни КЕТЕ ТІГЕ d prm вло) eneu ogly |7: екипа зепти РИМ х|х|х|х +" 4” ота | в01[831 9119305 чое чет uL БЕІН | ота pi вововіоо eursne[5r MIL ** gem x|xixixixixixIxIxixixix] > r. owa ; 841488 82410) чеца | "С ** сом х|х|х|х|х|х|х eSvipA јо SPIONI вАзш 897 "pre t" ezrey х1 xixix[x|x|X|x|X|x|X| зо ерер 29430 20 spog umo өлеЗүпл umor, unqop yoM ЧЕНИНИН lege “> отавівлв UST AA 89917108 В T T вошем 370 syjuou Ќа апйшыз у wu 21401 wan ероо 371 "$junoure 19859] ut e[q9|reA y х *8junoure ©8198] ut ejqepreAy x *yZingxoy Aq pesueAez ол? өѕәцт x| He ees чапдшвуї VET by |х|х|х|х|х| зо злемола ојфедодел peog | ^: size8ma еџеџодет | > ameg j e з ** рор X|x|x СЕСЕ undureyp ло врје umo | "' wmionbyue 915300100 | "" ето | ** umm jo водо] TT EXT XT XP Xx] о Hmdurey 10 spjoy ал | '* umionbyue вієвооо) | ** uemyp | пооцевр ume 14484 ы Шы ке шыл за rset сақа ** .. воде та 19430 .. "doo umy .. те .. **8uomo ‘soea Aq reou рех ] Х |] pue вломола epqu393eA [воот | ** | зпапеєціда stumong | ** Hexe | 1equmono әл чоөрү "АРАҒА WE АРН >> ' sogea Aq зәм | ^^ зе пл пп | 7" *puijey| uojeurrojua JO uro; ‘tandureyy вода пл Aq твоп Ср рер рр рх somo ` oeoa воот |: oded вудтопд | · змов": | urjdumd jo Комод "18284 LEE a —| чада) soea Aq uveou хх ххх | px [xXx [х | “воло 9148398а^ воот | '' wueZuojpourumuvwpog | '" ue3usg | -- qued 884 #ә1@01әбә д us "ho өе. m 19440 JO овоца х X|X|X|* X| рив влом ојавјодел jevo | ** stuiopsue вувлвивд | °° weg | ** uveq 301238 *1833'€ur хр” о ОЗ ОЗЕ | Ж 4 4% ота .. enyedry, ee vod jo AJB хіх |х | х ЇХ | х | х | х х | х | х х | з» um отша vh snorput snuefe) | +: reury |, vod џоөдід xL х|х|х|х| °° а ота s esueATS umsgtq |ң: WHEN | ** ved prer,r ххх ххх хх " ота вапоучтиоов епјооввца | °° чом |** вод еџчооу ЕЗ х |х|: х|х|х|х|х| = "я ота ** авозогрва snjoossyg | ** Suny xl x [XP bU ix]x)xIxixi б e оша "+ „овапш впјоовеца і *° рід 372 '(1equieoeqq 19318 pepyord) х|х| х|хро16БОГОГОГОГ ЇЇ ваемола ојачдодол peog 3041433 snuvydey "uw нәцетрен "водила | "вило хіхіхіх "ОЇ 19430 jo 81040148 ојавјодел | ‘лвл зооваою зотзвзая тдодача | ^ ломоцтпво — ""мтпацву "впечогаї х|х|: шолу eH] ‘Bera jo sprong | «rea ‘elajiq(nq BALOO nyeyey | ·· ured jo A30119A *19331q 30и XIx[' ота треоб "sHeqoes[p exrT | змецовур тату Aequinono elI хіх вломола өјфе}әдәл [8007 (4)=10108123 stumong вцроәѕ ләдшпәпә 0} резче "вограо х|хі|х врјеу ezreur 20 903409 UT | -ureur "ява opu вгшпопо вілоцову doquinono өңү uo[o]y | xl врә uj ({)в1л}вәйитвә snorredy тед узе вшоолцепИј *SI9MOIZ ојаједо дол хіхіхіх | eoo 'ezreur Чу рец имо вјрадивјпов вда" вва, panog jo ршу хіхіх врјец UMC Snares SIUININD "Inpy Joquinong "ооо ие чим имою) 'врјец чмо | · snjue[noso вповтатн tpurqg | * ; захо “imd . | xixi: «пер шолу орум| 'sp[eg 189071 «тдивлецо вограошоц| Boley | 7 uou 194918 - о a| >| ен! |р : 3BBEEEMHPHBE ("ppouoo)— X62; 07250 4 189118108 18207 ојавролв UAM soo1nog usriZuzr SOUISN (penurjuoo)—squow. fiq ndun y ил 290009 ғро0 т 373 —|—\— Ар 13334 хх хх X|X|X| иду рэма) ‘spy твдот (ure13) (од -могцо vag) | ің вчивҷо sdo} ump _ —|—|— "(Am — аду х|х|х|х хрхух| рома) ʻea јо ‘sprorg (seed 225') |гуув4 vx тезе souta vod јо sdoy, хх '* вріәп Аэыза рив увоцам пт шпатв шотродоџәҷо njeg 4003-05008 OYM 8.04012 o[qe30804A [ROOT (ystper 225') | түуей ің img do} usrpesr SpI9g Ut РПА SIPHIA воцзџвівшу | “е тегипеҷо "7 920904 Seare pojeAnjoun ut pit A (;)esord. вуоритицу, wnssv[ | 7? 9194104 рим y "рипола uo рвәлі soAwo[ PEM (оувоог оў e[quuQ) | ** зАцата оде твои spuod ut рим зотувпђе gaourodq мем 9194304 рим ғәдојәВәа /о27 xl Апвоогт €jeSorreA этазчизя | *° авицову | 7" Ieuyoey jo spng A "209 "eeuqvr ХІ” күк задвил 19440 рие Аүвәот `* вхќш зграю | ex —€eioquqwT Bye jo spng a "вуву во Адәт доб х "Їх мошеу | -19A *eooezepo — orsswaig | pueg зозазю | "" "" одвадео рез х|х|х|х утро] амо pue вломола ©9498, 5878384 воошодј | puebrexpreug | + 89048404 Joong "Хх! вломол8 o[qujo3oA [BOOT "7 (чырзләә$) | печатуиаи | 7 рой poes узгред L4 "(rid y-yoreyy popyord) хх ххх] вломола әраелодал | | BOOT 30180 snoneq | ** iefep | "" ** 8402280 xixixixi xix (рертота) врјоц umo | ** wurredsoSA1ojd вдипоуј | · • vufueg | >” цетрел овлон о 189 re "вломолха хх хх! `|х|х| ореол pue зреу имо umsoleqn, шпивіоя | ·· шу |: 590389404 374 ("uopoux вш +. х|х|х|х| ** seSeptA 19430 раз Апвоо | °° opu srunon) | ** znqueyy | 9exr]) ‘поем yoomg е 56 |x| x] х | х ота Ys звота тел впо | ** | пеба |7: | Tedeed jo soraogr х1" "|. X|x|x|x|x ома ds sopounreür o|Boy | ** Pa "тоҷӣ 2 {Хх Җи ^ ота B. 91103131 818688 ex впчерү | 7” 801088019 пуну ° sel пу м хіх ** едвјил JO вололку вувлошоја snot remp | °° 829 PIM ^ (‘Ame 10478 жшк ИЕ ы ae pornp ло реа) ‘som ххх ххх | хіх х | х| х| punore *eprspeorx Зпојв ‘вело зогрит влеуідиву; шу | '' веодивих подію иы ххх: . yerd чим имола АНепз | ** suopoeA€13 umuvpooneg | 7” зоб «а ша my Ge) fee ЗВО (Ame "штоовад х|х|х|х |х|х|х|х| —mdy репа) ‘врюу 18007 | -umuooj зпечо8 гај, год моолаппод ыыы (Атар ER xIxixix . хіх х| —-Чолерү рема) 'врјец {воот 5119 3эЧитво vorssuagr тлелвуј sdo} рәлеҙвт ‘randureyg јо 95043 хрр] pue вломола epqej93eA [0007 втїв8үпл ея хера | "еј sell тај oz ol > У SH] Н —војфеодол Aye ЗЕБНННЕБНЗЕНЫЫ a | 190108309 T990'T eurer (penuruoo)—8s6142u0u. fig “тамо y ut Nman $poo,T 375 x| Ж Ж ~ X X "вода А 19440 jo 9804$ рив вломола o[qejoZoA 18007 "umd “| -ueg pue одаша jo вололе) зе ел 19430 шогу Apoo "вәЗецтл 10430 јо 095043 рив 8олол8 [воот '* врџәу рив волого ** оебзціл jo волого әӘөрпл рипогу "" өЗеүл Jo Блог; ^. Cpudy—dqureoeq рої * * 88: твои 89011, e he La кебе mon ‘отл JO волого ота ота одаја рипогв soar, ота ол BON soera 19430 ‘OZE ш мод әӘе А OY} Jo волого suena вороною umquerdes ssn увод BIPION `+ вх ці BPI запа! snudAziZ вотрш впршавшивј, зпоррипјо snudÁzrz штутецаоје өтаоләд '' ртов 811310 v: terdos втлед зо) шпувивіЗ vorunq зивіодціві eruoSnyr sreuepe2uoq snot, ззцовпрегв erp завуцідовр xrueouq 31198918 хтаооца тлвХп2 umpisa "7 зпорабодит snd.isoog1y запали], чох BIO GV] тең *e1euqe[ eyeu пою 1939 M, ** тери вши 4 әңгі sumg]d от eqnfnf изгра] spod рицтошві, шпа РИД ә(44в poom өшті виәәц jo sorog әуепелдәшод шшщ е от -- изказая soroq w80219]y е С ова ** gamp зиму soe 376 х|х|х|х|х|х|х|х|х|х|х]х| С 2 SpreÁ ynog | ** SrputA виупец | ^* вшрод | ** "o шу ‘umusog хіхіхіхіх . * *u03300 10 39 Пим чим рец up um три 1918599} шы, рив тт, | eua рив хоз "Sprog х|х|х|х|х|х|х|х|х|х|х[х| &opreq. pue увоцм jo лорлод soounl eorssedgp | ** Tey 20 вт ** presny "Хојтва ххх ххх хх | Х | Х | Х | Х | чим poxu 10 *19p10q єв 59Л019 ** ВАТ вәт иеп "б рађепу "атподоцогр хіхіхіхіхіхіхіхіхіхіхіх euo[e Ajrensn ‘spjoy up | "вд 'sndeu — worssedgr en j: '** ейн "01918 10 “Хојлва х|х|х|х|х|х|х|х|х|х|х|х| чвеца чим sppy ш змо stiqsedureo вотвєвіЯ uosiBy ** presny х| х| ххх | өЗвүтА. Jo врея urmnreuroggo 02104840086 JI опволедп ie xlix обвал твои вршод esourdsrq ede], 31948018 jnujseqo 19e AA xixiXxIXxIXIXIXI ТЕЛЕ” x одел твои Spuod ut вмолк) 510401 өзоціш кг *"[odeqq ка| | Sal Б ЖО |> Sa] = > ЗЕВНИНЕНВЕНЕ spoof soio Teoruejoq. 19991 9118188 TOT AA воолпоб Чец8оя eureN (рорпјропоо )—еујиош fiq anda у ил әдат spoog 19551 WISER— FOODS ОЕ A NORTH INDIA VILLAGE 377 The preceding record tells us what foods our village people eat. And the fol- lowing diet records give us an idea of how much they eat. We cannot hope to arrive at accurate results, with the equipment of village homes and the preparation of illiterate village housewives. These two drawbacks we could have met. But custom rose up to complicate each step. To get a representative picture of village food consumption, I thought that we should have records from at least twelve homes each representing a different standard. I could not rely upon the women of these homes to do the recording, as they were not accustomed to measuring their food, and they could not have kept any records of the measurements if they had made them. Аз was mentioned in connexion with Food Sources, measuring is a casual matter, and is usually com- parative—one food balancing another. 'The two women who could write and whom I could have taught to do the measuring, were kept in strict seclusion within their own courtyards. I had lived in the village long enough to know that my interference in anything related to food would be most unwelcome, even among my best friends. I finally chose three Brahman boys to help in the project. They could write. They were still young enough to be acceptable in any of the homes, and their caste made them doubly welcome. They went into the project very seriously and importantly. One of the many surprises I had in connexion with it, was the willingness of the women to let the boys write down figures which they themselves could not read. Village men, and still more, village women, have a great fear of written records. In their experience the only exact records have been those kept by Government clerks, rent collectors and grain and money-lenders. Аза result they associate recorded figures with oppression. They become restrained at the sight of a note book and pencil. So I learned to make mental notes, to be written down after getting back to our tent. But with the boys, recording took on a different aspect. They flourished their pencils and paper—both rare in the village—and made a great fuss over their measurements and figures. And the women laughed and were willing to humour them. But before the second month was up, their old suspicions were aroused. Some one brought up the question which blocked us repeatedly. Why did we want such figures? They began to imagine that the quantities of food they ate might reveal their status in some way. And this dis- covery might in turn lead to an increase in taxation. To this road, all paths seemed to lead—fear of increased interest or taxation. Also, I realized that within certain seasons, the diet is monotonous enough to make records for short periods satisfactory. We waited for a new set of field crops. But in waiting, we missed the wild herbs, and the rainy season vegetables. These may be found in the partial food records for August and September, ap- pended to this chapter. We renewed our weighing and recording in November. One week was found enough to cover that season. There were still other complications which added to the interest of record- keeping. In our village we have three distinct systems of weights, for those who [Vor. 42 378 ANNALS OF THE MISSOURI BOTANICAL GARDEN aspire to weighing. One is called *kachcha." According to it, one "ser" is 1.4 avoirdupois pounds. The second is called “pakka.” According to it, one "ser" is 2.8 avoirdupois pounds. And a third is "sarcari" or "official" amounting to 2.12 pounds. We tried to keep each of our helpers to the system to which he was most accustomed. But they were apt to slip from one to another and back again, with- out noting the change. This required additional checking. In the light of these possibilities of error, I make no claims for the dietaries beyond the fact that they are a seriously made estimate of foods eaten in village omes. The next question we ask is—Do the amounts of these foods eaten meet the needs of the people who eat them. То estimate this, I figured the total daily food allowance necessary for each family, and compared this with the content of the foods eaten by the same family in one average day. 'The food allowance for each member of a family for a day is calculated according to the suggestions of Edith Hawley. (26). These are then totalled to give the need of the whole family for a day. Hawley does not include iron. So I have based the allowance for adults on Sherman's "Chemistry of Foods," (27) and that for children is calculated from the amount suggested by Rose and others in a special study of iron intake and output of a girl 31 months old, weighing 31 pounds. They recommend that chil- dren two to three years of age receive at least 0.75 milligrams of iron per 100 calories. (28). e needs of the pregnant | women were calculated from figures given by Rose (29). Those of the nursing mother were calculated from suggestions made by Rose in “Feeding the Family. (30).” (26), (27), (28), (29), (30) See References. 379 “ст e3e) vid euo 404 “св ‘Lp 8938) uou eor. (19202 B 0 sdo4 ureur) Күҳәәл) 5 31914 884 E: а т| 215 отт fot 3 uo[ourxsu]y ^ p unmdung № 0 ` чні = 0 15 Хәнең зані 0& веобинуї | Р 6 1equmon) | 8 р | пода | г: | | 0 85 звечм 19101, 8 £ Жорта її 9 звәодиву | 8 0 4equmon;) : : uM те #9 9 0 il І uopuysny | с 5 wydumg ios oe 9 is it P вәойиз |9 Т шашпа № 0 шец) : : : zl 65 "вод іс от ft + uoqpwysn |9 т Чата | 9 т woes : 1 4 Tou 85 їс H 01 зәодиву | 21 0 sdo} ше) | 9 п посла ы : "oou LZ әре от ? 9 £t c seoduew | с с ymd 335 Qe 8 © PAM 95 $3 її. 5 390843 | ZI 5 ходшпоп) | ZI I ЈЕ. 0 1 PAM се zo*qp |70 “4 5з '20 *qI 70741 20 "41 8661 (eeu) Cmm) "ву 7 әәлв pino қ 10430 өәц3 10 9 8 3195 10 10 “ш ama 591483980 A osmd 946 |әоты| З S| p9e1q ш вәлә) 9790 19808 | по 1033 па %% мәз ШИ | +8361 ‘Avy Јо угат ouo бимпр эпйшызуу fo fipwpf июшүрдү v fg uam роо] "(т 988) лофоц ouo 380 "(вуупош QT 088) quujur ouo (ст ‘LT ‘вәЗв) пошом omy, “(азолом лодипод jo 1943044—6 988) Код ouo (05 “95 ‘Lp 8988) uou COIN T,» 210 ‘5903 prejsny мор |9 1 Зи "(19909 9 I “MMe |212 ра SIZ әлеу | IL 15 өлер |) "взубіюм Ардоом) NUT | 19301, чт їм є | St ene рт п ойма uou | 8 $ foro soeg | 6 ££ oym $ -F ino мо prejsnyg ос сме "0 39 FYM eI "(Келе пөлід "ват голу, "1841503 205 sexo [eroedg) 9 1169189 Cc] у ^] "Pe e eee " 7 47 801 PAM п | 9 1 ‘sod 121 вои E i | uoa | с: [ато Aeg | 11 / ozen от (пиш 55 өлер | pony ouo) f]. oiv ро Да DE ни ү 9 8 M 6 9 т ‘sod .. .. .. .. .. .. .. .. чол 8 е А ezrepr g Р авочм 8 "IOMO в JBM il "је O° р зе т т Ж ле 0 |. отв L more ТЕ ptt e p ЕЕ MB Ld 8 0 Херея |9 8 ezrew 5. "20 "41 | '20 | '70'41 "20 'а1 '70 “QT '20 "І “20 "qi "20 'а1 “quoy „9940, 8999115 p. 19910 5 eoy3 = Я Е F à 3198 | зо | | зо | є |$^| & | 5949498ед | esmd 97486 | eog | тволео рецолвд | рвела ut [60102 ова 18808 | 8 | по EE MOY Ф 48661 ‘әдшәао fo yaan ouo Вилатр апашлаоу fo ћрлиој unwynsg v fig uama poog і 381 Че оде) 118 eut) (р 'о8в) Ход euo “(01 938) 1213 ouo “(06 ‘L ‘09 988) пошом соці, "(ро ‘gg 8988) пөш oa T. GI воодивуј "вод eI 6 жа! siqa ; лю Т uopouueyeA | PT L £ чоөйу 211 т LI amg 18304, LT |р ѕәоЗивуу | 05 — preyy | >> | >- (ег чоршавоц | urydumq | Є шв || Хрен | 6 L PNM 211 1 с за $3 8 seoguey | upjdunq d “рт Хере | 1 6 Және L 5 әрең 1 $ по 9% рази | `` | >> | % чојештту | udumga L 5 PUM 0% L с за $ по 9 I равјепуиј | со | ^* |1 uopouuo2eA urydumg L $ omg 6z 1 с sg $ по 91 равуепуј | 7: |: |; вәоЗивуг rydumn gq |8 Аерва | 1 с Дора 85 1 б вод "вод fs [9 воодиву $ uoesig | | 8 Хорея | 2 soeg 15 L 4 .epeg і "по y | разу | © |“: | у чојешхепу L С PUM 95 | L 6 gmg "по о 2 * 01828 9 % PsN | | | с чојошхепу urydumg | Є peuoreq |: L 6 PAM 96 20 чо 70 "zo “4 чо "zo "1 720 'qi 8261 "ву 599945 19q30 gm £| 100100 pueuq ло | вера | 0048 ю по |26 па этая se[qe302eA | өзүп 3105 рецолед | ur peered eyeq 16808 pen "ITE „8261 “Йәй fo рәт ouo бимтр anda у fo брат 42juodamy) v fiq uama роо ahi tenia TUE и сале vhi ы ш. . од ето . [7 ‘(OT 988) 118 euo | "(ро ‘ge зов) пөш oT, 9 төөчү 9 ST ШИ 6 ЖІ по 9 Z puny|z % ved (9 | вом |0 2 Хорея | 0 91 399A | |"'еуцаољ їр т | одизу | руту | '' | є волопі |9 rsousrpesr чоәйтд PONA | ст 0 өгей | 9 cc олву 1910, Oe vam y LIS її өлорү ofuey |9 Т eeup 0 1 AM eI | "вита Т ст по СІ сив | 1941489 Я ст о | одизуј | presny I = | ^ uend£S .. .. .. 6 99M | TT | 9 9 т ved 9 т 9011 дерти: .. 1 ввлепр) .. чоо8т репа 73 TIL 3901) от 9 І шов i " тета «а у 4” 52 0,1 Херея |0 2 Onen 6. 9 Т 840 в. Н по Чим є 0 | одағы | рати | ·· | ·· .. воцятрем ae s от Арея | 6 , өтей 8 9 т vod II 2 зри поет т eI 0 OZW 3891р) ГА £ | 9 т жәй fa 0 о8авр[ "m .. 1 .. ee 100214 ae ee 9 8 2241274 „88 ; *86 *20 *9]| 'zo о *qt "я "20-41 "zo 41 '20' 41 1041 о *qr «ходшолом пш ) өп ГҮҮ 3 29930 ind g ypu [99160 ртола Ч. 5 оде dyd eot aq Мел воррта 10 ПО 8 e acr yng | sepqeje2eA | osma pag ІН poqa Ut 89220 мең — 382 48261 “әдшәйом fo yaam auo бимтр sndwwoy fo yrunf аригаа0о v fiq uama роот 383 '(вчуиош 9 еде) Ход ouo “(9 өЗз) Ход euo "пт 989) 1113 euo *(15 988) чешом ouo "(86 989) usur өп), 9 | wei pn peuoreq отч ші) oom) 81 воодивуј 210 | wed с 2 Хоџа |0 pI Aog | > yom "zog | ZI EISE 9 uo[poudijj9M | 9 т punop uoodiq резон | 0 РІ звочм 19101, 126! шэл) PI Корея қ oon. "o Pme У ит FAM те ** 1% uo[eurioje uooztq з lero 1 M n1 ~ Ут — Аоред 5. - PII WAM 08 ** || чојошлоји 009814 a? т 9 1 шый 8 ! аи. i `+ |Z чојешшоја peuoreq вво - ^ 515 — Аза k "7 181 sSeoJueg | 9 1 pinoy х" Ја 515 увочм et 5 W E = т Aog e vod woes 1 PII увочм 15 Яс — Хорна 80 wed 0981 1 FIZ BM 95 515 (әрең візі M 212 WAM 9c "що 20 (по)! '20 "4 "20 91 "zo *qr "10 *q 10 "41 8201 “бор үзекен А se 80.100 рола мев | : | E] | ош IR] | уља | ваановод | ойр leong ~ чі 9939) оза ледпв | ® |З о | 2999 EH «8061 ‘foyr fo угат ouo бимпр audwuD у fo Кушту sassasd о un fiq uama роо] 384 “(9 038) Ход опо ‘(syguout 9 088) Хоа опо "(2 988) челом опо “(ІІ 939) n3 опо (ве сесі usu 0: (18303 КІ 8 0 PIN ! : : -490A) 8 т weg їс увоцм | е F Зәшш $ |с 1 | є чярву | ёт 9940) 0 сє мтму peuoreq родвво poxidg ву ом iei по 9 I uvudg|g Т тәй #5 шоо |б L PAM 19101, с вөойп®рү| preng c ZI 0зәцетрен цоэ | рецолва |, РІ отты CIS өтөрү з $ чире ү : i 8 0 рад ро ы 0 с PAM eT cI 9949 “ ne 0 & •шпів |510 ure (1819893) 0 Y paejsngy 081315439 peuoreq | 6 с PAM п шоо | $ р өш 2 вговиви I 9 I човшаѕ à fe рочыза.| = pomdg 01 ^ ” и $ Ро өтей 6 er 0 vod c 0 : uoesig || fe звоцм ројввоц |0 с өтер 8 510 тө . мова |" el c әтзей L "ва од i Н 5 210 чим |510 шва i i p1iejsnyy I зецятрву peuoreq : el c Oze 9 *20 хат | 20 - то "т "то “QI *20 "20 *qr . ш 193 8661 -mg) ‘19qUIOAO NT j99AS © ius, i "i 19910 > Е Ф [99109 рвола » 3188 10 вордота | 0948 10 по "294719 e| вәјдезәдәд өзүпа 3198 ш ова 18815 Я AEN peuoreq т [8910 Ма et 28561 офиовом fo yam ouo Bump ташу fo Ёүїшт/ 1258248 по up fiq uara pcos 385 "є еЗе) Ход euo “(004 989) свом euo "81 988) пвшом euo | "сс 988) usw өч) “(19309 edi оп PM | тн» © ZI | 7 f|ocseoSuew | 8 $ шадашта |g ү жәй пообід Р 8 Кәред |) MOL ET. Коня у вәоЗивуү | 9 т undung от 3394 М те эт Керең 9 I чита 9 1 авец M ог 9.1 «өрең 9 вәодперт ZI 0 wed чое 9 I AM 65 9 I Корея ZI 0 ‘sod чоәйт “18 T q99u M, 8c 9 I Корея 21 9 I увөч M 1% эт Корея $ фр seoSuepr Io p eoq M 95 ue "b $ 19 seouey | Zio umdumg 212 увочм 95 "(род В “20 дї "ZO а 20 “aI 20 “20 "20 | -1933ng) 8261 ‘ABT веома] ку а O = 1940] =" В E 5 по 2 88 greg | о | |8 < | чепа зголя 891483930 A esmd ydg o 9| peosq ur гесло) ча reS8ns| З 55 ти =. мен te „98261 ‘Avy fo yaam эио Вилатр snd ыт x Јо Прушој ојуротто un fig uama роот 386 “(е 938) Код euo “(00 988) томом euo “(вт 988) авшом euo | “(02 988) изм euo, 8 9 mu "(19303 98918 4 26 өлер || ‘вом 302A) 9 I sdo} 19301, 8 1 по чим |8 с ved 9 їст |6 soppa pwy | $ воцетре чое fg Корея | Р 9 | звочм 210 по $ $ одиву ргеҙвт | T p 88 WAM aI sdo} 210 по 91 чим (18414903) 8 $ о?пер рупии вәцетрезі я а 212 FBM п 210 ү» і .. .. woes 5 $ өтөрү ot ZIO sod {т of8uveg поет Ее өлер 6 ir ih I 2m 8 £ өлөрү 8 8 0 səd 2 725 mu T ; woes fo «әрең 48910 1 8 0 səd ZIZ mu фр оЗазу 1 000814 98910 9 *3b ; “(этш zo | 70 20 zo *qt +104 20-41 то "І "zo "zo “qT "8261 -mg) *iequioAoN. S 5 иш 3198 1; SOPOT 9948 оро |-1933nq вөдезәЗәдл | esmdqandg feory Bine рвота ur [9190 eyeq di “ПИ м «8061 “4әфшәаоду fo qoom ouo бимтр гтішыру fo ушо/ греюото ию fiq uama роот 1955] WISER— FOODS OF A NORTH INDIA VILLAGE 387 The content of the foods eaten had to be based on analyses of the same or similar foods in the West. I realize that their application to foods grown, stored and prepared in homes of North India, can only be an approximation. The food tables in Rose's Laboratory Manual, third edition, 1929, and Sherman's Chemistry of Foods, third edition, 1930, were my chief references, supplemented by Bulletins of the United States Department of Agriculture and of Hawaii, The calories, carbohydrate, fat, protein, calcium, phorphorus and iron con- tained in the total amount of each separate food eaten by a family during the week were found. These constituents were then totalled (from all foods eaten during the week) and divided by seven, to get the daily average for the family. These tables with their requirements based on the comparatively low weights of our village people, imply that the foods which they are getting are adequate, But as we compare their physique with that of the Sikhs, except in calcium. A admired by McCarrison, we know that they are low in weight and height. review of each food constituent may help explain the lack. Calories—According to the figures estimated, calories are more than sufficient for the oil-presser family, and very nearly adequate for the others, except the carpenters. These calories come chiefly from carbohydrates and protein, whic implies that an unusually large amount of both of these must be consumed. Their chief sources are cereals and pulses. The figures for cereals consumed are based on five-sixths of the quantity recorded as being taken out of the store-rooms by the women. We found in a number of experiments with toasted cakes that from A comparison of nutritive needs witb food consumption of a Brahman family in Karimpur Weights in grams пса Calories Protein Minerals Ca. P. Fe. Total amounts needed daily .. 2» 16,609 422.7 4.600 6.972 .104 Total amounts consumed daily (May, 1928) 13,855 466.8 3.699 | 13.216 144 Total amounts consumed daily (November, 16,865 440.66 | 2.227 | 11.539 .120 1928). [Vor. 42 388 ANNALS OF THE MISSOURI BOTANICAL GARDEN Estimated daily needs of individual members of family. Total need 18 based on these figures Weights in grams ыы Protein Minerals Са Ж Fe One man (age 47 years) .. 110 2,810 49:5 *550 *990 *015 One man (age 25 years) .. 105 2,205 47.3 “525 "945 "015 One man (age 20 years) .. 115 2,415 51:7 "575 | 1:035 *015 One woman (age 17 years) lac- 90 | 2,530 68:0 ` 781 "970 *018 tating. One woman (age 15 years) 85 | 1,985 45:0 “675 800 “018 regnant. One baby girl (age 10 тен 16 i s is о "і One boy (age 9 years) 54 1,944 70:2 “648 . 972 *018 One helper (age 12 years) .. 70 3,220 91:0 *840 | 1:260 *015 Total Ex 4% 16,609 | 422-7 | 4:600 | 6:972 *104 А comparison of nutritive needs with food consumption of carpenter family of Karimpur Weights in grams idi Calories Protein Minerals Ca. E. Fe. Total amounts needed daily сж 13,230] 366-9 | 4-075 6-725 “097 is с ун consumed daily (May, | 7,838 | 297-3 | 1:534 | 7-118 | -0924 Total wee consumed daily (Nov- У ^ i і ember. 1928.) у (Моу 12,878 364°5 | 1'5209 | 7 005 | 0771 1955] WISER— FOODS OF А NORTH INDIA VILLAGE 389 Estimated daily needs nA individual members of family. Total need based on these figures Weights in grams Weight | Calories in lbs. | Ргођет Minerals Ca. P Fe. One man (age 33 years) .. 110| 2,310| 495 550 990 .015 One man (аде 24 years) .. 105 | 2,205 |. 47:3 525 +945 .015 One woman (age 27 years) .. 90| 1,890 | 40-5 450 810 | -015 One woman (age 50 years) .. 85 | 1,785 | 38-3 | +425 760 | -015 One woman (age 20 years) .. 95 | 1,995 | 42:8 476 855 -016 One boy (age 4 years) T: 30 630| 42-0 -540 150 -006 One girl (age 3 years) i. 25 525 | 35:0 -450 625 -006 One girl (age 10 years) і, 55 | 1,890 71:6 -660 990 *010 Total 13,230 | 366-9 | 4-076 | 6-725 -097 A comparison of nutritive need with food consumption of an oil pressure family of Karimpur Weights in grams =— Calories Protein Minerals Ca. Ж Ее. Total amounts needed daily 6,606 | 196:4 2:148 | 3:420 " "Total amounts consumed deily (May, 8,325 | 325°85 993 | 8:399 | "00948 1928). Total amounts consumed daily (Nov- 7,479 | 231:14 | 1:796 | 6:441 0658 ^ ember, 1928). [Vor. 42 390 ANNALS ОЕ THE MISSOURI BOTANICAL GARDEN Estimated daily needs њу individual members of family. Total need based on these figures Weights in grams Weight à à Calories ж» Protein Minerals Ca. P. Fe One man (age 38 years) .. 100 | 2,100| 45% *500 *900 "015 One woman (age 27 years) .. 80| 2,160 36-0 *400 720 “015 One girl (age 11 years) ds 50 | 1,050 65:0 “600 :900 “010 One boy (age 6 years) = 36 | 1,296 | 50-4] 048 | -900 | +009 One boy (age 6 months) .. 12 Total EP 2% 6.606 1964 | 2.148 3:420 *037 А comparison of nutritive needs with food consumption of an outcaste family of Karimpur Weights in grams —— Calories Protein Minerals Ca. P. F е. Total amounts needed daily .. | 6,816 | 162:5| 1:868 | 3:165 * 050: ee ice consumed daily (May, | 4,468 | 170:19| :714]| 4-690 039 Total amounts consumed daily (Nov- 5,085 | 180:8 :862 | 3:658 *0312. ember, 1928). Estimated daily needs №, individual no = family. Total need based on these figu Weights in grams Weight : : — А Colories in lbs. Protein Minerals One man (age 22 years) .. 105 | 2,205 47-3 “595 -945 *015 One woman (age 50 years) .. 85 | 1,785 38:3 "425 "760 "015 One woman (age 18 years) .. 90 | 1,890 | 40-5 *450 -810 "015 One boy (age 3 years) PD 26 936 36:4 -468 -650 "005 Total 4% a 6,816 | 162°5 | 1-868 | 3:165 "050 1955] WISER— FOODS OF A NORTH INDIA VILLAGE 391 every three pounds of grain taken out to be used, eight ounces are lost in grinding and cleaning. The loss in the case of pulses was not recorded. In the Brahman home, the pulses are soaked and the husks removed, but in poorer homes, the whole pulse is usually cooked. The daily average of carbohydrates recorded as having been eaten by our four village families are as follows: Carbobydrates in dietaries of four Karimpur families Per cent. of р total calories D Family Month y am | Y average) 8 lin form of саг- : qus hydrates Brahman 425 -. | May E e 2,487.67 g. 72 November 2,874.22 72 Carpenter May KA 252 1,385.50 71 November 2,397.38 75 Oil-preser .. .. | May 1,371.80 75 November , 1,311.60 70 Outcaste .. | May 868.4 78 ЕЕ 1,013.3 68 McCay in his jail investigations made a study of co-efficients of absorption of carbohydrates. His table is as follows: Wheat à mun E .. (23 oz) 96.5 % absorbed. Barley a (23 oz.) 96.0 % Р Great millet 97.8 % x Spiked millet 4% i ащ" Сгат 22 Ge (5 о.) 97.7 % т Pigeon pea . . xt zi >. 968 % ” Of fats, McCay says, “We have made по investigations on the absorption of fats of the jail dietaries of the United Provinces. The quantity of fat in these diets is very small and did not appear of sufficient importance to warrant special enquiry." (31). Our figures for the average daily amount of fats used are as follows: (31) See References. (Мог. 42 392 ANNALS OF THE MISSOURI BOTANICAL GARDEN Fats in dietaries of four Karimpur families. рай а се total calories Family Month y 8° | (daily average) (grams) in form o; Brahman AR sap Мар E жа 222.34 14 November oe 2 239.56 15 Carpenter si ось МУ bes tg 110.95 13 November i3 ey 201.28 14 Oil-preser .. .. | May v3 I 86.50 9 November 23 нет 141.19 17 Outcaste :5 zu | МАУ Sia 2% 30.65 November с. i3 126.45 19 The larger amount consumed by all families in November is due to the advent of a special festival. Wheat and ghee were substracted from the original figures of the Brahman family, to allow for cakes distributed to dependants. А lesser amount of wheat and mustard oil was substracted from the figures given by the carpenter, as they too have begging callers at this time. But after these amounts have been deducted, the festival figures still stand high. The Brahmans of the village are reputed as using more ghee than others. This may be accounted for by their comparative prosperity, and by the greater demand for "pakka" food among them. This is discussed under Food Practices. In most proposed dietaries, the calories furnished by protein are listed sep- arately, whereas those of fats and carbohydrates are lumped together. Lusk has collected standard dietaries, a study of which may be useful. “The ‘standard’ dietaries are given below, not because they are inflexible re- quirements in any sense of the word, but merely for the convenience of the reader. The individual standard will ever be controlled by climate, the amount and kind of mechanical effort, by appetite, purse and dietetic prejudice." 1955] WISER—-FOODS ОЕ А NORTH INDIA VILLAGE 393 Standard dietaries for a man of 70 kilograms. (Weights in grams) па hod Voit Rubner Protein 123 Fat sx > 47% 46 Carbohydrates 377 Calories 2,445 Medium work— Protein 118 127 Fat 56 52 Carbohydrates 500 509 Calories 5s is ү 3,005 2,868 Hard work— Protein 222 x өза 5 145 165 Fat ES 42% T 100 70 Carbohydrates 2 MES : 500 565 Calories з оз ста > 3,574 3,36,232 [as in original] The calorie make-up of our village dietaries, compare with these standard dietaries as follows: Per cent. of total calories Constituent Voit Rubner Karimpur (May) Protein > 42% 15--17 18--20 14-- 16 Fat Aw си Y^. 16—25 16—19 6—14 Carbohydrates . . 56 ё. 65—67 62—71 71—78 in carbohydrates, while protein, and ght expect when cereals are used so able rather than animal sources. According to this, our village diets are high especially fat, аге low. This is what we mi extensively, and when proteins come from veget for family dietaries Sherman states, dual preferences as well as to to consider that from 10 to 15 » (33). The daily Protein—In writing of protein standards “To allow for varying conditions and for indivi provide a liberal margin for safety it is customary per cent of the total calories may be in the form of protein. average of protein in our tables is as follows: (33) See References. [Vor. 42 394 ANNALS OF THE MISSOURI BOTANICAL GARDEN Protein in dietaries of four Karimpur families Per cent. of Залы esee total calories Family Month y чи (daily average) (grams in form of protein Brahman m : Мау P са 466.8 14 November n is 440.7 13 Carpenter © ж os | May ко. 5% 297.3 16 November 2% 2% 364.5 11 Oil-preser .. +, | May 2% yx 325.9 16 November 15 e 231.14 15 Outcaste 42 .. | Мау AT Va 170.19 16 November .. зб 180.80 13 Although the protein according to figures in the record of food consumed appears to be more than the amount needed, with one exception (the carpenter family in May), it does not exceed the percentage of calories suggested as correct by Sherman (10 to 15 per cent.). McCay (34) gives the following table of percentage absorption of proteins of the food materials in use in the jail dietaries of the United Provinces. Food stuff Absorption of protein Wheat 2% Fa 22 67.10 Arhar "dal" (pigeon pea) plus vegetables . . um с 80.50 Arhar "dal" (pigeon pea) alone is 154 ids 81.84 Gram "dal" (split gram) 55% = is Pa 64.20 Barley йс» ~ a ја 57.60 Јчаг (great millet) “ч ie Lj La 53.00 "Urd" і с. Е Г. 69.20 “Bajra” (spiked millet) =. 49.40 From this he concludes: "It cannot ђе for the the of the збої that ап average of 34 grams of protein daily should remain incapable of absorption, which, while in the bowel, p a splendid culture medium for the growth of putre- factive micro-organisms." (35). He comments on the importance of the methods of treatment of foods, especially the split pulses, in relation to their absorption. “Gram dal, for instance, (34), (35) See References. 1955) WISER—-FOODS OF А NORTH INDIA VILLAGE 395 is usually given as a morning meal in the form of parched gram, i.e. simply fried or heated until the greater part of the contained moisture is got rid of; no steps whatever are taken to break up the grains so that the digestive juices may have a proper opportunity of carrying out their work satisfactorily. Investigations on the absorption of pea and lentil flour, properly cooked, show that the protein is all taken up except about 8 to 9 per cent. If, however, the pulse is not given in a state of fine division, the loss of protein has been found to rise to 40 per cent." (36). Not only must the absorption of proteins be considered, but also their ability to meet body needs after absorption. This ability apparently depends upon the amino-acids which the proteins yield on digestion. Therefore, "It seem logical to prefer to serve the needs of the body by providing it with the optimal amount of such proteins as can be utilized very effectively rather than to require it to digest and assimilate an excessive amount of proteins of low value." (37). А group of Indians are undertaking a study of the pulses which are the chief sources of protein in the Indian diet. Their first study was that of the globulins of the pigeon pea and field pea, both of which are used in our village. They con- clude that “Both the globulins compare favourably with casein and contain requisite amounts of arginine, histidine, and lysine. The arginine content of both the globulins is higher than that of casein, being nearly double in the case of the globulin of Pisum arvense field pea. The two globulins are, however, very deficient in both cystine and tryptophane. These legumes are usually taken with cereals and are useful in supplying the essential diamino acids especially lysine in which cereal proteins are usually deficient.” (38). Later they studied the values of the lentil (lens esculenta), cow pea (vigna catjang), the aconite bean (the “moth” of our village). They found that all three pulses contain lysine, histidine, arginine, cystine, tyrosine and tryptophane. Their comparison of their utility is: "The analysis of the three globulins shows that the globulin of lens esculenta is characterized by its higher content of lysine and arginine. The net protein values content of digestible protein Ж biological value 100 of the pulse calculated from the metabolism data are given in the following table: Total Protein Net Protein Pulse (N X 6.25) value Per cent. Per cent. Lens Esculenta PM > ЛА a 28.74 = 26 Vigna catjang PEE: v. 2 a 26.52 em Phaseolus aconitifolius эге т Hes 27.09 й (36), (37), (38) See References. [Vor. 42 396 ANNALS OF THE MISSOURI BOTANICAL GARDEN At a 10 per cent. level of intake, the proteins of lens esculenta are highly digestible while those of vigna catjang have a high biological value. Phaseolus aconitifolius is poor in both respects." (39). It is interesting that the aconite bean ("moth") is very little cultivated by our people. They seem to keep it especially for invalids. Lentils are not common in Karimpu McCollum and Simmonds contribute experimental evidence on the possibility of vegetarian diets. They state that in their own earlier experiments and those of Slonaker, vegetarian rats "grew to about half the size and lived half as long as did their fellows receiving animal food." (40). Later, McCollum and Simmonds found that it was necessary to introduce a very large quantity of leafy vegetables into a diet of seeds and legumes, to induce normal growth = development in omniverous animals. Their most satisfactory combination was, “maize 50, alfalfa leaf (dry) 30, and cooked dried peas 20 per cent. The ingredients were ground together so that these proportions were eaten.” (41). Their conclusion is: "It is possible that better mixtures of vegetable foods may be found, but these results show definitely that for the omniverous type of animal, whose digestive tract is small, the consumption of sufficient volumes of leafy foods is impossible." In regard to the effect of a purely vegetarian diet on men they conclude that, “It is possible to make a fairly satisfactory diet of foods derived entirely from vege- table sources, but it is not easy to do so. It would be difficult for man to eat enough leafy foods to enable him to сие with a strictly Бе барар diet. Such a diet will generally be low in protein." (43). Our village people supply the necessary protein from the pulses, or legumes. But a more successful diet has been worked out by the Sikhs and others further North, who occasionally eat meat and eggs, and drink large quantities of milk. The success of the Sikhs is in agreement with the suggestion of McCollum and Simmonds. ‘‘When one adheres to such a regimen (vegetarianism) the addition of small amounts of meat will be valuable, and eggs, because of their relative richness in vitamins A and B, will often be a more effective supplement. The same may be said of glandular organs." (44). The close of their discussion is in- teresting from the point of view of our study. "Many of the people of China and Japan, because of poverty and overcrowd- ing, can afford only small additions of fish, poultry, eggs, or the flesh of mammals to a diet which consists in great measure of vegetable foods. From an early age the children are schooled to eat large amounts of leafy vegetables. Many weeds are eagerly sought after and trees are stripped of their buds in spring to supplement the supply of spinach, cabbage, and other leafy foods." This might almost have been written of our village. It explains the gathering of wild potherbs and the picking of tops of growing plants as well as the eagerness for buds of plum trees. (39), (40), (41), (42), (43), (44) See References. 1955] WISER— FOODS OF A NORTH INDIA VILLAGE 397 Repeated studies have shown that whereas proteins of plant origin may be adequate for full-grown adults, they do not contain all of the amino acids which must be supplied in the building of strong new bodies, and they are therefore not sufficient for growing children. Children should have the protein of milk or milk products, and other animal proteins if possible. Calcium and phosphorus—Calcium is the most consistent deficiency shown in our comparison of food needs and food consumption. Calcium in the Brahman family is higher in May than in November. They also had milk during May, and not in November. The oil-presser's calcium rises in November along with the addition of two quarts of buttermilk. In the outcaste family there was a similar increase in calcium with the addition of milk. Our American dentist in India told us that there was a deficiency of calcium in the milk of Indian cows and buffaloes. He prescribed calcium lactate for our children. My own experience with teeth and pregnancy has caused me to wonder about calcium in Indian foods. There seems to be no lack of phosphorus. I lost two teeth with each new baby. I realized that it must be that my own body calcium was being used to build the baby's bones and drank quantities of milk to renew the supply. If I, with over a quart of milk a day, suffered from calcium shortage or from improper balance of calcium and phosphorus, the condition of village mothers should be worse. And it is. They suffer greatly with their teeth, and of course have never heard of a dentist. The cereals are secondary sources of calcium, and become important because of the large quantities in which they are eaten. Raw cane sugar also supplies calcium. And it is the only sugar used in the village. In the absence of definite information I have credited gbee with an amount of calcium equivalent to that in butter. But I feel sure that it should be less, as some calcium must be lost with the protein Which, separated during long cooking, is strained off. Phosphorus—It is apparent from our figures that we need not be concerned over a shortage of phosphorus in the diet. The cereals and pulses supply it in generous amounts. But we cannot count on American figures to represent accurately the amount of phosphorus in Indian foods. This is one element which is known to vary in different soils. Our chief interest in phosphorus is in its relation to calcium in the work of bone-building. The village children apparently get plenty of phosphorus, but their calcium is deficient. This increases the importance of sunshine in maintaining а proper balance between the two. This will be discussed further, in connexion with Vitamin D. Iron—Of the several surprises which came to me from the comparison of the needs and the foods consumed in the village, the greatest was the iron. I expected to find a striking deficiency here. But deficiency appears only in the carpenter family in November, and the outcaste family in both May and November. [Vor. 42 398 ANNALS OF THE MISSOURI BOTANICAL GARDEN Here again, as with phosphorus, the cereals and legumes are important. ‘Most of these grains (wheat, barley, millet, maize, rice) are eaten whole; they are not subjected to any milling or refining process before use. The outer layers of the grain and embryo, containing valuable dietary constituents, are thus consumed with the endosperm." (45 This use of the whole grains is important to people who are near the margin of deficiency of minerals, especially iron. The green foods which supply our people with iron are not prominent either in May or November. The oil-presser managed to get some spinach. Potatoes appear, in the Brahman family in November, and raw sugar appears in every family as a source of iron. But with iron as with other food elements, we cannot base our conclusions as to its utilization on figures representing its appearance in the foods we eat. "Recent research shows that the percentage of iron contained in a food is not by itself an adequate measure of the value of that food in meeting the iron require- ment of the body." (46). Experiments with different food combinations reported by Sherman have shown different degrees of assimilation of the iron. He points out that further research will be necessary before these variations in iron values from equivalent iron con- tents can be explained. Vitamins—The prevalence of eye infections prepares one for a deficiency of Vitamin А. And a study of the foods eaten shows that a deficiency exists. We do not know just how much A is lost in the heating of milk, its churning and standing and final long heating in the preparation of ghi. According to McCollum and Simmonds A is readily oxidized at higher temperatures, but is relatively stable under conditions generally maintained in the cooking of foods. (47). However, the supply of А from milk is not what we might count on in an American village where butter is used in place of ghi. Mustard oil is not an efficient source of A, as А is absent or present in but traces in fats or oils of vegetable origin. More- over, the percentage of total fat in the diet is low. Carrots are a possible source of Vitamin А. А report on a study of carrots made by Bills and McDonald showed that the most highly coloured carrots con- tained 9.6 mg. carotene per 100 g. White varieties contained only about one- eightieth of this amount. (48). e thin, green leaves of plants have been found to be richer in A than the thicker or whiter leaves. This may help to explain the popularity of the tender tips of leguminous plants. Vitamin B is supplied by the cereals and pulses which our people eat in abund- ance. Fortunately it is not harmed in drying and not easily destroyed in cooking. McCarrison demonstrates by experiment the high value of wheat, and the almost equal value of great millet and spiked millet, in Vitamin B as compared with rice. (49). The amount of B which village people get in their cereal flour is greater (45), (46), (47), (48), (49) See References. 19551 WISER—-FOODS ОЕ А NORTH INDIA VILLAGE 399 than that of the town dwellers who pride themselves on their thin cakes, made of white flour. In vitamin B as in A, the leafy vegetables, including mustard greens, make a small contribution. Meat can only be reckoned as a source of vitamins in a few cases. There is little chance of getting a satisfactory supply of В» for the women and children, since the men drink most of the butter-milk. They refuse meat, and the idea of yeast is repulsive to them. Indian mothers see the growth of our children and ask why they grow so much faster and are so much more vigorous than village children. Vitamins may be the answer. We have talked in terms of vitamin C during the past few years. But long before it was called this, an antiscorbutic value was attributed to foods containing it. As has been mentioned, the tamarind was regarded as an antiscorbutic food, in ancient Sanskrit medicine. McCay in his investigations quotes the rules from the jail code, made before 1911. “1061.--Вегусеп the first of April and the first of November when the supply of antiscorbutic vegetables is very indifferent, Provision should be made for the daily issuance of an allowance of lime-juice, tamarind pulp, or amchur (dried green mangoes) to each prisoner." (50). Mangoes are probably our most important available source of C. The follow- ing report promises further information on its potency. Several varieties of mango were studied. "The pulp of the “Alphonso” is stated to be one of the most potent known sources of vitamin C and compares favourably with lemon juice, while “cawaije Patel" is only slightly less active. The 'shundrya' variety is comparatively poor. "The vitamin A content of the pulp of all three varieties is comparable with that of butter. The "Alphonso" variety was found to be rather better than the other two." (51). 4 Not knowing what variety of mangoes we have, we hardly know what this study may mean to our village. But it promises further helpful information. = study ы the vitamin C content of Indian — made in the “or research —— tories, Coonoor, in 1935, adds ны other food. e list M sources of vitamin C. Among the foods. аи Т in our tee which supply vitamin o good ould inclu > uvas, um dex eaves, green mango pickle (tender, preserved in salt), green n chillis саза paka cabbage (fresh жоро leaves), bitter gourd, cauliflower, spinach (fresh), and turnip. Long standing, or мар in open vessel endangers the vitamin С, Our village babies get less vitamin C than their elders. Many babies have been brought to me who seem similar to those described by McCollum thus: “Тһе baby may become pale, lose its appetite and cease to gain. This condition has been termed ‘latent scurvy,’ and is more to be feared today than the florid, acute form.” (52). (50), (51), (52) See References. [Vor. 42 400 ANNALS OF THE MISSOURI BOTANICAL GARDEN With the balance between calcium and phosphorus as doubtful as it is in our village, vitamin D becomes especially important. Мо one in the village buys cod- liver oil, as it is too expensive. Fortunately for our children, they are generously supplied with D through exposure to sunshine. We are not certain of the optimal content in the diet of calcium and phosphorus, but we do know that "when the skin is irradiated with light containing a suitable content of ultra-violet rays, normal calcium and phosphorus metabolism will be maintained, even when the proportions of the bone-forming elements are unfavourable.” (53). This refers to unfavourable proportions, but may not apply in case there is actual shortage of calcium or phosphorus. In surveying the foods eaten by the people of Karimpur, in relation to their needs, we are aware of deficiency, but not any one specific deficiency. We know that the men are not as large, and have not as much resistance to infection as their countrymen, the Sikhs. We know that the women and children are not suffering from any striking illness caused by absence of certain food elements, but ‘they are not as strong as they should be. Figures do not demonstrate many con- spicuous faults in the diet. And yet we know that there is not the highest degree of nutrition which might be obtained from the foods available. Rainy Season diets—not analyzed. Diet of a Brabman family in Karimpur* à; butter- | Salt Date Cereal in bread Split pulse Vegetables milk (Butter- August, 1927 Ib Ib. oz. Ib. oz ilk) on qt. 13 Wheat Sr 57% Godd .. 5 8 17 .. 14 с 7 | Field peas j £6 i 2 Barley T 7 15 Whet .. 7 Mung 7. 2 15 б : Barley "an 1 16 ea .. 14 Pigeon pea 1 6 EE oe ж 17 Ж + 7 | Urd is 1-6 V 42.” СЕ 18 13 7 | Urd .. 2 15 | Mushrooms 0 12 Barley .. 7 ^ 19 У... м Egypt arum 2 13 2 16 Сона 1 $ Ји *Three women, three men, three children. (53) See References. 19551 WISER— FOODS ОҒ A NORTH INDIA VILLAGE Diet of a farmer family т Karimpur* 401 Milk Date Cereals Vegetables butter- | Fruit Salt ili (Butter-| (Dried August, 1927 milk) | mango) Ib. oz №. oz. | «t. 02. 02. 13 Wheat 4 4| Gourd ж.ж & 1} ““ 14 Barley 4 4| Mushroom .. 0 12 1} А 15 ” „4 й 4% ц 12 . 16 - . 4 4| Mushroom 0 12 1} ь 17 " с 44 Е 1} . 18 » ға es в . 19 i 4 4| Gourd 14 1} | 8 *Two men, one woman, one child Diet of an outcaste family in Karimpurt Split Date Cereal eld Vegetable Sweet | Salt pea) August, 1927 Ib. oz. Ib. oz. Ib. oz.| Ib. oz.| Ib. oz. 13 Wheat flour 6 0 А Egyptarum 2 0 4 .. 14 Ao .. T 9 ; Mushrooms 0 12 E .. 15 у Но 2 5 .. Wild potherb 1 6 oe se Field pea flour 2 5 (nari). ck pea flour 2 5 16 Wheat flour 2 5 1 6 v T .. Barley four 2 5 Chick pea flour 2 5 17 y flour 2 12 .. . .. Field pea flour 2 12 flour 1 6 18 Wheat flour 2 5 .. Wild potherb 1 6 .. Barley flour 2 5 (nari). Chick pea flour 2 5 19 Wheat flour 7 0 Ку Wild potherb 1 6 1 2 0 12 (nari). TThree men, one woman, two children. (Мог. 42 402 ANNALS OF THE MISSOURI BOTANICAL GARDEN Diet of a Muhammadan family in Karimpur* (Most of the food for this particular family is begged from Brahman farmers Date Cereals Split pulse Vegetables Meat Salt ugust, Ib. oz. OZ. Ib. oz. Ib. 02 1927 13 Wheat 38 . Egyptarum 0 12 .. 14 Wo 19 ix Wild potherb 0 12 ық 5% (пагі). Barley 1 2 Chick pea 1 2 15 Barley 110 .. Egypt arum 0 12 .. “ Foss .. 110 .. 16 Wheat з 8 Wild potherb 0 12 e» as 17 ээ ee 1 10 i eon .. Фе "uu ae Barley 110 18 Wo € кақ .. Egypt аппа 0 12 .. .. Peas .. 110 19 Wheat з 8 m 2 5 *Two men, one woman, one child. 1955] WISER—-FOODS OF A NORTH INDIA VILLAGE 403 VI—CONCLUSION RECOMMENDATIONS Several years ago an English deputy commissioner set out to improve conditions in a whole government district, not far from ours. His faith was great. In the record of his work he writes, “Тһе secret of all success in the villages is propaganda, and the campaign that has been going on for the last seven years has been wonder- fully successful." "Things undreamt of before, things supposed to be utterly opposed to every custom and sentiment, have come to pass easily and naturally, all owing to con- tinuous and intensive propaganda." (54). If my confidence were as great as his, I would go first to the village elders, and then to the mothers, with the suggestions which follow. To the elders I would suggest: (1) Start orange groves, like those in Kaimganj, fifty miles away. (2) Replace the old mango trees with new ones. (3) Grow more potatoes. (4) Learn from the co-operative society in Fatehgarh, forty miles away, how to store potatoes. (5) Allow the vegetable growers to raise tomatoes. (6) Advise the vegetable growers to add more leafy vegetables such as cabbage and chard to their crops. (7) Get information from the Mainpuri Government experiment farm, on how to get a better milk supply. We have made our contribution to this by securing a high grade bull from the provincial government. (8) Pass a ruling that fruits must not be picked, or knocked off of trees, until they are ripe. Explain to the men and boys that they will get more from the larger ripe fruit. If I could find mothers willing to forget custom, follows: (1) Use wheat for the children as far as possible, barley for festivals and adult visitors. (2) If there is little wheat, it is be rather than use it up and then give the others, (3) Grind all flour finer for the smaller children, cloth as you do now, for festival cakes. (4) Use pigeon pea in preference to ot ally, for variety, preferably “mung” or “urd.” I would advise them as and keep the millets and tter to mix some of it with other cereals such as millets, alone. and sift it through coarse her pulses. Use the others occasion- (54) See References. (Мог. 42 404 ANNALS ОЕ THE MISSOURI BOTANICAL GARDEN (5) Do more than split the pulses. Break them up with the light mill. Or better still, soak the pulse for one day and night; spread it out in a shady corner of the courtyard and keep it covered with a damp cloth until it sprouts. The children will like it. (6) Ask husbands to give the children more of the buttermilk. (7) Or give each child a glass of the fresh milk each day. Give each one two or three glasses, if possible. This takes for granted that there is а milk animal—as there is in most hom (8) On days when there is milk for the children, give them less pulse. (9) Rather than fry cakes in “ghi” for the children, toast the cakes and spread them with fresh butter. (10) If the butter used means a shortage of “ghi”, mustard oil can replace the "ghi" for most purposes. (11) Get eggs, if possible, so that each child will have one or two a week, and gradually increase to four or five a week. (12) Cook potatoes at least four times a week, if possible. They may be combined with other vegetables and cooked as at present. (13) Use dasheen (Egyptian arum) during the months when potatoes are not in season. (14) Cook whatever vegetable is in season, along with potatoes or dasheen. Most vegetables require less cooking than potatoes, and can be added later in the cooking process. (15) Let the children eat raw carrots, potatoes and cucumbers when there is no cholera in the district. (16) Try drying carrots, by slicing them thin and spreading them in the sun—as mangoes are dried. Use them in vegetable mixtures when gourds and pumpkins are the chief vegetables in season. (17) Have a green vegetable at the meal when "kachcha" food can be most easily served. The children can gather fresh potherbs in the rainy season and early fall; and fenugreek and the green tops of peas, gram and mustard plants in the late fall and winter. These same leaves can be dried for the spring, and supplemented with fresh radish tops. (18) Let the children gather and eat each fruit in its season but encourage them to let the fruit ripen, at home in the sun if not on the trees. (19) When husbands go to the bazar, ask them to bring home oranges in season, or tomatoes, if they are cheaper. (20) Plan the children's and your own meals something like this: Morning—Left-over cakes of wheat or mixed flour, or “satua” (parched cereal) ground finer than for the men. МИК or buttermilk with either one. Allow at least 1 pao (1 cup) for each of the children. Fruit, if there is any. Noon—Fresh cakes of wheat or mixed flour. Vegetables,—potato and one other. Pulse broken up and cooked thoroughly, with mango, fresh or dried, to flavour it. Or sprouted pulse. 19551 WISER—FOODS OF A NORTH INDIA VILLAGE 405 Night—Cakes left from noon, with green vegetable cooked in usual way, one cup of milk or buttermilk for each of the children. (21) Set regular meal hours for the children and yourselves, as for the men. Ask the children to bring home the eatables they collect from the fields and groves, and share them with you at meal time. (22) Let them play in the sunshine as they do now, with as few clothes as they now wear. For the mother of a new baby, I would add the following: (1) Nurse him for two years, if you want to, and if neither you nor he loses weight begin supplementing your milk with goat's or cow's milk, when he is about a year old, so that he will accept it when your milk is insufficient. (2) Eat the same kinds of food that you prepare for the children but eat more of them and drink milk if you can possibly get it. You are feeding the baby as well as yourself. (3) If the baby shows no gain in weight, or cries a great deal, start giving him other food, gradually. (4) Do not feed him whenever he cries, but try to feed him four times each day, when the shadows are at the same place. (5) When you start giving him other food do it gradually, beginning with boiled goat's milk diluted with water. You can give him some of the rice water which you drink. Put a little partly ground wheat in a small kettle if you have it, and let it stand beside the milk that is simmering. When it has cooked several hours, strain it and give the liquid to him. When he is about a year old, you can give him some of the potato and other vegetable which you cook, before adding the spices. Mash the vegetable with your cooking spoon. Give him some of the liquid from the green leaves which you cook, before adding spices. Let him have a taste of the egg-yolk prepared for one of the older children. As he gets used to it, add more, until he has one yolk each day. When he has plenty of teeth for chewing, give him a piece of toasted bread, hard enough so that he must chew it before swallowing. Be sure that the flour used in his bread is finely ground. ТЕ you cannot get tomato or orange or cabbage juice for him, give him some raw potato, scraped off with the bowl of the iron spoon. As he gets older, give him some of the split pulse cooked for the family, but cooked longer, and mashed. Better still, cook some of the sprouted pulse for two or three minutes for him. Cook the milder fruits which the children bring home, sweeten them and give him the juice. Later he can have the pulp. As he grows, increase the milk, to at least half a quart. This might go on indefinitely, with all of the detailed instructions used in America for new mothers. It may sound inadequate from the Western point of view. But it is so far beyond present practices in our village that it sounds like the chapter in Mr. Brayne’s “Remaking of Village India”, called “The Dream”. [Vor. 42 406 ANNALS OF THE MISSOURI BOTANICAL GARDEN The reasons for most of the foregoing suggestions are obvious. The suggestions to the village elders are explained by what I have already said of present custom and the shortcomings of the local food supply. The elders demand that each proposal will bring them profit, economic or religious. Health has not yet become equally important with these two. Furthermore, the elders must be satisfied that the proposals are worthwhile before anyone else in the village will consider carrying them out. All this demands more tact and resourcefulness than scientific infor- mation. The reasons for the suggestions concerning foods for children are as follows: Wheat yields more fat and more minerals per 100 calories than the other cereals do. Moreover experiments have shown that the proteins of wheat more success- fully supplement those of pulses, or legumes. McCollum and Simmonds in studying vegetable proteins conclude that "There are however, some remarkable instances of effective supplementing between certain cereals and certain legume seeds. Con- spicuous among these successful combinations is wheat and pea. . . . Maize and pea, fed in combination parallel to the wheat and pea relation just mentioned, was almost a failure for the nutrition of the rat. Animals confined to this diet, the only fault of which lay in its protein moiety grew slowly and remained stunted. Fertility was very low and the second generation increased very slowly in weight.” Of the combinations used in their experiments, only the wheat and the maize combined with peas are like our village diets. Our village friends tell us that they used to eat more wheat. But since foreign trade has been introduced, the price of wheat has gone up from 40 pounds to the rupee, to 16 pounds to the rupee. As a result, they sell as much wheat as possible, and hoard the money, while living on inferior grains. The pigeon pea is recommended, as a source of protein, partly because a larger percentage of it is absorbed than of other pulses. This was shown by McCay’s work, reported on page 396. Also, the studies of Niyogi and his co-workers have shown the pigeon pea to be a comparatively satisfactory source of protein, when used with cereals (page 397). However, the net value of the protein of Indian Pulses is shown to be only about 50 per cent. of the total protein content (page 397). This means great waste, and unprofitable expenditure of energy in its digestion and assimilation. For this reason, milk or buttermilk are proposed to replace a part of the protein now contributed by pulses. A larger percentage of the protein of milk is assim- ilated than that of pulses, and the amino acid content of milk protein is more complete. At the same time, milk supplies other materials, such as vitamin A, and provides the calcium in which village diets are conspicuously deficient. Be- cause of these qualities of milk, McCollum calls it a "protective food.” "Milk an the leafy vegetables are the only foods so constituted as to make good the deficien- cies of cereal grains, legume seeds, tubers, roots and muscle meats. For this reason, McCollum proposed that these foods should be designated "protective foods” in order to give them the prominence they deserve.” (55). (55) See References. 1955] WISER—FOODS ОЕ А NORTH INDIA VILLAGE 407 A number of experiments have been made, testing the effect of milk on chil- dren. The following from New Zealand where life is comparable to that in our village, is of interest to us. It is reported in Nutrition Abstracts and Reviews, Volume II, no. 2. А test group of 93 Maori children (age 5 to 14 years) received fresh milk daily (1 to 1 pint per child) for 13 weeks, during which period records were kept of their height and weight. Children of another school provided a con- trol group. The children in the test group showed twice the increase in height and two and a half times the gain in weight of those in the control group. A marked decrease in the incidence of scabies and impetigo was noted among the children receiving milk. There also seemed to be an improvement in their mental alertness. (56). Butter is suggested to replace ghee on the supposition that in the long standing and heating preparatory to the final production of ghee, there is a loss in vitamin A. And the village diet needs a supply of vitamin A badly. This is evidenced by the sore eyes and the corneal ulcers shown us by parents who hope that we can magically restore sight. When the butter is heated and strained, much of the protein is strained out, and some of the minerals undoubtedly go with it. We have no exact information on this. Ghee is such an important item in diets all over India that it will surely receive examination from Indian nutritionists. Egg yolk is suggested as another possible source of vitamin A combined with fat and iron and protein of high value. But the adoption of eggs is very doubtful among caste Hindus. For this same reason, meat was not suggested as a children’s food, in the list above. The prejudice against it is even greater than that against eggs. Moreover, the meat now available comes from old, decrepit animals, an Carrots are advised chiefly because of their vitamin A content. | Potatoes are suggested as a substitute for part of the cereal, now consumed in such large quantities. They are a cheap source of carbohydrate, and at the same time bring in certain constituents such as vitamin C, and iron, which cereals provide in lesser amounts, in proportion to calories supplied. : e germinating legume seeds, and green vegetables are easily obtained con- tributors of vitamin C, along with the mangoes and tamarinds, and certain other fruits. Oranges are within reach, and may come into use gradually. But the introduction of tomatoes, while simpler from the physical standpoint, will depend on influence other than soil, or nutrition. All of these suggestions are general, and will require specia individual cases. I have no illusions as to the difficulty of securing their adoption. Back in the village, I shall be bound by the circle of complicating factors. But each suggestion will find reception somewhere. In high caste homes, where eggs are taboo, there is usually an abundance of milk, and great respect for any product of the cow. In outcaste homes where there is little or no milk, chickens are kept. Even in these outcaste homes, a goat is possible, thanks to free grazing. | application in (56) See References. [Vor. 43 408 ANNALS OF THE MISSOURI BOTANICAL GARDEN The feeding of babies during weaning, as given in the list for new mothers above, is a task which we must undertake seriously. I have gone through the ex- perience with mother after mother, of watching a beautiful baby wither, and feebly slip away. And we have not known what to до. Also, I have gone through the suspense of the feeding of my own babies, and appreciate what it means to have guidance at this time. Ас present some of our babies gradually starve to death оп mother's milk, while their mothers pride themselves on the fact that the babies want no other food. Others find themselves suddenly replaced by а new brother or sister, with coarse cereal cakes and stewed pulse as their new fare. There is a disease called "sukhi" (drying up) among our babies. We could not locate its source, but now I believe that it can be traced to one or perhaps both of the above situations. There are various methods of treating “sukhi,” chiefly with charms and taboos. But I have found no suggestion of its relation to faulty nutrition. It is one phase of feeding which I want most to study on our return to the village. THE PROBLEM In approaching this problem of foods in our Hindu village of North India, we should inform ourselves on the following: (1) The people, their physical condition, including resistance to disease. (2) The attitude of the people towards suggested practices. (3) The foods available, locally, with some information on the comparative consumption of each. (4) The treatment of the foods, including customary practices. (5) The value of foods used. We have acquainted ourselves with the first, third and fourth of these through personal observation and inquiry. With the help of medical friends we hope to learn more of the physcal condition of the people when we return among them. Our information on the fifth point comes from various sources, chiefly Western. This field has scarcely been touched in India. The second opens up a whole new field which we hope to investigate. In trying to solve the problem of how to make the foods of our Hindu village more adequate, further points must be kept in mind. (1) The foods emphasized should be among those available and those ac- ceptable to at least a part of the community. (2) The proposals in regard to the treatment of food should be built on familiar practice and should be made with consideration for existing custom and prejudice. (3) The values of foods recommended, and the treatment of them sug- gested, should be built upon evidence presented by reputable scientists. I have kept these points before me in presenting solutions of the problem. With my own knowledge of the foods available, and of current food practices, and with the support of McCarrison, McCollum, Rose, Sherman and others, I know that my suggestions are harmless, at least. With the example of the Sikhs who live not far 1955] WISER—FOODS OF А NORTH INDIA VILLAGE 409 from us and who have practically the same food resources as our people, I know that we are striving toward something attainable and worthwhile. The problem has resolved itself to a simple matching of human needs and a comparatively satis- factory food supply. McCarrison has expressed the result very simply in his Primer intended for Indian school children. “The right kind of food for Indian children and indeed, for children in any country is one made up of the following simple things: (1) any whole cereal grain or mixture of cereal grains; (2) plenty of milk and the products of milk—curds, buttermilk, butter, ghee; (3) sprouted pulses; (4) eggs or liver, or meat, or fish occasionally, if religion permits their use; (5) tuber and root vegetables; (6) abundance of green leafy vegetables; and, (7) fruit.” It sounds very simple. And we have satisfied ourselves that it is right. But the hardest task of all remains. Our village folk are illiterate and their lives are directed by custom, and fear of spirits. We cannot give them the reasons which Science offers for adopting the new food practices recommended here. The propaganda method of the English commissioner did not prove itself lasting after his departure. Then, how are we going to help our people to feel a need for better food to such an extent that they will be ready to improve their food resources, and improve the methods of utilizing those resources? This step carries us from the present study into a whole new field. On our success in exploring this new field, the effectiveness of our solution of the: food problem reste Же must carry our information back to our village and translate it into village terms, with the help of educated Indian friends who can interpret for us. They know the world of spirits in which villagers dwell, better than we do. An illustration of this is one of the many stories told about a gallant Indian woman lawyer who goes into remote sections of the country in connexion wit her official duties. Once while visiting a home where her women clients were particularly superstitious, she found them in terror of the threatened visit of the smallpox goddess. The goddess had already brought tragedy to several of their neighbours. They were using all of the charms within their reach to persuade her to pass their house by. The educated Indian woman knew the futility of urging She asked the women if the smallpox goddess did not did. Then said the educated he smallpox goddess by blood vaccination in this home. like a sacrifice of blood. They agreed that she woman, "I know a spirit charmer who wards off t sacrifice. Each of you may share in the offering." They had already employed the services of local spirit charmers, and were ready to try another. The educated woman drove to the nearest town and brought the district vaccinator back with her. He was allowed to scratch the arm of every one in the family. They were warned that in a few days the goddess would come, but learning of their blood sacrifice, she would pass on without the usual terrible visit. Women like this, rather than scientists, can help us in the study of present attitudes, and in the task of interpreting in terms of those attitudes. Education is slowly reaching the villages. But until it is much better than at present, we do not want our people [Vor. 42 410 ANNALS OF THE MISSOURI BOTANICAL GARDEN to cast everything of the past aside. In so doing they would lose much of value which experience has taught them. I should like to close this study, with a suggestion of the future of nutrition in India. In 1928, the following statement was made by the Royal Commission on Agriculture: “The problems of human nutrition, which have only recently come into prom- inence are being investigated by Colonel McCarrison who is working under the Research Fund Association at Coonoor. Не lays emphasis on malnutrition as a problem facing those engaged in agricultural research. In his inquiries, Colonel McCarrison invokes the aid of the agricultural departments. The problems of human and animal nutrition are likely to assume such importance that we consider it desirable that work on human nutrition and on the nutrition of farm animals should be carried out in the closest co-operation, or, in other words, that there should be team work by workers with a knowledge of different branches of the science of nutrition. Continuity is also essential. The various workers on nutrition problems should be formed into a committee on nutrition which would meet at regular intervals to discuss common questions. This would assure the requisite close touch between different branches of the subject. In view of the importance of the subject, we recommend that a central institute of human nutrition should be established." (57) Since this report, the important work carried on at Coonoor has continued, under the direction of Sir Robert McCarrison, until his recent retirement. His successor, Dr. W. R. Aykroyd, has a record of thorough scientific research. То this he is rapidly adding an interest in ways of meeting the nutritional needs of poorer families in India. Не has already worked out an adequate diet which is low in cost, and available in villages. Meanwhile, further research is being con- ducted, and applied practically in the All-India Institute of Hygiene and Public Health, Calcutta. There are other centres of useful research, in Calcutta, Banga- lore, Bombay, and Allahabad. Indian food-stuffs are being analysed, so that we shall no longer depend upon Western tables. Ordinary diets are being tested, both in laboratories and in boarding institutions. e most recent move has been the correlaton of these scattered studies. During the last Indian Science Congress, in Indore, in February, 1936, an Indian Nutrition Committee was formed. And still more recently, our new Viceroy revived the above quoted suggestions of the Royal Commission of which he was chairman, at the opening meeting of the new Nutrition Advisory Committee of the Indian Research Fund Association. We look upon this committee as the beginning of active, unified interest in human nutrition in India Meanwhile, public interest in this subject has been growing; popular lectures on diet problems now have their place in gatherings to promote public welfare. Leaflets are being prepared like that issued by the Bombay Presidency Baby and Health Week Association on Balanced Diets. Newspapers and magazines now publish (57) See References. 1955] WISER—FOODS OF A NORTH INDIA VILLAGE 411 more articles on nutrition within a month than they did in a year when the Royal Commission first proposed an institute of human nutrition. АП of this is bound to influence the literate townsman. But much more must be done, to reach the illiterate villager and his wife. Perhaps establishing literacy is the first step. To quote from the Viceroy's speech at the opening of the Nutri- tion Advisory Committee: "I am entirely convinced that, whether we are con- cerned to advance among the rural population improved agricultural practice or better sanitation and hygiene or better nutrition, we shall find that our best en- deavours will achieve little that will endure unless and until literacy is imparted to the women of the countryside. We may bring about some considerable degree of improvement under existing conditions by means of intensive propa ganda and close supervision, but in default of female literacy it will be found that whenever super- vision is removed there will be a relapse into age-old customs and that, within a few months, nothing will be left of the better living that has been so laboriously inculcated." While literacy is being made possible, we must find ways of present- ing better nutrition in terms of a villager's experience. Не is suspicious of any- thing new or strange. If we press upon him, ideas which have not been carefully tested, or ideas which involve drastic change, we shall strengthen his prejudice against what he regards as interference. But he and his wife will gladly accept one simple suggestion at a time, if they understand its benefit and see that this benefit can be attained without additional cost. They are too poor to risk even a pice in experimenting. And while their simple, very practical learning is going on, we can educate their children toward an understanding and appreciation of the pur- poses and results of Science. REFERENCES* (1) Church, A. H. Food-Grains of India. Chapman and Hall, London, p. 73. (2) Mainpuri Gazetteer, p. 4. (3) Moreland, W. H., Indian Civil Service. by Government Central Press, Allahabad. (4) Mainpuri Gazetteer, p. 3. ‘Se Mukerjee, Radhakamal. The Rural Economy of India. ро 31. Agriculture of the United Provinces. Published Longmans, Green & Со, 1926, 6) Wiser, У. H. Social Institutions of а Hindu village of North India, p. 17. 7) Moreland, W. H. Agriculture of the United Provinces. Government Central Press, Allahabad, pp. 131—132. (8) Church, A. H. Food-Grains of India. Chapman and Hall, London, p. 98. (9) McCay, Major D. Investigations into the Jail Dietaries of the United Provinces. Scientific Memoirs (New Series), No. 48. Superintendent, Government Printing, Calcutta, 1911, p. ‚ (10) Blunt, E. A. H., Indian Civil Service. The Caste System of Northern India. University Press, Madras, 1931, p. 88. (11) Bose, Basanta C., M.A., вл. Hindu Customs in Bengal. Calcutta, p. 99. $ 1 -at-Law. The History and Status of Landlords and Tenants . The Pioneer Press, Allahabad, 1931, pp. 215 an Caste and Race in India. London, Kegan Paul, б-а The Book Company, Ltd., U; 216. (13) Ghurye, G. S., PH.D. Trench, Trubner and Co., Ltd., 1932, p. 7. (14) Blunt, Е. A. H., ор. cit, p. 89. * As in the rest of the paper, the citations are given exactly as in the original text. [Vor. 42 412 ANNALS OF THE MISSOURI BOTANICAL GARDEN (15) тергеген ~~ чення Mal Mouncha, Rai Bahadur, in "Hindu Ноте Life.” London Printing Press, Luck (16) Stieglitz, T а. ані in Medicine. Chemical Foundation, New York, 1928. Dietary "m pet by J. M. Gamble, M.D., (17) Annual Report ~ the om itary Cani sioner with the Government of India for 1880. ‚ Op. cit., р. and 336. [as in original.] ор. си., р. 3 (22) McCay, D., ор. cit., p. 23) McCarrison, В. pris Die and a Bad One: Ал шә! Contrast. Indian Journal of Medical Research. Vol. ps No. 3, Jnd 1927, p. (2 cCarrison, R. ood Diet and a Bad One: ye Experimenta Contrast. Indian Journal of Medical Research. Vol. Mic No. 3, Janvary, pes Р —654. (25 typed copy of this article was sent to m С el МсСагг (26) Dietary Scales and Standards for ти : Family's Мини ој Technical Bulletin No. 8, U. S. Department зі Agriculture, November, 1927. 27) этан Н.С. Chemistry of Food, Fourth editas p. 320 ff. (28) Rose, Marie: Robb ind Bloomfield. Journal of Nutrition, Vol. III, November, 1930. (29) Rose, Mary Swartz. Foundations of Nutrition. MacMillan & Co., 1930, pp. 170 and 183. (30) Xo Mary S. еі the Family, Third чы MacMillan & Со., 1929, pp. 118 and PT 1) — ме: эне, ре} (32) Lusk, The. е of the Science of Nutrition, Fourth edition. №. B. Saunders, а Желін 1526 р. 459. (33) Sherman. Chemistry of Food and Nutrition, Third edition, re-written, 1931. Мас- Millan & Co., p. 541. (34) McCay, op. cit., Т, 12). 13 (37) а and. ы. Newer Knowledge of Nutrition. MacMillan & Co. New York, =, (38) М 2% В., Narayana, N., Desai, В. С. dies in Nutritive Value of Indian. Vege- table Food- M Part L Nutritive Values T Pigeon X "Cajun: us Indicus) and Field pea (Pisum arvense, Linn.). Indian Journal of Medical — 1930-31 XVIII, p. 1 (39) Niyogi, S. B., Narayana, М, Desai, В. Studies in Микене Value of Indian Vege- table Food-Stuffs, Part III, Nutritive Values of il, Lens CMT Moench; cow pea, Vigna fo 7х and Асопие Bean, Phaseolus Aconitifolius, Jacq. . P. Niyogi, N. иу В. 6. ву. McCollum and Simmonds, op. cit. (41) McCollum and Simmonds, op. cit., p. 149. (42) McCollum and Simmonds, op. cit., p. 151. (43) McCollum and Simmonds, op. cit., p. m (44) McCollum and Simmonds, cit., р. ; x (45) McCarrison. Problems of Nutrition i in fen Nutrition Abstracts and Reviews, Vol. П, D Жур (46) Sherm , op. ез p. 321. (47) Мс Collum and Simmonds. On the Story of the зара of the Vitamins. Chemistry of Medicine. Edited by т е Chemical Foundation, New York. : (48) Bills, e H E nd McDonald, Е. С. The Carotene e of Ten Varieties of Carrots, Science, ds m oo эз (49) М The Nutritive Value of Wheat, Paddy and certain Other Food-Grains. Indian Tourn a of Medical Research, Vol. XIV, No. 3, January, 1927. (5 Pony А cit., p. 19 (51) Р Е. О. V Mer iva. S. S. Preliminary Report % Vitamin Content of the Mango, Nutrition нення and Reviews, Vol. II, No. 1, 1932, pp. 60, (52) McCollum and Simmonds, op. oe Р. 16 (53) McCollum and ери and Stieelitz, op. (54) Bravne, F. Remskine of Villas In dia. Tondon, із», р. 64. (55) Td and кендерді. ор. cit., Fourth edition, p. : (56) Turbott, H. B.. and Rolland, A. Е. The Nu oll Value of Milk. Experimental evidence from Maori school children. М. Z. Medical Tournal. 1932, 31, р. 109—111. dc ag п оп pisi асы in India. Abridged report. Government Central GENERAL INDEX ТО VOLUME XLII New scientific names of plants and the final members of new combinations are printed in bold face type; synonyms and page numbers having reference to figures and plates, in виђа; and all other matter in ordinary type A Acanthothamnus aphyllus, 285 Acarospora (Pachnolepia) Gri, 144 Adélie Land, lichen flora of, 14 Adlay, I ет morphology of, 1727: Mou. dt 45 Allosorus an тна 102 Alsodeia glabra, 2 Anaxagorea Allenii 151; crassipetala, 151; dolichocarpa, 151 Animal products, as a source of food in Karimpur, 309, 344 Annonaceae, Three new, from Panama, 151 Antarctic continent and the subantarctic islands, Lichenological notes on the flora of t Fe Tek 1 peto 113 Apluda айыы. 40% Argopsis cymosa, 134 Asiatic relatives of maize, Studies on, 103 Aspicilia арка 135; Лють 135 tha Australian Nat tum, дивне Research Ex- pedition, licen collected by, 1 B Balansia trinitensis, 136 Blastenia keroplasta, 135 "Blue Adlay”, 107, 108, 709; chromosome morphology of, 115, 127, 128 Botanical catalogues of Auguste de St. Hilaire, 153 Brahmans, food habits of, 307, 379, 400 — чоно, collectors in, 153 Briz Buellia 674 143, 146; — 146; Mawsoni, 143; subplicata, C Caloplaca crozetica, 1 Catalogues, botanical, 2m Auguste de St. Hilaire, 153 Catha benthamii, 244; monosperma, 249; vitiensis, 291 Catillaria mei ii есе 139 Ceanotbus var. ' papillosus, зат, уаг. punctatus, се И 236; ре: 224. 250; chef Til 244; АН ҒЫ 285; bivalvis, 285; bolivia. 285; bodinieri, 285; bullatus, 140; buxifolia, 285; и m 249; case folius, 282; траје , 265, 286; с mo 4, 246; а брини 282; chungii, "EI c col confertus, 286; crassifolia, 273; crenatus, 286; crenulatus, 286; crispulus, 266; cuneifolius, 286; cunninghamii, 286, var. parviflora, 286; dependent, 231; dilatatus, 286; discolor, 285; disperma, 286; diversi- folia, 286; dubia, 286; fant 267; emarginatus, 286; esquirolianus, 286; esquirolii, 287 ; Боба а 139 287; "eupble- adest 231; fasciculatus, 287; feddei, 7; finlaysonianus, 287; flagellaris, 276; floribundus 287; fournieri, is gemini- 287; hindsii, 249, var. d 249; sutus, 270; hookeri, : bookeri, 274; bypoglaucus, 257; bypoleucus 257; 5. Р. 287; кыю. gs lancifolia, 261; laotica, 231; latifolius, 239; leiocarpus, 267; lenti- cellatus, 278; leptopus, 287; 279; linearis var. gascariensis, 287; (413) [Vor. 42 414 ANNALS OF THE MISSOURI BOTANICAL GARDEN lineatus, 287; listeri, 288; loeseneri, 265; 291; beer 291; yunnanensis, 291; lokcbongensis, 258; longe-racemosus, 267 ; zeylanic longipes, 281; lucida, 291; lucida, 288; Cen "отус A ари 134; sarmentosa, і i 40 288; madagascariensis, 238; magellanicus, 2. used in North India, 329 288; mainsiana, 279; mairei, 288; malay- ^ Chailletia gelonioides, 285 ensis, 246; marianensis 250; mauritiana, Chionachne, 103, 106, 108, chromosome 255; meridensis, 284; merrillii, 250; тои а 30; mexicanus 107, 111, 112, 113, chromosome morph- olo ову of a 125, 130; Massii, 106; Нее 136; subtropica, 136 hromosome morpholog ву, ЈЕ Coix, Poly- h 231; w lorus, не esi 272, toca and ionachne 288; nepalensis, 288; novoguineensis, 234; nutans, 226% or iy и. 274; oblongi- romosome numbers а Сда 2233 folius, 250; obtusatus, 288; миди, in Сох, 113, in Selaginella, 1 288; ERE 288; opposita, 289; or- Cladia aggregata, 140 bulis f. major 261, f. е ills Cladonia Шекесі 140; Johnstoni, S: 261, var. aureo-arillata, 261; orbiculatus, Mawsoni, 140; phyllophora, 134; sarm 60; orbiculatus, 253, f. papillos sus, 261, tosa, 140 var. даро, А хаг. ад 261, var. Coenogonium subtorulosum, 137 punct 266; а, 289; ov alifolius, Coix: cytological study of, 107, materials 289; sends 289; poe CER 289; pachy- used, 107, methods, 113; history of, 103; ғасын, 279; pallidus, 289; dryer nomenclacure in the genus, 104 280; paniculatus, 229, var. andam Coix, 103; agrestis, 104; aquatica, 104, 230, var. Balansae, 236, ssp. multiflorus, 108, 113; arundinacea, 104, 106; barbata, е9 ssp. Mee e E iro 106; chinensis, 104; from Cuttack, India, 5 ; v ; i psoides, 106; exaltata, 104; ато, 289; parvifolius, 289; pauci- gigantea, 104, 109, 110, 116, 125, 128, aquatica, 104, var. aqu + 104 a trinitensis, 285; racemulosus 249; re- gigantea, 104, var. major, 104, f. aquat- Р - нн reticulatus 5; retusa, іга 1 "mayüen" : 289; rbombifo - 289; i Ouf. " E 7. latifolius, 289, 8. cocblearifolius, 289 о 104, ID т мар 4 іс 128, var. minor, 104, f. A 103, x ruja, жек be 253: salicifolia, stenocarpa, 104 , Var. typica “Blue Adlay,” з PE OR Re жч nent te 105, 108, 115, 124, 126-128; lingulata, 256; spinifolius, 290; spinosus, 290 stephanotiifolius, 261; striatus, 266, 290; strigillosus, 261; stylosa, 254, 265, var. T LJ " Collemopsidium pyrenuloides, 133 subspicatus, 236; fafarinowii, 260; tetra. Colubrina ^u oo d merus, 290; tristis, 290; зери, 290; Соух tubulosa, 104 vanioti, 256; ы үреді 290; venulosus, Cuvier’s account of St. Hu vente 154 49; versi 61; ем, 290; Cyathorhachis ~ 1 verticillatus, 290; vitiensis, 291; vulcani- Cyphella aeruginascens, 13 colus, 282; wallichienus, 291; wallichi, Cytology of Coix and its pee 115 19551 INDEX D "Dal Roti" in India, 5 Dale, Ernest E., Nort 17. Тікенек апд: Tassel modi знав iig in Zea Mays, 1 Deuterolichenes, 36 Diosma serra › 231 Doryo лы Е new species of, from Suri- nam, 213 Doryopteris, 213; palling, 214; conformis, 213, 214; lomariacea, 213; pedata var. multipartita, 214; sagittifolia, 214; vari- ans, 214 Dreuzy’ 5 рр to St. Hilaire's "Voyage à Rio G Lp e botanical M of Auguste de St; Hilaire, 15 E a Mays, morphology of, Ecological factors in relation to РЕА 8 региз, а кагуны вы Eleonurus candidus var tuus 167 Embelia ану з Ephelidiu g ae 136 Ephelis, "s ae 136 Erythvospermum й Беса 257 Itu neifolius, 286; à eupblebipbyllus Bi thunbergianus 285 Euonymoides, 227; гей, Е Ferns, See Pellaea and Selaginella Festival dishes of North India, 346 ~ ueggia -n 249 ood: Coix as, 105; of North India, 303; ms of Brahmans, 307, 379, of da castes, 381, s of a Hiadu Village in North India, 303 Forage plants, Зенас. псу а of three species of е Dic Robert E.: Three new Annonaceae rom Panama, 15 Fruits eaten in Indi 316, 341 а Gasparrinia Harrissoni, 146; macquariensis, ена басо of Coix, relation- р chromosome number and, "Ghi" as food in North India, 568 Glade in reda "— Garden Arbo- etum, 190; es list M fre- quencies yon under pei eran 415 Gouania wil eno 290 Gramineae, 103 Gray Summit, Mo., history of land usage at, 17 Greens used as food in India, 341 св feddei, end henryi, t "n им» утп ‚ 289; skr 286; diversifolia, 286; rn Ле 286; heyne- 8 289; ИУ 290; royleana, 290; rufa 290; trigy oro ; variabilis, 290; with ensis, 246. уч“ 1; ҰЙҚЫМ) 286, 289, 291 H ure рок Trufant: Comparison of Juni ations on an Oza и ags Фа Old Fields, 171, Appendix, pes Island: Additions to - піни flora ‚ 131; views — Has eppia antarctica, Heredity, Zea M as a tool in study of, 5 — village in North India, foods of a, How ien A revision of the genus Celas- нудним in Celastrus, 225 I Цех crenata, 285; cuneifolia, 287; macro- carpa, 290; purpurea, 285; scandens, 286; Vbi 290 i x fro om, 108, 109; Foods of a Hindu vi ве i in, 303; food plants of, eid аки of Zea May ys: staminate, Tassel; pistillate, see 5227 hy Ce се 219; diagrams сна ation in Juniperus, 171, 186 Iodina rhombifolia, 289 J Job’ s Tears, see Coix Juniper populations on an Ozark glade and old. fields, Comparison of, 17 i ‚171, variation in, ginian гасе, 171, 18 1, 186, var. typica, 171; vir sheis 186 Juniperus virginiana d J. Ashei, variation patterns in, 181, 1A. 184, 185 416 K Karimpur, India, foods of, Kramer, Karl U., and Rolla M. Tryon, Jr.: A new species of Doryopteris from Suri- nam, Kurrima sp., 289 Кони crozetica, 135 L Leaves of Juniperus, variations in, 181, 182, иса у. учи earns f: 134; их $445; p iseomarginata, 145; perrugosa, 140; versicolor, 137 134; Auberti > Lecanora chrysoleu 134; ab a antarctica, oe еш Е t Deightoni, 136 Lycopodium bryoides, 4 0, 67; carinatum, 82; ciliatum, 77; Dregei, 57; laxum, 82; rupestre, 64; мај. 81, 82 M Macquarie па additions to the Lichen flora of, Mac еа Land, additions to the Йога of, 14 Maize: in North India, 329; Studies on жин relatives of, 105; Tassel modifica- tion з 133 Manatee: in North India, 316 Maytenus bilocularis, 285; boaria, 285, 288, 290; buxifolius, Ge dress 289; conterrus; 286; cunninghamii, 286; dii rmus, 286; ‘dptendtaides 289; sis, 284; orbiculatis, 286, 288; 289; е ovata, 289; quadrangulatus 289 cemosu. 79; retusa, 286, 289; ibn cimi) 287; verticillatus, 290; williamsii, 282 Meiosis i in Е and related genera, 118 g rcumsorediata, 142 Миљан ый lena, 13 Microthelia Vise мираза 137 Microtropis bivalvis, 285 Bo — a A Arboretum, 172: ill, 175; The oe requ Old Field, 178 [Vor. 42 ANNALS OF THE MISSOURI BOTANICAL GARDEN —€— = Srg of Zea Mays, 195 Moya spin Mutants of eie й Zea Mays: cle of, 199; key to, 196; measurements of н d features, 197; spi iter in s, pl. 28; Club, Tassel-seed 7 and тај seed 8, pl. 26; Vestigial glume, pl. Myrsine жөніме, =. 290 М Neuropogon aeromelanus У. inactivus scabridulus, 143; insularis, 135; trachy- carpus, 135 Nickerson, Norton H., and Ernest Ay Tassel medication i in Zea Mays, Nirodi, Na = s tudies on Asiatic ‘awe of maize, Nostoc, sp., Nutrition practices in a Hindu village of North India, 356 о Old fields in тні Botanical Garden Arboretum, 17 , 190, 101; distribution of Ж-Е беліне іп, 180; variations о rs іп, 180, 181; продо ЈЕ with fre requencies for understory o Orixa japonica, 286, Ozark glade and e li ‘Comparison of Juniper populations on, Р PARARE dichroa, 134; роману 132 — Johnstoni ; pubescens v. con- 145; sblagubris 142; tenuirima, 142; dido. З Pellaes, А new, Dis на Africa, 101 — РеПаеа andromedaefolia, 102; intermedia, 54% ovata, 102; myrtillifolia, 101; rufa, » IO. Piin Lairdi, 158 People of dig tyloplaca, Phyllanthus sp., 290 Pittosporum je conet 288, 290 Placodium bicolor, 1 Placopsis bicolor, E еке 140 Pleurostylia согыс айк, 289, 291 estes 103, 108; barbata, 113; ' bracteata, cyathopoda, ; itata, 106; 1 зеза 106; macrophylla, 106, 107, 19551 INDEX 110,111,113, pe mre morphology of, 117, 126, I2 30; Massii, 106; semi- teres, 106; ME 106 Populations, Төзе, comparisons of, o lade and old fields, 171; уза репа! 92 Prema parvilimba, 2 Preservation of di ad AP. India, 323 E rot Hoa peus citrina, myrtifolia EU Оши ка о я glabra, 138 Psoroma versicolor, 1 Lern) use of, as food i in North India, 333, R Ramalina geniculata, 142 288; napalensis, 290; ман Rhizocarpo n Johnstoni, 134; kerguelense, es awsoni, 134 Medis aspicilina, 136; uec al es 146 Rudolph, Emanuel D., Carroll W. Dodge t Antarctic к. Е жуз и LIV, 5 Sabia рын var. nitidissima, 287; yun- nanens St. Hilaire, August de, The botanical cata- logues of, 153: numbers of c lis ted Жы “у — 159; ; page of Cit ta- logue C?, 166; page о list of collec- Pm numbers, 170; i попис то шы. 163; Жедде key to, 161, 169 Saurauja vaniota, 256 Schoepfia flexuosa, 287 Sclerachne, 103; cyathopoda, 106; punc- 11 mmersonii, 291; paniculata, 230 Selaginella rupestriy de its Selaginella, 16; acanthonota, 26; amazonica, ; Aracbavaletae, 34; arenaria, 26; arenicola, 23 acanthonot marii, 20, var. arwin. ур 67; “Соғу, 28; deflexa, 16; densa, 417 66, var. densa, 66, 07, 08, var. scopu- lorum, 97, var. балдыр, 71, 07; dentic- sea, „76 ; beste 11; Hin- ton Landii, 81, 99; lepidophylla, 8 bryoides, 74, 98; longipila, 52, 2; тасга- % У sis, limitanea, 46, var. m texana, 46; muticola X rupincola, 21, neomexicana, 21; nivea, 55, 96; njamn- jamensis, 52, ; oregana, 61, 97; Parishii, 80, 99; percei, 34; peruvi- ombeyan iana, “i porrecta, 37; Preis siena, 17; proxima, 56, 96; pumi mila, 17; pygmaea, 17; Quartiniana, 5 53; Riddellis, 24; Rosen- amazonica, ^ Dregei; 57, . Hay N © © N ~ f. j rigbtii, 48, ssp. Balansae, 27, var. acantbonota, 26, var. Balansae, 27, var. Bolanderi, 49, var. borealis, 81, var. brasiliensis, 34, var. brevipila, $1, var. 3 densa, 70, V і senii, 49, rva f. abyssinica, 53, f. angolensis, 53, f. capensis, 53, va longipila, 81, subvar aucina, 81, var. viridis, 81, va arp £5 77, г. montanensis, 45, У iana, 77, recurva, 57, Ї Divina, a tr Welwitscbiana, 57, var. rupi 20, var. Sartorii, 37, var. shakotenensis, 43, var. 5; 2 var. tropica, 82, var. W allacei, 4 cola, 20, 4; gonensis, Schmidtin ч 1, уаг. ‚ Krauseorum, 71; scop- uloru 7; selagin Sel 49, 95; struthioloides, 82; tortipila, 29, 04; uliginosa, 17; Underwoodii, 62, 97, 418 уаг. ота = emerge TA: 96; Vardei, 7; viridisima, 28, 94; сої Watsonii, 7 kieg 28, 94; Wightii, 39, var. Phil- lipsiana, 40, var. vetusta, 39, var. Wightii, 39; Vrightii, 48, 95; Wrightii, 48 Semarilla Sem темн 2 42 Simmondsia californica, 288 Siphulastrum cladinoides, 153 South Africa, A new Pellaea from, 101 Spices ies ooking food in North India Spo dum | 136 Stachygynandrum rupestre, 9 Steinera glaucella, 132; Wadi 132 Зено зав argodes, 140; Argus, 140; cymosu 134; pulvinere; 140; sub- о. 140 Stricta glabra, 138 Surinam, A new species of Doryopteris from, 213 + Tassel modifications in Zea Mays, 195 о ~ 1рз ы etween, ап Taxonomic monographs, role of St. Hilaire’s catalogues in, 163 Тыс мі macquariensis, 141 ие heardense, 131; praevalescens, 13 Tichothecium matoes, use ты in North India, 309 Trichilia јода, 287 Trilobachne, 1 Trinidad, Coix “Blue wd from, 108 EDO RUM menus З Tryon, : А new E from South Africa, 1 Tryon, Rol ‚ Jr.: Selaginella rupestris and its M E Karl U. Kramer and: A new species of Doryopteria from Surinam, 213 [Vor. 42, 1955] ANNALS OF THE MISSOURI BOTANICAL GARDEN U Umbilicaria antarctica v. subvirginis, m Hunteri, spongi pu v. subvirg 144; subcerebriformis, 1 United Provinces of India, pet of, 3 Unonopsis costaricensis, 152; кейс 1 A о сажу 52; арте 1 Pitti ; Schippii, 1 [кзы а үн гей 135 Ureceolina kergueliensis, 134 Urostacbys carinatus, 82 Usnea antarctica, 143; arida v. musc cicola, 143; Taylori, 135; torulo: е 142; trachy- carpa, 135; xanthopoga, V Variation in Juniperus at the Garden Arbo- retum, Vejetábles grown in North India, 315 Verrucaria kerguelena, 132; vedi pae 131 * wW W eddell's key to catalogues of St. Hilaire, 161; portion of page from, 109 Wiser, Charlotte Viall: Foods of a Hindu illage in vct 2 "i 303 Wimmeria mexican beri grown in hd. 330; bread from, X Xanthopyrenia heardensis, 2 Хап опа Mawsoni, 143, 146 2 Zea Mays: Tassel modifications in 195; tassel ај Ру illustrations of, Qa 29, key to MISSOURI BOTANICAL GARDEN | о ROBERT . .. Senior E. Ұоорѕом, Ja, | |