DE MUN Annals of the Missouri Botanical Garden ` Volume XXVIII 1941 With Twenty-nine Plates and Seventy Figures Published quarterly at Fulton, Missouri, by the Board of Trustees of the Missouri Botanical Garden, St. Louis, Mo Entered as second- сыын matter at the post- aps at Fulton, Missouri, nder the Act of March 3, 19. 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 MissouRI BOTANICAL GARDEN appears four times dur- ing the calendar year: February, April, September, and November. Four numbers constitute a volume. Subscription Price........ $6.00 per volume Single Numbers.......... 1.50 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. * Except No. 3 of Vol. 22, price $5.00, and No. 2 of Vol. 25, price $2.50. ANNALS OF THE MISSOURI BOTANICAL GARDEN TABLE OF CONTENTS Some Effects of Carcinogens on Yeasts.... . .. Carroll W. Dodge, Bertha S. Dodge and George Thomas Johnson A Monograph of the Genus Parmelia in North Ameriea, North of Mexico........ ОРОМУ ОО ОГ Edward Cain Berry Binary Variation in Tradescantia bracteata DO DURO Uh SR Edgar Anderson On the Flora of the Frontier Formation of Southwestern Wyoming enry N. Andrews and Cortland S. Pearsall The North American Asclepiadaceae. I. Perspective of the Genera * 9 9 9 9 c9 * c9 39 9 9 > € 9 e 9$ € 5 6 о в e» 6.6.6 $ & оге «v s» t ee V Two New Asclepiads from the Southwestern United States .. Bassett Maguire and Robert E. Woodson, Jr. A pm Survey of the Genus Tripsa- СИВ ар скрия Hugh C. Cutler and Edgar Anderson Miscellaneous New Asclepiadaceae and Apo- eynaceae from tropical America Robert E. Woodson, Jr. o o o 90662620626 ооо йо » 9». а ә € 9 € ^ * Mass Collections— The Technique and Use of Mass СоПес- tions in Plant Taxonomy..... ‘хат Anderson Camassia scilloides........... Ralph O. Erickson Rubus odoratus and В. parviflorus....... MM LEE EU DN Norman C. Fassett Dichophyllum Moorei and Certain Ass тале беби. оз Henry N. p tied: A Revision of the North American Species of the Genus Anisacanthus......Stanley H. Hagen Contributions toward a Flora of Panama. V. Collections chiefly by Paul H. Allen and by Robert E. Woodson, Jr. and Robert Be CBOE fe Wee er Pon rete .. Robert E. Woodson, Jr. and Robert W. Schery The Cytogenetics of Poa pratensis......... illiam L. Brown ө coo $4 $9 9 4 + 9 9 $ фе 9» о э 9943199 95 PAGE 1- 30 31-146 147—165 165-192 193-244 245—248 249—269 271-286 281—292 203-298 299—314 879—384 385—408 409—490 495-522 528—532 STAFF OF THE MISSOURI BOTANICAL GARDEN Director GEORGE T. MOORE HERMANN VON SCHRENK, EDGAR nke Pathologist Genetic JESSE M. GREENMAN, ROBERT E. WOODSON, JR. Curator of the Herbarium Assistant did of the Herbari CARROLL W. DODGE, NELL C. HORNER, Mycologist Librarian and Editor of Publications BOARD OF TRUSTEES OF THE MISSOURI BOTANICAL GARDEN President GEORGE C. HITCHCOCK Vice-President DANIEL K. CATLIN Second Vice-President EUGENE PETTUS L. RAY CARTER Й GEORGE Т. Moore DUDLEY FRENCH A. WESSEL SHAPLEIGH JOHN S. LEHMANN ETHAN A. H. SHEPLEY EX-OFFICIO MEMBERS GEORGE В. THRO WILLIAM DEE BECKER, сорорени em NES ae Mayor of the City of U sity St. Louis WILLIAM SCARLETT, A. H. TIMMERMAN, a A of the Diocese of President of the ый of uri Science of St. Lou JOHN A. FLEISCHLI, President of the Board of Education of St. Louis GERALD ULRICI, Secretary Annals of the Missouri Botanical Garden Vol. 28 FEBRUARY, 1941 No. 1 SOME EFFECTS OF CARCINOGENS ON YEASTS CARROLL WILLIAM DODGE Mycologist to the Missouri Botanical Garden Professor in the Henry Shaw School of Botany of Washington University BERTHA SANFORD DODGE Assistant in the Henry Shaw School of Botany of Washington Uniwersity AND GEORGE THOMAS JOHNSON John Simon Guggenheim Memorial Foundation Fellow to Central America Formerly Lecturer in the Henry Shaw School of Botany of Washington University With increasing knowledge of the chemistry of cholanthrene and related compounds and their action on mammalian cells, it has seemed desirable to study their effects on organisms with less complex structures and interrelations of parts. Yeasts were thought to be a group of plants of relatively simple or- ganization and to lend themselves to comparatively simple techniques. A preliminary survey of a half dozen genera of increasing complexity of polarity and morphology showed much less effect on morphology than anticipated, therefore subsequent study was confined to two strains of Saccharomyces ellipsoideus Reess emend. Hansen, one a champagne strain, the other a Tokay strain. No differences in the physiology of these strains have been detected in our experiments. We take this opportunity to acknowledge the generous financial assistance of the International Foundation for Cancer Research which has made this study possible; the careful work of Miss Helen Bramsch and of Stanley Hagen, who prepared sections of colonies for our studies of morphology in the early ANN. Mo. Вот. GARD., Vor. 28, 1941 (1) [Vor. 28 2 ANNALS OF THE MISSOURI BOTANICAL GARDEN stages of this work while research assistants; of Edward Cryder, who made most of the haemacytometer counts here reported and who prepared the graphs; and of Dr. Verne F. Goerger, who assisted in mierophotography in our preliminary attempts to secure photographie data in connection with pop- ulation studies. Mrs. Dodge has given much time during the four years of this investigation, while Dr. Johnson has been associated with us during the academie year 1939-40 only. We also wish to express our gratitude to Professors Philip Franklin and Norbert Wiener, of the Massachusetts Institute of Technology, for examining our graphs of population studies and for their assistance in interpreting them ; and to Dr. Irving Langmuir and Dr. Katharine B. Blodgett, of the General Elec- trie Company, for suggestions on our work with surface films. MonPHOLOGY An old stock culture of Saccharomyces ellipsoideus, which had been kept in the ice-box for nearly six months, produced on methyl cholanthrene agar (see p. 4) very curious colonies with a much more complex morphology than in the controls, or anything reported by Píšová ('34). These have been re- ported in a previous paper (Dodge and Dodge, '37). Repeated attempts to reproduce these curious colonies with their dif- ferentiated cortical and giant cells have so far failed. Some differences in morphology exist between the cultures with a earcinogen, especially benzpyrene, and the controls in Saccharomyces ellipsoideus Reess emend. Hansen, Zymonema capsulatus Dodge, Z. dermatitidis (Gilchrist & Stokes) Dodge, Mycocandida onychophila (Poll. & Nann.) Lang. & Tal., and Castellania tropicalis (Cast.) Dodge (a strain originally from Castellani's laboratory but perhaps not the original strain), but these differences are much less striking and difficult to describe. That the age of the colony from which the transfer is made and its previous history have some effects on its growth is well known, but so far we have been unable to duplicate our first observations on S. ellipsoideus and have turned our atten- tion to fermentation and growth studies. 1941] DODGE, DODGE & JOHNSON—CARCINOGENS AND YEASTS 3 GROWTH Growth as applied to unicellular microorganisms is com- monly restricted to multiplication or to increase in individ- uals, rather than to mere increase in size or differentiation and changes in tissues. Consequently, in any closed system such as a culture, it is essentially a measure of the increase in the number of cells between two instants of observation less the number that have died and disintegrated during the inter- vening time. In the absence of copulation and sporulation, multiplication is by binary fission, so that theoretically the number of cells should increase by powers of two, an ideal practically never reached in experimental conditions. Before discussing our results, it seems desirable to consider the methods which have been proposed and the sources of error and the interpretation of each method. METHODS AND SOURCES or ERROR Media.—In aiming to keep as many environmental and nu- tritional factors as constant as possible, the ideal medium would be a synthetic one where the concentration and inter- relations of each nutrient were known. From time to time such media have been proposed for yeasts, but after trying various formulae, using both inorganie and organie (amino acid) sources of nitrogen, growth was so slow that we reluctantly returned to peptone in spite of its variable composition and the presence of possible growth-promoting substances. We have tried, however, to keep our sources of error as constant as possible by using only two batches of Difco Peptone in all our experiments and we have failed to find any significant dif- ference between the two batches. Our liquid medium has consisted of 1 per cent Difeo Peptone and 2 per cent glucose. To this 2 per cent agar was added for solid media. Saturated aqueous solutions of the carcinogens! 1 The carcinogen crystals were suspended in large flasks of hot distilled water and allowed to stand for several days with occasional shaking and the remaining erystals filtered off through fine filter-paper; the resulting filtrate was considered as a saturated aqueous solution. , [Vor. 28 1 ANNALS OF THE MISSOURI BOTANICAL GARDEN were used to dissolve the other ingredients of the medium and distilled water for the controls. The medium has been auto- claved at 15 lb. pressure for 15 minutes. The 1,2,5,6 di-benz- anthracene, benz-pyrene, and methyl cholanthrene? have been used in the course of this study, using methyl cholanthrene for most of the experiments. The solubility of carcinogens is very slight, and so far we have been unable to secure a satis- factory method for determining it. We experimented with several monomolecular film methods developed or suggested by Langmuir and others without satisfactory results. The apparent monomoleeular film resulting from spreading the methyl cholanthrene dissolved in benzene was evidently not a true monomolecular film, and Dr. D. F. Waugh, of the De- partment of Zoology of Washington University, after a further study of our material, reports as follows: Attempts to spread methyl cholanthrene, dissolved in benzene, resulted in the formation of a large number of extremely thin platelets which formed at the periphery of the benzene drop. The continuous formation of such platelets caused the water surface, for a considerable distance around the benzene drop, to be covered by a film of methyl cholanthrene which was apparently too thin to show interference colors but which, however, was a number of molecules thick. The effect was similar to that obtained when paraffin, dissolved in benzene, is allowed to spread. No film measurements were attempted. Tn the solutions used, these carcinogens appeared to have a 2 The 1,2,5,6 di-benz-anthracene was secured from the Eastman Kodak Com- pany, the benz-pyrene from F. Hoffmann, Laroche & Co.A.G., Bale, and the methyl — from the International Foundation for Cancer Research. On check- melting point of our methyl eholanthrene after it stood three years on a хз чен shelf, it was found to melt at 168—174? C. instead of 176-177? C. It was reerystallized by dissolving in hot benzene, eooling and preeipitating with cold ether, when it melted at 176—177? C. It was also dissolved in benzene, pae acid dissolved in у» added, heated with bone black, cooled, ET ed, and dis solved in re benze Sodium earbonate was added, the flask shaken, filtered, the Жыр. ых ni reerystallized as above. 'This also melted p 176-177? C. A sample of methyl eholanthrene was also obtained from the Eastman Kodak Com- pany but as it melted at 171—174? C. it was not used in our experiments. A care- ful inspection of our data, obtained shortly before and after our purification, re- vealed no signifieant differences. Alsopp (740) has reported alteration of benz- pyrene and 1,2,5,6 di-benz-anthracene when irradiated with ultra violet light. 1941] DODGE, DODGE & JOHNSON— CARCINOGENS AND YEASTS 5 marked effect on surface tension. When Ше media were being made up in volumetric flasks and shaken well, the bubbles in the carcinogen media rose rapidly and broke almost instan- taneously, while those in the control media persisted as froth for along time. We have not yet had time to make quantitative measurements of surface tension. Measurement.—Five methods of measuring growth of microorganisms have been frequently used: diameter of col- ony, volume, weight (wet or dry), nephelometry, and counting. These will be discussed in this order. Diameter of Colony.—When an organism grows radially with equal rates in all directions, as is the case with many fungi, this method is rapid and satisfactory. It yielded some satisfactory results with Zymonema but was unsatisfactory for the moister colonies, as they were not of uniform height and contour, and was abandoned after preliminary studies. Volume.—The culture is centrifuged in graduated tubes and the volume of the closely packed cells is recorded. Variations in speed and time of centrifuging, as well as shape of the cells, are possible sources of error. This method has the inherent disadvantage that the cells may be injured in the process, so it does not lend itself readily to following a population over an extended period of time. It also requires a considerable period of time or a relatively large volume of culture media to secure sufficient growth. Since it did not seem to have ad- vantages over dry weight for our purposes, it was not tried. W eight.—Wet weight, where the organisms are filtered and weighed immediately, has given reasonably consistent results to some investigators (Satava, '18) but seemed to have little except speed to recommend it as compared with dry weight. In some of our preliminary work we filtered by suction, using filter-paper in Gooch erucibles before turning to alundum crucibles. The coarser crucibles (R.A. 98 and R.A. 360) were suitable for filamentous forms, but R.A. 84, with an average pore size considerably smaller than the diameter of our yeast cells, were used in our work with Saccharomyces ellipsoideus. [Vor. 28 6 ANNALS OF THE MISSOURI BOTANICAL GARDEN We used Erlenmeyer flasks of 125-ml. capacity, containing 50 ml. of medium inoculated with 3 drops of a uniform suspen- sion from a 5-ml. pipette, and incubated at 25° C. At the end of ineubation, the contents were transferred to 50-ml. centri- fuge tubes, using the usual precautions of chemical gravimetric technique. After centrifuging long enough to secure a clear supernatant liquid, most of the latter was decanted. Then the yeast, resuspended in the remaining liquid, was poured into the erucible and the tube thoroughly rinsed twice with distilled water. Filtration was by moderate suction (too great suction will crack the crucibles). The crucible was then placed in a drying oven at 70° C. Constant weight was usually reached in 8 hours, although 24 hours was allowed between weighings before the weight was accepted as constant. No attempt was made to read the balance closer than milligrams as it was felt that the next figure would be insignificant. After the final weighing, the crucibles were heated to redness to remove the yeast cells and were stored in a desiccator until we were ready to weigh and use them again. This method gave very consistent results, showing very little error in the technique. This method is a measure of the total cells, living and dead, but does not include those disintegrated during the time interval. Its chief disadvantage lies in the relatively long time that must elapse between inoculation and the first data as it takes about two days to produce sufficient growth to weigh with much accuracy. As we shall see later, each culture behaves as a separate individual after inoculation, so that the three values of a triplicate set were seldom identical although the variation was not great. If we try to take samples from a larger volume of culture, sampling error is introduced as it is difficult to se- eure a uniform suspension for sampling and by the very process of sampling the equilibrium of the culture is upset. Also the dry weight varies somewhat with the depth and the area of the surface of medium exposed to the air and hence with the size of the flask and the volume of the medium. Conse- quently, dry-weight methods are ill adapted to a study of growth rates. A further source of error was introduced in our preliminary experiments as all but Saccharomyces ellip- 1941] DODGE, DODGE & JOHNSON——CARCINOGENS AND YEASTS т soideus of the species studied were pathogenie for man; con- sequently the cultures were sterilized before filtration to avoid accidental infection of the laboratory workers. Since consider- able exosmosis is known to follow death, it is probable that exosmosis occurred during the autoclaving and subsequent cooling, probably roughly proportional to the time, a factor hard to control or evaluate accurately; hence our values of dry weight in the preliminary experiments are probably too low. This source of error was not present in our work with S. ellipsoideus. Nephelometry.—This popular method was not used, as it was thought that the great variation in the size of cells and the presence of oil droplets and refractive granules would tend to render it unsuitable for our work. Subsequent studies by Loofbourow and Dwyer (738) and Richards (740) seem to sup- port this conclusion. Census.—Counting of microorganisms has long been the chief source of information of growth rates as changes in pop- ulation may be observed at frequent intervals. This may be accomplished either by sampling a large culture at intervals or by direct observation of a small closed system under a microscope. In the former, the sample may be dispersed as an inoculum into fresh medium and the resulting colonies counted, a method commonly used with very small organisms such as bacteria which are not easily observed directly. It gives a measure of the total living cells in the volume of sample and, by implication, in the whole culture if sampling and dilu- tion technique is adequate. It does not measure the dead and disintegrated cells which may have formed during the given time interval. Steinhaus and Birkeland (739) have empha- sized the importance of this disintegration in growth studies on Serratia marcescens Bizio. As an alternative, the sample may be placed in a haema- cytometer and counted directly over different areas, the counts being averaged and computed for the whole sample and, by implication, for the whole culture. This will give the viable and dead cells but not the disintegrated. Differential counts may [Vor. 28 8 ANNALS OF THE MISSOURI BOTANICAL GARDEN be obtained by staining the dead cells. We used this method while studying the possible cell types of our cultures and at- tempted to obtain differential counts. In general, the curves roughly parallel the counts in colonies under continuous ob- servation, deseribed in the next paragraph, in spite of the pre- sumably greater oxygen tension of the medium. Both these techniques have inherent sampling errors and changes in equi- librium (much less than in dry-weight determination since the sample subtraets a much smaller volume from the culture). Finally, we may use a haemacytometer as a micro-culture dish and keep the closed system under theoretically constant observation, counting the cells lying in the volume of medium above the ruled squares. In our work we used standard haema- cytometers, taking a sufficient amount of culture medium to fill the counting cell without running into the grooves which were filled with vaseline. A suspension of yeast cells was made and diluted so that the volume used in the culture chamber would contain approximately 1 cell per small square when the cover-slip was lowered into place. No gas bubbles were produced under our experimental eonditions. Care was taken to exclude all air bubbles as their expansion and contraction with relatively slight variations in temperature and barometric pressure might cause convection currents. A count was made immediately, and if the cells were found to be unevenly dis- tributed or too many per small square, the culture was washed up and a new one started. Counts were made at hourly intervals during the working day. It is evident that the observer will need time to sleep and rest his eyes so that a continuous record could not be se- cured with the observers available. The carcinogens were counted from 8:30 a. m. to 4:30 p. m., and the controls from 9 a. m. to 5 p. m. each day. When there were indications of a sudden change in the slope of the curve toward the end of the day we returned to the laboratory in the evening for ad- ditional counts. On Sundays only two or three counts each were made, unless the counts showed considerable change from the previous ones. In the graphs based on these observa- tions, the daytime periods appear as sawteeth of varying 1941] DODGE, DODGE & JOHNSON—CARCINOGENS AND YEASTS 9 magnitude connected by straight lines. In this manner miero- cultures were observed more or less continuously up to 46 days. We probably missed many minor peaks and perhaps a few major ones, but our data were the best obtainable with the observers available. Several difficulties seem to be inherent in this method. Oxy- gen tension must be lower than in flask cultures, in fact ap- proaching anaerobic conditions. The volume of culture me- dium available per cell is probably less, although we have no data on this point. The error of counting is negligible until about 60 yeast cells per small square is reached when the cells are closely packed, and it inereases as larger counts are ob- tained, tending to smooth the peaks of the curve more than the depressions. Another source of error lies in the movement of single cells, or rarely pairs of cells, over the line from one square to the next as the rows of small squares counted were not contiguous. While the yeasts are supposed to be non- motile, several times we observed slow movements of trans- lation similar to those common in some of the Myxophyceae, and a study of photographie records and circumstantial evi- dence point to other examples. The magnitude of the error is probably not great, as in general the movements would tend to cancel each other. It has been suggested that such move- ments are due to convection currents, but this seems unlikely since in all observed cases cells very close to the moving cell showed no movement. Such movements were noted on slides which had been resting on the stage of the microscope for 12-18 hours, as well as on those resting on the table beside the microscope. No attempt was made to use a constant- temperature stage, but in every ease the slides lay side by side next to the mieroscope, covered by a small box cover to exclude dust and light when not under aetual observation. "Therefore, the temperature difference between the two halves of the same slide and between the two methyl cholanthrene cultures on one slide and the controls on the other were very slight, although there was some diurnal variation in the laboratory as well as some seasonal variation after the building was no longer heated, since these experiments covered a period from No- [Vor. 28 10 ANNALS OF THE MISSOURI BOTANICAL GARDEN vember 1938 to the middle of June 1939. These variations probably affected the total counts and consequently the growth rates slightly but less than other factors inherent in the ex- perimental conditions. The sampling error of the counts is probably low. It was not studied during the direct observation counts, but the photo- mierographs showed it to be less than other sources of error in that method (see p. 19), since curves obtained by plotting counts from different rows of four small squares each are parallel for the same mieroculture. On the other hand, in a preliminary experiment where a small amount of air was in- cluded between the vaseline seal and the medium, due to a faulty seal, growth was more rapid (probably owing to greater oxygen tension) and some cells moved into the field, vitiating the experiments. In our graphs, the points were plotted from the actual counts, and a faetor of 50,000 should be used to express cells per eubie millimeter. An automatic apparatus to produce photomicrographs at regular intervals suggests itself for reducing some of the errors of this method and giving a more continuous record than we have been able to secure by a limited number of ob- servers. Some attempts were made to construct such an ap- paratus but it presented financial and technical difficulties in the conditions under which we worked. One inherent difficulty of this method is that many cells observed easily by focusing from the upper to the lower surfaces of the layer of liquid be- tween slide and cover slip are missed by the focal plane of the camera, and the images of cells just below and just above the focal plane are so blurred that it is not easy to decide whether buds are attached to the cells or not. Therefore, one is forced to count only the cells in sharp focus. This increases the movement error much more than in direct observation, as well as the inherent difficulty of keeping a camera in the same focus over a period of several weeks while manipulating it to change spools of film, ete. It is probable that, with a specially constructed haemacytometer having a much thinner layer of 1941] DODGE, DODGE & JOHNSON—CARCINOGENS AND YEASTS 11 medium between slide and cover slip, a very bright light source of the flashing type, and a specially constructed camera, such a method might be devised but it was beyond our means. FERMENTATION Quantitative studies of fermentation fall into two main eate- gories on the basis of determination of carbon dioxide evolved. The gas may be measured volumetrically in a gas burette, applying the usual corrections for temperature and pressure, and the weight computed. The 50 ml. of medium was pipetted into a 125-ml. Erlenmeyer flask, sterilized, and inoculated with three drops of a suspension of Saccharomyces ellipsoideus. The flask was attached to the gas burette by means of previ- ously sterilized rubber stoppers and glass tubing. The gas was collected over acidulated water and read at regular intervals, recording temperature and barometric pressure from a mer- eury barometer hanging near the laboratory desk. After each reading the gas was allowed to escape and the water column returned to zero by means of a leveling bulb. Periods of a rapidly rising barometer complicated the process since they tended to drive the acidulated water back into the culture flask, causing the discontinuance of some experiments earlier than planned. This method has an advantage in allowing more fre- quent readings than gravimetrie methods, with the same de- gree of accuracy. In our later work we used gravimetric determination by col- lecting the gas in potash bulbs or soda-lime tubes and weigh- ing. After eliminating several sources of error in preliminary experiments, we set up trains as follows: soda-lime tube, wash- bottle of distilled water, tube with sterile cotton, large test- tube with 50 ml. of medium, tube with sterile cotton, anhydrous ealeium chloride tube, potash bulb or soda-lime tube which was weighed, closed filter flask (to prevent back seepage of water or earbon dioxide to the rest of the system), and water pump. Each time before weighing, the pump was started, and moist, carbon-dioxide-free air was drawn slowly through the system for ten minutes. The soda-lime tubes were weighed quickly [Vor. 28 12 ANNALS OF THE MISSOURI BOTANICAL GARDEN on a chainomatic balance. The tubes containing the medium were sterilized, inoculated, and rapidly attached to the system by previously sterilized rubber stoppers and glass tubing. No contaminations were detected. By the above system the air passing over the medium was moist, free of carbon dioxide and sterile, and the gas reaching the soda-lime tube to be weighed was dry, hence we were not weighing water vapor as well as carbon dioxide. A possible source of error is too rapid aeration, which might remove carbon dioxide from the soda- lime tube before it was completely absorbed. The final, practi- eally constant, weights obtained in our longer experiments seem to support the accuracy of this method. Potash bulbs with strong solutions of sodium hydroxide were early aban- doned, as water vapor is lost from the solution during the passage of air through it and in some of our early experiments the potash bulbs actually lost weight. While this method does not lend itself as readily to frequent readings, it is much simpler and involves less computation. Besides measurement of carbon dioxide evolved, an attempt was made to study the amount of sugar disappearing during the growth of the culture. In our early experiments we de- termined sugar by the Benedict and Folin-Wu methods. Later we returned to the Bertrand-Shaffer method in which the Fehling’s solution is reduced under standard conditions and the precipitate dissolved and titrated with standard potassium permanganate. The end-point of the titration is sharper by this method and gives more satisfactory results. Since other substances beside glucose reduce Fehling’s solution, the inter- pretation of results is sometimes puzzling. In general, the reducing power of the medium disappeared sooner than the cessation of carbon dioxide evolution, although in the final computations the sum of the carbon dioxide, the remaining reducing substance (computed as glucose), and the dry weight (assuming it was all carbohydrate which it certainly was not) accounted for only half to three fourths of the original glucose (determined from the uninoculated tube). The reactions in- volved need further study before the data are satisfactory. 1941] DODGE, DODGE & JOHNSON—CARCINOGENS AND YEASTS 13 EXPERIMENTAL DATA Colony diameter.—This proved a fair measure of growth for filamentous forms but less so for forms with a moist creamy colony. The methyl cholanthrene seems to have depressed DIAMETER OF COLONIES IN MILLIMETERS Cone. | Z. dermatitidis C. tropicalis M. onychophila S. ellipsoideus 10 days 20 days % 10 days} 20 days | 10 days | 20 days 0 10 25 | 27 | 28 5 7|8.88]71|2]|6 6161 6.5 |5|9 2 4 22 | 95 | 20 | 22 | 81 216 2 | 5.5 4 4 8 | 21 | 10 11 | 25128 3/5 215 10 7 17 | 23 | 18 | 20 | 80 416 |6/3/6 515160 1214 20 5 23 | 25 | 11 | 13 | 25126 6167 216 100 7 23 | 25 | 12 | 14118201515 1514571 888 136 growth of Castellania tropicalis and perhaps of Zymonema dermatitidis in concentrations above 4 per cent saturation. Dry-weight determinations given below seem to bear this out although the figures are less striking. The variable thickness of the colonies of Mycocandida onychophila and Saccharo- myces ellipsoideus results in seemingly erratic figures for these species. Dry weight.—In our study of the influence of concentration of carcinogen on dry weight we secured the following results with Castellania tropicalis and Mycocandida onychophila. It should be remembered that these were our earliest experiments before our technique was perfected and that exosmosis follow- ing autoclaving may have affected the results. DRY WEIGHTS IN MILLIGRAMS C Castellania tropicalis Mycocandida onychophila one. % mg. dry weight | Aver- | Error mg. dry weight Aver- | Error age age 0 |374|359|324 349 8.3 |37 27|20| |27/20 26.2 1.9 1 254/277|231| 254 7.5 75/57/58|75|58|57 65.0 2.8 2 |299 1254134013201 303 | 11.3 |39 148 |46 55 |55 |4846 48.1 1.2 4 34| 32 5 |316/315/276/284/318) 301 6.6 |46/54|23|54|46/23 39.0 | 3.3 10 |285|285|316|321|328| 312 6.3 [38|21|27|38|27|21 28.7 | 2.1 20 |340|326|307|332|309| 323 4.5 |93|105|10/61|27|61|27|10 49.1 | 8.2 50 32913471335| 337 3.1 11/30] 8|30|11| 8 16.3 3.1 100 | 191/221/279/339/330| 272 | 19.8 |59| 69/52|14|29|52|29|14 37.2 | 5.4 [Vor. 28 14 ANNALS OF THE MISSOURI BOTANICAL GARDEN In a subsequent series of experiments with Saccharomqyces ellipsoideus we found the following results (Graph 1 left): /0r во |- UPPER LINE. V LINE ANAEROBIC 8 T as 98 + х № № m! Е во RY WEIGHT с T T RENES SRS + ме A^ чо © in Ha =, | | П 1 Жам. 1 * 1 1 ESO 20 36 т-а 60 79. 007 жи WR ak XU XT " CONCENTRATION Graph 1. Left, growth in dry weight at varying concentrations of methyl cholanthrene. Medians plotted, means tabulated. Right, growth in dry weight with three carcinogens wog co = pper line represents cultures under aerobic conditions, the lower line represents cultures used for fermentation experiments hown in Graph 3, "ite conditions Ae essentially anaerobic The above graph is based on 5 to 15 determinations for each concentration of the carcinogen (expressed in percentage saturation) with a probable error of + 0.9-2.7 mg. While the results are not striking they seem to indicate that smaller concentrations depress growth as determined by dry weight over that of the controls, while concentrations over 50 per cent saturation tend to increase it to that of saturation. As the above averages include some of our earliest experiments be- fore we had secured the alundum crucibles, they are probably less accurate than those of our later work. If we exclude these experiments the curves are substantially the same except the value for saturated methyl cholanthrene (shown by dotted line in the graph). Effect of various carcinogens.—In connection with fermen- tation experiments, dry weights were determined in quadrupli- eate for methyl eholanthrene, benz-pyrene, and 1,2,5,6 di-benz- e 1941] DODGE, DODGE & JOHNSON—- CARCINOGENS AND YEASTS 15 anthracene and controls on cultures 7 days old. No significant differenees were noted, as we obtained the following figures: control 70.8 + 0.8 mg. ; methyl cholanthrene 70.8 + 0.8 mg. ; benz- pyrene 70.8 + 0.8 mg. ; and 1,2,5,6 di-benz-anthracene 70.4 + 0.8. In the light of our subsequent experiments, it seems likely that the time (7 days) chosen for our experiment was unfortunate, since the controls are the same as the carcinogen, probably due to erossing of curves. It is interesting to note, however, that if each separate dry weight is plotted, the successive weights for methyl cholanthrene are 64.5 + 1.7, 68.0 + 0.9, 71.5 + 2.1, 14.2 + 1.8, 77.2 + 1.8 and 72.5 + 1.5; 1,2,5,6 di-benz-anthracene is slightly aberrant with values 65.7 + 0.8, 68.2 + 1.3, 68.7 + 2.2, 79.6 + 2.0, 75.5 + 2.2 and 71.7 + 1.8 (Graph 1 right). This sug- gests some sort of cycle of dry weights such as we shall also see in fermentation. In these values the probable errors are higher owing to the averages being for 4 values instead of 24. The range of probable error for controls is + 0.9-2.6; methyl cholanthrene + 1.3-2.1; benz-pyrene + 0.8-2.1; and 1,2,5,6 di- benz-anthracene + 0.8-2.2 mg. In this series of experiments, half of the values were obtained in cultures freely exposed to the air as in our other dry-weight determinations. The other half were used to determine fermentations by gas vol- umes which in the conditions of our experiments were under lower oxygen tension and high carbon-dioxide tension ap- proaching anaerobic conditions. In the latter the absolute amounts were about 10 per cent less dry weight and more vari- able in suecessive experiments. 'The probable error of the determinations ranged from 0 (when two observed values were identical) to + 1.2 mg. Several long-time experiments were undertaken to attempt to follow changes in dry weight during the stationary and de- cline phases of cultures. Dry-weight determinations do not lend themselves readily to a study of the lag and the loga- rithmie phases. The differences between methyl cholanthrene eultures and controls are slight. As the curve levels off, it be- gins to descend slowly by a series of small sawteeth, similar to those of population counts but much less in extent, with an occasional high peak. While in general the curve for methyl [Vor. 28 16 ANNALS OF THE MISSOURI BOTANICAL GARDEN cholanthrene is slightly below that of the control, the curves frequently cross. An average of 118 determinations from the second to the twenty-fifth day gives a dry weight of 80.1 for the controls and 78.1 for the methyl cholanthrene cultures, with a probable error much less than + 1 mg. While only a single experiment was carried to 130 days, the results are similar to the above series earried only to 25 days (Graph 2). -E EM > > 4 / 70 20 30 40 §0 60 70 80 90 700 ог Hour. произ терм QU Ff KR J 4 3 4 хоум "W Nn ATUM CNN E aph 2. Growth in dry weights for first 25 days in long-term experiments. Gra Upper right, a xr population eurve with large inoculum. Note nearly complete absence of lag phas FERMENTATION In our first experiments we attempted to study the effects of methyl cholanthrene, benz-pyrene and 1,2,5,6 di-benz-anthra- cene. Six successive series were run and the volumes of carbon dioxide recorded. When the corrected volumes are plotted against time, in series I-IV (Graph 3) we note a continual in- crease in the amount of gas evolved, the close proximity of the curves of carcinogen and control, and the rising dry weights in each successive series. In series V the total carbon dioxide is below that of series IV in the control and methyl chol- anthrene, while that of 1,2,5,6 di-benz-anthracene is far ahead. 1941] 17 DODGE, DODGE & JOHNSON—CARCINOGENS AND YEASTS мар LLL ttt LI ГРЕЕ LJ Rungarn LE] annas ЕИЗ с ЖА БИРА щем AD H ЕР dnd цизвраи zor} i ЮЛ БЕШ H м РЕН im Я m FE YT: b та: 7 FEEFEE E HH eng H : BRE : г eo 1 i L] H Е f 1 S. Ei | tj į Д + TES = ЕЕЕ РЕ ВЕРЕ o 20 40 60 TIME iw Hous Cee LELI LEE zh LH 1111 ВЕЕ wt HH FH м 1 Ha кг, Е Е г s TERTA r хав k ERRE Дания ao PERRE EA TEET EREI AT n пашаш ЕНЕГЕ Зра Ez ениш 1i us пи HH Be H SHIRE «ЕЕ RN ЕҢ ? d ; as ci E $. уюгу Den Sas pine ЕЕ ШЕБИН : : ZH ЕЕН pu SHIT ёо кю 420 4420 ч SUE Е FEET | LELELI h EL LIL EE Ter 8 ^ deu 18 HF | TATH | Й Li LLLI Г ЕЕЕ E LLELE man Е ГЕН ж LLLLLLLE г я ин = TIT A в = " vt. иго + 4144 +- tt - ти T mmm soo Et tit ir A - 4 D з 40 "E 3 ПЕ ЕЕ їй ЁЁ - і шик n Lad ae i Е; ип «се За - же B н TTT I я 1 ЗЕТИ 2 ge + 5 і 1 2718 4 ny ї ө о © М ý ade " 40 + Wo Tm UE ~ A =P -— i tii "NNI ey THR A а: аро "un TIME iv Hours т 4 Я: ee Graph 3. Fermentation. Carbon-dioxide volumes plotted against time. [Vor. 28 18 ANNALS OF THE MISSOURI BOTANICAL GARDEN In Series VI the control curve becomes irregular and both benz-pyrene and 1,2,5,6 di-benz-anthracene show great in- creases over series V. From our later experience, the irregu- larity of the control curve suggests that had we continued a seventh series we would have found the yeast much more sus- ceptible to the influence of the carcinogens. An attempt was made to determine the reducing power of the solutions at the end of the experiments, hoping to check the carbon dioxide evolved against the decrease in glucose. The percentages for the different series were as follows for the controls: I, 38.1; IT, 40; IIT, 41.5; IV, 49; V, 48; VI, 45. There- fore it seems probable that part of the glucose is synthesized into a more complex carbohydrate (without or with part of the peptone) which remains in solution but which has lost its re- ducing power. Since difficulties were experienced with negative pressures toward the close of the experiments, we turned to gravimetric determinations during the next experiments and confined our attention to methyl cholanthrene (Graphs 4, 5 upper left). The weights of carbon dioxide were higher in seven consecu- tive series and consequently the proportion of the original glucose accounted for was higher, ranging from 27 to 55 per cent for controls and 29 to 57 per cent for methyl cholanthrene, yet there was no eonspieuous correlation between dry weights, weight of carbon dioxide, and the amount of glucose left. Since from other observations we had noted some differences in size and shape of cells, it seemed possible that the different types of cells might behave differently, and we attempted a partial separation by their different sedimentation rates (Satava, '18, further studied by Nielsen, 737). A 50-ml. culture to be used for inoculum was centrifuged until the cells, packed in the bottom, differed from those re- maining in suspension, as shown in haemacytometer counts. One of the samples was then diluted until haemacytometer counts showed approximately the same number of cells per unit volume. These were used as an inoculum for Series A-G, top; while the sediment was used for another series (A-G bottom, Graphs 5, 6 left). A similar separation was also if [T fe |. А ПИ п sd [XII ETT Ln UU ШЕШ ашшы ш ЕЕ рове Пи Hopes КЕТ dio ut «m ШШШ ai iis ia ААВ dede MEET E ЖИШШ вари L0: 208 dtu SIEUT ms A ОВК ih И a A ТҮҮ, M He in E pou 2 ао F LE pes ШИШИШИ Е. sod А Le А DT i аа ери | idi id ИШИН ЇЇ! 100 120 Ej iip LT ПШ RARIUS nit nL Hoher Tu Г О ШИ T UG ШИН ПШ ДЕ ШИ ТИН ШШ НЇН ан. Du Dr ДИТ! e| ас ELLE pees n E ве uc Dau pe рт мг үзө" om Ej SET ibada | ШИШИШИ Е: Е ЇНЇ ЙТ ШИН! НН. али, dusk. үш a aean | ЖШ ЕШ ШИП Ei ГЕШ = Lon T Eire ЧЕ ВАЗА МАНА А E - UH LE TE a Тай тайа SO ET и. ЕД v "ERE ^i | + ү! D LUE P NIIS LOI. EE ТЕРНЕР: j Г в ира TERI HE ; i SER E Е Mis qu j pa A пас A a ГГ d mu n LEES E I | dl те НА Euri GR Eh i 2 Е | " EINER. [Uu КУЧАТ ее H HE ЕО ti | Ls Ns. EE ———— 3° Ep un TU porem EUH [CE ста пе ЛӨП nur 2 TIL Graph 4. Fermentation. Another series with weights of earbon dioxide plotted against time. The last member of this series shown on upper left of Graph 5. төн ШШ ИШ: ШЫ: Їй КЕРИ ЕЕ Le MEA race I ELI m Г Г Ее " mb 22485. КН LE 297 ТИ TT "re а ОШ ШШ ЕЙ писа E т: E И ТҮГӨЙ ШИ a PI a ТР PARS EEATT] Е mmi И w ii Р р ^ tor t mm L nti i мны [^ it t m i TRE И мш ae LET таа d suner ШЙ! i НВ БЕШ — I. aoe зи Emir “= я Бинтти т мт ШШ, ШИ pt eet a: AR d - cU — = шш АМ ШШ ii Hri apegan та i ma На еи е ешш a a L Е i В ii! -- ai e üt d та tage TMs Ob rm accu {= i шан a “IRE "BR TE dint Д == i КТУ * min iit qnam ма вещ! unma деят fis аи M а LI ЕЕ 20 Сала ESE mmi IBI QU UA v EM и a à T 2. o eat ELTE] ! J LI Soar nn haa — LL] E Ri a а i үз р ip я КҮҮ Ads Boa Bb ME s (SEPAN IN а e Ea 3I PA Th ван i ART SH Ei вш f юй @ a HUS OR a ET КИ d T Ш үт zu Ba m ji Е ERE HE m d mi | i ате iii EBEN TM beta | | Ше LLLA tbe i "E ҮП "а КП E | Т тв m M Y з SE ZELTITS "EEG еВ кнава guessed ES 3 SEHE ЕСЕ Ва HE Li : |. Md Fr at EIS шие La В yer eee ES p ЛД Е ee Т Ж шеш p ИРИНА ЕЕ а I] Fm ten нашити ИШ ИШНИ TL КЕШИГИ ИШДИ et d HH te ШШ мо enm neve pm мне ү ШЗ po nea a ВОВ HET m Е Te AUN s Y. in СЕКЕ гише Ea SETTE ЕГЕТЕ К-ГА i LED M o vi d ЫШЫ ШҮП TER rye Lr | a n ИШЧИ EET m aia 4 co s aph 5. Fermentation. Upper left, continuation of series on Gra h 4. Series A- Е, aaa of fermentation by cells fro om the supernatant (Т) and sediment (В) in fraetional centrifugation studies. 1941] DODGE, DODGE & JOHNSON—CARCINOGENS AND YEASTS 19 attempted by natural selection with Gorodkova solution cul- tures and compared with our usual medium (Series H, Graph 6 upper right) and finally a comparison of the top (Ti, Ts) and bottom (Bs, Bs) cells with the whole (Si, Se, Ss, S4) culture (Series I, Graph 6 lower right). For discussion of the morpho- logical types found, see p. 25). With both the top and bottom cells the controls show less dry weight and leave more glucose in solution at the end of the experiment than in the methyl cholanthrene. The top cells of the controls show more of the glucose accounted for and slightly greater total fermentation than the methyl chol- anthrene cultures, while the bottom cells of the controls show less glucose accounted for and slightly less fermentation. Since the fermentation was allowed to go to completion, it would be expected that the difference in total fermentation would not be conspicuous. If we examine the amount at about 90 hours, when the curve begins to level off, the differences are more conspicuous, especially in the bottom methyl chol- anthrene. As we shall see later, there was no conspicuous сог- relation between these data and the various types of cells pres- ent. It seems probable that such relations exist but it will take more refined methods of classifying the morphology of cell types to show it clearly. POPULATION STUDIES In order to test the validity of our population counts, a given large square of the haemacytometer was photographed at 4—6-hour intervals for 150 hours and the resulting negatives were used to count the cells in each small square. Graph 7 left was based on the sum of rows 1 and 3 and of rows 2 and 4 to compare with our previous counts where we counted rows 1 and 3 in each of two adjacent large squares. The curves are closely parallel considering the difficulty of keeping the microscope at the same focus over a long period of time. "This fact should rule out serious error due to the move- ment of cells from one square to another, and points to a single stimulus or group of interrelated stimuli which act on the whole population within the time intervals between photographs. As [Vor. 28 20 ANNALS OF THE MISSOURI BOTANICAL GARDEN a further check upon our method we plotted, from a single field of our haemacytometer, the hourly counts selected at random from our mass of data and again found the curves parallel. In this case rows 1 and 3 in two adjacent large squares were counted. If we plot the average of rows 1 and 3 from each large square the parallelism is closer than if we compare the two rows farther apart, i.e., row 1 of the upper large square with row 4 of the lower large square (Graph 7 lower right), sug- gesting a stimulus diffusing from a focus which takes more than an hour to reach its maximum effect. Since these curves are extremely irregular with similar phase throughout the community, it is clear that there is some factor or group of interrelated factors acting uniformly upon the whole com- munity. In the eourse of our experiments 36 cultures were counted for varying periods of time. If we average all our counts where there were at least 10 data available at each hourly interval, we have curves approximating the classieal growth curve, since most of the fluctuations are averaged out (Graph 7 upper right). The control curve shows a lag of 12-15 hours, a logarithmic phase of about 30 hours and a decline phase of at least 60 hours, followed by a smaller secondary rise of at least 60 hours. The slope of the secondary rise is less steep and the total counts are considerably less. Richards (732, '40) has studied this secondary rise in much detail, using Saccharo- myces cerevisiae. Our data parallel that of Richards rather closely considering the differences of medium and species of yeast. Similarly, the average curve of the carcinogen is smooth, showing a lag of 20-22 hours, a logarithmic phase of about 80 hours, a short decline of about 20 hours, and secondary and tertiary rises of about 25 hours with declines of about the same length. The rises are less steep and the height of the first peak is less, but the decline phase tends to remain at a much higher level than in the control. Fishbein, Weaver and Scherago, in a paper read at the annual meeting of the Society of American Bacteriologists as this paper was in press, working with Escherichia coli and 1,2,5,6 di-benz-anthracene, showed essen- ECE ТЕБЕ ык La TEE jut | | 1, an Ш Реса ча ЗЕБИ EA Ae JETTI ШЕШИП, ситен! ij 1 220 і Т E p» munus АГЕЕВ" хост LE SRE T ET EHH SEAS ВЕЕ Sg" HEBEL E ва Ts a tj EEEE G58EB dou E MEM BEBE Oo gm z% 7 53 5 CIS nt Я Фон о BOCs Ф g а 525 асыш ч а ed ERE зз о $2869 вом Бонн Baar Bg os 5 Н ДЕФЕ Орад „В Б ex 434, noosa КЕКЕ ng ЗЕЕ ЕН os ЗА E Н шо [70] 2 = ag 2 QE IE a. aod SPH UBS Wd © О o.na Bates ZS 2.8 Бр Есен © cu og E, 2450505 ansa Е КЕ 5 оао e SRE г 3 ams ся Eu m ES по ат ван с Фор Е UD du. Graph 7. Lower graphs, cell counts from appe ig ge E 4—6-hour учен Left, sums of rows 1 and 3 plotted against rows 2 and 4 as in Graph 8 of hae eytometer counts. Note parallelism of eurves. Right, н only, row 1 plotted against row 4. Upper right, average curve for all eurves shown on Graph 8. 1941] DODGE, DODGE & JOHNSON—CARCINOGENS AND YEASTS 21 tially the same type of curves, although working with bacteria and colloidal suspensions of the carcinogen. In three cases where we started with higher inocula (about 200 cells per field, Graph 8 upper and extreme right, Graph 2 upper right) the curves rise, with no appreciable lag, to a much lower peak and level off for the first 100 hours in a long sta- tionary phase (senescent phase of Steinhaus and Birkeland). This curve somewhat resembles that reported by Hopper and Clapp (739) and closely that of Fishbein, Weaver and Scherago, mentioned above, when they used large inocula. These results agree essentially with those found in dry weights and fermentations where also we were dealing with large inocula. There is no appreciable difference in the slope of the curves between cultures grown in the haemacytometer and those grown in test-tubes and sampled for counting. Averages, however, often conceal some important data. If we examine the curves of cultures in the two fields of the haemacytometer having the same medium and inoculated with approximately the same number of cells from the same suspen- sion and subjected to the same environmental conditions but not physically connected, we find that they rapidly get out of phase and vary quite independently. If the fluctuations of the counts were due to genetic or to other factors inherent in the inoculum or factors of the external environment of the culture, we should expect the fluctuations to occur simultaneously in the separate cultures. This is not the case, and we may safely conclude that the fluctuations in the populations, resulting from division and disintegration of cells, are not the result of inherent variations in the inoculum nor in the external environ- ment of the culture but rather the result of stimuli arising within the culture itself. Steinhaus and Birkeland (’39) prob- ably observed similar fluctuations in Serratia marcescens Bizio but they present only average curves. They also emphasize the importance of disintegration of cells and the reutilization of the nutrients thus freed during the senescent phase. Our curves (Graph 8) show a high degree of randomness with large fluctuations compared to the mean, such as one en- counters in curves in stock-market reports where the deci- [Vor. 28 22 ANNALS OF THE MISSOURI BOTANICAL GARDEN sions of relatively few individuals in the community pro- foundly affect the curve of prices of stocks. Applying the usual tests, we find that our curves are neither those of random dis- tribution nor of Brownian movement. We should hardly expect random distribution since each new cell originates from a pre- vious cell at a definite point in time and space, and from our fragmentary observations when light conditions were espe- cially favorable, it seems to be included in a thin gelified sheath for some time after it has severed observable protoplasmic connection with the parent cell. While movement of single cells has been noted, presumably after the sheath has disintegrated, it has none of the characteristics of Brownian movement. If the phenomena causing the fluctuations were mass effect due to large numbers of cells, we should expect large smooth sweeps with a small degree of randomness. In a large culture such as we used in determining dry weights and fermentation, we find this type of curve, since the small individual fluctua- tions arising at different points in time and space rapidly average out and give a relatively smooth curve with only the extremely large fluctuations evident. We have already seen this when we averaged all our individual curves where at least 10 counts for a given hour were available (Graph 7). In the light of our experiments, it seems probable that we are not dealing with individuals in a community, but that the com- munity as a whole represents a liquid tissue in which the cells have lost their geometric arrangement in space but have still retained some differentiation in function, and that certain cells are differentiated to secrete some growth hormone. Per- haps this unlooked-for result should have been expected, since from studies of comparative morphology and phylogeny of the yeasts we have a definite degeneration series from organisms with very complex morphology and life cycles to the ‘‘unicellu- lar’’ condition in Saccharomyces (Dodge, ’35). The presence and persistence of the gelified sheath point in this direction but need much more study before their meanings are clear. We also need more continuous observation by some method which would give a continuous record of the fluctuation in number of cells, paralleled by frequent microscopic examina- ЕЕ БА БЭ | 2 1 ПМЕ IN Dars in cells Broken xpress 000 to e ith methyl cholanthrene. H c Yo) ++ o = © — [2] ВЕ Фа Ф Рд + ‚© ЕЕ i33 даа га Deon EE 3 on ©, ir Box N Ba о Ф 45 o 65 Е a. ен о Ф nE 8% о о, Ех се + съ aa сад ва е = 243 ue Я ее dag E o Su rg t». Bus non amm ос Я ES зоо — dH Бор 3 чаб эчә ou un = 8 2 Рио Cen 585 ои ФЕ gud 1941] DODGE, DODGE & JOHNSON—CARCINOGENS AND YEASTS 23 tions to locate the few individuals which are producing the fluctuation and to describe their morphology. If we examine the individual fluctuations, we also note that very rapid ‘‘birth’’ rates are followed by nearly as rapid dis- integration rates, so that the curve for a single fluctuation is nearly symmetrical. The very rapid disintegration of cells from one hour to the next has been evident in most of our work and seems to have been overlooked by many workers. This rapid disintegration apparently sets free nutrients which may be used again to allow more growth, for if we examine the bottoms of the depressions we find that the curves are again approximately symmetrical, although the peak reached after a depression is usually lower than the previous peak. When we examine the curves for methyl cholanthrene we find essentially similar conditions. In general, the initial lag period is longer and the peaks are broader as well as higher. It would appear that the stimulus is more continuously ap- plied in addition to that of the postulated hormone secreted by individual cells. The larger numbers of cells usually pres- ent after the first rise would also tend to smooth the curves, as there would be a greater chance of ‘‘hormone’’-secreting cells being present. If the methyl cholanthrene is a constantly acting stimulus, it would tend to mask the rapid disintegration rates, thus resulting in inereased growth rates during the loga- rithmic phase and apparently decreased disintegration rates. Since this does not appear to be the case, it is probable that the methyl cholanthrene stimulates cells to divide that ordi- narily would not do so, apparently injured or senescent cells. If a diffusible chemical impulse were the only factor, and all cells receiving the stimulation divided immediately, the peaks would be much greater and would reach approximately the same heights where the initial cultures contained approxi- mately the same number of cells. Hence it seems probable that at any one time only a portion of the cells are sensitive to stimu- lation. If such cell sensitivity were merely a function of age, the curves would rise in continuing peaks as all the cells came of suitable age, which is not the case. The differences in height, number, and time of appearance of the peaks suggest rather [Vor. 28 24 ANNALS OF THE MISSOURI BOTANICAL GARDEN a few sensitive cells which form centers of growth. The height may be determined by the number of sensitive cells and whether such cells are in our immediate field of vision. Their age may be important in determining the response to the stim- ulus. Where there are very few sensitive cells, as in a small inoeulum, the peak is conspieuous and sharp; where more of these cells are present, the peaks appear to be made up of a succession of sawteeth, less sharp as more sensitive cells are present until in large inocula we reach a broad and smooth curve similar to the classical growth curve. The other cells, which are present in far greater numbers than the sensitive cells, may be regarded as somatic cells. It is not clear at pres- ent whether they die slowly or persist until their function is taken over by a new erop of somatic cells. That these cells are merely reproductive cells which function without a definite stimulus is untenable, since if there were no diffusible stimulus there would be no parallelism between curves plotted from different regions of the same culture. We would have only isolated centers where multiplication and dis- integration would start quite independently, which is not the case. There is also an effect traceable to the history of the culture evident through a series of transfers, since otherwise there would not be as much parallelism between subcultures trans- ferred on the same day. Despite differences great enough to indieate that they vary independently, there is sufficient paral- lelism in population counts, and even more in dry weights and fermentation data, that duplicates, i.e., contemporaneous sub- cultures, are definitely more alike than cultures from the same stock culture made days or weeks apart. There is still another complicating factor. Evidence is ac- cumulating that there is a definite cycle of fluctuations in the physiology of the cultures. This cycle appears to be a long one, and while frequent transfer may somewhat modify the time it does not alter the fundamental process. It appears that instead of being a strictly unicellular organism, completing its life history in a comparatively short time, the yeast is a com- plex organism in which the whole culture behaves as an indi- 1941] DODGE, DODGE & JOHNSON— CARCINOGENS AND YEASTS 25 vidual, a liquid tissue with a much longer life history which the addition of fresh nutrient (transfer, which is similar to tissue cultures and cuttings of the higher plants) may alter only in a minor way. Taking long enough periods we may re- produce dry weights and fermentations very closely. Re- appearance of very high dry weights (over 90 mg. under our standard conditions) in approximately six-month periods suggests this as an approximation for the cycle of our stock culture. CELL DIFFERENTIATION Study of cells of individual cultures reveals a fairly con- stant morphology during the logarithmic phase. The inoculum, if taken from an actively growing 24-hour culture, consists of comparatively small, homogeneous cells. If inoculated from older cultures this form is also assumed during the lag phase and remains homogeneous during a portion of the logarithmic phase. Small vacuoles appear and granules of reserve food become visible in the cytoplasm. In still older cultures, the vacuoles become large and the granules more numerous, there being apparently little correlation as to whether the granules are located inside or outside of the vacuole itself. Elongate cells rarely occur in young cultures but are relatively more abundant as the culture ages, although their number is seldom large. In our medium, the elongate cells seem to revert easily to the spherical form but those isolated by a micromanipulator and transferred to fresh medium frequently retain the elongate form for several generations. Although cultures of the same age are relatively constant in morphology during the early stages after transfer, the morphology is progressively more variable in older cultures, perhaps due to different disinte- gration rates. In this preliminary study we recorded the cell types as per- centages of large spheres, small spheres, and elongate cells. We also recorded the spherical cells with buds, the elongate cells with buds, cells with large vacuoles with or without re- serve materials, and shadow cells. In our experiments with fractional centrifugation we recorded the following: [Vor. 28 26 ANNALS OF THE MISSOURI BOTANICAL GARDEN SAMPLE OF SUPERNATANT (“ТОР”) Spheres Buds Vaeuolate Series Elongate Shadow Large | Small Cells |Spheres| Elon- | Without| With | Cells ate Reserves Reserves A 13 78 9 13 7 6 0 0 B 7 82 11 11 8 4 0 0 С 9 81 10 14 8 2 0 2 р 20 73 7 20 5 0 0 5 E 10 80 10 15 4 2 0 1 F 18 67 15 9 6 6 3 + с 32 59 9 12 4 4 0 7 Average Inoeula 15.6 74.3 10.1 13.4 6 3.4 0.4 2.7 Control 42.1 53.3 3.6 18.7 2 9.0 4 0.8 42.1 54 3.9 18.4 2.4 9.9 4.5 0.2 SAMPLE OF SEDIMENT (‘‘BOTTOM’’) A 37 57 6 28 1 18 2 0 B 46 49 4 32 0 19 3 0 C 47 50 3 30 0 11 11 0 D 49 46 5 28 4 15 9 2 E 42 56 2 23 0 7 0 0 F 61 36 3 20 1 9 3 0 G 58 40 2 18 0 7 10 0 Average Inoeula 48.6 47.7 3.6 25.5 0.9 12.3 5.4 0.3 Control 43.8 52.1 3.3 17.4 1.7 11 3.4 0.3 2.5 53.8 3.6 21.0 2.2 8.4 3.8 0.4 It is apparent that the crude fractionation we used did not completely separate the cell types as we classified them, although a larger proportion of large spheres, vacuolate cells, and buds was found in the sediment and more elongate cells and shadow cells remained in the supernatant. No clear cor- relation has been evident between the cell types, as we have classified them, and physiologieal activity. At the end of the experiments, the cultures inoculated with either top or bot- tom cells approach the bottom types of the inocula obtained by fractional centrifugation, as shown in the next to last line in the above tables. The methyl cholanthrene cultures consist- ently show more elongate and vacuolated cells at the close of the experiments (last lines of the above tables). We also tried some media for possible selective action with reference to our cell types. Four solutions were inoculated with 1941] DODGE, DODGE & JOHNSON—CARCINOGENS AND YEASTS 2 two drops of a uniform suspension of a 24-hour culture which had been transferred every 24 hours for several days previous to this inoculation. From these solutions the yeasts were trans- ferred to fresh media every four days. Examinations were made after the third transfer. Sabouraud's solution (our usual medium) gave results similar to those we had obtained previ- ously; Williams? and Cohn's solutions favored the develop- ment of colonial forms of varied morphology; Gorodkova's solution gave a much larger number of elongate cells with more budding of these cells. Since these differences were greater than we had secured by fractional centrifugation, we at- tempted to compare growth in Gorodkova solution with that in the Sabouraud culture as a control. Four-day cultures of this third transfer to fresh media were shaken to form a uni- form suspension and the cell types and number of cells per ml. were recorded. A portion was then removed from each cul- ture, placed in sterile flasks, and diluted until the number of cells in each was approximately equal as determined by an- other count. During the dilution process the cultures from which the source of inoculum was taken were filtered through a Berkfeld filter and a portion of each filtrate equal to the amount taken to make the stock cultures was added to the other culture, so that the final results were strictly comparable, not only as to approximate number of cells, but also the component nutrients were identical (Series H). The inoculum and final results contained the following cell types: SERIES H и Inoculum End of Experiment Cell types а в Gorodkova Sabouraud Me. Chol. | Control | Me. Chol. | Control Large spheres 7 28 32 36 | 36 | 31 8 41 | 32 24 Small spheres 68 67 60 60 | 57 65 | 70 57 | 64 74 Elongate cells 25 5 8 4 ri 4 2 2 4 2 Spheres with buds 43 12 18 14 | 21 12 | 18 24 | 23 18 BONUS cells with 21 4 3 3 5 2 0 0 2 2 Cells with large vacuoles 2 4 10 9 if 9 И 4 8 7 Vacuolate cells with reserves 0 2 4 0 0 0| 3| 3| 5 2 Shadow cells 0 0 0 0 0 0 0] 0 НН. © [Vor. 28 28 ANNALS OF THE MISSOURI BOTANICAL GARDEN No correlation with cell types was noted beyond slightly more elongate cells persisting from the Gorodkova inoculum. Finally we compared flasks inoculated with a whole culture with the top and bottom fractions obtained by fractional cen- trifugation. (Series I). SERIES I Inoculum End of Experiment Cell types Top Bottom Whole Me. Chol. Control 1 | В, | 5, Sa Ts | B, S, БЛ Large spheres 12 38 24 |18 | 32| 41| 29 | 24 | 28| 38] 35 П spheres 80 60 69 |80 64 54| 69 | 75 | 70| 61| 63 Elongate cells 8 2 7 1 2 Spheres with buds 12 10 20 |16 | 28/17] 14 |13 | 21] 23] 17 Elongate cells with uds 5 2 3 216200210 Cells with large vac- 8 2 10 7 2|14|]|11|14| 81991 6 Vacuolate cells with reserves 0 T 1 0.1.0] 7144.8 BL 4| 1B Shadow cells 2 0 2 010101 0/0/|0]| 1| O0 We found that the culture as a whole was much closer to the bottom fraction in dry weight while closer to the top fraction in fermentation. Evidently something has happened in cen- trifugation as the fermentation curves in methyl cholanthrene rise more steeply, especially the curve of the top fraction; the dry weight is conspicuously higher in both methyl chol- anthrene and control. Further work will be necessary to decide whether this is due to mere separation of cells from the gelified sheath or to some type of injury. SuMMARY Studies of dry weight produced under standard conditions, and population studies with large inocula, revealed only small differences between cultures with methyl cholanthrene and the controls. The methyl-cholanthrene cultures tended to have a slightly longer lag phase and slightly less dry weight. Studies of population by various methods of counting, using smaller inocula, show a much longer lag phase, a gentler slope in the first logarithmic phase, and higher and more sustained peaks in the decline phase in cultures with carcinogens. Evi- 1941] DODGE, DODGE & JOHNSON—CARCINOGENS AND YEASTS 29 dence is presented that each culture behaves as liquid tissue of an individual with physiological as well as morphological dif- ferentiation of cells, with the probable secretion of a hormone by relatively few cells which stimulates cell division. Car- cinogens stimulate the division of senescent or injured cells which are not ordinarily susceptible to the action of the postu- lated hormone. Stock cultures have long cycles of physiological activity which can be modified only slightly by frequent transfer to fresh media. Attempted separation of morphological types of cells by fractional centrifugation and the selective action of media have been only partially successful, and it has been impossible to correlate these types with physiological activity. Movements of translation similar to those of the Myxo- phyceae have been noted; also the presence of a gelified sheath which holds the cells together in small colonies. LITERATURE No attempt has been made to review the extensive literature on carcinogens and their action, nor on factors influencing growth of yeasts. The following references include only those mentioned in the foregoing text. oe C. В. (’40). Photo-oxides of carcinogenic compounds. Nature 145: 303. 1 fig. Dodge, C. W. a Medical mycology, fungous diseases of man and other mam- mals. St. Lou Dodge, C. W. and B. S. Dodge (737). Some effects of methyl cholanthrene on the morphology and growth of yeasts. Ann. Mo. Bot. Gard. 24: 583-590. pl. 34. Fishbein, M., R. H. Weaver and M. Scherago (740). The effect of 1:2:5:6: dibenz- anthracene on the growth of Escherichia coli. [preprinted from Jour. Bact. 41: 27-28. 1941] Hopper, S. H. and D. B. Clapp (239). Effect of д and other hydro- carbons on the growth of Escherichia communior. Jour. Bact. 38: 13-22. 1 fig. Loofbourow, J. R. and C. M. Dwyer (738). кр consistency of weights and counts in determining LR орн by и г. Studies Inst. Div. Thomae 2: —135. 3 fig. Nielsen, N. (737). а а über die Ите der Hefe. Trav. Lab. Carlsberg, Ser. Physiol. Píšová, M. (734). Anatomická struktura Pu kolonií, Rozpravy II. Třídy Ceské Akad. 44:41: 1-15. 1 pl [Vor. 28, 1941] 30 ANNALS OF THE MISSOURI BOTANICAL GARDEN Richards, О. W. (732). The second cycle and subsequent growth of a population of iue Arch, f. Protistenk. 78: 263-301. (740). Yeast growth: a possible test organism for X-radiation. Radi- ees 38: 317-326. i. J. (718). O igen in formách kvasinek. 50 p. 3 pl. Praha. Steinhaus, E. A. and J. M. Birkeland (739). Studies on the life and death of bac- teria. I. The senescent оми in aging eultures and the probable mechanisms involved. Jour. Bact. 38: 249-261. A MONOGRAPH OF THE GENUS PARMELIA IN NORTH AMERICA, NORTH OF MEXICO! EDWARD CAIN BERRY Formerly Graduate Student in the Shaw School of Botany of Washington University INTRODUCTION This study has been conducted with the purpose of bringing about an orderly arrangement of the species of Parmelia and to clarify the literature of the genus for North America, north of Mexico. During this investigation, all obtainable types were studied. Where types have been lacking, original descriptions and illustrations have been studied and compared with exsic- catae and collections from the type localities for the purpose of correlating the description of those species with the ma- terial at hand. To prevent confusion, in the literature of the genus Zahlbruckner’s arrangement of the species in Engler and Prantl (Nat. Pflanzenfam. ed. 2, 8: 233. 1926) has been followed, except where the monographer believes that Zahl- bruckner is in error. History The genus Parmelia was established by Acharius in 1803, to supersede Imbricaria Acharius (1794). Imbricaria was used by various workers to designate the genus until 1930, when Parmelia Acharius was placed provisionally in nomina generica conservanda, and Imbricaria Acharius in nomina rijicienda (International Rules of Botanical Nomenclature, suppl. nomina generica conservanda proposita. 127. 1930). In Acharius’ treatments subsequent to 1803 (Lichenographia Universalis. 1810; and Synopsis Methodiea Lichenum. 1814), he removed considerable portions of the genus as first con- sidered. Of the species which he retained in this last treat- ment, most remain in the genus as it is understood to-day. ! An investigation carried out at the Missouri Botanieal Garden in the Graduate Laboratory of the Henry Shaw School of Botany of Washington University and submitted as a thesis in partial fulfillment of the requirements for the degree of Doetor of Philosophy. Issued February 26, 1941. (31) [Vor. 28 32 ANNALS OF THE MISSOURI BOTANICAL GARDEN Dodge and Baker also agree that the type should be selected from this group (Second Byrd Antarctic Expedition— Botany. Liehens and Lichen Parasites. Ann. Mo. Bot. Gard. 25: 589. 1938). This would conform with the present rules of Botanical Nomenclature since it would exclude the doubtful species and those which other workers have placed in related genera. The author feels that Lichen saxatilis Linn. meets all Ше require- ments recommended for the selection of the type, it being a species with which Acharius was thoroughly familiar and at the same time one species which was universally recognized by early writers as a member of the genus Parmelia. Conse- quently, in this monograph, the type of the genus Parmelia is designated as Parmelia saxatilis (L.) Ach., Meth. Lich. 204— 205. 1803. The description of the plant is given on page 82. EcoroavY The area considered in this monograph includes continental North America, the immediately adjoining islands north of the international boundary of Mexico, and the islands of the Gulf of Mexico. This area represents a variety of habitats. The elevation varies from sea level to 4419.5 m.; the rainfall from less than 12.5 em. to more than 508 ст. annually; the average annual temperature from 24° to 25° C., the temperature range from —45? to more than 56° С. The lichens growing on the peaks of the higher mountains may be considered as occupying an arctic environment, the plants from the southern end of Florida and the mouth of the Mississippi as tropical in their affinities and climatic condi- tions. All of the species of Parmelia found in the area defined will be considered in this monograph regardless of their affini- ties or environmental conditions. Species of Parmelia are found on rocks, bark of trees, fence- posts, buildings, fallen branches, logs, and rarely on the ground. Lichen growth is influenced by the environmental fac- tors of water supply, substratum, wind, smoke, and tempera- ture. Some species are tolerant to many environmental con- ditions and have a wide distribution; others, which have a very 1941] BERRY—THE GENUS PARMELIA IN NORTH AMERICA 33 limited tolerance to one or more of the environmental factors, are consequently more restricted. Parmelia conspersa, P. caperata, P. centrifuga and P. lineola require full sunlight, growing on rocks, the upper branches of trees, and on buildings. These species are never seen in vigorous condition in shade or in subdued light. Parmelia lineola is found in the West and Southwest, covering the tops of rocks where they are exposed to the intense light of the desert sun. Parmelia olivacea and its variety aspidota grow in shaded situations, the full light of the sun inhibiting their growth. Another group which requires some direct sunlight but is seldom found in full light consists of Parmelia rudecta, P. Bor- reri, P. cubensis, and P. erecta. These lichens are found grow- ing in positions where the full light of the sun falls on them for only a portion of the day. The moisture requirement of the different species 1s very variable. Parmelia lophyrea is endemic only in the region of the Northwest where the moist wind of the Pacific Ocean comes in contact with the foothills of the mountains, while P. lineola is found in the desert section of the continent. Parmelia en- teromorpha prefers a cool moist habitat, but also shows tol- erance for desert conditions. The species attains its greatest size and abundance in the redwood section of northern Cali- fornia, but is also found in reduced size and numbers in the deserts of southern California. The distribution of a number of species indicates a definite temperature limitation. Parmelia Cladonia is limited to the northeastern portion of the United States and the higher ele- vations of the Appalachian Highlands. It is common in the White Mountains of New Hampshire and continues south in the mountains at high elevation. The southernmost collection is from Clingmans Dome, Tennessee-North Carolina, an ele- vation of 6600 feet. Parmelia physodes is a common lichen in the timbered section of Canada, northern United States, and in the higher mountains to the south. In the lowlands, it ex- tends south to Florida and Texas, gradually decreasing both in the number of individuals and in size. Parmelia latissima, [Vor. 28 34 ANNALS OF THE MISSOURI BOTANICAL GARDEN P. tinctoria and P. livida are abundant in Central America and Cuba. They grow also in the southern part of the United States, but never in localities where severe freezing occurs. Parmelia caperata and P. Borreri are found in all parts of the area studied, with little change in appearance. Smoke is always harmful to the species of Parmelia, and in any considerable eoneentration will destroy the plants entirely.? Мовеногову Рагтейа is leaf-like in form, growing above Ше surface of the substratum. It is composed of an upper cortex, an algal layer, medulla, and lower cortex. The cortex, sometimes re- ferred to as a plectenchyma, is formed of small septate vertical hyphae which are woven into a firm layer. The distance be- tween the septae is about equal to the diameter of the hyphae, giving the cortex the appearance of being formed of small iso- diametrie cells. The hyphae are separated by secretions which form a matrix characteristic of the species. Cortex.—The thickness of the cortex in a species is char- acteristic, but it may be modified by age or environment. Old plants of some species in exposed situations may lose the entire upper cortex, in which case the whole upper surface becomes sorediose. In the younger portions of the thallus, the upper and lower cortex are essentially the same in thickness. Hither the upper or lower cortex is often broken by cracks which per- mit the exchange of gases with the gonidial and medullary hyphae. In those species which have the margin of the thallus raised, as in P. perforata and P. erecta, the lower cortex is sub- ject to the same factors of weathering as the upper cortex. The exposed portion of the lower cortex is approximately of the same thickness as that of the upper, while farther from the margin it is definitely thicker. The structure of the lower cor- tex is the same as that of the upper, except that from its surface either rhizinae or a cushion-like mass of hyphae arise. The outer layer of hyphal cells and the rhizoids are usually heavily 'Haugsjá, Pål К. Über den Einfluss der Stadt Olso auf die Flechtenvegetation der Baume. Nyt Mag. Naturv. 68: 1-118. 1930. 1941] BERRY—THE GENUS PARMELIA IN NORTH AMERICA 35 inflated with a dark brown or black pigment. A detailed study of the thickness of the thallus in 72 collections of P. rudecta shows that it varies from 400 to 810 и, with an average thick- ness of 638 и. The thickness of the upper cortex varies from 18 to 23 и, with an average of 20 и. The lower cortex varies from 65 to 84 и, with an average of 73 и. The variation of Ше algal layer is from 42 to 45 р, with an average of 44 и. The thickness of the medullary layer varies from 262 to 667 y, with an aver- age of 502 и. Isidia are coralloid branchlets and granules on the upper surface of such species as P. caperata, P. Borreri, P. crinita, and they are constant characters. They are the result of a con- tinued growth upward of the vertical hyphae of the cortex. Some of the algal cells from the gonidia are carried up with the growing hyphae, and a gonidial layer is formed in these struc- tures. Such outgrowths increase the photosynthetic area of the plant and also furnish an important means of vegetative reproduction. They contain all of the essential vegetative parts of the lichen body, and are easily broken off. They can be carried by wind, gravity, or by water to new locations, where they may produce new thalli. Soredia are constant structures on some species of Parmelia, but never found on others. They are uncovered masses of hyphae which contain algal cells, and they may take the form of small round dots, as in P. rudecta, or confluent globular masses along the margins of the thallus, as in P. cristifera. Soredia arise from the gonidial layer of the thallus and reach the surface by the breaking, cracking, or loss of the cortex. The medullary hyphae absorb large quantities of water, and the im- bibitional swelling causes them to exert pressure on the cortex when the plant is thoroughly wet, particularly after having been dried. The pressure of the medulla pushes the gonidial layer into the break and a soredium is formed. When the weakened portions of the cortex assume definite lines, the soredia are formed in a distinct pattern of reticulate ridges, as is seen in P. sulcata. In those species with definite marginal soredia, as P. physodes and P. cristifera, the upper and lower cortex are weakly joined together and the soredia are formed [Vor. 28 36 ANNALS OF THE MISSOURI BOTANICAL GARDEN by cracks at this line. The number and size of the soredia are determined by the extent of the weakness of the cortex and the amount of the swelling of the medullary hyphae. When the cortex is broken and the algae are placed in a position to secure more carbon dioxide, the rate of photosynthesis increases. Thus the food supply to the gonidial hyphae becomes greater and more rapid growth is produced. The soredia, being unpro- tected, are easily detached from the thallus by wind or rain. They thus serve as an important method of vegetative repro- duetion for those species in which they occur. Gonidia.—' The gonidial layer consists of algal cells (species of Protococcus) enmeshed in a tangle of thin-walled hyaline hyphae. The algae may form a continuous layer with the hyphae, or the layer may be lobed or even broken in spherieal masses of hyphal and algal cells. The hyphae usually partly surround the algal cells with disk-like appressoria, though in many instances it is possible to see haustoria penetrating the cell walls of the algae. The algae of the gonidial layer sup- ply the elaborated food for the fungus of the lichen. Medulla.—The medullary layer consists of a loosely woven web of septate hyphae running parallel to the surface of the thallus. The hyphae are thick-walled and usually hyaline or very brown or yellowish. In a few species, P. pertusa, P. phy- sodes and P. olivacea, some medullary hyphae are extremely large and dark brown or black. The medulla serves the purpose of holding the thallus together and providing a means of mov- ing the nutritive materials in the plant. In species where the inflation of the thallus occurs, there is a separation of the hyphae of the medulla. The medullary hyphae are usually heavily impregnated with the lichen acids, as is shown by the reactions given with potassium hydroxide and ealeium hypo- chlorite. The color of the medulla when exposed by the removal of the cortex is white or yellow. The loosely interwoven hyphae of the medulla also serve to provide a passage for gases through the thallus of the plant. Apothecium.—The apothecium of Parmelia consists of an open disk on the upper surface of the thallus. The margin con- 1941] BERRY—THE GENUS PARMELIA IN NORTH AMERICA Sii sists of two well-defined layers. The inner layer is a continua- tion of the hypothecium and is of the same origin, structure, and color. This covering is a continuation of the protective hyphae which formed around the ascocarp when it first began development below the upper cortex. The hypothecium never contains any algal cells. The algae in the gonidial layer below the hypothecium are continuous with those in the thallus, and are always present except in those species in which the apothecium is perforated. The perforations of the apothecium arise from the death of the algal cells. In turn the hyphal cells die from lack of food supply and drop out, leaving a hole in the apothecium and thallus below. The perforation of the apothe- cium may appear occasionally in any species, but in certain species this is constant enough to provide a secondary diag- nostic character. The apothecium may be sessile, attached di- rectly to the thallus by the under-side, or it may be short- stipitate, the short stalk raising the fructification slightly above the surface of the thallus. The amphithecium or outer protective layer surrounding the apothecium is similar in general appearance to the thallus, and is thalloid in origin. The hyphae forming it are a direct continuation of those of the thallus, but the variation in thick- ness is much less than in the thallus. The gonidia are continu- ous with the thalloid gonidia. The disk of the ascocarp is lo- cated just above the sub-apothecial gonidial layer and consists of a sterile base (the hypothecium) which is composed of small, closely woven, hyaline or light brown, septate hyphae. Immedi- ately above the hypothecium lies the thecium, composed of the ascogenous hyphae, and asci, with the ascospores and the sterile paraphyses imbedded in a gell. The asci are clavate, and usually contain eight ellipsoid asco- spores. In a few species the number of spores may be two or four, as in P. pertusa, or many, as in P. multispora. The asco- spores are hyaline, non-septate, and usually thin-walled. Oc- casionally the walls are thick, as in P. Herreana. The paraphyses are septate, sparingly branched, and usu- ally enlarged at the tip. The color of the disk is due to the color of the tips of the paraphyses and the surrounding gelatinous [Vor. 28 38 ANNALS OF THE MISSOURI BOTANICAL GARDEN matrix. In its development the epithecium, or disk covering, is not distinguishable from the open disk. The ascospores ma- ture continuously throughout the existence of the open apothe- cium. It is common to find both young asci and mature spores in the apothecium of any species. Spermagonium.—The spermagonia appear as black dots on the surface of the thallus or the apothecial margin. The black dot is the spermagonial opening, commonly called the ostiole. The spermagonium is a pyriform or spherical cavity with a black upper wall and a hyaline lower wall. Spermatophores arise from the hyaline portion of the wall and are sparingly branched. The spermatia are simple, straight, or cylindrical with a slight constriction in the middle. They are discharged through the ostiole along with a colorless gelatinous secretion when the thallus is wet. In Parmelia they are small and lack sufficient differentiation to be of taxonomic value in the genus. Rhizinae.—The rhizinae arise on the lower surface of the thallus as outgrowths of the hyphae of the lower cortex. They may be lighter, darker, or of the same color as the surface of the lower cortex. They usually consist of numerous thread-like strands, but they may be either papillose or united into com- pact strands terminating in a mucilaginous disk called a hap- torium. The rhizinae serve to attach the lichen to the sub- stratum. There is little reason to believe that they have any considerable importance as organs of absorption or conduction of minerals in solution. CHEMICAL REACTIONS The color reactions of potassium hydroxide, calcium hypo- chlorite, and paraphenylendiamin are the result of these re- agents on the specific lichen acids. A characteristic color reac- tion is an indication of the species of a lichen when taken in connection with morphological characters and distribution. A chemical reaction alone is not sufficient to differentiate a species. The technique followed in determining the color reac- tion was to make several ‘‘free hand’’ sections 40-100 р in thickness through the thallus so that both cortices of the lichen 1941] BERRY—THE GENUS PARMELIA IN NORTH AMERICA 39 were shown. The sections were placed on a slide and the re- action of the chemieal reagent on each part of the thallus was observed through a dissection microscope. CHART OF CHEMICAL REACTIONS Paraphenyl- Species KOH CaOCl, KOH-4CaOCHh | endiamin olivacea var. aspidota none none non yellow Bolliana none none light pink none cetrata none none none none cetrata var. subisidiosa none none none none cirrhata none none none yellow sphaerosporella none none none red-brown chloroch none none brown yellow cube none none none yellow encausta none none none none exasperata none none none none perata var. incorrupta none none none red Hubrichtti none none none none multispora none none none none sulphurosa none none none none olivace none none none yellow texan none none brown non caperata var. subglauca none none none none onspers yellow none on red-brown furfuracea yellow none yellow ellow tinctoria yellow none none none conspersa var. isidiata yellow none none none ata yellow none none red centrifuga yellow none yellow none erinita yellow none one none conspurcata yellow none red none endoxanth yellow none yellow yellow enteromorpha yellow none yellow yellow lophyrea yellow none one none leucochlora yellow none none none omphalodes yellow none none none perforata yellow none pink none perforata var. hypotropa yellow none pink none pertusa yellow none yellow none perlata yellow none none none perlata var. ciliata yellow none none none physodes yellow—then r none brown red proboscidea yellow none none none quercina yellowish none yellow sulcata yellow none yellow red-brown lineola yellow none none none saxatilis yellow yellow yellow red-brown saxatilis var. Anzi yellow yellow yellow red-brown cetrarioides yellow yellow yellow red-brown [Vor. 28 40 ANNALS OF THE MISSOURI BOTANICAL GARDEN CHART OF CHEMICAL REACTIONS (Continued) Paraphenyl- Species KOH CaOCl, KOH-CaOCL | endiamin erecta none yellow yellow yellow livida none brown brown orange Herrei none brown none none Cladonia brown none brown yellow praesignis brown none none gray incurva brown none none none aurulenta, brown none brown brown stygia none none none none laevigata brown none none orange proliza brown none none none sublaevigata brown none none none bescens brown none none none caroliniana brown yellow brown none tifera red none brown red-brown reticulata red none none none Herreana red none red none cetrarioides var. rubescens| red-brown none none none Finkii none red red none rudecta none red yellow gra soredica none red none none olivacea var. glabra none red none none Borreri none red none none latissima red red none brick-red ABBREVIATIONS The herbaria from which material has been studied and from which specimens are cited in this monograph are indicated by the following abbreviations: BSNH—Boston Society of Natural "cà — Mass. BPI—Bureau of Plant Industry, Washi D. C. D—Private Lichen Herbarium of Dr. C. T Dodge, St. Louis, Mo. F—Farlow Herbarium, Cambridge, Mass. M—Field Museum of Natural History, Chicago, Ill. IA—University of Iowa, Iowa City, Ia. M—University of Michigan, Ann Arbor, Mich. MIN—University of Minnesota, St. Paul, Minn. MBG—Missouri Botanical Garden, St. Louis, Mo. 1941] BERRY— THE GENUS PARMELIA IN NORTH AMERICA 41 MON'T—University of Montana, Missoula, Mont. NYBG—New York Botanical Garden, New York, N. Y. LSU—Stanford University, Palo Alto, Calif. O—University of Ohio, Columbus, Ohio. R—Russell Herbarium of Buffalo Soeiety of Natural History, Buffalo, N. Y. T—Tuekerman Herbarium, Harvard University, Cambridge, Mass. US—United States National Herbarium, Washington, D. C. W—Wellesley College, Wellesley, Mass. WIS—-University of Wisconsin, Madison, Wis. TAXONOMY Parmelia Acharius, Meth. Lich. 153. 1803; Lichenogr. Univ. 89, 456. 1810; Syn. Meth. Lich. 195. 1814; Rabenh., Prodr. FI. Neomarch. 299. 1804; Rohl., Deut. Fl. 3, Abt. 2: 18. 1813; Mart., Fl. Crypt. Erlangens, 205. 1817; S. Gray, Nat. Arr. Brit. Pl. 1: 436. 1821; Hook., Fl. Scot. 2: 52. 1821; Fée, Essai Crypt. Ecor. Offiein. 121, pl. 62. 1824; Fries, Syst. Orb. Veget. pars 1: 241. 1825; Lichenogr. Eur. Reform. 56. 1831; Duby, Bot. Gallie. 2: 600. 1830; Eaton, Man. Bot. N. Amer. ed. 6, 56. 1833; Wallr. Fl. Crypt. German. 3: 442. 1831; Taylor apud Mack. Fl. Hibern. 2: 141. 1836; Rabenh., Deutschl. Krypt. Fl. Rar. 2: 54. 1845; Tuck., Enum. N. Amer. Lich. 49. 1845; De Not., Mem. Acead. Sei. Torino, П. 10: 378. 1849; Schaer., Enum. Crit. Lich. Europ. 33. 1850; Nyl., Mem. Soc. Nat. Cherbourg 3: 174. 1855; Syn. Lich. 1: 375. 1860; Mudd, Man. Brit. Lich. 92. 1861; Müller-Argau, Mem. Soc. Phys. & d'Hist. Nat. Genéve 16: 372. 1862; Kickx, Fl. Crypt. Flanders 1: 218. 1867; Th. Fries, Lichenogr. Scand. 1: 111. 1871; Leighton, Lich. Fl. Gt. Brit. 122. 1871; Sydow, Flecht. Deutschl. 32. 1887; Wainio, Étude Lich. Brésil 1: 27. 1890; Crombie, Monogr. Lich. Brit. 1: 232. 1894; Calkins, Lich. ЕІ. Chicago & Vic. 18. 1896; Oliver, Expos. Lich. Ouest Fr. 1: 107. 1897; Schneid., Textbook Gen. Lichenol. 167. 1897; Zahlbr. apud Engl. & Prantl, Nat. Pflanzenfam. 1 Teil, Abt. 1: 211. 1907, and ed. 2, 8: 233. 1926; Fink, Contrib. U. S. Nat. Herb. 14: 190. 1910; Lich. Fl. U. S. 320. 1935; Herre, Proc. Wash. Acad. Sci. 12: 193. 1910; A. L. Smith, Monogr. Brit. Lich. 1: 118. 1918; Handbook Brit. Lich. 35. 1921; Dodge & Baker, Ann. Mo. Bot. Gard. 25: 589. 1938. [Vor. 28 42 ANNALS OF THE MISSOURI BOTANICAL GARDEN Imbricaria Acharius, Kgl. Svenska Vetensk. Akad. Nya Handl. 15: 250. 1794; Michaux, Fl. Bor.-Amer. 322. 1803; DC. apud Lam. & DC., Fl. Fr. ed. 3, 2: 385. 1815; Kickx, Fl. Crypt. Louvain, 72. 1835; Korb, Lich. Germ. Spec. 8. 1846; Krmphbr., Geschichte d. Lich. 2: 49. 1869; Arnold, Lich.-Fl. München. Ber. Bayr. Bot. Gesells. 1: 26. 1891. Parmelia sect. Lobaria Sprengel, Fl. Halensis, 332. 1806. Physcia S. Gray, Nat. Arr. Brit. Pl. 1: 455. 1821. Parmelia sect. Imbricaria Fries, Syst. Orb. Veget. pars 1: 242. 1825; Tuck. Enum. №. Amer. Lich. 49. 1845; Schaer., Flora 32: 291, 296. 1849. Parmelia Lobaria sect. Imbricaria Endlicher, Gen. Pl. 15. 1836. Parmelia sect. Parmelia Hooker, Handbook New Zealand РІ. 571. 1867. Parmelia subgen. Euparmelia Harmand, Bull. Вос. Sci. Nancy, II. 31: 216. [1896] 1897. The type of the genus is Parmelia saxatilis (L.) Ach., Meth. Lich. 204—205. 1803. Thallus foliose, appressed or ascending, laciniate, with rounded, linear, or filiform lobes, often imbricate; upper sur- face often sorediose or isidiose, lower surface either lighter or darker than the upper, usually covered with rhizoids except at the margins; upper and lower cortex of vertical hyphae woven into plectenchyma; rhizinae arising from a fusion of hyphae of the lower cortex; medulla of loosely woven peri- clinal hyphae; algae Protococcus; apothecium on upper sur- face never marginal, sessile or short-pedicellate, disk concave or flat, chestnut-brown, amphithecium prominent, hypothecium hyaline with algae below; paraphyses imbedded in a gell, usu- ally branched and septate, clavate or pointed at the tips; asci clavate, usually 8-spored, ascospores hyaline and unicellular, ellipsoid or spherical; spermagonia immersed in surface of thallus or amphithecium, spherical or pyriform, opening by black ostioles, wall black or dark brown above, light brown or hyaline below, spermatophores simple or branched, spermatia oblong with acute tips or sometimes fusiform. 1941] BERRY—THE GENUS PARMELIA IN NORTH AMERICA 43 KEY TO THE SUBGENERA OF PARMELIA A. Always with thick rough lower eortex; rhizinae never present. B. Thallus with oval perforations; spores 2-4, large. ............. Menegazzia B. Thallus never with oval perforations; spores 8, small............ Hypogymnia A. Never with thick rough lower cortex; rhizinae always present. B. Thallus foliose, lobes never сһаппејей......................... Euparmelia B. Thallus always fruticose, lobes channeled..................... Pseudevernia MENEGAZZIA I. Parmelia subgenus Menegazzia Wainio, Étude Lich. Bresil 1: 27. 1890. Menegazzia Mass., Neag. Lich. 3. 1854. The type used for the subgenus by Wainio is Parmelia pertusa. Thallus always without rhizinae, attached by mucilaginous secretions; lower part of medulla composed of large dark brown or black hyphae, upper part, of small hyaline hyphae; upper cortex always perforate; ascus clavate, spores 2-4, large, hyaline. The small number of large spores and distinctive medullary hyphae make this subgenus very different from the remainder of the Parmeliae. It is probable that the Menegaeziae are not closely related to the remainder of the genus. The Menegazziae are represented in North America by one species, Parmelia pertusa (Haller) Schaer., whose distribution is in the northern or mountainous areas. 1. Parmelia pertusa (Haller) Schaer., Lich. Helvet. Spic., sect. 10:457. 1840. Lichen pertusus Haller, Hist. Pl. Helvet. 2012. 1769; Schrank. Baier. Fl. 2: 519. 1789. Lobaria terebrata Hoffm., Deutschl. Fl. 2: 151. 1790. Lichen diatrypus Ach., Lichenogr. Suec. Prodr. 116. 1798. Imbricaria diatrypa DC. apud Lam. & DC., Fl. Fr. ed. 3, 2: 393. 1815. Schaerer’s type is ‘Lichenes Helvetici Exsiccati,’ 365, in Herbier Boissier, Université de Genéve, Switzerland, the type collection having been made in the Swiss Alps. Schaerer re- ferred to Schrank's Lichen pertusus (Baierische Flora 2: 519. [Vor. 28 44 ANNALS OF THE MISSOURI BOTANICAL GARDEN 1789) but Schrank used as his type the description of Lichen pertusus Haller (Historia Plantarum Helvetiae, 2012. 1769). Hoffmann cites Schrank's Lichen pertusus as synonymous with his plant. Acharius considered Lichen diatrypus (Li- chenographiae Suecicae Prodromus, 116. 1798) the same as Sehrank's Lichen pertusus. DeCandolle cites Lichen diatrypus Ach. as synonymous with his Imbricaria diatrypa. Thallus glossy, ash-gray to greenish-brown above, closely adnate, sparingly perforated with round to oblong holes, in- flated, lobes crowded, imbricated, with scattered confluent round white soredia on the rounded margins; upper and lower cortex easily separated, often pulling away and leaving the white medulla exposed, black below, brown at the margins, wrinkled ; apothecia rare, 2-9 mm. diameter, sessile, disks con- cave, chestnut-brown; K, yellow, C, none, K + C, yellow, P, none. Algal layer continuous, 25 и thick; upper cortex 12.76-14 и thick, lower cortex 15-20 и thick; medullary hyphae Ти diam- eter, a web of heavy dark brown hyphae 3.19 и in diameter ex- tending from the lower part of the medulla often through cracks in the lower cortex, the line of cleavage in the inflation of the thallus being at the junction of the small hyaline hyphae and the heavy dark hyphae; thallus 600-650 и thick; thecium 70-90 thick; asci clavate, spores 2-4, hyaline, ellipsoid, 20-30 x 50-60 и; spermatia straight, cylindrical, spermagonia sub- spherical, ostioles not raised on the surface of the thallus. Distribution: Nova Scotia to Maryland, west to British Co- lumbia, south to Washington. Nova Scotia: on tree trunks, Bridgewater, Aug. 10, 1910, Macoun 237 (MBG); on spruces, near Digby, May 20, 1910, Macoun (F); Rothrock (FM). NEW BRUNSWICK: on trunks, Yarmouth, June 6, 1910, Macoun (MBG). QUEBEC: on trees, Montmorency Falls, June 24, 1905, Macoun 13 (F). MAINE: KNOX 00.: on trunks of spruces, Mount Battie, Camden, May 1, 1910, Merrill (F). NEW HAMPSHIRE: GRAFTON CO.: Franconia Notch, Oct. 14, 1884, Faxon (F). coos CO.: on rocks, White Mountains, Calkins (FM); White Mountains, Willey (MIN). NT: ADDISON 00. :оп cedar tree, Seandon Swamp, Leicester, May 22, 1920, Dutton 1012 (MBG,FM,WIS). ORLEANS CO.: outlet Willoughby Lake, June 26, 1884, Faxon (WIS). 1941] BERRY— THE GENUS PARMELIA IN NORTH AMERICA 45 MASSACHUSETTS: BRISTOL CO.: trunks and rocks, New Bedford, Willey 59 (FM,MBG,R) N : HAMILTON CO.: Indian Lake, Peck 10 (R). WASHINGTON CO.: on bark, Archdale, Nov. 1871, Clinton (R). PENNSYLVANIA: LANCASTER CO.: on trunks, Bear Town, Ебу (MBG). MARYLAND: ANN ARUNDEL CO.: on holly at Boiling Spring, Dee. 12, 1905, Plitt (ВРІ). WEST VIRGINIA: POCAHONTAS CO.: Brown's Creek Road, Huntersville, Feb. 6, 1930, Sd 11880 (NYBG). MIN TA: COOK CO.: on cedars in swamp, Grand Marais, Jan. 29, 1902, Fink 5240 (XIN J: BRITISH COLUMBIA: on limbs of trees, Vancouver Island, May 5, 1909, Macoun 14.09 WASHINGTON: Washington Territory, Davis аи CHEHALIS СО.: on trunks of trees, Aberdeen, Apr. 1909, Foster 35 (F,MIN); on trees, Montesano, June 1919, Grant 3030 (F). KING CO.: on Cascara, Wynaco, m Grant (FM). FOREIGN AND LocaL ExsICCATAE: Claud. 19; Cromb. 145; Erb. I, ; Flag. Lich. Fr.-C. 17; Hav. Oecid. 18; Larb. Caes. 66; Маг. 225; "p A М 35; Nyl. Par. 32; Pieq. 74; Rab. Cent. 7; Schaer. 365; Trev. 156 252. KEY TO SPECIES IN SUBGENUS HYPOGYMNIA A. Lobes always inflated. B. Medullary hyphae all of one type; spores spherical, 3 x 3 or 4 x 4 y. Jab o co MAE EIS DEUS СЕСК ЕЛО А 4. Dia B. Medullary hyphae of two types. C. Usually sorediate; spores 4—6 x 6-8 и; K, yellow, then red, C, none EUREN EU UU 9. physodes C. Rarely sorediate; spores 4—6 x 7-10 и; К, greenish-yellow, C, none. EIDEM RO DU ANT OTT OR OG OM ECR ORE TOR MCI RETO 3. en ода A. Lobos never inflated; K; none, ©, n0n8.... 50504 еее 5. encausta HYPOGYMNIA II. Parmelia subgenus Hypogymnia Nyl, Flora 64: 537. 1881. Nylander's type for this subgenus is Lichen physodes L. Thallus gray, yellowish-brown, or black; under-side black, roughened and without rhizinae ; medulla often with a cavity; marginal or terminal soredia common except in Parmelia enteromorpha. 'The distribution of the subgenus is world-wide in the colder or mountainous regions. 2. Parmelia physodes (L.) Ach., Meth. Lich. 250. 1803. Lichen physodes Linn., Sp. Pl. 1144. 1753. Lichen laciniatus var. physodes Weis, Pl. Crypt. Fl. Goetting. 57. 1770. [Vor. 28 46 ANNALS OF THE MISSOURI BOTANICAL GARDEN Lobaria physodes Hoffm., Deutschl. Fl. 2: 150. 1795. Imbricaria physodes DC. apud Lam. et DC., Fl. Fr. ed. 3, 2: 393. 1815. Parmelia physodes var. vulgaris Th. Fries, Lichenogr. Scand. 1: 117. 1871. Menegazeia physodes Navas, Lich. de Aragon, 25, 1908. The type of the species used by Acharius is based on a col- lection in the Herbarium of Linnaeus. He cites as synonymous Jacquin ‘Collectanea ad Botanicam’ 3: 1.8, f. 2£ 3. 1789, Hoff- mann, ‘Enumeratio Lichenum,’ t. 15, f. 2a 6 b. 1784, and Dil- lenius, ‘Historia Muscorum,' t. 20, f. 49. 1741. The collection of the type species was from trees or rocks in northern Europe. The type of F. W. Weis, given in ‘Plantae Cryptogamicae Florae Goettingensis,’ 57. 1770 is based on t. 20, f. 49 in Dillenius, ‘Historia Muscorum,’ Acharius, ‘Lichenum Sue- cicae Prodromus,’ 115. 1798, and Wulfen apud Jacquin ‘Col- lectanea’ 3: t. 8, f. 2 Ф 3. Wallroth considered his Parmelia ceratophylla var. phyllodes the same as Dillenius’ plant. Thallus middle-sized to large, irregular in outline, 5-15 cm. diameter, loosely adnate, lobes inflated, long, narrow, di- ehotomously branched, imbrieated, usually with marginal soredia, often with slightly raised tips; upper surface smooth, greenish-gray to whitish, margins often brownish or black, lower surface black, without rhizinae, roughened and much wrinkled; cortex continuous except at junction of lower and upper cortex, where it is often broken exposing the white me- dulla; apothecium sessile or short-pedicellate, 2-15 mm. di- ameter, disk concave, chestnut-brown, margin of the amphi- thecium entire, smooth; K, yellow, then red, C, none, K + C, brown, P, red. Algal layer continuous or cells grouped, 10-15 y thick; upper cortex 12-18 и thick, cells of the vertical hyphae small and hyaline but separated by a dark brown matrix, giving the ap- pearance of a brown net; medulla of loosely interwoven hyphae of two types, (1) those on the upper side hyaline and of small diameter, (2) those on the lower side dark brown and of large diameter, the inflation of the thallus occurring at the junction of the two types ; lower cortex 12—20 и thick, matrix dark brown 1941] BERRY—-THE GENUS PARMELIA IN NORTH AMERICA 4T or black; thickness of the thallus extremely variable, 30-200 и; thecium 30-36 и thick ; ascus clavate or almost spherical, spores 8, 4-6 x 6-8 и; paraphyses branching once or twice, tips slightly enlarged. Distribution: Labrador to northern Alabama, west to Brit- ish Columbia, south to California. LABRADOR: on earth, Forteau, Sept. 19, 1889, Waghorne 69 (MIN) ; on branches of spruee, Aug. 23, 1896, Low (F EWFOUNDLAND: on trees, Goris Cove, Bay of Tiland, Sept. 15, 1856, Waghorne (MBG) ; on rocks, Hodges Cove, Trinity Bay, June 29, 1885, Waghorne 10 (MBG); Chance Cove, Trinity Bay, Мау 24, 1893, Waghorne (МВС); Chance Cove, Trinity Bay, Мау 26, 1893, Waghorne (MIN); on trees, ‘Вау of Bull Coeur,"" Мау 28, 1893, Waghorne (MIN) ; on trees, New Haven, June 21, 1893, Waghorne 32 (MIN); on gravelly sand, July 27, 1893, Waghorne (MIN) ; on trees, New Harbor, Sept. 21, 1895, Waghorne 69 (MIN); on rocks, Hodges Cove, Random, Mar. 22, 1895, Wag- horne (MBG); on trees, Cold Cove, Bay of Hand, Sept. 15, 1896, Wi: 69 (MIN). Nova Scotia: Rothrock (FM). New Brunswick: Little Branch, Miramichi, June 15, 1897, Fowler (F). QuEBEC: alluvial woods along Riviere Cap Chat, July 6, 1923, Collins $ Dodge (D); on rocks covered by mould, Ste. Anne, May 11, 1935, Lepage 143 (D); on rotten logs, Ste. Anne, May 30, 1935, Lepage 480 (D) ; dry peaty knowls and slopes by Lae Pierre, elev. about 1100 m., Table Top Mountain, Aug. 6, 1923, Fernald, Dodge § Smith (D) NTARIO: on rocks and trees, Ottawa, Apr. 26, 1891, Macoun (МВО); on trunks, Ottawa, May 16, 1903, Macoun 242 (MBG); on rocks and trees, Ottawa, Apr. 26, 1891, Macoun 77 (MIN,WIS) ; on fences and tree trunks, Apr. 24, 1892, Macoun (MBG) ; Rainy Lake, Aug. 20, 1930, Schmidt (MBG) ; on trees, Bao, July 18, 1901, Fink 719 (MIN AINE: JASEINOTON co.: June "br Markin (WIS). HANCOCK CO.: Winter E Apr. 30, 1923, Norton (F); Mount Desert, July 1928, Markin (WIS). OX CO.: on rocks, Mount Battie, Merrill (F). DIS HAMPSHIRE: CARROLL CO.: Chocorua, Aug.-Sept. ex Farlow 479b (F,WIS). GRAFTON с0.: on coniferous trees, Plymouth, Mar. 1, 1891, Cummings 9b (WIS,MBG) ; on coniferous trees, North Woodstock, Sonidos Чек July 27, 1891, Cummings 9a (WIS,MBG). SULLIVAN CO.: spruce and hemlock woods, elev. 1460 ft., Lempster, Aug. 1927, оне (ЕМ). VERMONT: WINDHAM 00.: Grout Pond, Stratton, July 4, 1925, Moore (MBG). ADDISON со: Middlebury, Apr. ote Dales 152 (D); on roots of Pinus Strobus, Middlebury, May 7, 1913, Dodge (D); Green Mountain Chain beyond Breadloaf, Ripton Hancock Road, ш 29, 1916, Dodge (D); on cedar trees, Leicester Brook Swamp, elev. 350 ft., Feb. 2, 1920, Dutton 966 (F); Bristol Foie June 24, 1927, Dodge (D); BIET OG Bop June 25, 1927, Dodge (D) ; Monkston, June 25, 1927, Dodge (D). RUTLAND CO randon, March 12, 1911, Рено» (MBG); on eedar trees in swamp, Bra NM a 26, 1921, Dutton 1216 (WIS,FM) ; Chipman Lake, Tinmouth, June 1938, Dodge (D). [Vor. 28 48 ANNALS OF THE MISSOURI BOTANICAL GARDEN MASSACHUSETTS: MIDDLESEX CO.: Pepperell Station, Oct. 1926, Dodge (D). RISTOL CO.: trees and rocks, New Bedford, Hall 38 (FM). WORCESTER со.: 1885, ври 424 (Е CONNECTICUT: WINDHAM CO.: on tree, Pomfret, Oct. 29, 1925, Evans 605 (Е). NEW YORK: ESSEX CO.: on mosses near Lake Plaeid, Sept. 1-16, 1898, Britton (Е); on spruce, Lake Placid, Sept. 27, 1914, Humphrey (WIS); on bein, Mount Marcy, near Lake Placid, elev. 5300 ft., Aug. 16, 1933, Lowe 2830 (Е). YATES CO.: Dundee, S. Wright (FM) ; on dead twigs of balsam firm, Dec. 31, 1894, Galway (F). Dundee, S. Wright (FM); on dead twigs of balsam fir, Dee. 31, 1894, Galway (Е). NoRTH CAROLINA: WATAUGA СО.: on bark of trees, Grandfather Mountain, June 16, 1923, ~~ (Е); SWAIN 00.: deep wet woods, Clingmans Dome, elev. 660 ft., «4 25, нетет В554 (МВС); woods, elev. 5500 ". Newfound Gap, des. 9 ety Tri t (MB dae MAHONING CO.: on old pad Ellsworth, Apr. 1910, Vickers 39 (BPI). WEST Fons POCAHONTAS CO.: on dead spruce, Bald Knob, July 19, Gray (Е); MERCER СО.: on roof of shed near Coopers Rock, July 28, 1907, Sheldon 1056 (F). MICHIGAN: CHEBOYGAN СО.: Mud Lake Bog, pepe of Burtt and Douglas lakes, June-Aug. 1923, Nichols (F). KEWEENAW CO. onifer, MeCargo Cove, Isle Royale, Aug. 4, 1930, Lowe 314 (F); Isle Pree а oak 10, 1901, Stuntz $ Allen (WIS); on birch, Rock Harbor, Isle Royale, Aug. 1904, Harper $ Harper 173 (FM INDIANA: COUNTY UNKNOWN: on trees, 1827, Boke mma (MBG). TENNESSEE: HAMILTON СО.: on oaks, t Mountain, Calkins 288 (FM); SEVIER CO.: deep wet woods, Clingmans Sa pm 6600 ft., Aug. 25, 1937, Hubricht (MBG); woods, elev. 5500 ft, Newfound Gap, Aug. 25, 1937, Hubricht В554 (MBG). WISCONSIN: VILAS CO.: valley of the Wisconsin River, near Lae Vieux Desert, June 13, 1893, Cheney 67 (WIS) ; Lost Creek, near Sayner, Aug. 30, 1938, J. Thom son (WIS). TAYLOR со.: valley of the Wisconsin River near Goodrich, summer 1894, Cheney (WIS). FOREST CO.: on soil east of Three Lakes, May 30, 1938, Wolf 6 (WIS). ILLINOIS: eounty unknown: Calkins (FM). MINNESOTA: LAKE СО.: on trees, Snowbank Lake area, July 21, 1897, Fink 863 (MIN). вт. LOUIS CO.: on vid in swamps, Harding, Aug. 17, 1901, Fink 1553 (MIN); on roeks, Tower, Aug. 24, 1901, Fink 1720 (MIN). CARLTON CO.: on trees, Kettle Falls, Aug. 12, 1901, Fink 1402 2 (MIN); on trees, Tafte, Carlton Peak, July 10, 1897, Fink 565 (MIN). KOOCHICHING CO.: on trees and rocks, Gunflint, July 2, 1897, Fink 383 (MIN); on trees, Koochiching, July 27, 1901, Fink 965 (MIN » on rocks, Koochiching, July 31, 1901, Fink 1050 (MIN); on rocks, Rainy Lake City, Aug. 1, 1901, Fink 1099 (MIN) ; on trees, Rainy Lake City, Aug. 7, 1901, Fink 1267 (MIN). BELTRAMI СО.: on tamaraek in swamp, Red Lake, July 27, 1900, Fink 913 (MIN) ; on pines, Bemidji, July 4, 1900, Fink 429 (MIN); in swamps, Beaudette, June 18, 1901, Fink 28 (МТМ); on eg Red Lake, a 3, 1900, Fink 1071 (MIN); on trees in swamps, Beaudette, June 18, 1901, Fink 6 (MIN); on me ges July 12, 1900, Fink 669 (MIN); on акы. and red cedar in a tamaraek swamp, Bemidji, July 9, 1900, Fink 514 (MIN). CASS co.: on trees and eel Grand Portage Island, Jui 23, 1897, Fink 143 (MIN). OTTERTAIL CO.: on tamaraek in 1941] BERRY—THE GENUS PARMELIA IN NORTH AMERICA 49 swamps, Henning, June 28, 1900, Fink 331 (MIN). ROSEAU CO. on tamarack in swamp, Warroad, June 28, 1901, Fink 246 (MIN). Зоотн DAKOTA: COUNTY INDEFINITE: Black Hills, Macbride (IA). MONTANA: CLARK CO.: on live байх, S. 11. T. 12 №. В 9 W., Jefferson Creek, Helena, Aug. 15, 1931, Flint (MONT). MISSOULA CO.: north slope of Mitten Moun tain, Dec. 13, 1932, Flint, Jr. (MONT). COUNTY INDEFINITE: Douglas fir bark, both dead and green trees, seet. 19. T. 12 N. R 6 W., elev., 5000 ft., May 28, 1931, Flint MONT New MEXICO: COLFAX CO.: on Pseudotsuga mucronata, vicinity of Ute Park, elev, 2200-2900 m., Aug. 21, 1916, Standley (Е). ALBERTA: Banff, woods, Oet. 1910, Swanson 354 (F); Sulphur Mountain, Oct. 18, 1918, Swanson 872 (Е IDA : by the highway near Locha River west of Lolo Pass, Sept. 15, 1937, Bary pu (MON T). sH CoLUMBIA: Ucluelet, on old logs and rocks, May 11, 1909, Macoun Tuum Lilloeet, July 1916, Macoun (F); New Westminster, Jan. 1903, Merrill (F). WASHINGTON: SPOKANE CO.: on bark of Pinus ponderosa, 5 miles southwest of Spokane, May 30, 1933, Flint (MONT). COLUMBIA CO.: on branches, Blue Moun- tains, Oct. 17, 1831, Howard 1620а (W). PIERCE CO.: on trees, Longmires Springs, Aug. 1906 Dia $ Harper 255 (FM); Sumner, June 1906, Harper $ Harper 254 (FM). ISLAND CO.: on log, Langley, Grant (FM); on trees, Langley, 1926, Grant (WIS) ; on trees, Langley, 1923, Grant (W); on rock, Langley, June 1923, Grant (W); Goose Rock, July 1923, Grant (W). SAN JUAN CO.: San Juan Island, July 1906, Harper § Harper (FM). CLALLAM CO.: on roof, Sequim, 1915, Grant 213 (MBG). CALIFORNIA: MARIN CO.: on trees, Mar. 29, 1893, Howe 25 (МВС); COUNTY IN- DEFINITE: on limbs of Manzanita bush, Cuyamaea Mountains, elev. 4500 ft., 1932, Cota MG s ND FOREIGN EXSICCATAE ry Cum. I 9a, b; Tuck. II-IV 72; P 52208 а Norrl, & Nyl. This species is distributed in the cooler and mountainous regions of both hemispheres. It commonly grows on rocks, but may be found on tree trunks and more rarely on earth. In North America, collections have been made from Labrador to North Carolina on the east coast, and from Alaska to central California on the west coast. Collections in the interior of the country, except in the mountains, have not been made south of Illinois. Workers with North American lichens have often rec- ognized Parmelia physodes var. labrosa Ach., Parmelia phy- sodes var. obscura Ach., Parmelia physodes var. platyphylla Ach. and Parmelia physodes var. vittata Ach. as distinct. These forms seem to be valid for European material, but not for that from North America. The plants from North America [Vor. 28 50 ANNALS OF THE MISSOURI BOTANICAL GARDEN bearing these names are environmental variants rather than distinet varieties of the species P. physodes Ach. 3. Parmelia enteromorpha Ach., Meth. Lich. 252. 1803. Lichen intestinalis Sm. apud Ach., Meth. Lich. 252. 1803. Parmelia physodes var. enteromorpha Tuck., Proc. Amer. Acad. Arts & Sci. 1: 220. 1848. Hypogymnia enteromorpha Nyl., Acta Soc. Sci. Fenn. 261°: 7. 1900. Imbricaria enteromorpha Jatta, Nuov. Giorn. Bot. Ital. N. S. 9: 471. 1902. Acharius’ type was Menzies’ collection from the west coast of North America, near Monterey, California. The type is now in Acharius’ Herbarium in Helsinki, Finland. Acharius cites as a synonym of his species J. E. Smith's Lichen intestinalis from Menzies’ type. Tuckerman used as the type of his variety enteromorpha a duplicate collection of Menzies’ which is prob- ably co-type material. The type of the variety is in Tucker- man’s Herbarium in the Farlow, at Harvard University. Nylander’s and Jatta’s synonyms were based on Menzies’ col- lection and arose from the renaming of the genus. Thallus large, loosely attached; upper surface green, dingy gray or brownish, smooth, with numerous black dots which are the open ostioles of the spermagonium; lobes numerous, long, narrow, inflated or occasionally broad at the tip; lower surface black, rough; lower cortex often cracked or broken, exposing the white medulla; apothecium abundant, short-pedicellate, 5-8 mm. diameter, disk concave, flat or sometimes convex, oc- casionally perforate at the center, pedicel always hollow, amphithecium entire or lobed; K, yellow or greenish-yellow, C, none, К + C, yellow, P, yellow. Algal layer continuous, 12-15 и thick; upper cortex 9-19 и thick ; medulla loosely interwoven with a central cavity, hyphae of two types, (1) large, brown, 2-3 y in diameter, and (2) hya- line, Тр or less in diameter, the large dark brown hyphae being on the lower side of the thallus and an inflation occurring at the junetion of the two types; lower cortex 24—27 y thick, matrix brown; thecium 21-25 и thick; ascus clavate, spores 8, 46 x 7—10 и; paraphyses branched. 1941] BERRY—THE GENUS PARMELIA IN NORTH AMERICA 51 Distribution: Labrador to Virginia, west to Alaska, south to California. LABRADOR: on earth, Oct. 18, 1894, Waghorne (MIN). NEWFOUNDLAND: on sand, July 27, 1893, Waghorne E ; on earth, Sept. 28, 1894, Waghorne UE on earth, Bay Is pex. PA 6, Waghorne (MBG) ; a cliff, Goose Arm, Bay of Hand, Sept. 2 ia: (MBG); on earth, ard 28, 1899, ee (US). MAINE: HANCOCK CO.: Beech Mountain, July 8, 1894, Merrill 89 (WIS). KNOX co.: on spruces, Camden, 1910, Merrill 89 (WIS). YORK со.: Saco, 1863, Blake New HAMPSHIRE: coos co.: White Mountains, 1863, Mann (FM); base of Mount Washington, Aug. 6, 1895, Harper $ Harper 35 (FM). GRAFTON CO.: coniferous trees, North сабо Franconia Mountains, July 27, 1891, Cummings VERMONT: LAMOILLE CO.: on twigs, Mount Mansfield, Aug. 11, 1851, Russell (R). VIRGINIA: WASHINGTON CO.: summit of White Top Mountain, elev. 5678 ft., May 28, 1892, Small (MIN NORTH CAROLINA: MITCHELL CO.: on trees, Roan Mountain, June 15, 1936, Schal- lert (B). MONTANA: FLATHEAD CO.: 2% miles up Glen Trail east from Flathead es Nov. 18, 1934, Flint (MONT); Columbia Falls, Sept. 26, 1892, Williams 2b (M MISSOULA CO.: dead twigs of conifer, north slope of Mitten Mountain, near и soula, elev. 3500 ft., Dec. 1930, Тш (MONT); on dead twigs of Douglas fir, Kitchen Gulch near Clinton , Nov. 25, 1931, ess Mp ; north slope of Mitten Mountain, elev. 3700 ft., near is Flint (MONT). WYOMING: on dead [ue of Pinus contorta, ORE River drainage of Yellow- stone за elev. 7000 ft., July 20, 1931, Flint (MONT). IDA EZ PERCE CO.: above Lake desc elev. 2000—3500 ft., ‚ June 3, 1896, EUM. à Eus Cas KOOTENAI CO.: on er branches of P ея head of Little Lost Fork, Big Creek, Coeur Pe ae Aug. 30, 1931, Wnt (M MACKENZIE TERRITORY: Arctic Ocean, Elephant Point, Eschscholtz = a 1880, Bean (FM). YUKON: on trunks in woods, The Cone, 1902, Macoun (MBG). dicet on hemlocks, Windham Bay, elev. 1700 ft., Oct. 4, 1905, Culbertson 126 (FM,F) ; Cooks Inlet, Fort Alexander, 1880, Bean (FM BRITISH COLUMBIA: Vancouver Island, Oregon Boundary Commission, 1858-59, Lyall (R,WIS); on trees and rotten logs, Victoria, Vancouver Island, May 1887, Macoun 69 (MIN,WIS) ; on trees, Donald, July 6, 1885, Macoun (MBG); on old wharf of Comox, June 23, 1893, Macoun (US); on ie of trees and bushes, Victoria, Vancouver Island, TY 23, 1887, Macoun (M ; Surrey, Hill (Е); near MeLeod Lake, June 27, 1875, Macoun (F); E мї of North Ameriea, WASHINGTON: SNOHOMISH CO.: Snohomish, Harper $ Harper 277 (FM). PEND D'OREILLE CO.: Pend d'Oreille, Oregon Boundary Commission from Fort Colville to Roeky Vc s 1861, к =. ЗРОКАМЕ CO.: high up in old yellow pine, 6 miles south of Spokane, Apr. 2 та Flint (M pois bark of Pinus ponderosa, 5 miles southwest of Spokane, A , 1939, Flint (MONT). WHITMAN CO.: very abundant on lower dead branches of ax pine, along а River near Palouse, [Vor. 28 52 ANNALS OF THE MISSOURI BOTANICAL GARDEN Aug. 12, 1938, Ownbey $ Ownbey (MBG). STEVENS CO.: on wood, Kettle Point, July 3, 1907, Cowles 318a (FM). PIERCE СО.: on trees, Longmires та Aug. 1906, Harper $ Harper (FM); Ashford, TE 1906, ФЕ и 4 Harper (FM); umner, June 1906, Harper $ Harper (FM). AND CO.: pinos dais 1922, Grant (WIS). SAN JUAN CO.: Rock Harbor, dur 3, bos, Fry ); San Juan Island, July 1906, Harper $ Harper (ЕМ); on wood, Friday Hashes, July 2, 1907, Cowles 318 (FM). CLALLAM CO.: on rocks, Port Angeles, Mar. 5, 1914, Foster 2636 (Е). CHEHALIS CO.: on trees near Aberdeen, Apr. 1909, Foster To COUNTY UNKNOWN: on ee Calkins sae кани е Davis OREGON: WALLOW n branches de a Mountains, EIL Cove and Minam River, e. pum ft., cgi > 1897, ‘Sheldon (FM,US). COUNTY IN- DEFINITE: on rails and trees, Calkins ws ; Coast Range, May 1905, Buhl 6 (FM); Caseade Mountains, Hall 18 (FM). CALIFORNIA: SAN BERNARDINO CO.: pine crest, San Bernardino Mountains, Aug. 1912, Hasse (Е); dead brush of yellow pine, San Bernardino Mountains, elev. 4500 north of Alpine, June 1932, Cota (US); shady dry limbs, Manzanita brush, elev. 4500 ft., 1932, Cota (US) ; Ward Canyon, Feb. 1933, Cota (FM). Los ANGELES CO.: on не, Santa Catalina Island, Nolava Canyon, June 30, 1920, Nuttall 514 (FM); on [Quercus ?] ‘‘Macrocarpa’’ bark, head of Benedict Canyon, Santa Moniea Mountains, elev. 1200 ft., Mar. 14, 1933, Wheeler 1554 (US). SAN ива: co.: Castle Rock Ridge, Aug. 10, 1906, Herre (MBG). MONTEREY CO.: Monterey, 1790, Menzies (probably Co-type) (T); on branches of Pinus kikata, E. Td Point Lobo Reserve, elev. 130 ft., Aug. 1, 1936, Wheeler (MBG). SANTA CLARA CO.: trees and shrubs, Blaek Мова, elev. 1500-2785 ft., July 6, 1903, Herre 122 (MIN,MBG); on trees, Kings Mountain, Purisima Creek, elev. 1900 ft., July 24, 1903, varie 179 (MIN); on fences, 4 miles west of Stanford Ll elev. 400 ft., Oet. 1, 1908, Herre 338 (MIN,MBG). SAN MATEO CO.: abundant on trees and болда Mong summit of range, Kings Mountain, May 19, 1902, м (WIS). SANTA CRUZ CO.: on old fences along summit of range, Santa Cruz Mountains, elev, 2000— 2200 ft, Apr. 5, 1905, Herre 633 (MBG,MIN). GLENN с0.: on trees, Plaskett BG). MARIN CO.: on trees, Маг. 29, 1893, Howe (US,MIN,WIS) ; on trees, Apr. 5, 1893, Howe (F,MBG). couNTY UNKNOWN: on Pinus insignis and oaks, 1864, Bolander (FM,US) ; on rails and trees, Calkins (ЕМ). The plant is entirely western and alpine in its distribution. It is typical when found in the more humid sections of the red- wood forests in California and Oregon. Those plants which come from the arid regions of California are small and de- pauperate. Collections of this plant are distinguishable from P. physodes by the larger size of the narrow lobes, the lack of soredia, and brown rather than gray color. 4. Parmelia lophyrea Ach., Meth. Lich. 198. 1803. Parmelia cribellata 'Tayl., Hook. London Jour. Bot. 6: 164. 1847. 1941] BERRY—THE GENUS PARMELIA IN NORTH AMERICA 53 Acharius’ description was based on Menzies’ collection from near Gray’s Harbor, Washington. The type is in his herbarium at Helsinki, Finland. The type for Taylor's plant was Menzies’ collection, and probably co-type material is now in the Taylor Collection at the Boston Society of Natural History, in Boston. Thallus small, greenish-gray to ashy, loosely adnate to bark of trees and shrubs; upper surface smooth or occasionally with scattered white soredia, lobes long, narrow, branched, margins eut, erenate, tips somewhat inflated; lower surface black and roughened, with reticulate pits or perforations; apothecium sessile, 3-8 mm. diameter, disk concave, reddish-brown ; amphi- thecium entire and not sorediate; K, yellowish-green, C, none, K + C, none, P, none. Algal layer continuous, 6-18 и thick; upper cortex 15-18 и thick, matrix dark brown, cells in the corticellular hyphae small; medulla of loosely interwoven hyphae, tending to form cavities in older portions of the thallus and at the tips of the lobes, hyphae all of one type, 1-3 р in diameter, hyaline; lower cortex 65-68 y thick, cells of hyphae rectangular, octagonal or spherical, 6-9 и in diameter, matrix brown; thecium 21-24 и thick, hypothecium hyaline, 6-8 y thick ; aseus short-cylindrieal, spores 8, spherical, 3 x 3 or 4 x 4 и; paraphyses very small in diameter, branched. Distribution: collected only in the Olympic Mountains of Washington. WASHINGTON: CHEHALIS CO.: Westport, Apr. 1908, Foster 509a (F); Nov. 22, 1904, Foster (Е). CLARKE CO.: Vancouver's Voyage, 1790—1792, Menzies (Co-type collection) (T); northwest coast, 1835-36, Nuttall (T). Menzies’ collection in Taylor’s Herbarium of the Boston Society of Natural History is the same as Tuckerman’s, but it bears the determination of Parmelia cribellata Taylor, of which it is the type. 5. Parmelia encausta (Sm.) Ach., Meth. Lich. 202. 1803. Lichen encaustus Sm., Trans. Linn. бос. London 1: 83. 1791. Imbricaria encausta DC. apud Lam. & DC., Fl. Fr. ed. 3, 2: 394. 1815. Parmelia physodes var. encausta Fries, Lichenogr. Eur. Re- form., 64. 1831. [Vor. 28 04 ANNALS OF THE MISSOURI BOTANICAL GARDEN Smith's type is no longer in existence, but he gives the draw- ing in Trans. Linn. Soc. London 1: t. 4, f. 6. Acharius’ type is in his herbarium in Upsala, Sweden; he cites as synonymous with his plant Smith's Lichen encaustus, Trans. Linn. Soc. London 1: 83. 1791. DeCandolle's type is not in existence, but be gives Smith's Lichen encaustus as a synonym of it. Fries cites Lichen encaustus and also lists Moug. & Nestl., Stirpes Crypt. 353, as his type. Parmelia encausta (Sm.) Ach. is found on rocks in the alpine regions of both North America and Europe. It is distinguished from Parmelia physodes (L.) Ach. by the narrowness of the lobes and the size of the spores. Thallus small, adnate to rocks, lobes long, narrow, dichoto- mously branched, imbricated, tips lighter-colored than the thal- lus, a dense layer of narrow, secondary, dichotomously branched, imbrieated lobes arising from the central portion of the thallus; under surface brown at the margins, becoming dark brown or black at the center, rough ; apothecium sessile, 1-6 mm. diameter; amphithecium with slightly concave crenu- late margin; no chemical reaction with either K, C, or P. Algal layer of scattered gonidia 5—6 и in diameter; upper cortex 9-10 и thick; medulla loosely interwoven, hyphae small and of one type; lower cortex 30—40 y thick; thallus 120-180 и thick ; Веста 45-50 и thick, hypothecium 18-20 y thick; ascus clavate, 8-spored, spores 4-7 x 5-9 y, hyaline; paraphyses branched. The external appearance of the plant is similar to P. cen- trifuga except that it is smaller and darker in color. Distribution: Labrador to North Carolina, west to Alberta and Montana. LABRADOR: on rock, Capstan Island, Aug. 1, 1889, Waghorne 32 (MIN). NEWFOUNDLAND: on rocks, Waghorne (MBG, FM). NEw HAMPSHIRE: COUNTY UNKNOWN: on rocks, Willey (FM). NORTH CAROLINA: FORSYTH CO.: Winston-Salem, May 1, 1926, Schallert (BPI). NTANA: LEWIS & CLARK CO.: bark of Douglas fir, near Helena, Dee. 1931, Flint (MONT). ALBERTA: оп rocks, Mount Fairview, elev. 9000 ft., Aug. 9, 1906, Fink (Е). FOREIGN AND LOCAL ExsICCATAE EXAMINED: Desm. ed. І. ser. І. 1493; ed. П. ser. I. 1593; Erb. I, 119; Funek, I, 374; Harm. 66; Harm. Lich. Loth. 319; Hav. 222; Lojk. Univ. 63; Moug. et Nest. 353; Roum. 40; Stenh. 67; Wartm. 740 1941] BERRY—THE GENUS PARMELIA IN NORTH AMERICA 55 EUPARMELIA IIL Parmelia subgenus Euparmelia Nyl. apud Hue, Revue de Bot. 4: 375. 1885-86. Nylander's type of the subgenus is Parmelia caperata (L.) ch. Thallus foliose, under-surface never with а thick, rough, black, lower cortex, rhizinae always present, either evenly dis- tributed or at the points of contact with the substratum ; me- dulla of loosely interwoven hyphae, never with a cavity or a central strand. The subgenus is world-wide in its distribution. KEY TO SECTIONS OF EUPARMELIA A. Rhizinae small and evenly distributed. Thallus dark some shade ot, тоу. eer TS M elanoparmelia . Thallus A some shade of imr green оо Е Xanthoparmelia А. ие Тат B. Rhizinae ke Ио оо Hypotrachyna B. Rhizinae at points of contact with substratum, margins nude....Amphigymna KEY TO SPECIES IN EUPARMELIA SECT. MELANOPARMELIA A. With eoralloid branchlets or soredia. B. Lobes broad, always wider than long. C. Amphithecium and upper surface hispid.......... 9a. olivacea var. aspidota (2A TwaysBorediate О бо alate dese 11. conspurcata B. Lobes narrow, always longer than wide. C. Marginal lobes fibrillose, cylindrieal................... зу: 6. pubescens ©. в lobes never fibrillose or cylindrical. . Lobes variously branched, imbricated at tips................. 8. prolixa Lobes digitately branched, tips rounded................ 12. exasperata А. ЖЫ О eoralloid branchlets or soredia. B. Lobes broad, always wider than long С. No reaction with either К or C. Пе SDOreS always Grids testes ee RUSSIE ES qu vice. 9. oli То Бротен 16-24 OF more; sow nee ee eS IUIS E 10. multispora C. Chemieal reaetion, K, AM (red. ese. ets 9b. olivacea var. glabra B. Lobes narrow, always longer than wide..............seseeecseees 7. stygia I. MELANOPARMELIA Parmelia subgenus Euparmelia sect. I. Melanoparmelia (Hue) Zahlbr. apud Engl. & Prantl, Nat. Pflanzenfam. 1 Teil, Abt. 1: 212. 1907. Parmelia sect. Melanoparmelia Hue, Nouv. Arch. du Mus. Paris, T, 4:138. 1899 The type of the section is Parmelia stygia (L.) Ach. [Vor. 28 56 ANNALS OF THE MISSOURI BOTANICAL GARDEN Thallus greenish-brown to blackish, under-surface sparingly eovered with rhizinae ; apothecia sessile. 6. Parmelia pubescens (L.) Wainio, Meddel. Зое. Fauna & Fl. Fenn. 14: 22. 1888. Lichen pubescens Linn., Sp. Pl. 88. 1753. Cornicularia pubescens Ach., Lichenogr. Suec. Prodr. 217. 1798. Wainio's type is Lichen pubescens in the Linnean Herba- rium, London, England. The type locality is in the mountains of Europe. Linnaeus based his name on the plant in his her- barium. Acharius' type is in his herbarium at Upsala, Sweden. He cites as synonyms Lichen pubescens Linn. and Lichen lana- tus Wulf. apud Jacq. (Misc. Aust.2: t. 10, }.5. 1778). The lichen has a limited distribution on the mountains of North America and Europe. It has often been misdetermined because of the fibrillose character of the margins of the thallus. Thallus small, loosely attached, somewhat ascending ; upper surface smooth, black, lobes long, round or slightly flattened on the under side, dichotomously branched, much imbricated, those at center of thallus wrinkled; lower surface smooth, lighter in color than the upper surface, rhizinae small; apo- thecium sessile, 0.5-1 mm. diameter, disk flat, dark brown or black, margin of amphithecium smooth; K, brown, C, none, K + C, none, P, none. Algal layer in the center of the lobes, very thin and scattered in the flattened portions at the center of the thallus ; upper eor- tex Зи thick, matrix yellow-brown; medulla loosely interwoven and with a central cavity; lower cortex 6 и thick, lower edge much roughened; thickness of thallus 60-80 и; theeium 55-60 и thick; ascus clavate, 8-spored, spores 4—5 x 4—6 и; paraphyses branched. | Distribution: Labrador, west to Washington, south to Cali- fornia. LABRADOR: on rocks, Bath Harbor, June 25, 1892, Waghorne 71 (MIN); on rock, Early Harbor, July 29, 1892, Waghorne 170 (MIN). NEWFOUNDLAND: on rock, Aug. 25, 1852, Waghorne (MBG). MONTANA: TETON CO.: Observation Mountain, Black Foot Indian Reservation, Aug. 25, 1897, Williams 107 (NYBG); on basalt rocks and occasionally on bark of yellow pine, near Fort Wright, Apr. 8, 1934, Flint (MONT). 1941] BERRY—THE GENUS PARMELIA IN NORTH AMERICA 51 COLORADO: COUNTY UNKNOWN: 10,000 ft. elev., May 1880, Brandegee 115 (BSNH). BRITISH COLUMBIA: on rocks, summit of Mount Benson, July 10, 1893, Macoun (MBG). WASHINGTON: SPOKANE CO.: on basaltie rocks, near Crater Basin, Dee. 1909, Bowser 88 (Е). YAKIMA CO.: Mount Adams, Suksdorf 147 (BSNH). KLICKITAT co.: thin earth over boulder, Goldendale, 1909, Foster 135 (WIS,FM). COUNTY UN- KNOWN: on rocks, Calkins 354 (F) OREGON: COUNTY UNKNOWN: Hall (MBG,FM). CALIFORNIA: TUOLUMNE CO.: rocks at Clawel's resort above Yosemite Valley, elev. 6000 ft., Apr. 1900, Hasse 571 (NYBG). FOREIGN ExsICCATAE EXAMINED: Malme, 405; P. lanata Erb. І 1221; Fellm. 82; Harm. Lich. Loth. 315; Hav. 181; Roum. 558; Wartm. 741; P. pubescens f. minuscula Fellm. $3. 7. Parmelia stygia (L.) Ach., Meth. Lich. 203. 1803. Lichen stygius Linn., Sp. Pl. 1143. 1753; Ach., Lich. Suec. Prodr. 109. 1798. Lobaria stygia Hoffm., Deutschl. F1. 2: 154. 1795. Imbricaria stygia DC. apud Lam. & DC., Fl. Fr. ed. 3, 6: 189. 1815. Acharius’ and Linnaeus’ types are not in either herbaria at present. Acharius (Lich. Suec. Prodr.) cites as synonyms of his plant Lobaria stygia Hoffmann, ‘Enum. Lich.,’ t. 14, f. 2. 1784, and Squamaria stygia Hoffmann, ‘Descr. & Adumbrat. Pl. Lich.’ 2: t. 25, f. 2. 1794. DeCandolle considers Parmelia stygia (L.) Ach., *Meth. Lich.' 203. 1803, and Lobaria stygia Hoffm. ‘Enum. Lich.,’ t. 14, f. 2, Ше same as his plant. The plants figured by Hoffmann came from the alpine regions of Europe, and should be considered as typical. Thallus small, loosely adnate to rock, irregular in outline; upper surface smooth, brown or shining black, lobes long, nar- row, flat, irregularly branched and twisted, often convex, tips raised, upper cortex brown; under surface black, wrinkled, and with many short black rhizinae; apothecium sessile, 2-4 mm. diameter, disk slightly concave or flat, dark brown; amphithe- eium crenulate, slightly lobed; К, none, C, попе, К + С, none, P, none. Algal layer continuous, 15-20 и thick; upper cortex 15-16 и thiek, matrix yellow ; medulla very loosely interwoven, hyphae large and of one type; lower cortex 15-18 и thick ; hypothecium 30-33 и thick; aseus clavate, 8-spored, spores 3-4 x 5-8 y; paraphyses branched and clubbed at the tip. (Vor. 28 58 ANNALS OF THE MISSOURI BOTANICAL GARDEN Distribution: Newfoundland to New York, west to Alberta and Washington. NEWFOUNDLAND: on rock, Venison Trickle, Oct. 15, 1893, Waghorne 17 (MBG) ; on cliff, Lark Harbor, Bay of Hand, Mar. 21, 1890, Момини (MBG); Sparrible Cove, Apr. 20, 1899, Waghorne (MBG). MAINE: PISCATAQUIS CO.: on rocks, Mount Katahdin, July 1856, Blake (FM). NEW HAMPSHIRE: coos co.: White Mountains, alpine region, 1863, Mann (FM); White Mountains, 1885, Farlow 428 (Е); Mount Moriah, Sept. 1885, Farlow (Е); Mount Жайманов, Sept. 1894, Farlow (MBG,FM); on rocks, Mount Lafayette, Hall 41 (FM). VERMONT: LAMOILLE CO.: on ledges, The Nose, region of Mount Mansfield, June 24, 1922, Merrill (F). EW SEX CO.: on roeks, Mount Marcy, near Lake Plaeid, elev. 5300 ft., Aug. 10, таа. Тоше 2601 (Е). ALBERTA: Mount Fairview, elev. 9000 ft., Fink (F). MONTANA: FLATHEAD СО.: Columbia Falls, Apr. 25, 1893, Williams (US). COLORADO: MONTROSE CO.: on rocks, elev. 5400 ft., Naturita, Oct. 1914, Payson (F); BOULDER CO.: on rocks, Mar. 1898, Morgan WASHINGTON: CLARKE CO.: on rocks of Pinnacle Peak, elev. 7200 ft., Sept. 8, 1909, Foster 1055 (F). LOCAL AND FOREIGN EXSICCATAE EXAMINED: Bartl. Dee. IV 7; Desm. ed. I, ser. I 1942, ed. IT, ser. I 1592; E. Fries, 166, 307, 337 ; Funck, II 107; Gar. I 8; Harm. 67; Harm. Lich. Loth. 313; Hav. 182; Leight. 365; ys. 2 Moug. et Nest. 315; Nyl. Lieh. Mont. Dor. 29; Nyl. Pyr. 17 ; Stenh. ?1; Tue The species is found only in the northern and бас regions of the northern hemisphere. It grows only on rocks, and its small size and shining dark brown or black narrow lobes make it distinct in appearance from either P. pubescens or P. prolixa. 8. Parmelia prolixa (Ach.) Róhl., Deut. Fl. 3, Abt. 2: 100. 13. Parmelia olivacea var. prolixa Ach., Meth. Lich. 214. 1803. Rohling’s type is in Acharius’ Бакі and is the plant used by Acharius as the type of his Parmelia olivacea var. prolixa. The type collection was made from a plant growing in the mountains of northern Europe. Thallus small, loosely adnate; upper surface wrinkled and with coralloid branchlets, brown or blackish-brown, lobes short, much divided, imbricated and dissected, tips slightly inflated ; under surface black with many short, black, flat rhizinae; apo- thecium sessile, 2-5 mm. diameter, disk flat or slightly concave, brown, margin crinkled, finely lobed; K, upper cortex yellow- green, medulla brown, C, none, K - C, none, P, none. 1941] BERRY—THE GENUS PARMELIA IN NORTH AMERICA 59 Algal layer continuous, 10-15 и thick; upper cortex 21-22 и thick; medulla loosely interwoven, hyphae large and of one type; lower cortex 20-22 y thick; thallus 350—400 и thick; the- cium 60-64 и thick; ascus clavate, 8-spored, spores spherical, 3x3or4x4y; paraphyses branched Distribution: Labrador to West Virginia, west to Oregon. LABRADOR: Packs Harbor, Hamilton Inlet, July 15, 1882, Waghorne 13 (MBG,MIN) NEWFOUNDLA on rock, sand beach, White Bay, Oct. 5, 1891, Waghorne 355 уста ; on m Sparrible Cove, White Bay, Apr. 20, 1894, Waghorne (MIN). UNSWICK: оп rocks, July 1879, May (FM). eg PISCATAQUIS CO.: on rocks in woods, side of Mount Katahdin, 1863, Blake (FM). кмох co.: Rockland, Sept. 20, 1915, Merrill (Е). NEW YORK: ESSEX CO.: on beech trees, elev. 1200 ft., June 20, таи Merrill (Е). WEST VIRGINIA: POCAHONTAS CO.: Dunmore, 1924, dray L309 MINNESOTA: LAKE CO.: on rocks, uber ads Lake area, July i 1897, Fink 832 (MIN); on rocks, Beaver Bay, July 13, 1897, Fink 703 (MIN). CARLTON CO.: on rocks, Carlton Peak, July 10, 1897, Fink 574 (MIN). KOOCHICHING CO.: on rocks, Gunflint, June 30, 1897, Fink 290 (MIN). AITKIN CO.: on rocks, Palisades, July 15, 1897, Fink 743 (MIN). RENVILLE CO.: on rocks, rare, Morton, July 7, 1899, Fink 315 (MIN). YELLOW MEDICINE CO.: on wood, Granite e July 11, 1899, Fink 405 (MIN). PIPESTONE CO.: on pipestone, Pipestone, July 19, 1899, Fink 643 (MIN). COUNTY UNKNOWN: on rocks, Blueberry Island, July 13, 1901, Fink 575 (MIN); on rocks, South Fowl Lake, ад 26, im Fink 197 (MIN); on rocks, Misquah Hills, June 5, 1897, Fink 491 (MIN). COLORADO: BOULDER CO.: on о rocks, P. Peak, Roeky Mountain Na- tional Park, elev. 12,600 ft., Sept. 17, "s Kiener (MBG). HINGTON: CLALLAM CO.: on rocks, Sequim, 1916, Grant (F). ISLAND CO.' Es Island, Е 3158 (Е). OREGON: COUNTY UNKNOWN: on trees, Lloyd n FoREIGN EXSICCATAE EXAMINED: Claud. 261; Fl. Hung. 19; Hav. 444; Lojk. m 119; Med 138 ; priam 4; Moug. et ae DIEA Nyl. Pyr. 54; Roum. ; Roum. Gen ; Zw. Parmelia prolixa (Ach.) Róhl. is found growing on stones and trees throughout most of the northern part of North Ameriea and Europe. It is distinguished from Parmelia oli- vacea (L.) Ach. by the dissected character of the lobes and the yellow-green reaction of the upper cortex when treated with K. 9. Parmelia olivacea (L.) Ach., Meth. Lich. 213. 1803. Lichen olivaceus Linn., Fl. Te 244. 1737; Sp. Pl. 1143. 1753. Lobaria olivacea Hoffm., Deutschl. Fl. 2: 150. 1795. [Vor. 28 60 ANNALS OF THE MISSOURI BOTANICAL GARDEN Imbricaria olivacea DC. apud Lam. & DC., Fl. Fr. 2: 393. 1815. Parmelia subolivacea Nyl. apud Hasse, Bull. Torr. Bot. Club 24: 445. 1897. The type of Linnaeus from Ше ‘Flora Lapponica,’ is un- known, but that from the ‘Species Plantarum’ is in his her- barium at London, England. Acharius' type is in his herba- rium at Upsala, Sweden. Acharius cites as the same as his plant #. 24, f. 77 of Dillenius, ‘Historia Muscorum,’ Hoffmann, ‘Enumeratio Lichenum,’ t. 13, f. 3 € 4, and Linnaeus’ ‘Species Plantarum,' 1143. Hoffmann's type is unknown but he cites as equal to his plant Linnaeus, ‘Species Plantarum,’ 1143, and Hoffmann, ‘Enumeratio Lichenum,’ t. 13, f. 3 #4. DeCandolle considered his plant the same as Linnaeus’ and Acharius’ and he lists as a synonym the illustrations in Dillenius, *Historia Muscorum’ (t. 24, f. 77). Nylander's type of P. subolivacea is in the Nylander herbarium at Helsinki, Finland. The type col- lection of P. subolivacea was made by Dr. Hasse from rocks in the San Gabriel Mountains of California, and a co-type col- lection is in the United States National Herbarium at Washing- ton, D. C. The types of the European colleetors came from the north- ern part of Europe. The species is distributed widely in the northern portions of both Europe and North America. Thallus small to medium-sized, 3-10 em. diameter, adnate to bark of trees; upper surface much wrinkled; cortex unbroken, olive-brown in color; apothecium sessile, 1-4 mm. diameter, disk concave, olive-brown, amphithecium slightly incurved, roughened with tiny lobules; K, none, C, none, K + C, none, P, yellow. Algal layer continuous, 20-22 y thick; upper cortex 18-21 1 thiek, with yellow-brown matrix ; medulla loosely interwoven with one type of hyphae; lower cortex 5-7 y thick, with black matrix; thallus 144—150 и thick; thecium 30-35 и thick; ascus clavate, 8 spores to an ascus, spores 3-4 x 5-6 и; paraphyses branched simply. Distribution: Newfoundland to New York, west to Cali- fornia. 1941] BERRY—THE GENUS PARMELIA IN NORTH AMERICA 61 NEWFOUNDLAND: on rock, Sparrible Cove, Apr. 21, peu Waghorne 81 (MBG); on trees, Capstan Island, Oct. 3, 1889, Waghorne 203 (MIN); on bark, Chappel, Trinity Bay, July, 1883, Waghorne 27 (MIN); on rocks, vut Bay, May 10, 1891, а аи QUEBEC: on mountain ash, Lac Trois Saumons, May 13, 1936, Lepage 303 Беси оп Norms pine, Ste. Anne, Apr. 29, 1936, Lepage (MBG). ONTARIO: on trees, islands in Lake Nipigon, July 11, 1884, Macoun (WIS) ; trees, Big E July 1894, Millan ф Sheldon 22458 (MIN); Collins Inlet, Algona District, Sept. 12, 1912, Klugh (F); on alder stem and fence rails in Ontario and on is former in British ома 1878, Macoun 24 (WIS AROOST 0.: on cae Di St. varie’ Aug. 1893, Cummings $ б ps (MBG ne ns JEMEN); ox co.: on rocks, Rockport, May 7, 1910, Merrill 136 (NYBG,MIN). CUMBERLAND CO.: Cumberland, 1863, Blake (FM). New HAMPSHIRE: CARROLL CO.: Chocorua, Aug. 1911, Farlow 478 (WIS); on trees, Chatham, July 1905, Riddle (F). coos co.: White Mountains, Willey (MIN). CHESHIRE C0.: Monadnoc, Russell (FM). VERMONT: ADDISON CO.: on yellow birch, Goshen, elev. 1500 ft., Dec. 21, 1922, Dutton (MBG). ORLEANS CO.: Elephants Head, June 18, 1883, Faxon (WIS). MASSACHUSETTS: BRISTOL CO.: trunks and rocks, common but infertile, New Bed- ford, Willey 27 (FM). NORFOLK Co.: on smooth bark, Wellesley, Oct. 1907, Rid- dle (F). NEW YORK: YATES CO.: Dundee, Wright 11 (FM); Penn Yan, Buckley (MBG). OHIO: ASHTABULA CO.: on rails and apple trees, Orwell, Apr. 27, 1895, Bruge 868 2 e oak, poplar, Sept. x: 1935, аи: (NYBG). PIKE CO.: on linden, Sept. 15, 1935, Wolfe 488 MICHIGAN: MACKINAC CO.: on beech, Mackinac Island, July 1899, iod (ENS n Fagus E Mackinac Island, July 11, 1899, Harper $ Harper 68 (FM). KEWEENAW CO. n bireh, Roc Rn Isle Royale, Aug. 1904, ie ф таи (ЕМ); оп p Isle Royale, July 26, 1904, Harper § Harper 169 (FM); Alnus, Rock Harbor, Isle Royale, July 19, 1904, Harper § Harper 176 (FM); Tale Royale, July 9, 1901, Stuntz $ Allen (WIS). COUNTY UNKNOWN: Sailors Encamp- ment, Aug. 3, 1897, Harper § Harper (FM). WISCONSIN: ASHLAND CO.: on trees, Oak Island, dei 10, 1901, Fink (MIN); ac Vie VILAS Cheney 69 (WIS). BARRON CO.: Barron, pel * ns Cheney 13022 (WIS). RACINE CO.: valley of the Wisconsin River near Lenwood Ferry, Summer 1894, Cheney 8586 (WIS). ILLINOIS: MENARD CO.: Athens, Hall (WIS,FM); limbs of Quercus alba, high up, Athens, Mar. 1862, Hall ane LA SALLE CO.: on trees, Calkins (FM). MINNESOTA: COOK CO.: alder, Gr poy Marais, July 22, 1902, Fink 5160 (MIN); on balsam, Grand М July 2 , 1902, Fink 5158 (MIN); on alder, Grand Marais, Jum 22, 1902, Fink 5157 дека ; оп roeks par the lake, Grand Marais, Aug. 5, 1902, Fink 5296 (MIN); on trees, Grand Portage, June 18, 1897, Fink 11 (MIN). LAKE CO.: on trees, Beaver Bay, July 13, je Fink 712 (MI NY on trees, Snowbank Lake area, pes 20, 1897, Fink 848 (MIN). ST. LOUIS CO.: on bark, Vermilion Lake, Lat. 48°, July 20, 1886, Arthur, Bailey § Holway А80 (MIN) ; on trees, Ely, July 28, 1897, Fink 1004 (MIN); Tower, Aug. 20, 1901, [Vor. 28 62 ANNALS OF THE MISSOURI BOTANICAL GARDEN Fink 1859 (MIN); € Aug. 20, 1901, Fink 1623 (MIN); on trees, Tower, psp $5 (MIN ; on roeks, Kettle Fi rede ,1901, Fink 1337 (MIN). KOOCHICHING C0.: on trees, Koochiching, July 25, 1901, Fink $89 (MIN); on rocks, Rainy Lake City, Aug. 8, 1901, Fink 1206 (MIN) ; Rainy Lake City, Aug. 8, 1901, Fink 1281 (MIN); on trees, Rainy Lake City, Aug. 8, 1901, Fink 1167 (MIN); on trees, Rainy Lake City, Aug. 30, 1901, Fink 1162 (MIN). HENNEPIN CO.: on bark, Minnehaha, Apr. 1891, Sheldon S4115 (MIN). BLUE EARTH CO.: on trees and old wood, Mankato, June 23, 1899, Fink 61 (MIN). BELTRAMI CO.: on tamaraek in swamp, frequent, ed Lake, July 30, 1900, Fink 991 e: ; on wood, Lake of the Woods, July 1896, Millan 6 (MIN); on trees in tamarack swamp, frequent, Ot at July 7, 1900, Fink 513 (MIN); on rocks, Bemidji, July 4, 1900, Fink 446 (MIN); on rocks along the lake, Red Lake, Aug. 1, 1900, Fink 1039 (MIN); on M in swamps, Beaudette, June 1, 1901, Fink 56 (MIN); on oaks, frequent, Bemidji, July 4, 1900, Fink 443 (MIN). oTTERTAIL CO.: on trees, Leaf Hills, June 26, 1900, Fink 243 (MIN); Battle Lake, June 19, 1900, Fink 31 (MIN). YELLOW MEDICINE CO.: on trees, Granite Falls, July 15, vdd Fink 537 (MIN). ROSEAU CO.: on trees, Warroad, June 25, 1901, Fink am (MIN). PENNINGTON CO.: on trees, Thief iver "Falls, July 20, 1900, Fink 830 (MIN). IOWA: FAYETTE CO.: on trees, Aug. 1893, Fink (WIS); on trees, Aug. 1898, Fink (MIN); on trees, 1896, Fink (MBG). ALBERTA: on Salix, Banff, Sanson = MONTANA: YELLOWSTONE CO.: bark of live Alnus tenuifolia, Buffalo Creek, Helena National Forest, Sept. 10, base Flint (MONT); bark of Douglas fir and on dead Salix, Bear Gulch, Helena National Forest, 8.20. T.7. М. R 4 E., July 28, 1931, Flint (MONT). COUNTY UNKNOWN: on alder, Lower Pattee "аним Jan. 15, 1934, Flint (MONT). WYOMING: ALBANY CO.: Sheep Mountain, July 3, 1897, A. Nelson (MBG). NATRONA CO.: Willow Creek, May 22, 1897, E. Nelson (MBG). NEW MEXICO: SOCORRO co.: shade, San Andres Mountains, Rhodes Pass, 47 miles west of Tularosa, June 6, 1938, Hubricht B999 (MBG). VALENCIA CO.: shade, 7 miles north of Trechado, June 9, 1938, Hubricht B1069 (MBG) ; open woods, 16 miles south of El Morro, June 10, 1938, Hubricht B1105 (MBG). MCKINLEY CO pine-juniper forest, 22 miles south of Gallup, June 11, 1938, Hubricht B1140 (MBG). BRITISH COLUMBIA: on bark of tree, June 1915, Macoun 15 (F); Lillocet, July 1915, Macoun 456 (F WASHINGTON: PIERCE CO.: on alder, Longmires Springs, Aug. 1906, Harper $ Pies u.a SLAND CO.: on Alnus, 1923, Langley, Grant (WIS). WHITMAN CO.: on lower dead branches of yellow pines, along Palouse River near Palouse, Aug. 12, 1938, Ownbey $ Ownbey (MBG). OREGON: COUNTY UNKNOWN: on Alnus oregona, 1921, Grant (FM). CALIFORNIA: SAN BERNARDINO CO.: on oaks, Wildwood Canyon above Yucaipa Valley, Mar. 24, 1936, Reed (MBG); on oak, San Bernardino Range, July 1895, Hasse (FM). RIVERSIDE СО.: on oaks, San Jacinto Mountains, elev. 1600 ft., 1903, Hasse 103 (MIN). MONTEREY CO.: on rock, Dells Camp, San Antonio Canyon, 1894, Hasse (F). MARIPOSA CO.: on roeks, Yosemite, May 11, 1896, Cummings (WIS). 1941] BERRY—THE GENUS PARMELIA IN NORTH AMERICA 63 SAN JOAQUIN CO.: on mossy sandstone, Castle Rock, elev. 3000 ft., June 16, 1906, Herre 744 (MIN); on sandstone, Castle Rock, elev. 3000 ft., Dee. 3, 1904, Herre (МВС); on sandstone, Castle Rock, elev. 3000 ft., June 2 1906, Herre (MBG). UTTE CO.: hills 4 miles east of Chico, on Quercus Douglasii, Jan. 28, 1914, Heller 11160 (MIN,MBG). SANTA CLARA CO.: on Quercus dt Black Mountain, elev. 1600 ft., Herre (MBG). COUNTY UNKNOWN: on oaks, Ano Nuevo Creek, elev. 100 ft., Aug. 9, 1906, Herre (MIN). 9a. Parmelia olivacea var. aspidota Ach., Meth. Lich. 214. 1803. Lichen aspidotus Hoffm., Enum. Lich., t. 13, f. 5. 1784. Parmelia aspidota Rohl., Deut. Fl. 3, Abt. 2: 100. 1813. Acharius’ type of the variety is no longer in existence, but he cites Hoffmann, ‘Enum. Lich.,’ t. 13, f. 5. 1784, and Dil- lenius, ‘Historia Muscorum,’ t. 24, f. 78. 1741. Rohling cites as the type of his plant, Acharius, Meth. Lich. 214. 1803 The variety is widely distributed in the northern portions of the temperate zones. It is distinguished from the species by the abundant covering of granules and coralloid branchlets on the upper surface of the thallus and the margin of the apothe- cium. Thalloid characters of such a nature do not merit the rank of species. Distribution: Newfoundland to North Carolina, west to New Mexico. NEWFOUNDLAND: sandstone, White Bay, May 10, 1891, Waghorne (MBG); Trinity Bay, Mar. 21, 1883, Waghorne (MBG). NEW BRUNSWICK: bark of trees, 1871, Fowler (FM). ONTARIO: on trees, islands in Lake Michigan, July 11, 1884, Macoun (MBG); on eedars in swamp, Emo, July 17, 1901, A 667 (MIN). VERMONT: COUNTY UNKNOWN: Pri BSNH). MASSACHUSETTS: COUNTY KR rd pP ene Path, June 5, 1882, Faxon (Е). NEW YORK: YATES CO.: Penn Yan, Buckley (MBG). NoRTH CAROLINA: COUNTY INDEFINITE: on trees, mountains of North Carolina, Calkins 377 (FM) OHIO: COUNTY UNKNOWN: on bark of tree, Bogue (FM). MICHIGAN: WASHTENAW CO.: on aspen, Whitmore Lake, Мау 16, 1931, Lowe (M). ALCONA CO.: on aspen, Mud Lake Bog, May 16, 1931, Lowe (M) WISCONSIN: WAUSHARA СО.: west of Coloma, Oct. 13, 1935, J. Thomson (WIS). LAFAYETTE CO.: Fayette, July 15, 1894, Cheney 9206 (WIS). ILLINOIS: MENARD CO.: high n. in hickory trees, 1878, Hall (FM,MBG,BSNH). MINNESOTA: BELTRAMI СО.: on trees, Bemidji, July 12, 1900, Fink 672 (MIN). OTTERTAIL CO.: on trees, Leaf dus July 2, 1900, Fink 385 (MIN). PIPESTONE CO.: on quartzite, Pipestone, July 18, 1899, Fink 621 (MIN). couNTY UNKNOWN: Oak Island, July 9, 1901, Fink 464 (MIN). [Vor. 28 64 ANNALS OF THE MISSOURI BOTANICAL GARDEN IOWA: FAYETTE CO.: on trees, 1896, Fink (MBG,F,WIS,MIN). MONTANA: CASCADE CO.: Great Falls, Mar. 31, 1888, Williams 25 (MIN). COLORADO: EL PASO CO.: on live elm, Palmer Lake, June 20, 1928, Laidig (F). New MEXICO: SAN MIGUEL CO.: on trees, Las Vegas, Feb. 18, 1926, Bro. Anect (M). 9b. Parmelia olivacea var. glabra (Schaer.) Linds., Trans. Roy. Soe. Edinb. 12: 236. 1859. Parmelia olivacea* corticola var. glabra Schaer., Lich. Helv. Spic. sect. 10: 466. 1840. Both Schaerer and Lindsay considered Schaerer’s ‘Lich. Helv. Exs.’ 370, the type of this variety. The variety differs from the species only in that the medulla becomes red with CaOCls. Distribution: North American collections of the plant have been seen from only a few localities in southern California. CALIFORNIA: SAN BERNARDINO CO.: on oaks, elev. 1600 ft., June 1899, Hasse 265 (F) ; on mossy sandstone, elev. 3000 ft., Castle Rock, June 16, 1906, Herre (F,MBG). LOS ANGELES CO.: on rocks, Camp Baldy, elev. 4700 ft., Hasse (F). 10. Parmelia multispora Schneid., Guide to Study of Li- chens, 154. 1898. Parmelia olivacea var. multispora Merrill, Bryol. 12: 73. 1909. Parmelia olivacea var. polyspora Herre, Proc. Wash. Acad. Sci. 12: 199. 1910. Schneider’s type is a plant collected in the mountains east of San Francisco in California, and is now in the United States National Herbarium. Merrill's type is in his herbarium at Harvard University, and is probably co-type material from Schneider. Herre's type is number 426 in the Lichen Herbar- ium at Leland Stanford University. Thallus small, 3-7 ст. diameter, adnate, lobes wide and short, little-branched, margins smooth; upper surface smooth or slightly wrinkled, olive-brown or almost black; under sur- face brown and smooth at the margins, black at the center, with numerous short black rhizinae; apothecia numerous, sessile, 1-8 mm. diameter, disks flat to slightly concave, chestnut- brown or darker, amphithecia thin, smooth; K, none, C, none, К + C, none, P, none. 1941] BERRY—THE GENUS PARMELIA IN NORTH AMERICA 65 Algal layer continuous, 4-8 и thick ; upper cortex 8-10 р thick, matrix brown; medulla loosely interwoven; lower cortex 12- 16 и thick, matrix black; thickness of thallus 48-120 и; thecium 60-65 и thick, ascus clavate, spores 16 to many, spherical, 1- 4 u diameter; paraphyses branched, enlarged at the tips. Distribution: Idaho, west to British Columbia, south to Cali- fornia. IDAHO: BONNER CO.: on Salix, Cooling, Sept. 18, 1919, Weir (Е). BRITISH COLUMBIA: trunks of trees, Sidney, Vancouver Island, 1912, Macoun — KLICKITAT CO.: on oaks, Goldendale, Oct. 16—23, 1919, Foster (Е). SPOKANE CO.: on alders, Dead Man Creek, | (Пе OREGON: MARION CO.: Salem, 1871, Ha CALIFORNIA: LOS ANGELES CO.: San drm Bonito ue 1906, Hasse (F); on oak, Garapito Canyon, Santa Monica Mountains, Sept. 1, 1920, Mozley 802 (BPI). MARIPOSA CO.: on oak, Yosemite Valley, 1900, Riddle (F). SANTA CLARA CO.: on oak, Black Mountain, Santa Cruz Mountains, Apr. 30, 1904, Herre (BPI,MIN); on Quercus Wislizenii, Black Mountain, Santa Clara Mountains, elev. 1600 ft., Apr. 30, 1904, Herre 426 (Type of variety) (LSU). 11. Parmelia conspurcata (Schaer.) Wainio, Meddel. Soc. Fauna & Fl. Fenn. 24: 22. 1888 Parmelia olivacea var. corticola f. conspurcata Schaer., Lich. Helv. Spic., sect. 10: 466. 1840 Schaerer took as the type of the form of his variety, Lich. Helv. Exs. 371, which was collected in the Swiss Alps. Wainio, using the same type, raised the form to the rank of a species. Thallus small to large, 2-10 cm. diameter, loosely adnate, gray to olive-brown or greenish-gray ; upper surface wrinkled, covered with small soredia, margin erinkled, slightly sorediate, lobes round, short, slightly and irregularly dissected; lower surface brown to black, with few black rhizinae; apothecium, sessile, 4-7 mm. diameter, margin slightly lobed and sorediate, disk concave, chestnut-brown; K, greenish-yellow then brown, C, none, K + C, red, P, none. Algal layer continuous, 12-15 и thick; upper cortex 15-16 u thick, matrix yellow; medulla loosely interwoven, hyphae of one type; lower cortex 15-20 и thick, matrix brown or black; thecium 36-45 y thick; ascus clavate, spores 5-7 x 10-14 u, hya- line; paraphyses branched and knobbed at the tip. [Vor. 28 66 ANNALS OF THE MISSOURI BOTANICAL GARDEN Distribution: Alberta to British Columbia, south to Cali- fornia. ALBERTA: on white spruce, Banff, Oct. 1, 1910, Sanson 323 (F). BRITISH COLUMBIA: ‘‘Biglon Bay,’’? New Denver, Dec. 25, 1920, MacFadden 12? (BPI ). CALIFORNIA: SAN MATEO CO.: on sandstone, elev. 1900 ft., Bear Gulch road, Santa Cruz Mountains, May 12, 1904, Herre 488 (F,MBG,MIN); on rocks and mosses, elev. 1000-1300 ft., San Bruno Mountain, July 13, 1906, Herre 885 (MBG,MIN). 12. Parmelia exasperata (Ach.) DeNot., Giorn. Bot. Ital., 2: 193. 1847. Collema exasperatum Ach., Lichenogr. Univ. 645. 1810. Acharius’ type was from the Valley of St. Nicolai in Swit- zerland. DeNotaris, using material determined by Acharius, placed the species in the genus Parmelia. Thallus small, 3-8 cm. diameter, adnate on stone or wood, wrinkled, smooth and shining or granular, lobes long, narrow, tips rounded, tightly appressed, imbricated, olive-brown to olivaceous-black ; K, none, C, попе, К + C, none, Р, none. Algal layer continuous, 70-75 и thick; upper cortex 15-20 и thick; medulla loosely interwoven, hyphae of one type; lower cortex 15-20 и thick, matrix brown; thickness of thallus 180— 200 и; thecium 30-35 и thick; ascus clavate, spores 8, 2-3 x 4— би; paraphyses branched. Distribution: Maine to West Virginia, west to California. MAINE: KNOX CO.: on trunks of trees, Camden, Sept. 1904, Merrill (Е). NEw HAMPSHIRE: MERRIMACK 00.: on old fence, North Sutton, July 1914, Riddle 1427 (F). WEST VIRGINIA: POCAHONTAS CO.: Wanless, Jan. 1923, Gray 1,231 (F). HIO: WAYNE CO.: on Pyrus Malus, Wayne, Apr. 2, 1895, Bogue 778 (NYBG). MICHIGAN: CHEBOYGAN 00.: vicinity of Burt and Douglas lakes, J une-Aug. 1911, Leeson (F). ALIFORNIA: COUNTY INDEFINITE: on Quercus, Tehachapi Mountains, June 1907, . Hasse (BPI). MARIPOSA СО.: on tree, Yosemite Valley, 1900, Hasse (F). SANTA CLARA CO.: on rocks, hills near Stanford, elev. 150 ft., Apr. 9, 1904, Herre (MBG). IL. XANTHOPARMELIA Parmelia subgenus Euparmelia sect. IL Xanthoparmelia Zahl. apud Engler & Prantl, Nat. Pflanzenfam. I Teil, Abt. 1: 212. 1907. 1941] BERRY—THE GENUS PARMELIA IN NORTH AMERICA 67 Parmelia sect. Xanthoparmelia Wainio, Etude Lich. Brésil 1: 60. 1890 Wainio considered Parmelia endoleuca Taylor as the type of the section Xanthoparmelia. Zahlbruckner considered Par- melia conspersa (Ehrh.) Ach. as Ше type for section Xantho- parmelia of Кирагтейа. Thallus small, lobes narrow and imbricated; lower surface with small rhizinae, apothecia sessile. KEY TO SPECIES IN EUPARMELIA SECT. XANTHOPARMELIA A. is light-eolored below. wider than long; spores spherical, 4 x 4.5—5 x 5 w....16. sphaerosporella 5 Lobes longer than wide. ое Брого т ОС uo Ero tute evo е TER eee 19. leucochlora а ГИ 14. centrifuga А. Thallus always dark bel . Never with soredia, isidis, or coralloid branchlets. Never on D. Upper surface always smooth, uncracked; spores 4-6 x 7-9 м. PM M e dU RLW on RM 17. conspersa D. Upper surface always cracked; spores 7-10 x 12-16 w........ 18. lineola C. Always on soil, often drifted о piles by the wind........ 15. chlorochroa B. Always with soredia, isidia, or eoralloid branchlets. C. With isidia or coralloid branchlets; K, yellow...17a. conspersa var. isidiata C. With yellow globose powdery soredia; K, brown............... 13. incurva 13. Parmelia incurva (Pers.) Fries, Nova Sched. Crit. 31. 1826 Lichen incurvus Pers., Neue Ann. d. Bot. 7: 24. 1794. Lobaria incurva Hoffm., Deutschl. Fl.2: 156. 1795. Imbricaria incurvus Ach., Lichenogr. Suec. Prodr. 107. 1798. Persoon's type is unknown, but probably came from near Leyden, as the plant has a wide distribution throughout Europe. Hoffmann cites Lichen incurvus Pers. as synonymous with his Lobaria incurva. Acharius’ Imbricaria incurvus was deseribed from material collected in the mountains of Europe. He cites Persoon's Lichen incurvus and Hoffmann's Lobaria incurva as being synonymous with his plant. Fries cites all three of the previous treatments in transferring the species to the genus Parmelia. Thallus small, 5-15 em. diameter, adnate on stones, margins [Vor. 28 68 ANNALS OF THE MISSOURI BOTANICAL GARDEN yellowish-green with center dark greenish-gray, lobes long and very narrow, imbricated, the tips appressed ; cortex smooth except for scattered, globular, powdery, yellow soredia; under- surface dark brown with short, thick, black rhizinae evenly dis- tributed, the tips of which are fused into a continuous layer of some extent; apothecium sessile, 3-10 mm. diameter, margin smooth and entire, disk concave to flat, chestnut-brown; K, upper cortex black, medulla brown, C, none, K + C, none, P, none. Algal layer discontinuous, gonidia in clumps 60-75 py in diam- eter; upper cortex 60-75 и thick, matrix brown; lower cortex 60-70 и thick, matrix black; medullary hyphae loosely inter- woven, hyphae all of one type, 1-2 и in diameter; thickness of thallus 1000-1017 и; thecium 55-65 и thick; ascus short-cylin- drical, spores 8, 5-6 x 8-14 и; paraphyses branched once or twice. Distribution: Newfoundland to New Hampshire and Ver- mont. NEWFOUNDLAND: on rock, Chance Cove, Trinity Bay, 1882, Waghorne (MBG); on rocks, east coast, July 20, 1891, Waghorne 36 (MIN AINE: KNOX CO.: on quartzite rocks, elev. 900 ft., Силна, Aug. 22, 1909, Mer- rill 75 (MIN). иконе COOS CO.: on roeks, White Mountains, 1854, Tuckerman ia Vaz T: WINDSOR CO.: table rock, Mount Horrid, Rochester, elev. 2500 ft., Sept. T 1006 Dutton 1415 (FM). FOREIGN AND LOCAL ExsiccaTAE EXAMINED: E. Fries, 260; Harm. 69; Harm. Lich. end 278; Hav. 78, 314, 390 ; Malme, 178; Merr. 75; Norrl. et Nyl. 27; Roum. 433 ; Salw. 237 ; Stenh. 123; Tuck. 76. 14. Parmelia centrifuga (L). Ach., Meth. Lich. 206. 1803. Lichen centrifugus Linn., Sp. Pl. 1142. 1753. Imbricaria centrifuga Ach., Lichenogr. Suec. Prodr. 118. 1798. Linnaeus based his Lichen centrifugus on European material in his herbarium. Acharius says that he saw the Linnaean plants and considered them the same as his Parmelia centri- uga. Thallus small, light yellow-green above, orbicular, closely adnate to rocks; upper cortex free from isidia or soredia, mar- gins always smooth, lobes dichotomously branched, narrow, 1941] BERRY—THE GENUS PARMELIA IN NORTH AMERICA 69 with rounded closely appressed tips; lower surface light- colored with numerous, small, light-brown rhizinae evenly dis- tributed; apotheeium sessile, 2-7 mm. diameter, disk convex, chestnut-brown, margin somewhat incurved and smooth; К, upper cortex lemon-yellow, medulla yellow, C, none, K + C, yel- low, P, none. Algal layer continuous, immediately beneath the upper cor- tex, 35 и thick; medulla of loosely interwoven hyphae Ти or less in diameter, with no differentiation into a medullary strand or central cavity ; upper cortex 14 и thick with vertically extending hyphae surrounded by a yellowish-green matrix; lower cortex 28 и thick, composed of vertically interwoven hyphae with a black matrix; theeium 33-35 и thick; ascus pyri- form, spores 8, 3-4 x 5-7 и, arranged irregularly; paraphyses branched, tips slightly enlarged. Distribution: Labrador to Tennessee, west to Alaska and British Columbia. LABRADOR: on boulders and other rocks, Aug. 27, 1896, Low (F); rocks, coast of north Labrador, Aug. 23, 1896, Low (F). NEWFOUNDLAND: on rocks, White Bay, Oct. 1, 1891, Waghorne 359 (MIN); on rocks, Harbor Deep, White Bay, Nov. 1, 1891, РЕ (MBG); on rocks, Tanners Head, Mar. 4, 1892, Waghorne (МВС); on rocks, Corke Rock, July 12, 1893, Wag- horne (MBG). NEw BRUNSWICK: on rocks, Lobewue River, July 1884, Hay (Е). QUEBEC: on rocks, Ste. Anne, Apr. 23, 1935, Lepage 146 (D); among dry granite roeks at 1200 m., northeast summit of Mount Dunraven, Tabletop Mountains, Aug. 1, 1923, Fernald, Dodge $ Smith (D) ; on rocks at Rimouski, May 22, 1937, Lepage 2 (D MAINE: PENOBSCOT CO.: on rocks, New Port, July 27, 1922, Plitt (BPI). KNOX co.: on quartz, Camden, elev. 900 ft., Aug. 22, 1909, Merrill 75 (Е); on ledges at 900 ft. elev., Camden, edt 24, 1909, Merrill 96 (MIN). NEW HAMPSHIR oos co.: Gate of Notch, June 16, 1883, Faxon (WIS) ; alpine regions, White RE ed Mann (ЕМ); Mount Washington, July 1884, Faron (FM); White Mountains, она mid Farlow 430 (F); White oe Willey 46 (MIN). GRAFTON CO.: sum f Mounts Lincoln and Lafayette, Franconia Mountains, Aug. 15, 1892, ee + Heller 107 (MBG,WIS,F D. VERMONT: LAMOILLE CO.: ledges, the Chin, Mount Mansfield, elev. 5000 ft., June 20—24, 1922, Dutton (MBG). NEW YORK: ESSEX CO.: on rocks, Mount Marcy, near Lake Placid, 5000 ft., Aug. 16, 1934, Yovs 4197 (NYBG). PENNSYLVANIA: LANCASTER CO.: on rocks, Bear Town, July 11, 1894, Eby (MBG) TENNESSEE: HAMILTON CO.: on rocks, Lookout Mountain, Calkins 7 (Е). [Vor. 28 70 ANNALS OF THE MISSOURI BOTANICAL GARDEN MINNESOTA: ST. LOUIS CO.: on rocks, Misquah Hills, July 5, 1897, Fink 496a MIN). COLORADO: CLEAR CREEK CO.: above Bear Creek, elev. 12,000 ft., July 6, 1896, ALASKA: Алай Ал Island, Behring Strait, Wright (F). UKON: on rocks, Bonanza Creek, July 18, 1902, Macoun 24 (Е). BRITISH COLUMBIA: on rocks, back of Glain House, elev. 4450 ft., Aug. 6-7, 1906, Fink 5866 (M: FOREIGN AND LOCAL ExsiCCATAE EXAMINED: Cum. I. 107; Elenk. 102a, b ; Fellm. 80; E. Fries, 48; Th. Fries, 6; Howe, 25; Lojk. Univ. 158; Malme, 67; Meresch. 54; Merr. 96; Norrl. et Nyl. 202; Rab. 471; Stenh. 125; Tuck. 78; Zw. 450 15. cwn chlorochroa Tuck., Proc. Amer. Acad. Arts & Sci. 4: 383. 1860. Tuckerman's type was collected on the upper Missouri River near the mouth of the Cannon Ball River by Hayden. This plant is now in Tuckerman's herbarium, at Harvard Univer- sity. Thallus small, 2-4 em. diameter, growing on arid soil, loose and easily blown about by the wind, lobes long, narrow, re- peatedly dichotomously branched, somewhat imbricate, con- vex, whitish-gray to yellowish-gray ; under surface dark brown or blaek with small black rhizinae evenly distributed; apo- thecia rare, sessile, 2-6 mm. diameter, disks concave to flat, dark brown, amphithecia smooth or erenulate, slightly in- curved at the margin; K, upper cortex yellow, medulla none, C, none in either the upper cortex or medulla, K + C, medulla brown, P, yellow. Algal layer discontinuous, in groups of cells which are 25- 40 u in diameter ; upper cortex 28-30 y thick ; medulla of loosely interwoven hyaline hyphae; lower cortex 20-25 и thick; thallus 266—506 и thick; thecium 48-60 thick; ascus clavate, spores 8, 3—4 x 6-8 и; paraphyses thin, branched. Distribution: North Dakota to Nebraska, west to Saskatch- ewan and New Mexico. E: NortH DAKOTA: MORTON CO.: on ground, abundant on right side of Missouri River near Cannon Ball River, Hayden ('T'ype Colleetion) (F). SouTH DAKOTA: PENNINGTON СО: Black Hills, Upper Pole Creek, Aug. 1856, Engelmann (MBG); on dry earth among digg ae 1907, Skinner (MIN); on eroded clay soil, Rapid City, Mar. 20, 1927, osh (M,F). WASHINGTON со.: abundant on dry soil, Indian Draw, July 18, Bes pne 212? (FM). 1941] BERRY— THE GENUS PARMELIA IN NORTH AMERICA 71 NEBRASKA: SIOUX CO.: on dry calcareous soil near War Bonnet Canyon, June 20, 1890, Williams 106 (F,MBG,FM,MIN). SASKATCHEWAN: on earth, Old Wives Creek, May 25, 1895, Macoun (MBG MONTANA: CASCADE CO.: on soil, р northwest, Great Falls, Mar. 22, 1932, Flint (MONT). LEWIS AND CLARK CO.: soil, very abundant on windswept site, bluffs, east shore of Lake Sewell, Helena, m 4000 ft., Oct. 1931, Flint (MONT). BEAVERHEAD CO.: E on dry gravelly soil in sagebrush near Bannack, June 1926, Flint (MON WYOMING: ALBANY CO.: Laramie, July 23, A. Nelson (MBG); hills 6 or 7 miles east of Laramie, Nov. 15, 1936, G. Ownbey (MBG). JOHNSON со.: on the floor of the Devil's Kitehen, Powder River, June 27, 1910, 4. Nelson (MBG). CARBON CO.: gravelly hills on the ground, 60 miles from foothills, Hall (FM); full sun, 1 mile north of Bairoil P.O., June 30, 1938, Hubricht B1304 (MBG). WASHAKIE CO.: in crevices of boulders, oldiéra Park, North Fork of Clear Creek, Big Horn Mountains, PAN 19, 1898, fcu (MBG). FREMONT со: Wind River Mountains, 1860, en (T). LN CO.: on earth, Gros Ventre Canyon, July 19, 1939, Andrews ae CONVERSE CO.: full sun, 7 miles northeast of Orin, July 1, 1938, Hubricht B1314 (MBG). COLORADO: CLEAR CREEK CO.: Rocky Mountains, Hall (T); ау es 1895, Bethel (W). GUNNISON CO.: Rogers Bridge, 1878, Brandegee (Е). PARK C on earth, South Park, 1873, Wolf (FM). CHAFFEE CO.: Mount piper m d m. Sept. 8, 1901, Clements 4: Clements (NYBG). JACKSON CO.: ack Water of the Platte, Engelmann (Т). RouTT CO.: on earth, Mount Meadow, pe 343 (FM). E EXICO: VALENCIA CO.: full sun, on ground, 16 miles south of El Morro, June 10, 1938, Hubricht B1107 (MBG); full sun, on ground, 7 miles north of Trechado, June 9, 1938, Hubricht B1085 (MBG). CATRON CO.: full sun, on ground, 11 miles north of Pie Town, June 9, 1938, Hubricht B1064 (MBG). ПОСА ExSICCATAE EXAMINED: Merr. 237; Cum. I 106, II 27. Parmelia chlorochroa 'Tuck. is endemie in the Rocky Moun- tains and Great Basin region of North America. In the liter- ature and herbaria it has been confused with Parmelia mol- liuscula Ach. and vagans Nyl. Acharius? Parmelia molliuscula was named from a sterile collection made by Thunberg on Table Mountain, South Afriea. Nylander discusses collections from Siberia, Peru, and North America as а Par- melia vagans; these are not the same as Tuckerman’s Parmelia chlorochroa. 16. Parmelia sphaerosporela Miill.-Arg., Flora 74: 378. 1891. The type was collected by Dr. Lyall in the Galton Mountains of British Columbia, and sent to the Kew Herbarium as the Plants of the Oregon Boundary Commission. [Vor. 28 12 ANNALS OF THE MISSOURI BOTANICAL GARDEN Thallus small, 4-8 cm. diameter, adnate to substratum ; upper surface much wrinkled but not broken, margins appressed, round, smooth, lobes short and wide, not branched, upper sur- face golden-yellow, medulla white; lower surface light (white or yellow with many thick, light-colored rhizinae) ; lower cortex much wrinkled but not cracked or broken; apothecia numerous, 1-6 mm. diameter, sessile, concave, light-brown; margin of amphithecium thin, smooth, and unbroken; K, none, C, none, К + С, none, P, red-brown. Algal layer continuous, 25-35 и thick; upper cortex 20-25 и thick ; medulla very loosely interwoven, hyphae hyaline, rough and of one type, 4—6 u in diameter; lower cortex 20-25 р thick, matrix light-colored or hyaline; thickness of thallus very var- iable, 400-1000 и; thecium 60-64 и thick, hypothecium hyaline, 48—60 и thick ; algal layer continuous under the thecium; ascus cylindrical or slightly clavate, spores 8, spherical, 4 x 4.5-5 x 9 и; paraphyses branched and enlarged at the tips. Distribution: Ontario and British Columbia. ONTARIO: on trees, Emo, July 18, 1901, Fink 684 (MIN). BRITISH COLUMBIA: Galton Mountains, Oregon Boundary Commission, 1861, Lyall (Co-type collection) (F,LSU). 17. Parmelia conspersa (Ehrh.) Ach., Meth. Lich. 205. 1803. Lichen conspersus Ehrh. in litt. apud Ach., Lichenogr. Suec. Prodr. 118. 1798. Imbricaria conspersa Ach., Lichenogr. Suec. Prodr. 118. 1798. Ehrhart based his deseription on the plants in the Linnean Herbarium. Acharius cites Ehrhart’s plant as synonymous with his Imbricaria conspersa and Parmelia conspersa. Hoff- mann's use of L. centrifugus to designate the plant is non valid, as centrifuga had already been applied to another plant of the group. Thallus large, wide-spread, adnate to substratum; upper surface smooth, without cracks, isidia or soredia, margins always smooth, bright green to yellowish-brown; lobes short, with margins much but not deeply dissected, either dichoto- mously or irregularly branched, arising from the center of the 1941] BERRY—THE GENUS PARMELIA IN NORTH AMERICA 73 thallus and growing out over the lower thallus, often forming several layers; lower surface black or brown, much wrinkled, lower cortex continuous, rhizinae evenly distributed, but papil- lose; apothecium sessile, 1-5 mm. diameter, concave, disk chest- nut-brown; amphithecium slightly incurved, with few vertical fissures on the margin; K, yellow, C, none, К + C, none, P, red- rown. Algal layer continuous with slight lobing at the lower side, 35-50 и thick; upper cortex of vertical hyphae with a dark matrix, 6.5-7 и thick; thallus 133-171 и thick; lower cortex 36— 60 u thick; medullary hyphae loosely interwoven, with no sug- gestion of acentral strand or central cavity, two types of medul- lary hyphae, the common type small, less than Ти in diameter and hyaline, and a few scattered dark hyphae 2-8 y in diameter ; thecium 100-125 и thick; ascus clavate, spores 8, 4-6 x 7-9 y; paraphyses branched and slightly flattened at the tips. Distribution: Newfoundland to Alabama, west to Washing- ton and California. NEWFOUNDLAND: on sea cliff, Half-way Point, Bay of Hand, Apr. 29, 1897, Wag- horne 315 (MIN); on rocks, Jan. 3, 1899, Waghorne (MBG). New Brunswick: Grand Hirman, July 1879, May (FM); Owen Sound, Hall M). UEBEC: on rock, Ste. Anne, May 5, 1935, Lepage 148 (D); on boulder, Ste. Anne, May s 935, Lepage 47 ( ONTARIO: on boulders in woods and open plaees, Ottawa, Oct. 16, 1892, Macoun 155 (WIS,MBG) ; on boulders in pastures, Apr. 27, 1903, Merrill (MBG) ; on rocks, Algonquin Park, July 12, 1900, Macoun 247 (MBG); on rocks, Emo, July 22, 1901, INE: CUMBERLAND СО.: on rock and stones, Cumberland, Nov. 30, 1855, Blake (FM). HANCOCK CO.: Mount Desert Island, July 25, 1928, Plitt (BPI); Mount Desert Island, Middle Triad Summit, Sept. 1894, Rand (D). NEW HAMPSHIRE: coos CO.: Mount Shelburne, Cabot, June 9, 1844, Schrenk (MBG). CHESHIRE CO.: on cai rock, Jaffrey, Aug. 1916, Riddle (F). VERMONT: ADDISON CO.: cedar trees in swamp, Leicester, Feb. 11, 1910, Dutton 337 (FM); on huge PX. Nope Apr. 26, 1913, Dodge 155 (0); on quartz, ‘Middlebury, M Fi 10, 1913, Dodge 148 (D). RUTLAND CO.: on wood-shingled roof, Brandon, Sept. Bon Dutton 471 (FM). ER DAMES ESSEX С0.: on rock, Annisquam, Арг. 2, 1892, Cummings (MBG,MIN,WIS,FM) ; ur July 12, 1895, Harper 4 Harper 36 ад MIDDLESEX CO.: Concord, 1863, Mann (ЕМ); on rocks, вы Арг. 15, 1895, Fink (MBG). SUFFOLK со.: ar Boston, Faxon (WIS). RFOLK CO.: ae Nov. 18, 1883, Cummings p (W). BRISTOL CO.: on ol e wood, New Bedford, 1862-1898, Willey (US) ; on roeks, New Bedford, Willey 42 (FM,MIN [Vor. 28 14 ANNALS OF THE MISSOURI BOTANICAL GARDEN NEW YORK: JEFFERSON CO.: vicinity of Fishers Landing, July 3, 1902, Robin- son $ Maxon 80 (FM). YATES CO.: rocks, Quarry Hill, Penn Yan, Sept. 1850, Buckley (MBG iced maging PIKE CO.: Matamoras, Sept. 24, 1896, Schneider (MIN). n rocks, 1878, Rothrock (FM). LANCASTER CO.: on rocks, Chestnut Min ne pi Eby (MBG); on rocks, Mar. 1884, Eby (MBG). DELAWARE: NEWCASTLE СО.: on dry rocks, Faulkland, Feb. 7, 1887, Cummons (NYBG). MARYLAND: MONTGOMERY CO.: rock outcrop, A knoll, Plummers Island in Potomae River, near Cabin John, Nov. 6, 1938, Leonard (US); rocks, Sligo ie 1900, Williams (US,WIS). PRINCE GEORGES CO.: mE in stream, Oet. 5, Schulz (M). District OF COLUMBIA: Lugger (MIN). ORT DAVIDSON CO.: on exposed rock, Friedberg section, July 1, 1922, Schallert (WIS). CHATHAM CO.: on exposed rock, Bennetts Stone Quarry, May 3, 1929, Schallert (WIS). FLORIDA: ORANGE CO.: on Quercus, Sanford, Apr. 1922, Rapp (BPI). ALABAMA: LEE CO.: on rock, Auburn, Feb. 20, 1897, Earle $ Baker (MBG,MIN). MICHIGAN: KEWEENAW CO.: on stones, Isle Royale, July 1904, Harper $ Harper 112 (FM). COUNTY UNKNOWN: Sailors Encampment, Aug. 22, 1899, Harper $ Harper 39 (FM); loosely eovering rocks, Sailors Encampment, Aug. 27, 1899, Harper $ Harper 27 (FM) OHIO: HOCKING со.: on sandstone boulders, Oct. 20, 1935, Wolfe 355 (0). JACK- SON CO.: on sandstone at edge of cliff, May 6, 1934, Wolfe $ Соиер (О). ATHENS CO.: on rock, Apr. 4, 1936, Wolfe 567 (O) INDIANA: CRAWFORD CO.: on sandstone roek, along Little Blue River near old Canes Mill, about 2 miles south of Grantsburg, June 7, 1929, Deam 46 (BPI). TENNESSEE: HAMILTON СО.: on rocks, Lookout Mountain, Calkins 3 (MBG); on rocks, Lookout Mountain, Calkins 7 (МТМ); Lookout Mountain, Calkins 1 (FM). ISCONSIN: RACINE CO.: Valley of the Wisconsin River, near Linwood Ferry, Summer 1894, Fits 3485 (WIS). ADAMS CO.: bluffs, 10 miles east of Friendship, Sept. 13, 1935, Thomson (WIS) ILLINOIS: JACKSON CO.: on sandstone boulders near Bat Cave, 5 miles southwest of Pomona, Apr. 14, 1940, Hubricht B1735 (MBG). MINNESOTA: COOK СО.: on rocks, Grand Portage Island, June 23, 1897, Fink 108 27, , roeks, Harding, Aug. 19, 1901, Fink 1602 2 (MIN); on bern пон, Та к. dH sn 1743 (MIN). CARLTON CO.: on rocks, Та е, Carlton Peak, July 10, 1897, Fink 1 (MIN); on rocks, Kettle Falls, Aug. 9, 1901, Fink 1330 (MIN). KOOCHICHING с0.: on rocks, Gunflint, June 30, 1897, Fink 289 (MIN); on rocks, Palisades, July 15, 1897, Fink 755 (MIN). BLUE EARTH CO.: on rocks, Mankato, June 26, 1899, Fink 105 (MIN). LAKE OF THE WOODS co.: Lake of the Woods, July 1896, Millan 2 (MIN); BELTRAMI CO.: on granite, Bemidji, July 16, 1900, Fink 745 (MIN). YEL- LOW ae co.: on rocks, Granite Falls, July 12, 1899, Fink 449 (MIN). ROSEAU со.: he big rock, Warroad, July 4, 1901, Fink 368 (MIN). PENNINGTON СО.: on Puer granite roek, Thief River Falls, July 20, 1900, Fink 832 (MIN). CLAY 1941] BERRY— THE GENUS PARMELIA IN NORTH AMERICA T9 : rocks, Ulen, July 5, 1899, Fink 269 (MIN). COUNTY UNKNOWN: on rocks, нарани Island, July 13, 1901, Fink 567 (MIN); оп rocks, Misquah Hills, July 5, 1897, Fink (MIN). IowA: CLAYTON CO.: on sandstone, Sept. 1894, Fink (MIN). POWESHIEK CO.: in protected but not disintegrated places on sandstone, Grinnell, Mar. 13, 1904, Fink 147 (FM). LYON co.: on rocks, northwest corner, Aug. 6, 1896, Shimek (IA); on rocks, northwest corner of county, near White farm, June 1897, Shimek (IA). SOURI: WAYNE CO.: on granite rocks, Clark Mountain, C. Russell (MBG). ST. LOUIS СО.: on rocks, Ranken Estate, Apr. 30, 1939, Berry (MBG). IRON со.: on bend rocks, top of Pilot Knob, Dee. 1900, C. Russell (MBG) ; full sun to semi- shade, Royal Gorge, 3 miles east of Areadia, Nov. 19, 1939, Hubricht B1721 (MBG). ST. FRANCOIS CO.: on rocks, open woods, Koester, June 10, 1939, Hubricht B1501 (MBG); surface of sandstone ledge, glade, Pickle Springs, May 21, 1938, и (MBG). WASHINGTON с0.: exposed limestone cliffs, Vineland, Mar. 1898, С. Rus sell 137 (MBG). FRANKLIN CO.: on limestone ledges, Gray Summit, Nov. 15, 1936, Berry (MBG). REYNOLDS CO.: on granite rocks, north fork of Black River, Nov. 18, 1938, Schmitt (MBG). GREENE с0.: rocky wooded bank of James River, 8 miles south of Springfield, Apr. 7, 1939, Dodge, Berry & Johnson (MBG). ARKANSAS: WHITE CO.: near Judsonia, Dee. 28, 1938, Anderson (MBG). BOONE co.: rocky wooded bluff, above dry creek bed, Bear Creek Spring, .5 miles north of Francis, Apr. 7, 1939, Dodge, Berry 4 Johnson (MBG). GARLAND со.: dry exposed rock, Hot Springs, Dec. 25, 1899, C. Russell 15 (MBG) ; on rocks, open slope, North Mountain, Hot Springs, Nos 12, 1938, Schull (MBG). NEWTON CO.: high bluff and long steep wooded hillside, Lookout Point, 7 miles south of Jasper, Apr. 8, 1939, Dodge, Berry § Johnson (MBG) ; steep dry wooded bluff with rocky flat top, 1 mile south of Ponea, Apr. 8, 1939, Dodge, Berry ф Johnson KANSAS: DOUGLAS CO.: on sandstone, Jan. 1895, Bridwell 111 (F). OKLAHOMA: DELAWARE CO.: Dripping Springs, near Mosby, Apr. 19, 1936, Fas- set (WIS EXAS: BREWSTER CO.: full sun to semi-shade, Castle Rock, south end of Chisos Mountains, Мау 28, 1938, Hubricht B869 (MBG) ; full sun to semi-shade, 18 miles south of Marathon, May 27, 1938, Hubricht B856 (MBG). PRESIDIO CO.: full sun to semi-shade, Wire Gap, head of South Canyon, 15 miles southeast of Casa Piedra, May 30, 1938, Hubricht B962 (MBG). CULBERSON CO.: full sun to semi-shade, Beach Mountain, 8 miles north of Van Horn, Hubricht B963 (MBG) WYOMING: WASHAKIE CO.: in erevices of boulders, Баира rs Park, North Fork of Clear Creek, г. 2B Mountains, Aug. 19, 1898, Williams 316 (MBG). YEL- LOWSTONE PARK , Hawkins 17a (WIS). COUNTY INDEFINITE: Rocky Moun- tain Flora, Eu ГА Най (ЕМ). COLORADO: EL PASO CO.: Pikes Peak, Aug. 1901, Harper $ Harper (FM); o roeks, Colorado о с 6, 1904, Clements $ Clements (MIN NEW MEXICO: VALE .: on rock, semi-shade, 16 miles iod of El Morro, June 19, 1938, onu 3001 (MBG). ARIZONA: COUNTY UNKNOWN: under andesitie precipice, Barfoot Peak, Oct. 8, 1906, Plummer 1469 (FM). WASHINGTON: FERRY CO.: on rocks, Republic, 1913, Foster 103 (FM). CALIFORNIA: LOS ANGELES CO.: on rocks, moss and earth, mountain top, Santa [Vor. 28 76 ANNALS OF THE MISSOURI BOTANICAL GARDEN Catalina Island, Nuttall 434 (FM). SANTA CLARA СО.: on rocks, Los Trancos Road, elev. 400 ft., Mar. 22, 1904, Herre (MIN); on roeks in foothills, elev. 150 ft., Apr. 9, 1904, Herre 392 (MIN). LOCAL AND FOREIGN EXSICCATAE EXAMINED: Bal. 4234; Bartl. VIII 6; Breut. 402; Claud. 15; Cromb. 26 ; Cum. I 10, II 86; Desm. ed. II, ser. II 586; Elenk. 54a, c; Fellm. 79; Flk. 34; E. Fries 167; Funek, II 109; Hav. 153; Howe, 27; Johns. 303; Larb. Caes. 62; Malbr. 270; Malme, 204; Mand. 11; Mass. 314; Meresch. 55; Merr. 251; Norrl. et Nyl. 26; R. & S. 11; Roum. 44; ie. 144; Schl. Cent. IV 46; Stenh. 122; Wain. 106; West. 1041. 17a. Parmelia conspersa var. isidiata (Anzi) Berry, n. comb. Parmelia conspersa f. isidiata Anzi, Cat. Lich. Sondr. 28. 1860. Imbricaria conspersa f. isidiata Arn., Flora 67: 163. 1884. Anzi's type of the form is Anzi, Lich. Exsic. Ital. 110. Arnold cites Anzi in transferring the form to Imbricaria. The type of the new combination is Flint, University of Montana. The new combination was made because the outgrowth of coralloid branchlets from the thallus is considered to be a character of the order of a variety. The use of the form is reserved for less specialized variations. The upper surfaee of the thallus is covered with coralloid branchlets, otherwise the characters of size, habit, habitat, chemical reaction, and spore measurements are the same as those of the species. Distribution: Nova Scotia to Tennessee, west to California. Nova Scotia: on rocks, Yarmouth, June 10, 1910, Macoun (MBG). VERMONT: RUTLAND CO.: on lime rocks, Brandon, elev. 600 ft., Dec. 1, 1922, Dut- ton 1970 (MBG MASSAOHUSETTÉ: BRISTOL CO.: on old fenee, eountry road, New Bedford, Dec. 1895, Willey (US). TENNESSEE: HAMILTON СО.: Lookout Mountain, Calkins 1 (MBG). WISCONSIN: SAUK CO.: Parferys Glen, Oct. 1936, J. Thomson E TEXAS: GILLESPIE C0.: Young Blood Creek, Sonia 431 (MBG MONTANA: MISSOULA CO.: from rocks, Lower Pattee Canyon, ш; 15, 1934, Flint (MONT). FLATHEAD CO.: on rocks at northwest corner Gravelly Range Lake, Apr. 14, 1931, Flint Eos. IDA R CO.: shale rock of Shoshone Range, Little Sentinel Peak, Aug. 30, d vine c (Type of variety) (MONT); on rock, near Cabinet, Nov. 4, 1934, Flint = ONT). CAL NIA: LOS ANGELES CO.: rough sunny lava, Puddingstone Dam, San Jose Hills, pim 900 ft., May 5, 1933, Wheeler 1713 (MBG). FOREIGN BXSCCATAR EXAMINED: Johns. 221. 1941] BERRY—THE GENUS PARMELIA IN NORTH AMERICA 77 18. Parmelia lineola Berry, sp. nov. Thallus large, orbicular in outline, adnate to surface of rock; upper surface brownish-gray in the center to olive-green at the margins, much wrinkled, with many cracks in the cortex ; black ostioles of the spermagonia numerous on the younger portions of the thallus; lobes narrow, subdichotomously branched, im- brieated, margins thin, smooth; lower surface dark brown to black, rhizinae black, small and thinly scattered; lower cortex much wrinkled but not broken, with cracks as on the upper cor- tex; apothecium sessile, 1-6 mm. diameter, disk black, concave or flat, margin thin, smooth, and somewhat crinkled; K, yellow then brown, C, none, K + C, none, P, none. Upper cortex of vertical hyphae very dense, almost pseudo- parenchymatous, matrix yellow, 45-50 и thick; algal layer dis- continuous, gonidia formed of scattered groups of from 2 to 12 algal cells, located directly beneath the upper cortex; algae Protococcus; medulla of very loosely interwoven hyaline hyphae; lower cortex 36-38 и thick, matrix yellow or brownish- yellow; thecium 80-85 y thick; ascus clavate, spores 8, 7-10 x 12-16 и, hyaline; paraphyses branched once or twice, tips en- larged. Thallus orbieulatus ad saxa adnatus; superficies superior rugosa, cortice fisso; spermagonia numerosa in thallo juvenili ; laciniae lineares, subdichotome ramosae, olivaceo-virides ad griseo-brunneae; superficies inferior nigro-brunnea ad nigra, rhizinis parvis, nigris, tenuibus, sparsis; medulla KOH ad- dito flavescens, CaOCl:, K et C addito non mutans; apothecium sessile, 1-6 mm. diametro, discus niger, concavus aut planus; amphithecum tenue, undulato-crenulatum; cortex superior densus, 45-50 и erassitudine; algae Protococcideae; stratum eonidiale discontinuum, gonidia catenulata, 2-12 algarum cel- lulis; medulla laxe reticulata, hyphis hyalinis uniformibus ; cortex inferior 36-38 y crassitudine; thecium 80-85 и cras- situdine; asci clavati, sporae 8-nae, 7-10 y crassitudine, 12— 16 u longitudine, ellipsoideae; paraphyses cum ramis binis vel ternis, apicibus dilatatis. Distribution: Montana to Texas, west to Utah and Arizona. [VoL. 28 78 ANNALS OF THE MISSOURI BOTANICAL GARDEN TEXAS: CULBERSON CO.: full sun to semi-shade, north side of summit of Van Horn Mountains, 15 miles south of Van Horn, June 2, 1938, Hubricht B937 (MBG). MONTANA: FLATHEAD CO.: two and one half miles up Glen Trail east from Flat- head Lake, Nov. 18, 1934, Flint (MONT). WYOMING: YELLOWSTONE PARK, 1922, Hawkins 17с (WIS). своок co.: Sundance Mountain, July 3, oe А. Nelson (MBG). FREMONT 00.: Pole Creek, July 1, 1895, A. Nelson (MBG). NTY UNKNOWN: on the rocks of the cliffs, Gibbon Meadow, ще 28, 1899, 90 А shies 6192 (MIN,MBG). MEXICO: SOCORRO CO.: semi-shade, San Andres Mountains, Rhodes Pass, 47 di west of Tularosa, June 6, 1938, Hubricht B1006 (MBG). VALENCIA СО.: semi-shade, 7 miles north of Trechado, June 9, 1938, Hubricht B1074 (MBG) ; semi- shade, 2 miles southeast of Ramah, June 10, 1938, Hubricht B1114 (MBG). SIERRA CO.: full sun, 9 miles west of Engle, June 7, 1938, Hubricht B1019 (MBG). ARIZONA: APACHE CO.: on rock, open woods, 3 miles northwest of Ft. Defiance, June 12, 1938, Hubricht B1170 (Type collection) (MBG); on log, open woods, 3 miles northwest of Ft. Defiance, June 12, 1938, Hubricht B1181 (MBG); canyon wall, Massaere Cave, Canyon del Muerto, ius 14, 1939, Cutler (MBG). UTAH: SAN JUAN CO.: full sun to semi-shade, 5 miles north of Blanding, June 22, 1938, Hubricht B1223 (MBG). 19. Parmelia leucochlora Tuck., Syn. N. Amer. Lich. 1: 64. 1882 Tuckerman's type is the collection of Veiteh made in the state of Mississippi. Thallus oval or irregular in outline, small to medium-sized, adnate to bark of trees; upper cortex wrinkled, but free from any type of outgrowth, pale yellow or straw color above ; lobes rounded, short and rather wide, deeply dissected, with the edges imbricated, lower surface olive-brown, with many small evenly distributed rhizinae of lighter color than the lower cortex; apothecia sessile, 1-5 mm. diameter, slightly concave, disks chestnut-brown, margin slightly erenate; K, medulla and upper cortex yellowish-green, C, none, K + C, none, P, none. Algal layer continuous, slightly lobed on the under side, 60— 75 и thick; upper cortex of vertical hyphae with a yellow ma- trix, 30-40 и thick ; medulla of loosely interwoven hyphae of one type, 1-2 и in diameter ; lower cortex with a yellow matrix, 18- 20 и thick; thallus 280-300 и thick; thecium 60-70 и thick; spores 8 to an ascus, almost spherical, 7-10 х 9-12 y, ascus cla- vate; amphithecium with a continuous layer of algal cells; paraphyses clavate at tips, branched once or twice. The plant differs from P. conspersa (Ehrh.) Ach. in the size 1941] BERRY—-THE GENUS PARMELIA IN NORTH AMERICA 79 of the spores, the color of the under cortex, and in chemical reaction. The distribution is definitely southern and south- eastern. Distribution: South Carolina to Florida, west to Texas. SOUTH CAROLINA: BEAUFORT CO.: Beaufort, 1868, Mellichamp (Т). FLORIDA: VOLUSIA CO.: New Smyrna, Apr. 18, 1921, Kelly (NYBG,BPI). DUVAL co.: on oaks, Fort George Island, Calkins 6 (W,MBG). ORANGE CO.: on oaks, San- ford, 1905, Rapp 18 (F). COUNTY UNKNOWN: Miss Wilson (BSNH). MISSISSIPPI: COUNTY UNKNOWN: 1845, Dr. Veitch (Type collection) (T). LOUISIANA: COUNTY UNKNOWN: 1853, Hale (T). RKANSAS: NEWTON CO.: steep dry wooded bluff with B flat at top, 1 mile south of Ponea, yis 8, 1939, Dodge, Berry 4 Johnson (MBG). TEXAS: AUSTIN CO.: оп cedar twigs, Austin, Apr. 1928, РИН 34 (BPI). > ПТ. HYPOTRACHYNA Parmelia subgenus Euparmelia sect. III. Hypotrachyna Zahl. apud Engler & Prantl, Nat. Pflanzenfam. I Teil, Abt. 1: 212. 1907. Parmelia section Hypotrachyna Wainio, Etude Lich. Brésil 1: 38. 1890. Zahlbruckner’s type of the section is Parmelia sublinearis Wainio. Wainio’s type is Parmelia acanthifolia Pers. Thallus dark greenish-gray or darker, under surface black, except in P. rudecta, P. Borreri and P. cubensis; rhizinae always present and evenly distributed; medulla loosely inter- woven, without a cavity, hyphae always of one type, hyaline and of small diameter. The distribution of the section is world- wide, with the greatest number of species from the tropics. KEY TO SPECIES IN EUPARMELIA SECTION HYPOTRACHYNA A. Thallus grayish-green or darker. B. Thallus light-eolored below. . With soredia above; К, none, C, попе........................ 28. rudecta с. With coralloid branchlets above; K, yellow, C,red.............. 24. Borreri B. Thallus brown to black bel С. я With EUN ae above. Lobes more or less branche F. Thallus small, e enl blaek rhizinae; K, none, C, red. СОС а D rp E ООА 34. Finkié F. Thallus large, with many black rhizinae; K, yellow, C, yellow. [Уог. 28 80 ANNALS OF THE MISSOURI BOTANICAL GARDEN G. Thallus light to dark but not Ъаек................. 21. sazatilis С. Thallus black and granular............... 21a. saxatilis var. Anzi E. Lobes rarely branched, sometimes bearing fibrils on upper surfaee; К, brown, ©, |. HEP 35. caroliniana D. Without eoralloid branchlets above. E. Lobes dichotomously bra ; K, none, C, brown......... 23. Herrei E. Lobes irregular or subpinnately р һе Е. Rhizinae papillose; К, none, C, попе................... 27. texana F. Rhizinae strong, numerous. G. Medulla sulphur-yellow; K, none, C, none......... 41. sulphurosa С. Medulla white; К, none, C, orange-brown............-. 33. livida C. Lobes wide. D. Margins flat or very slightly raised. E. With marginal soredia or black cilia. F.W th marginal soredia; К, none, C, попе................ 30. cetrata F. With heavy black жылы cilia; К, none, C, поме........... еее 30a. каки var. subisidiosa E. Without marginal soredia Е. Rhizinae numerous; К, red-brown, C, none............ 29. E Е. Rhizinae few; К, yellow, С, попе.................... D. Margins raised, under side E. With marginal cilia; К, none, C, уеПож................... 32. erecta E. Without marginal eilia; K, cortex yellow-green, medulla none, M ee виси (C SYLT TET ee ce TL IT 31. Hubrichtii A. Thallus de; меча: green to yellowish-brown. . Thallus yellowish-brown or darker, lobes smooth or with coralloid branch- let И О о ка VIO tC SERE . omphalodes B. Thallus pale yellow-green to yellow-brown. C. Lobes narrow. D. Closely adnate. E. Lower surface light brown; К, попе..................... . cubensis „ Lower surfasé DISSE; К, ОРОМ, оне roa ht X 36. sublaevigata D. Loosely adnate; lower аи black ыз» е Гө к PN VETE. VAS 38. laevigata C. Lobes D. "usi PH smooth, unbroken. E. Medulla yellow, K, brown; lower surfaee dark.......... 40. aurulenta edulla white; K, yellow; lower surfaee light......... 39. endozantha D. Upper cortex pitted, or retieulately sorediate. E. Upper cortex pitted; lower surface light; K, none........ 25. Bolliana E. Soredia in retieulate ridges; lower surface blaek; K, yellow Е аР И НАИ а врана а TEL TS 2. sulcata 20. Parmelia omphalodes (L.) Ach., Meth., Lich. 204. 1803. Lichen omphalodes Linn., Sp. Pl. 1143. 1753. Parmelia saxatilis var. omphalodes Fr., Lichenogr. Eur. Reform. 62. 1831. 1941] BERRY— THE GENUS PARMELIA IN NORTH AMERICA 81 Acharius’ type was from Linnaeus’ herbarium. The type is now in the Linnean Herbarium in London. The plant used by Linnaeus was probably collected in central Europe. Fries cites as typical plants of his variety, E. Fries, Lich. Suec. Exs. 108, and Moug. & Nest. Exs. 348. Thallus 4-10 em. diameter, loosely adnate to stones or rarely to firm earth, shining olive-brown to dark brown or black; upper surface rough with wrinkles and thickly covered with согаПоіа branchlets or granules and flat lobules, lobes long, narrow, much branched, dissected, and imbricated; lower sur- face wrinkled but unbroken; cortex black, numerous strong black rhizinae to the margins of the thallus; apothecium 1- 4mm. diameter, sessile, disk slightly concave or flat, chestnut- brown, margin of amphithecium covered with coralloid branch- lets; K, upper cortex and medulla yellow then brown, C, none, К + C, none, P, none. Algal layer continuous, irregular in thickness, 5-50 и thick; upper cortex 20-24 и thick, matrix brown; medulla loosely interwoven, hyphae of one type, hyaline; lower cortex 20-86 и thick, matrix black or dark brown; thallus 160-180 y thick; thecium 48—60 и thick; ascus clavate, spores 8, 3-4 x 5-6 y; paraphyses branched. Collections of this species from Europe are usually larger in external respects than American collections, and very rarely have the coralloid branchlets on the upper surface. There is a tendency to form overlapping layers of the thallus which helps to distinguish this species from Parmelia saxatilis (L.) Ach. in the field. Distribution: Labrador to New Hampshire, west to Alaska and British Columbia. LABRADOR: on rocks, Red Bay, Aug. 12, 1889, Waghorne (US). QUEBEC: on wet stones, Ste. Anne, Apr. 30, 1936, Lepage 386 (MBG). NEW HAMPSHIRE: COUNTY INDEFINITE: on rocks, mountains, Calkins (ЕМ); on trees, mountains, Calkins 376 (FM). LASKA: on the Arctic Ocean, ten miles east of Cape Lisbon, Aug. 22, 1880, Bean FM). | e COLUMBIA: on roeks, Deer Park, Columbia River, June 1890, Macoun 86 (WIS) ; on rocks, Ucluelet, Vancouver Island, July 4, 1908, Macoun (MBG). [Vor. 28 82 ANNALS OF THE MISSOURI BOTANICAL GARDEN 21. Parmelia saxatilis (L.) Ach., Meth. Lich. 204. 1803. Lichen saxatilis Linn., Sp. Pl. 1142. 1753. Гората saxatilis Hoffm., Deutschl. Fl. 2: 145. 1795. Parmelia saxatilis var. leucochroa Wallr., Fl. Crypt. German. 3: 499. 1831. Linnaeus’ type was a European collection, which is still in his herbarium in London. Acharius accepted this plant as the type, transferring it from the genus Lichen to Parmelia. He cites Jacquin, ‘Collectanea’ 4: 281. $. 20, f. 2. 1790, which is designated Lichen saxatilis, as synonymous with his plant. Hoffmann cites Jacquin’s figure as synonymous with his Lo- baria saxatilis. The type of Hoffmann is unknown. Wallroth cites Linnaeus' plant and the figure of Jacquin as synonymous with his concept of the variety leucochroa. Thallus foliose, 6-10 ст. diameter, appressed; upper sur- face smooth, slightly cracked or with few scattered soredia, lobes linear, narrow, imbricated, dichotomously or subdichoto- mously branched, margins very slightly ascending; lower sur- face black to the margins, rhizinae black, numerous, papillose at the margins, long at the center of the thallus; apothecium sessile, 3-10 mm. diameter, disk concave, chestnut-brown, mar- gin crenulate, dissected irregularly and usually sorediate; K, upper cortex yellow, medulla yellowish-green then reddish- brown, C, medulla and upper cortex yellow, K + C, yellow, P, red-brown. Algal layer continuous, 16-18 и thick; upper cortex 24-28 y thick ; medulla of loosely interwoven hyphae of one type; lower cortex 32-33 и thick; thickness of thallus 131-138 и; thecium 40—44 u thick ; ascus clavate, spores 8, 8-9 x 15-16 и; paraphyses branched, enlarged at tip; spermagonia immersed in surface of thallus lobes, spherical, walls black above, brown or hyaline be- low, opening by black ostioles, spermatiophores simple, sper- matia fusiform. Distribution: Newfoundland to Florida, west to Washington and California. EWFOUNDLAND: on trees, Bar Harbor, June 24, 1893, Waghorne (MBG); on rocks, Western Cove, White Bay, Dee. 11, 1891, Waghorne (MBG); Chance Cove, 1941] BERRY—THE GENUS PARMELIA IN NORTH AMERICA 83 Trinity Bay, Sept. 12, 1882, Waghorne (MBG); on trees, New Harbor, Sept. 21, 1893, Waghorne 39 (MIN) Nova Scotia: on rocks, Springfield, Aug. 17, 1910, Macoun (Е); on twigs and branches, Rothrock (FM). QUEBEC: mossy woods between Пас Mont Louis and the North Fork of Madeleine River, July 31, 1923, Fernald, Dodge $ Smith (MBG); on rocks covered by leaf mould, Ste. Anne, May 5, 1935, Lepage 135 (D) ; on roeks, Ste. Anne, May 5, 1935, Lepage 139 (D) ; adherent to roeks, Ste. Anne, May 11, 1935, Lepage 115 (D); dry roeks, Ste. Anne, May 15, 1935, Lepage 138 (D) ; on boulders, Ste. Anne, May 30, 1935, Lepage 269 (D). ONTARIO: on rocks, Emo, July 20, 1901, Fink 768 (MIN); quite common on trees, dead wood, and rocks, Belleville, 1871, Macoun 73 (WIS,MIN,MBG) ; on trees and old logs, Ottawa, icd 26, 1896, Macoun (MBG Е: HANCOCK CO.: on rock, Mount Dont Island, Cliffwalk, Seal Harbor, July 12, d Merrill (D). KNOX co.: on trunks of trees, Camden, May 21, 1912, Merrill (WIS,MBG,FM). CUMBERLAND CO.: on trees, Cumberland, 1855, Blake (FM); on fun St. Thomason, Sept. 17, 1910, Merrill (F). NEW HAMPSHIRE: COOS CO.: near top of Mount Washington, Aug. 1926, Kirk и. VERMONT: WINDSOR CO.: ledges, Mount Horrid, Rochester, elev. 2700 ft., Oct. 3, 1920, zs Ди 1110 (FM); base of shaded ledges, Mount Horrid, Soi. Oet. 3, 1920, n (MBG). RUTLAND CO.: on dead wood, Pawlet, Mar. 22, 1913, Dodge 157 z^ MASSACHUSETTS: BARNSTABLE CO.: on trees, near Morrison, Cape Cod, July 14, 1937, Berry Кыры, ESSEX CO.: on rocks, Rockport, Apr. 7, 1899, Cummings 274 (FM). BRIS : on bark, New Bedford, Willey (MIN); on rocks and trunks, New И ites а 5 ; on moss, New Bedford, Willey (MBG). New YORK: ESSEX CO.: Newcomb, June 8, 1922, House (Е). YATES CO.: on wood, Penn Yan, Buckley (MBG) ; Dundee, Wright (FM). PENNSYLVANIA: PIKE CO.: Sept. 24, 1896, Schneider (MBG). HUNTINGTON со.: on trees, Pennsylvania State College Nature Camp, 16 miles southeast of State Col- lege, acd 23, 1937, J. Thomson (WIS VIRGINIA: LEE CO.: edge of wood, Hunt Gap, Powell Mountain, 2.7 miles north of irs и Aug. 23, o Hubricht B404 (MBG). H CAROLINA: ALEXANDER CO.: on bark pud Mount Pisgah, Sept. 9, 1922, Schallert 3094 (WIS). мїтс ven ek June 1 Schallert (B) ; on Roan в ae 18, 1801, ES (M SOUTH CAROLINA: COUNTY UNKNOWN: on rocks, ern an Jan. 14, 1928, Evans Я FLORIDA: ALACHUA CO.: on red maple, low hammock, near Gainesville, Маг. 14, 1938, Murrill (D). COUNTY INDEFINITE: on trees, Black Forest, Sept. 1913, Rapp F MICHIGAN: KEWEENAW CO.: Isle Royale, Aug. 12, 1901, Stuntz E Allen (WIS); south of рае Isle ола Aug. 12, 1901, Stuntz $ Allen (WIS) ILLINOIS: LASA 0.: on trees and rocks, Calkins (F). MENARD CO.: Athens, Hall (FM). KENTUCKY: LAUREL CO.: deep wet woods 8.4 miles north of East Bernstadt, Aug. 18, 1937, Hubricht B275 (MBG). [Vor. 28 84 ANNALS OF THE MISSOURI BOTANICAL GARDEN заанууд HAMILTON 00.: оп trees, Lookout Mountain, Calkins 285 (MIN). A: COOK CO.: on young spruce trees near water, Grand Marais, July 21, их, Fink 5138 (MIN); on soil-eovered rocks, Grand Marais, July 21, 1902, Fink 5133 (MIN). LAKE C0.: Baptism River, Aug. 1896, Elfman (WIS). ST. LOUIS CO.: on inn in swamps, Harding, Aug. 16, 1901, Fink 1506 (MIN). KOOCHICHING co.: on cedars in swamps, Koochiching, July 29, 1901, Fink 999 (MIN). BELTRAMI 0.: on tamarack in swamp, Bemidji, July 7, 1900, Fink 621 a ; on old pine logs, Red Lake, July 26, 1900, Fink 891a (MIN); on trees in swamp, Beaudette, June 18, 1901, Fink 2 9 (MIN) ; on cedar in swamps, Beaudette, Jose 22, 1901, Fink 112 (MIN); on cedars in swamp, June 24, 1901, Fink 129 (MIN). HUBBARD CO.: on balsam, Park Bay, Aug. 29, 1902, Fink (MIN). BECKER CO.: on roeks, Oak Island, July 9, 1901, Fink 435 (MIN). OTTERTAIL CO.: on tamarack in swamp, Hen- ning, June 25, 1900, Fink 190 (MIN). WISCONSIN: FOND DU LAC CO.: two miles south of Hamilton, Мау, 1936, J. Thom- son (WIS). viLAS CO.: valley of the Wisconsin River, near Lac Vieux Desert, Sum- mer 1893, Cheney (WIS). DANE CO.: east side of Picture Rock, Oct. 2, 1938, Skinners (WIS). JUNEAU CO.: north of Necedah, Apr. 1936, J. Thomson (WIS). SAUK CO.: on roeks, Devils Lake, Oet. 23, 1936, J. Thomson (WIS). IOWA: BREMER CO.: on trees, July 1894, Fink (MBG). MISSOURI: WAYNECO.: on granite rocks, Piedmont, Mar. 1899, C. Russell (MBG). ARKANSAS: CHICOT СО.: near Macon, Dee. 29, 1938, E. Anderson (MBG). MONTANA: MISSOULA CO.: on granite rocks, vates spots, Blaek Bear Moun- tain, Bitter Root Mountains, Ра 4, 1931, Flint ( WYOMING: YELLOWSTONE PARK, 1922, Hawkins dan COLORADO: WELD CO.: over siliceous rocks and mosses, p St. Vrain Canyon elev, 7200 ft., Oct. 7, 1938, Kiener (MBG). EL PASO 00.: on moss-covered rock, elev. 8300 ft., Мала, near Pikes Peak, July 1919, 1. M. Johnston (Е); оп dead spruce, Muir Lake, July 9, 1928, Laedig (F). NEW MEXICO: RIO ARRIBA CO.: northward face of cliff, vicinity of Brazos Canyon, Sept. 4, D Standley $ Bollman (US ALA : logs in Cross Sound, June 18, 1880, Bean (FM); Unalaska, July 28, 1882, a (FM); on Picea, near Kink, Aug. 1913, Cheney 159 (MBG); on bark of trees, Quinhagak, Nov. 19, 1921, Stecker (ЕМ); at Cape Lisbon, Aretie Ocean, Bean BRITISH rehome IA: on trees and old fences, Comox and Victoria, Vancouver Is- land, May 1893, Macoun 150 (MIN, d ; on boulders, Sidney, Vancouver Is- land, Sept. а 1913, Macoun (Е); Vancouver m Анн vc Commis- sion, spiri Lyall (F); New Жоан. Jan т (Е). Was TON: WHITMAN i on >и — Paliman, EX 22, 1901, Cotton (F). Ба SPOKANE CO.: On granitic Dee т (F). KLICKITAT CO. : LE 2 of не p uada cca С газ 24, 1909, ter (Е). AND CO.: roek, Langley, June 1923, Grant (D); on trees, tesano, fr 1919, has ЧЁ), CHEHALIS C0.: near Aberdeen, Apr. 4, 1909, Fos- er (D). CALIFORNIA: NAPA CO.: on oak trees 1 mile south of Pope Valley Post-office, July 9, 1938, Ownbey ^ Ownbey (MBG). SHASTA CO.: on bark of tree, region of upper бетй, July 24-Aug. 10, 1894, Howe (US). SAN MATEO CO.: оп sand- 1941] BERRY—THE GENUS PARMELIA IN NORTH AMERICA 85 stone hills 4 miles west of Stanford University, elev. 300 ft., July 22, 1903, Herre 172 (MIN жет. ;on eos ра Creek, elev. 200 ft., 2 miles from Pacific Ocean, June 30, 1906, Herre 830 (MIN,F). COUNTY renee on sandstone rocks, Searsville Ridge, Pun 500 9n July 22, ine Herre (F). CAL AND FOREIGN EXSICCATAE EXAMINED: Bohl. 11; Claud. 173; Cromb. 27; Cum. I 352; Cum. II 274; Desm. ed. I. ser. I 1940, ed. II. ser. I 1590; Elenk. 53 а-с; Erb. I 118; Fellm. 77; Е Fries 1n Funck, II 108; Hav. 179, 223; тъ Herb. 211; Leight. 203; Malme, 265; Moug. & Nest. 249; ‘Oli 1116 1 9:457 Rab. Cent. 16; Salw. 140; Stenh. 155 (saxicola), 156 RU: Trev. 267. 21a. Parmelia saxatilis var. Anzi Del. apud Duby, Bot. Gallie. 2: 602. 1830. Parmelia horrescens Tayl. apud Mack., Fl. Hibern. 2: 144. 1836. Delise's type was from France and is in his herbarium. Taylor's type of P. horrescens was from County Kerry, Ire- land, and is in his herbarium at the Boston Society of Natural History in Boston. Thallus grayish-white, densely clothed with grayish-brown isidioid excrescences, otherwise as in the species; K, upper cor- tex and medulla yellow, medulla becoming red-brown ; C, upper cortex and medulla yellow, K + C, yellow, P, red-brown. The distribution of the variety is limited to the northern part of United States and southern Canada. It is found on wet or shaded rocks. Distribution: Newfoundland to Maine, west to British Co- lumbia. NEWFOUNDLAND: on rock, Western Cove, Nov. 12, 1881, Waghorne (MIN); on rocks, M Sallon,’’ July 23, 1893, Waghorne 11 (MIN); on rocks, Red B Aug. 12, 1899, Waghorne 43 (MIN); Chance Cove, Trinity Bay, Sept. 11, 1593, Бри Aa QUEBEC: АВЕ wet rocks and chimneys, elev. 825-1125 m., Rivière Ste. Anne des More De Mountains, Aug. 5, 1923, Fernald, Dodge $ Smith (D). MAINE: HANCOCK CO.: Salisbury Cove, Aug. 29 gura Plitt (BPI). KNOX co.: on granite, South Thomaston, Sept. 17, 1910, Merrill (MIN). New HAMPSHIRE: COOS CO.: ааа regions, Wu Mountains, 1863, Mann (FM MINNESOTA: YELLOW MEDICINE CO.: Granite Falls, July 17, 1899, Fink 561 (MIN). CLAY co.: on rocks in shaded places near Ulen, July 5, 1899, Fink 268 (MIN) BRITISH COLUMBIA: Oregon Boundary Commission, Vancouver Island, 1858—59, Lyall (WIS). [Vor. 28 86 ANNALS OF THE MISSOURI BOTANICAL GARDEN 22. Parmelia sulcata Taylor apud Mack., Fl. Hibern. 2: 145. 1836. Taylor's type was from County Kerry, Ireland, and is in his herbarium in the Boston Society of Natural History at Boston. Thallus loosely adnate, 5-12 em. diameter; upper surface with reticulate ridges of white soredia, light greenish-gray or light brown, lobes short, wide, rounded, imbricated, dichoto- mously branched; lower surface black to the margin, with strong black rhizinae to margin of thallus; lower cortex un- broken; apothecium sessile, 1-15 mm. diameter, margin unfis- sured but slightly sorediate; K, upper cortex yellow, medulla yellowish-brown; C, upper cortex yellow, medulla попе; К + С, yellow; P, red-brown. Algal layer continuous, 20-24 и thick; upper cortex 20-28 и thick; medulla of loosely interwoven hyphae of one type, hya- line and of small diameter; lower cortex 16-32 и thick; thallus 120—160 и thick ; thecium 64-80 и thick; ascus clavate, spores 8, 4—5 x 12-14 и; paraphyses branched. Distribution: Newfoundland to Tennessee, west to Washing- ton and California. NEWFOUNDLAND: On rocks, € 7, 1893, Waghorne (MBG,MIN); on rocks, Outer Cove, Dec. 11, 1891, Waghor e (W). Nova Scotia: on old boards, ata and trunks, Sable Island, July 27, 1899, Macoun pe p MAINE: WAS TON CO.: July 1929, Markin (WIS). KENNEBEC CO.: on rock maple, Albion, Ma. 23, 1922, Parlin 5 (D). CUMBERLAND CO.: on tree trunks and ч Nov. 7, 1855, Blake (ЕМ HAMPSHIRE: COUNTY INDEFINITE: summit of Mount Willard, June 2, ig pe (W dp ctt WINDSOR CO.: on ledges, Mount Horrid, Rochester, elev. 2700 ft., Oct. 3, 1920, Dutton 1110 (WIS). ADDISON CO.: on Acer rubrum, college campus, ees Аре 18. pos — 100 (D). ND CO.: on eedar in swamp, Brandon, elev. 475 ft., Feb. 1921, Dutton pb (WIS, РМ); Chipman Lake, Tinmouth, June 1938, dbi "d MASSACHUSETTS: ESSEX CO.: on rocks, Rockport, Apr. 7, 1899, Cummings 325 (WIS). MIDDLESEX со.: Concord, 1863, Mann OHIO: C : Springfield, 1879, Mrs. E. J. Spence (F). MICHIGAN: KEWEENAW CO.: Isle Royale, Sept. 9, 199, Stuntz ф Allen (WIS). TENNESSEE: COUNTY UNKNOWN: on roeks, Calkins 344 (FM). WISCONSIN: ADAMS CO.: bluffs 10 miles east of Friendship, Oct. 13, 1935, J. Thomson (WIS). MONROE CO.: 9 miles west of Tomah, Nov. 14, 1935, J. Thomson 1941] BERRY—THE GENUS PARMELIA IN NORTH AMERICA 87 (WIS). RICHLAND CO.: valley of the Wisconsin River near Eagle River, June 27, 1893, Cheney 870 (WIS ILLINOIS: LASALLE CO.: on rocks, Calkins (FM). соок co.: on trees, Calkins M (FM). MINNESOTA: ST. LOUIS CO.: on rocks, Tower, Aug. 23, 1901, Fink 1710 (MIN). CARLTON CO.: on Men Kettle Falls, Аш 12, 1901, Fink 1880 (MIN). KOOCHICH- : Gunflint, June 30, 1897, Fink 240 (MIN); on old ите Battle Lake, Fane B 1900, E 134 (MIN); on ind Rainy Lake, xg 1901, Fink 1088 (MIN trees, Koochiching, July 2 ‚ 1901, Fink 886 (M N). BLUE EARTH СО.: on old hee Mankato, June 28, 1899, Fink 165 (MIN). LAKE OF THE WOODS CO.: on old wood, Lake of the Woods, July 1896, Millan 3 (MIN). BELTRAM trees, Red Lake, Aug. 2, 1900, Fink 1049 (MIN); on dead pine logs, Ва pue 4, 1900, Fink 423 (MIN); oaks, Bemidji, July 4, 1900, Fink 447 (MIN); on tamarack, in swamp, Bemidji, July 7, 1900, Fink 525 (MIN); on roeks along Red Lake, Aug. 1, 1900, СЕ 1038 (MIN). YELLOW MEDICINE CO.: on shaded roeks, Granite Pall, July 17, 1899, Fink 586 (MIN). ROSEAU CO.: on trees, Warroad, June 25, 1901, Fink pA (MIN). MARSHALL CO.: on old logs, Thief River, July 19, 1900, Fink i^ (MIN). ROCK CO.: on Mound Creek, Sept. 1901, Wheeler (MIN). COUNTY UNKNOWN: on trees, Flag Island, July 12, 1901, Fink 536 (MIN); on bark, Flag Island, Aug. 1897, Millan ф Sheldon (MIN); on dead pine, 1900, Fink (F). IOWA: BREMER CO.: on trees, July 1894, Fink (MIN). боптн DAKOTA: MEAD CO.: Piedmont, 1892, Pratt (Е). MONTANA: LEWI CLARK CO.: on forest duff and humus, sect. 7-8 T. 12 W. R. 9W., Helena zig d aOR). MISSOULA CO.: Mitten Mountain, near Mis- soula, sey, ae ft., Dec. 1930, Flint (MONT); Pattee Canyon, close to aneri house, n , 1934, Flint (MONT); on denis stem, Pattee Canyon r Mis soula, e mud ft., y 1, 1935, Flint (MONT); on granite бй, sheltered NN Blaek aud эма т , Bitterroot Montis Aug. 14, 1931, Flint (MONT). WYOMIN оок Sundance mises July 3, 1896, A. уйл. (MBG); Hayden re 97 J. М. rus er (US). COLORADO: WELD CO.: over iih roeks and mosses, south St. Vrain Canyon, elev. 7200 ft., Oct. 7, 1936, Kiener (MBG). EW EXIO0: COLFAX CO.: moist bank, vicinity of Ute Park, Aug. 29, 1916, Standley (F). IDAHO: NEZ PERCE CO.: elev. 1000 ft., Apr. 30, 1896, Heller $ Heller (MBG, MIN) BRITISH COLUMBIA: on the old wharf, Comox, Vancouver T June 24, 1893, Macoun санани ,US) ; on trees and old fences, Victoria, May 1893, P (Е): WASHINGTON: ISLAND CO.: on rocks, Langley, May 1920, Grant (WIS); оп ie FR 1923, Grant (WIS) ; on Tsuga log, Oct. 1922, Langley, Grant (WIS) ; n limbs of trees, т 1921, Grant 282 (Е). SAN JUAN CO.: Вап Juan, July 27, 1008 Fink (FM). OREGON: MULTNOMAH CO.: on bark of maple, Portland, Apr. 29, 1906, Foster F) CALIFORNIA: SANTA CRUZ CO.: on sandstone, Devil's Canyon, elev. 2300 ft., July 28, 1905, Herre 702 (MIN). FOREIGN AND LocAL ExsICCATAE EXAMINED: Claud. 17; Elenk. 106a, b; Harm. Lich. Loth. 297 ; Hav. 180; Johns. 69; Malme 230; Merr. 27; Mudd, 66; Salw. 168. [Vou. 28 88 ANNALS OF THE MISSOURI BOTANICAL GARDEN 23. Parmelia Herrei Zahlbr. apud Herre, Proc. Wash. Acad. Sei. 7: 353. 1903. The type was collected on earth and sandstone, Pilarcitos Creek Canyon, 2 miles from the ocean, in San Mateo County, California, by A. C. Herre, and is now in the United States Na- tional Herbarium. Thallus gray to yellowish-gray in herbarium material, lobes long, narrow, tips rounded or wavy, somewhat sorediate, center of thallus becoming imbrieate with the tangling of the lobes, margin fringed with long black cilia, below densely covered with long black rhizinae; apothecia not known; К, none, С, brown, K + C, none, P, none. Algal layer continuous, 15-18 и thick, upper cortex 34-38 и thick, matrix light-colored ; medulla loosely interwoven; lower cortex 18—26 и thick, matrix black; thallus 66-90 y thick. The plant resembles P. saxatilis but is more rugged in ap- pearance. It has been collected only from the type locality and by A. C. Herre. Distribution: Plant unknown except in type locality. CALIFORNIA: SAN MATEO CO.: on earth and sandstone, Pilarcitos Creek Canyon, 2 miles from the ocean, elev. 200 ft., May 28, 1904, Herre (Type collection) (US) ; on earth and sandstone, Pilareitos Creek Canyon, 2 miles from the ocean, elev. 200 ft., June 30, 1906, Herre (MBG,F,MIN,LSU,M). 24. Parmelia Borreri Turn. apud Sm. & Sowerb., Eng. Bot. 25: tab. 1780. 1807. Lichen Borreri Sm. apud Sm. & Sowerb., Eng. Bot. 25: tab. 1780. 1807. Parmelia Borreri var. hypomela Tuck., Syn. N. Amer. Lich. 1: 58. 1882. The plants used as types by both Turner and Smith are un- known, though their material came from Sussex, England. The plates cited by Smith in Smith and Sowerby’s ‘English Bot- tany' 25: ¢. 1780. 1807, and Turner, ‘Transactions Linnean So- ciety London’ 9: tab. 13, f. 2. 1808, are well made and repre- sent the descriptions of the plants. Tuckerman’s type of the variety is a collection made in Illinois and is in his herbarium at the Farlow. Thallus light grayish-green to olive-green, adnate, wrinkled, 1941] BERRY—THE GENUS PARMELIA IN NORTH AMERICA 89 margins of lobes round and usually wavy, from which short flat dichotomously branched sorediate lobules arise, surface of thallus glaucous, black ostioles of the spermagonia prominent on the younger portions of the upper surface; lower surface, dirty white to dark brown, with numerous white or darkening rhizinae; apothecia numerous, sessile or short-pedicellate, 0.5—14 mm. diameter, disks deeply to slightly concave, chestnut- brown to dark brown, margins regular with vertical grooves marking the outside of the amphithecia, occasionally sorediate ; K, upper cortex yellow, medulla none; C, medulla red; K + C, none; P, none. Algal layer continuous, 20-81 и thick; upper cortex 20-24 и thick; medullary hyphae loosely interwoven, hyaline; lower cortex 20-30 и thick; ascus clavate, spores 8, 8-10 x 10-12 и; thecium 60-96 и thick ; thallus 400—600 и thick. Distribution: Vermont to North Carolina, west to British Columbia and California. ONTARIO: on trunks, Algonquin Park, May 27, 1900, Macoun 241 (MBG,US). VERMONT: ADDISON CO.: on Abies canadensis, Monkton, Dec. 2, 1878, Faxon (MBG) MASSACHUSETTS: MIDDLESEX СО: Blue Hills, Nov. 6, 1883, Faxon (WIS). NORFOLK CO.: Weymouth, Willey ood NoRTH CAROLINA: BUNCOMBE CO. es, арсен Feb. 1899, C. Russell (M). MicHIGAN: EMMET CO.: Harbor utis Sept. 1931, 8 (M). NDIANA: UNION CO.: on bark, July 25, 1917, Fink d inh 75 (F). WISCONSIN: BURNETT CO.: valley of the Wiseonsin River near Webster, sum- mer 1894, Cheney (WIS). oNEIDA CO.: Aug. 27, 1936, Peck (MBG). GREEN LAKE co.: on shagbark hickory, above granite outcrop, 2 miles south of Marquette, Sept. 4, Dc (WIS) ILLINOIS: COOK CO.: on oaks, Calkins (NYBG); River Forest, June 22, 1893, Harper "n UT (FM); on bark of trees, Palos Park, May 30, 1900, Harper $ Harper 99 (FM). CHAMPAIGN CO.: on oak Mirum коты Feb. 8, 1879, Earle (NYBG). LASALLE CO.: on o ‘Calkins 2 (F). ARD CO.: on old fences, Б Hall (ЕМ); Athens, y Hall (Type of сони s variety) (MBG). KENTUCKY: MADISON CO.: glade, 10 miles north of Richmond, Aug. 18, 1937, л с (MBG). IINNESOTA: HENNEPIN CO.: On , Lake Harriet, Apr. 1890, Frost (MIN). BLUE EARTH CO.: on trees, Mankato, bob 22, 1899, Fink 11 (MIN). BELTRAMI CO.: on trees, Red Lake, July 26, 1900, Fink 893 (MIN); on pines, AI Е TE 1901, Fink 658 (MIN). MEEKER CO.: bark, Litehfield, June 1892, Frost F46 (MIN). REDWOOD CO.: on E Redwood Falls, July 8, 1899, Tin 335 (MIN). OTTERTAIL on pe n Hills, Nd 27, 1900, Fink 281 (MIN); on trees, ttle T Tite 2 , Fink 43 (MIN). PENNINGTON CO.: on trees, Thief River Falls, July 19, ior E $24 (MIN); Red River, May 1856, Lapham (WIS). [Vor. 28 90 ANNALS OF THE MISSOURI BOTANICAL GARDEN ТОМА: CLAYTON CO.: on trees, е. 1894, Fink (WIS); оп E Sept. 1894, Fink (MBG). JOHNSON co.: Iowa City, 1885, Macbride (IA); uercus alba, March 13, 1896, Fitzpatrick $ Fitzpatrick (FM). FAYETTE CO.: on pisi Sept. 20, 1892, Fink (MBG) MISSOURI: ST. LOUIS со.: bark of tree, Allenton, May 5, 1938, Berry (MBG bark of tree, 2 miles southeast of Centaur ra Jan. 1, 1937, Hubricht 812 (MBG). IRON с0.: on granite rocks, Buford M ‚ Nov. 1897, C. Russell 4 (MBG). BOONE С0.: bark of tree, Sept. 17, 1930, ue (MO, MBG). MORGAN CO.: bark of hickory tree, Oct. 25, 1930, ps (MO,MBG). SALINE CO.: fence rails, Emma, 1892, Demetrio (MBG). JOHNSON со.: bark of tree, Warrensburg, Jan. 3, 1938, Berry (MBG). Мовтн DAKOTA: CASS CO.: on 2 адл, Fargo, June 1914, Brenckle 1015 (Е). SOUTH DAKOTA: BROOKINGS : trees and dead wood, Brookings, June 13, 1891, Luke (MBG); on ier trees, eines Apr. 1895, Williams 155 (MBG,WIS, MIN,FM TEXAS: GILLESPIE CO.: Jermy (MBG). LIMESTONE со.: Grossbeck Mountain, summer 1931, Jones $ Hedrick (M); C. Wright, 1848-50 (T). REAL CO.: shade, Frio Canyon, 11 miles north of Leakey, Мау 23, 1938, Hubricht В??? (MBG). DEN- TON CO.: in group of ** Quercina minor,’’ 3 miles southwest of Denton, Feb. 1935, Bussard V (MBG BRITISH COLUMBIA: on old wharf, Comox, June 24, 1893, Macoun 149 (MBG). CALIFORNIA: SANTA CRUZ CO.: on oaks, Ano Nuevo Creek, Santa Cruz Moun- D FOREIGN EXSICCATAE Bxixnn: Malbr. Lieh. Nor. 269; Harm. Lich, тича id Flag. Lich, Ег.-С. 73; Mus. Krypt. Exs. Vind. 3163; Moug. & Nest. 634; Cum. Lich. Bor. Am. 85, 165. The citation of Parmelia dubia by Schaerer as a synonym of Parmelia Borreri was evidently based on a typographieal error. Lichen dubius Wulf. apud Jacq. (Collectanea 4: 275, t. 19, f. 1. 1790) has powdery soredia on the margins and clearly is not the plant described as P. Borreri by Turner. 25. Parmelia Bolliana Müll.-Arg., Flora 60: 78. 1877. The type was collected by J. Boll in Van Zandt County, Texas, and sent to Müller-Argau for determination. It is now in the Herbier Boissier, Université de Genéve, Switzerland. Thallus greenish-brown, much wrinkled, with reticulate sec- ondary ridges, many little white dots where the upper cortex has been lost and the white medulla is exposed, lobes short and round, margins smooth; under surface white at the margin, be- coming light brown beneath, with thinly scattered, small, light brown rhizinae ; apothecia numerous, 1-2 mm. diameter, sessile, deeply concave to almost flat, disks chestnut-brown ; amphithe- 1941] BERRY—THE GENUS PARMELIA IN NORTH AMERICA 91 eium marked with deep vertical lines or ridges; K, none, C, none, K + C, pink, P, none. Algal layer continuous, 20-25 и thick; upper cortex 20-22 в thick ; medulla loosely interwoven, diameter of hyphae 1.1-2 р; lower cortex 20-25 u thick, matrix light-colored; thallus 400— 570 и thick; ascus clavate, spores 8, 7-9 х 12-14 и; paraphyses branched. Distribution: Missouri, Texas, and Arkansas. MISSOURI: WAYNE СО.: on trees, Clark Mountain, Nov. 1, 1898, C. Russell (MBG). ARKANSAS: BOONE СО.: rocky wooded bluff, above dry ereek bed, Bear Creek Spring, 0.5 mile north of Francis, Apr. 7, 1939, Dodge, Berry § Johnson (MBG). TEXAS: VAN ZANDT CO.: on trees, J. Boll (Co-type) (T). GILLESPIE CO.: Jermy (MBG). BREWSTER CO.: full sun to semi-shade, 18 miles south of Marathon, May 27, 1938, Hubricht B853 (MBG). CULBERSON CO.: full sun to semi-shade, north side of summit, Van Horn Mountains, 15 miles south of Van Horn, June 2, 1938, Hubricht B946 (MBG). 26. Parmelia quercina (Willd.) Wainio, Termeszetr. Fuze- tek 22: 279. 1899. Lichen quercinus Willd., Fl. Berol. Prodr. 353. 1787; Ach., Lich. Suec. Prodr. 124. 1798. Lichen tiliaceus Hoffm., Enum. Lich. 97. 1784. Parmelia tiliacea Ach., Meth. Lich. 215. 1803, not Lobaria tiliacea Hoffm. Imbricaria quercina DC. apud Lam. & DC., Fl. Fr. ed. 3, 2: 390. 1815. Willdenow used Lichen quercinus to designate this lichen in *Florae Berolinensis Prodromus,' 353. 1787. Hoffmann, in his *Deutsehlands Flora? 2: 149. 1795, designates it Lobaria tiliacea and cites his ‘Enumeratio Lichenum,’ f. 16, f. 2. 1784, where the epithet Lichen tiliaceus was used. Hoffmann cites Willdenow and considers Lichen quercinus as synonymous with Lobaria tiliacea. Acharius (Lichenographiae Suecicae Pro- dromus, 124. 1798) uses the epithet Lichen quercinus of Will- denow, considering Hoffmann’s Lichen tiliaceus аз belong- ing to another plant. He questioned the identity of Lobaria tiliacea given by Hoffmann (Deutschl. Fl. 2: 149. 1795) as a synonym of Lichen quercinus Ach. (Lich. Suec. Prodrom. 124. 1798). In his ‘Meth. Lich.’ 215. 1808, Acharius applied the epithet Parmelia tiliacea, citing Lichen quercinus Willd. (Fl. [Vor. 28 92 ANNALS OF THE MISSOURI BOTANICAL GARDEN Berol. Prodr. 353. 1787), but considering the figure and de- scription in Hoffmann's ‘Enum. Lich.’ 1784, as that of Li- chen scorteus. This would make Willdenow's *Florae Bero- linensis Prodromus,' 353. 1787, the oldest valid description of the plant, and his name, quercinus, should be used. Willde- now’s collections were from the vicinity of Berlin, and his type isin Berlin. Acharius’ herbarium, at Helsinki, contains collec- tions of the species which he determined. Except for the plates, there seems to be none of Hoffmann’s material left. DeCandolle (Fl. Fr. ed. 3, 2: 390. 1815) cites Lichen quercinus (Willd., FI. Berol. Prodr. t. 7, f. 13. 1787; and Ach., Lich. Suec. Prodr. 124. 1798) as synonymous with his plant. DeCandolle's speci- mens are in Genéve, Switzerland. Thallus 5-15 cm. diameter, closely adnate, usually wrinkled at the center, greenish-gray, lobes short and round, occasion- ally elongated and subdichotomously branched, margins ir- regular, erenate; apothecium sessile, 4-12 mm. diameter, disk slightly to deeply concave, light chestnut-brown; amphithe- cium entire or dissected; K, upper cortex yellowish-green, medulla greenish ; C, none, K - C, none, P, yellow. Algal layer continuous, 48-52 и thick; upper cortex 10-16 џи thick, matrix yellow-brown; medulla loosely interwoven, hyphae hyaline and of small diameter; lower cortex 40-44 п thick, matrix black; thallus 288-300 и thick; thecium 24-28 и thick; ascus clavate, spores 8, 4-5 x 6-8 и; paraphyses branched, enlarged at tips. Distribution: New Brunswick to Florida, west to Wyoming and California. New BRUNSWICK: on trees, Willey (BSNH). ONTARIO: оп tree trunks in wood, Ottawa, Мау 2, 1900, Macoun (MBG); on tree trunks, Ottawa, June 24, 1898, Macoun (MBG); on trees, Emo, July 18, 1901, Fink 693 (MIN). MAINE: WASHINGTON CO.: June 1929, Markin (WIS). ARoosTooK CO.: St. Francis, Aug. 1893, Cummings 102 (MIN). PISCATAQUIS CO.: on high river bluffs with red maple, poplar, and balsam, Camp Sunday, Medford Township, Aug. 28, 1905, Merrill (NYBG). KNOX CO.: on various trees, Warren, Apr. 30, 1911, Mer- rill 221 (MIN). № HAMPSHIRE: 0005 CO.: Shelburne, Aug. 1894, Farlow toes, ; on pine trees, base of Mount Washington, Aug. 6, 1895, Harper $ Harper VERMONT: ADDISON CO.: on dead wood, Middlebury, Apr. 22, rs Dodge (D). 1941] BERRY—THE GENUS PARMELIA IN NORTH AMERICA 93 RUTLA : on maple Pu Pawlet, Mar. 22, 1913, Dodge (D); Chipman Lake, ET cung Fins 1938, Dodge (D). bosse ери co.: Hazelwood, Маг. 20, 1882, Faxon (WIS). BRIS- , New se 1878, Willey (US); bark of trees, New Barton, aie p MBG) ; w Bedford, ios eet Willey (US); trunks, New Bedford, Willey 32 ads COUNTY INDEFINITE: New England Lichens, Wright (MBG). EW YORK: QU co.: on Acer, East Port, Long Island, June 26, per Schrenk (US). YATES CO.: Penn Yin, Buckley (MBG); Dundee, S. Wright 40 (FM). PENNSYLVANIA: MIFFLIN CO.: Sept. 1879, Rothrock (FM). CHESTER CO.: June 27, Ше Rothrock d Rothrock (FM). MAR ERY CO.: rock outcrop, cliffs near Cabin Hill, Plummers island PUR vs near Cabin John, Nov. 6, 1938, Leonard 2821 (US); base of oaks, West Knob, Plummers Island, Potomae River, near Cabin John, Nov. 6, 1938, Leonard 2775 (US); pes e West Knob, Plummers Island, Potomae River, near Cabin John, Nov. 6, 1938, Leonard 2764 (US). Yo. PRINCE WILLIAM 5 on old walnut logs near ruins of old house, bet П Piga Point vndis and ridge east, Bull Run Mountain, irn 23, 1937, Allard 77682 (US). co.: edge of woods, Hunter's Gap, P ] Mountain, 2.7 miles ОДА of Баева. Aug. 23, 1937, Hubricht B412 (MBG). я NoRTH CAROLINA: HENDERSON СО.: bark of tree, Flat Rock, July 4, 1921, Schal- lert 2026 (FM); base of tree, Grandfather Mountain, June 13, 1936, Schallert (MBG) GEORGIA: THOMAS СО.: deme 1906, Mrs. A. P. Taylor (F). FLORIDA: ORANGE CO.: ving eypress in swamp, йыш, July 1906, Rapp 17 (Е). LEON CO.: 11 pos 1891, Farlow (Е). COUNTY UNKNOWN: Calkins 4 mid ; Upsala, Nov. 1921, Rapp 663 (F). ALABAMA: MOBILE CO.: Mobile, Mohr (US). MARSHALL CO.: open woods, З miles AT of Boaz, us 21, 19 39, Hubricht B1543 (MBG). a ATHENS CO.: on oak, ра 22, 1938, еси 257 (О). PERRY CO.: on tree, Apr. 4, 1936, Wolfe 546 (О). co aple, Mar. 14, 1936, Wolfe 515 (0). Es VAY CO.: on locust, Ng D. 1936, "Wolfe 631 (O). PIKE CO.: on linden, Md 15, 1935, М (NYBG); on linden, Sept. 15, 1935, Wolfe 215 (О). JACK- n dea neh, Nov. 3, 1935, Wolfe 335 (O). CLERMONT CO.: on bark, i p D inim p Hill 610 (O). COUNTY UNKNOWN: on a poke tree, Apr. 12, 1918, Fink (D MICHIGAN: COUNTY UNKNOWN: on trees, Sailors Encampment, Aug. 3, 1897, Harper $ Harper 55 Bones KENTUCKY: LAUREL CO.: deep wet woods, 8.4 miles north of East Bernstadt, Aug. 18, 1937, Hub те B280 (MBG). кхох co.: 1.4 miles north of Jarvis Store, Aug. 19, 1937, Hubricht B316 (MBG). TENNESSEE: HAMILTON CO.: on elms, Pag ын Calkins 4 (FM); on trees, Missionary Ridge, Calkins 1 (MIN). SEVI woods, elev. 3600 ft., “Тһе Loop,’’ 10 шо south of Gatlinburg, Aug. ны n Hubricht B518 (MBG). Wisc co.: valley of the Wisconsin River, near Lac Vieux Desert, RACE Тый, DN 41 (WIS). PORTAGE CO.: valley of the Wisconsin River, near Stevens Point, Summer 1894, Cheney 3498 (WIS). TAYLOR CO.: valley of the Wis- eonsin River near Goodrich, Summer 1894, Cheney (WIS [Vor. 28 94 ANNALS8 OF THE MISSOURI BOTANICAL GARDEN ILLINOIS: LAKE CO.: оп oak, Lake Bluff, May 1906, Harper $ Harper 100 (FM). LASALLE CO.: on trees, Calkins (F,FM). MENARD CO.: on trees, 1878, Н (FM); c 1878, Hall (MIN,MBG,WIS). MINNESOTA: COOK CO.: on cedar, Grand Marais, Aug. 22, E Fink 156 (MIN). LAKE CO.: On dice Beaver Bay, July 14, 1897, Fink 724 (MIN); on trees, Snow- bank Lake area, July 23, 1897, Fink 896 (MIN). вт. LOUIS 00.: on bark of Frazi- nus, Vermillion Lake, July 19, 1886, Bailey 4 Holway A226 (MIN); on trees, Ely, July 28, 1897, Fink 1018 (MIN); on trees, Harding, Aug. 17, 1901, Fink 1560 (MIN); on trees, Tower, Aug. 23, 1901, Fink 1698 (MIN). CARLTON CO.: on trees, Kettle Falls, à 13, 1901, Fink 1417 (MIN); on иу Tafte, Carlton Peak, July 10, 1897, Fink 620 (MIN). KOOCHICHING CO.: on trees, Koochiching, July 26, 1901, Fink 922 ipii ; on trees, Gunflint, July 2, 1897, Fink 407 ( . AITKIN СО. on Pinus Strobus, Mille Laes Indian Reservation, June 1892, Sheldon 82328 (MIN); on Quercus macrocarpa, Mille Lacs Lake, June 1, 1892, Sheldon S2289 MIN). crow WING co.: on Quercus tinctoria, Garrison, June 1892, Sheldon 82236 е pee EARTH CO.: on trees, Mankato, June 23, 1899, Fink 56 (MIN). BEL oaks, Bemidji, July 4, 1900, Fink 449 (MIN); on trees, Red Lake, duy: M 1900, Fink 1008 (MIN); on bin in swamp, Beaudette, June 18, 1901, Fink 9 bagging ARSHALL CO.: on trees, Thief River o July 23, 1900, Fink 874 (MIN). orTERTAIL CO.: on trees, Leaf Hills, June 26, 1900, Fink 208 (MIN); on tes, Ta — dec 18, 1900, Fink 16 (MIN). IOWA: JOHNSON C on К, Jan. 8, 1898, Fink $ Savage (ТА). HENRY CO on bark, ти 1898, аа ied BREMER CO.: on tree ped 29, 1895, end (D). FAYETTE CO.: on па Sept. 1892, Fink ng WIS). MISSOURI: ST. LOUIS CO.: Cliff Cave, Aug. 26, 1899, Norton (US); on Crataegus tomentum, covering "v branehes, Crescent May 1896, = т (MBG) ; bark i oak tree, Ranken Estate, Nov. 20, 1936, Бети (MBG). FRANCOIS CO.: of trees, Piekle Springs, May 21, 1938, dece (MBG). WAYNE 00.: on oak he: Clark Mountain, Mar. 16, 1892, C. Russell (US). BUTLER CO.: on bark of Ostrya, deca Dee. 25, igi C. Russell (MBG № SAS: BAXT dry upland woods, 1 mile west of Big Flat, Apr. 9, 1988, jen Ави уч iid (MBG). SEARCY CO.: open wooded hillside, 4 miles west of Marshall, Apr. 9, 1939, Dodge, Berry $ Johnson (MBG). BOONE co.: rocky wooded bluff above dry creek bed, Bear Creek Spring, 0.5 mile E of Francis, Apr. 7, 1939, Dodge, Berry 4 Johnson (MBG). NEWTON CO.: steep dry wooded bluff with roeky flat top, 1 mile south of Ponea, Apr. 8, 1939, vb em Berry $ John- son (MBG); moist woods on gentle slope near Marble e bes pos south of Harrison, Apr. 8, 1939, Dodge, Berry § Johnson (MBG). on dead oak tree, Carpenters Road, Hot Springs, June 30, 1939, sly руй с MBG); on dead branches, North Mountain, Hot Springs, Dec. 15, 1939, Scully 1407 (MBG TEXAS: GALVESTON CO.: Galveston, Hall 362 Ja WYOMING: CROOK CO.: hapaa Mountain, July 3, 1896, A. Nelson (MBG). CALIFORNIA: SANTA CL CO.: trees and shrubs, summit of Black Mountain, elev, 2785 ft., July 6, de riva (US). RIVERSIDE CO.: on oak, San Jacinto Moun- М AND LOCAL ExsiccaTAE EXAMINED: Claud. 506; Cum. I, 25, 108, п, 3 22; Tuck. TIT & IV, 70; P. quercina var. sublaevigata Bal. 4176, 4210; Spruce, 136, 156; P. quercina var, vicinior Merr. M, теч 1941] 7 BERRY— THE GENUS PARMELIA IN NORTH AMERICA 95 27. Parmelia texana Tuck., Am. Jour. Arts & Sci. II. 25: 424. 1858. Tuckerman’s type was based on Charles Wright 228, from Blanco, Texas. The type is now in the Tuckerman Herbar- m. Thallus light greenish-gray, wrinkled and covered with many small powdery soredia, margins of lobes round and irregular; lower surface black, margins lighter, with papillose rhizinae toward the center ; apothecium sessile, 1-6 mm. diameter, disk concave, chestnut-brown, amphithecium thick, wavy, and more or less sorediate ; K, попе, C, none, K + C, brown, P, none. Algal layer discontinuous, groups of algae 40-42 u in diam- eter ; thickness of upper cortex 20-21 и, of lower cortex 20-21 п; medulla of loosely interwoven hyaline hyphae; thallus 700- 712 и thick; thecium 88-90 y thick, ascus clavate, spores 8, 2-4 x 7—12 и; paraphyses unbranched and slightly inflated at the tips. Distribution: South Carolina to Florida, west to Oklahoma and Texas. SoUTH CAROLINA: COUNTY UNKNOWN: Curtis (NYBG). GEORGIA: THOMAS CO.: AS Aug. 1906, Taylor (F). FLORIDA: ORANGE CO.: oak trunks, Sanford, Nov. 1906, Rapp 19 (F); on Sabal Palmetto, Mar. и 0 Барр (ВРІ,Е); Sanford, Jan. 1911, Rapp 42 (F); on Magnolia glauca, Santora. Apr. 1911, Rapp 39 (F); Sanford, Apr. 1922, Rapp 323 (Е ALABAMA: LEE СО.: on fence rail, Auburn, Mar. 21, 1897, Earle 4: Baker (MBG, MIN). OHIO: BROWN CO.: on fence, Apr. Е 1936, Wolfe $ Hill 605 (О). CLERMONT CO.: on oak, Apr. 18, 1936, Wolfe $ Hill 609 (0). CLINTON 00.: on tree, Apr. 18, 1936 Wolfe $ Hill 626 (О). FAIRFIELD CO.: on rock, Apr. 27, 1935, Wolfe 93 (0); MARION CO.: June 30, 1892, Bogue (O). PIKE CO.: on oak, Sept. 15, 1935, Wolfe 209 (0). m co.: on maple, May 3, 1936, Wolfe $ Hartley 657 (О). VINTON со.: on red m ‚ Mar. 14, 1936, Wolfe 474 (0). WARREN CO.: on hickory, Apr. 18, 1936, ors Hil igs KENTUCKY: CO.: open woods, 1.4 miles north of Jarvis Store, Aug. 10, 1937, de pos B321 (MBG). TENNESSEE: HAMILTON CO.: on rocks, Lookout ара Calkins 204 (NYBG). ARKANSAS: CHICOT CO.: near Macon, Dee. 29, , Anderson (MBG). OKLAHOMA: NOBLE СО.: on dead juniper, ism Bliss, Apr. 18, 1917, Mrs. J. Clemens (MBG). EXAS: HARRIS CO.: on living Quercus Phellos, Houston, Jan. 9, 1939, Fisher 392 (MBG). BLANCO CO.: on dead wood, Blanco, Fink 11,427 (М); Blanco, Wright 228 (Type eolleetion) (T). [Vor. 28 96 ANNALS OF THE MISSOURI BOTANICAL GARDEN 28. Parmelia rudecta Ach., Syn. Meth. Lich. 197. 1814. Parmelia Borreri var. rudecta Tuck., Enum. N. Amer. Lich. 49. 1845. Parmelia leiocarpa Tayl., Hook. London Jour. Bot. 6: 170. 1847. Acharius’ type description was from a collection of Muhlen- berg's made in Pennsylvania. Tuckerman's type of the variety was based on a collection from the White Mountains of New Hampshire, which is in his herbarium at the Farlow. Taylor's P. leiocarpa was based on a North American collection and is now in his herbarium at the Boston Society of Natural History. Parmelia rudiata Eaton (Manual of Botany for North Amer- ica, ed. 6, pt. 2:56. 1833) is a typographical error which ap- peared in the sixth edition and was continued through the eighth edition. Previous to the sixth edition Eaton used Par- melia rudecta. Thallus olive-green or lighter, wrinkled, loosely adnate, older portions eovered with confluent soredia, granular isidia, or eoralloid branchlets, lobes round, margins entire or slightly disseeted ; lower surface white to light brown or darker; apo- thecium 3-5 mm. diameter, sessile, disk deeply concave, chest- nut-brown, amphithecium entire, wavy or irregular; K, upper cortex yellow, medulla none; C, upper cortex none, medulla red; К + C, yellow; P, gray. Algal layer continuous, 42—45 р thick; thickness of upper cortex 18-23 и; medullary hyphae loosely interwoven, large, 3—4 y in diameter ; thickness of lower cortex 65-84 y, thickness of thallus 400-810 и; thecium 60-67 и thick, ascus clavate, spores 8, 8-10 x 10-12 y. Distribution: Nova Scotia to Florida, west to Washington. Nova Scotia: on tree, Sandy Cove, Aug. 1908, Lowe (Е); Rothrock (ЕМ). QUEBEC: on oak trunk, Hull, Sept. 28, 1907, Macoun (F); adherent to rocks, Ste. Anne, May 2, 1935, Lepage 137 (D) ; on trees, Ste. Anne, May 5, 1935, Lepage 142 (D). ONTARIO: on bark of trees, Ottawa, Apr. 20, 1891, Macoun 85 (MIN,MBG,WIS) ; on trunks and old rails, Ottawa, Oct. 15, 1893, Macoun 40 (MBG); on ground, fallen trees, base of trees, in shaded peris of woods, Rainy Lake, Aug. 20, 1930, Schmidt 11 (US); on rocks, Emo, July 20, 1901, Fink 777 (MIN). MAINE: HANCOCK CO.: on paths to cliffs, Seal Harbor, Sept. 20, 1894, Merrill 1941] | BERRY—THE GENUS PARMELIA IN NORTH AMERICA 97 (D). KNOX co.: on old birch trunks, Thomaston, Apr. 1912, Merrill 2 (MBG). OXFORD CO.: on black ash, Buckfield, July 16, 1924, Parlin 7419 (F). NEW HAMPSHIRE: Coos CO.: White Mountains, Tuckerman (Туре of P. Borreri var. rudecta) (Т). VERMONT: LAMOILLE CO.: on maple tree, Cambridge, June 21-24, 1922, Dutton 1713 (MBG). WINDHAM CO.: on apple tree, Pike Hollow, Wardsboro, July 3, 1935, Moore (MBG). ADDISON CO.: over ledges, Mud Pond, Leicester, Aug. 28, 1922, Dutton 1878 (F). RUTLAND CO.: on cedar tree in swamp, elev. 475 ft., Brandon, Feb. 26, 1921, end (WIS,FM). iA CE IUSETTS: BARNSTABLE CO.: on limb of white oak, West Falmouth, Apr. 11, Midi Linder (F). BRISTOL CO.: on trunks, New преса еи 26 (МВ CONNECTICUT: WINDHAM CO.: on trees, Woodstock, Oct. , Evans (F). NEW YoRK: QUEENS CO.: on trees, Long Island, 1914, MM "7 (Е). ҮАТЕЅ C0. : diim Samuel Wright (FM). PENNSYLVANIA: CHESTER CO.: on chestnut rail, Nov. 27, 1897, Rothrock 4: Roth- rock (FM). LANCASTER CO.: along fence bed ae ee Mar. 14, , Ебу MBG); along mill creek, Lancaster, Nov. 22, 1892, Heller (F); on red cedar trees, Bear Town, Sept. 16, 1895, Eby (US); on trees, err к. 23, 1994, Eby (US); on tree, Bear Tomi: July 11, 1894, Eby (MBG); on tree, ти 1896 Eby PUMA PIKE CO.: Rae Sept. 24, 1896, Schneider (MB YLA PR GEORGES ase of juniper tree, pine Pie. southeast of Decatur Em реки Sept. 26, 1938, Leonard (US). MONTGOMERY CO.: logs and bases of trees, Takoma Park, 1896, Williams (MBG,MIN,WIS) ; hickory tree, Plummers Island, Nov. 6, 1938, ida (US). I CAROLINA: STOKES C irk of trees, Germantown Road, Mar. 1, 1922, proi 971 (WIS). FORSYTH CO.: Bennetts Rock, Apr. 10, 1936, Schallert (MBG SouTH CAROLINA: COUNTY UNKNOWN: Ravenel (US). GEORGIA: THOMAS CO.: Thomasville, Aug. 17, 1906, Taylor (F). FLORIDA: ORANGE CO.: on trunks of trees, Sanford, Mar. 1910, Rapp 40.11 (F). ABAMA: MARSHALL CO.: open woods, 3.2 miles north of Boaz, July 21, 1939, Hubricht B1546 (MBG). MISSISSIPPI: DE SOTO CO.: 2 miles north of Coldwater, Dec. 28, 1938, Anderson d OHIO: VINTON CO.: on red maple, p 14, 1936, Wolfe 474 (O). HOCKING со.: on tree trunk, Dee. 9, 1934, Wolfe (NYBG); PIKE CO.: on oak, Sept. 15, 1935, Decl 209 (O). FRANKLIN CO.: Geo eas zu T0 = Bogue (W). UN maple, May 3, 1936, ша 657 (O). CLARK со.: p diss Feb. 9, E pes come (W). CLINTON CO.: on tree, Apr. 18, 1936 Oran 6 (O). BROWN CO.: on fence, Apr. 18, io ЕА 605 ae FAIRFIELD CO.: on iid Apr. 27, 1935, Wolfe 93 (O). CLERMONT CO.: k, Apr. 18, е Wolfe 609 (O). WARREN CO.: on hickory, Apr. 18, 1936, T Hill 624 (O WEST VIRGINIA: LOGAN CO.: on Peach m Pulaski, Cummings (W). MICHIGAN: CHEBOY 0: on Betula, Pap Island, vicinity of Burt and Douglas lakes, June-Aug. cae poss (F). INDIANA: FRANKLIN CO.: on wood near Brookville, July 25, 1917, Fink (D). KNOX 00.: edge of woods, 1.3 miles east of Friehton, Aug. 16, 1937, Hubricht B161 MBG) [Vor. 28 98 ANNALS OF THE MISSOURI BOTANICAL GARDEN KENTUCKY: BALLARD CO.: Kevil, Apr. 1936, Fassett (WIS). ROCKCASTLE CO.: bark of trees, open pine woods near Rockcastle River east of Livingston, Aug. 18, 1937, Hubricht B238 (MBG). TENNESSEE: SEVIER CO.: semi-shade, boulders in stream bed, elev. 3600 ft., * The Loop,’’ 10 miles south of Gatlinburg, Aug. 25, 1937, Hubricht B507 (MBG). HAMILTON CO.: on pine tree, Lookout Mountain, Calkins 3 (F). WISCONSIN: DODGE CO.: northeast of Kekoskee, May 1936, J. Thomson (WIS). WAUSHARA CO.: sand plains south of Plainfield, Oct. 13, 1935, J. Thomson (WIS). MARQUETTE CO.: 12 miles east of Westfield, Oct. 13, 1935, J. Thomson (WIS). COLUMBIA CO.: Black Hawk Lookout, opposite Prairie du Sae, Sept. 28, 1935, J. Thomson (WIS). ADAMS CO.: Leola, Oct. 27, 1935, J. Thomson (WIS); bluffs of the Wisconsin River, Oct. 6, 1935, J. Thomson (WIS). BURNETT CO.: valley of the Wisconsin River, near Webster, summer 1894, Cheney (WIS ILLINOIS: LA SALLE CO.: on trees, Calkins 401 (F). MENARD I on fences and old wood, 1878, Hall (FM). couNTY UNKNOWN: on limbs, Mar. 9, 1879, Earle (NYBG). MINNESOTA: COOK CO.: on rocks, Grand Portage, June 24, 1897, Fink 188 (MIN). LAKE со.: on wood, Snowbank Lake, July 20, 1897, Fink 839 (MIN). ST. LOUIS CO.: on rocks, Ely, July S Бази eg 983 (MIN); on eedars, Tower, Aug. 26, 1901, Fink (MIN). on trees, Tafte, Carlton Pea July 10, 1897, Fink 612 (MIN); on вен "Kettle ' Falls, Aug. 12, 1901, Fink 1895 (MIN). KOOCHICHING CO.: on rocks, Rainy Lake City, Aug. 3, 1901, Fink 1136 (MIN); on eedars, Koochiching, July 29, 1901, Fink 989 (MIN); on rocks, gn flint, July 1, 1897, Fink 369 (MIN). AITKIN CO.: on roeks, Palisades, July 1 1897, Fink 744 (MIN); on Acer saccharum, Mille Laes Lake, June 1892, Shel ax 82622 (MIN). BLUE EARTH CO.: on old wood and trees, Mankato, June 22, 1899, Fink 12 (МТМ); on trees, Mankato, June 23, 1899, Fink 58 (MIN). BELTRAMI CO.: on cedars in swamps, Beaudette, June 22, 1901, Fink 114 (MIN); on shaded gran- ite, Bemidji, July 1900, Fink 674 (MIN); on cedars in swamps, Beaudette, June 22, 1901, Fink 111 (MIN). YELLOW MEDICINE CO.: on rocks and near trees, Granite Falls, J "d 11, 1899, Fink 383 o ). ROSEAU CO.: on cedars in swamps, Warroad, June 29, 1901, Fink 267 (M IOWA: IOWA. об on bark of various mu north exposure, moist canyon Тома A EN May 30, 1937, Kiener (MBG). OURI: ST. CHARLES CO.: bark of tree, bluff = Missouri River, Mar. 12, ibaa, tia (MBG). вт. LOUIS CO.: on oak tree, Centaur, 5 miles beyond Clay- ton, Apr. 24, 1898, Ferguson (MBG) ; on boulders, erae Estate, Sept. 28, 1935, Moore (MBG). JEFFERSON CO.: on Juniperus, Herculaneum, Oct. 25, 1935, Moore MBG). ST. FRANCOIS CO.: bark of trees, Pickle Springs, May 21, 1938, Berry (MBG). IRON со.: Iron Mountain, July 17, 1915, Emig (MBG). FRANKLIN CO. : bark of tree, Missouri Botanieal Garden Arboretum, Gray Summit, May 2, 1938, Berry (MBG); Boat Rock, Oct. 20, 1935, Moore (MBG); Elmont, May 20, 1914, Pes. (MBG). CALLAWAY CO.: bark of tree, Sept. 25, 1930, Berr кс па -— MILLER : bark of т. рер Creek, Iberia, ния (MBG). on ia wooded bank of James River, 8 miles south of Spent, rer А Bes Dodge, Berry $ Johnson (MBG). JOHNSON CO.: bark of tree, Knobnoster, June 9, — Ви (М KANSAS: WHITE CO.: near Judsonia, Dec. 28, 1938, Anderson (MBG). NEW- TON со.: top of dry wooded bluff with rocky flat top, 1 mile south of Ponea, Apr. 8, 1941] BERRY—THE GENUS PARMELIA IN NORTH AMERICA 99 1939, Dodge, Berry $ Johnson (MBG); high bluff and long steep wooded hill- side, Lookout Point, 7 miles south of Jasper, Apr. 8, 1939, Dodge, Berry $ Johnson (MBG); steep wooded slope with limestone bluff at top, 4 miles south of Boxley, Apr. 8, 1939, Dodge, Berry ф Johnson (MBG). BOONE CO.: rocky wooded bluff Pu creek bed, Bear Creek Spring, 0.5 mile north of Francis, Apr. 7, 1939, Dodge, erry 4 Johnson (MBG). AE HARRIS CO.: Houston, Feb. 24, um Fisher T4 DE tree, Harris- burg, Feb. 17, 1875, Joor кйш (MBG). BEXAR esquite woods, 1.5 miles southeast of junction of Jolly Lane and Atkins- GEM Lane, 18 miles east of San Antonio, May 17, 1939, pun B1887 (MBG). WASHINGTON: JEFFERSON CO.: Olympic Mountains, 1915, Grant (Е). 29. Parmelia reticulata Tayl. apud Mack., Fl. Hibern. 2: 148. 1836. Taylor's type was a collection from the rocks near Dunker- ron, County Kerry, Ireland. Thetypeisin Taylor's Herbarium at the Boston Society of Natural History, Boston. Thallus large, 15-20 cm. diameter; upper surface smooth, margins rarely sorediate, without cilia, rhizinae extending from the lower side, giving a superficial appearance of mar- einal cilia; lobes round and slightly dissected; lower surface brown at the margins, black with many black rhizinae at the center; apothecia rare, sessile, 2-6 mm. diameter, disks con- cave, chestnut-brown, margins of amphithecia smooth and reg- ular; K, upper cortex yellow, medulla red-brown; C, none; К + С, none; P, none Algal lave Codi qe 160-180 и thick; upper cortex 36- 40 и thick, matrix light brown or yellowish; DAS SIT. of loosely interwoven hyphae of small diameter and of one type; lower cortex 48-60 и thick, matrix black; thickness of thallus 304— 350 и; thecium 80-90 y thick, ascus clavate, spores 8, 6-8 x 12- 16 и; paraphyses branched, enlarged at the tips. Distribution: Pennsylvania to North Carolina, west to Min- nesota. PENNSYLVANIA: LANCASTER CO.: Small (MBG); Russell Hill, July 1895, Eby BG). NoRTH CAROLINA: HENDERSON CO.: on the bark of trees, Flat Roek, Mar. 1, 1921, сеа (MBG). TENNESSEE: HAMILTON CO.: Lookout Mountain, Calkins (MBG). MINNESOTA: YELLOW MEDICINE CO.: on shaded rocks, Granite Falls, July 17, 1899, Fink 574 (MIN); on rocks, Redwood ‘Falls, July 8, 1899, Fink 360 (MIN). [Vor. 28 100 ANNALS OF THE MISSOURI BOTANICAL GARDEN 30. Parmelia cetrata Ach., Syn. Meth. Lich. 198. 1814. Acharius based his description on collections from North America, probably Muhlenberg's from Pennsylvania. The type is in Helsinki, Finland, and there is a co-type in Upsala, Sweden. Thallus large, 10-20 cm. diameter, loosely adnate ; upper sur- face wrinkled, but unbroken except at the margins where it is usually sorediate and somewhat ciliate, lobes short and wide or somewhat elongated and branched, usually ascending to- ward the margins; lower surface black with numerous black rhizinae, margins brown; apothecia rare, 4-15 mm. diameter, subpedicellate, disks deeply сопсауе, chestnut-brown, com- monly perforate at the center, amphithecia entire or fissured, regular or crenulate ; K, none, C, попе, К + C, none, P, none. Algal layer continuous, 15-21 и thick; upper cortex 12-15 y with a yellow matrix; medulla of small, loosely interwoven, hyaline hyphae of one type; lower cortex 9-20 и thick, matrix black or dark brown; thallus 165-180 и thick; thecium 24-26 1 thick, ascus clavate or oval, 8-spored, spores 6-7 x 7-9 y. Distribution: New Hampshire to Florida, west to Alaska and British Columbia. New HAMPSHIRE: CARROLL CO.: on bark of very old Fagus in deep woods, Jaf- frey, Aug. 1918, Riddle (F). NEW YORK: COUNTY UNKNOWN: Port Jervis, Buckley (US). лу: MONTGOMERY CO.: on eedar bark, Plummers Island, Aug. 18, 1907, k (M); dead limb, west side of Cabin Hill, Plummers Island in the Potomae ei near Cabin John, Nov. 6, 1938, Leonard (US). Тивтълог ог COLUMBIA: on rotten log, wooded bank of Rock Creek, northern eorner of D. C., Apr. 25, 1908, Mary Miller (US). VIRGINIA: COUNTY UNKNOWN: July 17, 1891, Seymour 1591 (F). NORTH CAROLINA: HENDERSON CO.: on the bark of trees, Flat Rock on War- burg Road, June 1, 1922, Schallert (MBG,FM,WIS). MITCHELL CO.: on tree roots, Roan Mountain, June = л Schallert (MBG). GEORGIA: THOMAS CO.: ville, Aug. 1906, E. B. Taylor (F). DE KALB CO.: on the summit of 2 Mountain, elev. 1686 ft., Small (M). FLORIDA: ORANGE rees, Sanford, Mar. 1922, Rapp (BPI); on trees, D. vicinity of Oviedo, May 1998, pd (BP ALABAMA Auburn, Jan. 9, 1897, Earle $ Baker (U MISSISSIPPI: HANCOCK СО.: on pine bark, Bay of St. Louis, Mar. 28, 1885, Lang- lois pate NA: RTIN CO.: on trunk of tulip tree, near swamp, St. Martins- die ває 22, we. ни (NYBG); Dec. 7, 1898, Langlois (US). 1941] BERRY—THE GENUS PARMELIA IN NORTH AMERICA 101 OHIO: CHAMPAIGN CO.: in cedar swamp, Werner L426 (O). FRANKLIN CO.: on sycamore trunk, Slate Run, o a Nov. 7, 1907, H.S.B. (O). KENTUCKY: BELL CO.: woods, north side of Pine Mountain, 5.5 miles east of Pineville, pos 19, 1937, bro B337 (MBG). MADISON CO.: glade, 10 miles north of Richmond, Aug. 18, 1937, Hubricht B200 (MBG). ROCKCASTLE CO.: open oods, 3.3 miles south of 7 on Madison-Rockeastle county line, Aug. 18, 1937, Hubricht B227 (MBG); bark o open pine hd near Rockeastle River, east of Livingston, Aug. 18, 1937, Pies B253 (MBG). TENNESSEE: COUNTY UNKNOWN: on roeks, Calkins (FM). ISCONSIN: LINCOLN CO.: on trees, Merrill Springs, May 30, 1893, Head 4 Buell. (WIS) ILLINOIS: JACKSON CO.: on sandstone boulders near Bat Cave, 5 miles south- west of Pomona, Apr. 14, 1940, Hubricht B1736 (MBG). LA SALLE СО.: on bark of trees, Oia (FM). соок co.: on trees, Calkins (FM MINNESOTA: KOOCHICHING CO.: on roe unflint, cay 1, 1897, Fink 362 (MIN). BELTRAMI CO.: on oak, Bemidji, “Fay d 2, 1900, Fink 691 (MIN). REDWOOD CO.: on cedar, Redwood Falls, July 8, 1899, Fink 329 (MIN). YEL- LOW MEDICINE CO.: on shaded rocks, Granite Falls, July 17, 1899, Fink 573 (MIN Towa: CLAYTON CO.: on trees, ius 1894, Fink (MIN); on trees, July 1897, Freuring (MBG). FAYETTE CO.: on trees, ion 1894, Fink (MBG). EMMETT СО.: on base of tree, bank of Mud Lake, bu 17, 1917, Walden (BPI). MISSOURI: ST. FRANCOIS CO.: on bark of tree, open woods, Pickle Springs, May 21, 1938, Berry (MBG). LINCOLN CO.: Juy 1897, Freuring (MBG). GREENE CO.: rocky wooded bank of James River, 8 miles south of Springfield, Apr. 7, 1939, PAM Berry ф Johnson (MBG). ARKANSAS: GARLAND CO: da ridge, water works, Hot Springs, June 10, 1940, Scully (MBG). BAXTER dry upland woods, 1 mile west of Big Flat, Apr. 8, 1939, Dodge, Berry М pouf (MBG). SEARCY CO.: open wooded hillside, 4 miles west of Marshall, Apr. 9, 1939, Dodge, ES d Johnson (MBG). BOONE CO.: rocky wooded bluff, above e creek b Bear En Spring, 0.5 mile north of Francis, Apr. 7, 1939, Dodge, Berry x SUMUS (MBG). NEWTON со.: ius bluff and long steep wooded hillside, v qund Point, 7 umi south of Jasper, А8; 9, Dodge, Berry $ Johnson (MBG); moist woods on gentle slope near и. Falls, 11 miles south of Harrison, Apr. 8, 1939, Dodge, Berry $ Johnson (MBG); steep, dry wooded bluff with rocky flat top, 1 mile south of Ponca, Apr. 8, 1939, DOO Berry ф Johnson (MBG). OKLAHOMA: MURRAY CO.: Arbuckle Mountains, Nov. 29, 1914, Emig (US). MC CURTAIN CO.: on dead a grassy lake, 13 miles southeast of Broken Bow, Stratton 78 (M) TEXAS: BRAZOS CO.: on oak, 8 miles south of College Station, Apr. 21, 1936, did sous GILLESPIE CO.: Jermy (MBG) ALA Juneau, Dec. 5, 1924, Fink (M). pees еее New Westminster, Jan. 1903, Macoun (Е). 30a. Parmelia cetrata var. subisidiosa Miill.-Arg., Engler's Bot. Jahrb. 15: 256. 1894 The type of the variety was from Cuba, and is in Geneva, Switzerland. [Vor. 28 102 ANNALS OF THE MISSOURI BOTANICAL GARDEN Margin of thallus beset with heavy black cilia; lower surface blaek with heavy black rhizinae; spore measurements, inter- nal morphology of the thallus, and chemical reactions same as in the species. Distribution: South Louisiana. LOUISIANA: NEW ORLEANS CO.: on hardwoods near Indian Village, New Or- leans, Dec. 31, 1931, Dodge (D). 31. Parmelia Hubrichtii Berry, sp. nov. Type: Missouri Botanical Garden Herbarium, No. 154725. Type locality: open woods, 1.4 miles north of Jarvis Store, Knox County, Kentucky. Thallus 2-8 em. diameter, loosely adnate to bark of trees; lobes small, round, dissected and somewhat imbricated, margins flat, light greenish-gray becoming darker at the center; upper cortex slightly wrinkled except at the margins, completely covered with small isidia; lower surface dark olivaceous- brown to black with a lighter margin, rhizinae black, numerous and long at the center of the thallus, at the margins smaller but not papillose; apothecium sessile, 1-5 mm. diameter, deeply concave, disk chestnut-brown, amphithecium covered with isidia or coralloid branchlets; K, upper cortex yellowish- green, medulla unchanged, C, none, K + C, none, P, none. Algal layer continuous, 25-40 и thick; upper cortex 12-16 и thick, matrix yellow ; medulla loosely interwoven, hyphae hya- line and of one type; lower cortex 25-30 y thick, matrix black ; thickness of thallus 120-150 и; thecium 52-56 и thick; ascus ovate-clavate, spores 8, 8-9 x 12-16; paraphyses branched, tips with oval enlargements ; spermagonia spherical, immersed, upper half of walls brown, lower half hyaline, 128 y diameter, spermatia slender and straight. Thallus 2-8 em. diametro, adnatus ad corticem ; laciniae min- imae rotundae, marginibus planis undulato-crenulatis et parvis isidiis tectis; superficies supra viridis, subtus nigra, rhizinis multis nigris; stratum superior 12-16 y crassitudine; medulla laxe retieulata, hyphis hyalinis ; stratum gonidiale continuum, 25-40 и erassitudine; algae protococcideae; stratum inferior 25-30 и crassitudine; thallus 120-150 и crassitudine; stratum 1941] BERRY—THE GENUS PARMELIA IN NORTH AMERICA 103 superior KOH addito flavo-virescens, medulla non mutans, CaOCls addito stratum superior et medulla non mutantes; KOH et CaOCle addito non mutans; apothecium 1-5 mm. di- ametro, sessile, discus concavus rufo-brunneus, amphithecium ramulis coralloideis vel isidiis ramosis tectum; thecium 52-56 и erassitudine; asci ovato-clavati, sporae 8-nae, 8-9 y crassi- tudine, 12-16 и longitudine, ellipsoideae; paraphyses ramosi, hyalini, apieibus globosis; spermagonia immersa, globosa, muris brunneis, 128 и diametro, spermatia tenuia et recta. Distribution: Kentucky to Alabama and west to Arkansas and Texas. MA: MARSHALL CO.: open woods, 3.2 miles north of Boaz, July 21, 1939, nn B1545 (MBG). MISSISSIPPI: MONROE CO.: 4 miles south of Nettleton, July 16, 1939, Hubricht В1532 Сана KENT open s, 1.4 ei north of Jarvis cai n 19, 1937, pons Е por E (MBG). ROCKCASTLE CO. n bark of trees, open pine woods, near Rockcastle River, east of Livingston, T 18, Du RKANSAS: NEWTON CO.: dry wooded hillside, 9 miles south of Boxley, Boston Mountains, Apr. 8, 1939, Dodge, Berry $ Johnson (MBG). coNWwAY co.: Рей 7 a T E E [2] in ч Len] в о 3 > A te © TEXA 0.: Jer (MBG). 0.: mesquite woods, 18 miles ent of San Antonio, May 17, 1940, си ie (MBG). 32. Parmelia erecta Berry, sp. nov. Type: Missouri Botanical Garden Herbarium, No. 154724. Type locality: moist woods on gentle slope near Marble Falls, 11 miles south of Harrison, Newton County, Arkansas. Thallus large, 10-20 em. diameter, loosely adnate to bark of trees and bushes, habit of growth distinctly fruticose; upper surface smooth, margins crenulate, not sorediate, and beset with black cilia, but always thin and without a break between the upper and lower cortex; lobes wide, short and rounded, ir- regularly incised, light mineral-gray when dry, becoming erape-green when wet, younger portion covered with black ostioles of spermagonia; lower surface finely reticulate on raised portions, margins colored like the upper surface, when in contact with the substratum dark brown to black and thickly covered with short black rhizinae; apothecia numerous, 1- 22 mm. diameter, subpedicellate and usually perforate when [Vor. 28 104 ANNALS OF THE MISSOURI BOTANICAL GARDEN mature, disk flat, light brown, margin of amphithecium ir- regularly crenulate and smooth; К, none, C, yellow, К + С, yellow, P, yellow. Algal layer continuous, 21-22 y thick; upper cortex 18-20 и thick; lower cortex 27—30 y thick at the center of the thallus and 15-23 р thick at the margins, matrix black; thallus 515- 530 p thick; thecium 30-35 y thick, ascus clavate, spores 8, 5-6 x 8-9 и; paraphyses simply branched; spermagonia spherical, 185-192 y diameter, wall light-colored or hyaline below, dark above, spermatia straight and pointed. Thallus laxe adnatus, pallidus aridus, vito-viridis humidus, juventate habitu fruticosissimo, lobi marginibus rotundis, sub- lati, rugosi, nigro-ciliati, partes juniores loborum ostiolis nigris spermagoniorum tectae; cortex superior 18-22 и cras- situdine, pseudoparenchymaticus, non sorediatus sed semper tenuis et sine ruptis inter illum et inferiorem corticem ; cortex inferior minute reticulatus marginibus viridibus ut in cortice superiori, dilutior siecatus, rhizinis nigris tectus; medulla hyphis hyalinis dispersis; stratum gonidiale 21-22 и сгаз- situdine; cortex inferior cellulis pachydermaticis, nigrescens, 27—30 и crassitudine ; apothecia numerosa, sessilia, rotunda vel angulosa, 1-22 mm. diametro, marginibus non ciliatis, gonidiis sub exeipulo et hymenio; thecium 30-35 и crassitudine, asci clavati, octospori, зрогае 5-6 и erassitudine, 8-9 и longitudine, hyalinae; spermagonia numerosa, immersa, globosa, 185-192 и. Distribution: North Carolina to Georgia, west to Texas. NoRTH CAROLINA: COUNTY UNKNOWN: on bark, Oct. 23, 1932, Schallert (MBG). YADKIN с0.: on tree branch, Lime Rock, May 12, 1936, Schallert ALABAMA: LEE CO.: on limbs, Auburn, Jan. 1, 1897, Earle $ Baker (MIN). LOUISIANA: ST. MARTIN CO.: on trees, St. Мыя, 189 sal denm (MIN). MISSISSIPPI: MONROE CO.: 4 miles south of Nettleton, July 16, 1939, Hubricht MBG). ENTUCKY: MADISON CO.: glade, 10 miles north of Richmond, Aug. 18, 1937, Hubricht B1406 (MBG). OX CO.: open woods, 1.4 miles north of Jarvis Store, Aug. 19, 1937, Hubricht ве "i (MBG). TENNESSEE: HAMILTON 00.: on bark, р gd Calkins 284 (MIN). ILLINOIS: COUNTY UNKNOWN: on trees, 1878, Hall (FM). MISSOURI: ST. FRANCOIS CO.: bark of tree, Pickle Springs, May 21, 1938, Berry (MBG). IRON co.: Royal Gorge, 3 miles east of Arcadia, Nov. 19, 1939, Hubricht 1941] BERRY—THE GENUS PARMELIA IN NORTH AMERICA 105 B1726 (MBG). WAYNE co.: old trees, Williamsville, Mar. 1898, C. Russell (MBG) ARKANSAS: IZARD CO.: on bark of trees, rocky hillside, 10 miles north of Melbourne, Apr. 9, 1939, Dodge, Berry $ Johnson (MBG). SEARCY 00.: on bark trees, open wooded hillside, 4 miles west of Marshall, Apr. 9, 1939, Dodge, Berry ф Johnson (MBG). BOONE CO.: on bark of tree, rocky wooded bluff above dry ereek bed, Bear Creek Spring, 0.5 mile north of Francis, Apr. 7, 1939, Berry ф Johnson (MBG). NEWTON CO.: on bark of tree, moist woods on dig slope near Marble Falls, 11 miles south of Harrison, Вера Mountains, Apr. 8 1939, Dodge, Berry Ф digi on (Type M e (MBG); on branches of а dien tree, high bluff and long steep wooded лд, Lookout Point, 7 miles south of Jasper, Boston Mountains, yee E 1939, Dodge, Berry $ Johnson (MBG); on пр ра steep dry wooded bluff with roeky flat top, 1 mile south of Ponea, Apr. 9, Dodge, Berry $ Johnson (MBG); dry bc hillside, 9 miles south of Md Apr 8, 1939, Dodge, Berry $ Johnson (М TEXAS: WALLER CO.: on living Xanthozylum, Hemstead, elev. E ft., Feb. 15, 1939, Fisher 3928 (MBO). BEXAR CO.: mesquite woods, 1.5 miles southeast of junction ыы Jolly Lane and три Elmendorf Lane, 18 miles east of San Antonio, 1940, Hubricht B1889 (MBG). BRAZORIA CO.: on oak fence post, Brazoria, elev. 50 ft., Feb. 8, 1939, Fisher n (MBG); on dead branches of Quercus Ls 5 West Columbia, elev. 40 ft, Feb. 8, 1939, Fisher 3914 (MBG). BRAZOS CO.: dead branches of trees, 8 EA south of сам Station, Apr. 21, 1936, Hubrioht (MBG). GILLESPIE с0.: Jermy (MBG). AL CO.: shade, Frio баў, 11 miles north of Leakey, May 23, 4 Hubricht pue (MBG). 33. Parmelia livida Tayl. Hook. London Jour. Bot. 4: 171. 1847 The type was collected near New Orleans, Louisiana, and is now in the Taylor Herbarium at the Boston Society of Natural History. The distribution of this species is almost entirely tropical. It has been collected abundantly in Central America and north- ern South America. Thallus small, adnate, upper surface smooth, shining, light gray to greenish-gray; lobes long, narrow, subdichotomously branched, margins smooth, closely adhering to the bark of trees; lower surface black with numerous black rhizinae which extend beyond the margin of the thallus; apothecia rare, small, 4—8 mm. diameter, disks flat, dark brown or blackish, amphithe- eia smooth, margins unbroken by fissures; K, none, C, me- dulla orange-brown, K + C, brown, P, orange. Algal layer continuous, 12-15 и thick; upper cortex 9-12 и thick, matrix yellow or light brown; medulla loosely inter- я [Vor. 28 106 ANNALS OF THE MISSOURI BOTANICAL GARDEN woven, hyphae hyaline and of one type; lower cortex 12-14 и thick, matrix black; thallus 120—155 и thick; thecium 45-60 и thick; ascus cylindrical, spores 8 in a single row, 4—6 x 9-12 р; paraphyses branched, very thin and pointed. Distribution: Florida to Louisiana. FLORIDA: ALACHUA CO.: red maple, near Gainesville, Mar. 14, 1938, Murrill (MBG). DUVAL co.: Jacksonville, Calkins (MBG). ORANGE CO.: on trees, San- ford, Mar. 1918, Rapp 641 (F). LOUISIANA: ORLEANS CO.: New Orleans, Taylor (Type collection) (BSNH). ST. LUCIE CO.: Bois Levert, July 12, 1894, Langlois (US). 34. Parmelia Finkii Zahl. apud Hedr., Mycologia 26: 162. 1934. The type was collected by C. Russell at Williamsville, Wayne County, Missouri, and is No. 8943, Herbarium of University of Michigan. Thallus small, 20-25 mm. diameter, closely adnate to bark of trees; upper surface granular with small soredia, isidia or eoralloid branchlets; upper cortex cracked into areolate sec- tions, small irregular breaks being left by the detachment of the isidia ; lower surface black with many short black rhizinae; apothecium small, 0.4-3 mm. diameter, disk concave, chestnut- brown, margin of amphithecium dissected or formed by flat lobes which are covered with soredia and isidioid branchlets ; К, none, C, medulla faint red; К + C, red; P, none. Algal layer continuous, 16-24 и thick; upper cortex 20-25 и thick; medulla loosely interwoven, hyphae of small diameter and of one type; lower cortex 40—45 y thick, matrix dark brown or black; thickness of thallus 140—200 и; thecium 40—45 и thick ; ascus clavate, spores 8, 5-6 x 8-10 и; paraphyses branched, en- larged at the tips. Distribution: Found only in Missouri. MISSOURI: WAYNE CO.: on bark, Williamsville, Mar. 1898, C. Russell (Type collection) (М); on ealeareous rocks, Williamsville, Mar. 3, 1898, C. Russell 119 (MBG). JEFFERSON CO.: on Juniperus virginianus, Vineland, June 1898, C. Rus- sell (MBG). 35. Parmelia caroliniana Nyl., Flora 68: 614. 1885. The type was from Aiken, South Carolina, and is Ravenel 404 in Tuckerman's Herbarium. 1941] BERRY—THE GENUS PARMELIA IN NORTH AMERICA 107 Thallus light grayish-green, wrinkled, covered with coral- loid branchlets, lobes narrow, with dissected wavy margins; lower surface dark brown to black, with scattered rhizinae; apothecium 2-4 mm. diameter, disk concave when young, flat when mature, light brown, amphithecium with isidia or branch- lets; K, brown, C, yellow, K + C, brown, P, none. Algal layer discontinuous, groups of algal cells 6-15 и in diameter ; upper cortex 22-24 y thick; lower cortex 7-9 y thick ; medulla of loosely interwoven hyphae; thallus 127—154 y thick ; thecium 20-26 и thick; aseus clavate, spores 8, 5-7 x 8-9 y; paraphyses branched once or twice. Distribution: South Carolina to Florida, west to Alabama. SOUTH CAROLINA: AIKEN CO.: Aiken, 1854, Ravenel 404 (Type collection) (T). COUNTY UNKNOWN: Curtis (NYBG). FLORIDA: ORANGE CO.: on wild cherry, Sanford, Rapp (Е); on oak, Sanford, Dee. 1923, Rapp 410 (F). ALABAMA: LEE CO.: Auburn, Jan. 9, 1897, Earle $ Baker 4 (NYBG). TENNESSEE: COUNTY UNKNOWN: on trees, Calkins (FM). 36. Parmelia sublaevigata Nyl, Ann. Sei. Nat. Bot. V, 7: 306. 1867. Parmelia tiliacea var. sublaevigata Nyl., Syn. Lich. 1: 383. 1860 Nylander's type of the variety was from South Africa, and is now in his Herbarium in Helsinki, Finland. Thallus variable in size, 4-20 em. diameter, upper surface smooth, greenish-gray, without soredia or wrinkles; lobes al- ternately dissected, with a black fringe of the rhizinae extend- ing from under the margin ; lower surface with a dense covering of black rhizinae ; apothecium 2-10 mm. diameter, sessile, disk flat or deeply concave, chestnut-brown ; K, brown, C, none, K + C, none, P, none. Algal layer continuous, 40-50 и thick; upper cortex 28— 30 p thick, matrix light brown; medulla of loosely inter- woven, hyaline hyphae of small diameter; lower cortex 32- 36 и thick, matrix black; thallus 120-140 y thick; thecium 52-58 y thick; ascus clavate, spores 8, 5-6 x 6-8 и; paraphyses branched. Distribution: North Carolina to Florida, west to California. [Vor. 28 108 ANNALS OF THE MISSOURI BOTANICAL GARDEN Мовтн CAROLINA: MITCHELL CO: on bark, Roan Mountain, June 15, 1936, Schal- lert (MBG); 1921, Schallert (Е). SOUTH CAROLINA: COUNTY UNKNOWN: Ravenel (MBG,BSNH). FLORIDA: ORANGE CO.: on trees, Sa а Apr. 1923, Rapp (BPI). DUVAL со.: on Taxodium, near Jacksonville, Calkins MBG,F) ; nig hate Apr. 1923, Rapp (MBG); on trees, Jacksonville, pa 2 (MIN) MA: BALDWIN CO.: on trees, Fairhope, Oct. 17, 1924, Evans (F,NYBG); G). LOUISIANA: TANGIPAHOA CO.: on fence of old cypress pickets, Mar. 7, 1894, Langlois (NYBG) ; р m swamp, 3 miles south of Ponchatoula, Mar. 26, 1937, Johnson 215 ipei KENTUCKY: 0. he ag woods, 1.4 miles north of Jarvis Store, Aug. 19, 1937, Sara) pea (apa TENNESSEE: SEVIER CO.: woods; elev. 1600 ft., Great Smoky Mountains, 3 miles southwest of oe Aug. 25, 1937, Hubricht B466 (MBG). LLINOIS: FULTON C on oak, Canton, Wolf (NYBG). MISSOURI: BOONE CO.: on bark of tree, Ashland, May 10, 1936, Wilson (B). ARKANSAS: AND CO.: on bark of pine tree, Hot fori , Jan, 2, 1939, Scully pem Oso, CHICOT CO.: near Macon, Dee. 29, 1928, E. re son (MBG). BOONE CO.: on rock, rocky wooded bluff, above dry creek bed, Bear Creek Spring, 0.5 mile north of Francis, Apr. 7, 1939, Dodge, Berry ф Johnson (MBG). NEWTON co.: on bark, I Niro on gentle slope near Marble Falls, 11 miles south of Harrison, Apr. 8, 1939, Dodge, Berry 4: Johnson (MBG); on bark, high blu long steep w Ne! hillside, Lookout Point, 7 sides south of Jasper, Apr. 8, 1939, Dodge, Berry $ Johnson (MBG); bark of tree, dry wooded hillside, 9 miles south of Boxley, Apr. 8, 1939, pc Berry ф Johnson (MB EXAS: BRAZORIA CO.: on branch of oak, West Gotami, Feb. 1939, Fisher (MBG). GALVESTON CO.: Аа, Hall (US CALIFORNIA: SAN DIEGO CO.: San Diego, Orcutt 4782 (US). 37. Parmelia cubensis Nyl., Flora 68: 611. 1885. The type is Charles Wright, Lichenes Cubae, 76, which was collected on the southeast coast of the Island of Cuba, and is now in Nylander's Herbarium, at Helsinki, Finland. Thallus small, adnate, lobes narrow, imbricated from the cen- ter of the thallus, with small rounded lobules arising from the margins; upper surface smooth, greenish-yellow, well covered with the black ostioles of the spermagonia; lower surface light brown, thickly covered with short brownish or black rhizinae ; apothecium sessile, 0.75-2 mm. diameter, margin of amphithe- cium smooth or occasionally dissected, disk concave, light brown; K, none, C, none, K + C, none, P, yellow. Algal layer discontinuous, groups of algae 5-19 и in diam- eter, upper cortex 22-25 y thick ; medulla of loosely interwoven hyphae of one type; lower cortex 7-9 и thick; thallus 150-217 и E 1941] BERRY—THE GENUS PARMELIA IN NORTH AMERICA 109 thick ; thecium 41-44 р thick; ascus clavate, spores 8, 5—6 x 7- ди; paraphyses branched, thin and pointed at the tip. Distribution: Virginia to Florida, west to Texas. VIRGINIA: JAMES CITY CO.: Williamsburg, Apr. 17, 1911, Farlow (Е). FLORIDA: ORANGE CO.: on trunk of palmetto, сога) Карр (ВРІ). ALABAMA: BALDWIN СО.: on tree, near Fish River, Mar. 22, 1925, Evans (NYBG); on live oak, Yupon Point, Mar. 1, 1925, Evans 201 (F); on tree, Mar. 27, 1925, Evans 371 (F) LOUISIANA: PLAQUEMINES C0.: on live oak, Pointe a la Hache, July 4, 1885, Langlois 101 (NYBG). TEXAS: BREWSTER CO.: common on rocks, north side of Mount Emory, Feb. 18, 1934, Warwick (MBG). 38. Parmelia laevigata (Sm.) Ach., Syn. Meth. Lich. 212. 1814. Lichen laevigatus Sm. apud Sowerb., Eng. Bot. 26: tab. 1852. 808 Parmelia sinuosa var. laevigata Schaer., Enum. Crit. Lich. Eur., 43. 1850. Smith's type was collected by Rev. Hugh Davies in Caern- avonshire, Wales. The present location of the type is unknown. Smith presents in Sowerby's ‘English Botany’ a drawing which represents the description. In transferring the species to the genus Parmelia, Acharius evidently saw Smith's type, but there is no record of the type in Acharius’? Herbarium. Schaerer cites ‘Lichenes Helvetici Exsiecati,' 561, as the type of the variety. Thallus 5-15 cm. diameter, thin, loosely attached to sub- stratum; upper surface free from wrinkles or ridges, some- times slightly sorediate; lobes narrow and somewhat imbri- eated, subdichotomously branched, rhizinae extending from beneath the margins, margins never ciliate; lower surface black to the margin, with numerous small black rhizinae; apo- thecium small, 3-6 mm. diameter, sessile, disk concave, chest- nut-brown, amphithecium entire or serrate; K, medulla brown, upper cortex yellow; C, none, K + C, none, P, orange. Algal layer continuous, 16-20 и thick; upper cortex 20-25 и thick, matrix brown; medulla loosely interwoven, of hyaline hyphae of small diameter; lower cortex 20-80 и thick, matrix [Vor. 28 110 ANNALS OF THE MISSOURI BOTANICAL GARDEN black; thecium 52-55 p thick; ascus clavate, spores 8, 5-7 x 9- 16 и; paraphyses branched. Distribution: Nova Scotia to Vermont, west to Missouri. Nova Scoria: Menzies (US). VERMONT: COUNTY UNKNOWN: on bark of tree, Knowlton (US). MISSOURI: PULA о.: on chert rocks, open woods, 2 miles north of Hanna, Oct. 8, 1938, wand po. (MBG). TEXAS: BRAZORIA СО.: on dead branches of Quercus sp., elev. 40 ft., West Columbia, Feb. 8, 1939, Fisher (MBG). Т. AND FOREIGN EXSICCATAE EXAMINED: Bohl, 110; Claud. 258; Cromb. ; Harm. Lich. Loth. 289; "m Caes. 64; Larb. Herb. 124, 293; Malbr. 268, ory Salw. 235; Spruce, 132, 17 39. Parmelia endoxantha Merrill, Bryol. 12. 73. 1909. The type is Rapp ХИ, т the Farlow Herbarium. The collec- tion is from Sanford, Florida. Thallus large, loosely adnate, surface smooth, dark shiny brown, becoming crustose and granular in the center; lobes radiating from the center, margins smooth and crinkled; lower surface olivaceous-brown, with many small light-colored rhi- zinae ; cortex unbroken; apothecium sessile, 1-3 mm. diameter, disk slightly concave or flat, chestnut-brown, margin of amphi- thecium slightly incurved and dissected; K, medulla and upper cortex greenish-yellow; C, none, К + C, yellow, P, yellow. Algal layer continuous, 8-14 и thick; upper cortex 15-21 и thick, hyphae of small diameter ; medulla of very loosely inter- woven large hyaline hyphae; lower cortex 15-18 и thick; the- cium 60—64 и thick; ascus clavate to oval, spores 8, 6-7 x 8-9 и, almost globose; paraphyses branched and enlarged at the tips. Distribution: found only in Florida. FLORIDA: ORANGE CO.: on bark, Sanford, July 1907, Rapp XII (Type collee- tion) (F); on trunk of Magnolia glauca, Sanford, Oct. 6, 1907, Rapp 5 (F); on — ne May 3, 1908, Rapp 1 (F); on palmetto, Sanford, July 1917, Rapp 16 (Е). LEE CO.: on cypress trunk, vicinity of Fort Myers, Mar. 4, 1916, tia ah 13084 (UB). DUVAL CO.: on cypress bark, Jacksonville, Calkins (FM, MIN). 40. Parmelia aurulenta Tuck., Amer. Jour. Arts & Sci. II, 25: 424. 1858 The type was from Crawfords Gap, in the White Mountains of New Hampshire, and is now in Tuckerman's Herbarium at the Farlow. 1941] BERRY—THE GENUS PARMELIA IN NORTH AMERICA 111 Thallus varying from small to large, lobes short and round, margins dissected; upper surface wrinkled and covered with scattered or confluent, yellowish-green soredia; lower surface, brown to black with numerous small black rhizinae to the mar- gin; apothecium 4-12 mm. diameter, sessile, disk concave, chestnut-brown, amphithecium often sorediate; K, medulla brown, C, none, K + C, brown, P, brown. Algal layer continuous, 8-12 р thick; upper cortex 9-12 в thick ; medulla of loosely interwoven, yellow hyphae; lower cor- tex 14-16 и thick, matrix black; thallus 70-88 и thick; thickness of thecium 20-25 и; ascus clavate, spores 8, 9-10 x 12-17 y; paraphyses unbranched. Distribution: Quebec to Florida, west to Minnesota. QUEBEC: on spruce, Ste. Anne, May 5, 1935, Lepage 133 (D). ONTARIO: on trunks, Ottawa, May 10, 1907, Macoun 200.07 (F). MAINE: KNOX с0.: boulder in shaded places, Megunticook Lake, Camden, Aug. 13, 1913, Merrill (F). NEw HAMPSHIRE: С005 Co.: White Mountains, near Crawford's Gap, 1844, Tuckerman as E or C); R co.: on Ostrya virginiana, Haystack Mountain, Pawlet, Aug. 25, SOT Pis E Ne CONNEOTIOUT: TOLLAND CO.: Union, June 18, 1927, Evans 1180 (F). MIDDLE- SEX CO.: on shaded rocks, East poda. Мау 14, 1927, Evans 1053 (Е). FAIR- FIELD CO.: Oct. 26, 1926, Evans 853 MARYLAND: BALTIMORE CO.: On us near Baltimore, E Рин (WIS,MBG, FM); on rock in M E guis Gwynns Falls path, Sept. 19, 1905, Merrill 101b (F). MONTGOM n hiekory tree, west knoll, Plummers inu near Cabin John, Nov. 6, ed poe 2781 (MBG). VIRGINIA: FAIRFAX CO.: Great Falls Park, July 25, 1931, Plitt (BPI); Harpers p са Vies (US). Nor HENDERSON CO.: on exposed rock, Flat Rock School House, June | Eros pom 6153 (WIS); on bark of tree, Flat Rock, May 5, 1922, Schallert а FORSYTH CO.: bark of tree, June 13, 1934, Schallert (MBG); ain ee 917Т, (Е). CHATHAM CO.: Bennetts Rock, Apr. do n hehe pu UNTY UNKNOWN: on bark of tree, South Moun- tain, аш 1, 1934, Schallert (Mud). Sou SI INA: COUNTY UNKNOWN: on trees, d (FM). bin ORANGE CO.: on Myrica, дог аја Rapp 317 (F). ALABAMA: COUNTY UNKNOWN: 1852, ers "d d COUNTY UNKNOWN: 1880, ж: шюр Inn UNKNOWN: Hall (BSNH). MENARD CO.: bark of trees, MU s pu сем); on old trees, 1878, Hall (ЕМ). MINNESOTA: CLAY CO.: on shaded roeks, near Ulen, July 5, 1899, Fink 287 (MIN). IowA: EMMET CO: on rock, Esterville, Oct. 1926, Fink (BPI). [Vor. 28 112 ANNALS OF THE MISSOURI BOTANICAL GARDEN ARKANSAS: GARLAND CO.: on oak tree, Little Rock Highway, 3 miles from Hot Springs, July 23, 1939, Scully 1375 (MBG). NEWTON 00.: on bark of tree, wooded hillside, 11 miles south of Harrison, Apr. 8, 1939, Dodge, Berry $ Johnson (MBG). 41. Parmelia sulphurosa (Tuck.) Fink, Lich. Fl. U. S., 328. 1935. Parmelia tiliacea var. sulphurosa Tuck., Syn. N. Amer. Lich. 1: 57. 1882. The type of the variety cited by Tuckerman was collected by A. H. Curtiss in Florida. This plant is now in the Tucker- man Herbarium at the Farlow Herbarium. Fink used the same collection as a type in the transfer of the variety to the rank of species. Thallus 3-10 em. diameter, closely adnate to bark of trees, lobes short and rounded, slightly imbricated; upper surface smooth or covered with yellow soredia ; medulla always yellow ; lower surface black with brown margins, numerous black rhizinae ; apothecium sessile, 1-9 mm. diameter, concave, light brown, amphithecium smooth or sorediate; K, none, C, none, К + C, none, P, none. Algal layer continuous, 16-20 и thick; upper cortex 8-12 и thick, matrix yellow; medulla of loosely interwoven, hyaline hyphae of small diameter; lower cortex 8-16 thick, with a black matrix; thickness of thallus 100-120 и; thecium 40-48 u thick ; ascus clavate, spores 8, 3—4 x 6-8 и; paraphyses branched and enlarged slightly at the tips. Distribution: Florida to Louisiana. FLORIDA: COUNTY UNKNOWN: 1878, Curtiss (Type collection) (T). LOUISIANA: PLAQUEMINES CO.: Grand Bayou, July 17, 1885, Langlois (US). IV. AMPHIGYMNIA Parmelia subgenus Euparmelia sect. IV. Amphigymnia Zahlbr. apud Engler & Prantl, Nat. Pflanzenfam. I Teil, Abt. 1: 213. 1907. Parmelia sect. Amphigymnia Wainio, Étude Lich. Brésil 1: 28. 1860. Parmotrema Mass., Atti I. R. Ist. Veneto, IIT. 5: 248. 1860. Zahlbruckner's type is Parmelia caperata (L.) Ach. Wainio selected Parmelia perlata (Huds.) Ach. as the type of his sec- 1941] BERRY—THE GENUS PARMELIA IN NORTH AMERICA 113 tion of Ше subgenus Huparmelia. Massalongo cites Parmelia perforata (Wulf.) Ach. as the type of Parmotrema. Thallus either smooth or ciliate at the margins ; lower surface with a heavy mat of rhizinae at the center; marginal rhizinae either absent or papillose; apothecia sessile or very short- stalked. The distribution of the section is world-wide in the warmer portions of the earth. KEY TO SPECIES IN EUPARMELIA SECTION AMPHIGYMNIA A. Margins always ciliate. pper surface never with lobules or isidia. с Closely adnate. I Lower surisee ПЕН ае 50a. perforata var. hypotropa ЕР: BUTLACO RATED oo Ev AD olan Ev ODER qe m 50. perforata Loosely. а о ьо аон a e 46. perlata B. Always with lobules or isidia. C. Lobules always present on upper surface and amphithecium. .51. proboscidea C. Isidia or eoralloid branchlets always present. D. Thallus loosely adnate, orbicular in outline................. 48. crinita D. Thallus closely adnate, never orbieular in outline................ АОН ВИО е Т 46a. perlata var. ciliata A. ко never eiliate. pper pent. it ШЫ, СЕЕ р. Баа. white. E. Soredia globose, confluent; K, reddish, C, none ......... 49. cristifera E. Soredia not globose, scattered, never confluent. . Thallus light gra G. K, yellow-green, с, УПИ 47. cetrarioides тп попе 0:709. ат 47a. cetrarioides var. rubescens llus yellowish; К, попе, C, тей.................... 44. soredica D. Soredia yellow BI. EX ooo VOTE E DUO 45. Herreana C. Without marginal soredia. D: Е ради yellowish-gree below; Ко поде С, оде 5530. ce tale e 43. praesignis = Saa i Ваше E red; C Mitoni тей. ise ees 53. latissima D. Thallus dark green; K,n Попе 42b. caperata var. subglauca on B. Upper surface with BENE огей: or isidia. C. With seattered soredia. RK yellows: С mon crys eoe ре 42. caperata DZEK попе С, поле 42a. caperata var. incorrupta C. With seattered ы, Rc UO E PN SI ОРНА ИОА ЗА 52. tinctoria 42. Parmelia caperata (L.) Ach., Meth. Lich. 216. 1803. Lichen caperatus Linn., Sp. Pl. 1147. 1753. Platisma caperatum Hoffm., Deser. & Adumbrat. Pl. Lich. 2: 50. 1794 [Уог. 28 114 ANNALS OF THE MISSOURI BOTANICAL GARDEN The type of Linnaeus was from central Europe, and the plant is now in the Linnean Herbarium in London. Hoffmann's type is unknown but he cites Linnaeus' Lichen caperatus as synony- mous with his Platisma caperatum. Acharius’ type is in Hel- sinki, Finland. Thallus 5-25 em. diameter, adnate on trees, stones, buildings, and fences; upper surface wrinkled and often sorediate, lobes short, rounded, imbrieated margins erenate and often slightly ascending, seldom with soredia; lower surface black at the center with light brown margins, rhizinae black and heavy at the center, very small and light-colored at the margins ; apothe- cium rare, sessile, 2-15 mm. diameter, disk chestnut-brown, concave, amphithecium entire, often with coralloid branchlets, isidia, or soredia; K, yellow, C, none, K + C, none, P, red. Algal layer continuous or discontinuous, 15-30 и thick ; upper cortex 15-18 и thick, matrix yellow-brown, cells small; medulla of loosely interwoven hyphae, of one type, hyaline and of small diameter; lower cortex 12-20 и thick, matrix black or dark brown; thallus 114—120 y thick; thecium 80-100 и thick; ascus clavate, spores 8, 12-16 x 16-20 и; paraphyses branched. Distribution: Ontario to Florida, west to Washington and California. ONTARIO: on old fence rails, Ottawa, Apr. 26, 1891, Macoun 79 (MIN,WIS, MBG); on trees, stumps, and fallen trees, Rainy Lake, Aug. 20 ap Schmidt 9 (MBG); on trunks and old rails, Ottawa, Apr. 26, 1900, Macoun 50 (MBG); on trees in cedar swamps, ae July 17, 1901, Fink 646 (MIN); on ira Emo, July 20, 1901, Fink 769 (MIN). uw WASHINGTON CO.: June 1929, Markin (WIS). KENNEBEC CO.: Albion, Mar , 1922, Parlin 3 (D). CUMBERLAND CO.: on rocks, Cumberland, Nov. 7, 1855, pq (FM); Peaks Island, Aug. 31, 1893, Curtis (NYBG). COUNTY UN- KNOWN: 1859, DeLeashi (FM); 1863, Nd. FM EW HAMPSHIRE: COOS CO.: on roeks, Cherry Мода, Aug. 8, Sb Harper $ т. me COUNTY UNKNOWN: Wiley Brook, Faxon (MBG) WINDHAM CO.: on apple tree, Pike Hollow, балоо, July 3, е. "vere 1358 (MBG); Grout Pond, Stratton, July 4, 1935, Moore (MBG). ADDISON CO.: Monkton, May 6, 1879, Жоро» гаа, on Pinus Strobus, Chipman an Middlebury, Apr. 13, 1913, Dodge (D). RUTLAND CO.: cedar tree in swamp, .6 E Nov. 11, 1922, Dutton (MBG); on rock, Tadmer Hill, Pawlet, Apr. 14, 1923, Dodge (D); Ln Lake, Rp cue June 1938, Dod са ). MASSACHUSETTS: ESS co.: on s, Annisquam, AW 1892, — 955 (WIS,MBG). MIDDLESEX CO.: peace 1863, Mann e NORFOLK 1941] BERRY—THE GENUS PARMELIA IN NORTH AMERICA 115 Wellesley, Cummings 108 (NYBG); on juniper, Wellesley, May 4, 1892, Cum- mings 55 (MBG); on Juniperus communis, Wellesley, Nov. 1895 and 1896, Сит- mings 170 (FM). BRISTOL co.: New Bedford, Willey (BSNH); on rocks and trunks, New Bedford, Willey 35 (FM); New Bedford, Willey 40 (MIN). WORCES- TER CO.: on chestnut tree, near Haynes Reservoir, Leominster, Feb. 6, 1922, Fassett CPI PLYMOUTH со: Hingham, Russell (FM). COUNTY UNKNOWN: Russell (СА ae YORK: YATES : Dundee, дал 22 (FM). ONONDAGA СО.: on fence, Syraeuse, Nov. 20, о. ‘Underhill (NYBG). PENNSYLVANIA: LANCASTER CO.: on chestnut rails, Aug. 27, 1879, Rothrock (ЕМ); on trees, Lancaster, Маг. 4, 1894, Eby (MBG); on rocks, ИИ, i 15, 1892, Eby ); on bark, Beartown, June 1895, (MBG); rails, Rheinholds, Mar. 11, 1893, Eby (МВС); on rocks, Smithville, May js ini Eby (MBG); on trees, Crow Hill, Mar. 17, 1894, Eby (MBG). ADAMS CO.: on tree, Rock Creek, Gettysburg, May = nox ще E. C. White 215 (US). FRANKLIN СО.: 1320 ft., May 2 E s. E. C. White 205 (US). couNTY UN- KNOWN: Brand Creek, 1855, HR i (Е ib MARYLAND: MONTGOMERY СО.: Ее outerop, west knoll, Plummers Island, in the Potomae River, near Cabin dis Nov. 6, 1938, Leonard 2789, 2763 (US, MBG); base E hiekory tree, west k Plummers Island, in the Potomac River, near Cabin John, Nov. 6, 1938, me 2788 (US,MBG); rock outerop, west end Plummers Island, in TM Potomae Riv e near Cabin John, Nov. 7, 1937, Leonard 2666 (US,MBG) ; vertical surface cliff, east side of Cabin Hill, Plummers Island, n the Potomae River, Nov. 6, 1938, Ae 2823 (US,MBG); rock outerop, west oed Plummers DATE in the Potomae River, Nov. 6, 1938, Leonard 2796 (US). COUNTY UNKNOWN: on rocks, Sept. 21, 1902, Maxon (M). VIRGINIA: HIGHLAND CO.: on white oak, uic Sept. 1, 1935, Darrow 983 (FM). вмутн CO.: on Walker M us elev. 3300 ft., June 13, 1892, Leeming (W). WASHINGTON CO.: on summit of White Top Mountain, May 28, 1892, Small (MBG,MIN). couNTY INDEFINITE: Blue Ridge, southwest Virginia, elev. 5670 ft., May 38, 1892, Small (MIN). NoRTH CAROLINA: STOKES CO.: on exposed rocks, Moore's Spring, May 3, 1930 Schallert (WIS). COUNTY UNKNOWN: on bark of trees, Nissen Park, July 4, 1922, LAKE CO.: Eustis, June 16—30, 1895, Nash (MBG). ORANGE CO.: San- ford, May 1911, Rapp (BPI) | LOUISIANA: COUNTY UNKNOWN: On pickets, Ye" 3, 1894, Langlois (M OHIO: BELMONT CO.: on tree trunks, Apr. 30, 1938, Wareham 933 one GEAUGA co.: on trees, Oct. 9, 1937, Wolfe 911 mo магов co.: on rock, Apr. 5, 1936, Wolfe 516 (О). ATHENS CO.: on roek in woods, Sept. 22, 1935, Wolfe 258 (О). RICHLAND co.: on bark of tree, Apr. 20, 193 | Waren 84 (О). HOCKING CO.: on tulip tree, Ash Cave, Oct. 13, 1935, Wolfe 278 (O). JACKSON CO.: on north facing of cliff, Nov. 3, 1935, Wolfe 322 (O). FAIRFIELD CO.: on face of cliff, Apr. 27, 1935, Wolfe co о CO.: оп elm, Мау 1935, Wolfe 183 (0). PIKE Mags on bark, Sept. 15, 1935, Wolfe 207 (O). WYANDOTTE CO.: on white oak, y 29, т hd 229 (O). UNION CO.: on trees near road, May 3, 1936, Wolfe v Paese p 6 (O). ADAMS CO.: on ar, Lynn, May 28, 1937, А $50 (O). LOGAN CO.: on elm, Мау 3, 1931, white Я Hartley [Vor. 28 116 ANNALS OF THE MISSOURI BOTANICAL GARDEN 660 (О). CLINTON CO.: on oak, Mar. 18, 1936, Wolfe $ Hill 588 (О). BROWN CO.: on elm, Apr. 18, 1936, Wolfe $ Hill 583 (О). CHAMPAIGN 00.: willow tree along Storms Creek, home farm 14 miles east of Thackery, July 9, 1934, Leonard 1722 (US,MBG). CLERMONT CO.: on tree, Apr. 18, 1936, Wolfe $ Hill 612 (О). WARREN со.: on tree, Apr. 18, 1936, Wolfe ү" Hill 602 (О). DARKE CO.: on trees, New Madi- son, Apr. 2, 1938, Wallace 922 (O). PREBLE CO.: on granite, open woods, Eaton, Mar. 9, 1914, Fink (M). BUTLER CO.: on stone, Oxford, July 6, 1924, Walters (M). MICHIGAN: KEWEENAW СО.: Isle Royale, Sept. 13, 1901, ey $ Allen (WIS). ALGER CO.: on cedar in swamp, Grand Marais, Aug. 2, 1902, Fink (M). MARQUETTE CO.: on trees, Mud Lake Bog, Whitman Lake, June 17, 1930, Hedrick M). INDIANA: TIPPECANOE CO.: on rocks in open wood near Lafayette, Aug. 31, 1917, Fink $ Tuson (D). FOUNTAIN CO.: on sandstone rock in the Bear Creek Canyon just south of Fountain, Apr. 22, 1917, Deam 22.252 (BPI). KENTUCKY: ROCKCASTLE CO.: bark of trees, open pine woods near Rockcastle River east of Livingston, Aug. 18, 1937, Hubrioht B253 (MBG). BALLARD CO.: Kevil, Apr. 26, 1936, Fassett (WIS). ESSEE: HAMILTON CO.: on rocks, Lookout Mountain, Calkins ? (MIN). WISCONSIN: FOND DU LAC CO.: on granitie roeks in pasture at north end of Wolf Lake, July 17, 1938, Schinners (WIS); bark of trees, edge of tamarack bog, northeast end of Wolf Lake, July р. 1938, Schinners — DODGE CO.: on rocks, northeast of Kekoskee, May pio . Thomson (WIS). USHARA CO.: 2 miles east of Plainfield, Oet. 13, 1935 ig n (WIS); sand gine за abandoned 7 years, south of Plainfield, Oct. 13, is 35, J. Thomson (WIS). MARQUETTE CO.: in Larix swamp, 12 miles east of Westfield, Oct. 13, 1935, J. Thomson ghd DANE CO.: common on trees and fences, Madison, May 10, 1893, Head $ Buell (WIS). VILAS co.: valley of Wisconsin River near Lae Vieux Desert, Summer, 1893, eney 9 SAUK CO.: оп bark, Devils Lake, Oet. 27, 1935, J. Thomson (WIS). U со.: Apr. 1936, J. Thomson (WIS). MONROE CO.: 9 miles west DA Tomah, Nor, p 1935, г. Tisha (WIS). GRANT CO.: bluffs of Wisconsin S: Potosi, Oet. 6, 1935; J. Thomson (WIS). LA CROSSE CO.: on trees, La Cross . 189 Panne (MBG). WASHBURN СО.: on bark of dead, white Dire tree, nés on moss of forest floor near Spooner, = 1932, Flint (MONT). ILLINOIS: COOK CO.: on oaks, Calkins (FM). JACKSON CO.: on sandstone boulders near qe INA 5 dm southwest of Pomona, Apr. 14, 1940, Hubricht B1734 (MBG). ARD CO.: on old trees, 1878, Hall (FM); Hall 8 (FM); on rock, Athens, May y" 1879, Hall (FM). ADAMS CO.: on boulders, Quincy, May 30, кнн Harper $ Harper 34 (FM). NESOTA: COOK CO.: on rocks, Grand Portage Island, June 23, 1897, Fink 916 ON): on eedar in swamp, Grand Marais, Aug. 2, 1902, Fink 5284 (MIN). LAKE C0.: Two Harbors, Aug. 25, 1893, Harper (FM); on trees, Beaver Bay, J ee 14, 1897, Fink 728 (MIN); on trees, Snowbank Lake, July 19, 1897, Fink t on trees, Snowbank Lake area, July 24, 1897, Fink 916 (MIN). вт on trunk of Fraxinus americana, Vermil Lake, July 19, 1886, а Вайеу Јт. $ кадыы 425 (MIN); on p Tow wer, Aug. 24, 1901, Fink 1718 (MIN). CARLTON CO.: on trees, Tafte, Carlton Peak, July 10, 1897, Fink (MIN). KOOCHICHING CO.: on о Gunflint, June 30, 1897, Fink 254A ge on rocks, Koochiching, July 31, 1901, Fink 1059 (MIN). AITKIN CO.: on rocks, Palisades, July 15, 1897, Fink 768 (MIN); on Pinus Strobus, Mille Lacs Indian Reservation, June 1892, Sheldon 52636 (MIN). HENNEPIN CO.: on bark, Minne- 1941] BERRY—THE GENUS PARMELIA IN NORTH AMERICA Ilf жа e 1891, Sheldon 54108 (MIN). BLUE EARTH CO.: on trees and rocks, Mankato, Ju E 23, n Fink 53 (MIN); on trees, Mankato, June 27, 1899, Fink 153 (MIN). WOODS CO.: on bark, Lake of the Woods, July 1896, Millan 4 QNT BELTRAMI CO.: on trees, Red Lake, July 26, 1900, Fink 898 IN); on tam araek in swamp, Bemidji, July 7, 1900, Fink 527 (MIN); on 1 Aug. 1, 1900, Fink 1033 (MIN); on old trees, Beaudette, June 19, 1901, Fink 20 (MIN). MEEKER CO.: on bark of trees, Litchfield, June 1892, Frost F115d (MIN). OTTERTAIL CO.: on trees, Battle Lake, June 19, 1900, Fink 26 (MIN); on tamarack in swamp, Henning, June 25, 1900, Fink 195 (MIN); on tamarack in swamp, Henning, July 2, 1900, Fink 409 (MIN). ROSEAU CO.: on trees, Warroad, June 26, 1901, Fink 182 (MIN). BECKER CO.: on trees, Oak Island, July p! RU Fink 524 1900, Fink MIN). PENNINGTON CO.: on trees, Thief River Falls, July 20, , Fink 832 (MIN); on trees, aged ds sii cT 20, 1900, Fink 841 (MIN). POLK со.: Red River, May 1857, k (MIN). CLAY co.: rocks near Ulen, July 5, E Е 285 (MIN). PIPESTONE CO.: on rocks, Pipestone, "eS Manzel (MIN). OUNTY INDEFINITE: on trees, Misquah Hills, July 3, 1897, Fink 403 (MI s on E Pen Lake Island, July 1894, Millan 4 puta 1040 peat IOWA: MUSCATINE CO.: on o Wild Oat Glen, Nov. 15, 1897, Savage 4. Shink Ша ae HENRY СО.: on bark, Jan. 1, 1898, Savage (IA). JOHNSON CO.: on apple , Mar. 20, 1896, Fitzpatrick 4 На nb. FAYETTE CO.: on trees and RR "n Us k (MIN,MBG). POWESHIEK CO.: on sandstone, Grinnell, Mar. 13, 1904, Fink 246 (FM). MISSOURI: DUNKLIN СО.: on tree, Sept. 28, 1897, Bertig (MBG). STE. GENE- VIEVE CO.: on soil over rocks, River Aux Vases, Apr. 10, 1932, Moore (MBG). on trees, Valley Park, May Me (MBG); on bark of tree, Ranken Es- tate, Valley Park, Apr. 20, T Вет тай, on trees, Allenton, Apr. 15, 1905, Schulte 4: Jones (MBG). вт. FRANCO rocks, open woods, Knie, June 10, 1939, Hubricht B1508 (MBG); no pe а “Pickle 8 prings, May 21, 1938, Berry MBG) ; on sandstone ledges in woods, Pickle Springs, May 21, 1938, в (MBG). WASHINGTON CO.: rock-covered hillside, near Antonio, Oct. 29, 1939, Berry (MBG). IRON CO.: full sun to semi-shade, Royal Gorge, 3 pre east of Arcadia, Nov. 18, 1939, Hubricht B1722 (MBG); near Arcadia, May 1925, Greenman (MBG). LIN- COLN CO.: July 1897, Freuring (MBG). FRANKLIN CO.: sandstone ledge in sunny glade, Missouri Botanieal Garden е Gray Summit, May 2, 1938, Berry (MBG). MONTGOMERY CO.: bark of tree, Montgomery City, Nov. 23, 1930, Drouet (MBG). BOONE CO.: on decayed Савана "oodd hillside, 9 miles east of бе сае, Oct. 5, m Berry 240 (MBG). PULASKI CO.: on chert, open woods 2 miles north of Hanna, Oct. 8, 1939, Hubricht B1709 (MBG). GREENE с0.: on bark of tree, rocky wooded bank of James River, 8 miles south of Springfield, Apr. 7, 1939, Dodge, e $ Johnson ро G). RKANSAS: BOONE CO.: on twigs and bark, rocky wooded bluff above dry creek bed, Bear Creek Spring, 0. i mile north of Francis, Арг. 7, 1939, Dodge, Berry $ Johns BG). NEWTON CO.: on bark of tree, ida bluff and long steep wooded n (M cee Lookout Point, 7 miles south of Jasper, Apr. 8, 1939, Dodge, иН 4 Johnson (MBG); bark of Mes steep dry wooded bluff with rocky flat at top, 1 mile south of Ponea, Apr. 8, 1939, Dodge, Berry 4 Johnson (MBG); wooded hill- side, 11 miles south of н н Apr. 8, 1939, Dodge, Berry $ Johnson (MBG). [Vor. 28 118 ANNALS OF THE MISSOURI BOTANICAL GARDEN 0.: on rocks on ridge, water-works, Hot — July 19, iet Scully per queni WHITE CO.: on bark of tree, near Лазо ec. 28, 1938 Ander- son (MBG). LOGAN CO.: open woods, 0.6 mile east of e. сас panic n Maga- zine Mountain, May 4, 1940, Hubricht B1905 (MBG). SoutH DAKOTA: PENNINGTON CO.: on bark of trees, Rapid City, Арг. 27, 1927, Lee (F). OKLAHOMA: MURRAY СО.: on dead cedar, Prices Falls, Apr. 30, 1926, Stratton (M). TEXAS: DENTON CO.: on oaks, 3 miles southwest of Denton, Feb. 1935, Bussard (MBG). BREWSTER CO.: practically covering the trunks of pines on the north side of Mount Emory, Chisos Mountains, July 2, 1937, Warnock (MBG). COLORADO: EL PASO CO.: on log of Pseudotsuga taxifolia, elev. 8300 ft., Pikes Peak, July 1919, 1. M. = ton (F). LA PLATTE CO.: on dead wood, 14 miles south of Durango, ак 00 ft., Aug. 920, Reeside Jr. (BPI). WELD CO.: over silieeous roeks, South St. pire белу, elev. 7200 ft., Oct. 7, 1936, Kiener 4526 (MBG) New MEXICO: COLFAX CO.: on moist igneous е io of Ute Park, elev. 2200—2900 m., Aug. 23, 1916, Standley (F). SAN MIGUEL CO.: on trees, Las Vegas, 1926, Bro. Anect (M). TORRANCE CO.: Manzano ME July 28, 1932, Jones (M). ARIZONA: COCHISE CO.: Carr Peak, Huachuca Mountains, July 16, 1909, Good- ding ( WASHINGTON: WALLA WALLA CO.: on bluffs of columnar lava, along the Co- lumbia River, gorge below Wallula, elev. 200 ft., Foster (F). OREGON: COUNTY INDEFINITE: in forest, Wor. 10, 1894, Il (NYBG). CALIFORNIA: SAN FRANCISCO CO.: on rocks and twigs, San Francisco, near Golden Gate, cm 50 ft., July 1904, Herre (LSU,MIN,MBG). SANTA CLARA CO.: On trees, Blaek иа. Santa Cruz Mountains, June 16, 1903, Herre (M,LSU,MBG). SANTA CRUZ CO.: Santa Cruz, C. S. Anderson NH). MONTEREY CO.: on shaded granite, eypress headland, Point Lobos Reserve, July 11, 1936, Wheeler (MBG). LOS ANGELES CO.: DM: oak, Bull Rush Canyon, Santa telón Island, Knopf (F). EGO CO.: shade i-shade, forming thiek mats on trunks of oak, pine e^ cypress, Ида ТЕ 1988, Cota (MBG). 42a. Parmelia caperata var. incorrupta (Moore) Berry, n comb. Parmelia incorrupta Moore, Ann. Mo. Bot. Gard. 19: 503. 1932. Moore's type was collected in the Guadaloupe Mountains of Texas, and is now in the Missouri Botanical Garden Her- barium, No. 1020340. This plant differs from the species only in that it gives no color reaction with KOH, and this, in the opinion of the writer, does not constitute a specific difference. Distribution: found only in Texas. 1941] BERRY—THE GENUS PARMELIA IN NORTH AMERICA 119 TEXAS: CULBERSON CO.: on bark of Pinus flexilis, ridge above McKittrick Can- yon, elev. 2300 m. mE. Mountains, July 17, 1931, Moore $ Steyermark 3490 (Type р. (MBG); on Quercus, slopes of Emory Peak, elev. 2000 m., June 22, 1931, Moore ф BO 3220 (MBG). EL PASO co.: Fort Bliss, Mrs. Jos. Clemens (MB 9). 42b. Parmelia caperata var. subglauca (Gasilien) Nyl., Lich. Envir. Paris, 35. 1896. Parmelia sub glauca Gasilien, Jour. de Bot. (Morot) 8: 126. 1894. The type of Gasilien was from Saint Omer, France. The type is unknown, but a co-type collection is in Nylander’s Herbarium at Helsinki, Finland. Nylander did not change the type in his transfer of the species to a variety. This plant differs from the species in its small size, glossy upper surface, the abundance of the apothecia, and in the ab- sence of any reaction with KOH. Distribution: found only in Texas. TEXAS: NUECES CO.: on trees, vicinity of Corpus Christi, Mar. 1894, Heller 197, 260 (MIN,MBG,WIS,FM). GILLESPIE CO.: Big Branch, Jermy (MBG). BEXAR CO.: San Antonio, Langlois (US) ; mesquite woods, 18 miles east of San Antonio, May 17, 1940, Hubricht 1927 (MBG); mesquite woods, 1.5 miles southeast of junction of Jolly Lane and Atkins-Elmendorf Lane, 18 miles east of San Antonio, May 17, 1940, Hubricht B1891 (MBG). HARRIS CO.: Houston, Hall (FM). 43. Parmelia praesignis Nyl., Bull. бос. Linn. Norm. II, 6: 270. 1872. The type was collected in the Valley of Mexico, and 1s Bour- geau, No. 1361, in the Paris Museum. Thallus D. 10-20 ст. diameter, loosely adnate to bark of trees ; upper surface slightly winter. cortex often broken by irregular eracks or round holes; lobes wide, short, round, and slightly dissected, margins flat, smooth and slightly brownish in color; lower surface brown at the margins, black at the cen- ter, with a few black rhizinae at the points of contact with the substratum ; apothecium 1—5 mm. diameter, sessile, disk deeply concave, margins of amphithecium thick, cortex often missing in irregular areas; K, upper cortex yellowish-brown, which soon disappears, medulla none, C, none, K + C, none, P, me- dulla gray. [Vor. 28 120 ANNALS OF THE MISSOURI BOTANICAL GARDEN Algal layer continuous, 12-15 и thick; upper cortex 16-20 џ thick, matrix light brown ; medulla of loosely interwoven, large, hyaline hyphae; lower cortex 35—40 y thick, matrix black; thal- lus 243-260 и thick; thecium 100-125 и thick; ascus clavate, spores 8, 6-8 x 14-16 и; paraphyses branched. Distribution: Arizona and New Mexico. New MEXICO: CATRON CO.: shade, 11 miles north of Pie Town, June 9, 1938, Hubricht B1062 (MBG). ARIZONA: PIMA CO.: on trees, near Tucson, Fink 12974 (M). This plant is widely distributed in Mexico and Central America. 44, Parmelia soredica Nyl., Flora 68: 605. 1885. The type was collected in Saskatchewan, Canada, and is in Nylander’s Herbarium, Helsinki, Finland. The collector is un- known. Thallus large, adnate to substratum, yellowish-green to yel- lowish-brown, margins and upper surface with many white soredia which often become large and are easily detached ; lobes wide, round, and sparingly branched, margins slightly ascend- ing, wrinkled and somewhat inflated; lower surface brown at the margin, black at the center with very few scattered rhi- zinae; apothecium rare, 3-12 mm. diameter, sessile, disk con- cave, chestnut-brown, amphithecium sorediate; K, none, C, medulla red, K - C, none, P, none. Algal layer continuous, 20-30 и thick; upper cortex 20-24 и thiek, matrix yellow, cells of cortex very small; medulla of loosely interwoven, hyaline hyphae of one type; lower cortex 20-30 и thick, matrix black; thallus 320—342 и thick; thecium 38—40 и thick, ascus clavate, spores 8, 6-8 x 12-16 и; paraphyses branched and slightly enlarged at the tip. Distribution: British Columbia to New Mexico and Cali- fornia. MEXICO: SOCORRO CO.: shade, 4 miles west of Magdalene, June 9, 1938, Hubricht B1046 (MBG); shade, San Andres Mountains, Rhodes Pass, 47 miles west of ит : une 6, 1938, Hubricht B992 (MBG). BRITISH COLUMBIA: on rocks, 2000 ft. elev., Macoun (FM). CALIFORNIA: CONTRA COSTA CO.: on Pinus радио, Mount Diablo, May 18, 1932, Herre (LSU); on Quercus sp., Mount Diablo, May 19, 1932, Herre (LSU). 1941] BERRY—THE GENUS PARMELIA IN NORTH AMERICA 121 ALAMEDA CO.: Oakland Hills, elev. 1000 ft., Mar. 14, as. Herre (LSU). SANTA CLARA CO.: on Quercus lobata, Stanford (шү, elev. 75 ft., Feb. 4, 1903, olin (MBG,MIN) ; on roof of house, Mayfield, elev. 35 ft., Ad Ores Peninsula, Apr. L 1904, Herre (MBG,MIN); on Quercus lobata, Los Gatos, Mar. > 1905, Herre (LSU); on twigs, Searsville Ridge near Stanford a ены Mar. 3, 1936, Herre (LSU); on bark of Catalpa, ed of Stanford University, Nov. 14, 1939, Herre (LSU). SAN BENITO CO.: re near Aromas, July 23, pont Herre (LSU); ; on trees, Searsville, elev. 318 i n 4, 1903, Herre (MBG,MIN); bark of Ficus, Goff ranch, Kenwood, Jan. 27, 1936, Goff "26 (LSU). LOS ANGELES CO.: Santa Monica Range, Feb. 1895, Hasse (FM). SONOMA CO.: 1893, X (LSU). MARIN CO.: bark of tree, Mill Valley, Nov. 20, To. Fg erre (LSU). SAN MATEO CO.: Kings Mountain, Santa Cruz Mountains, elev. 1900 ft., Spring 1930, am (LSU); on live oak, elev. 2000 ft., Santa Cruz o Ms pus 12, 1938, Herre (LSU); on Quercus agrifolia, hills near Woodside, elev. 800 ft. ig Щи 18, 1939, Herre (LSU) ; on oaks, Santa Cruz Mountains, 12 miles west of Manford бока, elev. 1500 ft., Apr. 26, 1933, Herre (LSU). 45. Parmelia Herreana Zahl., Cat. Lich. Univ. 6: 239. 1930. Parmelia perlata var. flavicans Tuck., Lich. Calif., 13. 1866. Parmelia flavicans Tuck., Syn. N. Amer. Lich. 1: 55. 1882 (non Ach.). Tuckerman's type was from Oakland, California, and is now in his herbarium in the Farlow. The specific name flavicans is non-valid because of its prior use by Acharius. Zahlbruckner applied a valid name to the type. Thallus large, 8-15 em. diameter, loosely adnate to bark of trees and stones; upper surface smooth, greenish-yellow or pale yellowish, with yellow soredia, lobes wide or narrow with erenate and sorediate margins; lower surface black with a chestnut-brown margin, scattered black rhizinae at the points where the thallus touches the substratum ; apothecium 3-6 mm. diameter, disk chestnut-brown, slightly concave, amphithecium often sorediate ; K, red, C, none, К + C, red, P, none. Algal layer continuous, 22—24 и thick, or of thinly scattered gonidia 20-22 y in diameter; upper cortex 40—60 и thick; me- dulla of loosely interwoven hyphae of one type; lower cortex 20—30 и thick, matrix brown or black; thallus 120—130 и thick; thecium 88-108 y thick; ascus clavate, spores 8, 6-8 x 16-20 и; paraphyses branched and enlarged at the tips. Distribution: Idaho to Arizona, west to California. AHO: NEZ PERCES CO.: Lake Waha, elev. 2000-3500 ft., June 29, 1896, Heller $ Pies (MBG [Vor. 28 122 ANNALS OF THE MISSOURI BOTANICAL GARDEN ARIZONA: YAVAPAI CQ.: n July 1898, Kanza (NYBG). CALIFORNIA: ALAM n bark of trees and among mosses, deep canyon, Oakland, Bolander gr n (T). SANTA CLARA CO.: sandstone, elev. 200—300 ft., Pilareitos Creek Canyon, Aug. 14, 1903, Herre 212 I ; on mossy sandstone, foothills near Stanford University, elev. 300 ft., Aug. 9, 1936, Herre ; sandstone covered with moss, hills near Stanford University, elev. 400 ft., Mar. 22, 1008, Herre (MBG). SAN BENITO CO.: on sandstone, big rock on Pinedale eutoff, Mar. 31, 1935, Herre (LSU). LOS ANGELES CO.: on mossy bank, elev. 2200 ft, mountains north of Claremont, Oct. 22, 1816, I. M. Johnston 3039 (LSU). SONOMA CO.: on sandstone, stony point near не В, July 9, 1903, Herre (NYBG, MIN,MBG). RIVERSIDE CO.: on wet moss, barranea switeh back, of highway, San Jacinto Mountains, elev. 4000 ft., Feb. 23, 1935, Reed (MBG). SAN DIEGO CO.: San Diego, 1933, Cota (MBG). 46. Parmelia perlata (Huds.) Ach., Meth. Lich. 216. 1803. Lichen perlatus Huds., Fl. Anglica, 448. 1762. Lobaria perlata Hoffm., Deutschl. ЕІ. 2: 148. 1795. Hudson's type was collected in England, its present location being unknown. Acharius eites Hudson's plant as synonymous with his, and also cites as a synonyn, Wulfen apud Jacquin, ‘Collectanea’ 4: f£. 10. 1790 Thallus 5-20 em. diameter, loosely adnate; upper surface smooth except at the margins where it is commonly sorediate, margins wavy and somewhat dissected dichotomously at the tips of the lobes, slightly ciliate ; lower surface black with scat- tered black rhizinae which often give the appearance of black cilia, margins brown; apothecium rare, sessile or very short- pedicellate, 4-12 mm. diameter, disk chestnut-brown, concave, amphithecium entire and never sorediate; K, yellow then red, C, none, K + C, none, P, none. Algal layer continuous, 15-20 и thick; upper cortex 12-16 y thick, yellow-brown; medulla loosely interwoven, hyphae of one type, hyaline and of small diameter; lower cortex 20-24 y thick, matrix dark brown or black; thickness of thallus 132— 140 и; thecium 88-94 и thick; ascus clavate, spores 8, 8-10 x 10- 12 и; paraphyses branched and enlarged at the tip. Distribution: Quebec to Florida, west to British Columbia and California. QUEBEC: on twigs, woods along River Ste. Anne des Monts, July 22, 1923, Col- lins $ Dodge (Е MAINE: WALDO CO.: on boulder, Freedom, Sept. 24, 1922, Parlin (F). p Erp | po TT RE 1941] BERRY—THE GENUS PARMELIA IN NORTH AMERICA 128 New HAMPSHIRE: Coos СО: White Mountains, 1885, Farlow 415 (Е). в co.: on rocks, North ы ка аи Franconia Mountains, July 1891, Cummings СР). VERMONT: WINDSOR CO.: exposed ledges, Rochester, Dutton (F). RUTLAND CO.: on Acer, Pawlet, jen 1913, Dodge (D). MASSACHUSETTS: WORCESTER CO.: on rocks, Lunenburg, ue Sargent 416 (F). PENNSYLVANIA: LANCASTER CO.: Welsh Mountain, Feb. 27, , Small (MIN). VIRGINIA: SMYTH CO.: along Necks Creek, June 4, 1892, me E P NoRTH CAROLINA: HENDERSON СО.: on bark of trek, Flat Rock, Schaller (WIS). FORSYTH CO.: on bark of tree, Bennetts Rock, A , 1936, ом (MBG). нлуүоор CO.: оп bark of гор Balsam Mountains, E. 8, 1922, Schallert 6753 (FM); Eagle Nest, elev. 4700 ft., Aug. 29, 1902, Curtis (F); gl of Eagle Nest near Waynesville, elev. 900—1500 m., Bion dley (D). age с0.: 0 woods, 3 miles south of Ace RUE Aug, S 1937, пром B583 (MBG). SOUTH CAROLINA KEN : Aiken, Ravenel (F). COUNTY UNKNOWN: on twigs, sand dunes, Зама T MAS pua 3, 1936, did (MBG). FLORIDA: ORANGE CO.: on palmetto logs, Sanford, Mar. 1906, Rapp VII (F); n Myrica, in dig Sanford, Rapp (BPI). DUVAL CO.: on Taxodium, Jackson- Е 1892, Calkins (Е). LAKE CO.: on bark of trees, Eustis, May 28 to June 15, 895, Nash не (MIN,WIS). Marion co.: Ocala, Jan. 29, scm Underwood (F). ESCAMBIA СО.: on oak bark, Pensacola, Dec. 1937, Fassett (WIS). ALABAMA: MARSHALL CO.: open woods, 3.2 miles north of Boaz, July 21, 1939, Hubricht B1548 (MBG). LOUISIANA: ACADIA CO.: on trunk of trees, Sept. 17, 1894, Langlois 997 (D). mus FRANKLIN CO.: Columbus, July 12, 1892, Bogue L84 ); Georgesville, July 15, 1892, Bogue L65 (O). cLINTON co.: New Antioch, Jan. 16, 1901, Vande- mark ( MICHIGAN: OAKLAND CO.: on dead cedar, Thomasville, Aug. 26, 1902, Fink (F). KEWEENAW CO.: on rock face at tip of mat of spruce needles in shade, Tobins Har- bor, зв 14, 1930, Lowe (Е). KEN NOX CO.: open п, 1.4 miles north of Jarvis Store, Aug. 19, 1937, 0 B308 (MBG). 0.: de yin wet woods, 8.4 miles north of East Bernstadt, Aug. 18, 1937, me Jos (MBG). ESSEE: HAMILTON CO.: on pine trees and rocks, Lookout PoS Cal- зи РК (MI N). CARTER CO.: рони Mountain, Thaxter (Е). SEVIER CO.: edge о woods, 2.3 miles north of Gatlinburg, Aug. 25, 1937, Hubricht poo (MBG). CLAIBORNE СО.: open woods near mouth of Indian Creek, Aug. 21, 1937, Hubricht B366 (MBG) ILLINOIS: MENARD CO.: Athens, Hall (WIS). MINNESOTA: LAKE CO.: on rocks, Snowbank Lake, June 26, 1897, Fink = CARLTON СО.: on trees, Tafte, Carlton Peak, July 10, 1897, Fink 527 (MIN IOWA: CLAYTON CO.: on trees, Sept. 1894, Fink (WIS). MISSOURI: ST. FRANCOIS CO.: on trees, Pickle pos May 21, 1938, Berry 1341 E on sandstone ledge, Pickle Springs, May 21, 1938 (езда 1338 (MBG). tree, rocky wooded bank of lu Beer, 8 miles south of EENE “ЕЕЕ э ie Т, mr Eck Berry $ Johnson (MB ARKANSAS: NEWT n bark of tree, steep dry сая bluff with rocky flat top, 1 mile mo x T p 8, 1939, Dodge, Berry 4 Johnson (MBG ys high bluff and long steep peeled hillside, Lookout Point, 7 miles south of Jasper, Apr. 8, 1939, Dodge, Berry 4 Johnson (MBG); moist woods on gentle slope near [Vor. 28 124 ANNALS OF THE MISSOURI BOTANICAL GARDEN falls, 11 miles south of Harrison, Apr. 8, 1939, Dodge, Berry § Johnson (MBG). SEARCY CO.: open wooded hillside, 4 miles west of Marshall, Dodge, "s ф John- son (MBG). BOONE CO.: rocky wooded bluff, above dry creek bed, Bear Creek Spring, 0.5 mile north of Francis, pter 7, 1939, Dodge, Berry & Fanum ү TEXAS: HARRIS CO.: on oak, Houston, Jan. 7, 1917, Fisher (F). BRITISH COLUMBIA: on trees, west coast of Vancouver Island, June 18, 1906, Macoun (F). WASHINGTON: SAN JUAN СО: on log lodge, Olga, July 16, 1906, Fink 344 M). CALIFORNIA: MONTEREY СО.: cypress headland, ia Lobos Reserve, elev. 75 ft., July 11, 1936, Wheeler (MBG). LOS ANGELES CO.: on live oak, divide at head of Зек anning's Canyon, Santa Catalina Island, Apr. 24, n Knopf (F). couNTY INDEFINITE: On roof of an old house on sea beach near Pillar Point, elev. 10 ft., Aug. 14, 1903, Herre (LSU,F OCAL AND FOREIGN EXBIOOATAE EXAMINED: Cum. I 8; Elenk. 52 a-c; тр Lich. Fr.-C, 16; E. Fries, 335; Funck, e Pe Hav. rs prn 23; Larb. Herb. 291 Leight. 76; Lojk. Univ. 111; Malbr. 65; Mig. 90; Oliv. 9; Rab. 912; Rab. Cent. 22; Roum. 38; Ват. 148; ione. e a ae Tuck. 15; Wain. 538b ; Welw. 46a. Parmelia perlata var. ciliata Duby, Bot. Gallic. 2: 601. 1830. Lobaria perlata var. ciliata DC. apud Lam. & DC., Fl. Fr. ed. 3, 2: 403. 1815. Parmelia submarginalis Eaton, Man. Bot. N. Amer. ed. 5, 316. 1836 (non Ach.). Parmelia ciliata Nyl., Flora 61: 247. 1878. DeCandolle used as a type of the variety a plant collected in Switzerland, which is now in his herbarium in Geneva, Switzer- land. Duby cites DeCandolle in his transfer of the variety to the genus Parmelia, but gives no reference to the actual ex- istence of a type. The plants he studied were from eastern France. Nylander's herbarium in Helsinki, Finland, contains a collection designated Parmelia ciliata which may be the type of his species. There is a collection in Tuckerman’s herbarium designated Parmelia submarginalis which is evidently co-type with Eaton's type in Kew. This plant is Parmelia perlata var. ciliata. In his ‘Cat. Lich. Univ.’ 6: 244. 1930, Zahlbruckner cites Parmelia Michauxiana as superseding Parmelia submar- ginalis. Parmelia Michauxiana Zahl. is non-valid because of prior use by Acharius (Lichenogr. Univ. 492. 1810) The variety is the same as the species except that the mar- 1941] BERRY—THE GENUS PARMELIA IN NORTH AMERICA 125 gins of the variety are sparingly beset with strong black cilia and occasionally there are coralloid branchlets and cilia arising from the upper surface of the thallus. The chemical reaction and internal morphology are the same as those of the species. Distribution: Tennessee to Minnesota. TENNESSEE: HAMILTON CO.: on rocks and trees, Lookout Mountain, Calkins 12 (MBG). MINNESOTA: CARLTON CO.: on rocks, Kettle Falls, Aug. 15, 1901, Fink 1496 (MIN); on cedar, Thomasville, Aug. 25, 1902, Fink 5411 (MIN). FOREIGN EXsICCATAE EXAMINED: Malbr. 315; Cromb. 30; Hepp. Zür. 41; Larb. Caes. 17; Larb. Herb. 86. 47. Parmelia cetrarioides Del. apud Duby, Bot. Gallic. 2: 601. 1830. Parmelia olivaria f. cetrarioides Merrill, Bryol. 11: 95. 1908. Delise's type was from northern France, and is now in his herbarium in Strasbourg. Merrill's type was collected in south- ern Canada by J. Macoun and is in the Farlow Herbarium. Thallus large, 8-20 cm. diameter, loosely adnate; upper sur- face somewhat wrinkled, cortex often cracked or broken, ex- posing a white medulla, lobes short, wide, round, greenish-gray to greenish-brown, margins crenulate, slightly raised, without cilia, occasionally with scattered soredia; lower surface black with brown margins, few small black rhizinae at points of con- tact with the substratum; apothecia rare, 2-10 mm. diameter, disks coneave or flat, chestnut-brown, amphithecia thin, smooth, only slightly dissected; K, cortex and medulla gen- erally yellow-green; C, medulla yellow, K + C, yellow, P, red- brown. Algal layer continuous, 28-36 и thick; upper cortex 16—20 и thick ; medulla of loosely interwoven hyphae of large diameter ; lower cortex 32-36 u thick; thallus 180-200 u thick; thecium 48—60 и; ascus clavate, spores 8, 6-8 x 12-16 и; paraphyses branched. Distribution: Nova Scotia to Florida, west to California. A SCOTIA: CUMBERLAND CO.: on trees, Truro, May 10, 1891, Macoun (Type of P pend f. cetrarioides) (F). COUNTY UNKNOWN: Rothrock (FM). QUEBEC: on rock, wet when collected, Ste. Anne, Apr. 27, 1935, Lepage 200 (D); adherent to dry rock, Ste. Anne, May 2, 1935, Lepage 140 (D); on mossy rocks, Ste. Anne, May 15, 1935, Lepage 489 (D) [Vor. 28 126 ANNALS OF THE MISSOURI BOTANICAL GARDEN ONTARIO: on large trunks in woods, Ottawa, May 15, 1891, Macoun (MBG); on IR Goulais Point, Aug. 4, 1869, Macoun (MBG); on trees in a swamp, Beech- ood, Apr. 27, 1892, Macoun (MIN). "Mar INE: KNOX C0.: on trunks of various trees, Rockland, Sept. 28, 1909, Merrill 99 (MIN). EW HAMPSHIRE: 0005 CO.: on rocks, Cherry Mountain, Aug. 8, 1895, Harper $ Harper (FM); base of Mount Washington, Aug. 6, oye Harper $ Harper 47 (FM); near Gate of Notch, June 1882, Faxon (MBG,WIS). GRAFTON CO.: on rocks, Franconia Mountains, July 1891, Cummings (MBG) ; ; on rocks, North Wood- sto ins Franconia 2 July 1891, Cummings 8 (WIS). BELKNAP CO.: Gil- ma 71868, Hall 80 ( . HILLSBORO CO.: Lapham (WIS). CHESHIRE CO.: on eran tenia, z 1850, Russell (FM). VER RUTLAND CO.: порака ge ie Pico, elev. 2500 ft., May 12, 1911, бинн (WIS); Todmer area, , June 26, 1913, Dodge (D); bare ledges, region of the Silver Mine, Birch uL paean July 3, 1920, Dutton (MBG). MASSACHUSETTS: MIDDLESEX CO.: on rocks, New Bedford, Willey 34 (MIN). NEW YORK: YATES CO.: Жү nn Yan, Buckley (MBG). PENNSYLVANIA: PIK : Mat decas 24, 1886, Schneider (MBG,MIN). LANCASTER C0.: Mo далайын May is, 1998, Е 4 (MBG); Conewago, Apr. 23, 1892, Small (MBG,MIN) ; on rocks, Rheinholds, nie 3, 1894, Eby (MBG); Bear Town, Sept. 16, 1895, Eby (MBG); on rails, Landis Valley, Oct. 1892, Eby (MBG); on rocks, Bear Town, Oct. 1895, Eby (M pei ; Kissel Hill, July 17, 1895, Eby (MBG) ; on trees, Kissel Hill, Oct. 1895, j^ (MBG). NORTH nion MITCHELL С0.: on bark of tree, Roan Mountain, June 15, 1936, Schallert (B). HAYWOOD CO.: on oak trees, vicinity of yos I near Wa Хаана elev. 1000—1500 m., Sept. 12, 1910, Standley (US). woods, Clingmans Dome, elev. 6600 ft., Aug. 10, 1939, Hubricht UN preteen MC DOWELL C0.: Mount Mitchell, June 5, 1929, острие 9 (BPI). COUNTY INDEFI- NITE: bark of tree, Great Smoky Mountains, June 13, ues Schallert (B). FLORIDA: ORANGE CO.: on trunks of Myrica, Sanford, 1907, phus (MBG). OHIO: CHAMPAIGN CO.: in dipsa swamp, Aug. 9, ago Were L453 (О) MICHIGAN: KE : Isle Royale, Aug. Stuntz $ Allen 33 (WIS) ; covering пзе реў cliffs, Rock Harbor, gon беен Aug. 1904, Har- N: VILAS CO.: valley of the Los River, near Conover, June 26, 1893, Cheney 798 (WIS). MARATHON CO.: valley of the Wisconsin River near Granite Heights, Summer 1894, Cheney a (WIS). couNTY INDEFINITE: valley of the Wisconsin River, near Doherty Lake, Summer 1893, Cheney 1150 (WIS). MINNESOTA: COOK CO.: on rocks, Grand Portage Island, June 24, pon Fink 169 (MIN); on eedar, Grand Marais, Aug. 2, 1902, Fink 5284 (MIN). со. roeks, Snowbank Lake, June 26, та Fink (D); on trees, Beaver Dar; July ii 1897, Fink 725 (MIN). ST. LOUIS CO.: on mossy rocks, frequent, Harding, Aug. 19, 1901, Fink 1612 (MIN). CARLTON CO.: on roeks, common, Kettle Falls, Aug. 10, 1901, Fink 1366 (MIN). KOOCHICHING CO.: on cedars in swamp, rare, Koochich- ing, July 29, 1901, Fink 998 (MIN); on "di Gunflint, July 1, 1897, Fink 368 (MIN); on pad к D Rainy Lake City, Aug. 5, 1901, Fink 1214 (MIN). COUNTY UN N: › South Fowl Lake, June 26, 1897, Fink 205 (MIN); on trees, ole fitis, dos 5, 1897, Fink 543 (MIN). 1941] BERRY—THE GENUS PARMELIA IN NORTH AMERICA ТЕТ WASHINGTON: SNOHOMISH CO.: on trees, Maryville, Jan. 1926, Grant (Е). ISLAND CO.: Goose Rock, Langley, July 1923, Grant (D). CALIFORNIA: ORANGE CO.: Laguna Beach, elev. 50 ft., June 10, 1930, Reed (MBG) 47a. Parmelia cetrarioides var. rubescens (Th. Fries) DR., Nyt. Mag. 62: 75. 1924. Parmelia perlata f. rubescens Th. Fries, Lichenogr. Scand. 1: 112. 1871. Fries’ type was collected in Norway and is in Upsala. Du- Rietz's type was from the mountains of southern Europe, and is now in the Paris Museum, Franee. The variety has a very limited distribution and may be only an environmental variant. It differs from the species only in the chemical reaction with KOH. In the variety the medulla be- comes reddish-brown when treated with KOH, while CaOCls gives no reaction. Distribution: found only in the southern part of Pennsyl- vania. PENNSYLVANIA: PHILADELPHIA CO.: on fence rails, Chestnut Hill, Oct. 1886, Eby (MBG). LANCASTER СО.: on fences, Kissel Hill, Oet.-July 1895, Eby (MBG). FoREIGN EXSICCATAE EXAMINED: Claud. 66; Harm. Lich. Loth. 288; Mus. Krypt. Exs. Vind. 1365. 48. Parmelia crinita Ach., Syn. Meth. Lich. 196. 1814. Imbricaria crinita Ach. apud Arn., Flora 67: 159. 1884. The type of Parmelia crinita was a collection of Muhlen- berg’s from eastern North America, probably Pennsylvania. The type is now in Acharius’ herbarium in Helsinki, Finland. Arnold cites Imbricaria crinita Ach. as a related species. Nothing is known of a type for Imbricaria crinita. Type locality: eastern North America (probably Pennsyl- vania). Thallus 4—18 em. diameter, loosely adnate; upper surface with many granules or isidia, pale green or gray-green, lobes broad, irregular, margins slightly ciliate, wavy and dissected ; lower surface brown at the margins, black with many rhizinae at the center ; apothecia not common, 3-13 mm. diameter, disks chestnut-brown, concave, amphithecia irregular, often with [Vor. 28 128 ANNALS OF THE MISSOURI BOTANICAL GARDEN isidia or granules, sometimes ciliate ; K, yellow, C, none, К. + С, none, P, none. Algal layer continuous, 28-30 и thick; upper cortex 24-28 и thick, matrix light brown; medulla loosely interwoven with hyaline hyphae of small diameter; lower cortex 35—40 и thick, matrix black or brown ; thickness of thallus 304—320 и; Шести 40-60 и thick ; ascus clavate, spores 8, 6-8 x 12-16 и; paraphyses branched with the tips enlarged. Distribution: Ontario to Florida, west to British Columbia and Washington. ONTARIO: on trees in woods, Brighton, Oct. 18, 1893, Macoun 175 (WIS); on trees in оа Brighton, Oct. 18, 1893, Macoun (MBG,MIN). MAINE: ox co.: Rockland, Sept. 8, 1909, Merrill 124 (MIN); Rockport, Nov. 1904, Mw. 844 (F); on the trunks of an old bireh, solitary, Roekland, Aug. 10, Merrill (Е); 1839, DeLaski (FM ASSACHUSETTS: BRISTOL CO.: on trunks and rocks, New Bedford, Willey 28 (FM); on trees, Hingham, Russell (FM); New Bedford, Willey 38 (FM); New Bedford, 1862-1898, Willey (US). PLYMOUTH со: Plymouth, Mar. 1884, Faxon (F). VIRGINIA: LEE CO.: edge of woods, = Gap, Powell Mountain, 2.7 miles north of Blackwater, Hubricht B393 (M Мовтн CAROLINA: ROWAN СО.: Dunn's ro ntain, - 21, 1894, Small (MIN). HENDERSON СО.: on bark of tree, Flat Rock, Mar. 1, 1921, Schallert 1969 (FM). HAYWOOD CO.: Sunburst, elev. 3200 ft., June 1913, hos (NYBG). rie CAROLINA: AIKEN CO.: Мма, Ravenel (MBG). GEORGIA: THOMAS CO.: Thomasville, Aug. 1906, Taylor (Е); on trees, Thomas- ville, Taylor (F FLORIDA: LAKE CO.: Eustis, May 28, 1895, Nash (MBG). Pind BALDWIN CO.: on magnolia, ана Маг. 29, 1925, p (NYBG on trees, Fai — оа 7, 1924, Evans (NYBG); trees on the beach, Daphne, od 15, 1924, Evans (N tat base of trees near Rock Creek, Feb. 27, 1995, Evans (F). MISSISSIPPI: HARRISON CO.: Biloxi, Sept. 15, 1891, Seymour (F). TATE CO.: 2 miles north of Coldwater, Dee. 27, 1938, E. Anderson (MBG). LOUISIANA: ST. MARTIN CO.: on tree trunks, St. Martinsville, Langlois (NYBG). HIO: FRANKLIN : Georgesville, July 15, 1892, Bogue (O). MICHIGAN: FAYETTE CO.: on tree, 1893, Fink (MBG). TENNESSEE: HAMILTON CO.: on trees, Lookout Mountain, vage e MBG). ILLINOIS: MENARD СО.: on tree, 1878, Hall (FM); Е. Най WISCONSIN: WALWORTH CO.: on trees, Springfield, May 27, A T $ Buell (WIS). MARATHON со.: valley of ак "Wisconsin River, near Knowlton, Summer Keds Veni (WIS). ESOTA: COOK CO.: on rocks, Grand Portage, June 23, 1897, Fink (M); on rocks, Grand Portage, June 25, 1897, Fink 114 (MIN). BLUE EARTH CO.: on trees, e, Mankato, June 23, 1899, Fink 47 (MIN); on trees, rare, Mankato, June 27, 1899, ха 133 (MIN). YELLOW MEDICINE CO.: оп trees and rocks, Granite Falls, 1941] BERRY—THE GENUS PARMELIA IN NORTH AMERICA 129 July 11, 1899, Fink 439 (MIN). CLAY CO.: on trees, rare, near Ulen, July 4, 1899, Fink 228 (MIN); on ipid rocks = shade, rare, near Ulen, July 5, 1899, Fink 258 (MIN). KOOCHICHING CO.: on rocks, Gunflint, July BS. Fink 362 (MIN). CARL- CO.: on trees, fatis, July "n p Fink 133 (MIN). IOWA: MUSCATINE CO.: on bark; Wild-cat-den, m 1897, Shimek (IA). FAY- ETTE CO.: on trees, Mar. 1893, Fink (MIN,WIS,M). MISSOURI: DUNKLIN со.: on tree trunks, St. Francis, Aug. 20, 1897, Trelease BG). WRIGHT CO.: on dead Acer saccharum, 2 miles southwest of Mansfield, elev. 1480 ft., June 5—12, 1911, Lansing Jr. 3200 (FM). ARKANSAS: BOONE CO.: roeky wooded ee above dry creek bed, Bear Creek Spring, 0.5 mile north of Francis, rus 7, 1939, Dodge, Berry 4 Johnson (MBG). KANSAS: DOUGLAS CO.: on trees, Nov. bod Bridwell 1185 ad TEXAS: GILLESPIE CO.: Jermy 424 (M да BEXAR CO.: mesquite woods, 1.5 miles southeast of junction of Jolly Lane and Atkins- де Lane, 18 miles east of San Antonio, May 17, 1940, СЕ В1886 (MBG). BRITISH COLUMBIA: on the old wharf, Comex, Vancouver Island, June 24, 1893, Macoun 149 (MIN); June 1915, Macoun 19 (F); on trunks in woods, Ucluelet, Vancouver Island, May 11, 1909, Macoun (MB WASHINGTON: SKAGIT CO.: on rocks, Aso July 12, 1908, Frye (D). 49. Parmelia cristifera Tayl., Hook. London Jour. Bot. 6: 165. 1847. The type of the species was collected by Wallich in Caleutta, India. The type plant is in Taylor's Herbarium in the Boston Society of Natural History. Thallus 5-15 em. diameter, adnate; upper surface smooth, greenish-gray to ashy or white, lobes round and broad, mar- eins raised, bearing globose powdery soredia; lower surface with brown margins, naked, black at the center with a few black rhizinae; apothecia rare; K, upper cortex yellowish-green, medulla reddish-brown; C, none, K + C, medulla brown, P, red-brown. Algal layer continuous, 16-20 и thick, matrix hyaline; me- dulla hyphae loosely interwoven, hyaline; lower cortex 20—28 р thick, matrix black; thickness of thallus 120-125 и; thecium 60- 96 u thick; ascus clavate, spores 8, 2-4 x 6-8 и; paraphyses branched. Distribution: North Carolina to Florida, west to California. Мовтн га FORSYTH CO.: Bennetts Rock, EL 10, pis Schallert (MBG). k HENDERSON CO.: Hendersonville, Apr. 1910, Plitt 268 (F). CKSON CO.: on bar of tree, Balsam Mountains, Mee 8, i Schallert зад (WI = GEORGIA: THOM oak bark, Thomasville, Aug. 1906, Taylor (F). FLORIDA: DUVAL CO.: on bdo Edi rg Calkins 10 (US). ORANGE CO.: on [Vor. 28 130 ANNALS OF THE MISSOURI BOTANICAL GARDEN fence, Mar. 1907, Rapp Ln m Mar. 1907, Rapp (F); on Melia, uns а May 1905, Rapp 9 (F). : Eustis, June 16, 1895, Nash 2037 (US). BORO CO.: north side of унну vani фе, Aug. 1898, Ferguson (US). LEE CO.: on cypress trees, Fort Myers, Mar. 17, 1916, Standley 326 (MBG). ALABAMA: BALDWIN CO.: on Melia, Fairhope, Feb. 25, 1925, Evans (Е). ARKANSAS: CHICOT CO.: near Macon, Dec. 29, 1938, E. Anderson (MBG). WHITE co.: near Judsonia, Dec. 28, 1938, E. Anderson (MBG). OKLAHOMA: COMANCHE CO.: vieinity of Fort Sill, May 20, 1916, Mrs. J. Clemens (D). ХАБ: HARRIS СО: on Pinus glabra, Clinton, Mar. 22, 1925, Thurow (Е). DENTON CO.: on Quercus minor, 3 miles southwest of Denton, Feb. 1935, Bussard VII $ VIII (MBG). COMAL CO.: rante io 500 ft. above the bed of Guadalupe ннан 1916, ги (MBG). BEXAR CO.: semi-shade to full sun, 1 mile south- of Selma, May 22, 1938, Hubricht Bret (MBG). GILLESPIE CO.: NA 873 (MEG). LA ал23 co.: Millett, Nov. 1897, Trelease (US). CULBERSON CO.: full sun to semi-shade, north side of summit of Yas Horn Mountains, 15 miles south of Van Horn, June 2, 1938, Hubricht B951 (MBG) CALIFORNIA: SAN DIEGO CO.: on dry bushes, Camp Kearney, Mesa, elev. 600 ft., 1932, Cota (MBG). 50. Parmelia perforata ( Wulf.) Ach., Meth. Lich. 217. 1803. Lichen perforatus Wulf. apud Jacq., Collect. 1: 116. 1786. Platisma perforatum Hoffm., Descr. & Adumbrat. Pl. Lich. 1: 65. 2. 13, f. 1. 1790. Wulfen's type was from a plant collected in central Europe. Its present location is unknown, but the plate shown by Jaequin is in agreement with the description. Hoffmann cites Wulfen's Lichen perforatus as synonymous with his plant. His type is unknown. Acharius' plant was collected by Michaux in the eastern part of North America. His type is in Helsinki, Fin- land, with a co-type in Upsala, Sweden. Thallus large, loosely adnate; upper surface smooth, green- ish-gray to brownish-gray, lobes round, wide, with raised mar- gins, becoming subfruticose, dissected and imbricate, with a fringe of black cilia, spotted with numerous black ostioles of spermagonia; lower surface black at the center, with numerous black rhizinae, margins naked, white or brown; apothecia numerous, 1-15 mm. diameter, short-pedicellate, lower sides reticulate, ridged, disks brown to chestnut-brown, concave or flat, usually perforate in the center, margins wavy and broken; K, yellow becoming pink or red, C, none, K + C, pink, P, none. Algal layer continuous, 40-45 и thick; upper cortex 28-30 и 1941] BERRY—THE GENUS PARMELIA IN NORTH AMERICA 131 thick, matrix brown; medulla hyphae loosely interwoven, of one type, hyaline and of small diameter; lower cortex 16- 32 и thick, matrix black; thallus 250-266 и thick; thecium 26- 40 и thick; ascus clavate, spores 8, 5-8 x 9-12 и; paraphyses branched, enlarged at the tip. Distribution: Massachusetts to Florida, west to California. MASSACHUSETTS: BRISTOL CO.: on trunks, common, New Bedford, Willey 31 (FM); trunks, New Bedford, Willey 30 (FM); New Bedford, 1908, Everhart MBG). PLY со: Hingman, 1846, Russell (FM). WORCESTER CO.: rocks, Lunenburg, ДМ. Sargent 219 (F). NORFOLK со.: Wellesley, Dec. 21, 1883, (WIS). Cummings COUNTY UNKNOWN, 1863, Mann (FM); on rocks, Russell (WIS). v tinh NEW HAVEN ps New Haven 1855, Eaton (F). YORK: pics co.: East Port, st: Island, June 24, 1894, Schrenk Hu YATES CO.: ds Buckle NEW JERSEY: CUMBERLAND СО.: on various trees, Vineland, Dee. 1893, Walker 22 (FM,MBG,WIS,MIN). PENNSYLVANIA: CHESTER CO.: on rocks, Nov. 10, 1879, т (ЕМ); о living and dead trees, common, Маг. 27, 1879, May $ Rothrock (FM). LANCASTER CO.: on oak trees, EE Nov. 25, 1852, Жой (Е); оп MN Rheinholds, Mar. 14, ie ы (MBG ND: BALTIMORE CO.: on fence x Agen 19, 1909, Plitt (F). CAROLINE Co.: on Чет tree, Denton, Apr. 1898, Waugh (Е). NORTH CAROLINA: FORSYTH CO.: оп Mess Niford Rock, June 1, 1934, Schallert (B); on bark of tree, Winston-Salem, Apr. 12, 1921, Schallert 1 EI BUNCOMBE CO.: vicinity of Montreal, Sept. 1, ка Standley Ф В an (Е). HENDERSON CO.: оп bark of tree, Flat Rock, Маг. 1, 1921, Schallert Wem (FM); on bark of tree, Flat Rock, Mar. 1, 1922, mid га (FM); ; on exposed rock, Flat Rock, Mar. 12, 1921, Schallert 1289 (FM SouTH CAROLINA: BEAUFORT CO.: iiid sand dunes, Pawleys Island, Aug. 2, 1936, Schallert (B). ec e co.: Glenn Springs, Ravenel (Т). FLORIDA: ORANGE CO.: trunk of orange tree, Lake т Feb. 14, 1909, Rapp (Е). DUVAL со.: shrubs, mE Calkins (MBG). AY CO.: June 21, 1879, Martin (FM). couNTY UNKNOWN: winter of fy gne Su (FM ALABAMA: MOBILE CO.: Mobile, 1852, күр e (м); Mobile, Mohr (MBG). LEE CO.: limbs, porn Jan. 9, 1897, E (o Toi r (MBG). LOUISIANA: sionis co.: on trees, Pointe a la Hache, 1884, Langlois 36 MBG). LA FOURCHE Kraemer, je 18, 1936, Hubricht (MBG). couNTY UN- KNOWN: m 7, 1889, Langlois 863 (M pir Он Mineral рад Oct. зе Kellerman (O). CLARK со.: D der т они, Teb. Mos е (O); near Springfield, 1872, Spence (O). CHAMPAIGN CO.: eds reali Sept. E 1892, Weaver (W). WEST VIRGINIA: COUNTY UNKNOWN: on trees, Gray (F). TENNESSEE: CLAIBORNE CO.: open woods, 8 miles east of Tazewell, Aug. 22, 1937, Hubricht B368 (MBG). HAMILTON CO.: on pine bark, Lookout Mountain, Calkins 7 (FM). [Vor. 28 132 ANNALS OF THE MISSOURI BOTANICAL GARDEN WISCONSIN: LAFAYETTE CO.: Fayette, psd 15, “ger shes А 9220 (WIS). Fein MENARD СО.: on trees, 1878, Hall (FM); M) ; Athens, 1864, 1 (FM); Athens, 1878, Най (FM,T,BSNH dd COUNTY UNKNOWN: on zn Calkins (FM). MINNESOTA: BLUE EARTH CO.: on trees, rare, Mankato, June 26, 1899, Fink 134 (MIN). ТОМА: HENRY CO.: on bark, Jan. 1, 1898, Savage (ТА). MISSOURI: WAYNE СО.: old trees, Williamsville, Mar. 1898, C. Russell (MBG). FRANKLIN СО.: bark of tree, Gray Summit, Nov. 15, 1936, Berry (MBG). ARKANSAS: NEWTON CO.: high bluff and long € wooded SPUREN Lookout Point, 7 miles south of Jasper, Apr. 8, 1939, Dodge, (M steep dry wooded bluff with rocky flat top, 1 mile кена P ba. Apr. 8, 1989, Dodge, Berry ф Johnson (MBG); wooded hillside, 11 miles south of Harrison, Apr. 8, 1939, Dodge, Berry 4: Johnson (MBG). conway co.: Petit Jean State Park, May 5, 1940, Hubricht B1917 (MBG). KANSAS: DOUGLAS CO.: on trees, Nov. 1896, Bridwell 87 (F). OKLAHOMA: COMANCHE CO.: vicinity of Fort Sill, May 20, 1916, Mrs. J. Clemens (MBG) TEXAS: GALVESTON CO.: Galveston, Hall (FM). HARRIS CO.: aud A: (FM); on living oaks, La Port, elev. 15 ft., Jan. 9, 1939, Fisher (MBG). TON co.: three miles southwest of — 1935, Bussard (MBG). GILLESPIE CO.: uos Creek, Jermy 423 (MBG). BEXAR CO esquite woods, 1.5 miles southeast of june- tion of Jolly Lane and Atkins- Eimendor? Lane, 18 miles east of San Antonio, May 17, 1940, Hubricht B1892 (MBG); semi-shade to ful i sun, 1 mile southeast of Selma, May 22, it p B768 (MBG CALIFORNIA: SONO : mossy sandstone, Stony Point near Searsville, ме 400 ft., July 9, 1903, Sick (F). LOS ANGELES CO.: reus top, Santa Catalina Island, June 12, 1920, Nuttall (FM). SAN DIEGO 00.: on uL Alpine Creek, elev. 1000 ft., July 28, 1903, Herre (MIN) ; San Diego, 1888, Palmer (MBG). 50a. Parmelia perforata var. hypotropa Nyl., Syn. Lich. 1: 318. 1860. Nylander's type of the variety was collected on trees in Texas, and is in his herbarium in Helsinki, Finland. The variety differs from the species only in the small size of the thallus and the light lower surface. Distribution: Florida west to Texas. FLORIDA: DUVAL CO.: on trees, mouth of St. Johns River, Mar. 31, 1921, Kelly (BPI); PALM BEACH CO.: 1898, Thaxter 2043 (F). TEXAS: BLANC .: on oaks, Blanco, 1848, Wright (T). BRAZOS co.: Bryan, June 22, 1919, Psi. (BPI). 51. Parmelia proboscidea Tayl. apud Mack., Fl. Hibern. 2: 143. 1836. Parmelia frondifera Merrill, Bryol. 11: 91. 1908. 1941] BERRY—THE GENUS PARMELIA IN NORTH AMERICA 133 Taylor's type was collected in Dunkerron, Ireland, and is now in his herbarium in the Boston Society of Natural History. Merrill’s type was collected by Macoun at Peele Point, Ontario, Canada, and is now in the Farlow Herbarium. Thallus large, round, adnate, upper surface covered with finely divided lobules, white globose soredia common on the older parts, margins with fine cilia; lower surface brown at the margins, black at the center, with many black rhizinae; apothecium 9-15 mm. diameter, pedicellate, abundant, disk deeply concave, chestnut-brown, not perforate, amphithecium sometimes fissured with vertieal ridges; K, yellow, C, none, К + C, none, P, none. Algal layer continuous, 18-30 и thick; upper cortex 36-38 и thick, matrix brown; upper portion of medulla of loosely inter- woven, small, hyaline hyphae, lower portion of brown hyphae of large diameter; lower cortex 24—30 и thick, matrix dark brown or black; thickness of thallus 105-120 и. Distribution: Ontario to Florida, west to Texas. ONTARIO: on trees, Peele Point, 1890, Macoun (F). MASSACHUSETTS: BRISTOL CO.: on trunks, common, New Bedford, Willey 31 M). MARYLAND: ANNE ARUNDEL CO.: on decaying trunk of tree in woods, Cornfield Creek, Aug. 26, 1906, Plitt (D). FLORIDA: ORANGE CO.: on trees, Sanford, May 5, 1922, Rapp (D) ; on oak trunks, Sanford, Aug. 12, 1906, Rapp (D); on trees, Sanford, Mar. 1920, Rapp (WIS); Sanford, Jan. 24, 1909, Rapp 26 (F). лее сие COUNTY UNKNOWN: оп bark of elms, Calkins 2 (FM). 0.: semi-shade to full sun, 1 mile southwest of Salem, May 22, Ка pote 25 (MBG). | FOREIGN EXSICCATAE EXAMINED: Wain. 400. 52. Parmelia tinctoria Despr. apud Nyl., Flora 55: 547. 1872. Parmelia praetervisa Müll.-Arg., Flora 63: 276. 1880. Despreaux's type was collected in the Canary Islands and is now in the Paris Museum, France. Müller-Argau's type was collected by Zollinger in Java, and is now in Geneva, Switzer- land. Thallus large, 15—35 em. diameter, loosely adnate to bark or wood; upper surface with isidia or coralloid branchlets, wrinkled, light gray to white or ashy in color, lobes large, [Vor. 28 134 ANNALS OF THE MISSOURI BOTANICAL GARDEN round, slightly erinkled, imbrieated, with smooth margins; lower surface brown at the margins, black toward the center with very few scattered black rhizinae at the places of contact with the substratum; apothecium 2-8 mm. diameter, subpedi- cellate, disk concave, light brown to chestnut-brown, amphithe- cium smooth or slightly isidiose; К, upper cortex yellow; С, medulla none, K + C, none, P, none. Algal layer continuous, 16-20 и thick; upper cortex 16-18 и thick, matrix yellow; medulla loosely interwoven, with one type of hyphae; lower cortex 18-20 и, matrix dark brown or black; thallus 160-200 и thick; thecium 6-70 y thick; ascus clavate, spores 8, 4-5 x 8-10 и; paraphyses branched. Distribution: North Carolina to Florida, west to Texas. NoRTH CAROLINA: FORSYTH CO.: on exposed rock and base of tree, Box Moun- tain near Walkertown, July 11, 1922, Schallert 6767 (WIS). GEORGIA: THOMAS СО.: on trunks, Thomasville, Taylor (Е); Thomasville, Aug. 24, dodici Mood (Е). : DUVAL С0.: Jacksonville, Calkins (MBG). ORANGE CO.: on fence, San- Pi. ay 1921, Rapp (WIS); on trunks of trees, Sanford, 1909, Rapp (F); on fence, Sanford, Dee. 1906, Rapp (Е). LAKE CO.: Eustis, May 28 to June 15, 1895, Nash (F,MBG). HILLSBORO CO.: north side of dead stump, Tampa, Aug. 1898, Ferguson (MBG); north side of small oak tree, Tampa, Aug. 7, 1898, Ferguson (MBG). ALABAMA: BALDWIN on tree, Shelby’s, Mar. 17, 1925, Evans (F). MOBILE 00.3 Il Jan. 1905, one (MBG). LOUISIANA: ORLEANS : hardwoods near Indian Ме d near New 2 Dee. 31, po Dodge (ИВ Do. LA FOURCHE CO.: on tree, Kraemer, sin ер. Hubricht (D). ST. MARTIN CO.: mostly on trunks of trees, St. Marti ville, gies 1894, — LIE on fences and various trees, St. Martingville, 1893- oe Langlois (MBG, ARKANSAS: CHICOT CO.: near Macon, Dec. 29, 1938, Anderson (MBG). TEXAS: HARRIS CO.: Clinton, Mar. 5, 1925, Thurow (F). 53. Parmelia latissima Fée, Suppl. Essai Crypt. Ecor. Offiein. 119. 1837. Fée's type was collected on the island of Jamaica and is now in the Paris Museum, France. Thallus large ; upper surface with soredia, ashy to greenish- gray, lobes short, wide, margins round, smooth, somewhat ascending ; lower surface chestnut-brown, black at the center, rhizinae scattered, black; apothecium 2-10 mm. diameter, ses- sile, disk concave, chestnut-brown, amphithecium smooth; K, red, C, intense red, K + C, none, P, brick-red. 1941] BERRY—THE GENUS PARMELIA IN NORTH AMERICA 135 Algal layer continuous, 20-40 и thick; upper cortex 2-24 и thick ; medulla of closely interwoven hyphae; lower cortex 20- 24 y thick, matrix black, thallus 200-250 и thick; thecium 140- 181 и thick; ascus clavate, spores 8, 12-16 x 32-36 и; paraphyses branched, enlarged at the tips. Distribution: Florida west to Louisiana, north to Arkansas. FLORIDA: DUVAL с0.: on trees, Jacksonville, Calkins 341 (FM). ORANGE СО.: on trunks of trees, Sanford, 1909, Rapp (MIN); Sanford, Jan. 24, 1909, Rapp 26 (F); on logs, Sanford, Mar. 1921, Rapp (F); on trees, Sanford, Oct. 1923, Rapp (BPI); on logs in damp woods, Apr. 1906, Rapp (BPI); vicinity of Oviedo, Dee. 1922, Rapp 660 (BPI). CLAY co.: Jan. 20, 1879, Martin (FM); COUNTY UN- KNOWN: on cypress swamps, Calkins (NYBG). ALABAMA: COLBERT CO.: base of tree, near Rock Creek, Feb. 27, 1925, Evans (NYBG). LOUISIANA: ST. MARTIN CO.: on fences and various trees, St. Martinsville, 1893— 1894, Langlois (WIS,F). ARKANSAS: NEWTON CO.: 1899, Trelease (MBG); high bluff and long steep wooded hillside, Lookout Point, 7 miles south of Jasper, Apr. 8, 1939, Dodge, Berry ф Johnson (MBG). PSEUDEVERNIA IV. Parmelia subgenus Pseudevernia (Zopf) Berry. n. sub- genus. Parmelia section Canalicularia Ach., Meth. Lich. 254. 1803. Parmelia section Everniiformes Hue, Nouv. Arch. du Mus. Paris, ТУ.1: 130. 1899. Pseudevernia Zopf, Beih. z. Bot. Centr. 14: 124. 1903. Parmelia subgenus Ецрагтейа section 1. Everniaeformes Zahlbr. apud Engler & Prantl, Nat. Pflanzenfam. I. Teil, Abt. 1: 212. 1907. Acharius used the section Canalicularia in his ‘Meth. Lich.’ 254. 1803, to segregate those lichens which have long and narrow erect lobes. This group included Parmelia villosa Ach., which has been placed in the genus Teloschistes by Norman (Nyt. Mag. 7: 309. 1853), and Parmelia furfuracea (L.) Ach., which has been cited as typical by all subsequent workers on this group. Of the remaining twenty species included by Acharius, all have subsequently been placed in other genera. Parmelia furfuracea (L.) Ach. is typical of Hue's Коегт- iformes. Zopf established Pseudevernia as a genus to include the lobate forms, and considered Parmelia furfuracea (L.) [Vor. 28 136 ANNALS OF THE MISSOURI BOTANICAL GARDEN Ach. as the type. Zahlbruckner used Everniaeformes to desig- nate the section of the subgenus Euparmelia т which he placed the group. His type was Рагтейа furfuracea (L.) Ach. The writer believes that the lobate forms of the genus Parmelia are distinct enough to warrant their being placed in a distinct sub- genus, and has designated the use of Pseudeverma Zopf in this category, with Parmelia furfuracea (L.) Ach. as the type. Thallus lobes long and narrow, more or less fruticose in growth, usually imbricated, and in some species with strong black marginal cilia; lower surface channeled, rhizinae papil- lose or absent except at the center of the thallus; spores 8, to an ascus. The subgenus is distributed throughout the world in the mountainous regions. KEY TO SPECIES IN SUBGENUS PSEUDEVERNIA A; Always With mpna ЗИНА: ooo ace rra Rr e Sr e o n 54. cirrhata A. Never with marginal cilia. B. Lobes flat at tip; plant coarse, usually with soft scales....... 55. furfuracea B. Lobes cylindrical at tip; plant always delieate................ 56. Cladonia 54. Parmelia cirrhata Fries, Syst. Orb. Veget. pars 1: 283. 1825. Evernia kamtschadalis Mont., Ann. Sei. Nat. Bot. IT. 18: 17. 1842. Everma americana Mey. & Fw., Nova Acta Acad. Leopold.- Carol. 14: 211. 1843. Parmelia Nepalensis Tayl., Hook. London Jour. Bot. 6: 172. 1847. Parmelia americana Mont., Ann. Sci. Nat. Bot. IIT. 18: 309. 1852. Parmelia camtschadalis f. americana Nyl., Ann. Sci. Nat. Bot. ТУ. 11: 215. 1859. Fries’ type was collected in Nepal, India, and at present is in Upsala, Sweden. Montagne described Evernia kamtschadalis from a plant collected in Chile, South America. He considered it as closely related to Borreri furfuracea Ach. (Lich. Univ. 500. 1810) but with thinner lobes and ciliate margins. Montagne’s type is in Ше Paris Museum, France. Meyer and Flotow based 1941] BERRY—THE GENUS PARMELIA IN NORTH AMERICA 137 their Evernia americana on a plant collected in Bolivia, South America. Their type is in Berlin, Germany. Parmelia Nepa- lensis of Taylor was based on a collection from Nepal, India, by Wallich, and the type is in Hooker's Herbarium in London, England. Montagne considered the same type in Parmelia americana as he used to describe his Evernia kamtschadalis. Parmelia camtschadalis f. americana of Nylander was based on collections from Bolivia and Peru, South America. His type is in Helsinki, Finland. Parmelia cirrhata Fr. represents a species with wide distribution in the mountains of the tropies andis an extremely variable species. Thallus fruticose, upper surface smooth, without soredia or cracks in cortex, lobes flattened, more or less erect, from 2 to 9 mm. broad, tips dichotomously forked, margins always with numerous strong black cilia; lower surface black, smooth or with many strong black rhizinae; apothecium 4-14 mm. di- ameter, concave to flat, chestnut-brown, margin of amphi- thecium broken or entire; K, none, C, none, K + C, none, P, yellow. Algal layer continuous, 7-20 y thick; upper cortex 40-50 и thick, matrix of yellow secretion ; medulla hyphae loosely inter- woven, hyaline, of one type and less than 1 и diameter; lower cortex 30-40 и thick, matrix black; thallus 185—200 y thick ; the- cium 35—45 y thick; ascus clavate, spores 8-10 х 18-20 y, ellip- soid ; paraphyses branched. Distribution: Collected but once in North America, north of Mexico. CALIFORNIA: SANTA CRUZ CO.: on trees and shrubs, Black Mountain, elev. 1500 to 2700 ft., July 9, 1903, Herre (F,LSU). This species has a wide distribution in Mexico and Central America. 55. Parmelia furfuracea (L.) Ach., Meth. Lich. 254. 1803. Lichen furfuraceus Linn., Sp. Pl. 1146. 1753. Lichenoides furfuraceum Hoffm., Desc. & Adumbrat. PI. Lich. 1: 45. 1.9, f. 2. 1790. Lobaria furfuracea Hoffm., Deutschl. F1.2: 144. 1795. [Vor. 28 138 ANNALS OF THE MISSOURI BOTANICAL GARDEN Pseudevernia furfuracea Zopf, Beih. z. Bot. Centr. 14: 124. 1903. The type of Linnaeus was collected in the mountains of cen- tral or northern Europe, and is now in the Linnean Herbarium in London. Hoffmann (Desc. & Adumbrat. Pl. Lich. 1: 45. t. 9, f. 2. 1790) gives a description and presents a plate which is clearly the same plant designated Parmelia furfuracea by Acharius. Hoffmann considered Lichenoides furfuraceum as synonymous with his Lobaria furfuracea. Acharius cites Li- chen furfuraceus Linn. and Lichenoides furfuraceum Hoffm. as synonymous with Parmelia furfuracea Ach. Zopf con- sidered his Pseudevernia furfuracea as synonymous with the Linnaean plant. Thallus greenish-gray or ashy ; lobes long, narrow, dichoto- mously and subpinnately branched, often imbricated, much divided toward the tips, usually bearing soft scales of coralloid branchlets; upper cortex smooth, unbroken by cracks; lower surface light at tips, becoming dark, channeled toward center, without rhizinae except in dark central portion, rhizinae black, short, thick ; apothecium short-pedicellate, almost marginal, 3— 15 mm. thick, flat or convex, disk brown to brownish-black, abundant, amphithecium smooth, regular, few or no fissures or lobes; K, upper cortex yellow; C, none; K + C, upper cortex yellow, P, medulla yellow. Algal layer continuous, 21-35 и thick; upper cortex 20-30 и thick, matrix yellowish-brown ; lower cortex 20-30 и thick, ma- trix black or dark brown; medulla of two types of hyphae, those of the central portion of thallus hyaline, 0.5-1 и thick, loosely interwoven, those at tips of lobes scattered, dark or black, 2— Зи thick; thallus 340—400 и thick; thecium 25-30 y thick; ascus clavate, 8-spored, 3—5 x 5-8 и, arranged irregularly in ascus. European collections of this species resemble the American collection in every respect. In no instance was it possible to distinguish the plants on the basis of the origin of collection. Distribution: Maine to South Carolina, west to California. MAINE: KNOX CO.: on large branches of oak, Pine Hill, Rockport, Oct. 16, 1909, Merrill (F). CUMBERLAND CO.: on Pinus rigida, West Falmouth, Mar. 31, 1925, Morton (F). NEW HAMPSHIRE: CARROLL CO.: Mount Whitefaee, Aug. 1934, Dodge (MBG). 1941] BERRY—THE GENUS PARMELIA IN NORTH AMERICA 139 MASSACHUSETTS: BRISTOL CO.: New Bedford, 1908, Everhart (MBG). CONNECTICUT: TOLLAND CO.: on trees in a bog, Willington, June 15, 1927, Evans PENNSYLYAM: PIKE 00.: Matamoras, Sept. 22, 1896, Schneider (MBG). LAN- CASTER CO.: fence rail, Chestnut Hill, Oct. 15, 1894, Hby (MBG); Chestnut Hill, in век vid (MBG). ROLINA: AVERY CO.: on eoniferous trees, Grandfather Mountain, Aug. 10, Хая, eae (MBG). BUNCOMBE CO.: pine trunk, vicinity of Montreal, Sept. 1, 1913, Standley $ Bollman 10239 (MBG). COUNTY INDEFINITE: mountains of North Carolina and Georgia, Buckley (MBG). SOUTH CAROLINA: GREENVILLE С0.: Hog Back, Dee. 29, 1892, Green (MBG). WEST VIRGINIA: POCAHONTAS CO.: Charles ONE Oet. 1923, Gray L245 (F). KENTUCKY: BELL CO.: woods, north side of Pine Mountain, 5.5 miles east of Pine- ville, Aug. 19, 1927, арен B342 (MBG). MINNESOTA: COOK CO.: Grand Marais, cedars in a swamp, Aug. 4, 1902, Fink (MIN). TEXAS: BREWSTER CO.: on Pinus cembroides on north side of Mt. Emory, E Mountains, July 2, 1837, Warnock (MBG); on Pinus ponderosa, ridge between Manning and Little Ajuga Canyon, June 17, 1931, Moore $ Steyermark 3300 cwm Сот, EL 0: common on spruce, Palmer Lake, Bettel (F); near Papa Pei Ars NU Baker 95 ( NEW MEXICO: EDDY CO.: on tree, near Deen: Aug. 12-30, 1924, Standley 40685 (Е). COLFAX CO.: moist soil, vicinity of Ute Park, Aug. 20, 1916, Standley 13369 (F). SANDOVAL CO.: on branches of balsam and spruce trees, Sandia Mountains, Aug. 8, 1914, Ellis (MBG). socorro со.: shade, San Andres Mountain, Rhodes Pass, 47 miles west of Tularosa, June 1938, Hubricht B987 (MBG). ARIZONA: PIMA CO.: dead branches overhanging ioc AE D Rincon Mountains, Oct. 12, 1909, Plummer 3470 (MBG). UNKNOWN: rolling andesitie pine land recently lumber es Barfoot ud pu 26, 106, Plummer 1429 B MBG). CALIFORNIA: SAN BERNARDINO bark and dead wood, Pseudotsuga macro- carpa, San Antonio Canyon, San Gabriel Mountains, elev. 4800 ft., Mar. 12, 1933, Wheeler 1526 (MBG). FOREIGN AND LOCAL ExsiccaTAE EXAMINED: Ekart, 8; E. e 140; Funck, I = II 113; Mus. Krypt. Exs. Vind. 156; perg ped Oliv. 165; Schaer. 387; P. furfuracea E ler Mus. Krypt. Exs. Vind. ; P. furfuracea var. isidi- ym. Mus. Krypt. Exs. Vind. 876 a, b; P. Mosa subspec. olivetorina Mus. . 1046; ra В, 189; 5 . curta Bri ; E. furfuracea f. nuda Britz. 523; E. furfuracea a. platyphylla Rab. pn p De E. furfuracea f. scobicina Hav. 252; Johns. 66; E. furfuracea var. soralifera Harm. 64 56. Parmelia Cladonia (Tuck.) DR., Svensk Bot. Tidskr. 18: 390. 1924. Evernia furfuracea (L.) Mann var. Cladonia Tuck., Syn. Lich. New England, 12. 1848. [Vor. 28 140 ANNALS OF THE MISSOURI BOTANICAL GARDEN Tuckerman's type of the variety was based on a collection made in the White Mountains of New Hampshire, and is now in his herbarium in the Farlow Herbarium at Cambridge, Massachusetts. DuRietz cites Tuckerman’s type in his trans- fer of the variety to specific rank in the genus Parmelia. Thallus light greenish-gray, with ascending habit of growth, branching dichotomously, branches smooth, very slender, round and light yellow above but flattened and black toward the base; apothecia unknown; K, brown, C, none, K + C, brown, P, yellow. Algal layer broken into masses 42-44 y in diameter; upper cortex 31.5 и thick ; lower cortex 31.5 и thick ; medullary hyphae loosely woven, 2.1 и in diameter; thallus 222—240 y thick. Parmelia Cladonia ('Tuck.) DR. is distinguished from Par- melia furfuracea (L.) Ach. by having narrow smooth branches, without isidia or coralloid branchlets. No collections have been made which have apothecia. Distribution: Nova Seotia to North Carolina. Nova ScoTIA: on spruce tree on gypsum cliff, Ste. Croix, Hants County, July 27, 1928, Prince дун ). МА! 0.: on Ma Camden, Oet. 1885, Merrill (F) ; on branches, Mt. eae с Denes. May 1 910, Merrill (F). OXFORD co.: on black spruce in deep bog, Canton, July 3 er hap (BPI). w icd indt CARROLL CO.: Whiteface Mountain, Aug. 1934, Dodge (MBG). GRAFTON CO.: on gene ee trees, Mt. Liberty, ыа Mountains, Aug. 1892, Cummings $ Seymour 49 (MBG). coos со: White Mountains, top of Mt. Моб, Sept. 1885, Farlow 397 (Е); White Mountains, Tuckerman (Туре) VER T: LAMOILLE twigs of spruce, summit of Mt. Mansfield, 1922, SE pie MBG); Mt. жойы dead tree, June 20, 1924, Dutton 1694 (MBG) ; on dead balsam, region Mt. Mansfield, Hell Break Trail, Dutton 1684 (F). WINDHAM CO.: old barn timbers, Pike Hollow, Wardsboro, July 3, 1935, Moore (MBG). NEW YORK: ESSEX с0.: on balsam, Mt. Marcy, Lake Placid, June 10, 1935, Darrow 858 (M у MARYLAND: ALLEGHENY CO.: swamp, beyond Frostburg, Aug. 30, 1914, Рин 472 (BPI). NORTH CAROLINA: MITCHELL CO.: on twigs, Roan Mountain, June 15, 1936, Schallert 11355 (B). мс DOWELL со: Mt. Mitchell, June 15, 1929, Youngken ip Y 00.: on twigs of trees, elev. 5000 ft., Grandfather Mountain, June 16, 1923, "Baal ri (F West VIRGINIA: POCAHONTAS CO.: Greenbank, Spring 1929, Gray 11126 (NYBG). 1941] BERRY—THE GENUS PARMELIA IN NORTH AMERICA 141 TENNESSEE: SEVIER CO.: deep wet woods, Clingmans Dome, elev. 6600 ft., Aug. 10, 1939, Hubricht B1581 (MBG) ; woods, Newfound Gap, elev. 5500 ft., Aug. 25, 1937, Hubricht B552 (MBG); cool wet deep forest, Clingmans Dome, Aug. 25, 1937, Hubricht B567 (MBG). SUMMARY On the basis of a critical morphological study, four sub- genera, fifty-six species, and ten varieties are recognized in North America, north of Mexico. Three new species have been deseribed and two new combinations formed. Reducing Zopf's genus Pseudevernia to subgenerie rank a new subgenus was established to include those lichens placed in the subgenus Euparmelia section Everniaeformes by Zahlbruckner. DOUBTFUL AND EXCLUDED SPECIES Parmelia ambigua (Wulf.) Ach., Meth. Lich. 207. 1803 = Parmeliopsis ambigua (Wulf.) Nyl., Syn. Lich. 2: 54. 1860. Parmelia ambigua var. Halei Tuck., Syn. N. Amer. Lich. 66. 1882 = Parmeliopsis ambigua var. Halei ('Tuck.) Zahlbr., Cat. Lich. Univ. 6: 13. 1929. Parmelia arizonica (Тиск.) Nyl., Lich. Jap. 104. 1890, should be in the family Gyrophoraceae. Parmelia atrofusca (Schaer.) Crombie, Grevillea 7: 99. 1879, insufficient data. Parmelia colpodes ( Ach.) Nyl., Syn. Lich. 404. 1858 = Anzia colpodes ( Ach.) Stizen., Flora 45: 243. 1862. Parmelia Frankliniana Tayl. Hook. London Jour. Bot. 6: 167. 1847, insufficient data. Parmelia Halseyana Tuck., Boston Jour. Nat. Hist. 3: 442. 1841, belongs in the genus Lecanora. Parmelia margaritata Hue, Nouv. Arch. Mus. Paris IV. 1: 193. 1899, insufficient data. Parmelia stuppea Tayl., Hook. London Jour. Bot. 6: 175. 1847, belongs in the genus Lecanora. ACKNOWLEDGMENTS The writer wishes to express his appreciation to those who have so generously helped in this study. He is greatly indebted [Vor. 28 142 ANNALS OF THE MISSOURI BOTANICAL GARDEN to Dr. George T. Moore, Director of the Missouri Botanical Garden, for the privileges extended in the use of the library and herbarium. Thanks are also due Dr. David Linder and the other members of the staff of the Farlow Herbarium for their kindness in allowing the use of their herbarium and library in this study. The writer wishes especially to express his grati- tude to Dr. C. W. Dodge, under whose guidance this study has been conducted, for giving so liberally of his time, making con- structive criticisms, and extending the use of his private libr ary and herbarium. List or ExsrccATAE, WITH ABBREVIATIONS Usrp Bal: Balansa, Plantes du Paraguay, 1888. 4176, 4210, 4234. Bartl.: р & Hampe, Vegetabilia Cellularia, 1832-1845. Пес. ТУ 7, VIII 6. Bohl.: Bohler, Liehenes Britanniei or Specimens, 1835-1837. 11, 110. Breut.: LE Flora ‘tz м с 1903-1. А . Claud.: Claudel & Harmand, Lichenes Gallici praecipui Exsie- eati, 1903-1913. 15, 17, 19, 66, 117, 173, 258, 261, 506. Cromb.: Crombie, Lichenes Britanniei oni 1874. 26, 27, 30, 139, 141, 145. Cum.: Cummings, Decades of North American лы, 1892-1903, I 8,19 а-, Т 10,11 * I 23, I 102, I 106, I 107, I 274, І ps ; II 23, II 27, II 56, II 274. Cum. Lieh. Bor. Am.: Cummings, . Desm. ed. II. Ser. Я games de Poi 1836-1851. I ll I 1592, I 1593; II 586. Ekart.: Ekart, Kryptogamische Gewächse Koburgs. 8. Elenk.: Elenkin, Lichenes Florae Ros- siae, 1904. 5, 52 a-c, 53 а-с, 54 ас, 102 a, b, 106 a-b. Erb.: Erbario Crit- байо бамы 1858-1807. I 17, I 118, I 119, I 465, 2 832, I 931, E 1221. те llm Fellman, Lichenes Arctici, 1863. 77, 79, 80, 82, 83. lag. Alg.: M: Lishendo Algerienses, 1892. 11. Flag. Lich. Fr.-C.: Flagey, pes de d he-Comté, 1882-1888. 16, 17, 72, 73. Flk.: Е Floerke, pu Lichenen, 1815-1 ian. $4. Fl Hung.: Flora Hungarica Exsiccata, 13. E. Fries: Fries, E., Lichenes Sueeiae Exsiecati, 1818-1852. 48, 140, 166, = 168, 169, 260, 307, 335, 337. Т. Fries: Fries, eodor M., Lichenes Gonaitinaviae, 1859-1865. 6. Funck: Funck, Cryptogamische Gewiic ме p^ Fiehtelgebirg's, 1801-1838. I 11, I 141, I 374, I 661; II 107, II 108, II 109, II 113. Gar.: Garovaglio, Lichenes Provinciae Сидра et УаШв-ТеШпае, 1837—49. I 8. fr rm.: fee armand, Lichenes Gallici Rariores Exsiccati, 1909. , 67, 69. Harm. Lich. Loth.: Harmand, Lichenes in Lotharingia. 258, 278, 288, 289, 294, 297, nus p 319. Hav.: Havaas, Li- chenes Norvegiae Exsieeati, 1901-1914. 78, 153, 179, il 181, 182, 222, 223, 251, 252, 314, 325, 390, 444. Hav. Occid.: Havaas, Liche IE" ee Exsieeati, 1912-1913. 18. Hepp, Zür.: Hepp, Fl ar Züri 1. Liehenes Novae Angliae, 1911—1914. 27, 28. Johns.: J ae ES of Haglund Lichen-Herbarium, 1894-1910. 23, 66, 67, 69, 221, 303. Larb. Cae a ae Lichenes Caesarienses et Sargienses Exsiccati, 1867-1869. 17, e, 65, 6. Larb. 1941] BERRY—THE GENUS PARMELIA IN NORTH AMERICA 143 Herb.: Larbalestier, Lichen-Herbarium, 1879-1881. 86, 124, 211, 291, 293. Leight.: m Lichenes Britanniei Exsieeati. 76, 203, 365. Lojk. Hung.: Lojka, Li- chenes Regni Hungarici, 1884. 119. Lojk. Univ.: Lojka, Lichenotheca Uni versi 62, 63, 111, 158. Malbr.: Malbranche, " Lichens de Normandie, 1863. , 268, 269, 270, 315, 370. Malme: Malme, р a Exsic- 23 E 61, 66, 67, 133, 176, 178, 204, 230, 265, 405, an s de Madére. 11. Mass.: Massalongo, Lichenes Italiei Е 1855- 1856. "y prety ОКУ, Lichenes Rossiae Exsieeati. 4, 54, 55. Merr.: Мег- rill, Lichenes Exsiccati, 1909-1912. 27, 35, 75, 96, 221, 287, 251. Mig.: Migula, Cryptogamae Germaniae, Austriae et Helvetiae Exsiccatae, 1902—1904. 90. Moug. et Nest.: Mougeot et Nestler, Stirpes Cryptogamae Vogeso-Rhenanae. 249, 315, 353, 634, 1428. Mudd.: Mudd, Lichenes Britannici Exsiccati, 1861. 66. Mus. rypt s. Vind.: Kryptogamae E. editae Museo Palatino Vindobonensi, 18041-1915. 156, 768, 876 а $ b, 1046, 1365, 3163. Norrl. et Nyl.: Norrlin et Ny- lander, Herbarium Lichenum SR 26, 27, 202, 209, a, ч Nyl Par.: Nylander, Herbarium Lichenum FE 1855. 32. Nyl Mont-Dor.: Nylander, Li- enes Mont-Dorienses, 1856. yl Pyr.: Nylander, Lichenes Pyrenaeorum Е 18 , 54. Om. Olivier, Herbier des Lichens de l'Orne et du Calvados, 1887-1884. 9, 116, 165, 208. Pieq.: Piequenard, Lichens du Finisterre. 74. Rab. Cen : Rabenhorst, Krypto peor то Centurien. ?, 16, 22. Ra aben- horst, ede Europae ри 1855-1879. 250, 471, 912. В. & S.: жаы. baeh et Schubert, Lichenes Exsieeati, ия 11, 35. Roum.: Varo ai Lichenes Galliei Exsiceati, 1879. 38, 40, 44, 186, 433, 558. Roum. Gen.: guére, Genera Lichenum Europaeorum Вдова, 1895. 27. Salw.: Salwey, Li- nes Centum 8 v chl. , Plantae Cryptogamicae Helvetiae и I 6. Spruce: Bid es Amazoniei et Andini, 1846. 127, 132, 136, 156, 176. Stenh.: Stenhammar, Liehenes Sueciae Exsiceati. 67, 71, 122, 123, a 154, 155, 156. Trev.: риб Lichenotheca Veneta, 1869. 156, 267. Tuck.: Tuckerman, Lichenes Americae Septentrionalis Exsiccati, 1854. 15, 17, 70, 72, 76, 78. Wain.: Wainio, Lichenes Brasilienses. 106, 400, 538b. Wartm.: Wartmann, Schenk & Winther, RM ische togamen, on 1882. 740, 741. Welw.: Welwitsch, Cryptotheca queen 18 - 77. estendorp, Herbier Cryptogamique Belge, 1845-1859. 1041. ee Zwack- ной Lichenes Exsiceati, 1850-1894. 252, 450, 569, 913. GENERAL INDEX Accepted scientific names are in Roman type; synonyms, in italics; new names, new combinations and principal references in bold face type. Page Page Borreri furfuracea ль... 136 Н de enteromorpha ...... 50 ollema exasperatum ............ 6 а О eee de eene 42 Corniculari E Mec qe eua 56 e Uh: T cen о 68 Evernia americana .......... 136, zs "t CRURA NOTER DEO CLTC 72 furfuracea var. Me para О conspersa f. isidiata ........... 76 kamtschadalis ............ 136, e [CALCUL ONDE ORE CNN о 127 [Vor. 28 144 ANNALS OF THE MISSOURI BOTANICAL GARDEN Page Page zc ЗАИР елен noce 43, 44 DRhUSDOSE" crus PELA TIERE зе 46 a ovo Dole Sae Б РИН Ivana FAX CH Rr ai NUR ЗОНА ases ecd oen Rn а О РОЯ 79 enteromorpha ................. 50 О ео edn ak 141 ОШОО. ..................... 67 ambigua var. Halei ............ 141 GOROG арка cx 1 98 9 Een 60 eta ian aga EEE TTT Ee 136, 137 MAUS OOER Lu vase E T 46 Amphigymnia ( Euparmelia sect. ) 112 ШТЕЙ зз укыса» E UCEA 91 E С еее 112 Id И RA SA ТГ СТ BT. ° ООД A haa sae bes ck Sareea 141 Lichen Gepidotus ............... 63 а idet RES Tara Pd ERI i 63 ВОГ ra css ky eor Ye 88 ды Vnde RS 141 WUE И 113, 114 о ЕШ „ууз з + куш» +» 110 cenirifuguas ................ 58 PROUD И око аме E баси A ООП8РЕТВИВ ................... 72 ОТТОЙ АСИ 79, 88, 90 diairypus eee 43, 44 Воттеті var. hypomela ......... SOLITI 90 Borreri var. ds ой .......... 96 enoaustus oo. cess ee 53, 54 camtschadalis f. americana 136, 137 furfuraoeus ............... 137, 138 зни; СОУ оа 135 МИМИ: а еа в ене течан 67 Lo Tic; hi QM c TE 55, 112, 113 intestinalis ................... 50 dd var. incorrupta ...... 118 laciniatus var. physodes ........ 45 eaperata var. subglauca ........ 9 ОАФ... еььеньььь 109 ЗАО opi ek сах а 106 т ЖЛ kk VA LS 56 БПО. е RUNE 54, 68 с nics ssn arre EY RR ER 59 Mini e cn var. phyllodes .... 46 omphalodes .................. 80 cebrarioldes .................. 125 perforatua ................... 130 наут var. rubescens ..... 127 ДОНОРИ 122 ПОТИ ЖЕЛЕ УТ еерее на 100 ПОТОЦИ |... роон о таа 3 48, 44 cetrata var. subisidiosa ........ 101 DIAJjSOGER |. oo kk cs vise Saban 45 ое BRI asa Fass 70, 71 ОШООО Leser ek Rs Fon Radia * 56 ПИ. apie eee T EE 24 QUOTOUAUS, .........„...ь›. 91, 92 СТОИ iacob baa dr 136, 137 тРНК 82 СПАЛО. ez err vaio 139, 140 RIS НИРО 92 БӨӨДӨ c uua Ima ca UAR RC RR 141 А PROP A PRA SR TE 57 о Cosa ror 72, 78 ИИЩОВИВ oiu raa aco Pea Inde 91 conspersa var. isidiata ........ 76 Lichenoides furfuraceum ..... 137, 138 conspersa. f. isidiata ........... 76 Lobaria furfuracea .......... 137, 138 BONSDETORIA eid duse ee ees 65 оса ee ee eee стажа 67 ОНОН M Sigs ее ач 52, 53 — Uer VECES E TRENT 59 GHI о. ро 127 eaa eiua ЛАС ТҮГҮ. 122 aristiforA. 2 SoG kos ch obo + „+ ANON E var, се ............ 124 спрата аа. 79, 108 ЗООР 46 РОА Р. аеро е RIP 90 ФаФаНИЯ ааа eee 82 ы рун И teas. 53, 54 Е И 57 CUAGIGUOR, ао Lea mand 67 foroebrdia! 1... 43 engoxantba an a з. э» 110 1941] BERRY—THE GENUS PARMELIA IN NORTH AMERICA 145 Page Euparmelia (subgen.) Hue ..... 55 Euparmelia (subgen.) Harm.... 42 Everniaeformes (Euparmelia BOCE atn ве 135, 136, 141 Everniiformes vera уе 135 Everniiformes (sect) ......... 135 OXAS PEL Abts, МИ 66 ДВ abi ЫЛ LOT 106 Са ео. 121 Franklumnana uo т ынк е ECCE PER furfuracea ..135, 136, 137, 138, |, 22037 AMAT 141 Шестата c m eom Ie. 121 теї оо C RUE 88 ПОТЕ В S ео: 85 Hubrichtii C НЕ 102 ymnia (subgen.) ........ Hypog a (su 45 вт орт sect.) 79 Hypotrachyna (вес 79 Imbriana (sect.) 45... ese 42 А ze oles oerte ses 118 ANCULTA gery А 67 lavigat: еее ecl 109 latissima. eee о а 134 Сара: 96 löucochlora m о. 78 lineola! е а: 77 МОА s 105 TOD AGAR (вес) 42 Lobaria sect. Imbricaria ....... 42 Ва ТЕ 52 шагкагпара EU UNE 141 porc DM (Euparmelia duc USC 5 о Савов.) ces 55 Menegazzia (subgen.) ......... 42 ichausxianag 2... cece were eens 124 О аи ats ce зок yi Оаа Ес I LONGER p acer sitis 136, 137 оуан 59 olivacea var. aspidota ......... 63 olivacea var. corticola f. conspur- DOLOS V а tra Tes olivacea var. glabra ........... 64 olivacea* corticola var. glabra.. 64 Page olivacea var. multispora ....... 64 olivacea var. polyspora ........ 64 olivacea var. prolixa ........... 58 olivaria f. cetrarioides ......... 125 RIOD ок о. perforsia hana do eost 113, 130 aes var. hypotropa ...... 132 S IO TIE. 112, 122 md VAT. апаа. 12 perlata var. flavicans .......... 121 perlata f. rubesoens ........... 127 pertusa o9 описа is ees 43 physodes ......... 45, 50, 52, 54 ев VAT. ПОП ......... 53 physodes var. enteromorpha 50 physodes var. labrosa ......... 49 physodes var. obseura ......... 49 physodes var. dog m гората 49 physodes var. vittata .......... 49 physodes var d oni SUME 46 ПЕЧЕНИ ак. 119 p оны 133 PEU co er DES 132 БА DE CERE PEN Coe eae ‚ 59 ei iri M (subgen.) ...135, 136 е ГАВ E sis e's 56, 58 CEA HS EL TET RE 91 ТЕСШ ое 99 ООСО. ЗАМ. 79, 96 а ао е ев 96 Бата ив: 42, 81, 82, 88 ТАНЕ VEPs Ашар. tees 85 saxatilis var. leucochroa ....... 82 saxatilis var. omphalodes ...... 80 sin де: ......... 109 о AER ens 120 па ре а. 71 WIE ОО ЛЕЧЕ а ЕО 141 р Айыз жээк ENT Tu 4% 55, 57 SUBDIONOG те 119 sublaevigaba ^. ESO 5: 107 ВоВе си ае 79 submarginalia scs -entea een. 124 зиройтаовй: „оно. 2 гас 60 8 p ERES DESEE M) MD Mm 86 SulDhurosni cris ы. >и oes 112 и э Мз 95 146 Page Bde iussis 91 tiliacea var. sublaevigata ....... 107 tiliacea var. sulphurosa ........ 112 р: о, 133 а 71 р Vari E арааа 135 Xanthoparmelia (Euparmelia (eaa Maa dI QM) (vs 66 [Vor. 28, 1941) ANNALS OF THE MISSOURI BOTANICAL GARDEN Page Xanthoparmelia (sect.) ........ а, О Eb vagis 112, 113 Vela. cere A EE Cd ES ES Platisma caperatum ......... 113, 114 DTT OTOP: Or tede PROUMOVETIAG. odes ets 135, 141 Уорс 138 Annals of the Missouri Botanical Garden Vol. 28 APRIL, 1941 No. 2 BINARY VARIATION IN TRADESCANTIA BRACTEATA EDGAR ANDERSON Geneticist to the Missouri Botanical Garden Professor in the Henry Shaw School of Botany of Washington University Although it is peculiarly variable, Tradescantia bracteata is ordinarily one of the most distinctive and easily recognized of the dozen or so species which are closely related to T. virgin- iana. In addition to various minor characters, it differs from all these species by its long rhizomes which permit it to grow into large mat-like clones, even in prairie sod. Plants of this species in an experimental garden may spread to a diameter of two or three feet, while in the same period the growth of other species is to be measured by inches, if at all. In spite of its several distinctive features, T'radescantia brac- teata is somewhat variable in a peculiar sort of way. If one confines his study to the selections from populations which find their way into herbaria, the variation seems no more extreme or extensive than in the other widespread species of Trades- cantia. If living populations (or mass collections) are ex- amined, however, it will be found that they vary around two quite different centers which are illustrated in figs. 1 and 2. In the prairie populations which we have studied the bulk of the individuals are more or less similar to Type ‘A’. They have unusually vigorous rhizomes and form large clones. The flowering stems are unbranched, and there are usually four ANN. Mo. Вот. GARD., Vor. 28, 1941 (147) [Vor. 28 148 ANNALS OF THE MISSOURI BOTANICAL GARDEN long internodes above the ground level. The leaves are narrow with a peeuliar yellowish cast, and the pubescence tends to be very short and scattered. The other center (Type В’) is made — WS AY, ZAY 1 таа I \ Y | i Fig. 1. Habit sketch of Tradescantia bracteata, type ‘A.’ up of plants which tend to be characterized by broad leaves, fewer nodes to the stem, vigorous secondary branches, and a lesser capacity to spread into large clones. If collections are made only by the tens and twenties these two centers are apt 1941] T ANDERSON—BINARY VARIATION IN T. BRACTEATA 149 . Habit sketch of Tradescantia bracteata, type ‘B.’ Figs. 1 and 2 d to the same scale by Ruth P. Ownbey, from plants collected at Portage des срама Мо. (VoL. 28 150 ANNALS OF THE MISSOURI BOTANICAL GARDEN to appear, but in every collection of 100 or more plants from the same colony they are certain to be evident. Tradescantia bracteata is therefore unique among the Amer- ican Tradescantias because it exhibits extreme intra-regional variation. There are other species of Tradescantia which show as much variation when collections are made from widely dif- ferent regions within their ranges, but we have found no other species which varies in this way within single populations. To analyze this variation, mass collections of T. bracteata were made at a number of points. The collection at Portage des Sioux was studied the most intensively since it was nearest the laboratory, but the general features discovered there are simi- lar at the other points where a large population was analyzed. Extremes of the two types illustrated in figs. 1 and 2 were studied carefully, and from them an index was constructed for classifying the entire population with reference to the approach of each individual to these two types. The method was origi- nally worked out for studying hybrid populations of T'rades- cantia (Anderson, '36) and has been found to be generally applicable to such cases. The particular characters used in this index and the values assigned to each are as follows: Maximum width of floral leaves: var Зо MENU EE bo sR REE ERA oes 0 Пади 19 MIS, оса... 4 Sepal color: Lg. РЕТ ETTETUIST Ло АК TERI 0 It med И О АИИС 1 о И ER PY ката 2 Stomata (under a hand lens): More conspicuous than surrounding cells ......... 0 Scarcely as conspicuous as surrounding cells ...... 1 No more conspicuous than surrounding cells ...... 2 Number of elongated internodes on the flowering stems: 1941] ANDERSON—BINARY VARIATION IN T. BRACTEATA 151 Branching of the flowering stem: ОРИ Ges еко Tq rl 2 Wit Sere Огайснев n.. рое beta лт eee 1 Branches bearing flowers ....................... 0 This index was used in classifying four populations of Т. brac- teata, with broadly similar results in each case (fig. 3), but with significant special features. At each of these localities the spe- cies was found to be morphologically duplex. That is to say, that it fluctuated around two different centers, Type ‘A’ and Type В’. That these two types are due to inherent differences in the germplasm and not to environmental or age differences is shown by the fact that transplants have consistently main- tained their original type in the experimental garden. Plants moved from various midwestern points to Boston and from the Dakotas to St. Louis have held to their original type over a period of years. Although connected in each population by a manifold series of intermediates, the extremes when sorted out by means of the index were most surprising. Individually or as a group, the extreme plants of Type ‘B’ are morphologically very sim- ilar to Tradescantia hirsutiflora of the Gulf Coast, a species which is today completely unknown within the range of T. brac- teata. Tradescantia hirsutiflora (or at least one element in it) is itself so similar to T. virginiana that it might almost be con- sidered a geographically localized variety of that species. While the ranges of T. virginiana and T. bracteata overlap slightly, extreme plants of Type *B' resemble T. hirsutiflora more closely than they do Т. virginiana. In any explanation of the binary variation of T. bracteata, there are several critical pieces of evidence. 1. While the differences between the two types are manifold, there is enough relationship physiologically between them to suggest that there might be basically a single difference in rhizome vigor, to which all the other differences are secondary. Active rhizomes would produce large clones with many shoots, among which there would be more root competition beneath [Vor. 28 152 ANNALS OF THE MISSOURI BOTANICAL GARDEN O-l 2-3 4-5 6-7 8-9 10-И 12 Fig. 3. Frequency diagrams for Index No. 1 of four populations of T. bracteata. Upper pi from — gb in sod: dotted line, = plants from unplowed swamp ра es at Por ge des Sioux, Mo.; solid two populations ra roadways: dot ка e 60 plants along roadway at edge of field, Portage des Sioux, Mo.; solid line, 60 plants from railroad right-of-way "ала A Vietor, to, 1941] ANDERSON—BINARY VARIATION IN T. BRACTEATA 153 the soil and more shading above the soil. It is therefore con- ceivable that the narrower, less-branched shoots and the nar- row leaves and smaller flowers might be merely a secondary eonsequence of more active rhizomes. 2. Though the differences between the two types might re- sult from a single physiological difference, there is little in the variation to suggest that it is the work of a few genes. The differences are too manifold and the variation of the intermedi- ates resembles the varied recombinations of second-generation species crosses rather than simple Mendelian segregation. 3. The differences are apparently not due to age, since trans- plants of both types have maintained their original character when brought into experimental gardens. One of the clones which was originally selected because of its remarkable capac- ity to spread repeated this behavior when transplanted from Towa to Massachusetts. 4. The peculiar characteristics of Type ‘A’ are unique in this group of species. From this evidence we conclude that Type ‘A’ and Type ‘B’ owe their differences to segregating elements within the germ- plasm, to differences of the order of whole chromosomes or of chromosome arms. As to how this variation was introduced into the species we have no evidence. There are at least two very different ways in which it might have occurred. On one view, Tradescantia bracteata would originally have been only of Type “В”; on the other hypothesis, the original element in the species would have been Type ‘A’. It may be that the bi- nary condition is as old as the species. On that view, the pe- culiarities of Type ‘A’ appeared within the original primitive stock of Type *B' (by whatever process or group of processes specifie differences are achieved in this genus) with the added peculiarity that the original stock tended to be carried along in the population. Quite another hypothesis is suggested by the resemblance of the variation of T. bracteata to that which we have studied in species hybrids of T'radescantia. On this view, Type ‘A’ was originally differentiated as a unique spe- cies and Type ‘B’ then originated by hybridization between Type ‘A’ and T. hirsutiflora at a time when the ranges of one 154 [Vor. 28 ANNALS OF THE MISSOURI BOTANICAL GARDEN or both of these species were different enough from their pres- ent distributions to bring them into contact. If this happened comparatively recently we might expect to find Type *B' lim- ited to the southern or southeastern extremity of the range of 105 100 95 90 os во 75 ¢ ‘Cee опека ма $ | EX e gr. F q № & [ Em. AN Oe з 7 VI LCS. ^ ч. \ \ aot? e$ | 7 I H - 5 2” и а q i „7 i —— a е ; tmu y- \ | | | ( SE Ка. i | кака D | № — — | “в ад --@ KI | / | | 957: о" | t PI ал Е ERA в | $. „Р Ad | РЗ ШИШ | Ter I H ^d p і | \ Н { j^ | E , > | d В ам E | SKA КК d',9 9 RKAS эмеес. XT v * е е 7 9 3 000 25.000 | — US 299 390. 409 200 | SQUARE м$ MILES / rmx) — — inni син ри rs е M"ILOMETAES z“ N ш ALOCA'S COUAL-AACA PROJECTION 100 95 90 es LONGITUDE wes Map of middle-western United States showing the known distribution of X. bracteata (black dots) in relation to that of T. hirsutiflora (diagonal lines). Each dot rep collections were ma sents an herbarium specimen; open circles, points at which mass ade. 1941] ANDERSON— BINARY VARIATION IN T. BRACTEATA 155 304 24- 18- WASTA 12- 6- 18 ] WORTHINGTON 125 64 12- ROGKERVILLE о ' LI Li BE. ncn ad 2] HARRIS'S GROVE 64 о M т , ' Li U 0-1 2-3 4-5 6-7 8-9 10 Fig. 5. Frequency roy ч ried Index No. 2 of four populations of T. bracteata. Populations arranged ae jording to аве per- centage of open soil in the habit [Vor. 28 156 ANNALS OF THE MISSOURI BOTANICAL GARDEN T. bracteata. If, however, the hybridization is a matter of very long standing, then Type ‘B’ might be expected to occur throughout the range of T. bracteata. For determining this point collections from the Black Hills would be particularly signifieant since they are at the opposite end of the range of T. bracteata. Mass collections were accordingly made at the points shown in fig. 4. Frequency distributions for these col- lections are shown in fig. 5. The index used in deriving the fre- quencies for fig. 3 could not be used since it was based in part on characters which can be scored only in living material. The following index was therefore constructed for dealing with the pressed material: Maximum width of floral leaves: From mid-vein to margin, over 8 mm. ............ 0 From mid-vein to margin, 6-7 mm. .............. 1 From mid-vein to margin, 5 mm. or less .......... 2 Length of hairs om pedicel: i24 5.9. 0 ое. 0 ПОТИ RAT DEM. (Losses EXERCERE Г 1 БӨРЕ ОВ: ана два кил 2 Number of elongated internodes on flowering stems: “АР А ОМА Е ic he ane wane ОТО РИУ Bossa vida ELO Lio dd cs RES E oes 1 A OE a. TETUR HEC ЛЕККЕ ГЕТ АЛТ 2 Branching of the flowering stem: IDONEAM ПИ РИ gag 2 Wi меко БОВЕ... 1 Branches bearing ПОГ... скань нь 0 It will be seen that this index differs from that previously used in the following ways: (1) In measuring the maximum width of the floral leaves (braets) one-half the width rather than the whole width had to be used because the specimens were pressed. This, as well as the changes in drying, increases the percentage of error. (2) The color of the sepals and the character of the stomata could not be ascertained in pressed material. In their place the length of the pubescence of the 1941] ANDERSON—BINARY VARIATION IN T. ВВАСТЕАТА 157 pedicel was substituted, since it is prevailingly thick and long in Type ‘B’ and short and scattered in Type ‘A.’ It was meas- ured to a tenth of a millimeter, using a camera lucida. The measurement was accurate, but a study of plants in the experi- mental garden has shown that this character is affected by temperature and humidity. Index No. 2 therefore, while it is more or less similar to Index No. 1, is not as accurate, and since itis based on fewer characters does not separate the two centers as well in the frequency distributions. However, it gives a roughly similar result as will be seen from fig. 6, which shows i LI "T e 10 e е «2 а | ее 8 6 4| е ji ва 2 e ө v LJ т Ц Li т || 2 4 6 8 10 Fig. 6. Correlation between index values of 20 plants from Portage des i as measured by Index No. 1, vertieal seale, and Index No. 2, hori- zontal seale. the correlation between the two indices for 20 plants which were measured by each index. Population frequencies for Index No. 2 are graphed in fig. 5. It will be seen that Type ‘B’ not only occurs in the opposite corner of the range of T. bracteata from that which comes closest to the territory of T. hirsutiflora but that it even makes up a higher percentage of the population than it does in the south. This is perhaps correlated with open-soil habitats and is discussed below on page 159. The curves are also less bi- modal than in the Missouri and Iowa populations. This is in part due to the inferiority of the index but probably also re- [Vor. 28 158 ANNALS OF THE MISSOURI BOTANICAL GARDEN flects a real difference in the northwestern populations. How- ever, although plants of Type ‘A’ were not present in large enough numbers to produce a bimodal curve, at least a few were encountered at every locality as well as various inter- mediates. Type ‘B’ therefore most probably originated be- fore T. bracteata moved into the territory it now occupies. The entire group of species to which it belongs is centered upon the Edwards Plateau in east-central Texas. It is quite possible that at one time T'radescantia bracteata grew within this same general area and could then have differentiated from, or hybridized with, T. hirsutiflora. This latter species still occurs there and is the most aggressive species in that area and the one most apt to hybridize with other species. To summarize: Associated with typical Tradescantia bracteata throughout its range and connected with it by a manifold series of intermediates is a Tradescantia morpho- logically very similar to T. hirsutiflora, a species of the Gulf Coastal Plain. It is suggested that this variant originated by partial differentiation from, or hybridization with, T. hirsuti- flora at a time when the two species were in closer contact than now, and presumably before T. bracteata moved into the territory it now occupies. Since the chief effect of the phe- nomenon is to produce a species which fluctuates around two centers, instead of one, we are calling it binary variation. There now remains to discuss: (1) Its selective advantages. (2) The cytology of T. bracteata. (3) Its probable occurrence in other groups of organisms. (1) The selective advantages of binary variation.—As will be seen from the map in fig. 4, T'radescantia bracteata is a spe- cies of the prairies and northern great plains. This region is notorious for a climate which fluctuates widely from decade to decade as well as from year to year, and within that area Tradescantia bracteata is most commonly found in marginal habitats which are peculiarly susceptible to climatic fluctua- tions. In the prairie states it grows at the edges of swales or in slight depressions which are just damp enough to discourage 1941] ANDERSON—BINARY VARIATION IN T. BRACTEATA 159 some of the prairie grasses. A single dry year has a marked effect upon the vegetation. In a fluctuating climate, in such a habitat, a species which varies around two centers would be at a great selective advantage, particularly if the permanence of the two-centered condition was assured by some cytogenetic mechanism. In this connection the frequency curves of fig. 3 are rather suggestive since they demonstrate that with a slight change in habitat there are changes in the proportions of Type ‘B’. Two populations were studied in Missouri and two in Iowa. In each case one of the populations was from a more or less natural prairie and the other was from a man-made habi- tat. At Portage des Sioux, Missouri, in rich bottomland near the junction of the Mississippi and Missouri rivers, is a low grassy pasture which is apparently a remnant of one of the river-bottom savannahs once common to the region. One col- lection was made in the pasture and another along a cart track which ran beside it. Of the Iowa collections, one was from the edges of a grassy swale in an unplowed pasture and the other from along a railroad track adjacent to rich bottomland, all of which was under cultivation. It will be seen that the curves of the two prairie habitats are very similar, as are also those of the two trackways, but that the proportion of Type ‘B’ is much higher in the open-soil habitats, so much so that the make- up of the species differed much more markedly between neigh- boring habitats than it did between Missouri and Iowa, for the same habitat. This impression is confirmed by the frequencies for Index No. 2, diagrammed in fig. 5, where they are arranged from above to below according to the prevalence of grass at each loeality. At Wasta, North Dakota, there was practieally no grass, and many of the plants of T. bracteata were growing in soll as open as if they had been cultivated in a garden. At Har- ris's Grove, the other extreme, the plants were growing in dense prairie sod and were so overtopped by the grass that they were hard to find. The frequency diagrams show that the grassier the habitat, the greater the percentage of Type ‘A’ and inter- mediates resembling it. This is not surprising since in at least two ways Туре ‘A’ is evidently better adapted to such an en- [Vor. 28 160 ANNALS OF THE MISSOURI BOTANICAL GARDEN vironment. In the first place, its longer, more numerous inter- nodes allow it to compete for sunlight with the grasses in situ- ations where plants of Type ‘B’ would be completely sub- merged. In the second place, its extremely active rhizomes would allow it to compete with the notoriously vigorous rhi- zomes of the prairie grasses. (2) The cytology of T. bracteata.—T radescantia bracteata is prevailingly diploid, with 6 pairs of chromosomes. "Table lists the chromosome determinations which have been made as well as those previously listed by Anderson and Sax (736) and summarizes the total frequencies. The species is overwhelm- ingly diploid, although polyploid individuals have been col- lected in nature at two points. It is therefore unlike all the other widespread species closely related to Т. virginiana, since they are either known only as tetraploids or have diploid races confined to a relatively small area. There are other diploid species in this group but, without exception, they are of very limited distribution and several of them are on the verge of extinction. In our opinion diploidy has persisted in T. brac- teata because it is actually at a selective advantage in prairie habitats by reason of its effect upon the flowering season. Like many other prairie plants, T'radescantia bracteata dies down rapidly after it has flowered and spends the summer in a dor- mant or semi-dormant condition. In an experimental garden TABLE I CHROMOSOME NUMBERS e a ER v T. BRACTEATA COLLECTED THE WIL Loealities 2n 3n 4n Loealities 2n 3n 4n Houlton, Wisconsin 1 St. Louis, Missouri 1 Preston, Minnesota 1 | Huron, South D woe 1 R Co., Minnesota} 1 Kennebec, 1 rthington, Minn 1 Murdo, А 1 Grinnell, Iowa 3 2 Rockerville, S. Dak.) 1 Kellogg, Iowa 1 Wind Cave, S. Рак. 1 она Iowa 1 Wasta Я 2 s 1 Overton, Nebraska 1 ена 3 Royal, Nebraska 1 Victor, € 1 Lawrence, Kansas 2 Portage des Sioux, Manhattan, Kansas 3 issouri 8 1941] ANDERSON— BINARY VARIATION IN T. BRACTEATA 161 it withers so quickly that one who did not know the species would assume that the plants were badly diseased. Tetraploidy would be disadvantageous to such a species since it favors longer flowering seasons. In nature, in the experimental gar- den, and in the greenhouse, tetraploid Tradescantias differ from their related diploids, among other things, by the greater length of their blooming periods. Polyploid strains, such as the plants reported in Table r, would be at a selective disadvantage since they would come into flower more slowly and carry their flowering into the unfavorable drought and heat of the sum- mer months. The two tetraploid plants reported above were discovered under precisely those circumstances. When orig- inally collected they were the only plants still in flower at that locality, all the neighboring plants having withered and died down. Aside from the gross information concerning chromosome number, little is known with regard to the cytology of T. brac- teata. Sax (737) and Darlington (737) have reported inver- sions in this species, and Swanson (240), from these facts and his own evidence, has concluded that the occurrence of inver- sions in T'radescantia is very widespread. The fact that both Type ‘A’ and Type ‘B’ have been found in every population which has been examined leads us to suspect that the binary condition is being maintained by some fairly precise cytogenetie mechanism. Otherwise it is difficult to see how both varieties could be present in every population. There must at least be some mechanism by which plants which are phenotypically like one variety can yield offspring resembling the other when they are intercrossed. One would need to postu- late no more intricate mechanism than those already demon- strated for cereal rogues or the complexes of Oenothera. Although the morphological differences between the two species seem to be manifold, it is possible, as we have pointed out above, that only a small proportion of the germplasm is responsible for the change. Given the basic difference in rhizome growth, many of the other differences might auto- matically follow. A large matted clone with many flowering stems might be expected to have narrower leaves and less sec- [VoL. 28 162 ANNALS OF THE MISSOURI BOTANICAL GARDEN ondary branches than an isolated plant such as Type ‘B,’ with only a few flowering stems. If only one or two chromosome segments are responsible for the bulk of the differences be- tween the two varieties it would be comparatively easy for the binary condition to be maintained. (3) Probable occurrence of binary variation in other groups of organisms.—Binary variation, in our opinion, is probably fairly widespread in both the plant and animal kingdoms. Its frequency can scarcely be discussed until more species have been examined by populations. Up to the present not more than a handful of species has been studied in this way. Certain of the Drosophilae studied by Dobzhansky, Spencer, Patterson, and their students are perhaps to be classified in this category though data with regard to populations are still too meagre to permit a definite opinion. The relationship between the black and white races in the southeastern United States is very similar and differs only in the fact that there are numerous communities which fluctuate around only one center and that the chief isolating factor (social prejudice) has caused nearly all the intermediates to be classified as black instead of being recognized as intermediates. Among the higher plants binary variation probably occurs fairly frequently within a part of the range of widespread species, but cases such as 7’. bracteata where a species is binary throughout its range are probably rare. In eastern North America many, if not most, of the deciduous trees which are relatively uniform in the North and East include other ele- ments within their populations in the Ozarks and the South- west. Acer saccharum, for instance, forms relatively uniform populations to the northeast, while in the Ozarks usually it includes the variety Schneckii and is united with it in that region by a series of intermediates. SUMMARY 1. Associated with typical T'radescantia bracteata through- out its range and connected with it by a manifold series of intermediates is a T'radescantia morphologically similar to T. hirsutiflora. 1941] ANDERSON—BINARY VARIATION IN T. BRACTEATA 163 2. It is suggested that this variant originated by partial differentiation from, or hybridization with, the Т. hirsutiflora- virginiana stock at a time when the ranges of these Trades- cantias were confluent. 3. The phenomenon is named ‘‘binary variation." Its selec- tive advantages and probable occurrence in other species are diseussed. 4. The cytology of Tradescantia bracteata is discussed. It is suggested that the exceptional maintenance of widespread diploidy in this species is due to the fact that in a prairie and great-plains environment the shorter blooming season of the diploids may be at a selective advantage. BIBLIOGRAPHY Anderson, Edgar (1936). Hybridization in American Tradescantias. I. A method for measuring species hybrids. II. qe gen between T. virginiana and T. canaliculata. An . Bot. Gard. 23: 511-525 and Karl Sax (1936). A TEE кн of the American species of Pi on Bot. Gaz. 97: 433—4 Ба C. D. (1937). Chromosome behaviour WR structural hybridity in the трет II. Jour. Genet. 35: 259—280. Sax, и (1937). Chromosome behavior and nuelear development in Tradescantia. Genetics 22: 523-533. Swanson, Carl В. (1940). The distribution of inversions in Tradescantia. Genetics ON THE FLORA OF THE FRONTIER FORMATION OF SOUTHWESTERN WYOMING! HENRY N. ANDREWS Instructor, Henry Shaw School of Botany of Washington University AND CORTLAND 8. PEARSALL Geologist to the Tunnel Division, Metropolitan Water Commission, Boston, Mass. INTRODUCTION In 1917 Dr. F. H. Knowlton described a small flora from the Upper Cretaceous Frontier formation of southwestern Wyo- ming. Attracted by that author's figures of Microtaenia pauci- folia (Hall) Knowlton, we made a small collection of the fructifications in the summer of 1934. Nothing of exceptional note was obtained at that time and our collections did not re- ceive immediate study. In the summer of 1939, when the senior author revisited the locality, certain undescribed fructi- fieations were discovered, but as time did not permit extensive exploration we returned the following summer and prospected the outcrop carefully for about 4 miles north and 1% miles south of Little Muddy Creek. We were fortunate in obtaining nearly complete fertile pinnae referable to Anemia Fremonti Knowlton and in finding certain other fossil plants or parts thereof not previously described. The plant-bearing horizon was originally discovered by Capt. John С. Fremont in 1843 while exploring for a better emigrant route to the Northwest. Fremont's collections were turned over to James Hall, New York State Paleontologist, who published descriptions in 1845. Other collections were made at later dates by members of the U. S. Geological Survey and were presented in Knowlton's paper of 1917. Of the 25 described species composing the flora there are 7 ferns, 1 Equisetum, and 17 Angiosperms, the latter including Quercus, Ficus, Salix, Aralia, Dewalquea and Cinnamomum. ТА study financed in part by a grant from the Penrose Fund of the American Philosophical Society of Philadelphia. Issued April 28, 1941. (165) [Vor. 28 166 ANNALS OF THE MISSOURI BOTANICAL GARDEN Certain of these genera are undoubtedly valid ; others are ques- tionable as Knowlton himself admitted. The present paper will not be concerned with the Angiosperm element of the flora. GEOLOGY The section of the Frontier formation discussed in this paper is situated in the southwestern corner of Wyoming, about 15 miles south of the town of Kemmerer, in the vicinity of Cum- berland Gap. The formation, which consists of a series of light-colored coal-bearing sandstones and shales approxi- mately 2500 feet thick, extends for many miles both north and south of Cumberland Gap and is characterized by a series of prominent hogbacks. It is underlain by the dark-colored oil- bearing Aspen shales, and overlain by the dark-colored Hill- iard shales. The most prominent hogback is formed by the Oyster Ridge member, which is an oyster (Ostrea soleniscus )- bearing sandstone about 200 feet thick near the top of the formation. Numerous coal seams are contained in the lower half of the formation. The plant material was collected from a series of white to light blue-gray shales about 50 feet thick. These are located stratigraphically about 75 feet above the generally accepted base of the formation. Small sandstone beds directly above the plant-bearing shales contain a great deal of silicified wood. The plant shales break poorly along the bedding plane and are in the nature of argillite which in thin section is a typical clastic, fine-grained sediment consisting of approximately 70 per cent clay minerals, 28 per cent angular quartz, 1 per cent feldspar, 1 per cent zircon, hornblende, biotite and carbonate (Veatch, ’07). METHODS The cellulose transfer technique proved very satisfactory with the compressions of the fertile pinnae of Anemia Fre- monti (see p. 168). Fragmentary specimens were selected in which the fructifications appeared to be well preserved. A SO, =~ 1941] ANDREWS & PEARSALL—FRONTIER FORMATION FLORA 167 rather heavy coat of nitrocellulose solution (Darrah’s for- mula) was then applied to the surface of the matrix and ex- posed plant material. The entire specimen was then placed in hydrofluoric acid. It was found advisable not to cut away the apparently excess rock in back of the compression because in some cases this contained additional fragments of the fructi- fications which dissolve away from the rock and serve as ex- cellent material for maceration. The disintegration of the rock can be accelerated by remov- ing the block from the acid each day and scraping away the partially dissolved matrix. Specimens as thick as 114 inches were treated in this way, and the time for complete dissolution TABLE I GEOLOGIC COLUMN SHOWING THE RELATION OF THE FRONTIER FORMATION TO OTHER CRETACEOUS FORMATIONS M EATCH, '07, TABLE FACING p. 501) System and Group Formation Eocene Unconformity Lower Laramie Adaville formation, with basal Lazeart sandstone Montana Hilliard formation Niobrara В Frontier formation with the Oyster © Ridge sandstone member 2 Benton 4 А = о Е = Aspen formation о Bear River, . > Dakota?, and Bear River formation Lower Cretaceous? Beckwith formation Jurassic Twin Creek formation [Vor. 28 168 ANNALS OF THE MISSOURI BOTANICAL GARDEN of the rock was in no case longer than 10 days. The last frag- ments of matrix may be removed with a camel's-hair brush under water. This transparent celloidin technique had, with our material, two very important advantages: first, the back side of the fossil may be studied as well as the front when it is mounted; and second, much better photographs could be ob- tained than when the rather grayish rock matrix constituted the background. If figs. 2 or 3 (photographs of specimens in the matrix) be compared with figs. 1 and 5 (photographs of specimens transferred to celloidin) the difference is quite ap- parent. Maceration of the isolated fragments of the fructifications was done in the usual way with strong nitric acid and potassium chlorate. SCHIZAEACEAE Anemia Fremonti Knowlton forma fertilis Andrews, forma nov. It is not often that well-preserved spore-bearing parts can be referred with certainty to any of the numerous sterile fern species that go to compose the fossil record. Associated with foliage which Knowlton (’17) described as A. Fremonti were found fructifications of an undoubted schizaeaceous nature and, as will be made clear, in all probability referable to the genus Anemia. The significance of the association of fertile and sterile parts will be considered after the former have been described. The most complete specimen that we have found is shown in fig. 3. It is pinnately compound, the branching taking place in one plane. Figures 1, 2, 4, 5 are isolated primary branches (as will be shown below, these are secondary pinnae, accord- ing to usual fern terminology) photographed at a higher mag- nification. The secondary branches, which are quite short, bear two or three flattened globose bodies, two being borne laterally and a third terminating the branch. In figs. 2 and 5 only one of these bodies may be seen attached to each secondary branch; in fig. 4 two may be seen on some branches while others have only one, both the terminal and a lateral having been lost. 1941] ANDREWS & PEARSALL—FRONTIER FORMATION FLORA 169 Well-preserved fragments of the primary branches isolated from the matrix show three in all cases. Within one minute after such fragments had been treated with strong nitrie acid and potassium chlorate each of the small black bodies opened up into about seven ‘‘valves.’’ If maceration was not then stopped by diluting the acid with water, the fragments very soon almost completely disinte- grated. Due to the extreme fragility of the material at this point it has not been possible to prepare permanent prepara- tions nor to obtain satisfaetory photographs. Figure 36 shows a secondary branch with its three **bodies"" opened out. The segments referred to as valves are in fact pinnules and bear sporangia on their inner surface. In only one or two cases has it been possible to detect the segmented (pinnular) nature of the globose bodies prior to maceration. 'The separation of the individual pinnules upon maceration is, however, always uniform and can in no wise be attributed to a maceration artifact. In only two instances have all of the sporangia been observed in place on a pinnule, one of which is shown in fig. 38. The sporangia are arranged in two rows, forming the typical monangial sori of the Schizaeaceae. No evidence of an indusium has been observed. Upon further maceration the sporangia go almost completely to pieces. The maceration residue, however, reveals two structures of con- siderable interest, spores and more or less complete annuli. The spores (figs. 6-12) are beautifully preserved, the entire exine being sculptured with characteristic ridges. These are well brought out in the photographs of opposite sides of the same spore (figs. 7 and 8). The triradiate commissure is well defined. Of 50 particularly well-preserved spores that were measured, the average diameter was found to be 40 и. There is, however, considerable range in size of the spores (from 25 to 47 u in diameter), as may be noted from the figures where all are shown at the same magnification. The annuli are also abundant in the residue, the rest of the sporangia disintegrating almost completely. Due to their opaque nature the annuli (fig. 33) are not the best of photo- [Vor. 28 170 ANNALS OF THE MISSOURI BOTANICAL GARDEN graphie subjects and are always split open, apparently along the stomium. It may be seen then that the structures referred to above as primary and secondary branches are, according to usual fern terminology, secondary and tertiary pinnae, and the sporangia are borne on the under-surface of the ultimate divisions or pinnules. The structure of the pinnules, arrangement of the sporangia, nature of the annulus, and characteristic seulptur- ing of the spores leave no doubt as to the schizaeaceous affini- ties of the fossil, and of the four living genera of the family, it is clearly more closely related to Anemia than any of the other three. The specimen shown in fig. 3 is comparable in every respect with one of the two basal fertile pinnae char- acteristic of the species of that genus. The authors are well aware of the inadvisability in, general of placing great weight on the association of isolated parts of fossil plants. In this case, however, the above-described fructi- fications were found associated with, and only with, the foliage described by Knowlton as Anemia Fremonti (figs. 35, 39). The two were found together at three different points between И and }% mile south of Little Muddy Creek. A comparison of the sterile frond (fig. 39) with certain living species such as Anemia adiantifolia Swartz in itself leaves little doubt as to the validity of Knowlton’s generic identification. This similarity, combined with our discovery of fertile pinnae whose characteristics clearly point to the same genus and which are constantly associated with the foliage, stands as a credit to Knowlton’s original identification and seems to render superfluous a new specific name for the fertile speci- mens. The evidence at hand strongly supports the supposition that these sterile and fertile parts are one and the same species, but in view of the lack of organic connection it is advisable to em- ploy some sort of nomenclatorial segregation. To assign a new specific name to the fertile parts tends to defeat rather than enhance the ideals of such an investigation, and, moreover, implies a distinction that the available evidence does not sup- port. In accordance with the remaining element of doubt we 1941] ANDREWS & PEARSALL—FRONTIER FORMATION FLORA 171 propose to designate the fertile specimens as a distinet form, namely forma fertilis. The problem of interpreting the associ- ation of scattered parts of fossil plants is one that the paleo- botanist frequently encounters, and it is probably true that most workers are inclined to place too much weight upon such associations. We believe that the procedure employed here does not exaggerate the truth yet does reflect the implications that the evidence affords. It has been assumed in the restoration (fig. 40) that the specimen shown in fig. 3 is a nearly complete fertile pinna and that the pinnae were borne in pairs as in the modern species. It is quite certain that fig. 3 represents a fairly young pinna since fragments of comparable size when macerated still con- tained large quantities of spores; furthermore other pinnae have been found, a single secondary one being shown in fig. 4, in which the whole has elongated considerably. The largest fragment of foliage that we have found is shown in fig. 39. This and numerous other specimens served as the basis of the restoration of the sterile portion of the frond. It may be that the lower sterile pinnae were more deeply dis- sected than is actually represented. Fossil history of the Schizaeaceae.— Our knowledge of the earliest members of the Schizaeaceae has been eonsiderably augmented by the recent researches of Radforth (’38, 39). He has shown that Dactylotheca Sturi Sterzel and D. plumosa Artis are actually annulate and con- sequently referable to S'enftenbergia. There has, in the past, been some doubt as to the affinities of that genus but Radforth's investigations leave no doubt as to its proper inclusion in the Schizaeaceae. The discovery that Senftenbergia (Dactylotheca) Sturi has annulate sporangia extends the history of the family back into Lower Carboniferous (Carboniferous Limestone Series) times. The apical annulus consists of a single row of cells in the modern genera Anemia, Schizaea and Lygonium. How- ever, Radforth has shown that Senftenbergia pinnaeformis, an Upper Carboniferous form, had an annulus of two rows of [Vor. 28 172 ANNALS OF THE MISSOURI BOTANICAL GARDEN cells, while S. Sturi, from the Lower Carboniferous, has a less regular annulus 4 to 5 cells deep. Of particular interest to the present discussion is his conclusion ‘‘that a close phylo- genetic relationship exists between these fossil Schizaeaceae and the living Schizaeaceae, and of the latter, particularly the genus Anemia." In Jurassic rocks the genus Klukia is an undoubted repre- sentative of the family and seems to have been rather wide- spread, specimens having been reported from Yorkshire, Po- land, Caueasia and Korea (?). From the Lower Cretaceous of Virginia, Berry (11) has described a fertile frond under the name of Schizaeopsis ex- pansa (Font.) Berry. This fossil is rather closely comparable to living species such as Schizaea elegans, judging from the general morphology of the frond and the sculpturing of the spores. Stopes and Fujii (210) described a schizaeaceous fern, Schizaeopteris mesozoica, from the Cretaceous of Japan which seems closest to Anemia. Much just eritieism has been aimed at the determination of fossil plants based on sterile material but confirmation of Knowlton's identification of Anemia Fremonti indicates that even with sterile foliage all is not guesswork. In view of this eonfirmation in the ease of the Frontier formation species we have checked through Knowlton's (719) list of supposed species of Anemia from America. Judging from the published figures there is a reasonable degree of certainty that the fol- lowing are correctly referred to that genus: Anemia elongata (Newberry) Knowlton— Laramie formation, uppermost Cretaceous: Erie, Colo. (?), Point of Rocks, Wyo. (7). (Knowlton, Е. H., U. В. Geol. Surv. Prof. Paper 130: pl. 2, fig. 2. 1922). Aneimia eocenica Berry— Lagrange formation, basal Eocene: Puryear, Tenn. (Berry, E. W., U. S. Geol. Surv. Prof. Paper 91: pl. 10, fig. 2, pl. 11, figs. 1, 2. 1916). 1941] ANDREWS & PEARSALL—FRONTIER FORMATION FLORA 173 Anemia hesperia Knowlton— Fruitland formation, Upper Cretaceous: San Juan Co., N. M. (Knowlton, F. H., U. S. Geol. Surv. Prof. Paper 98: pl. 84, fig. 3. 1916). Anemia occidentalis Knowlton— Raton formation, Paleocene: Trinidad, Colo., Yankee, N. M. (Knowlton, F. H., U. S. Geol. Surv. Prof. Paper 101: pl. 54, fig. 2. 1917). Anemia supercretacea Hollick— Vermejo formation, Upper Cretaceous: Rockvale, Colo. (Knowlton, F. H., U. S. Geol. Surv. Prof. Paper 101: pl. 30, fig. 5. 1917). It is thus clear that the family Schizaeaceae, and particu- larly the genus Anemia, once enjoyed a much more northerly distribution than at present. Anemia is now confined to the American tropics and subtropics (with the exception of one species from South Africa), extending northward only into the southern part of the United States. Anemia sp. Fragments of fronds have been found at a number of points along the outcrop, which are closely comparable with certain living species of Anemia. Although similar in certain respects to A. Fremonti they are much less robust and have not been found associated with the fertile pinnae of the latter. The most complete specimens that we have discovered are shown in figs. 30-32, 34. Those shown in the first two figures probably represent nearly complete fronds which are char- acterized by a rather long slender stalk, twice pinnate, with a tendency toward a tertiary division in the basal pinnae (fig. 31). It is quite possible that these fronds are simply small specimens of A. Fremonti but since they have not been found closely associated with that species it seems best to describe them separately. There is, moreover, a striking similarity to the Wealden fern, Ruffordia Goepperti, which Seward ('94) likewise includes in the Schizaeaceae. [Vor. 28 174 ANNALS OF THE MISSOURI BOTANICAL GARDEN GLEICHENIACEAE Gleichenites coloradensis (Knowlton) Andrews, n. comb. Dryopteris coloradensis Knowlton, U. S. Geol. Surv. Prof. Paper 108-F : 83, pl. 30, figs. 3, 4. 1917. This apparent fern was included by Knowlton in the Poly- podiaceae, a justifiable conclusion considering the fragments of foliage that he had available for study. Although our speci- mens do not bear reproductive structures, a considerable amount of information has been obtained concerning the struc- ture of the frond as a whole. As may be judged from the fol- lowing deseription, it is strikingly similar vegetatively to cer- tain species of Gleichenia, the resemblance being sufficiently great to warrant its transference from Dryopteris to the genus Gleichenites of Goeppert. Collections were made at three localities along the plant- bearing outcrop, at points approximately 11%, 134 and 2 miles north of Little Muddy Creek. Most of our material came from the last two localities where it occurred as a ‘роге stand,” there being no other associated fossil plants. Knowlton did recognize that it was a plant of considerable size. He writes, ‘‘From the presence of large pieces of stems intermingled with the fronds and presumed to belong to them, it is assumed that this fern was probably of large size, but the direct evidence is only suffieient to say that it is at least bi- pinnatifid." In all probability the ‘‘large pieces of stems"' that Knowlton mentions are fragments of the rachis. The most distinctive feature of the frond lies in its successive trichoto- mous-like branchings. The rachis is of considerable size, as may be noted in figs. 20 and 26, a number of fragments similar to that shown in fig. 26 having been found. The central member of the ‘‘trichotomy’’ may grow out (fig. 20) or remain abortive (figs. 21, 24), but it does seem to have been more generally de- veloped in the fossil than in living species such as Gleichenia pectinata. In the reconstruction of the frond in fig. 21 it has been as- sumed that c corresponds to the primary subdivisions of the specimens shown in figs. 20 and 26. The specimen shows three 1941] ANDREWS & PEARSALL—FRONTIER FORMATION FLORA 175 further divisions, the last of which lies on the counterpart of the specimen and is shown at b, fig. 22 (the negative has been reversed so as to show it in the same orientation as fig. 21). The pinnules are arranged on the ultimate (fourth order) sub- divisions as shown in figs. 24, 27. Figure 29 has been included to show more clearly the typical pinnule morphology. The removal of this species from the genus Dryopteris seems fully justifiable in view of our present knowledge of the struc- ture of the frond. Had Knowlton had more complete specimens he would undoubtedly have placed it in the Gleicheniaceae. Be- cause of the fact that sporangia still remain to be discovered it seems best to refer the plant to Gleichenites. Of the various species of Gleichenites that have been de- seribed (Hirmer, '27, pp. 623-4), G. coloradensis compares most closely with G. Gieseckiana Heer. The characteristic mode of branching, as well as the morphology of the pinnules, is very similar to specimens of the latter described by Seward (226) from the Cretaceous of Greenland. It was apparently a widespread species during Cretaceous times, specimens having been reported from Spitsbergen, England, France, Germany, Russia, Sakhalin (Japan), Dakota and New Jersey (Seward, '26, p. 147). In view of this cireumpolar distribution of G. Gieseckiana, it is not surprising to find the very similar, if not identical, G. coloradensis in the Upper Cretaceous of south- western Wyoming. Where reconstructions of fossil plants or parts thereof can be prepared with a reasonable degree of accuracy it seems de- sirable that the plant be presented in that fashion. The figured specimens, as well as numerous others from which information has been drawn, bear out the accuracy of our restoration of the frond as we believe it appeared in life (fig. 41). Whether or not the entire frond is represented is, of course, not certain. As our material indicates but four orders of branching, only that many have been shown, and since no rachis fragments larger than those shown in figs. 20 and 26 have been found it is likely that the entire frond is represented. Some of the terminal branches have been eliminated in the restoration in [Vor. 28 176 ANNALS OF THE MISSOURI BOTANICAL GARDEN order that the structure of the terminal pinnae might be brought out more clearly. The distinetive morphology of the fronds, as well as the sori and sporangia of Gleichenia, has enabled its scattered fossil remains to be assembled into one of the most interesting stories of geographical distribution presented by an extant genus. Confined now to the tropics and sub-tropics, Gleichenia once enjoyed a range far north of its present confines. It has been described from Jurassic and Cretaceous rocks of Green- land, and from the Cretaceous period numerous localities in North America, to mention a few: Maryland, New Jersey, Kansas, Colorado, Wyoming, California, British Columbia. In 1935 Seward wrote, ‘‘The sight of well preserved fronds of Ferns with forked arms exposed on a slab of shale on the beach of Upernivik Island (71° N. lat.) took me back to a hill- side above Penang in the Malay Peninsula, where living Gleichenias in company with Dipteris formed a wonderful tangled carpet of luxuriant growth on the edge of a tropical forest.’’ No more striking contrast to either the arctic or tropi- eal climates could be found than the present semi-desert sage- brush hills of southwestern Wyoming whose climate must have been much more like that of Penang in Upper Cretaceous times than it is at present. CYATHEACEAE-DICKSONIACEAE Microtaenia Knowlton In his account of the Frontier flora Knowlton described two species under this genus, M. variabilis and M. paucifolia (Hall). In both cases apparently fertile fronds were described and figured but no actual evidence of their fertility was given. As we have succeeded in isolating spores from both species a further account of these interesting ferns is presented here. Microtaenia paucifolia (Hall) Knowlton, U. S. Geol. Surv. Prof. Paper 108-F : 82-83, pl. 30, figs. 1, 2. 1917. Fertile fragments of this species were found to be fairly abundant, particularly within a radius of № mile to the north and south of Little Muddy Creek. The greatest part of the 1941] ANDREWS & PEARSALL—FRONTIER FORMATION FLORA 177 fragments apparently had matured previous to fossilization since of the numerous ones collected only two have yielded spores. The largest specimen (fig. 16) was deposited before spore dispersal as spores have been found in all of the sori studied. When the large globose bodies terminating the pin- nules were removed from the rock and macerated in nitric acid and potassium chlorate for a few minutes, a residue of spores resulted (fig. 17). The triradiate commissure is clearly shown. The exine is rather delicate and smooth, with no evidence of sculpturing of any sort. All the spores are distinctly triangular as shown in the figure, and although this may be due in part to collapse it may represent the actual shape of the spore in life. There is very little variation in spore size, all being about 26 и in diameter. No evidences of sporangia have been found in the globose terminal supposed sori. There remains the possibility that the ‘‘sori’’ are actually single terminal sporangia. There is a striking resemblance of these structures to the Jurassic genus Coniopteris, particularly of C. hymenophylloides (see Thomas, ’11, pl. 3, figs. 1, 4, 5; Seward, ’00, pl. 17, fig. 8). How- ever, no foliage has been found in the Frontier formation which compares at all closely with that of Coniopteris. Because of the close similarity between Microtaenia paucifolia and the fertile pinnae of Coniopteris hymenophylloides, which is gen- erally accepted as being referable to the Dicksoniaceae, it seems advisable to include Microtaenia within that family in- stead of the Polypodiaceae following Knowlton’s classifica- tion. Microtaenia variabilis Knowlton Only two fertile fragments of this plant are included in our collection. Figure 13 shows one of these magnified nearly 5 times and fig. 14 is a portion magnified 24 times. It differs from M. paucifolia in the broader, more leaf-like nature of the pinnae. In fig. 14 a single vein may be seen passing out to each sorus. Sporangia have not been observed but a few spores have been obtained. Their walls are delicate and wrinkled (fig. 15) (VoL. 28 178 ANNALS OF THE MISSOURI BOTANICAL GARDEN and there is no evidence of sculpturing or a triradiate commis- sure. EQUISETACEAE Equisetum sp. In his report on the Frontier plants Knowlton described and figured a species of Equisetum based on a supposed under- ground stem. In our collection of the past summer we have a specimen referable to his species, but it is far from being a convincing representative of the genus. Other specimens have been found, however, of a much less doubtful nature. A nodal diaphragm is shown in fig. 18. This was associated with rather poorly preserved stem fragments which, although a specific name would be meaningless, does attest to the presence of the genus in the flora. PLANTAE INCERTAE SEDIS Baiera sp. (?) The problematieal specimens deseribed under this name were found at only one locality, about 115 miles north of Little Muddy Creek. 'The most complete specimen that was collected is shown in fig. 28. It may be noted that this dichotomizes five times and, as in the other specimens illustrated (figs. 19, 23, 25), the branching is equal or nearly so in all cases. In the right-hand primary fork of the specimen shown in fig. 28 a single vein may be discerned in each subdivision. The vein divides some distance before reaching the dichotomy of the lamina (this term is employed on the assumption that the fossil does represent a leaf of the Baiera type). Unfortu- nately, the fossils are preserved in a coarse sandy shale, there being no cuticular remains. There is a close superficial similarity between our specimens and certain species of Валета, particularly B. spectabilis (Walton, '40, fig. 124a) and a specimen tentatively referred by Seward to B. Lindleyana (Seward, '26, pl. 10, fig. 101). There are, however, other possibilities that merit consideration. In 1895 Seward deseribed, under the name of Becklesia anomala, a fossil plant from the Wealden of England, consist- 1941] ANDREWS & PEARSALL—FRONTIER FORMATION FLORA 179 ing of a central axis with pinnately arranged branches which in some cases ‘‘appear to bifureate close to the point of attach- ment to the central axis." He compared his specimen with the peculiar forked leaflets of the living Macrozamia heteromera Moore. A study of herbarium specimens of the latter revealed a rather striking similarity although the Frontier fossils are somewhat larger and appear to have been more laxly disposed. SUMMARY The preceding is an account of certain fossil plants from the Upper Cretaceous Frontier formation of Wyoming. Well- preserved compressions of fertile pinnae of a schizaeaceous fern are described and shown to be referable to Anemia Fre- monti Knowlton. Since the sterile and fertile parts of the frond have not been found in actual organic connection the latter are described as forma fertilis. Fronds of the supposed polypodiaceous fern Dryopteris coloradensis Knowlton are shown to possess vegetative char- acters distinctive of Gleichenia and are redescribed as Gleichenites coloradensis. Spores have been obtained from the fertile fronds of Micro- taenia variabilis Knowlton and M. paucifolia (Hall) Knowl- ton. The presence of Equisetum in the flora is recorded and fossils referable to Baiera are described. ACKNOWLEDGMENT For the carefully prepared drawings composing plates 6 and 7 we are especially grateful to Mr. Albert A. Heinze. LITERATURE CITED Berry, E. W. (1911). A Lower Cretaceous species of Schizaeaceae from eastern North America. Ann. Bot. 25: 193-198. Hirmer, M. (1927). Handbuch der Palaobotanik. München. Knowlton, F. H. (1917). A fossil flora from the Frontier formation of southwestern yoming. U. S. Geol. Surv. Prof. Paper 108-F: 73—94. , (1919). A eatalogue of the Mesozoic and Cenozoie plants of North America. U. S. Geol. Surv. Bull. 696. [Vor. 28, 1941] 180 ANNALS OF THE MISSOURI BOTANICAL GARDEN Radforth, N. W. (1938). An analysis and comparison of the structural features of Dactylotheca plumosa Artis sp. and Senftenbergia ophiodermatica Goeppert Sp. Trans. Ro oy. Soe. Edin a 9). Further ай to our knowledge of the fossil Schizaeaceae; genus Senftenbergia. lbid. 59: 745-761. Seward, A. C. 1894). А Wealden Flora. Cat. Mes. Pl. Geol. Dept. Brit. Mus. І. — — — , (1895). d. II. , (1900). m Jurassie Flora. I. The Yorkshire Coast. Ibid. III. ‚ (1926). The Cretaceous A bearing roeks of western Greenland Phil. Trans. а Кос. Lond. B215: 35). Selections from the story of plant migration revealed by fossils. Sei. pipi 30: 193-217. Stopes, M. C. and К. Fujii (1910). Studies on the structure and affinities of Cestaccons plants. Phil. Trans. Roy. Soc. Lond. B201: 1-9 Thomas, H. H. (1911). On the spores of some Jurassic ferns. Proc. Cambridge Phil. Soe. 16: 384—388 Veateh, A. C. (1907). Geography and geology of a portion of southwestern Wyo U. 8. Geol. Surv. Prof. Paper 56: 1- Walton, fuhr (1940). An introduction to the study of fossil plants. London EXPLANATION OF PLATE PLATE 1 Anemia Fremonti Knowlton forma fertilis Andrews Fig. 1. Two primary branches of fertile pinna. Photograph from а nitro- cellulose transfer. No. 1312, x Fi 1 — of fertile pinna, Only one pinnule eluster is shown o each Vitus branch, the гд Be iet been removed with the on гэ т prior to na Menus; or are imbedded in the matrix. > id 3. А nearly complete њой fertile pinna. No. 1313, x 3. Primary branch of a somewhat older fertile pinna, as evidenced by its iin size and more expanded пат of the pinnule elusters. Two of the latter may be clearly distinguished on most of the secondary branches. No. 826, x 3.2. Fig. 5 rimary ще of a ins pinna. Photograph from a nitrocellulose алой, No. 1311 PLATE 1 ANN. Mo. Bor. Garp., Vor. 28, 1941 & PEARSALL—FLORA OF FRONTIER FORMATION ANDREWS [Vor. 28, 1941] 182 ANNALS OF THE MISSOURI BOTANICAL GARDEN EXPLANATION OF PLATE PLATE 2 Figs. 6-12. Spores of Anemia Fremont forma fertilis, Figs. 7 and 8 are of 50. Figs. 13-15. Microtaenia variabilis Knowlton. Fig. 13. Portion of fertile rond. No. 1308, x 4.8. Fig. 14. Portion вари same more highly magnified, x 24. Fig. 15. ае и Fig. 16. ditionis paucifolia Knowlton. Portion of fertile frond showing the large, terminal, globose sori (or sporangia?). No. 1310, x 1.7. Fig. 17. Spore of same, x 750. Fig. 18. Equisetum sp. nodal diaphragm, x 2.3. ANN. Mo. Bor. Garb., Vor. 28, 1941 PLATE 2 = e REN ANDREWS & PEARSALL—FLORA OF FRONTIER FORMATION [Vor. 28, 1941] 184 ANNALS OF THE MISSOURI BOTANICAL GARDEN EXPLANATION OF PLATE PLATE 3 19,23. Baiera sp. Fig. 19. No. 1302, x 1.3. Fig. 23. No. 1301, x 1. . 20, 21, 22, 24, Gleichemites coloradensis (Knowlton) Айген, Fig. 20. Basal gine of frond showing primary branching. No. Fig. 21. Portion of frond showing secondary and tertiary branch. No. 1293. Fig. 22. Photograph of the counterpart of the specimen shown in fig. 21. The negative was reversed in order to show the two in the same orientation. No. 1294. Fig. 22a corresponds to fig. 21a, while the fourth order of branching is shown at fig. 22b. Fig. 24. Final (fourth) branching showing pinnule morphology. No. 1298. All approx. x .8. ANN. Mo. Bor. Gard., Vor. 28, 1941 PLATE 3 йл 90 о ANDREWS & PEARSALL—FLORA OF FRONTIER FORMATION EXPLANATION OF PLATE PLATE 4 Figs. 25, 28. жиз» Fig. 25. No. 1304, x 1. n 28. No. 1303, x 1.4. Gleichenites coloradensis. Figs. 26, 27, 29 Fig. 26 branching of frond. No. 1292, x 1. Fig. p T eme No. 1291, x 1. Fig. 29. Por- tion of a terminal branch showing pinnule morphology. No. 1297, x .9. ANN. Мо. Вот. Garp., Vor. 28, 1941 PLATE 4 ANDREWS & PEARSALL—FLORA OF FRONTIER FORMATION [Vor. 28, 1941] 188 ANNALS OF THE MISSOURI BOTANICAL GARDEN EXPLANATION OF PLATE PLATE 5 Figs. 30, 31, 32, 34. Anemia sp. Portions of sterile та Uc en in text. Fig. 30. No. 1300, x 1. Fig. 31. No. 1299, x 1. Fig. 32. 1307, x ig. 34 No. 1295, x .9. Fig. 33. Anemia Fremonti forma fertilis. Annulus, Fig. 35. Anemia Fremonti. Terminal portion of m -A 1305, x 1. ANN. Mo. Bor. Garp., Vor. 28, 1941 PLATE 5 ANDREWS & PEARSALL—FLORA OF FRONTIER FORMATION [Vor. 28, 1941] 190 ANNALS OF THE MISSOURI BOTANICAL GARDEN EXPLANATION OF PLATE PLATE 6 Figs. 36-38. Anemia Fremonti forma fertilis. Fig. 36. Portion of primary branch of a fertile pinna showing a secondary branch after treatment with macerat- ing fluid, x 15. Fig. 37. A single pinnule cluster in side view showing the sporangia arranged in two rows along the under-surface of the pinnules, x 30. Fig. 38. Under- surface of a single pinnule, x 30. Fig. 39. Anemia Fremonti. Part of sterile portion of frond. No. 1306, x 1. Fig. 40. Restoration of eomplete frond. PLATE 6 ANN. Mo. Bor. Garp., Vor. 28, 1941 FLORA OF FRONTIER FORMATION & PEARSALL ANDREWS [Vor. 28, 1941] 192 ANNALS OF THE MISSOURI BOTANICAL GARDEN EXPLANATION OF PLATE PLATE 7 Fig. 41. Restoration of frond of Gleichenites coloradensis. NOILVWUYOd WA3LLNO?H4 ЧО VAOTA—TIVSAVAd № SMAAANV af ae Ea L $ 327.3 р 34 О jw" № 3r Ge МА А. Yal 4 ATANAN, ARE у Pi à У 9 T Р ND a ALS PAS ANA P Ne ay, : A cw, ` Y KANER A 2 20 аус a ST © Vs ES рМ P2 ar BR) ae ae RN CS сър es 4 а MY? VR HSS я УА zb Is Иа ety sis S Я PAE s o N нар I 2 Буу R eB 2 Я ZR | А AQ VY FERR Reh eek sk Tm E IS FR ZRSR S) B. BR ZR SIRIS (D 6 5) M 57 S SR $ P v ДАТ) BER ; DASS. ar, M: А РАЙС а) ВАА - ссггор е ло? BS а SBS AS г мее SN ASA e y я 143 Д — (<< : у сч, 42 < t - = Ar EQUO SS 7 ЭТУУ nf se ea Av N Эу А, T c дар ‘> ~. ON ADN СИЕ » RAS TIS a A И ^. УД Mis й ff 2 0 р 4) № (5 C^ р. eue IX fa Fur а d ewe Дъ "a ke 73 PS AN AD * ү, ж К затоа SOP E? => WR XS A SE dom Nem = TON E EY Shae EET Nw OTE S O NE o ЗАД PLATE 7 " o riv p г И V ~ AAA pet Sa оса " [X e nw S Б (6 ps ‚ы de J = pS T Я Ф T R че, 5205 THE NORTH AMERICAN ASCLEPIADACEAE I. PERSPECTIVE OF THE GENERA ROBERT E. WOODSON, JR. Assistant Curator of the Herbarium, Missouri Botanical Garden Assistant Professor in the Henry Shaw School of Botany of Washington University INTRODUCTION The problem of the North American genera of Asclepiada- ceae has been summarized well by Standley (Fl. Costa Rica 3: 949. 1938): Тһе family is noteworthy for the complicated strueture of the flowers, more complicated, probably, than those of any other family of plants. The Mexican and Central American members of the group have never been monographed properly, and their classification is at present in a decidedly chaotic state." It is unfortunate that Charles Darwin did not compose a study on the floral structure and pollination of Asclepiads as a companion to his famous observations on Orchids. In both groups, as distantly related as Dieotyledons and Monocoty- ledons may be, entomophily has produced elaborate and vari- able innovations from group to group. But apparently with the same intention: for the attraction of insects and cross- fecundation by them, through means of a pollinium. The Orchids have responded to entomophily largely by zygomorphy and elaboration of the perianth. In the Ascle- piads, on the other hand, aetinomorphy has been retained, and, although corolline modifications are obvious, diversification is due chiefly to the struetural plasticity of the staminal fila- ments. These elaborations of the stamens are known as the corona. Since the anthers of Asclepiads are invariably five in number, it obtains that the corona also is five-parted, the seg- ments being quite free or mutually coherent, and diversified in ways that are too numerous for description. Although a staminal corona is the rule for the group, a few entities possess Issued April 28, 1941. (193) [Vor. 28 194 ANNALS OF THE MISSOURI BOTANICAL GARDEN simple filaments. In any event, the center of the flower is oc- cupied by a composite structure known as the gynostegium, comparable to the column of Orchids. This consists of the five coherent stamens surrounding the style of the bicarpellate pistil, with an enlarged terminal ‘‘stigma head’’ at the free end. The whole stigma head is not truly stigmatic, the recep- tive surfaces being five narrow strips of glandular tissue be- tween the contiguous anthers. The anthers of American Asclepiads are bilocular, and within each cavity the pollen grains are grouped at maturity into a glutinous, sac-shaped pollinium of definite form and size. In the subfamily Cynanchoideae, containing all the New World species, these pollinia are combined by pairs through a more or less elaborate yoke mechanism called the translator. The translators consist of two arms* attached to their respective pollinia, and are themselves joined by a roughly sagittate body called the ‘‘gland.’’* A pair of pollinia, therefore, consists of the contents of adjacent anther cavities of contiguous anthers. The translators, with their glands, are formed between the neighboring anthers. The current explanation of the formation of the translators is that their substance consists of the solidified secretion of special glandular cells located upon the stigma head. Investi- gations of my own, not yet completed, lead me to assume that in the early development of the pollinia the tapetal liquid, char- acteristically abundant in young anthers, seeps through small pores of the neighboring anthers to gather in the commissural grooves of the two organs. This liquid later solidifies outside the anther as well as about the pollen, instead of being absorbed as is usually the case. Superficial support is given this view by the fact that the uniting ‘‘gland’’ always is distinctly 2-parted. According to either view, the translators are moulded when still in the liquid state by the available spaces between the young anthers. *<¢¢Translator агт’? and ‘‘gland’’ frequently are called ‘‘retinaculum’’ and ‘í corpusculum?’ respectively in the literature. The former are chosen here as being somewhat more easily associated with the appearance of the structures as well as more adaptable to English: the latter are more useful in Latin diagnoses. 1941] Е WOODSON—NORTH AMERICAN ASCLEPIADACEAE. I 195 When the pairs of pollinia are mature they are easily re- moved from the dehiscent anthers by means of the yoke-like translators. A visiting insect may be observed to pull them out and carry them away suspended from its legs. On the same or another flower, the pollinium must be inserted between the anthers at the stigmatic surface, in order to insure pollination. At that time, in response to the stigmatic secretions, pollen tubes germinate from the pollinium while still intact, to travel down the style to the ovules. The significance of the pollinia and their method of fecundation were fully appreciated for the first time by Robert Brown (Mem. Wern. бос. 1: 12-58. 1809) in his celebrated paper distinguishing the Asclepiads from the Apocynads. The method of pollination in Asclepiads which Robert Brown disclosed appears at first sight to be of design so cun- ning that great fecundity of the plants should be assured, as well as frequent hybridization. Every conceivable innovation would seem to have been made by the flowers to insure success- ful insect ministrations. But such does not necessarily seem to have resulted. The family as a whole is outstanding for the constancy of its species and the rarity of obvious hybridity. Furthermore, the amount of fruit produced scarcely seems to be commensurate with the ‘‘efforts’’ taken to insure it, to speak anthropocentrically. Accounts of the pollination of Asclepiads seldom mention the actual difficulties intervening: the danger of the pollinia being broken or only partially withdrawn by the insect’s casual visits; the horny margins of the anthers which seclude the stigma, frequently with entangling hooked or grooved decora- tions. Germinated pollinia are conspicuous, even in dried and boiled material, because of the abundant, felty pollen tubes. Amongst the hundreds of flowers of various genera and species which I have dissected under magnification, I have found very few indeed bearing germinated pollinia in the stigmatic cham- ber between the anthers. Of those few, the companion pol- linium was usually in place in its mother anther sac. In nu- [Vor. 28 196 ANNALS OF THE MISSOURI BOTANICAL GARDEN merous instances I have found germinated pollinia while still within the anther sac. Whether such pollen tubes can reach the ovules successfully has not been ascertained, but it is not impossible. Such observations explain to some extent the rarity of hybridization, but fail to shed much light on the use of the pollinia as agents for pollination by insects. Another feature of the Asclepiad flower which would appear to have little positive survival value is the division of the stigma into five rather restricted receptive regions. This ap- parently is responsible for the fact that only one follicle com- monly develops from the two carpels of the pollinated pistil. Thus, even though three stigmatic surfaces were to receive pol- linia, fecundation of only one carpel might result: but it must be confessed that conversely were only two pollinia deposited development of both follicles might ensue. A pair of follicles, nevertheless, is seldom encountered. Perhaps it is safe to as- sume that the reduction in number of fruits is compensated by - the volatility of the comose seeds. The classification of Asclepiadaceae into major divisions is based upon the nature and position of the pollinia, and thus owes its foundations to the classical observations of Robert Brown (Mem. Wern. Soc. 1: 12-58. 1809). That versatile genius separated the family as it is now recognized into five tribes: Periploceae, with open cornucopia-shaped pollinia and adhesive translators somewhat as in the Orchids ; Secamoneae, with 4-locular anthers, the yoke-like translators bearing two pollinia upon each arm; Asclepiadeae Verae, with 2- locular anthers, the yoke-like translators bearing a single pendulous pollinium upon each arm; Gonolobeae, with pol- linia similar to those of Asclepiadeae Verae, but horizontal; and Stapelieae, with similar, but erect pollinia. Recent sys- tems (cf. К. Schumann, in Engl. & Prantl, Nat. Pflanzenfam. 4?: 209. 1895) have divided the family into two subfamilies, Periplocoideae and Cynanchoideae; the latter having four tribes, Asclepiadeae, Secamoneae, Tylophoreae (Stapelieae of Brown), and Gonolobeae. In the New World only the Cynan- choideae are encountered in the native flora, with the three tribes Asclepiadeae, Gonolobeae, and Tylophoreae. 1941] WOODSON—NORTH AMERICAN ASCLEPIADACEAE. I 197 In these studies no attempt will be made to subdivide the tribes into the smaller subdivisions advocated by Schumann, for these are manifestly unnatural. Thus the genus Sarco- stemma R.Br. (sensu stricto) was placed in Ше subtribe Cy- nanchinae of Asclepiadeae by that author (K. Sch. in Engl. & Prantl, Nat. Pflanzenfam. 4?: 245. 1895), but the two certainly synonymous genera, Philibertia HBK. and Funastrum Fourn., in the Glossonematinae and the Asclepiadinae respectively (pp. 225, 231. loc. cit.). I am inclined also to ignore Schu- mann’s subtribe Astephaninae, characterized by the absence of a corona, since in all cases which I have observed the absence of a crown is the only character separating species groups of obviously close affinity, notably in the inclusive treatment of Cynanchum adopted here. One monotypic genus, Ту годопна Griseb. (treated as a division of Cynanchum here), Schumann includes in the Astephaninae, but I have observed the presence of definite corona segments in an isotype deposited in the herbarium of the Missouri Botanical Garden (Wright 2964). Neither can Oxypetalinae of Schumann be maintained by the appendages of the translator arms, since several South Ameri- can species of the single genus, Oxypetalum, do not have those structures. The separation of the American Asclepiads into tribes some- times is a bit difficult, since the position of the pollinia usually must be observed while they are still within the anther зас. Once withdrawn and placed upon the dissecting stage, the deli- cate translator arms are apt to twist into any conceivable atti- tude, frequently giving a false impression of their natural posi- tion. The Asclepiadeae and Tylophoreae usually are quite easy to detect; but the Gonolobeae have given trouble from the start. I am willing to be convinced that the normal position of the Gonoloboid pollinium may be truly horizontal, although my interpretation of the tribe includes forms with pollinia that range in position from pendulous to ascending. By far the best means that I have found to separate this tribe is by means of the structure, not the position, of the pollinia. In all the Asclepiadeae and Tylophoreae examined by me, the faces of the pollinium are uniformly rounded or flattened on [Vor. 28 198 ANNALS OF THE MISSOURI BOTANICAL GARDEN either side, and the translator arm makes a sharply definite connection with them. In the Gonolobeae, on the other hand, the faces of the pollinium are more or less dissimilar: one con- vex or rounded and the other flattened or somewhat excavated, frequently very strikingly so. In some cases both sides are deeply furrowed or excavated. In the Gonolobeae, also, the translator arm usually engages the pollinium more gradually, the attachment being marked by a more or less conspieuous hyaline indentation or margin (always the upper). Without the use of this structural criterion, Schumann placed such genera as Macroscepis and Fischeria in the Asclepiadeae and Metalepis in the Gonolobeae, whereas other characters of the plants show their natural positions to be exactly the reverse, as shall be explained presently. At this writing ninety-seven genera of Asclepiadaceae have been described from North America, based wholly or in large part upon the structure of the corona. Authors of the great majority of these genera seem not to have understood what a versatile feature this organ is, for it is only when its multi- farious aspects are interpreted in a broad and comparative manner that a natural and conveniently referable classifica- tion can result. North American students of tropical Ascle- piads, particularly, would do well to consider the floral vari- ability of the native Milkweeds before attempting the descrip- tion of exotic genera. The prospective Asclepiadologist should understand from the first that differences of the corona sepa- rate species in this family, which would form remarkable generic criteria, say of the **dise'' in other families. After the first shock of surprise at the few genera recognized for North America, readers of the following key to genera may notice the fairly incidental use of corona characters. These are reserved chiefly for subgeneric and sectional distinctions, and will be discussed in that connection. The term ‘‘corona’’ has been taken in a rather restricted sense in these studies. Amongst earlier literature frequent mention is made of ‘‘double’’ or even of ‘‘triple’’ coronas. This terminology is very confusing except to the authors concerned. Morphologically, the corona, as interpreted here, 1941] WOODSON—NORTH AMERICAN ASCLEPIADACEAE. I 199 consists of various elaborations or enations of the staminal filaments only. These enations may occur as separate bodies attached to the staminal column, as in Asclepias, to cite a famil- iar example, or may consist of a more or less entire ring of tis- sue adnate to the bases of both staminal column and corolla throat. Where ‘‘double’’ or ‘‘triple’’ coronas have been de- scribed by various authors, these usually will be found to con- sist of a faucal annulus of the corolla tube, or sterile append- ages of the anthers, which are rather to be called such than to be referred to as corona. The most cogent reason for avoiding the use of ‘‘corona,’’ except in a simple sense, is that many readers of the literature are apt to interpret a corona as double, whilst the author may not intend such a view. Several instances of possible or pub- lished confusion arising from an author's use of ‘‘double eorona" come to mind. In the subgenus Chthamalia of Matelea, for example, the corona consists of an annular struc- ture of five more or less united and variously constituted seg- ments, each of which bears internally a more or less conspicu- ous ligular process, sometimes surpassing the height of the subtending segment. Contrary to what one might suppose, this complex structure is not considered by the literature as a ‘double eorona." The flowers of Gonolobus (sensu stricto) are very difficult to identify with standard keys, for they usually possess (1) a faucal annulus of the corolla, (2) a fleshy, true corona, (3) dorsally appendaged anthers. Some authors interpret these structures as constituting a ‘‘triple”’ and some as a ‘‘double’’ corona. Those maintaining the dual nature may construe either the corolline annulus or the anther appendages as the supernumerary corona, in which cases the true corona will be spoken of as either the ‘‘inner’’ ог the ‘‘outer’’ corona, respectively. The segregate genus Podostigma, of the southeastern United States, is keyed by K. Schumann (in Engl. & Prantl, Nat. Pflanzenfam. 4?: 235. 1895) on the basis of having an **inner corona"! borne high upon the column, alternate with the anthers. When these tiny objects are examined, however, they are found to be nothing but the minute ligules commonly alter- [Vor. 28 200 ANNALS OF THE MISSOURI BOTANICAL GARDEN nating with the corona hoods in Asclepias proper, their normal position doubtless having been disturbed by the elongation of the anther stipe* above the insertion of the hoods. In Fischeria, also a **double согопа” is described in all texts. But when carefully dissected, the bladder-like ‘‘inner corona segments’’ are found to be nothing but the hypertrophied anthers themselves. This peculiar mark of Fischeria may be described more accurately and more clearly upon the basis of anther structure than by introducing an ambiguous and morphologically indefensible reference to supposed ‘‘corona.’’ The North American Asclepiadaceae have never had the attentions of a specialist devoted primarily to their study, ex- cept for the period between 1897 and 1904 during which a series of eight short papers by Miss Anna Murray Vail appeared in the ‘Bulletin of the Torrey Botanical Club.’ These papers con- sist only of short notes, with the exception of rather casual re- visions of Acerates, the Asclepias verticillata complex, and Rouliniella (i.e. Cynanchum subgenus Mellichampia, in part, of my studies). Even from these small and interrupted begin- nings, it is obvious that Miss Vail possessed a very consider- able knowledge and insight into the North American Ascle- piads, and it is regrettable that her retirement to France terminated her ambitions for more comprehensive works. I must confess that more than once, when I was apparently undergoing swift mental attrition because of the damnably variable coronas and pollinia, I have reflected on my prede- cessor's retreat to the Riviera. The South Ameriean and Antillean Asclepiads have had sev- eral specialists, including Decaisne, Fournier, Schlechter, and Malme; to these must be added the contemporary South Ameri- ean botanists, A. G. Schulz and F. C. Hoehne. The work of these eminent systematists, however, has nearly always been in the field of regional floristies, with few exceptions, such as Malme's monograph and subsequent emendations on Osy- * [ have found it eonvenient in dealing with some genera to distinguish two regions of the staminal column: that below the attachment of the corona as the **eolumn,?" as in most species of Asclepias proper, and that above, between the corona and the anthers proper, as the ‘‘stipe,’’ as in Podostigma. 1941] \ WOODSON—NORTH AMERICAN ASCLEPIADACEAE. I 201 petalum. But these contributions have affected the northern Asclepiads only in rather minor particulars. Regional floras of North America have included thoughtful interpretations of the Asclepiadaceae in several instances, but have been hin- dered by their restricted scope and the divided interest of their authors. Special mention should be made here to the monograph of Marsdenia by Rothe (in Engl. Bot. Jahrb. 52: 354-434. 1915), not only since it is the only comprehensive account of an Ascle- piad genus in both American continents, but because the author’s solution of the problems of this group closely parallel my own for the family. Confronted by an aggregation of eight segregate genera, nearly all in current use but with very eva- sive criteria when seen as a whole, Rothe courageously com- bined them as sections. The result, in spite of some imperfec- tions, is the one bright spot in the determination of tropical American Asclepiads from the existing literature. In the Tylophoreae, at least, one needs no longer to puzzle over the generic interpretation of floral innovations, nor fear that the specimen he is trying to determine may already be described as anew genus in some remote section of the taxonomic indices. At the same time, a natural system of species groups is de- fined and provided with names which may (or should) be used in the description of novelties, or to which additions may even be made should sufficiently diserepitant plants be found. A lifetime would be too short to perfect a complete system of the North American Asclepiads. But after an intensive study of several years, I have arrived at a definition of the genera which I believe may well lead to a more adequate under- standing of the complex as a whole. The result is a revolution- ary change in nomenclature which surely will not be regretted more by the reader than by the author. The change clearly is unavoidable because of the long neglect of the family. In this group of plants apparent entomophily has produced a floral structure even more complicated and variable than in the Orchids. Morphological details separate species and even varieties here, which would serve to distinguish genera in other groups of Flowering Plants. The student of the Asclepiads is [VoL. 28 202 ANNALS OF THE MISSOURI BOTANICAL GARDEN impelled to the defeatist attitude that only two choices are available in classification: to ‘‘lump’’ genera or to ‘‘split’’ them. If he continues the study of Milkweeds sufficiently long, he probably will find himself tossed from horn to horn of the dilemma. After several such harrowing experiences, I have recognized that a narrow concept of coronal structure as a criterion of genera, if pursued consistently, will lead to a multitude of monotypic entities—as a matter of fact, that few genera will consist of more than two or three species—and that each new species will introduce anew the question of generic differenti- ation. This is scarcely an exaggeration when one remembers that an inclusive concept of North American Asclepias, alone, involves the listing of twenty generic synonyms. If he contemplates an extended association with the Asclepiadaceae, with yearly increments of specimens brought to his attention for classifying, even the most hardened ‘‘liberal’’ taxonomist will take heed before embarking upon such a disastrous course. Although the results of my studies wreak havoc upon the existing classification, I feel confident that the groups pro- posed are natural, and that they will stand the test of many years. The desideratum of stability and convenience will be served much better in having a few, almost infallibly recogniz- able genera even though they may be large: the natural affini- ties within these groups may be indicated sufficiently by the erection of subgenera and sections. This paper is intended as the forerunner of a series devoted to the North American Asclepiadaceae. It is composed as a general survey of the generic elements involved, together with an indication of representative changes in nomenclature made necessary by my visualization of the natural system of the family. These species are drawn only from those with which I am familiar at this time, and should not be considered as complete synopses; neither do they include species not re- quiring adjustment in terminology. The synonymy, also, is not complete, consisting merely of the name-bringing and other more prominent synonyms. To some readers, the inclusion of the many nomenclatural changes upon such a scanty prepara- 1941] WOODSON—NORTH AMERICAN ASCLEPIADACEAE. I 203 tion may seem precipitate. The decision to include them here, rather than to withhold them for subsequent, complete revi- sions, has been made in view of the advantages of concrete il- lustration of the generic elements proposed, the uncertainty of our ability to obtain European types necessary for a complete study within the near future, and the validation of obviously necessary names for the purpose of present determinative work. Should the future permit, this paper will be followed by a monograph of the North American species of Asclepias, and by subsequent revisions of the other important genera, in each ease with all recognized species illustrated by analytical drawings. To be of wide use, a critical study of the Asclepiadaceae should be illustrated profusely. I feel that an apology is in order, therefore, because of the total lack of analytical draw- ings in this paper. Their absence is due in part to the condi- tions of publication at the moment, and in part to plans which I have made for their inclusion in future studies. After all, this paper is scarcely more than an annotated key to genera, and of little interest save to specialists. Such readers will be suffieiently familiar with the structures to which I call atten- tion not to miss the figures. KEY то THE TRIBES AND GENERA Pollinia strictly pendulous, their faces uniformly flattened or rounded, uni- formly fertile to the attachment of the translators........... ASCLEPIADEAE Pollinia usually horizontal or essentially so, occasionally ascending or descend- ing, but one or both faces more or less excavated, and with a sterile hyaline margin or indentation near the attachment of the translators... .GONOLOBEAE Pollinia strictly erect, their faces uniformly rounded, uniformly fertile to the attachment of the translators... с Eee wos пежо TYLOPHOREAE ASCLEPIADEAE Erect or decumbent perennial herbs; pollinia very strongly flattened; corona of 5 eueullate, ealeeolate, or clavate hoods, usually with an internal horn | КК oon о оо eor ch xe КОКС. E "ASCLEPIAS Lianas or twining undershrubs; pollinia faces broadly rounded or only e aie compressed. of translators conspicuously thickened and тр еи near their неке to the gland.. ево... . OXYPETALUM [Vor. 28 204 ANNALS OF THE MISSOURI BOTANICAL GARDEN Arms of translators not as above. Corona of 5 separate or united, laminate to filiform scales, ос- easionally eompounded or with internal processes, rarely wholly lackin "e re By oe IY TI ЧҮҮ e ООН, III. CYNANCHUM f the fede ace ay has cas ates Seb iGbs) OA IV. BLEPHARODON Corona of 5 closed inflated рако joined at the bases by a fleshy ring adnate to the corolla-throat.............. V. SARCOSTEMMA G'ONOLOBEAE Anthers relatively simple, not ag раты vesieular, nor with dorsal ap- pendages; corolla lobes various, but not erisped............ VI. MATELEA Anthers very conspicuously чи ИДИ В and vesicular throughout; corolla lobes stri vg DEDE IAS UE LL гое VII. FISCHERIA Anthers with spreading, more or less laminate, fleshy dorsal appendages; corolla lobes various, but not егізрей.................. VIII. GONOLOBUS 'TYLoPHOREAE и. ЛО ИА TI T CIPIT C IX. MARSDENIA I have been unable to interpret two enigmatie genera of Baillon, Microstelma (Hist. Pl. 10: 286. 1891) and Stelmago- num (loe. cit. 287), both placed in the Gonolobeae and thus probably synonymous under Matelea. I. Авспертав L. Sp. Pl. 214. 1753. Anthanotis Raf. Fl. Ludov. 52, 149. 1817. Anantherix Nutt. Gen. N. Am. Pl. 1: 169. 1818. Stylandra Nutt. loc. eit. 170. 1818. Otaria HBK. Nov. Gen. 3: 192. 1819. Acerates Ell. Sketch Bot. S. Carol. 1: 316. 1821. Podostigma Ell. loc. cit. 326. 1821. Acerotis Raf. New Fl. N. Am. 1: 49. 1836. Oligoron Raf. loc. cit. 4: 60. 1836. Otanema Raf. loc. cit. 61. 1836. Onistis Raf. loc. cit. 63. 1836. Polyotus Nutt. Trans. Amer. Phil. Бос. n.s. 5: 199. 1837. Asclepiodora A.Gray, Proc. Amer. Acad. 12: 66. 1877. 1941] WOODSON—NORTH AMERICAN ASCLEPIADACEAE. I 205 Schizonotus A.Gray, loc. cit. 1877, non Lindl. Solanoa Greene, Pittonia 2: 67. 1890. Solanoana O.Ktze. Rev. Gen. 2: 421. 1891. Oxypteryx Greene, Pittonia 3: 234. 1897. Podostemma Greene, loc. cit. 285. 1897. Biventraria Small, Man. Fl. 1072. 1933. Asclepiodella Small, loc. cit. 1073. 1933. Gomphocarpus of American authors. KEY ro THE SUBGENERA A. кр rotate to rotate-subcampanulate, the lobes reflexed to somewhat une corona attached to the column immediately beneath the Mid: B. Hoods us or cucullate-spatulate to calceolate, rarely strongly conduplieate, ereet to обе: the alternate lobules more or less deeply 2-cleft, rarely a C. Corolla throat without enum eallous processes. D. Hoods usually more or less substipitate as seen from within, the basal attachment shallow, eonduplieate but not deeply saccate 239/290 2 0. 872 ПВР 1. EUASCLEPIAS DD. Hoods very sessile, the basal attachment deeply saceate. E. Hoods entire to somewhat cleft ventrally, open but occasionally appressed against the column. F. Base of hoods not ma c RN or merely somewhat keeled laterally. G. Hoods with a more or less conspicuous internal horn or crest. H. Horn or crest compressed radially ; ме not deeply sae eate between the hoods............ . ASCLEPIODELLA HH. Horn or crest compressed — очи: qi deeply eate between the hoods......... . PENTASTOMATIA GG. Hoods puma ut a horn or crest. H. Hoods with very pronouneed marginal lobes, the orifi . ASCLEPI not appressed against the column. IOPHANES HH. Hoods with very inconspieuous acu lobes, if any, the rifiee appressed against the column...... 5. ACERATES FF. Base of hoods with conspicuous, external, laterally excurrent mbranous аррепйаф@ев...................., 6. POLYOTUS EE. un E pups. IT ventrally, elosed, Фр еп- closing the horn if presemt.....-..-... senses ee ae . SOLANOA CC. Corolla with an interrupted pre annulus of 5 conspicuous callous processes alternating with the hoods; hoods strongly conduplicate, wi onspicuous one potett erest....8. ASCLEPIODOLUS BB. Hoods. tiens involute-clav C. Corolla lobes reflexed ; мыз with deeply bifid alternating lobules; Ld arms very long. со. 9. ANANTHERIX [Vor. 28 206 ANNALS OF THE MISSOURI BOTANICAL GARDEN CC. Corolla lobes ascending; hoods with entire alternating lobules; trans- lator arms of moderate length................. 10. ASCLEPIODORA AA. Corolla rotate-subtubular, the lobes ereet or only slightly spreading; anther head borne on a slender stipe high above the eorona......... В он eee ие NEES 11. PODOSTIGMA This key will not be of great use to those who are not thoroughly familiar with the varying structure of the Milk- weed flower. The monograph of the genus which I am prepar- ing will contain the discussion and illustrations of the criteria necessary for general use. To many readers familiar with the segregate genera of Asclepias in a restricted region of the United States, the redue- tion of such well-established entities as Acerates, Asclepi- odora, and Podostigma may appear as an admission of casual superficiality. Let them study the scores of species of Ascle- pias represented in North America, intensively, for several years, however (and above all, let them attempt to compose an adequate key even to subgenera and sections), and I am con- vinced that those without prejudice will appreciate the practi- eal and theoretical advantages of ‘‘lumping.’’ Fortunately, very few new combinations are necessary by interpreting the genus in a broad sense, for practically all species have been described under Asclepias at one time or another. The rela- tively few exceptions amongst the species with which I am fa- miliar at present are treated under their respective subgenera as follows: SUBGENUS 1. EUASCLEPIAS AscLEPIAS hypoleuca (A.Gray) Woodson, comb. nov. Gomphocarpus hypoleucus A.Gray, Proc. Amer. Acad. 17: 222. 1881-82. AscLEPIAS Phenax Woodson, nom. nov. Acerates humilis Benth. Pl. Hartw. 291. 1848, non Ascle- pias humilis Schltr. AscrEPIAS Pringlei (Greenm.) Woodson, comb. nov. Acerates Pringlei Greenm. Proc. Amer. Acad. 34: 570. 1899. 1941] WOODSON—NORTH AMERICAN ASCLEPIADACEAE. I 207 SUBGENUS 5. ACERATES Ascuepias hirtella (Pennell) Woodson, comb. nov. Acerates hirtella Pennell, Bull. Torrey Club 46: 184. 1919. SUBGENUS 6. POLYOTUS | Asciepias Engelmanniana Woodson, nom. nov. Acerates auriculata Engelm. in Torr. Rept. Bot. Mex. Bound. Surv. 160. 1859. Asclepias auriculata (Engelm.) Holzinger, Bot. Gaz. 17: 125, 160. 1892, non HBK. SUBGENUS 7. SOLANOA AscrEPIASs Solanoana Woodson, nom. nov. Gomphocarpus purpurascens A.Gray, Proc. Amer. Acad. 10: 76. 1874, non A.Rich. Schizonotus purpurascens A.Gray, loc. cit. 12: 66. 1877. Solanoa purpurascens (A.Gray) Greene, Pittonia 2: 67. 1890. Solanoana purpurascens (A.Gray) O.Ktze. Rev. Gen. 2: 421. 1891. SUBGENUS 8. ASCLEPIODOLUS AscLEPIAS insignis (Brandg.) Woodson, comb. nov. Asclepiodora insignis Brandg. Zoe 5: 253. 1908. SUBGENUS 10. ASCLEPIODORA AscLEPIAS circinalis (Dene.) Woodson, comb. nov. Acerates circinalis Dene. Ann. Sci. Nat. Bot. ii. 9: 322. pl. 10, fig. c. 1838. Asclepiodora circinalis (Dene.) Fourn. loc. cit. vi. 14: 369. 1882. AscLEPIAS Fournieri Woodson, nom. nov. Acerates gomphocarpoides Dene. Ann. Sci. Nat. Bot. ii 9: 323. 1838, non Asclepias gomphocarpoides Schltr. Asclepiodora gomphocarpoides (Dene.) Fourn. loc. cit. vi. 14: 369. 1882. Ascuepias zanthodacryon (Smith & Harris) Woodson, comb. nov. Asclepiodora zanthodacryon Smith & Harris, Contr. Gray Herb. 114: 12. 1936. [Vor. 28 208 ANNALS OF THE MISSOURI BOTANICAL GARDEN IL OxvrETALUM R.Br. Mem. Wern. Бос. 1: 41. 1809. Apparently О. cordifolia (Vent.) Schltr. is the only repre- sentative of this troublesome genus in Central America and the Antiles. Fortunately, this species has strongly appendaged translator arms which enable it to be separated easily from Cynanchum. But in South America there are many species without this diagnostic structure, the importance of which will present one of the major problems of the Asclepiads in the southern continent. III. Cynancuvum Г. Sp. Pl. 212. 1753. Ditassa R.Br. Mem. Wern. Soc. 1: 49. 1809. M etastelma R.Br. loc. cit. 52. 1809. Enslenia Nutt. Gen. N. Am. Pl. 1: 164. 1818, non Raf. Ampelamus Raf. Amer. Monthly Mag. 4: 192. 1819. Lyoma Ell. Sketch Bot. S. Carol. 1: 316. 1821, non Nutt. Seutera Reichenb. Consp. 131. 1828. Enslinia Reichenb. loc. cit. 1828. Roulima Dene. in DC. Prodr. 8: 516. 1844, non A. Brongn. Orthosia Dene. loc. cit. 526. 1844. Tassadia Dene. loc. cit. 579. 1844. Irmischia Schlecht. Linnaea 19: 738. 1847. Nanaturis Turez. Bull. бос. Nat. Mose. 211: 254. 1848. Amphistelma Griseb. Fl. В.М. 417. 1861. Tylodontia Griseb. Cat. Pl. Cub. 175. 1866. Metalepis Griseb. loc. cit. 179. 1866. Mellichampia A.Gray, Proc. Amer. Acad. 22: 437. 1887. Pattalias S.Wats. loe. cit. 24: 60. 1889. Tainionema Schltr. in Urb. Symb. Ant. 1: 263. 1899. Decastelma Schltr. loc. cit. 264. 1899. Rouliniella Vail, Bull. Torrey Club 29: 662. 1902. Basistelma Bartlett, Proc. Amer. Acad. 44: 631. 1909. E picion Small, Man. Fl. 1075. 1933. Astephanus of American authors. Besides the preceding, several synonyms eventually will have to be added from the South American flora. Very few of 1941] WOODSON—NORTH AMERICAN ASCLEPIADACEAE. I 209 these synonymous genera are absolutely co-extensive, since they are based for the greater part upon variations of the corona. Because these differences are multitudinous but of the same general motif, the practieal solution of the problem ap- pears to lie in an inclusive treatment, such as that adopted for Asclepias. The principal characters used to separate the seg- regates just enumerated are aestivation of the corolla lobes (whether contorted or valvate—extremely difficult to observe in most species), corona (whether simple or compound, deeply or less deeply divided, or absent), and structure of the inflo- rescence (whether racemiform or umbelliform—since the basic structure is cymose for all, this distinction resolves itself into a comparison of internode length). These characters combine kaleidoscopically in so many manners, and frequently result in the separation into different genera of so many species of ob- vious affinity, that they appear of very doubtful value. I am inclined to discount particularly the value of corolla aestiva- tion and corona structure, and to group the North American species under the following subgenera. These I am illustrating with representative adjustments in nomenclature. KEY ro THE SUBGENERA A. Cymes racemiform to corymbiform, with definitely manifest internodes. B; d ко nulate to rotate-subeampanulate, the throat not con- t the orifice. ©: uicit lobos relatively elongate, free or united only at the bases. D. Corona lobes acuminate, entire or with very obscure lateral lobules ————— ССГ. c 1. MELLICHAMPIA DD. Corona lobes deeply їйӣ....................... 2. AMPELAMUS CC. Corona от broadly and shallowly lobed, the lobes eR arginate oe ovre В A РНЕ 3. METALE BB. Gristle Дарение the throat constricted at the orifice... .4. а АА. Cymes umbelliform, usually without the appearance of definite internodes. B. Corolla lobes ascending or spreading............. 5. METASTELMA BB. Corolla lobes inflexed-euecullate................ eese 6. CLEISTOLOBUS SUBGENUS 1. MELLICHAMPIA Сумлмсном jaliscanum (Vail) Woodson, comb. nov. Rouliniella jaliscana Vail, Bull. Torrey Club 29: 668. 1902 [Vor. 28 210 ANNALS OF THE MISSOURI BOTANICAL GARDEN CynancuuM jamaicense (Griseb.) Woodson, comb. nov. Enslenia jamaicensis Griseb. Fl. B. W.I. 418. 1861. Roulimella jamaicensis (Griseb.) Rendle, Journ. Bot. 74: 340. 1936. CvNaNcHuM lignosum (Vail) Woodson, comb. nov. Rouliniella lignosa Vail, Bull. Torrey Club 29: 666. 1902. Сүхлхоном ligulatum (Benth.) Woodson, comb. nov. Enslenia ligulata Benth. Pl. Hartw. 290. 1848. Mellichampia rubescens A.Gray, Proc. Amer. Acad. 22: 437. 1887. Ampelamus ligulatus (Benth.) Heller, Contr. Herb. Franklin & Marshall Coll. 1: 79. 1895. Mellichampia ligulata (Benth.) Vail, Bull. Torrey Club 26: 425. 1899. HRoulinia ligulata (Benth.) Pittier, Contr. U.S. Nat. Herb. 13: 111. 1910, as to synonymy. Cynancuum Rensoni (Pittier) Woodson, comb. nov. Roulima Rensoni Pittier, Contr. U.S. Nat. Herb. 13: 101. 1910. Cynancuum saepimentorum (Brandg.) Woodson, comb. nov. Vincetoxicum saepimentorum Brandg. Univ. Cal. Publ. Bot. 4: 381. 1913. Cynancuvum sinaloense (Brandg.) Woodson, comb. nov. Roulima sinaloensis Brandg. Zoe 5: 243. 1908. Mellichampia sinaloensis (Brandg.) Kearney & Peebles, Journ. Wash. Acad. Sci. 29: 488. 1939. Cynancuum unifarium (Scheele) Woodson, comb. nov. Gonolobus unifarius Scheele, Linnaea 21: 760. 1848. Roulima unifaria (Scheele) Engelm. in Torr. Rept. Bot. Mex. Bound. Surv. 160. 1859. Roulimella unifaria (Scheele) Vail, Bull. Torrey Club 29: 665. 1902. CynancHum Watsonianum Woodson, nom. nov. Roulima Palmeri S.Wats. Proc. Amer. Acad. 18: 115. 1883, non Cynanchum Palmeri (Wats.) Blake (Pat- talias Palmeri Wats.). Roulinella Palmeri (S.Wats.) Vail, Bull. Torrey Club 29: 664. 1902. 1941] WOODSON—NORTH AMERICAN ASCLEPIADACEAE. I 211 Ampelamus, Rouliniella, and Mellichampia can be recog- nized as genera only by extremely tenuous criteria. As far as I am aware, the first would have to be supported by the bifid corona segments and the rostrate stigma head; the second by a flat (or scarcely conical) stigma head and acuminate crown segments, although they are slightly emarginate in some species and not infrequently show a tendency toward the de- velopment of marginal lobules; the third would have to depend upon somewhat larger flowers than Rouliniella, somewhat more nearly campanulate corollas, and crown segments some- what more united at the base. Mellichampia also has a stigma head intermediate between that of Ampelamus and that of Rouliniella. The crown segments of Ampelamus and of Melli- champia are flat; those of Rouliniella are usually somewhat cucullate or infolded at the base. The primary division of the key to subgenera advocated here is one that will have to be used in some cases with a certain degree of experience, if not prejudice. Whilst the inflorescence of the Mellichampia-complex may be described as racemiform generally, and that of the Metastelma-complex as umbelliform, ambiguous species occur in both groups quite frequently enough to support my contention that it is impractical to main- tain a series of distinct genera. Whilst the flowers of the Mellichampia-complex are smaller, as a rule, than those of the Metastelma-complex, this is in the nature of a guiding ‘‘ preju- dice’’ only. I have been unable to discover supporting struc- tural characters of merit. SUBGENUS 2. AMPELAMUS CYNANCHUM LAEVE (Michx.) Pers. Syn. 1: 274. 1805. Gonolobus laevis Michx. Fl. Bor. Am. 1: 119. 1803. Enslenia albida Nutt. Gen. N. Am. Pl. 1: 165. 1818. Ampelanus albidus (Nutt.) Britton, Bull. Torrey Club 21: 314. 1894. The interpretation of this species has been singularly con- fused. In 1803 Michaux (loc. cit.) founded Gonolobus with three species, two of which previously had constituted the whole complement of the earlier genus Vincetoxicum Walt. [Vor. 28 212 ANNALS OF THE MISSOURI BOTANICAL GARDEN (Fl. Carol. 104. 1788): G. macrophyllus (V. gonocarpos Walt.), and G. hirsutus (V. acanthocarpus Walt.). The third species, G. laevis, originated with Michaux. It is important in this connection to emphasize certain morphological characters of the fruit as embodied in the generic and specific descriptions by Michaux. For the genus as a whole: ''Folliculi plerumque costati seu angulosi. . .’’ The follicles of the three species were described as follows: G. macrophyllus—‘folliculis costato-angulosis’’; for G. hir- sutus—‘folliculis oblongis, sparsim muricatis’’; but for G. laevis—‘folliculis laevibus.’’ It is obvious that the smooth condition of the fruit of G. laevis—not angled as in macro- phyllus, nor muricate as in hirsutus—was particularly out- standing in the estimation of Michaux, and that it suggested the specific adjective itself. The type specimens of G. laevis, if they may be called such, have been examined both by Dr. Gray (Proc. Amer. Acad. Sci. 12: 75. 1877) and by Miss Vail (Bull. Torrey Club 26: 427. 1899), and both agreed that at least two elements are repre- sented amongst the fragments, namely, flowers and angled fruits of G. suberosus (L.) R.Br.—of which G. macrophyllus Michx. (G. gonocarpos (Walt.) Perry) probably is no more than a variety—and leaves of the slender vine currently known as Enslenia albida Nutt. or Ampelanus (correctly Ampelamus ) albidus (Nutt.) Britton. But Dr. Gray chose the flowers and fruit as authentic, whilst Miss Vail chose the leaves. At first glance Miss Vail’s selection may appear to be with little foundation, until it is recalled that Michaux stipulated for his species ''folliculis laevibus," exactly the condition of Ampelamus; and that in choosing the material with angled pods, Dr. Gray denied the diagnostic character of the species. In such an instance, of obviously accidental mixture, the evi- dent solution is to accept, as typical, material known to со- incide with the original literature. Had she studied her problem under existing rules of nomen- clature, Miss Vail certainly would have placed laevis in a new combination under Ampelamus. Under the interpreta- tions current at the New York Botanical Garden during that 1941] WOODSON—NORTH AMERICAN ASCLEPIADACEAE. I 213 time, however, she invoked the ‘‘Doctrine of Residues,’’ re- turning С. macrophyllus and С. hirsutus to Vincetoxicum Walt., from which they had been taken by Michaux, and re- taining the name Gonolobus for G. laevis. In this connection, however, it should be noted that Michaux embodied in his diagnosis of Gonolobus two characters which do not coincide with the present interpretation of G. laevis: ** Stylus discoideo-5-gonus. . . КоШсий plerumque costati seu angulosi."" The absence of a ** Doctrine of Residues’’ notwith- standing, therefore, I feel that Miss Perry (Rhodora 40: 283. 1938) is quite correct in considering the Walterian element as typical of Gonolobus, choosing G. macrophyllus as the standard species. Nevertheless, one feels that sentiment may have induced her to perpetuate Dr. Gray's interpretation of G. laevis rather than the more logieal conclusion of Miss Vail. Although it does not bear directly upon the question of G. laevis, it might be well to explain here, for future refer- ence, that although Vincetoxicum Walt. clearly antedates Gonolobus Michx., as well as Vincetoxicum Moench (Meth. 717. 1794), the nomenclatorial powers at Kew, invoked by Miss Perry (loc. cit. 281-282. 1938), ‘‘temporarily’’ have placed Vincetoxicum Moench on the list of Nomina Con- servanda, tantamount to final acceptance, since ‘‘the European Vincetoxicum is so widely used that there is every chance of its being conserved”’ (passage of letter by Mrs. T. A. Sprague, as quoted by Miss Perry). I suspect that the European Ascle- piadologists, having had Vincetoxicum Moench conserved for them, will scarcely know how to use it: its distinction from Cynanchum is extremely evasive. SUBGENUS 3. METALEPIS CynancHum cubense (Griseb.) Woodson, comb. nov. Metalepis cubensis Griseb. Cat. Pl. Cub. 179. 1866. The inclusion of Metalepis within the Gonolobeae by Schu- mann (in Engl. & Prantl, Nat. Pflanzenfam. 42: 297. 1895) is a conspieuous instance of the ambiguity of the current defini- tions of the tribes of Asclepiadaceae. The pollinia in the species of this subgenus are quite pendulous, with the equally [Vor. 28 214 ANNALS OF THE MISSOURI BOTANICAL GARDEN rounded surfaces characteristic of the Asclepiadeae, but the translators are very long and horizontal, a condition found exaetly duplieated in a few species of the subgenus Melli- champia. Aside from this character, the subgenus depends upon the racemiform inflorescences and the eyathiform corona, the inadequacy of which, as generic characters, has been noted previously. SUBGENUS 4. TYLODONTIA CvNANCHUM Grisebachii (Maza) Woodson, comb. nov. Tylodontia cubensis Griseb. Cat. Pl. Cub. 179. 1866, non Cynanchum cubense (Griseb.) Woodson (Metalepis cu- bensis Griseb.). Astephanus Grisebachii Maza, Cat. Periant. 276. 1894. Tylodontia has been included as a synonym under Astepha- nus by Schumann (in Engl. & Prantl, Nat. Pflanzenfam. 4?: 224. 1895), but dissection of flowers from an isotype (Wright 2964) in the herbarium of the Missouri Botanical Garden shows the presence of a corona of five erose, truncated scales. The subgenus is outstanding in the Mellichampia-complex be- cause of the small flowers having urceolate corollas with sharply reflexed lobes. SUBGENUS 5. METASTELMA At present I do not feel sufficiently familiar with the syn- onymy of Metastelma to attempt the many nomenclatural changes necessary for inclusion in Cynanchum. This is the most difficult of the American groups of Asclepiads because the flowers are so small. Within the tiny corollas, averaging only about 2 mm. in length, are packed as many structural details and variations as within the larger flowers of other genera. Once when I was complaining to Dr. Standley of the eye and nerve strain necessary in dealing with the group, he laughingly agreed that it should be studied by an algologist or a mycologist. Metastelma will be handled very much easier, I believe, if it is treated as a single subgenus under Cynanchum. Unques- tionably the various entities included here are much more closely related to one another than to any other complex. The 1941] WOODSON—NORTH AMERICAN ASCLEPIADACEAE. I 215 whole aspect of the plants is so unmistakable that I feel, even with convenience particularly in mind, that ‘‘lumping’’ is natural as well. As has been remarked in a previous paragraph, Astephanus is difficult to maintain upon a sure foundation even with the higher powers of a binocular dissecting microscope. Amongst several specimens ascribed to A. pubescens Greenm. in the herbarium of the Missouri Botanical Garden are two entities, one with a manifest corona and one without, the anthers of the two species differing considerably; yet the superficial aspects of the plants are all but undistinguishable. Natural genera should not be so easily confused. Many readers may be surprised at the inclusion of Ditassa and Decastelma with Metastelma. But the cleaving of the corona scales or the presence of a ventral tooth or ligule, which distinguish these entities from Metastelma (sensu stricto), are found amongst the Old World Cynancha as well as in the Mellichampia-complex of Central America. Similar variation also is found in Asclepias. SUBGENUS 6. CLEISTOLOBUS Cynancuum utahense (Engelm.) Woodson, comb. nov. Astephanus utahensis Engelm. Amer. Nat. 9: 349. 1875. The inflexed-cucullate corolla lobes of this species, unlike those of any other known to me, were interpreted by Dr. Engelmann as a modification to compensate for the absence of a corona. IV. BrgPHanopow Dene. in DC. Prodr. 8: 603. 1844. The nine species assigned to Blepharodon by Decaisne fall into two groups superficially marked by volubile or ereet her- baceous habits, but accompanied by interesting differences in the corona and pollinia as well. Although I may seem incon- sistent in view of my treatment of generic lines generally in this family, I feel that two distinct genera may be represented. The North American species are all of the volubile group (with the exception of B. nerüfolium Dene., which I believe [Vor. 28 216 ANNALS OF THE MISSOURI BOTANICAL GARDEN probably will be found to be an Asclepias when the material is avallable for study), and since this group seems to have been more typical of the genus in the opinion of Decaisne, I am reserving the use of the generic name for that connection. In respect for the memories of the many genera that I have reduced to synonymy during these studies, I am deferring judgment on the generic status of the erect species until I have the opportunity to devote my attention to the South American Asclepiads. V. SancosrEMMA R.Br. Mem. Wern. Soc. 1: 50. 1809. Philibertia НВК. Nov. Gen. 3: 195. 1819. Pentagonium Schauer, Nova Acta Acad. Caes. Leop. Nat. Cur. 19. Suppl. 1: 364. 1843. Zosima Phil. Sert. Mendoz. Alt. 29. 1871. Funastrum Fourn. Ann. Sci. Nat. Bot. vi. 14: 388. 1882. Cystostemma Fourn. in Mart. Fl. Bras. 6*: 204. 1885. Philibertella Vail, Bull. Torrey Club 24: 305. 1897. Ceramanthus Malme, Ark. Bot. 44: 2. 1905. A few additional synonyms will have to be added from the South American flora. I am very reluctant to disregard Schlechter’s separation of Philibertia and Funastrum (in Fedde, Repert. 13: 279-287. 1915), but feel that it is necessary to do so in order to maintain balance within the family. As far as I am aware, the only character really separating the two entities, upon the basis of present speciation, is the structure of the corolla, whether campanulate with shallow lobes in the former, or rotate with more deeply divided lobes in the latter. Other definitive structures apparently are lacking, and the striking similarity, if not identity, of the anther, pollinia, and corona characters is impressive. The high attachment of the corona bladders to the anthers, cited by Schlechter for Phili- bertia, manifestly does not hold. The annulus connecting the bladdery corona segments is usually more pronounced in Funastrum than in Philibertia, but is present in both. Sarco- stemma, currently interpreted as including only Old World species, was erected by Robert Brown to include species now 1941] WOODSON—NORTH AMERICAN ASCLEPIADACEAE. I 211 relegated to Funastrum and Philibertia as well. I can find no tangible distinetion between the three, although they were placed in different subtribes by Schumann, as has been ex- plained previously. Few North American species have been described since the prevalence of the restricted definition of Sarcostemma, and only the following transfers appear to be necessary : SarcostemMa tomentella (Brandg.) Woodson, comb. nov. Philibertia tomentella Brandg. Univ. Cal. Publ. Bot. 4: 90. 1910. SancosrEMMA Torreyi (A.Gray) Woodson, comb. nov. Philibertia Torreyi A.Gray, Proc. Amer. Acad. 12: 64. 1877. Philibertella Torreyi (A.Gray) Vail, Bull. Torrey Club 24: 309. 1897. Funastrum Torreyi (A.Gray) Schltr. in Fedde, Repert. 13: 287. 1915. VI. Mareea Aubl. Fl. Guian. 1: 277. 1.109. 1775. Hostea Willd. Sp. Pl. 1: 1274. 1798. Macroscepis HBK. Nov. Gen. 3: 200. 1819. Pherotrichis Dene. Ann. Sci. Nat. Bot. 11. 9: 322. 1838. Ibatia Dene. in DC. Prodr. 8: 599. 1844. Polystemma Dene. loc. cit. 602. 1844. Dictyanthus Dene. loc. cit. 604. 1844. Chthamalia Dene. loc. cit. 605. 1844. Ptycanthera Dene. loc. cit. 606. 1844. Trichosacme Zucc. Abh. Akad. Wiss. München 42: 11. 1845. Tympananthe Hassk. Flora 30: 757. 1847. Rytidoloma Turez. Bull. Soc. Nat. Mose. 252: 319. 1852. Callaeolepium Karst. Fl. Col. 2: 123. 1865. Poicilla Griseb. Cat. Pl. Cub. 176. 1866. Himantostemma A.Gray, Proc. Amer. Acad. 20: 294. 1885. Rothrockia A.Gray, loc. cit. 295. 1885. Tetracustelma Baill. Hist. Pl. 10: 292. 1891. Urostephanus Robins. & Greenm. Amer. Journ. Sci. iii. 50: 159. 1895. [Vor. 28 218 ANNALS OF THE MISSOURI BOTANICAL GARDEN Prosthecidiscus Donn.Sm. Bot. Gaz. 25: 149. 1898. Labidostelma Schitr. Bull. Herb. Boiss. ii. 6: 843. 1906. Microdactylon Brandg. Zoe 5: 252. 1908. Amphorella Brandg. Univ. Cal. Publ. Bot. 4: 91. 1910. Poicillopsis Schltr. in Urb. Symb. Ant. 7: 339. 1912. Pachystelma Brandg. Univ. Cal. Publ. Bot. 7: 330. 1990. Cyclodon Small, Man. Fl. 1075. 1933. Odontostephana Alexander, in Small, loc. eit. 1076. 1933. Edisoma Small, loc. cit. 1078. 1933. Heliostemma Woodson, Amer. Journ. Bot. 22: 689. 1935. Gonolobus and Vincetoxicum of North American authors, in large part. To this imposing list of synonyms eventually will have to be added numerous generie names from the South American flora. Very few of these entities are strietly co-extensive, nearly all having been based upon some more or less striking variation of the corona. Loud protestations probably will be made by botanists familiar with such extremes as Dictyanthus and Macroscepis that at least those familiar genera should be allowed to stand inviolate, but I should like to lay a curse on the man who revives them without at least as much study as I have devoted. And I feel compelled to enjoin all future sys- tematists to ponder long and deeply the intricacies of the gonoloboid corona before proposing additional novelties. The corona of Matelea, as is almost universal in the Cynan- choideae, consists of a unit enation of the anther filament, fundamentally subtending an additional enation (such as the hood and horn, respectively, of most species of Asclepias). The outer units may be separate, essentially entire, or with variously elaborated marginal lobules; at other times they may be concrescent into an inconspicuous, fimbriate skirt at the base of the gynostegium, into a ring, or into a massive ro- tate dise adnate to the corolla throat. The inner units may be small appendages of the anther head stipe rather high above the outer units, or may be combined, either essentially free or completely adnate, with the outer units. The result is a kaleidoscopic medley that can be conveyed in print only by the use of critical illustrations for each of the dozens of species. 1941] WOODSON—NORTH AMERICAN ASCLEPIADACEAE. I 219 At one time during the studies which preceded this paper, an intensive study was made of the anther structure of the Gonoloboids, for I felt that the position of the anthers with respect to the stigma head (whether beneath or about the mar- gin) would aid in the establishment of some of the favorite genera apparently doomed to synonymy. Here numerous trends were obvious, but nothing of sufficient stability for generie distinetion. A long consideration of pollinium struc- ture also was ended in rather bitter frustration. A peculiar vegetative character which links practically all the species groups of Matelea (although by no means every species of the groups) is the mixed indument consisting of long eglandular hairs and short, bulbose emergences. These emer- gences usually appear to be somewhat glandular, and may be white, dark brown, or black. Since no other group of Ascle- piads with which I am familiar possesses such an indument, with the exception of the closely neighboring Fischeria main- tained provisionally and with some misgivings, I regard its occurrence in Matelea as extremely significant. I am соп- vineed that the elements which I have included in that genus are inextricably related. Heretofore, Matelea has been a genus of perhaps four South American species poorly represented in North American her- baria. Its expansion as advocated here consequently entails a shocking number of new combinations. Only a few of the generic synonyms (such as Macroscepis and Dictyanthus) contain as many as half a dozen species, however, and in view of that fact and the many nomenclatural changes necessary in any event it seems scarcely worth while to ask conservation for any of those names. Then, too, such a drastic reformation may better be served by a poorly known generie name than by one with more definitely established associations. Recent tropical American collections are bringing to light more novelties in Matelea than perhaps in any other genus of Asclepiads. Therefore I have prepared the following key to subgenera and sections, both to illustrate the reformation of the genus and as a temporary aid to herbarium study. I have treated several groups as subgenera simply because they have [Vor. 28 220 ANNALS OF THE MISSOURI BOTANICAL GARDEN familiar names and in spite of the fact that their characters are less important than certain others which are treated as sec- tions, having no previous generie name (for I am wary of treating them as nouns lest some floristie student raise them to genera). As accumulation of Asclepiads from tropical orth America proceeds, additional entries will have to be inserted in the key. That should accentuate the greater convenience of an ultra-conservative generic concept. As a final word, it should be emphasized that the species groups show great variation amongst their constituents, and that they cannot safely be raised to generic rank as they now exist. Certain species show affinities for more than one group. Where I have made an arbitrary disposition in such cases I usually have included notes or references to aid their identi- fication. Ккү то THE SuBGENERA AND SECTIONS A. Pollinia more or less descending from the translator arms. B. Corolla subeampanulate-rotate, without a faucal annulus, the limb merely tape А the dilation of the throat and much longer than it; corona , the 5 constituent segments Е. = 3-lobed ; бете + Ати prolonged beyond the attachmen the tr де arms, excavated on one face only............ a ae BB. Corolla very broadly campanulate, the lobes very broad and scarcely as long as the shallow open throat, without a faucal annulus; corona rotate, of 5 segments consisting of 2 faleate outer lobes and a median inflexed lobule; pollinia conspicuously excavated on both Pl. corn orcia sach ad cad ВАРЕНО: ns dec 2. LABIDOSTELMA ВВВ. Corolla salverform-rotate, usually with а faucal annulus, the limb abruptly spreading or somewhat reflexed from the short tube. C. Corolla lobes without caudate appendages; corona of 5 digitate seg- ments adnate to the corolla tube, each with a more or less distinct callous boss supporting the stigma head; pollinia without apical processes; plants variously pubescent, but not white-woolly...... АЛИНА ee 3. MACROSCEPIS CC. Corolla lobes with long plumose caudate appendages; corona annu- lar, shallowly 5-lobed, each lobe with an inconspicuous internal ligule; pollinia with a slender apical process; entire plant v densely үћһіфе-тғооПу.......................... 4, TRI CHOSACME AA. Pollinia essentially horizontal, but occasionally arcuate, with the tips somewhat ascending or descending. B. — species (corolla rotate; if deeply campanulate, see Pachy- telma). 1941] WOODSON—NORTH AMERICAN ASCLEPIADACEAE. I 221 C. Corona of 5 broad barely united segments each bearing a conspieu- ous ineurved hook or ligule; anther head very Mc stipitate NRI с dece tisse ы» АЕ Pay sie Totes sre iie ЕРЕ POICILLA СС. Corona not ав яроте |: с. ie vies vere ее 6. dU conus D. Anther head de SUpitate..-- v... ce vine = delere e $. Pauciflor D. Anther head T т eea E a cert $. Variifoliae D BB. Continental specie C. Pollinia Ее ог oblong-reniform with a narrow hyaline TRL corona of 5 more or less united segments each with an internae pulo огаронн еее 7. CHTHAMALIA CC. райна faleiform or areuate, the tips markedly ascending, hyalin rgins or indentations ае жр ог laeking. D. Ereet ог prostrate herbs; corolla lobes ascending. E. Corolla rotate-subeampanulate, the orifice not constr p corona segments not strongly adnate to the corolla throa TERE Pu ERI RE E E RIEN UEM ОЕ 8. bios EE. Corolla urceolate-campanulate, the orifice markedly constricted ; corona strongly adnate to the corolla throat...9. A ELLA DD. Lianas or twining undershrubs (except M. caudata) ; eorolla lobes reflexed or sprea E. Corolla rotate, id long narrow ascending lobes; corona of 5 eld separate pectinate segments; d head with a appendiculate stipe.............. TIARASTEMMA EE. ODE нсә campanulate, with relatively 3d Tor "i т: с of 5 fimbriately compound TNCS anth ead sessile. esoe cocco eo rnb ves reese ses POLYSTEMMA EEE. eos ое РА with ob atti Decal widely spreading or reflexed lobes; corona of 5 broad, more or less earuneulate segments usually strongly adnate to the corolla throat; anther head sessile............... 12, HELIOSTEMMA CCC. Pollinia subtriangular-pyriform, with a экы аер hyaline indenta- tion or margin and broad winged translator ar D. Corolla rotate to subeampanulate-rotate; anther head round to very oadly 5-gonal; corona more or less annular or rotate......... о ее ВЕЕР Fa Eo раде 13. EUMATELEA E. Corolla without a faucal annulus exterior to the true eorona . Anther head with a strongly manifest fluted stipe. .§. Л РИИ FF. Anther head US sessile; corona with 5 broad parti- tions superposed. G. Corona PA and pna, Пири аба есе 0.3 $. Violac GG. Corona dark and сагипешайе............. ‚ а бое ЕЕ. a throat with an mi ae 5-angled забез] annulus e or to the true: сото ОТИТ $. Vi ridiforar DD. ix deni Napa anther head saliently 5-gonal; coro of 5 mpound digitate segments. E. e iiec digitately compound, adnate to the column but essentially free from the corolla; corolla lobes strikingly pilose-barbate within................. . MICRODACTYLON [Vor. 28 222 ANNALS OF THE MISSOURI BOTANICAL GARDEN EE. Corona segments simply digitate or merely somewhat caruncu- late, adnate to the corolla throat at least at the base; corolla lobes essentially glabrous within. F. Corona lobes relatively short and thiek, adnate to the eorolla only at the base; eorolla lobes flat...... 15. PACHYSTELMA FF. Corona lobes long and narrow, wholly adnate to the eorolla throat and imbedded in similarly shaped fleshy corolline excrescences; corolla lobes sharply revolute, especially at eee ree rr re eee ТУРУ 16. DICTYANTHUS SUBGENUS 1. IBATIA MarELEA araneosus (Donn.Sm.) Woodson, comb. nov. Gonolobus araneosus Donn.Sm. Bot. Gaz. 47: 257. 1909. Marexea atrocoronata (Brandg.) Woodson, comb. nov. Vincetoxicum atrocoronatum Brandg. Univ. Cal. Publ. Bot. 6: 372. 1917. Млтегвл chrysantha (Greenm.) Woodson, comb. nov. Gonolobus chrysanthus Greenm. Proc. Amer. Acad. 32: 299. 1897. Vincetoxicum chrysanthum (Greenm.) Standl. Contr. U.S. Nat. Herb. 23: 1190. 1924. MarzrEA cordifolia (A.Gray) Woodson, comb. no Rothrockia cordifolia A.Gray, Proc. Amer. ad. 20: 295. 1885. MarELEA fruticosa (Brandg.) Woodson, comb. nov. Rothrockia fruticosa Brandg. Zoe 5: 165. 1903. Матегвл gonoloboides (Robins. & Greenm.) Woodson, comb. nov. Urostephanus gonoloboides Robins. & Greenm. Amer. Journ. Sei. iii. 50: 159. 1895. MarELEA inconspicua (Brandg.) Woodson, comb. nov. Gonolobus inconspicuus Brandg. Univ. Cal. Publ. Bot. 3: 387. 1909. Матегвл maritima (Jacq.) Woodson, comb. nov. Cynanchum maritimum Jacq. Stirp. Amer. 83. pl.56. 1763. Gonolobus maritimus (Jacq.) R.Br. Mem. Wern. Soc. 1: 35. 1809. Gonolobus floccosus Bertol. Opuse. 4: 521. 1823. Gonolobus suberosus Spreng. Syst. 1: 846. 1825, non R.Br 1941] WOODSON—NORTH AMERICAN ASCLEPIADACEAE. I 223 Гана maritima (Jacq.) Dene. in DC. Prodr. 8: 599. 1844. Lachnostoma maritimum (Jacq.) Nichols. Dict. Gard. 2: 236. 1884. Ibatia muricata Griseb. Fl. B.W.I. 421. 1861. МАтЕгЕА mollis (Griseb.) Woodson, comb. nov. Ibatia mollis Griseb. Cat. Pl. Cub. 177. 1866. Lachnostoma molle (Griseb.) Maza, Cat. Periant. 270. 1894. Ptychanthera mollis (Griseb.) Schltr. in Urb. Symb. Ant. 1: 280. 1899. Матегел patalensis (Donn.Sm.) Woodson, comb. nov. Gonolobus patalensis Donn.Sm. Bot. Gaz. 47: 256. 1909. Marzrza petiolaris (A.Gray) Woodson, comb. nov. Gonolobus petiolaris A.Gray, Proc. Amer. Acad. 21: 397. 1886. Vincetoaicum petiolare (A.Gray) Standl. Contr. U.S. Nat. Herb. 23: 1189. 1924. Matera porphyrantha (Standl.) Woodson, comb. nov. Vincetoxicum porphyranthum Standl. ex Yuncker, Field Mus. Publ. Bot. 17: 387. 1938. MarELEA Pringlei (A.Gray) Woodson, comb. nov. Himantostemma Pringlei A.Gray, Proc. Amer. Acad. 20: 294. 1885. Матегжд Prosthecidiscus Woodson, nom. nov. Prosthecidiscus guatemalensis Donn.Sm. Bot. Gaz. 25: 150. pl.12. 1898, non Matelea guatemalensis (K.Sch.) Woodson. МатегвА pueblensis (Brandg.) Woodson, comb. nov. Vincetoxicum pueblensis Brandg. Univ. Cal. Publ. Bot. 4: 91. 1910. MarELEA Purpusii (Brandg.) Woodson, comb. nov. Gonolobus Purpusii Brandg. Univ. Calif. Publ. Bot. 3: 387. 1909. Mare tea tristiflora (Standl.) Woodson, comb. nov. Vincetoxicum tristiflorum Standl. Field Mus. Publ. Bot. 17: 272. 1937. MarELEA umbellata (Brandg.) Woodson, comb. nov. Rothrockia umbellata Brandg. Zoe 5: 165. 1903. [Vor. 28 224 ANNALS OF THE MISSOURI BOTANICAL GARDEN MarELEA Warscewiczii (Karst.) Woodson, comb. nov. Callaeolepium Warscewiceu Karst. Fl. Col. 2: 123. pl. 165. 1865. Fimbristemma Warscewicew (Karst.) Benth. & Hook. Gen. Pl. 2: 768. 1876. Ibatia is placed in the key as having ‘‘pendulous’’ pollinia since their vertical dimension is at least equal to the horizontal and usually is greater. Some confusion with Chthamalia is likely to occur, but may be avoided by a comparison of the eoronas, that of the former subgenus being without internal ligules or bosses adnate to the segments and that of the latter always having such structures. SUBGENUS 2. LABIDOSTELMA MarELEA Quirosii (Standl.) Woodson, comb. nov. Cynanchum rotatum Sesse & Mociño, Fl. Mex. 76. 1887, non Vell. Labidostelma guatemalense Schltr. Bull. Herb. Boiss. ii. 6: 843. 1906, non M. guatemalensis (Donn.Sm.) Wood- son. Vincetoxicum Quirosu Standl. Field Mus. Publ. Bot. 188: 959. 1988. SUBGENUS 3. MACROSCEPIS MaTELEA calcicola (Greenm.) Woodson, comb. nov. Gonolobus calcicola Greenm. Proc. Amer. Acad. 40: 30. 1904. Vincetoxicum calcicola (Greenm.) Standl. Contr. U.S. Nat. Herb. 23: 1191. 1924. Млтегкл congesta (Dene.) Woodson, comb. nov. Gonolobus congestus Dene. in DC. Prodr. 8: 597. 1844. Vincetoxicum congestum (Dene.) Standl. Contr. U.S. Nat. Herb. 23: 1189. 1924. Matera congestiflora (Donn.Sm.) Woodson, comb. nov. Cynanchum hirsutum Kesse & Mociño, Fl. Mex. 76. 1887, non Vell. Macroscepis congestiflora Donn.Sm. Bot. Gaz. 25: 149. 1898. 1941] WOODSON—NORTH AMERICAN ASCLEPIADACEAE. I 225 MarELEA diademata (Edwards) Woodson, comb. nov. Gonolobus diadematus Edwards, Bot. Reg. 3: pl.252. 1817. Vincetoxicum diadematum (Edwards) Standl. Contr. U.S. Nat. Herb. 23: 1188. 1924. Мате вА magnifolia (Pittier) Woodson, comb. nov. Gonolobus magnifolius Pittier, Contr. U.S. Nat. Herb. 13: 104. fig.13. 1910. Vincetoxicum magnifolium (Pittier) Standl. loc. cit. 23: 1188. 1924. MaTELEA obovata (HBK.) Woodson, comb. nov. Macroscepis obovata HBK. Nov. Gen. 3: 200. 1819. SUBGENUS 4. TRICHOSACME Maree lanata (Zuce.) Woodson, comb. nov. Trichosacme lanata Zuce. Abh. Akad. Wiss. Munchen 4?: 11. 1845. An explanation, if not an apology, is in order for reducing Trichosacme from generic status. M. lanata is quite unmis- takable because of the copius lanate pubescence and long plu- mose corolla lobes; but aside from these particulars it is mani- festly intermediate between other groups, having somewhat the corona of Chthamalia, the pollinia of Грана, and the corolla (except the lobe appendages) of a miniature Macroscepis. SUBGENUS 5. POICILLA Martera ovatifolia (Griseb.) Woodson, comb. nov. Poicilla ovatifolia Griseb. Cat. Pl. Cub. 177. 1866. Ptychanthera ovatifolia (Griseb.) Schltr. in Urb. Symb. Ant. 1: 279. 1899. МхлткгЕА tamnifolia (Griseb.) Woodson, comb. nov. Poicilla tamnifolia Griseb. Cat. Pl. Cub. 176. 1866. SUBGENUS 6. PTYCANTHERA $. Pauciflorae Matera acuminata (Griseb.) Woodson, comb. nov. Orthosia acuminata Griseb. Cat. Pl. Cub. 175. 1866. Ptychanthera Berterii ace. to Schltr. in Urb. Symb. Ant. 1: 279. 1899, non Dene. [Vor. 28 226 ANNALS OF THE MISSOURI BOTANICAL GARDEN Poicilla acuminata (Griseb.) Schltr. loe. cit. 5: 469. 1908. Poicillopsis acuminata (Griseb.) Schltr. loc. cit. 7: 339. 1912. MarELEA nipensis (Urb.) Woodson, comb. nov. Gonolobus nipensis Urb. Symb. Ant. 9: 491. 1995. MarkLEA oblongata (Griseb.) Woodson, comb. nov. Orthosia oblongata Griseb. Cat. Pl. Cub. 176. 1866. Ptychanthera oblongata (Griseb.) Schltr. in Urb. Symb. Ant. 1: 280. 1899. Poicilla oblongata (Griseb.) Schltr. loc. cit. 5: 470. 1908. Poicillopsis oblongata (Griseb.) Schltr. loe. cit. 7: 339. 1912. Матв pauciflora (Spreng.) Woodson, comb. nov. Gonolobus pauciflorus Spreng. Syst. 1: 846. 1825. Ptycanthera Berterii Dene. in DC. Prodr. 8: 606. 1844, not acc. to Schltr. in Urb. Symb. Ant. 1: 279. 1899. $. Varüfoliae MarELEA bayatensis (Urb.) Woodson, comb. nov. Gonolobus bayatensis Urb. Symb. Ant. 9: 420. 1995. MarrLEA Ekmanii (Urb.) Woodson, comb. nov. Gonolobus Ekmanii Urb. Symb. Ant. 9: 492. 1925. MarkrzA Sintenisii (Schltr.) Woodson, comb. nov. Gonolobus Sintenisii Schltr. in Urb. Symb. Ant. 1: 288. 1899. Vincetoxicum Sintenisii (Schltr.) Britton, Sci. Surv. Porto Rico & Virgin Isl. 6: 100. 1925. Mare.ea tigrina (Griseb.) Woodson, comb. nov. Gonolobus tigrinus Griseb. Pl. Wright. 520. 1862. Maren variifolia (Schltr.) Woodson, comb. nov. Gonolobus varufolius Schltr. in Urb. Symb. Ant. 1: 286. 1899. Vincetoxicum varüfolium (Schltr.) Britton, Sci. Surv. Porto Rico & Virgin Isl. 6: 100. 1925. The subgenus Ptycanthera was largely the deciding factor in the inclusive treatment of Matelea which I have adopted. 1941] WOODSON—NORTH AMERICAN ASCLEPIADACEAE. I Ser Were it not for the Antillean species, one might compose a fairly respectable key to several genera upon the continent, fol- lowing generally the lines of the key to subgenera and sections of Matelea as it now appears. However, the Antillean species cut so sharply across most of the distinctions between the con- tinental groups that the only way I can distinguish them from the latter in print is by the geographical factor. Were it not for their ambiguity with respect to the continental entities, one might obliterate Ptycanthera entirely, distributing §. Pauciflorae and $. Varüfoliae, perhaps, to the continental sub- genera Humatelea and Heliostemma respectively. Variation of the corona in Ptycanthera appears largely to be responsive to the development of the anther head stipe. Where the latter is fairly pronounced, in $ Pauciflorae, the corona adopts much the same fluted appearance with subtend- ing skirt as in Humatelea $ Reticulatae; but where the anther head is sessile ($ Varufoliae), the corona becomes a rotate dise as in Heliostemma, or a shallow 5-angled ring, in either case with 5 partitions, bosses, or hooks corresponding to the fluted buttresses of $ Pauciflorae. In M. Sintenisii the com- pression of the anther head and corona apparently has led to the formation by the corolla throat of a rather prominent faucal annulus. The anther and pollinium structures are some- what more variable even than in the continental species. Yet, withal, the superficial aspect of the various species is so similar in most cases that I cannot but view them as ex- tremely closely related, particularly in view of their restricted geographical distribution. The exasperations attendant upon their attempted segregation are illustrated graphically by Schlechter’s vacillation between Роса, Ptycanthera (spelled Ptychanthera by him), Poicillopsis, and Gonolobus, with the familiar Vincetoxicum-motif supplied by Dr. Britton. SUBGENUS 7. CHTHAMALIA MaTELEA adenocardium (Standl.) Woodson, comb. nov. Vincetoxicum adenocardium Standl. Field Mus. Publ. Bot. 17: 267. 1937. МатегЕл Baldwyniana (Sweet) Woodson, comb. nov. [Vor. 28 228 ANNALS OF THE MISSOURI BOTANICAL GARDEN Gonolobus Baldwynianus Sweet, Hort. Brit. ed.2. 360. 1830. Vincetoxicum Baldwinianum (Sweet) Britton, Mem. Tor- rey Club 5: 265. 1894. Odontostephana | Baldwiniana (Sweet) Alexander, in Small, Man. Fl. 1077. 1933. MaTELEA biflora (Raf.) Woodson, comb. nov. Gonolobus biflorus Raf. New Fl. N. Amer. 4: 58. 1836. Chthamalia biflora (Raf.) Dene. in DC. Prodr. 8: 605. 1844. Vincetoxicum biflorum (Raf.) Heller, Contr. Herb. Frank- lin & Marshall Coll. 1: 79. 1895. MarELEA brevicoronata (Robins.) Woodson, comb. nov. Gonolobus parviflorus var. brevicoronatus Robins. Proc. Amer. Acad. 26: 169. 1891. MarkrEA camporum (Brandg.) Woodson, comb. nov. Vincetoxicum camporum Brandg. Univ. Calif. Publ. Bot. 4: 185. 1911. MarzrEa carolinensis (Jacq.) Woodson, comb. nov. Cynanchum carolinense Jacq. Coll. 2: 288. 1788. Gonolobus carolinensis (Jaeq.) R.Br. Mem. Wern. Soc. 1: 35. 1809. Vincetoxicum carolinense (Jacq.) Britton, Mem. Torrey Club 5: 265. 1894. Odontostephana carolinensis (Jacq.) Alexander, in Small, Man. FI. 1077. 1933. Marevea cynanchoides (Engelm.) Woodson, comb. nov. Gonolobus cynanchoides Engelm. in Engelm. & Gray, Bos- ton Journ. Nat. Hist. 5: 251. 1845. Vincetoxicum cynanchoides (Kngelm.) Heller, Muhlen- bergia 1: 2. 1900. Mar&ELEA decipiens (Alexander) Woodson, comb. nov. Odontostephana decipiens Alexander, in Small, Man. FI. 1077. 1933. Gonolobus decipiens (Alexander) Perry, Rhodora 40: 286. 1938. Martera flavidula ( Chapm.) Woodson, comb. nov. Gonolobus flavidulus Chapm. Bot. Gaz. 3: 19. 1878. 1941] WOODSON—NORTH AMERICAN ASCLEPIADACEAE. I 229 Gonolobus hirsutus var. flavidulus (Chapm.) A.Gray, Syn. Fl. ed.2, 21: 404. 1886. Vincetoxicum flavidulum (Chapm.) Heller, Muhlenbergia 1: 2. 1900. Odontostephana flavidula (Chapm.) Alexander, in Small, Man. Fl. 1078. 1933. МАтЕгЕА floridana (Vail) Woodson, comb. nov. Vincetoxicum floridanum Vail, Bull. Torrey Club 26: 428. 1899. Odontostephana floridana (Vail) Alexander, in Small, Man. Fl. 1078. 1933. Marera Greggii (Vail) Woodson, comb. nov. Vincetoxicum Gregg Vail, Bull. Torrey Club 26: 431. 1899. MarELEA LeSueurii (Standl.) Woodson, comb. nov. Vincetoxicum LeSueuri Standl. Field Mus. Publ. Bot. 17: 270. 1937. The pollinia of this species are not the subquadrate-reniform sort typical of Chthamalia, but are more nearly triangular- pyriform. Nevertheless, the affinities of M. Гебиеиги unmis- takably are with that subgenus, particularly with M. Num- mularia. MarELEA Nummularia (Dene.) Woodson, comb. nov. Chthamalia Nummularia Dene. in DC. Prodr. 8: 605. 1844. Marteka obliqua (Jacq.) Woodson, comb. nov. Cynanchum obliquum Jacq. Coll. 1: 148. 1786. Gonolobus obliquus (Jaeq.) R.Br. Mem. Wern. Soe. 1: 35. 1809. Vincetoxicum obliquum (Jacq.) Britton, Mem. Torrey Club 5: 266. 1894. Odontostephana obliqua (Jacq.) Alexander, in Small, Man. Fl. 1077. 1933. MarELEA parviflora (Torr.) Woodson, comb. nov. Lachnostoma (?) parviflorum Torr. Rept. Bot. Mex. Bound. Surv. 165. 1859. Gonolobus parviflorus (Torr.) A.Gray, Proc. Amer. Acad. 12: 79. 1877. [Vor. 28 230 ANNALS OF THE MISSOURI BOTANICAL GARDEN Vincetoxicum parviflorum (Torr.) Heller, Muhlenbergia 1: 2. 1900. МАтЕгЕА parvifolia (Torr.) Woodson, comb. nov. Gonolobus parvifolius Torr. Rept. Bot. Mex. Bound. Surv. 166. 1859. Gonolobus hastulatus A.Gray, Proc. Amer. Acad. 12: 78. 1877. Vincetoxicum hastulatum (A.Gray) Heller, Muhlenbergia 1: 2. 1900. Gonolobus californicus Jepson, Man. 771. 1925. Млтегел pedunculata (Dene.) Woodson, comb. nov. Chthamalia pedunculata Dene. in DC. Prodr. 8: 605. 1844. MarELEA producta (Torr.) Woodson, comb. nov. Gonolobus productus Torr. Rept. Bot. Mex. Bound. Surv. 165. 1859. Vincetoxicum productum (Torr.) Vail, Bull. Torrey Club 26: 431. 1899. MaTELEA prostrata ( Willd.) Woodson, comb. nov. Cynanchum prostratum Willd. Sp. Pl. 1: 1257. 1798. Gonolobus prostratus (Willd.) R.Br. Mem. Wern. Soc. 1: 35. 1809. Lachnostoma prostratum (Willd.) Dene. in DC. Prodr. 8: 602. 1844. Tetracustelma prostrata (Willd.) Baill. Hist. Pl. 10: 292. 1891. Матетвл pubiflora (Dene.) Woodson, comb. nov. Chthamalia pubiflora Dene. in DC. Prodr. 8: 605. 1844. Gonolobus pubiflorus (Dene.) Engelm. in Engelm. & Gray, Boston Journ. Nat. Hist. 5: 252. 1845. Vincetoxicum pubiflorum (Dene.) Heller, Muhlenbergia 1: 2. 1900. Edisonia pubiflora (Dene.) Small, Man. Fl. 1078. 1933. MarELEA Schaffneri (A.Gray) Woodson, comb. nov. Gonolobus Schaffnert A.Gray, in Hemsl. Biol. Centr.-Am. Bot. 2: 334. 1882; A.Gray, Proc. Amer. Acad. 21: 399. 1886. Gonolobus bifidus Hemsl. Biol. Centr.-Amer. Bot. 2: 330. 1882. 1941] WOODSON—NORTH AMERICAN ASCLEPIADACEAE. I 231 МАтЕгЕА Shortii (A.Gray) Woodson, comb. nov. Gonolobus obliquus var. Shortw A.Gray, Syn. Fl. 2!: 104. 1878. Vincetoxicum Shortw (A.Gray) Britton, Mem. Torrey Club 5: 266. 1894. Gonolobus Short A.Gray, Bot. Gaz. 8: 191. 1883. Odontostephana Short (A.Gray) Alexander, in Small, Man. Fl. 1077. 1933. MaTELEA Stenopetala ( A.Gray) Woodson, comb. nov. Gonolobus stenopetalus A.Gray, Proc. Amer. Acad. 21: 398. 1886. Vincetoxicum stenopetalum (A.Gray) Standl. Contr. U.S. Nat. Herb. 23: 1190. 1924. MarELEA Vailiana Woodson, nom. nov. Gonolobus acuminatus A.Gray, Proc. Amer. Acad. 21: 399. 1886, non Matelea acuminata (Griseb.) Woodson. Vincetoxicum acuminatum (A.Gray) Vail, Bull. Torrey Club 26: 431. 1899. MarELEA Wootonii (Vail) Woodson, comb. nov. Vincetoxicum Wootonii Vail, Bull. Torrey Club 28: 485. pl.30. 1901. This is the characteristic group of Matelea in northern Mexico and the southern United States. The pollinia are well illustrated for the species ‘‘in the Manual Range’’ by Miss Perry (Rhodora 40: pl.494. 1938) where the contrasting fea- tures of true Gonolobus pollinia are shown also. I believe that Chthamalia is perhaps most closely related to Ibatia amongst the subgenera of Matelea, the ‘‘pendulous’’ pollinia of the latter frequently appearing much like the ‘‘horizontal’’ pol- linia of the former (merely of greater vertical dimension). The corona of the two subgenera, also, are much alike, but the internal ligules of the former are merely adnate (sometimes nearly free) to the outer segments, whereas the fusion is virtu- ally complete in the latter. SUBGENUS 8. PHEROTRICHIS MarmELEA Balbisii (Dene.) Woodson, comb. nov. [Уог. 28 232 ANNALS OF THE MISSOURI BOTANICAL GARDEN Asclepias villosa Balb. Mem. Accad. Sci. Torino 7: 386. 1803, non Mill. Cynanchum villosum (Balb.) R. & S. Syst. 6: 103. 1819. Lachnostoma Вази Dene. in DC. Prodr. 8: 602. 1844. Gonolobus pogonanthus Hemsl. Biol. Centr.-Am. Bot. 2: 333. 1882. | Pherotrichis Balbisu (Dene.) A.Gray, Proc. Amer. Acad. 21: 400. 1886; Syn. Fl. ed.2. 21: 462. 1886. Pherotrichis Schaffneri A.Gray, loc. cit. 1886. MarELEA chihuahuensis (A.Gray) Woodson, comb. nov. Gonolobus chihuahuensis A.Gray, Proc. Amer. Acad. 21: 398. 1886. Vincetoxicum chihuahuense (A.Gray) Standl. Field Mus. Publ. Bot. 17: 270. 1937. Mare ea leptogenia (Robins.) Woodson, comb. nov. Pherotrichis leptogenia Robins. Proc. Amer. Acad. 29: 319. 1894. SUBGENUS 9. AMPHORELLA MarELEA castanea (Brandg.) Woodson, comb. nov. Amphorella castanea Brandg. Univ. Calif. Publ. Bot. 4: 92. 1910. SUBGENUS 10. TIARASTEMMA MarEL£A calcarata Woodson, comb. nov. Vincetoxicum calcaratum Woodson, Amer. Journ. Bot. 22: 689. pl.1. д.8. 1935. MarELEA belizensis (Lundell & Standl.) Woodson, comb. nov. Vincetoxicum belizense Lundell & Standl. in Standl. Field Mus. Publ. Bot. 17: 268. 1937. SUBGENUS 11. POLYSTEMMA MarELEA Decaisnei Woodson, nom. nov. Polystemma viridiflora Dene. in DC. Prodr. 8: 602. 1844, non М. viridiflora (G.F.W.Mey.) Woodson. Mareea rupestris (Brandg.) Woodson, comb. nov. Polystemma rupestre Brandg. Univ. Cal. Publ. Bot. 7: 330. 1920. MarELEA scopulorum (Brandg.) Woodson, comb. nov. 1941] WOODSON—NORTH AMERICAN ASCLEPIADACEAE. I 233 Polystemma scopulorum Brandg. Univ. Calif. Publ. Bot. 6: 189. 1915. SUBGENUS 12. HELIOSTEMMA Marteka caudata (A.Gray) Woodson, comb. nov. Gonolobus caudatus A.Gray, Proc. Amer. Acad. 21: 399. 1886. Vincetoxicum caudatum (A.Gray) Standl. Contr. U.S. Nat. Herb. 23: 1190. 1924. Marteka crenata (Vail) Woodson, comb. nov. Vincetoxicum crenatum Vail, Bull. Torrey Club 26: 429. 1899. This species might almost as well be placed in Гфайа upon the basis of the pollinia alone, for they are only slightly falei- form. The corona segments, also, are scarcely typical of Нейо- stemma in that they are not strongly adnate to the corolla throat. However, they do have a conspicuous internal boss which is characteristic of the latter subgenus and never found in the former. МАТЕА cyclophylla (Standl.) Woodson, comb. nov. Vincetoxicum cyclophyllum Standl. Contr. U.S. Nat. Herb. 23: 1191. 1924. Мате ка nigrescens (Schlecht.) Woodson, comb. nov. Gonolobus nigrescens Schlecht. Linnaea 8: 522. 1833. Vincetoxicum nigrescens (Schlecht.) Standl. Contr. U.S. Nat. Herb. 23: 1192. 1924. MarErEA picturata (Hemsl.) Woodson, comb. nov. Gonolobus picturatus Hemsl. Biol. Centr.-Am. Bot. 2: 332. 1882. Maretea pilosa (Benth.) Woodson, comb. nov. Gonolobus pilosus Benth. Pl. Hartw. 289. 1848. Gonolobus atratus A.Gray, Proc. Amer. Acad. 22: 436. 1887. Gonolobus suberiferus Robins. Proc. Amer. Acad. 27: 181. 1892. Vincetoxicum Grayanum Standl. Contr. U.S. Nat. Herb. 23: 1191. 1924. [Vor. 28 234 ANNALS OF THE MISSOURI BOTANICAL GARDEN Vincetoxicum pilosum (Benth.) Standl. loc. cit. 1924. Vincetoxicum suberiferum (Robins.) Standl. loc. cit. 1189. 1924. Vincetoxicum quercetorum Standl. Field Mus. Publ. Bot. 17: 271. 1937. Матеткл Pittieri (Standl.) Woodson, comb. nov. Vincetoxicum Pittieri Standl. Field Mus. Publ. Bot. 17: 270. 1937. Vincetoxicum discolor Woodson, Ann. Missouri Bot. Gard. 20: 833. fig.1. 1938. SUBGENUS 13. EUMATELEA $. Reticulatae MarzrzaA alabamensis (Vail) Woodson, comb. nov. Vincetoxicum alabamense Vail, Bull. Torrey Club 30: 178. pl.9. 1903. Матегкл campechiana (Standl.) Woodson, comb. nov. Vincetoxicum campechianum Standl. Carnegie Inst. Wash. Publ. 461: 82. 1935. Матегжл Gentlei (Lundell & Standl.) Woodson, comb. nov. Vincetoxicum Gentlei Lundell & Standl. Field Mus. Publ. Bot. 17: 269. 1937. Млтегел lanceolata (Dene.) Woodson, comb. nov. Gonolobus lanceolatus Dene. in DC. Prodr. 8: 598. 1844. Vincetoxicum stenophyllum Standl. Contr. U.S. Nat. Herb. 23: 1188. 1924. Marerra reticulata (Engelm.) Woodson, comb. nov. Gonolobus reticulatus Engelm. ex A.Gray, Proc. Amer. Acad. 12: 75. 1877. Vincetoxicum reticulatum (Engelm.) Heller, Bot. Expl. Texas, 79. 1895. MarELEA Tuerckheimii (Donn.Sm.) Woodson, comb. nov. Gonolobus Tuerckheimii Donn. Sm. Bot. Gaz. 44: 116. 1907. Vincetoxicum trichoneuron Standl. ex Yuncker, Field Mus. Publ. Bot. 17: 388. 1938. MarzrzA velutina (Schlecht.) Woodson, comb. nov. Gonolobus velutinus Schlecht. Linnaea 8: 521. 1833. 1941] WOODSON—NORTH AMERICAN ASCLEPIADACEAE. I 235 Vincetoxicum velutinum (Schlecht.) Standl. Contr. U.S. Nat. Herb. 23: 1189. 1924. $. Pseudobarbatae МАтТЕГЕА guatemalensis (K.Sch.) Woodson, comb. nov. Gonolobus velutinus var. calycinus Donn.Sm. Bot. Gaz. 13: 189. 1888. Gonolobus guatemalensis K.Sch. in Engl. & Prantl, Nat. Pflanzenfam. 42: 302. 1895. MarErza pinguifolia (Standl.) Woodson, comb. nov. Vincetoxicum pinguifolium Standl. Journ. Wash. Acad. Sci. 17: 13. 1927. Млтегел pseudobarbata (Pittier) Woodson, comb. nov. Gonolobus pseudobarbatus Pittier, Contr. U.S. Nat. Herb. 13: 105. fig.14. 1910. Matelea nigrescens shows a strong affinity with this section, but has been placed in Heliostemma because of its falciform pollinia. §. Viridiflorae MarErzEA grandiflora (Standl.) Woodson, comb. nov. Vincetoxicum grandiflorum Standl. Carnegie Inst. Wash. Publ. 461: 83. 1935. MarELEA viridiflora (G.F.W.Mey.) Woodson, comb. nov. Cynanchum viridiflorum G.F.W.Mey. Prim. Fl. Esseq. 141. 1818. Gonolobus viridiflorus (G.F.W.Mey.) R. & S. Syst. 6: 61. 1819. Vincetoxicum viridiflorum (G.F.W.Mey.) Standl. Contr. U.S. Nat. Herb. 27: 311. 1928. This little group illustrates the maddening complications that result from attempts at segregation in the Gonolobeae. M. viridiflora and M. grandiflora quite obviously are very closely related, as evidenced particularly by the interrupted eorolline annulus. Other characters, however, point strongly to relationship with other groups of Matelea, with the excep- tion of the pollinia, which are scarcely similar to those of Eumatelea, and can be construed as ‘‘triangular-pyriform”’ [Vor. 28 236 ANNALS OF THE MISSOURI BOTANICAL GARDEN only with a considerable streteh of the imagination. But on the other hand, the pollinia of the two species are so dissimilar that I can find no descriptive term to eover both, and at the same time to distinguish them from Humatelea. SUBGENUS 14. MICRODACTYLON MarELEA cordata (Brandg.) Woodson, comb. nov. Microdactylon cordatum Brandg. Zoe 5: 252. 1908. SUBGENUS 15. PACHYSTELMA MaTELEA bicolor (Britton & Wilson) Woodson, comb. nov. Marsdenia bicolor Britton & Wilson, Bull. Torrey Club 50: 47. 1923. Gonolobus bicolor (Britton & Wilson) Urb. Symb. Ant. 9: 421. 1925. MaTELEA crassifolia (Standl.) Woodson, comb. nov. Vincetoxicum crassifolium Standl. Field Mus. Publ. Bot. 8: 36. 1930. MarELEA megacarpha (Brandg.) Woodson, comb. nov. Vincetoxicum megacarphum Brandg. Univ. Cal. Publ. Bot. 4: 381. 1913. Pachystelma cordatum Brande. loc. cit. 7: 330. 1920. Dictyanthus brachistanthus Standl. Field Mus. Publ. Bot. 8: 38. 1930. SUBGENUS 16. DICTYANTHUS MarzrzA altatensis (Brandg.) Woodson, comb. nov. Gonolobus altatensis Brandg. Zoe 5: 244. 1908. Матегвл ceratopetala (Donn.Sm.) Woodson, comb. nov. Dictyanthus ceratopetalus Donn.Sm. Bot. Gaz. 18: 208. 1893. MATELEA dictyantha Woodson, nom. nov. Rytidoloma reticulata Turez. Bull. Soc. Nat. Mose. 25?: 320. 1852, non Matelea reticulata (Kngelm.) Woodson. Dictyanthus reticulatus (Turez.) Benth. & Hook. Gen. PI. 2: 765. 1876. MarELEA diffusa Woodson, nom. nov. Dictyanthus prostratus Brandg. Univ. Cal. Publ. Bot. 7: 329. 1920, non Matelea prostrata (Willd.) Woodson. 1941] WOODSON—NORTH AMERICAN ASCLEPIADACEAE. I 231 MarrELEA Hemsleyana Woodson, nom. nov. Dictyanthus parviflorus Hemsl. Biol. Centr. Am. Bot. 2: 329. 1882, non Matelea parviflora (Torr.) Woodson. MarErEA Pavonii (Dene.) Woodson, comb. nov. Dictyanthus Pavoni Dene. in DC. Prodr. 8: 605. 1844. Tympananthe suberosa Hassk. Flora 30: 758. 1847. MarELEA Standleyana Woodson, nom. nov. Dictyanthus tigrinus Conzatti & Standl. in Standl. Contr. U.S. Nat. Herb. 23: 1183. 1924, non Matelea tigrina (Griseb.) Woodson. MarELEA stapeliaeflora (Reichb.) Woodson, comb. nov. Dictyanthus stapeliaeflora Reichb. Sel. Sem. Hort. Dres- den, 4. 1850. MarELEA tuberosa (Robins.) Woodson, comb. nov. Dictyanthus tuberosus Robins. Proc. Amer. Acad. 27: 180. 1892. MarELEA yucatanensis (Standl.) Woodson, comb. nov. Dictyanthus yucatanensis Standl. Field Mus. Publ. Bot. 8: 37. 1930. Dictyanthus aeneus Woodson, Amer. Journ. Bot. 22: 691. 1935. I am very reluctant to merge Dictyanthus with Matelea, since the various species here show considerably more coher- ence and distinction than in other generic groups. However, generic lines cannot be drawn sharply due to such groups as Pachystelma, Labidostelma, and Macroscepis. The only really unique feature of Dictyanthus is that the faucal callus, or an- nulus, of the corolla is digitate, as are the segments of the corona. I am prepared for vigorous opposition from those who are accustomed to think of Dictyanthus as exemplified by D. Pavonu. М. altatensis provides a very suggestive link with either Pachystelma or Eumatelea, according to one's view- point at a particular time. УП. Fiscuerta DC. Cat. Hort. Monsp. 112. 1813. I am maintaining Fischeria with misgivings, and only pro- visionally, since its affinities with the broad interpretation of [Vor. 28 238 ANNALS OF THE MISSOURI BOTANICAL GARDEN Matelea are unmistakable, as shown particularly in the pol- linium structure and the mixed indument with interspersed bulbose, glandular hairs. In the Central American species the corona appears outstanding, but is scarcely different from that of Eumatelea $ Pseudobarbatae. The anthers, also, show the inflated vesicular development mentioned in the key to genera. But in F. viridis Moldenke, recently discovered in Colombia, the corona is resolved into a more typical structure for Matelea, and the anthers are not vesieular. Should more species comparable to Е. viridis appear in Ше rich South Amer- ican flora, it is difficult to foresee what characters could be found to support the genus other than the crisped corolla lobes. Since Fischeria is not well represented in North America, Ше species there being quite easily distinguished, I am leaving the question of its validity until I am more familiar with the South American congeners. VIII. Сохоговов Michx. Fl. Bor. Am. 1: 119. 1803. Vincetoxicum Walt. Fl. Carol. 13, 104. 1788, non Moench. Fimbristemma Turez. Bull. Soc. Nat. Mose. 25?: 320. 1852. Exolobus Fourn. in Mart. Fl. Bras. 64: 318. 1885. Trichostelma Baill. Hist. Pl. 10: 287. 1891. Lachnostoma of North American authors. The complications concerning the use of Gonolobus Michx. and Vincetoxicum Walt., the ‘‘temporary’’ conservation of Vincetoxicum Moench. over the latter, and the typification of Gonolobus by G. gonocarpos (Walt.) Perry have been dis- cussed in previous paragraphs (p. 213). It may be remem- bered that the original element of Vincetoxicum Walt. (or Gonolobus, through the rejection of the earlier name) consists of two species, G. gonocarpos (Walt.) Perry, the type, and G. carolinensis (Jaeq.) Schultes (V. acanthocarpos Walt.). Until recently no question has been raised against the con- sideration of these two species as congeneric, and Gonolobus became the general catch-all for gonoloboid species without sufficiently striking innovations of the corona to prompt their 1941] WOODSON—NORTH AMERICAN ASCLEPIADACEAE, I 239 designation as separate genera. At about the opening of the present century, however, the earlier Vincetoxicum Walt. was revived by Dr. Britton and Miss Vail at the New York Botani- eal Garden. Transfers from Gonolobus were then in order, and few species of the complex have been described for that genus since. It seems rather severe that American botanists now will have to adjust themselves anew to the conservation of the dubiously valid Vincetoxicum Moench of Europe. In 1933 (in Small, Man. Fl. 1076.) E. J. Alexander separated G. gonocarpos and G. carolinensis generically, founding the genus Odontostephana with the latter, together with other species from the southeastern United States. Alexander quite naturally used Vincetoxicum Walt. in place of Gonolobus, the judgment of the authorities at Kew not yet having been passed. The key characters used to separate Vincetoxicum from Odontostephana were taken from the coronas, whether ** Crown disk-like or saucer shaped, obscurely if at all ridged,” or ‘‘ Crown cup-shaped or incurved at the tip, usually crested or appendaged on the inner side.’’ In the descriptions of the two genera, attention was called also to the characters of the fruits, longitudinally winged in the former, and muricate in the latter, as observed by Walter in 1788. It is regrettable that Odonto- stephana is antedated by Tetracustelma Baill. if a narrow generic concept is adopted, and that a broad concept will sub- merge it in Matelea Aubl. I have tried in vain to find characters of the corona which will separate Gonolobus and Matelea, but the structures of the anthers of the two genera appear to me as amply sufficient for distinction. The anthers of both vary considerably in size, po- sition relative to the anther head, hyaline apical appendage, and nature of the marginal ‘‘wings.’’ But in the species that I have assigned to Gonolobus the anther proper bears a more or less conspicuous, fleshy, usually laminate dorsal appendage which I never have found indicated to any degree in species assigned to Matelea. These dorsal appendages vary greatly from species to species, and although rather poorly developed as a rule in the temperate representatives, are very conspicu- [Vor. 28 240 ANNALS OF THE MISSOURI BOTANICAL GARDEN ous in the tropies. Typically, they are somewhat reniform in outline, but frequently deeply 2-lobed or entire, and rarely cleft into three divisions. The dorsal anther appendages are indicated in Alexander's figure for Vincetoxicum Walt. (loc. cit. 1933), although their significance is not mentioned, and a splendid illustration of them is provided by Baillon (Hist. Pl. 10: 235. fig.177. 1891) for G. erianthus. The dorsal anther appendages of Gonolobus have been over- looked in most species of North America, but they were made the chief foundation for the genus Exolobus by Fournier (cf. Mart. Fl. Bras. 6*: pl.94. 1885), who designated them as “corona superior." The species assigned to Gonolobus by Fournier all are of the element represented by G. hirsutus Michx. which I am relegating to Matelea. K. Schumann (in Engl. & Prantl, Nat. Pflanzenfam. 42: 298, 301-302. 1895) fol- lowed Fournier in the separation of Gonolobus and Exolobus, but appears to have been remarkably confused, for the dorsal anther appendages of my terminology are described as ‘‘inner corona” (‘‘corona . . . innere an den A. befestigt, nach aus- sen strahlend, fleischig.’’) for the former, but for the latter simply as appendages of the anthers (‘‘Stb. mit einem mitt- leren, fleischigen, nach aussen gerichteten Fortsatze. . .’’). As I have explained previously (p. 199), flowers of Go- nolobus as interpreted in this paper customarily have three structures which may be called ‘‘corona’’ if the terminology is loose: a corolline faucal annulus, frequently ciliate, a fleshy true corona borne at the base of the staminal column, and the dorsal appendages of the anthers proper. In his key to the Gonolobeae, Schumann (loc. cit. 298) places both Gonolobus and Ехоюфиз under the division ‘‘II. Corona doppelt,’’ and distinguishes them by ‘‘1. Aussere Corona kahl’’ and “2. Aus- sere Corona gewimpert,’’ respectively, apparently referring to the corolline annulus. In his descriptions of the two genera, however, Schumann reveals by his ambiguity how confused was his concept: ** Corona ringfórmig, einfach oder doppelt, die àussere zuweilen noch von einem hautigen Rande umgeben, so dass man fast 3 1941] s WOODSON— NORTH AMERICAN ASCLEPIADACEAE. I 241 Coronen unterscheiden kann, oft ringförmig, der Blkr. an- geheftet, kurz, gestutzt oder gelappt, zuweilen durch 5 Gewebe- platten mit dem Gynostegium verbunden; innere an den A. befestigt, nach aussen strahlend, fleischig’’ (Gonolobus, p. 301) ; **Corona doppelt: üussere in der Form eines sehr nied- rigen, behaarten, aus den Blkr. vortretenden Ringes, die innere aus 5 freien Schuppen bestehend, welche der Róhre des Gynostegiums angeheftet sind. Stb. mit einem mittleren, fleischigen, nach aussen gerichteten Fortsatze und einem háutigen Mittelbandanhang’’ ( Exolobus, p.302). The coup de grace is given these verbal circumlocutions by a consideration of fig. 92, N-T, provided by Schumann (p.301) to illustrate Gonolobus and Exolobus. The floral structures are found to be equivalent in all important particulars, even to a ‘‘gewimpert corona" (ciliate corolline annulus) for G. riparius. The extended quotations from Schumann should be suff- cient to illustrate the ambiguity that results from treating the term “согопа” in a complex sense. The dorsal anther ap- pendages of Gonolobus are not equivalent to any structure found in Matelea. At one time during these studies, I suspected that they might represent the adnation to the anther of the inner ligules commonly accompanying the corona segments of the latter genus. This is shown to be erroneous, however, by the fact that these ligules almost invariably occur in Gonol- obus in addition to the dorsal anther appendages, which ap- parently are more directly comparable to the vesicular tissue of the anthers of Fischeria. I am not yet sufficiently informed to know positively what will result when this character is applied extensively to the South American gonoloboids, but wide observations indicate its validity. Supporting charaeters are found amongst the relatively few species of Gonolobus and Matelea for which fruit is known, the follicles of the former being longitudinally winged, infrequently quite smooth, and those of the latter muricate or infrequently smooth. The smooth condition would appear to be derived from both series. In Matelea, as has been explained previously, a peculiar indument of eglandular and [Vor. 28 242 ANNALS OF THE MISSOURI BOTANICAL GARDEN interspersed glandular hairs characteristically occurs; such an indument never is found in Gonolobus. Amongst the North American species of Gonolobus with which I am familiar at present, the following divisions into subgenera appear both natural and convenient. As usual, I have illustrated them with necessary transfers. KEY ro SuUBGENERA A. Corolla with widely spreading or reflexed — with a more or less pro- nouneed faueal annulus exterior to the tru B. Anthers only about half included beneath p rie 5-gonal stigma BÉ. urere Te TS Ee әт КЕККЕ Ул бск Е 1. PTEROLOBUS BB. Кыл wholly (ехсер Le dorsal appendages) Те beneath the saliently 5-gonal stigma Һеай....................... . EUGONOLOBUS AA. Corolla with erect or Sara m lobes, without a faueal а TEE PES TREE Co ECKE CE TRAP о 3. PSEUDOLACHNOSTOMA I should like to include T'richostelma Baill. in this key either as a subgenus or as а section under Eugonolobus, but the height of the corolline annulus appears to be far too variable. SUBGENUS 1. PTEROLOBUS Сокоговов chiapensis (Brandg.) Woodson, comb. nov. Vincetoxicum chiapense Brandg. Univ. Cal. Publ. Bot. 6: 190. 1915 In addition to the preceding, Pterolobus includes such species as G. niger R.Br., G. Salvinii Hemsl., and G. nemorosus Dene. The name refers to the undulated wings of the follicles of the few species where I have observed them. SUBGENUS 2. EUGONOLOBUS GowoLonvus albomarginatus (Pittier) Woodson, comb. nov. Exolobus albomarginatus Pittier, Contr. U.S. Nat. Herb. 13: 108. fig.16. 1910. Сохоговоѕ aristolochiaefolius (Brandg.) Woodson, comb. nov. Fischeria aristolochiaefolia Brandg. Univ. Cal. Publ. Bot. 6: 190. 1915. бо Pann calycosus (Donn.Sm.) Woodson, comb. nov. Trichostelma ciliatum Baill. Hist. Pl. 10: 288. 1891, non G. ciliatus Schltr. Damen "TENSILE 1941] WOODSON—NORTH AMERICAN ASCLEPIADACEAE. I 243 Fimbristemma calycosa Donn.Sm. Bot. Gaz. 16: 196. pl. 16. 1891. Сохоговоѕ cteniophorus (Blake) Woodson, comb. nov. Vincetoxicum cteniophorum Blake, Contr. Gray Herb. 52: 84. 1917. Vincetoxicum ? tortum Brandg. Univ. Cal. Publ. Bot. 10: 414. 1924. Vincetoxicum Lundellii Standl. Field Mus. Publ. Bot. 8: 148. 1930. ; Сохотовов dasystephanus (Blake) Woodson, comb. nov. Vincetoxicum dasystephanum Blake, Contr. Gray Herb. 02: 84. 1917. Сомотовоз Lasiostemma (Hemsl.) Woodson, comb. nov. Lachnostoma Lasiostemma Hemsl. Biol. Centr.-Am. Bot. 2: 335. 1882. Gono.Losus oblongifolius (Donn.Sm.) Woodson, comb. nov. Trichostelma oblongifolium Donn.Sm. Bot. Gaz. 48: 296. 1909. Сохоговоѕ stenanthus (Standl.) Woodson, comb. nov. Vincetoxicum stenanthum Standl. Field Mus. Publ. Bot. 4: 255. 1929. Gono.osus stenosepalus (Donn.Sm.) Woodson, comb. nov. Fimbristemma stenosepala Donn.Sm. Bot. Gaz. 18: 208. 1893. SUBGENUS 3. PSEUDOLACHNOSTOMA Gono.osus arizonicus (A.Gray) Woodson, comb. nov. Lachnostoma arizonicum A.Gray, Proc. Amer. Acad. 20: 296. 1885. Gono tosus gonoloboides (Greenm.) Woodson, comb. nov. Lachnostoma gonoloboides Greenm. Proc. Amer. Acad. 39: 84. 1903. IX. Marspenta R.Br. Mem. Wern. Soc. 1: 28. 1809. Nephradenia Dene. in DC. Prodr. 8: 604. 1844. Ecliptostelma Brandg. Univ. Cal. Publ. Bot. 6: 371. 1917. I have been unable to find characters to distinguish Neph- radenia except the erect, fruticose habit. Ecliptostelma molle [Vor. 28, 1941] 244 ANNALS OF THE MISSOURI BOTANICAL GARDEN Brandg. (loc. cit. 1917) is identical with Marsdenia Gilgiana Rothe, which seems to be an extraordinarily variable species, even from the present meager representation, and which pos- sibly should be merged with the very closely neighboring M. mexicana Dene. For further generic synonyms, reference should be made to the monograph by Rothe (in Engl. Bot. Jahrb. 52: 354—434. 1915). Following that standard work, new seetions or subgenera probably will have to be erected to include both the following species, although I am deferring that action for the present. Marspenia astephanoides (A.Gray) Woodson, comb. nov. Vincetoxicum astephanoides A.Gray, Proc. Amer. Acad. 22: 435. 1887. Cynanchum astephanoides (A.Gray) Standl. Contr. U.S. Nat. Herb. 23: 1177. 1924. Marspenta neriifolia (Dene.) Woodson, comb. nov. Blepharodon meriüfoliwm Dene. in DC. Prodr. 8: 604. 1844. Nephradenia neriifolia (Dene.) Benth. & Hook.; Hemsl. Biol. Centr. Amer. Bot. 2: 336. 1882. Nephradenia fruticosa Donn. Sm. Bot. Gaz. 16: 196. 1891. TWO NEW ASCLEPIADS FROM THE SOUTHWESTERN UNITED STATES BASSETT MAGUIRE Associate Professor of Botany, Utah State Agricultural College ROBERT E. WOODSON, JR. Assistant Professor in the Henry Shaw School of Botany of Washington University Assistant Curator of the Herbarium, Missouri Botanical Garden AscLEPIAS Ruthiae Maguire, spec. nov. Herbae perennes; caudices multicipitales, caulibus nonnullis simplicibus sub- erectis 1.0-1.5 dm. altis erispo-pilosulis ; folia opposita vel ap- proximata late ovata apice aut acuminata aut abrupte mu- eronulata basi obtusa 2.2-4.5 em. longa 1.5(1.0)-2.3 ст. lata, lamina sparse margine densius pilosula, petiolis 0.2-0.5 cm. longis minute pilosulis; inflorescentia aut terminalis aut lat- eralis umbelliformis pauciflora (2-7) sessilis, pedicellis fili- formibus 1.5-2.5 em. longis pilosulis; calycis lobi lanceolati acuti 0.2-0.3 em. longi pilosuli violacei; corolla rotata pallide cano-violacea extus minute puberulo-papillata, lobis ovatis ob- tusis ca. 0.4 ст. longis patulis; staminum columna са. 0.1 cm. alta viridula inter foliolas coronae saccata, antherae purpureae 0.15-0.23 cm. longae apicibus seariaceis obtusis integris; pol- linia compresse elongato-pyriformia ca. 0.05 cm. longa; co- ronae foliolae saccatae са. 0.2 ст. longae lobo dorso obtuso lobulis lateralibus obtusis longiore corniculo incluso roseo- purpureae ; folliculi erecti faleate ovato-fusiformes 3-4(5) cm. longi dense pilosuli vel subglabrati. Uram: frequent, sandy soil, vie. water tank, Calf Springs Canyon, alt. 5000 ft., San Rafael Swell, 18 mi. southeast of Castle Dale, Emery Co., May 10, 1940, Ruth £ Bassett Maguire 18310 (Herb. Utah State Agricultural College, Tyre; Herb. Missouri Bot. Garden, ІѕотүрЕ); Calf Spring Canyon, San Rafael Swell, June 5, 1940, B. Maguire 18423 (Herb. Utah State Agr. Coll.; Herb. Missouri Bot. Gard.) ; same locality, June 7, 1940, B. Maguire 18477 (Herb. Utah State Agr. Coll.; Issued April 28, 1941. (245) [Vor. 28 246 ANNALS OF THE MISSOURI BOTANICAL GARDEN Herb. Missouri Bot. Gard.); in the desert, 20 mi. south of Green River, June 8, 1939, C. L. Porter 1799 (Rocky Mtm. Herb., Univ. Wyoming). This is another member of the complex represented by A. brachystephana, A. uncialis, and A. Cutleri (cf. Woodson, Ann. Missouri Bot. Gard. 26: 262. 1939), but is distinguished by its broadly ovate leaves and short lateral hood lobules. It Fig. 1. Asclepias Ruthiae Maguire. Flower, pollinia, and hood in longitudinal section. gives me much pleasure to dedicate it to my wife, Ruth R. Maguire, who was its discoverer and who so frequently is my helpful companion in the field. Ascuepias Sperryi Woodson, spec. nov. Herba suffrutescens ramosissima ca. 1.5 dm. alta habitu A. macrotem simulans; caules gracillimi suberecti compositi minutissime pilosuli vel glabrati, internodiis са. 1.5-2.0 em. longis; folia linearia 1-5 em. longa ca. 0.05 em. lata sessilia glabra; flores solitarii lat- erales, pedicellis са. 1 em. longis glabris; calycis lobi ovato- nose - 1941] MAGUIRE & WOODSON—TWO NEW ASCLEPIADS 247 laneeolati acuti 0.2-0.25 em. longi papillati; corolla dilute viridula plus minusve purpureo-tincta, lobi ovato-elliptici acuti 0.5-0.6 em. longi patuli; staminum columna vix manifesta, co- rona sessili; gynostegium ca. 0.3 cm. altum, antheris 0.1 cm. longis apicibus scariaceis inflexis, stigmate depresso ca. 0.25 em. diam.; coronae foliolae habitu floseulo Aristolochiae similes gilvae vel dilute roseae dense papillatae parte basali patula ca. 0.2-0.3 em. longa involuta margine alata corniculo umbonato incluso, parte superiori erecta laminata apice 3- ` ON Fig. 2. Asclepias Sperryi Woodson. Flower, pollinia, and hood in longitudinal section. lobulata; folliculi deflexi anguste fusiformes ca. 6 cm. longi 0.5 em. erassi laeves glabri. Texas: east of Nichols Ranch House, Glass Mountains, Brewster County, Aug. 1, 1938, C. La Motte, O. E. Sperry, & B. H. Warnock T553 (Herb. Missouri Bot. Garden, ТУРЕ; Herb. Sperry, 1soryPe) ; about 10 miles east of Alpine, Brew- ster County, May 21, 1936, Sperry T301 (Herb. Missouri Bot. Garden; Herb. Sperry). The superficial habit of this species is almost identical to that of A. macrotis, but the 1-flowered cymes and peculiar form [Vor. 28, 1941] 248 ANNALS OF THE MISSOURI BOTANICAL GARDEN of the corona hoods are quite exceptional in the genus as repre- sented in North America. It has been named for Professor Omer E. Sperry because of his activity in the botany of west- ern Texas and in appreciation for his aid to my study of Asclepias. Annals of the Missouri Botanical Garden Vol. 28 SEPTEMBER, 1941 No. 3 A PRELIMINARY SURVEY OF THE GENUS TRIPSACUM HUGH C. CUTLER Research Fellow in the Henry Shaw School of Botany of Washington University AND EDGAR ANDERSON Geneticist to the Missouri Botanical Garden Engelmann Professor in the Henry Shaw School of Botany of Washington University When Mangelsdorf and Reeves! demonstrated in 1931 that Zea could be hybridized with Tripsacum, it became evident that a detailed monograph of the latter genus was of practical and theoretical importance. When in 1938? they advanced the hypothesis that Tripsacum had played an important role in the development of North American maize, such a monograph be- came a scientific necessity. The available evidence, taxonomic, genetic, and cytological, suggested that the relationships of the various entities in Tripsacum were very poorly understood and that they might be quite intricate; extensive field work, cytological examination of living material, and routine tax- onomic techniques would all be necessary if an adequate under- standing of the genus was to be reached. A comprehensive program was accordingly outlined, and two successive grants from the Penrose Fund of the American Philosophical Society ! Journ. Hered. 22: 329—343. 2 Proc. Nat. Acad. Sci. 24: 303-312. Issued September 20, 1941. ANN. Mo. Bor. Garp., Vor. 28, 1941 (249) [Vor. 28 250 ANNALS OF THE MISSOURI BOTANICAL GARDEN enabled Dr. Cutler to visit the chief eastern herbaria and to colleet extensively in the United States, Mexico, and Guatemala in 1940 and 1941. We are also indebted to various other institu- tions and individuals for coöperation and support in the de- velopment of this program. For information, herbarium speci- mens, living plants, transportation, etc., acknowledgment should be made to G. T. Barrusta, P. C. Mangelsdorf, Mariano Pacheco, Ulises Rojas, H. B. Parks, and numerous others. We are under a very special obligation to J. H. Kempton, who not only supplied us with much pertinent information but even turned over to us his own unpublished notes on the genus. It is now apparent that our original estimate was correct and that anything like a final judgment on the entities which make up the genus Tripsacum must await the collection and integra- tion of evidence from several fields. We are therefore publish- ing this preliminary survey as a center about which the efforts of those interested in the problem may be coórdinated. A col- lection of living plants is being assembled at the Missouri Bo- tanical Garden, largely for cytological examination. We shall be grateful for plants or viable seeds from known localities or for chromosome counts made on plants of known derivation. We shall also appreciate herbarium specimens from localities other than those cited below. Plants, seeds, or specimens should be sent to Edgar Anderson, Missouri Botanical Gar- den, St. Louis, Missouri. Tripsacum is unfortunately one of those genera which pre- sent special difficulties to the collectors and have consequently been rather neglected by them. Making an accurate and com- plete record of a T'ripsacum plant on an ordinary herbarium sheet is like attempting to stable a camel in a dog kennel. By selecting portions of the plant and supplementing the specimen with pertinent information, an acceptable substitute for a com- plete specimen ean be made, however. Where possible, such a record should include: (1) a terminal or lateral inflorescence (labelled as such since the terminal is usually more branched); (2) one or two successive internodes, with the sheaths, au- rieles, and lower portions of the leaf-blades attached; (3) a 1941] CUTLER & ANDERSON—THE GENUS TRIPSACUM 251 complete leaf from one of the lowermost nodes, labelled as such (though neglected by many collectors this is particularly important since the species differ markedly in the degree to which the blade is constricted above the sheath on the lower leaves); (4) notes as to the height of the plants, number of nodes, and number of lateral inflorescences. The available cytological evidence suggests that phylo- genetic relationships within the genus may be quite complex. We are therefore deliberately postponing final judgment on many of the entities until chromosome counts can be made on a much larger scale and until our field studies are completed. Since we now know that the genus extends to South America it may be a matter of some years before the evidence is as- sembled. We have accordingly been conservative in proposing any new names, though under each species we have discussed briefly those variants which might ultimately receive recogni- tion. The new species described from South America is ad- vanced to this rank because it has previously been confused with a species to which in our opinion it bears no very close relationship. Specimens are cited geographically according to states and alphabetically according to collectors under the states. The first time a collector’s name appears his initials are given, if they be two or more; if not, the whole name where this is known. The following abbreviations have been used in refer- ring to specimens in the various herbaria: Ba—Bailey Hortorum, Cornell University. F—Field Museum of Natural History. Gray—Gray Herbarium, Harvard University. Ha—Museum of Economic Botany, Harvard University. MBG—Missouri Botanical Garden. Mich—University of Michigan. NY—New York Botanical Garden. USDA—United States Department of Agriculture. USNH—United States National Herbarium. Wisc—University of Wisconsin. Yale—Yale University. [Vor. 28, 1941] 252 ANNALS OF THE MISSOURI BOTANICAL GARDEN Tripsacum L. Syst. Nat. ed. 10. 1261. 1759. Digitaria Heist. ex Adans., Fam. Pl. 2: 38, 550. 1763. The genus has been a little-known one, and the variations within the species and the distribution have been left un- studied. A. S. Hitchcock? published a synopsis of Tripsacum in 1906, and in 1909 G. V. Nash* outlined the species for the ‘North American Flora.’ Since these two works appeared, many specimens have been collected, and although there are not yet enough to provide a firm basis for a complete revision of the genus, there are sufficient to indicate the problems which will be encountered. 1. T. floridanum Porter ex Vasey, Contrib. U. S. Nat. Herb. 3: 6. 1892. T. dactyloides var. floridanum Beal, Grasses N. Amer. 2: 19. 1896 This species is apparently distinct, and its delicate habit re- sembles that of Manisuris cylindrica more than that of any Tripsacum species. The Texas collection is probably introduced from Florida as no collections between the points have been made. RIDA: on road and glade at Crossman's, Dade Co., 9 Dee. 1903, 4. A. Eaton а ©: ay, USNH) ; rocky soil recently "D Garratt, Dade Co., 25 Feb. 1905, Eaton 1243 (Gray); Miami, June 1877, A. P. Garber 77 (Gray, USNH TYPE); а, March 1903, A. S. Hitchcock Кш, ; on eoral roek, Homestead, 2 April 1906, Hitchcock 686 (USNH) ; rocky soil near river, Miami, 4 April 1906, Hitch- cock ?26 Bisnis between Florida City and Royal Palm State Park, 30 Oct. 1935, . L. Loomis (MBG, USN ы oist diteh in tropieal hammock, 3 miles west of South Miami, 30 Jan. 1933, Hugh O'Neill 7610 (NY, USNH); Miami, 4-7 April 1898, С. L. Pollard $ С. N. Collins 272 (NY, USNH); pinelands near Nixon-Lewis Hammock, Dade Co., 15 Jan. 1916, J. K. Small 7350 (Gray, MBG, NY, USNH); in мана сирак пеаг Ше dat railroad grade between Cocoanut Grove and Cutler, 31 Oct.-4 1903, Small $ J. J. Carter 597 (NY); pinelands, Homestead to Big Hammock ан, 15-17 Feb. 1911, Small, Carter, G. К. Small 3433 (NY, USNH) ; in pinelands near the Homestead Trail, near Camp Longview, 13-16 May 1904, Small $ P. Wilson 1698 (NY); in — southern peninsular Florida, 6-7 May 1903, Small § Wilson 1727 (NY); rocky pine forest, open places and along bri sides, Homestead, 15 Feb. 1935, J. в, Swallen 5225 (USNH); Miami, 8 May 1904, S. M. Tracy 9318 (Gray, MBG). 3 Bot. Gaz. 41: 297-298. * М. Аш. Fl. 17: 79-81. "поцтрпоз [ensn oq] шогу SsuorjerieA AJIUSIS вовацуиете I y "логу а эзецәәтрәй 03 (әвотта) әўвүотуәа пләц}лоп 'sorpu e[rssoq (е-г згаицвтрит впотавро | $9A99[ I9MO'T 08—05 T-I 189A “ләшү 390 Сепввовдив) | (33-4) prds в1вшоувио eje[[oorpoq. CL ззицета -oyB[NoIeqny, | ejv[orjed yon 0017—08 9—5 03 озткхей (115598) | (GI-I) "ymo Kpop “етеш әұецәәтрәд 9 jourjsrpug вполаето | әұе[отјәй oN | 08—08 | #-Т -e3unp Y OIN o[rssosqns 03 (9—5) еГешәјвпо o[rssog I yourystput Клод | oynsity Kposredg | ejeporjed ON | 08-0т | @—6°0 03 зпоху jourjsrpqus | атм 9503991103 eyeporjod e[resog УТ 03 39013810 -oso[nut'T фецмешов 0$—01 $ вотлоцгу "в (епввовапв) | (9-6) (sno1quos) (OF) ‘SO Tey 915895 2-1 jourjsrpug snoiquw[) | eje[ored ок | 08-6Т | G-T -U39 рие 119358 Сепввовапв) (€-3) e[rssog г jourjsrput «лед snoiqw[p | eje[ored 40N єт I-9'0 ертлогя о mec m - = SER [|ESEZ Е RS ЕЕ MES z пос oBBB я. а, ENS HE ва: A = g oes e 28g Фе Б zm 8. a m Ф в го m во га ве c я o Eg ES] A E н = << g н- O Е [эл = е: 5 Br Ф = B што] а unsond `I wnjpnosp/ ‘Т, n401020W] `L IWS Г; s2pojfi;onp ‘Т wnunpr.oy 1; »NOOVSdIUL NI вяятоучуно 014199485 I W'IHViL [Vor. 28 254 ANNALS OF THE MISSOURI BOTANICAL GARDEN TEXAS: damp sands, Beaumont, 24 April 1903, J. Reverchon 4188 (Gray, MBG, USNH). 2. T. dactyloides (L.) L., Syst. Nat. ed. 10. 1261. 1759. Coix dactyloides L., Sp. Pl. 972. 1753. C. angulatus Mill., Gard. Dict. ed. 8. Coix no. 2. 1768. Ischaemum glabrum Walt., Fl. Carol. 249. 1788. Т. monostachyum Willd., Sp. Pl. 4: 202. 1805. T. dactyloides var. monostachyon Eaton & Wright, N. Amer. Bot. ed. 8. 461. 1840. T. compressum Fournier, Bull. Soc. Roy. Bot. Belg. 15: 466. 1876. T. dactyloides var. В monostachyum Fourn., Mex. Gram. 68. 1886. Dactylodes angulatum Kuntze, Rev. Gen. Pl. 2: 773. 1891. T. dactyloides var. monostachyum Vasey, Contrib. U. S. Nat. Herb. 3: 6. 1892. Dactylodes Dactylodes (L.) Kuntze, Rev. Gen. Pl. 3: 349. 1898. This species is almost as variable as those to the south. There are at least five groups which ean be distinguished al- though only one of these is distinet. (1) The specimens from along the Atlantic coast, which include the type, are less vari- able than any other group. The spike is usually solitary or binate, the staminate glumes stiff, bluntly tipped and the backs of them almost straight. (2) In Florida and adjacent Alabama and Georgia is found a wide-leaved and robust type with hairs at the ligule and on the inner side of the sheath and leaf ad- jacent. (3) On the prairies and plains and down to the Gulf coast of Texas and Louisiana is found a very variable form which, unlike all the other groups, with haploid chromosome numbers’ of 36, has haploid numbers of 18. This group has staminate glumes which are more variable, more pilose and scabrous, and with the back greatly curved. In large colonies of this group it is usually possible to find several plants which have the upper staminate spikelet barely pedicellate. (4) From * Mangelsdorf & Reeves. Texas Agr. Exp. Sta. Bull. 574. 1939. 1941] CUTLER & ANDERSON—THE GENUS TRIPSACUM 200 Central Illinois there are three specimens (the only ones from that area) with sub-pedicellate staminate bracts and with glumes softer and occasionally tapering. More material of this form is particularly desirable. CONNECTICUT: Guilford, 16 Aug. 1907, A. I. Bartlett (Yale) ; edge of salt marsh, Guilford, 19 Aug. 1906, G. H. Bartlett (Gray); Bridgeport, 1832, H. C. Beardslee (Yale) ; salt marsh at Pond Point, Milford, 27 July 1900, C. H. Bissell 2033 (Yale, Gray); waste ground, Raton Point, Norwalk, 23 Aug. 1901, Bissell 5389 (Yale); border of salt marsh, Guilford, 14 Aug. 1906 & 3 Sept. 1917, Bissell (Yale) ; in dry field by salt marsh, Stratford, 9 July 1912, A. E. Blewitt 318 (Yale); Pond Point, Milford, 24 Aug. 1909, H. S. Clark (Yale); dry bank of Housatonie River, in colonies along coast, Stratford, 31 July 1893, E. H. Eames (Gray, Yale); south end, East Haven, 1882, D. C. Eaton (Yale); along shores and eoves, East Lyme, Groton and Waterford, 1882-1889, C. B. Graves (Yale); brackish meadows, Bran- ford, 3 Sept. 1902, В. W. Woodward (Gray, Yale). w Уовк: near garbage reduction plant, Staten Island, 16 July 1932, J. A. Кы 8309 (MBG). PENNSYLVANIA: Safe Harbor, 2 Aug. 1882, Small (NY). New JERSEY: sandy beach along Delaware River, near Elsinboro Point, 3 miles southwest of Salem, 31 Oct. 1933, J. M. Fogg, 6280 (Gray); dike bordering tidal marsh, 1.5 miles southwest of Harrisonville, 29 Oct. 1934, Fogg (Gray); along stream, Woodbridge, 13 bay 915, L. H. Lighthipe (MBG); Cape May, 4 Aug. 1909, F. W. Pennell 2225 (USNH). MARYLAND: Great Falls, 4 Sept. 1899, C. E. Ball 68 (Gray) ; d View, 9 Aug. 1891, F. Blanchard (USNH); dry sandy soil, boron ry, Sept. W. M. Canby 173 (USNH) ; sandy beach, Millstone, 4 Aug. , Hitchcock 0 Е БМН); оп roeks at Great Falls, 2 Sept. 1898, T. Hol Eu, Mattawoman Creek, 3 Aug. 1914, I. Tidestrom 7223 (Gray, MBG, USNH). DELAWARE: near Wilmington, July 1866, Canby (Gray, Mich, MBG, Yale); darum Beach, 11 Aug. 1911, J. В. Churchill (Gray). VIRGINIA: roadside, Arlington Farm, near Rosslyn, 13 Sept. 1933, H. A. Allard 58 (Gray); Bedford Со., 15 Aug. 1871, A. H. Curtiss 9848 (MBG); dry roadside (Gray) ; rieh alluvial woods and thiekets baek of sand beach of James River, Clare- mont Wharf, Surry Co., 13 June 1938, Fernald $ Long 8097 (Gray); swales, Mun- den, 3-19 Sept. 1905, K. K. Mackenzie 1743 (Gray, NY) ; Portsmouth, 4 July 1897, E. B. Noyes 3495 (Yale). NORTH CAROLINA: Chapel Hill, May, W. W. Ashe (Yale); open moist low soil along Hiteheoek Creek, Rockingham, Riehmond Co., 25 Aug. 1936, D. S. Correll ?124 (Gray); open bank of Lake Raleigh, Wake Co., 20 May 1938, Е. K. Godfrey 4014 (Gray, NY) ; sandy soil, marsh bordering Brice Creek, 2 miles southwest of Jam City, Craven Co., 11 July 1922, L. Е. $ Е. Е. Randolph 868 (Gray); Asheville, i Aug. 1891, A. B. Seymour 2 (Gray, MBG); ditch bank, Pullman Park, Raleigh, 27 June 1927, K. M. Wiegand $ W. E. Manning 120 (Gr SovTH CAROLINA: damp soil, near Troy, аы d 8 May 1921, John Davis [Vor. 28 256 ANNALS OF THE MISSOURI BOTANICAL GARDEN 2006 (MBG) ; in moist ditch along roadside, near the Clementia Tourist Camp, 14 miles south of Charleston, 8 Nov. 1929, H. N. Moldenke 142 (NY) ; swale, south side of Santee River, north of Bonneau, Berkeley Co., 12 July 1927, Wiegand $ Manning 121 (Gray). GEORGIA: Yellow River, Gwinnett Co., 27 J uly 1897, H. Eggert (MBG); sandy soil near Middle Oconee River, Clarke Co., alt. 620 ft., 29 June 1900, R. M. Harper 99 (Gray, NY); field by Bobbin Mill Creek, Athens, 16 June 1934, L. M. Perry 755 (NY); Yellow River near MeGuire's Mill, Gwinnett Co., alt. 750 ft., 2 July 1895, Small (NY). Mississippi: Grand Batture Island, 23 May 1911, A. H. Howell 748 (USNH); Long Beach, 19 Aug. 1891, J. F. Joor, (MBG); near Starkville, 27 Sept. 1896, T. H. Kearney 59 (Gray, USNH) ; Agrieultural College, Oktibbeha Co., 11-17 Aug. 1896, C. L. Pollard 1272 (Gray, MBG, NY, USNH) ; low ground, Milton, July 1931, William Rhodes (Gray); in ditches along roadside, Agrieultural College, 17 July 1902, P. L. Ricker 848 (USNH); Miller, De Soto Co., 12 July 1923, L. E. Wehmeyer (Mich). FLoRIDA: near Tallahassee, №. К. Berg (МУ); hammock south of Miami, 8 Sept. 1907, Agnes Chase 3903 (USNH); Homosassa, Robert Combs 943 (Gray); along edge of saw grass and lakes in fertile hammock, not uncommon, Grasmere, Orange Co., 20 Sept. 1898, Combs $ С. Е. Baker 1043 (USNH); large open prairies near small stream, uncommon or rare, Bradenton, Manatee Co., 1898, Combs 1254 (USNH); low moist soil near Osprey, Sarasota Со., 9 July 1936, Correll 5889 (Gray); open dry soil along edge of ditch, near Oviedo, Seminole Co., 8 Aug. 1936, along Tamiami Trail, 25 miles west of Miami, Dade Co., 26 Dee. 1927, Moldenke 3743 (NY); clay soil, vicinity of Eustis Lake, Lake Co., 1-15 April 1894, G. V. Nash 374 (Gray, MBG, NY, USNH) ; edge of a cypress swamp, 4-7 ft. tall, Lake Harris, near Oeklawaha River, 5 July 1895, Nash 2140 (NY); Lake City, Columbia Co., 11-19 July 1895, Nash 2207 (Gray, USNH) ; Little River, 26 March 1923 D. С. Peattie 1915 (USNH); Lake City, 29 June 1894. P. H. Rolfs 806 (MBG, USNH) : 1842, F. Rugel 438 (MBG, USNH) ; hammock, Cutler, 15 Nov. 1906, Small 4- Carter (NY); Ft. Myers, Lee Co., 1 June 1916, J. Р. Standley 215 (Gray, MBG, USNH) ; moist place among pines and scrub palmetto, Titusville, 29 May-3 June 1926, Swallen 212 (MBG, USNH) ; edge of old cultivated field, Marianna, 25-29 June 1926, Swallen 492 (USNH) ; Long Key, 25 May 1901, S. M. Tracy ?775 (Gray, USNH) ; edge of salt marsh, Ormond, 8 Aug. 1896, H. J. Webber 472 (USNH). ALABAMA: Auburn, 12 June 1897, G. W. Carver 48 (USNH); in a swale, near 1941] CUTLER & ANDERSON—THE GENUS TRIPSACUM 257 Selma, 29 June 1895, Kearney 10 (Gray, MBG, NY, USNH); low rich Г ара Mobile, 1 June 1883, Charles Mohr 603 (USNH) ; Alex. Winchell 236 (USNH). LoviSIANA: U. S. Riee Experiment Station, 1 mile west of Crowley, В. Parish, 21 Мау 1940, H. C. Cutler 3156 (Ва, Е, MBG, МУ, Ha, ОБРА); 1 mile east of Vinton, alt. 60 ft., 3 May 1941, Cutler 4813 (MBG); in salty swamps, Pointe a la Hache P. O., 4 July 1885, A. B. Langlois (USNH) ; prairie, Welsh, Jefferson Davis Parish, 17 Мау 1915, Е. J. Palmer 7652 (NY); large clump in swamp, Cam eron Co., 11-13 June 1931, Swallen 1891 (USNH) ; east bank of lower Bayou Little Caillou, 27 miles below Houma, near Cocodrie, 8 June 1913, E. C. Wurzlow (USNH). MicHIGAN: along railroad tracks, Utica, 11 Aug. 1922, O. A. Farwell 6297 (Mich, Gray, USNH). INDIANA: common along a ditch through a cultivated field about 5 miles east of Lincoln City, Spencer Co., 10 Oct. 1931, C. C. Deam 51560 (USNH). TENNESSEE: Nashville, A. Gattinger (Gray, USNH); along the French Broad River between Paint Rock and Del Rio, Cocke Co., 10 Sept. 1897, Kearney 938 (NY, USNH) ; low places, Spring City, July 1929, Rhodes (Gray). ILLINOIS: wet prairies about Salem, July 1860, M. 8. Bebb (Gray, Yale) ; Kicka- poo Prairie, near Washington, wet ground, June 1835, George Engelman (MBG) ; Spoon River bottom, Fulton Co., 6 June 1842, 0. B. Mean (USNH) ; Hancock, 1842, Mean (Wise). Iowa: ах in wet plaees, Deeatur Со., 14 July 1897, T. J. $ М. Е. І. дема (ЧҮ MISSOURI: uncommon, rocky river banks, Greene Co., 4 Sept. 1892, B. Е. Bush 430 о Montier, 30 June 1894, Bush 877 (MBG) ; оп Skinker's Wege, St. Louis, 29 June 1875, Eggert (MBG); Springfield, 1897, S. A. Hoover (Gray); Jerome, 6 June 1914, J. H. Kellogg (MBG) ; St. Clair, Franklin Co., 12 June 1928, Kellogg 2103 (MBG); Washington, Franklin Co., 25 June 1888, L. H. Pammel (MBG) ; 1056. Julian Steyermark 11221 (MBG) ; sandy alluvium along Eleven Point River, 1% mile north of MeCormack Hollow, Oregon Co., 27 July 1936, Steyermark 12322 (MBG) ; lower wooded slopes bordering field along King’s River, southeast of Allen Ford, in Barry Co., 22 June 1937, Steyermark 22577 (МВС); open limestone slopes along Indian Creek near Holy Cliff, 3% miles northeast of Topaz, Douglas Co., 19 July 1937, Steyermark 23366 (MBG). ARKANSAS: low wet fields, Monticello, Drew Co., Delzie Demaree 13690 (NY); wet plaees, Miller Co., 23 July 1896, Eggert 155 (USNB); ; Monticello, 24 July 1881, G. W. Letterman (USNH) ; open field adjacent to station, Stuttgart, 30 July 1932, D. M. Moore 32805 (NY). Kansas: along railroad, 5 miles northwest of Lawrence, Douglas Co., 18 Jun 1938, Edgar Anderson (MBG); Florence, 28-30 July 1903, David Griffiths bis H) ; sandy roadside, 2 miles west of St. George, Riley Co., 3 June 1908, Hitch- cock 2527 (USNH); 2-3 ft. high, low prairie, Manhattan, 24 June 1913, Hitch- cock 10420 (USNH) ; wet places, Riley Co., 20 June 1895, J. B. Norton 580 (Gray) ; schoolhouse hill & town reservoir, Pleasanton, Linn Co., 19 June 1929, P. A. Ryd- berg & Ralph Imler 85 (NY); Big Spring, 3 or 4 miles south of Bilby’s Lakes, 16 July 1929, Rydberg $ Imler 1072 (NY); high prairie, Lawrence, Aug. 1892, W. C. Stevens 62 (USNH). [Vor. 28 258 ANNALS OF THE MISSOURI BOTANICAL GARDEN OKLAHOMA: Cimarron Valley, Cherokee Outlet, 1891, M. A. Carleton 225 (USNH) ; in the Indian Terr., chiefly on the False Washita, between Fort Cobb & Fort Arbuckle, 1868, Ed. Palmer 422 (Yale) ; grassy valley near Fairvalley, Woods Co., 28 May 1913, С. W. Stevens 753 (Gray, USNH) ; in large gravelly bar in Spring River, near Ottawa, Ottawa Co., 29 Aug. 1913, Stevens 2510 (Gray) ; grassy sandy valley near Guthrie, Logan Co., 14 June 1914, Stevens 3284 (Gray). Texas: Leon Springs, Bexar Co., 19 Мау 1911, Mr. $ Mrs. J. Clemens 21 (MBG, USNH); Soil Conservation Service Nursery, San Antonio, 23 May 1940, Cutler & Reeves ‘‘San Antonio’’ Tripsacum, 23 May 1940, Cutler 3169 (MBG); Kemah, alt. 20 ft., 3 July 1926, G. L. Fisher 267 (USNH) ; 6 miles west of Raywood, Liberty Co., alt. 60 ft., 6 May 1941, Cutler 4815 (MBG); between Ft. Bend and Harris Cake. 11 miles southwest of Houston, alt. 60 ft., 6 May 1941, Cutler 4816 (MBG) ; Houston, 11 July 1915, Fisher 1711 (USNH) ; wet ground, Houston, 20 April 1872, Elihu Hall 844 (Gray, NY, USNH); roadside, Waxahachie, 19 May 1936, H. E. Hazard (Gray); Harvester, 24 April 1906, Hitchcock 1198 (USNH); edge of m near river, New Braunfels, 20 June 1910, Hitchcock 5240 (USNH); bank of stream, San Antonio, 24 June 1910, Hitchcock 5255 (USNH); 1888, G. C. рач (Gray, NY, USNH) ; rocky creek banks, Austin, Travis Co., 12 Ма ay 1918, Е. J. Palmer 18659 (USNH) ; rich damp lands, Dallas, May 1879, Reverchon 1156 (USNH); dry sands, Lindale, e May 1902, Reverchon 2804 (NY); Boerne Road at Bexar Co. line, 31 May 1931, W. A. Silveus 128 (USNH) ; open places, Bracken- ridge Park, San Antonio, 4 Oct. 1933, Sister Mary C. Metz (NY); in water, Knickerbocker Ranch, Dove Creek, Tom Green Co., May 1880, Frank Tweedy (Gray, Yal (5) The fifth group is distinct and occupies a separate area on the western edge of the range of the species. All the speci- mens seen have been from west of the Pecos River. It is prob- able that there are intergrades but the group is decidedly dif- ferent from specimens from central and eastern Texas. 2a. T. dactyloides var. occidentale, n. var.? Similar to the species but the staminate glumes more than 9 mm. long, softer, and tapering to an acute tip. TEXAS: shaded ledges at base of bluff, Little Aguja Canyon, Davis Mts., Jeff Davis Co., alt. 1520 m., 15 June 1931, J. A. Moore 4: Julian Steyermark 3092 (Gray, Mich, MBG түре, NY, USNH); Chisos Mts., 5 Aug. 1931, C. H. Mueller 7891 d ; rocky partially shaded ground, along Зы, near Alpine, Brewster Co., 8 June 1926, Palmer 30584a (MBG); rocky banks of creek, in deep с anyon, Oak Phim Chisos Mts., Brewster Co., 24 May 1928, Palmer 24159 (Gray, MBG, * T. dactyloides var. occidentale, var. nov., speciei simile sed glumis stamineis plusquam 9 mm. longis, mollioribus et mueronatis ad apicem. 1941] CUTLER & ANDERSON—THE GENUS TRIPSACUM 259 NY); rocky plains, partially shaded situations amongst syenite boulders, Davis Mts., Jeff Davis Co., 4 Oct. 1926, E. J. Palmer 31923 (MBG, NY, USN F Davis, 1881, V. Havard (USNH). 3. Tripsacum australe, n. sp.” T. dactyloides subsp. hispidum Hitchcock, Bot. Gaz. 41: 295. 1906, in part. Plant slender to robust, nodes usually enlarged; leaves 1—4 em. wide, somewhat petiolate, blades smooth, usually glabrous, sheath with distinct to semi-distinet auricles, outer surface glabrous below, lanulose-tomentose above, at maturity barely clasping the culm, culm lightly to heavily lanulose-tomentose ; inflorescence of 1-4, rarely more, spikes, staminate spikelets sessile. This species is readily distinguished from T. dactyloides by the lanulose tomentum investing portions of the culm and sheath, the tendency of the leaves to become petiolate, and its range. From T. latifolium it is distinguished with difficulty in the northern part of its range but the specimens may be de- termined by the presence of the tomentum, and (on the her- barium specimens, at least) the less petiolate leaves. The type of tomentum is considerably different from that found in the specimens of T. lanceolatum with which the Morong specimen was grouped to form a subspecies of T. dactyloides. In the South American plants (Т. australe) the tomentum is soft and felted while in those of Mexico the hairs are ax thieker and harsh. SOUTH AMERI OLIVIA: among 8 x Coroieo, Nor-Yungas, alt. 1560 m., 25 Dee. 1923, Hitch- ps 22721 (Gray, USNH) ; marshy edges of forest, M Amazon Basin, alt. 1000 Не: 22 Пес. 1921, О. E. White 2324 (NY, USNH турк). : Burchell 9066 (Gray, USNH); in savannas among shrubs, Pará-Marajo Mes г a River, June 1914, Andre Goeldi 87 (USNH); Morrinho do Lyra Linha alse, Маш Grosso, May 1918, T. С. Kuhlmann 1833 (USNH) ; Г. Riedel 1279 (Gray, MBG). ''Tripsacum australe, sp. nov. Planta gracilis vel robusta; nodis plerumque tumidis; foliis 1—4 em. latis, aliquid petiolatis, laminis levibus plerumque glabris; vaginis eum auriculis plerumque distinctis, superficie exteriore basi glabra, apice lanuloso-tomentosa; culmo lanuloso-tomentoso; infloreseentia plerumque 1—4 spicis, spieulis stamineis sessilibus. [Vor. 28 260 ANNALS OF THE MISSOURI BOTANICAL GARDEN BRITISH GUIANA: Kanuku Mts, behind Parika, Rupunini Distriet, Jan. 1934, J. G. Meyers 4842 (USNH). COLOMBIA: coarse herb, up to 6 ft., open erm Dept. Santander, upper Rio Lebrija valley, northwest of Bucaramanga, alt. 400-700 m., Eastern би, 29 Пес. 1926, E. P. Killip & A. C. Smith 16279 mia ХЕ, USNH); 1760—1808, J. C Mutis 5489 (USNH) ; 5-6 feet tall, local and rather rare on hillsides in sheltered ne Van on the border of forest below 2500 ft., near Masinga, Santa Marta, „ 27 Oct. 1898, Н. Н. Smith 2745 (= 119) (Gray, MBG, NY, USNH). BOUADOR: partly shaded slope, large bunches, 4—6 ft., between Huigra and Naran- japata, Prov. Chimborazo, alt. 600-1200 m., 17 July 1923, Hitchcock 20643 (Gray, NX, DN AGUAY: Cerro de Tobatí, 14 Jan. 1903, K. Fiebrig 746 (Gray) ; Uferwaldrand Hirn bei Gestein, zwischen Rio Apa und Rio Aquidaban, 1908-1909, Fiebrig 4618 (Gray) ; in regione versus superioris fluminis Apa, 1901-2, E. Hassler 7901 (Gray) ; in altaplanitie et declivibus, Sierra de Amambay, Dee. 1907, Hassler 9953 (USNH) ; Central Paraguay, 1888-1890, Thomas Morong 675 (MBG, NY, USNH); on the Paraná, 26°-24° S. lat., April 1883, D. Parodi 53 (NY). VENEZUELA: on slopes, forms large tufts, ascent from Motatan Bridge to Car- vajal,near Valera, Trujillo, H. Pittier 10768 (Gray, NY, USNH); near Eseuque, Trujilla, in savannas on road to Valera, 11 Jan. 1929, Pittier 13151 (USNH). 4. Tripsacum lanceolatum Rupr. ex Fourn., Mex. Gram. 68. 1886. T. lanceolatum Rupr. in Benth., Pl. Hartweg. 347. 1857, nomen nudum. T. acutiflorum Fourn., Bull. Soc. Roy. Bot. Belg. 15: 466. 1876, nomen nudum. T. Lemmon Vasey, Contrib. U. S. Nat. Herb. 3: 6. 1892. T. dactyloides var. Lemmoni (Vasey) Beal, Grasses N. Amer. 2: 19. 1896. zT. ei age var. angustifolium Scribn. in Urbina, Cat. PI. Mex. 376. 1897. T. dact Пощ subsp. hispidum Hitchc., Bot. Gaz. 41: 295. 1906, in part. While Nash! (1909) accepted the description by Fournier in 1876 as constituting valid publication, no real distinctions be- tween species were made in Fournier’s article and it there- fore cannot be accepted. Fournier’s later description is based on specimens and is detailed enough to distinguish between the species then known. VN. див PFL 14: 81. 1941] CUTLER & ANDERSON—THE GENUS TRIPSACUM 261 It is possible to separate the specimens of T. lanceolatum into three general groups but these have some intergrades, and until further collections are made and more entire plant speci- mens observed, it will be futile to attempt to delimit these groups. The type specimen is from Aguas Calientes in Central Mex- ico, and the group from the central plateau is characterized by а large amount of anthocyanin coloring, narrow leaves, and solitary or paired spikes in the inflorescence. Closest to this group are those specimens found on the west slopes of Mex- ieo, which have broader leaves, solitary to ternate inflorescence spikes, and are the most robust. From these two groups it is easy to separate those specimens from southern Arizona which formerly went under the name of T. Lemmoni. These are characterized by an inflorescence lacking anthocyanin and much divided, with as many as nine spikes, usually pedicellate upper spikelet, and narrow leaves. With present material, this group may be readily separated by the gap between its representatives and those of the other groups. The gap must not, however, be interpreted as a real absence of the plant but as an absence of collections from northern Mexico. T. Lemmoni is probably distinct enough to be considered a good species, but until the extent of the variations within the species T. lanceolatum has been determined and until collec- tions have been made at more points in northern Mexico it will be better to consider T. lanceolatum as a variable species with T. Lemmoni as one of several groups within it. There are several specimens, as that of Brandegee from El Taste, Baja California, 1 Nov. 1902, which do not fit well into any of the three groups but are well within the species as interpreted herein. There is some resemblance to T. dacty- loides in the northeastern Mexican specimens, for example, in Wynd & Mueller 536, with larger glumes and wider leaves. ARIZONA: % mile north up lateral canyon, 8 miles down Sonoita Creek from Patagonia, Santa Cruz Co., 28 April 1941, H. C. Cutler $ J. D. Freeman (MBG); Mule Mts., about 5000 ft., 20 Sept. 1929, G. J. Harrison 4 T. H. Kearney 6101 [Vor. 28 262 ANNALS OF THE MISSOURI BOTANICAL GARDEN (Gray, USNH) ; grown in Washington, D. C. greenhouse by J. H. Kempton, from seeds secured in southern Arizona, probably the Mule Mts., by Kearney. U.S.D.A No. T29-29, picked Oct. 1934 (MBG, USNH); on a high peak with southern slope, near moist rocks, Huachuca Mts., 21 Sept. 1882, J. G. Lemmon 2932 (Gray, USNH type of T. Lemmoni) ; near Patagonia, 23 Feb. 1930, H. Е. Loomis 6409 (USDA, USNH) MEXICO: AGUAS CALIENTES: Aguas Calientes, 1839, Theodor Hartweg 252 (NY cotype, complete speeimen, USNH, fragments from the Steudel and the Trinius herbaria, cotypes, and Boissier Herb. T BAJA CALIFORNIA: El Taste, 13 Sept. 1893, T. S. Brandegee 4 (NY); Sierra de San Franeisquito, 29 Sept. 1899, Brandegee 6 (USNH); El Taste, 1 Nov. 1902, Brandegee CHIHUAHUA: infrequently scattered, occurring as small groups, 2 or 3 m. high, pine oak country; Sierra Canelo, Rio Mayo, 29 Aug. 1936, H. S. Gentry 2496 (Gray, MBG); in large elumps in arroyo bed by running water, Sierra Charuco, Rio Mayo, 1 Oct. 1936, Gentry 2914 (Gray); roeky ravine, 8000 ft., Sanehez, 12 Oct. 1910, Hitchcock 7702 (USNH); Rio Bonito ‘hot country,’’ 25 Aug. 1936, Harde Le Sueur Mex-093 (USNH). COAHUILA: moist stream side, Hacienda Piedra Blanca, Canyon de Sentenela, Sierra del Carmen, Villa Acuña, 6 July 1936, Е. L. Wynd $ С. Н. Mueller 536 (Gray, MBG, NY, USNH). COLIMA: large bunches on rocky cliff by seashore, Manzanillo, 20 Sept. 1910, Hitchcock, Amer. Gr. Nat. Herb. No. 230 (Gray, MBG, NY, USNH); rocky hill- side, alt. 1500 ft., Alzada, 21 Sept. 1910, Hitchcock 7082 (Mich, NY, USNH), 7083 TM GO: rocky hill, Iron Mt., Durango, alt. 6200 ft., 6-8 Oct. 1910, Hitchcock в Совы, 7648 (NY, USNH) ; Га Bajada, тушай, 300—600 m., Nov. 1921, J. б. Ortega 4334 (USNH) ; city of Durango and vicinity, Àpi-Nos. 1896, Ed. Palmer 587 (Gray, MBG, NY, USNH). GUERRERO: 33 km. south of Chilpancingo on Mexico-Aeapuleo road, alt. 1360 m., 24 Sept. 1940, Cutler 3918 (Ва, Ha, Е, MBG, USDA) ; on rocks of cliff, alt. 1500 ft., Balsas, 9 bie 1910, Hitchcock 9816 (USNH). JALISCO: Arenal, 9 Oct. 1923, Collins $ Kempton 79 (USNH); Barranca de Oblatos, фарб 12 Oct. 1923, Collins $ Kempton 81, 85, 88 (USNH) ; 0.5 km. 4018 (Ba, Ha, F, MBG, USDA); Platanar, on railroad 53 km. north of Colima, alt. 1000 m., 10 Oet. 1940, Cutler 4087 (MBG, Ha) ; walls of barranea, 1 km. north- east of Ciudad Guzman (Zapotlan), alt. 1520 m., 10 Р 1940, Cutler 4088 (MBG), along creek, same locality and date, 4104 (MBG), a 1600 m., 4105 (Ba, F, Ha, MBG, NY, USDA); side of Barranca de Oblatos, Guna alt. 5000-6000 ft., 27-28 Sept. 1910, Hitchcock 7358 (USNH); La B , Guadalajara, 21 Nov. 1930, М. E. Jones 27628 (MBG, NY, USNH) ; ане de Ам, сенін 15 Oct. 1921, Kempton $ Collins (USNH Nos. 1064495-1064497); Hacienda San Diego, Cocula, 21 Oct. 1921, Kempton $ Collins (USNH) ; mts. near Guadalajara, 16 Dee. 1889, Pringle 2610 (USNH); road between Juejuquilla & Mesquitee, 25 Aug. 1897, Rose 3570 (Gray, USNH). 1941] CUTLER & ANDERSON—THE GENUS TRIPSACUM 263 MEXICO (including DISTRITO FEDERAL): Ваггапса de Dolores, Lomas de Cha- pultepee, Aug. 1940, G. T. Barrusta (MBG); pedregal near Mexico, 26 June, M. Bourgeau 444 (Gray); Tlalpam, alt. 7480 ft., 3 Aug. 1924, Fisher (MBG); San Angel, alt. 7350 ft., 2 Aug. 1926, Fisher 53 (USNH); Temasealtepee, alt. 1750 m., 30 Aug. 1932, G. у Hinton 1444 (Gray); hill, Tejupileo, 1340 m., Tema- зсаЦерес, 4 Sept. 1932, Hinton 1600 (Gray); copse, edge of a Tacubaya, 27-30 July 1910, Hitchcock 5909 (USNH); Mexico City, Oct. 1896, E. W. D. Holway 8 (USNH); Pedregal de San Angel, Sept. 1927, E. Lyonnet 61 ps MBG, NY USNH); rare on voleanie soils, pedregal, Tlalpam, alt. 2300 m., 30 Oet. 1928, M. St. Pierre 828 (USNH) ; Taeubaya, J. G. Schaffner 41 (USNH, fragment from Paris Herb.). MICHOACAN: near Cerro de las Nalgas, alt. 1900 m., vicinity of Morelia, 9 Sept. 1909, Bro. G. Arsène 2572 (USNH) ; Cerro de las Nalgas, alt. 800 m., 9 Sept. 1909, Arsène (USNH) ; near La Huerta, 1950 m., vicinity of Morelia, 1 Sept. 1910, Arsène 5576 (MBG, NY, USNH) ; easeade near Loma de La Huerta, alt. 1950 m., vieinity of Morelia, 1 Sept. 1910, Arséne 7006 (USNH). MORELOS: small clumps, rocky cliffs, alt. 4500 ft., Cuernavaca, 10-11 Sept. 1910, Hitchcock 6840 (Mich, USNH); Valle de Tepeite, ^ Sept. 1938, Lyonnet 2421 (USNH) ; Teposteco, 22 Sept. 1938, Lyonnet 2552 (USNH); Xochitepee, 24 Sept. 1938, Lyonnet 2645 (USNH) ; Barranca de San rd near Cuernavaca, 28 Oct. 1904, Seler 4348 (USNH). yARIT: Los Fresnos, Tepic, 2 Oct. 1923, Collins $ Kempton 135 (USNH); EE del Cruz, Tepic, 1 Oct. 1923, Kempton ф Collins T17 (USNH NUEVO LEON: Diente Canyon, Sierra Madre, Monterrey, 29 July 1933, С. Н. $ M. T. Mueller 368 (Gray, USNH); in moist places, dense woods bordering stream, Canyon Marisio Abajo, Raneho Las Adjuntas, Municipio de Villa Santiago, 27 June 1935, zu 2069 (Mich, Gray, USNH). ocky cliff, Oaxaca, 5000 ft., 12-13 Aug. 1910, Hitchcock 6160 (USNH) ; via lis pun 1842, F. M. Liebmann 547 (USNH). PUEBLA: source au dessus de la finca Guadalupe, alt. 2121 m., 20 Nov. 1906, Arséne 73 (USNH); Mayorazgo, sur l'Atoyae, alt. 2120 m., vie. Puebla, 7 July 1907, Arséne 1328 (MBG, USNH); near Hacienda Batan, Barranca de la Alseseea, alt. 2120 m., faced of Puebla, 13 June 1907, Arsène 1472 (MBG, USNH) ; entre les haciendas Santa Barbara y Cristo, sur 1’Alseseca, alt. 2150 m., vie. Puebla, 27 June 1907, Arséne SUED USNH); Aeatzineo, Distrito 2 Dupuis, vie. Puebla, alt. 2110 m., July 1907, Arséne 2266 (MBG, USNH); Mayorazgo, alt. 2120 m., vie. Puebla, 4 July 1907, Arsène 10106 (USNH) ; El Riego, July 1905, E A. Purpus 1227 (MBG SAN LUIS POTOSI: limestone ridges, San Jose Pass, 15 Aug. 1890, Pringle 3447 (USNH); roeky hills, Las Canoas, 14 Aug. 1891, Pringle 3811 (Gray, MBG, NY, USNH түрЕ of T. dactyloides subsp. hispidum); Bargre, Aug. 1911, С. A. Purpus 5433 (USNH). SONORA: small infrequent colonies, 2-3 m. high, oak hill slope, ^ oae Rio Fuerte, 22 Oct. 1936, Gentry 2953 (USNH); Guadalupe Canyon, 27 Aug. 1893, E. C. Merron 2035 (USNH) ; Santa Rosa Canyon, near Bavispe, ki реа Sonora, 19 July 1938, S. S. White 601 (Gray) ; small valley in granitie hills, 14 miles north of Babiacora on road to Cumpas, 22 Sept. 1934, I. L. Wiggins 7392 (Mich, USNH). [Vor. 28 264 ANNALS OF THE MISSOURI BOTANICAL GARDEN TAMAULIPAS: Sierra de San Carlos, vicinity of San Jose, alt. 3040 ft., 13 July 1930, H. H. Bartlett 10310 (NY, USNH). YUCATAN: edge of old field, Chichen Itza, 7-13 July 1932, Swallen 2482 (USNH). GUATEMALA: Lake Retana, north of Jutiapa, 18 Oct. 1935, Kempton $ Wil- son Popenoe (MBG, USNH); San Pedro, alt. 5600 ft., 28 Oct. 1935, Kemp- ton ф Popenoe (MBG, USNH); below San Pedro, alt. 4600 ft., 29 Oct. 1935, Kempton 4$ Popenoe (MBG, USNH); hills above San Lucas, 4500 ft., 9 Nov. 1935, Kempton 4: Popenoe (MBG, USNH). 5. Tripsacum fasciculatum Trin. ex Ascherson, Verh. bot. Ver. Prov. Brandenb. 17: 79. 1875. T. fasciculatum Trin. ex Steud., Nomencl. Bot. 2: 712. 1841, nomen nudum; Gram. 1: 363. 1855, nomen nudum; ex Rupr., Bull. Acad. Roy. Brux. 9: 243. 1842, nomen nudum. T. laxum Nash, N. Amer. Fl. 17: 81. 1909. This species has a rather wide range of variation but is easily distinguished in most cases by the smooth sheaths, the wide and not petiolate leaves, and the robust habit. Specimens from cultivation vary considerably, and this may be due to a more favorable environment. In central Guatemala T. fasciculatum is eut for forage, probably from native stands. This species apparently has a high degree of sterility. Her- barium specimens of it or of T. pilosum never reveal the shiny seeds which in T. dactyloides are indicative of a well-developed endosperm. Nearly a thousand seeds were collected in Mex- ico and Guatemala but all those examined had undeveloped embryos, and when planted have so far failed to germinate. The amount of variation in time of flowering and size and num- ber of parts of large colonies in both Mexico and Guatemala, however, suggest that the members of a colony are not all of the same clone. In northern Guatemala and in western Mexico no intergrades have been found between T. pilosum and T. fasciculatum, al- though both of these oceur in the region. They have never been reported from the same spot but their habitat require- ments are similar. Specimens with solitary terminal spikes are infrequent, and although such plants usually have narrower leaves than those with a much branched inflorescence, the leaves are still wider than those of T'. lanceolatum and the plant more robust. 1941] CUTLER & ANDERSON—THE GENUS TRIPSACUM 265 Mexico: COLIMA: rocky, grassy hillside, 1500 ft., Alzada, 21 Sept. 1910, Hitchcock 7103 (Mic GUERRERO: 36 km. south of Chilpancingo, alt. 1360 m., 24 Sept. 1940, Cutler 3915— 3917, 3921 (Ba, F, Ha, MBG). VERA CRUZ: Mirador, Aug. 1841, Liebmann 549 (MBG, USNH) ; [Fortin], Zaeua- pan, 1917, Purpus 8027 (Gray, MBG, NY, USNH); Hacienda de la Laguna, Bar- rio de Tinxedo, 1836, C. T. Schiede 947 (TYPE, not seen). GUATEMALA: Jardin de Don Mariano Pacheco, plants from Coban, Alta Verapaz, 8 Nov. 1941, Cutler 4301 (Ha, MBG); field at e 5 km. on Guatemala-Barberena E alt. 6500 ft., 9 Nov. 1940, Cutler 4306 (Ba, F, Ha, Mich, MBG, NY, USDA, USNH); same окай, T7; Nie 1940, Cutler ps (MBG) ; below San Pedro, alt. 4600 ft., 29 Oct. 1935, Kempton 4. Popenoe (MBG, USNH) ; J. J. i re- ceived 21 June 1916, grown in quarantine house, U. S. D. A. No. 42967 (USNH); edge of pine em ^ bd: Dept. Izabal, alt. 75-225 m., 15-31 May E Standley 24256 (U SALVADOR: San pum 1922, Salvador Calderón 1332 (Gray, NY, MBG, USNH); seed from C. Deusen, grown in greenhouse, Washington, D.C., 14 March 1924 (USNH); cultivated as forage, 21 Nov. 1916, Carlos Renson 1, 2 (USNH); eultivated, vieinity of San Salvador, 650—850 m., 30 March-24 April 1922, Standley 22631, 23637 (Gray, MBG, USNH). PANAMA: cultivated for forage, Coclé, alt. 600 m., Paul Allen 2252 (MBG). 6. T. pilosum Scrib. & Merr., U. S. Dept. Agric. Div. Agrost. Bull. 24: 6. 1901. We have never seen this species or T. fasciculatum growing without having found T. lanceolatum near by. T. pilosum is more selective in its habitat, and in localities on the west slope of Mexico it will be found in more protected sites while T. lanceolatum will extend beyond the margins of the best locali- ties. Thus, T. pilosum is usually restrieted to damp spots in open woods, along the railroads and on the slopes of deep canyons, but T. lanceolatum will grow in these places and in addition will fringe the upper rim of the canyon and eneroach upon the dry plains and hillsides. There is a local distinction between these two species. T. pilosum is known as maiscillo, T. lanceolatum as sacaton, a general term for coarse grasses. The irritating hairs of T. pilosum have been noted by Kempton (MS.), and we found that the hand which grasped the basal sheaths of this species during a day of collecting would remain swollen and sore for several days. [Vor. 28 266 ANNALS OF THE MISSOURI BOTANICAL GARDEN This species differs from Т. lanceolatum in its more robust growth habit, the densely tuberculate-hispid leaf sheaths, and the large numbers of spikes in the inflorescence. From Т. fas- ciculatum it differs mainly in the hispid character of the leaf sheaths but the pedicels of T. fasciculatum are, in addition, usually shorter and thicker, the number of spikes smaller, and the plant less robust. open grassy ground among rocks, steep slope of ravine, Alzada, 21 Sept. 1910, 7 Tada ck 7088, or Amer. Gr. Nat. Herb. No. 231 (Gray, MBG, NY, USNH). NGO: Sierra Madre, alt. 5200 ft., 15 Aug. 1897, J. N. Rose 3513 (USNH). GUANAJUATO: garden of college, Guanajuato, 1901, Alfredo Dugès (Gray). JALISCO: frequent widely separated clumps, 3.7 m. high, 0.5 & 5.0 km. north of Tonilita, alt. 1000 m., 9 Oct. 1940, H. C. Cutler 4013 (MBG) & 4019 (Ba, F, Ha, MBG, USDA); песма on walls of bar а 1 km. northeast of Ciudad Guzman (Zapotlan), alt. 1520 m., 19 Oct. 1940, s i 4079, 4080, 4081, 4082 (MBG); same data, 4089 (Ha, MBG); same 0106006 у, 22 Oct. 1940, Cutler 4090 (Ha, MBG) ; same data, 4091, 4092 (MBG) ; 3 m. tall, frequent along ereek, northeast edge of Ciudad Guzman, alt. 1520 m., 22 Oet. 1940, Cutler 4110 (Ha, MBG), 4111 (Ba, Ha, MGB, USDA) ; inflorescence collections from single clones 1 km. northeast of Ciudad Guzman, 22 Oct. 1940, Cutler 4117, 4118 (MBG); side of Ваггапса Oblato, Guada- lajara, alt. 5000-6000 ft., 27-28 Sept. epis Hitchcock 7348, 7361, 7366 (USNH) ; La Barranea, Guadalajara, 19 gi 1930, M. E. Jones 27629 (MBG); Barranca de Oblatos, Guadalajara, 15 Oet. La. ғ Зан фа. №. Collins (USNH Nos. 1064498, 1064499, ШП iis) — Etzatlan, 22 Oct. 1921, Kempton $ Collins (USNH) ; tlan, 23 Oct. , Kempton $ Collins (USNH) ; Los Teo- sintes, Ampaso, 23 ч p dh ade рн ан (USNH) ; Ciudad Guzman, 27 Oct. 1921, Kempton 4 Collins (USNH); Arupara, 23 Oct. 1921, Kempton $ Collins (USNH); Rio Blanco, June-Sept. 1886, Ed. Palmer (Gray, NY, Yale, USNH) ; hills near Guadalajara, 29 June 1889, Pringle 2611 (USNH) ; canyons near Guada- lajara, 3 Dec. 1889, Pringle 2623 (USNH); road between Coatlan and Bolaños, 7-9 Sept. 1897, Rose 2841 (Gray, USNH түрк); Rose $ Robert Hay 6278 (USNH). MICHOACAN: Cerro San Miguel, near Morelia, 10 Feb. 1912, Arséne 9938 (USNH). NAYARIT: Los Fresnos, Tepie, 2 Oct. 1923, Collins E Kempton T34 (USNH); Jala, near Ahuaeatlan, 7 Oct. 1923, Collins $ Kempton 80 (USNH) ; Cerro del Cruz Tepic, 1 Oct. 1923, Kempton $ Collins (USNH Nos. 1646091—1646093, 1646115- 1646118). OAXACA: Cerro San Felipe, alt. 1800 m., Distrito del Centro, 12 Aug. 1906, C. Con- zatti 1615 (USNH) ; Las Sedas, Distrito de Etla, alt. 2000 m., 29 Aug. 1909, Con- gatti 2523 (USNH) ; Cañada San Juan, Zimatlan, alt. 1925 m., 2 Oct. 1931, Conzatti 4638 (Mich). SAN LUIS POTOSI: limestone ledges, Tinanul, 24 July 1891, Pringle 3993 (USNH) ; hills, Las iron 24 July 1891, Pringle 3993 (USNH). GUATEMALA: Kalkberge, Quien Santo, Huehuetenango, 23 Aug. 1896, E. Seler 2723 punt USNH). 1941] CUTLER & ANDERSON—THE GENUS TRIPSACUM 267 7. Tripsacum latifolium Hitche., Bot. Gaz. 41: 294. 1906. It is only with diffieulty that some of the herbarium speci- mens of this species may be distinguished. The petiolate char- acter of the large lower leaves seems to be distinctive in this species, but most collectors take only the upper portions of the plant with leaves small enough to put on a herbarium sheet. The rarely collected complete plant series, as Bartlett 11888, serve to connect the extremes of the species found in the West Indies, as Ekman 16226, and the terminal collections with the type. West Indian plants are very slender, small, with solitary terminal inflorescences, usually deeply colored with purple. They must, however, be considered as a variant of the species not worthy of distinction at this time. British HONDURAS: 10 feet tall, edge of ravine, Mountain Pine Ridge, El Cayo District, 1 March 1931, H. H. Bartlett 11888 (Mich, МҮ); near river bank, El Cayo District, Vaca, 4 March 1938, P. H. Gentle 2298 (USNH) GUATEMALA: 18 ft. high, near Seeanquim, trail to Cahabon, 25 Nov. 1904, G. P. Goll 44 (USNH) ; Sierra del Mieo, between Los Amates & Izabal, alt. about 750 ft., 23 Feb. 1907, W. A. Kellerman 6242 (USNH) ; between San Marcos and San An- dreas, alt. 4500 ft., 2 Nov. 1935, Kempton ф Popenoe (MBG, USNH) ; La Libertad, Petén Distriet, 18 April 1933, C. L. Lundell 2836 (Mich); vie. Seeanquim, Alta Verapaz, alt. 550 m., 6 e s H. Pittier 261 (USNH); uu. ot Dept. Alta Verapaz, alt. 350 m., Jan. , H. von Tuerckheim 8333 (U YPE) ; 12-15 ft. tall, rich upland iub Mice nor of Quirigua, 2 March pe E: eim 90 (1703 я Honpuras: Tela, 14 Feb. 1931, Collins $ Kempton (USNH); plants 6—15 ft. high, forming dense eolonies, in open swamp, Uluita Station, 24 Jan. 1928, Standley 54941 (USNH); San Pedro Sula, Dept. Santa E alt. 1600 ft., 1887, Carl pn 5595% (USNH); rich soil along river, 6 km. west of Siguatepeque, 1200 m. , T. G. Yuncker, В. Е. Dawson, H. Е. Youse, 6391 (Mich, MBG, USNH). gon cultivated, 1931, pies Expedition (MBG, USNH). WEST INDIES: HAITI: open slopes, vicinity of Mission, Fonds Varettes, alt. about 1000 m. and above, 17 April-4 May 1920, E. C. Leonard 3945 (Gray, USNH) ; dry ravine north- east of West Indies Co. Plantation, vieinity of St. Michel de l'Atalye, Dept. du Nord, 350 m., 18 Nov. 1925, Leonard 7157 (USNH). SANTO DOMINGO: Cordillera Septentrional, Prov. Santiago, Santiago, Cuesta de Piedras, alt. 200 m., 23 Nov. 1930, E. L. Ekman 16229 dics ; vast — steep slope near the top: R. р, Cordillera Central, Prov. Santo Domingo, У Ша graeia, Loma Marian Chicle, 800 m., 6 Jan. 1930, Ekman pie (USNH); prope La Salinas in via do Los Cerareos, 700 m. ай., Prov. Barahona, Sept. 1911, Padre Miguel Fuertes 1424 (NY). TRINIDAD: La Brea, 9 March 1915, W. E. Broadway 4982 (USNH). (VoL, 28 268 ANNALS OF THE MISSOURI BOTANICAL GARDEN EXCLUDED SPECIES The taxonomy of Rottboellia and Manisuris is in such con- fusion that, for the most part, we have not been able to go be- yond the ‘Index Kewensis’. Mistakes in copying generic names (Index Kew. Suppl. 2: 187. 1904; and Kunth, Enum. P1: 1: 467, 602. 1833), mistakes in identification (Schlecht. & Cham., Linnaea 6: 40. 1831), adoption and listing of herbarium label names (Urbina, Cat. Pl. Mex. 376. 1897), and other errors have introduced names which must be included in this list of excluded species or under synonymy. It is unfortunate that many of these mistakes were published in the ‘Index Kewensis.’ Tripsacum aegilopoides Kunth, Enum. Pl. 1: 467, 602. 1833 - Rottboellia hirsuta Vahl, Symb. Bot. 1: 11. 1790. T. aristatum Rasp., Ann. Sei. Nat. I, 5: 306. 1825 - Ischae- mum aristatum L. Sp. Pl. 1049. 1753. T. avenacea Scribn. & Merr. (U. S. Dept. Agric. Div. Agrost. Bull. 24: 23. 1901) aec. to Index Kew. Suppl. 2: 187. 1904 - Tristachya avenacea. T. ciliare Rasp., Ann. Sci. Nat. I, 5: 306. 1825 - Elionurus tripsacoides НВК. Nov. Gen. et Sp. 1: 192. 1. 62. 1816. T. compressum Rasp., Ann. Sci. Nat. I, 5: 306. 1825 = Hemar- thria compressa R. Br., Prodr. 207. 1810. T. cylindricum Michx., Fl. Bor. Amer. 1: 60. 1803 - Mani- suris cylindrica Kuntze, Rev. Gen. Pl. 2: 779. 1891. T. distachyum Poir., Кисус. 8: 114. 1808 = Ischaemum ru- gosum Salisb., Ie. Stirp. Rar. 1: /. 1. 1791. T. distichum Rasp., Ann. Sci. Nat. I, 5: 306. 1825 - Ischae- mum rugosum Salisb., Ic. Stirp. Rar. 1: f. 1. 1791. T. fasciculatum Rasp., Ann. Sci. Nat. I, 5: 306. 1825 = Chloris radiata Sw., Prod. Veg. Ind. Oce. 26. 1788. T. giganteum Rasp., Ann. Sci. Nat. I, 5: 306. 1825 = An- thistiria gigantea Cav., Ic. 5: 35. 1799. T. granulare Rasp., Ann. Sci. Nat. I, 5: 306. 1825 = Hacke- lochloa granularis Kuntze, Rev. Gen. Pl. 2: 776. 1891. T. hermaphroditum L. Syst. Nat. ed. 10. 1261. 1759 - Anthe- phora hermaphrodita Kuntze, Rev. Gen. Pl. 2: 759. 1891. 1941] CUTLER & ANDERSON—THE GENUS TRIPSACUM 269 T. hirsutum Rasp., Ann. Sci. Nat. I, 5: 306. 1825 - Rottboel- lia hirsuta Vahl, Symb. Bot. 1: 11. 1790. T. laxa Seribn. & Merr. (U. S. Dept. Agric. Div. Agrost. Bull. 24: 23. 1901) aec. to Index Kew. Suppl. 2: 187. 1904 = Tristachya laxa. T. mucronatum Rasp., Ann. Sci. Nat. I, 5: 306. 1825 = Tra- chys mucronata Pers. Syn. 1: 85. 1805. T. muticum Rasp., Ann. Sci. Nat. I, 5: 306. 1825 - Ischae- mum muticum L. Sp. Pl. 1049. 1753. T. myuros Rasp., Ann. Sci. Nat. I, 5: 306. 1825 = Rottboellia Myurus Benth., Journ. Linn. Soc. 19: 68. 1881. T. paniceum Rasp., Ann. Sci. Nat. I, 5: 306. 1825 - Pogon- atherum saccharoideum Beauv., Agrost. 176. #. 11. f. 7. 1812. T. pubescens Lichenst. ex Nees, Fl. Afr. Austr. 1: 74. 1841 - Anthephora pubescens Nees, Fl. Afr. Austr. 1: 74. 1841. T. pubescens Willd. ex Steud., Nom. Bot. ed. 2, 1: 317. 1841 - Cenchrus pubescens Steud., Nom. Bot. ed. 2, 1: 317. 1841. T. semiteres Wallich, Cat. no. 8628. 1828 - Polytoca semi- teres Benth. in Benth. & Hook. f., Gen. Pl. 3: 1113. 1883. MISCELLANEOUS NEW ASCLEPIADACEAE AND APOCYNACEAE FROM TROPICAL AMERICA ROBERT E. WOODSON, JR. Assistant Curator of the Herbarium, Missouri Botanical Garden Assistant Professor in the Henry Shaw School of Botany of Washington University APOCYNACEAE RavworriA sarapiquensis Woodson, spec. nov. Arbor са. 15 m. alta omnino glabra; ramulis erassiusculis post exsicca- tionem plus minusve angulatis cortice luteo-brunneo tectis, internodiis 3-4 em. longis. Folia ternata inaequalia in axillis glandulosa, petiolis 1.5-2.5 em. longis; laminis obovato-ellip- ticis abrupte acuminatis basi sensim attenuatis 14-20 em. longis 5.0-6.5 em. latis coriaceis, venis secondariis subhori- zontalibus. Inflorescentia foliis dimidia brevior subterminalis ternata, pedunculo primario ca. 6 em. longo deinde repetite 4- vel 3-natim composito. Flores albi; pedicello 0.4 em. longo; calycis laciniis ovato-subreniformibus rotundatis 0.25-0.3 em. longis; corollae tubo cylindrico 0.5 em. longo са. 0.15 em. dia- metro, lobis oblongis rotundatis adscendentibus 0.4 em. longis са. 0.15 em. latis; staminibus prope fauces affixis 0.1 ст. longis apiculatis; ovario synearpo 0.15 em. longo nectarium annulare minute crenatum ca. bis superante; stylo gracili 0.15 em. longo, stigmate capitato 0.05 em. longo dense papillato apice minute 2-lobato. Ваесае ignotae.—Costa Rica: Vara Blanca de Sara- piquí, north slope of Central Cordillera, between Poás and Barba volcanoes, alt. 1460 m., Febr., 1938, А. Е. Skutch 3556 (U.S. Nat. Herb., TYPE). Somewhat intermediate between the sections Grandiflorae and Andinae, and not obviously related to any species known to occur in Central America. Quadricasaea Woodson, gen. nov. (Apocynaceae-Plume- roideae-Tabernaemontaninae). Calyx inaequaliter 5-partitus intus multiglanduligerus. Ovarii carpellae liberae ovariis Issued September 20, 1941. (271) [VoL. 28 212 ANNALS OF THE MISSOURI BOTANICAL GARDEN numerosis multiseriatim positis neetario annulato valde ad- nato in stylo gracili gradatim productae stigmate subumbrac- uliformi obseure 2-lobato. Corolla salverformis limbi lobis 5 sinistrorse obtengentibus. Antherae sessiles anguste sagit- tatae omnino fertiles.—F'rutiees volubiles. Folia opposita eglandulosa. Inflorescentia bostrycino-racemosa aut lateralis aut terminalis. Genus in honorem el. J. Cuatrecasatis dedi- eavitur. Species typica prima sequitur. Quapricasaka inaequilateralis Woodson, spec. nov. Frutex volubilis sat magnus. Folia opposita breviter petiolata late elliptiea apiee breviter acuminata basi inaequilateraliter ob- tusa 20-30 em. longa 7.5-9.0 em. lata firme membranacea gla- bra, petiolis 0.4 em. longis. Inflorescentia lateralis bostry- cino-racemosa, pedunculo ca. 2 em. longo, pedicellis ca. 1 em. longis, omnino glabra; ealycis lobis valde inaequalibus ovatis vel late ovato-oblongis apice rotundatis 0.7-1.0 em. longis ut dicitur subpetalaceis albidis glabris; corollae flavae extus gla- brae tubo ca. 3.25 em. longo basi ca. 0.45 em. diam. paulo infra medium usque 0.2 em. diam. constricto ibique staminigero deinde gradatim ampliato, faucibus ca. 0.4-0.45 em. diam., lobis oblique elliptico-oblongis obtusis ca. 2.5 em. longis patulis; antheris angustissime sagittatis acuminatis basi acutissime 2-lobatis sessilibus glabris ca. 0.6 em. longis ; ovariis ovoideis glabris са. 0.3 ст. longis in stylo gracili gradatim produetis, nectario annulato adnato ca. 0.1 em. alto, stigmate са. 0.1 em. longo. Fructus ignoti—CoLomBIA: COMISARIA DEL CAQUETA: Florencia, entre matorrales residuales de monte, alt. 400 m., March 29, 1940, J. Cuatrecasas 8814 (U. S. Nat. Herb., TYPE). Quapricasaka caquetensis Woodson, spec. nov. Frutex volu- bilis sat magnus. Folia opposita breviter petiolata elliptico- oblonga apice anguste acuminata basi aequilateraliter acuteque cuneata cum petiolo 14—20 em. longa 2.5-5.5 em. lata firme membranacea glabra. Inflorescentia subterminalis bos- trycino-racemosa pauciflora, pedunculo са. 0.5 em. longo, pedicellis ea. 1 em. longis, omnino glabra; calycis lobis valde inaequalibus ovatis vel late ovato-oblongis apice rotundatis 1941] WOODSON—ASCLEPIADACEAE AND APOCYNACEAE 273 0.7-1.2 em. longis ut videntur subpetalaceis albidis glabris; corollae flavae extus glabrae in alabastro sat maturo tubo са. 3 em. longo.—CoLoMBIA: COMISARIA DEL CAQUETA: Cordillera Oriental, vertiente oriental, Sucre, bosques entre 1000 y 1300 m. alt., April 4, 1940, J. Cuatrecasas 9062 (U.S. Nat. Herb., TYPE). Quadricasaea is particularly outstanding because it is the only genus of the American Tabernaemontaninae with the habit of a liana, all other genera being trees or shrubs. The structure of the corolla, inflated at base and throat and con- stricted at the insertion of the stamens, is also noteworthy. Although it appears somewhat dangerous, upon short acquaint- anee, to deseribe two species of such close affinity from the same general region, the leaves of Q. inaequilateralis and Q. caquetensis are so dissimilar that it is quite logical to do so. The inflorescences are probably specific as well. ManpEviLLA јаѕтіпійога Woodson, spec. nov. Frutex са. 3 m. altus; ramulis crassiusculis teretibus valde lenticellatis glabris. Folia opposita petiolata oblongo-elliptica acuminata basi obtusa 3.5-8.0 em. longa 1.0-1.25 em. lata subcoriacea margine post exsiccationem revoluta supra glabra nervo medio in longitudinem sparse glanduligero subtus inconspicue pi- losula, petiolo ca. 0.8-1.5 em. longo glabro. Inflorescentia lateralis racemosa pluriflora, peduneulo ea. 3.0-8.5 ст. longo parce pilosulo, pedicellis tenuibus ea. 0.7 em. longis pilosulis, bracteis scariaceis minimis; calycis laciniis ovato- lanceolatis acutis extus ca. 0.2 cm. longis pilosulis, squamellis oppositis multifissis; corollae flavae campanulatae extus glabrae tubo 0.3 cm. longo basi ca. 0.1 cm. diam. faucibus ca. 0.3 cm. diam. intus pilosulo, lobis fere erectis ovato-ellipticis acutis ca. 0.35 em. longis; antheris 0.2 em. longis basi obscure 2-lobatis gla- bris; ovariis ovoideis ca. 1 em. longis glabris, nectariis 5 ovariis fere aequilongis, stigmate umbraeuliformi longe apiculato ea. 0.2 em. longo. Follieuli ignoti.—CoroMBiA: carpas: Quindio, alt. 2900 m., Febr., 1937, E. Dryander 2145 (U. S. Nat. Herb., TYPE). Although clearly in the subgenus Exothostemon, this species is quite unlike any other known at present, both because of its relatively large, shrubby habit and because of the small [Vor. 28 214 ANNALS OF THE MISSOURI BOTANICAL GARDEN flowers of unusual shape which recall the Mexican M. Syring Woodson, а Папа of the subgenus Humandevilla. MawpEviLLA nerioides Woodson, spec. nov. Frutex ut dicitur altitudine ignotus ; ramulis erassiusculis tenuiter alatis glabris cortice rubro-brunneis. Folia opposita brevissime petiolata lineari-lanceolata acuminata basi obscure cordata eum petiolo 4—7 em. longa 0.3-0.6 em. lata subcoriacea glabra nervo medio supra in longitudinem pauciglanduligero. In- florescentia terminalis subspicata pauciflora, pedunculo 2.0- 4.5 em. longo glabro, pedicellis са. 0.2 ст. longis, bracteis minutissimis; calycis laciniis ovatis anguste obtusis ca. 0.25 cm. longis glabris, squamellis oppositis multifissis; corollae speciosae albido-roseae extus glabrae tubo proprio 1.5 cm. longo basi ca. 0.1 cm. diam., faucibus conico-campanulatis 2 cm. longis, ostio ca. 1.25 cm. diam., lobis oblique obovatis 2 cm. longis patulis; antheris 0.45 cm. longis ellipticis basi brevissime auriculatis glabris; ovariis ovoideis ca. 0.1 cm. longis, nectariis 5 ovaria ca. dimidia aequantibus, stigmate umbraculiformi brevissime apiculato ca. 0.1 cm. longo. Fol- liculi 12той.—СотомвтА: RIO GUAVIARE: San José del Guaviare, terrenos graniticos, 270 m. alt., Nov. 12, 1939, J. Cuatrecasas 7674 (U. S. Nat. Herb., түрк). Closely related to M. lancifolia Woods., but differing in its stouter habit, larger flowers, and broader leaves with cordate bases. MaLovETIA Cuatrecasatis Woodson, spec. nov. Arbuscula altitudine ignota; ramulis crassiusculis glabris maturitate cortice griseo-brunneo bene lenticellato. Folia opposita petio- lata oblongo-elliptica apice obtuse acuminata basi late cuneata 4—6 cm. longa 1-2 cm. lata omnino glabra supra nitidula subtus ораса in axillis nervi medii rare foveata, petiolo 0.4-0.5 em. longo. Umbellae laterales 1— vel pauci-florae, pedicellis 1.2 em. longis glabris; calycis laciniis ovatis rotundatis vel ob- tusis 0.2-0.25 em. longis subfoliaceis glabris vel indistincte papillaltis, squamellis solitariis alternatis; corollae salver- formis albidae tubo са. 1 em. longo basi ca. 0.1 ст. diam. sub medio usque 0.25 em. diam. dilatato deinde gradatim contracto, 1941] WOODSON-—ASCLEPIADACEAE AND APOCYNACEAE 275 faucibus abrupte campanulatis са. 0.2 ст. diam., ostio con- spieue ealloso-annulato, lobis elliptico-ovatis obtusis 1.2 cm. longis extus papillatis intus dense minuteque puberulis; an- theris valde exsertis са. 0.2 ст. longis dense puberulo-papil- latis; ovariis ca. 0.15 cm. longis puberulo-papillatis, nectariis concrescentibus ovaria subaequantibus. Folliculi ignoti.— CotomBia: vauPÉs: Mitu, bosque, 200 m. alt., Oct. 20, 1939, J. Cuatrecasas 7285 (U. S. Nat. Herb., түрк). Closely allied to M. lata Mgf. of the lower Amazon Valley, but differing in the blunt calyx lobes of heavier texture and in the longer, narrower eorolla tube. ASCLEPIADACEAE Oynancuum (Meraeris) subpaniculatum Woodson, spec. nov. Frutex volubilis; ramulis crassiusculis glabris. Folia longe petiolata late ovato-cordata sinu aperto apice abrupte acuteque subcaudato-acuminata 14-16 em. longa 10-12 cm. lata glabra firme membranacea nervo medio supra basi glanduligero, petiolo 9-10 em. longo glabro. Inflorescentia axillaris foliis ca. duplo longior bostrycine spicato-paniculata mutiflora, pedunculis sparse pilosulis; calycis laciniis ob- longo-elliptieis aeutis 0.35 em. longis extus minute papillatis; corollae rotatae lobis ovato-elliptieis acuminatis 0.4—0.45 cm. longis extus papillatis intus papillatis prope apicem conspicue barbatis plus minusve retro-revolutis; gynostegio breviter stipitato apice са. 0.5 cm. diam. stigmate umbonato; polliniis leviter reniformibus ca. 0.09 em. longis, caudiculis gracilibus ea. 0.1 em. longis, corpusculo minuto; corona pentagulo-patelli- formi ea. 0.5 em. diam. intus dense papillata, segmentis late emarginatis. Follieuli ignoti—Cotomsia: vicinity of Santa Marta, alt. 2000 ft., June, 1898-1901, H. H. Smith 2410 (Herb. Missouri Bot. Gard., TYPE). Metalepis cubensis Griseb. (Cynanchum cubense (Griseb.) Woods.), under which this specimen was distributed, differs conspieuously in the smaller, nearly triangular leaves, in the uniformly papillate corollas, and in structural characters of the gynostegium and pollinia. A discussion of Metalepis as a ГУоп, 28 216 ANNALS OF THE MISSOURI BOTANICAL GARDEN subgenus of Cynanchum will be found in Ann. Missouri Bor. Garp. 28: 213-214. 1941. Cynancuum (Merauepis) Haughtii Woodson, spec. nov. Frutex volubilis graciliusculus; ramulis minutissime pilosulis glabratisve. Folia petiolata ovato-cordata basi vix lobata fere rotundato-subtruncata 9-12 cm. longa 7-8 сш. lata apice abrupte angusteque acuminata glabra membranacea nervo medio supra basi glanduligero, petiolo 2.5-3.0 em. longo mi- nutissime pilosulo vel glabro. Inflorescentia axillaris panicu- lata multiflora foliis ca. dimidia longior, pedunculis dense puberulis, pedicellis 0.4 em. longis puberulis; calycis laciniis oblongis obtusis 0.3 em. longis sparse pilosulis; corollae ro- tatae viridulae lobis ovatis acutis ca. 0.25 em. longis supra subcucullatis prope apicem incrassatis ibique minute pilosulis caeterumque papillatis vel glabris; gynostegio subsessili apice ca. 0.2 em. diam., stigmate umbonato; polliniis leviter reni- formibus са. 0.075 em. longis, caudiculis 0.04 em. longis, cor- pusculo minuto; corona pentagulo-rotata ca. 0.295 em. diam., segmentis angustatis emarginatis conduplicatis inflexis Fol- liculi ignoti.—Ecvapon: roadside near Pajan, alt. ca. 200 m., Dec. 10, 1939, O. Haught 2985 (Herb. Missouri Bot. Gard., TYPE). Outstanding in the subgenus because of the small flowers and peculiar subeucullate corolla lobes. MarELEA (НеглозтеммА) inops Woodson, spec. nov. Frutex volubilis; ramulis graciliuseulis ferrugine pilosis, internodiis sat elongatis. Folia opposita petiolata ovata apice acuminata basi late cordata 5.5-9.0 em. longa 3-5 em. lata membranacea utrinque ferrugine puberulo-papillata, petiolis 2.5-6.0 em. longis similiter vestitis. Inflorescentia extra-axillaris um- belliformis pluriflora, peduneulo 0.5-0.7 em. longo, pedicellis 1.5-2.0 em. longis puberulo-papillatis. Calycis lobi ovati acuti 0.4 em. longi puberulo-papillati pilis longioribus interspersis. Corolla rotata ut creditur viridula ca. 2 em. diam. extus in- tusque sparse pilosula; lobis ovatis acutis ca. 0.6 em. longis. Corona simplicissima disciformis discolorata ca. 0.35 em. diam. sub antheris inconspicue ligulata. Gynostegium sessile; 1941] WOODSON—ASCLEPIADACEAE AND APOCYNACEAE 277 stigmate late 5-gono ca. 0.08 em. diam. depresso; antheris sub stigmate positis, polliniis obpyriformi-subfaleiformibus valde excavatis cum caudiculo plieate alato fere aequilongo ca. 0.07 ст. longis, corpusculo mediocri—MeExico: cHiaras: Cascada Siltepec, Aug. 5, 1937, E. Matuda 1731 (Herb. Missouri Bot. Gard., түрк; Herb. Univ. Michigan, 1soryprs). Noteworthy amongst the other liana species of subgen. Heltostemma be- cause of the very simple discoid corona, somewhat resembling that of the fruticulose M. caudata. MarELEA (HELIOSTEMMA) tinctoria Woodson, spec. nov. Frutex volubilis; ramulis graciliusculis glabris, internodiis sat elongatis. Folia opposita longe petiolata late ovata apice abrupte angusteque subcaudato-acuminata basi rotundata vel obseurissime cordata 14-17 em. longa 7-11 em. lata mem- branacea glabra post exsiccationem sordide purpureo-discol- orata, petiolis ca. 5 em. longis. Inflorescentia extra-axillaris longe peduneulata racemiformis pauciflora, pedunculo ca. 8 em. longo, pedicellis ca. 2 cm. longis dense minuteque fer- rugineo-papillatis; calycis lobis ovato-lanceolatis acuminatis ca. 1.2 em. longis extus minutissime ferrugineo-papillatis; corolla eampanulato-rotata viridula glabra vel extus indis- tincte papillata post exsieeationem dilute livido-discolorata, faucibus late campanulatis ca. 0.5 em. altis ca. 1.2 cm. diam, lobis triangularibus acutis ca. 1.5 em. longis patulis; gynostegio breviter stipitato, stigmate late 5-gono ca. 0.35 cm. diam.; antheris sub stigmate positis sed appendicibus apiealibus cero- eandidis super eo convergentibus, polliniis triangulo-falei- formibus in eaudieulo gradatim attenuatis ca. 0.1 em. longis, eorpuseulo mediocri; corona carnosa rotata simplici profunde 5-lobata per partitiones 5 earnosas ad gynostegium annexa. Folliculi ignoti—Cosra Rica: Vara Blanca de Sarapiquí, north slope of Central Cordillera, between Pods and Barba vol- canoes, alt. 1890 m., Febr., 1938, А. Е. Skutch 3589 (U.S. Nat. Herb., түрк). Closely related to M. picturata (Hemsl.) Woods. and M. Pittieri (Standl.) Woods., but probably more closely to the latter, from which it is distinguished by its much larger flowers and leaves as well as by the more simple structure of [Vor. 28 278 ANNALS OF THE MISSOURI BOTANICAL GARDEN the corona. The leaves of these three species, as well as others of the subgenus Heliostemma, produce a peculiar blackish purple dye upon wilting, as do the flowers also to a lesser ex- tent. The peculiar whitened anther appendages are noteworthy also. Matera (EuMATELEA?) cynanchiflora Woodson, spec. nov. Suffrutex volubilis dense ferrugineo-pilosulus; ramulis gra- ciliusculis, internodiis sat elongatis. Folia opposita oblongo- elliptiea apice anguste acuminata basi obtusa 6.0-10.5 em. longa 2-4 cm. lata membranacea utrinque ferrugineo-pilosa, petiolis 1.0-1.25 em. longis. Inflorescentia extra-axillaris ses- silis umbelliformis pluriflora, pedicellis 1.0—1.3 em. longis fer- rugineo-pilosis; calycis lobis ovatis acuminatis 0.15 em. longis extus ferrugineo-pilosulis; corolla rotata luteo-albida ca. 1.2 em. diam., lobis late ovatis rotundatis ca. 0.4 em. longis extus sparse pilosulis intus minute papillatis; corona cyathiformis ca. 0.15 em. alta 0.3 em. diam. margine subintegra per parti- tiones 5 earnosas ad gynostegium annexa; gynostegio mani- feste (ca. 0.15 em.) stipitato, stigmate late 5-gono ca. 0.2 em. diam.; antheris sub stigmate positis, polliniis oblique obpyri- formibus fere subfaleiformibus eum caudiculo ea. 0.075 em. longis. Follieuli ignoti.—Cosra Rica: vicinity of El General, Prov. San José, alt. 640 m., Jan., 1939, А. F. Skutch 4071 (U.S. Nat. Herb., түрк). The deep cyathiform corona is quite un- usual for Matelea, and the whole superficial aspect of the plant suggests Cynanchum. The pollinia, on the other hand, show the species to be Gonoloboid. It is not closely related to any other known species of Humatelea, and may eventually neces- sitate the erection of a separate subgenus. MarELEA (EuwATELEA) Steyermarkii Woodson, spec. nov. Frutex volubilis ; ramulis graciliusculis ferrugineo-pilosis tan- dem glabratis, internodiis sat elongatis. Folia oblongo-ovata apiee anguste acuminata basi sat indistincte cordata 5-9 em. longa 1.5—3.0 em. lata membranacea utrinque ferrugineo-pilosa, petiolis 1.5-2.5 em. longis ferrugineo-pilosis. Inflorescentia umbelliformis pauciflora, peduneulo 2.5 em. longo, pedicel- 1941] WOODSON—ASCLEPIADACEAE AND APOCYNACEAE 279 lis 1.5 em. longis ut in pedunculo ferrugineo-pilosis; calycis laciniis oblongo-ovatis acuminatis 0.7-0.8 em. longis extus fer- rugineo-pilosulis; corolla rotata alba venis venulisque insig- niter viridibus extus intusque glabra vel minutissime papillata, lobis ea. 1.2 em. longis late ovatis rotundatis patulis; gyno- stegio 0.4 em. alto; antheris sub stigmate positis, polliniis exca- vato-pyriformibus cum caudiculo pellucido са. 0.1 em. longis, corpusculo medioeri; corona complicata urceolata ca. 0.3 ст. alta ca. dimidia superiori profunde multifissa basi paulo in- flata integra (an corollae fauces ?) minutissime papillata per digitos 9 carnosos anguste oblongos ca. 0.2 em. longos ad gyno- stegium annexa. Folliculi ignoti.—GUATEMALA: DEPT. SAN MARCOS: upper south-facing forested slopes of Volcán Taju- mulco, between Canojas and top of ridge, 7 mi. from San Se- bastián, alt. 3300—3900 m., Febr. 16, 1940, Steyermark 35810 (Herb. Missouri Bot. Gard., rvrE).— This species well il- lustrates the tremendous complexity of the corona of most species of Matelea. The radial partitions of the corona are so conspicous that they are apt to be mistaken for the dorsal anther appendages of Gonolobus, from which they are quite distinet morphologieally. Marteka (EuMaATELEA $ RETICULATAE) tenuis Woodson, spec. nov. Frutex volubilis; ramulis tenuibus inconspicue pilosulis, internodiis sat elongatis. Folia opposita elliptico-oblongata apice acuminata basi obtusa 5-7 cm. longa 1.5-2.3 em. lata delicate membranacea glabra; petiolis 0.5-0.7 em. longis incon- spicue pilosulis. Inflorescentia extra-axillaris racemiformis pauciflora, peduneulo ea. 0.3-0.5 em. longo minute papillato, pedicellis ca. 0.5 em. longis minute puberulo-papillatis. Flores ut dicuntur virides. Calycis lobi late ovati acuti 0.25 em. longi extus puberulo-papillati. Corolla rotato-subcampanulata са. 1 em. diam.; lobis ovatis obtusis ca. 0.3 em. longis extus in- tusque minute papillatis. Corona patelliformis obscure 5- lobata caeterumque integra ca. 0.4 cm. diam. per partitiones 9 latas adscendentes ad gynostegium annexa. Gynostegium са. 0.2 em. stipitatum; stigmate late 5-gono са. 0.2 ст. diam. ; an- theris sub stigmate positis, polliniis horizontalibus obpyri- [Vor. 28 280 ANNALS OF THE MISSOURI BOTANICAL GARDEN formibus profunde excavatis eum caudieulis ca. 0.1 em. longis, corpusculo mediocri. Folliculi ignoti—GuaTEMALA: IZABAL: vicinity of Quirigua, alt. 75-225 m., May 15-81, 1922, P. С. Standley 24036 (Gray Herb., түрк). Outstanding amongst the species of $ Reticulatae because of the thin, narrow leaves and the small, subeampanulate corollas. MarELEA (EvuMarELEA) violacea Woodson, spec. nov. Frutex volubilis; ramulis ferrugineo-pilosis, internodiis sat elon- gatis. Folia late oblongo-ovata apice anguste acuminata basi late cordata 9-13 em. longa 4-6 cm. lata membranacea utrinque pilosa, petiolis 3.0-3.5 em. longis ferrugineo-pilosis. Inflores- centia subsessilis pauciflora, pedunculo ca. 0.5 em. longo piloso, pedicellis ea. 1 em. longis pilosis; calycis laciniis late ovatis anguste acuminatis extus ferrugineo-pilosis; corolla rotata violacea venis venulisque reticulatis brunneo-purpurascent- ibus extus laxe ferrugineo-pilosa intus minute papillata, lobis late ovatis obtusis ca. 0.5 em. longis patulis; gynostegio ses- sili; antheris pro parte sub stigmate positis, polliniis excavato- pyriformibus cum caudiculis latis pellucidis ca. 0.15 em. longis, corpusculo mediocri, stigmate subplano late 5-gono са. 0.3 em. diam. ; corona rotata patula late 5-lobata са. 1 em. diam. com- plieate multifissa per digitos 5 ineonspieuos ad gynostegium annexa. Folliculi ignoti.—GUvATEMALA: DEPT. SAN MARCOS: be- tween Canjulá and La Unión Juárez, near southeast portion of Volcán Тасапа, alt. 2000-3000 m., Febr. 22, 1940, J. А. Steyer- mark 36445 (Herb. Missouri Bot. Gard., түрк; Herb. Field Mus., isorvrE).—Beside the peculiar coloration of the corolla, this species is distinguished by much the same fimbriation of the corona as in M. Steyermarkii, but in rotate form and with very much less conspicuous radial partitions. MarkLEA (ГлвтрозтеимА) Hintoniana Woodson, spec. nov. Frutex volubilis sat validus; ramulis ferrugine strigoso-pilosis, internodiis sat elongatis. Folia opposita late ovata apice acuminata basi late cordata 6-10 em. longa 3.5-6.5 em. lata membranacea utrinque pilosula, petiolis 4.5-6.0 cm. longis sparse pilosulis. Inflorescentia extra-axillaris laxe racemi- 1941] WOODSON—ASCLEPIADACEAE AND APOCYNACEAE 281 formis pluriflora, peduneulo 3.0-3.5 em. longo pilosulo, pedi- cellis 2.5-3.0 em. longis similiter vestitis; bracteis ovato- lanceolatis 0.5-0.7 ст. longis subfoliaceis pilosulis; calycis lobis elliptico-lanceolatis acuminatis са. 1.2 ст. longis sub- foliaceis pilosulis; corolla late eampanulato-rotata fere plana ca. 4 em. diam. speciosa praecipue ad lobos roseo-tincta et more subgen. Dictyanthi reticulata; lobis late triangularibus acutis ca. 1 em. longis; corona complicate 5-lobata, lobis 3-lobulatis more M. Ощтози (Labidostelma guatemalense) compositis; gynostegio subsessili, stigmate obtuse 5-gono depresso. Fol- liculi ignoti—Mexico: Bejucos, alt. 610 m., District of Temas- ealtepee, Mexico, Aug. 24, 1932, G. B. Hinton 1450 (Gray Herb., TYPE). Closely related to M. Qwrosu (Standl.) Woods. ( La- bidostelma guatemalense Schltr., which has somewhat smaller flowers with ovate-acuminate lobes, not conspicuously reticu- late as in M. Hintomana. MarELEA (Isara 7) glaberrima Woodson, spec. nov. Frutex volubilis glaberrimus ; ramulis validiusculis, internodiis longis- simis. Folia ovato-elliptica apice acuminata basi rotundata 13-18 em. longa 5-11 em. lata; petiolis 1-3 em. longis. Inflores- centia umbelliformis pauciflora, pedunculo subnullo, pedicellis са. 1 em. longis. Flores ut dicuntur virides brunneique. Caly- cis lobi ovato-lanceolati acuminati 0.5 em. longi extus minutis- sime puberulo-papillati. Corolla rotata extus glabra intus puberulo-papillata ca. 1.75 em. diam.; lobis ovato-ellipticis acutis ca. 0.8 em. longis patulis. Corona poculiformis valde 5- gona margine undulata, sinis margine 2-lamellatis ibique ligula inconspicua munitis, anguli margine revoluti. Gyno- stegium substipitatum; stigmate valde 5-gono depresso; an- theris sub stigmate positis, polliniis a corpusculo aliquantulum descendentibus cum caudiculis brevibus са. 0.15 em. longis, cor- puseulo mediocri. Folliculi ignoti.—GvaATEMALA: PETEN: Uax- actun, March 24, 1931, H. H. Bartlett 12300 (Herb. Missouri Bot. Gard., түре; Herb. Univ. Michigan, тзотурЕ). This species is rather intermediate between Ше subgenera Грайна and Mac- roscepis. In my key to the subgenera of Matelea, it will fall into the former after some persuasion, but it is probable that [Vor. 28 282 ANNALS OF THE MISSOURI BOTANICAL GARDEN a new subgenus will have to be erected for it eventually, since it is quite unlike any other described species, especially with regard to the peculiar corona, as well as the large coriaceous leaves. Сомоговоз lanugiflorus Woodson, spec. nov. Frutex volu- bilis omnino fulvo-hispidus; ramulis crassiusculis. Folia late oblongo-ovata apice abrupte angusteque subcaudato-acumi- nata basi late cordata 9-14 em. longa 4.5-7.5 em. lata mem- branacea, petiolis 2.0-3.0 em. longis. Inflorescentia umbelli- formis subsessilis, pedicellis са. 1 em. longis. Flores viriduli ; calycis lobis lanceolatis acuminatis ea. 1 em. longis; corollae rotatae extus intusque fulvo-hispidulae lobis ovato-lanceolatis anguste acuminatis ca. 1.5 em. longis, annulo faucium bene manifesto ca. 0.05 em. alto minute fulvo-hispidulo integro; corona rotata per saepta 5 radialia 5-partita margine mi- nutissime crenulata glabra; gynostegio subsessili ea. 0.45 em. diam; antheris circum stigma positis, appendiculis bene mani- festis carnosis lobis valde divaricatis obtusis ca. 0.1 em. longis, polliniis subhorizontalibus exeavato-pyriformibus eum cau- dieulo lato pellucido са. 0.15 em. longis, corpuseulo magno sagittato ca. 0.05 em. longo, stigmate 5-gono umbonato. Fol- Пеш ignoti.—GvATEMALA: DEPT. SAN MARCOS: vicinity of town of Tajumulco, northwestern slopes of Volcán Tajumulco, alt. 2300—2800 m., Febr. 28, 1940, J. A. Steyermark 36906 (Herb. Missouri Bot. Gard., туре; Herb. Field Mus., 1sorypr).—The uniformly dense, fulvous indument and subsessile cymes of this species are quite different from any species of the genus known to me. GowoLosvs longipetiolatus Woodson, spec. nov. Frutex volubilis; ramulis gracillimis laxe fulvo-hispidulis, internodiis sat elongatis. Folia oblongo-elliptica vel -ovata apice subcau- dato-acuminata basi obtusa vel indistincte cordata 4.5-9.0 em. longa 2—4 ст. lata delicate membranacea glabra; petiolis tenu- ibus 2-5 em. longis glabris. Inflorescentia subracemiformis, peduneulo 2.0-2.5 em. longo papillato, pedicellis aequalibus; calycis lobis anguste lanceolatis longe acuminatis ea. 0.7 em. 1941] WOODSON—ASCLEPIADACEAE AND APOCYNACEAE 283 longis extus minute papillatis intus glabris; corollae rotatae glabrae viridis lobis ovato-oblongis acuminatis 0.9 em. longis marginibus valde involutis interioribus albis inter se conspicue ealearatis, faucibus leviter annulatis; corona conspicue 5- lobata, lobis erectis late oblongis ca. 0.2 cm. longis apice truncatis vel paululo emarginatis marginibus subcondupli- eatim revolutis; gynostegio longiuscule (0.2 cm.) stipitato; antheris sub stigmate positis, appendiculis ovatis integris ca. 0.15 em. longis earnosis patulis, polliniis excavato-pyriformi- bus eum eaudieulo lato pellucido ca. 0.1 cm. longis, corpusculo minuto anguste sagittato; stigmate anguste 5-gono ea. 0.35 em. diam. conspicue rostrato. Folliculi ignoti.—GvATEMALA: DEPT. SAN MARCOS: above Finca El Porvenir, up Loma Bandera Влас, lower south-facing slopes of Volcán Tajumulco, alt. 1300- 1500 m., March 9, 1940, J. А. Steyermark 37428 (Herb. Field Mus., ryrE).—Closely related to such species as G. Lundellii and G. ctemiophorus, but differing in the entirely glabrous corolla and peculiar erect corona lobes. Сокоговов Steyermarkii Woodson, spec. nov. Frutex volu- bilis; ramulis sparse fulvo-hispidulis, internodiis sat elongatis. Folia ovato- vel oblongo-elliptiea apice subeaudato-acuminata basi late subauriculato-cordata 7-25 ст. longa 2.5-12.0 cm. lata firme membranacea utrinque pagina sparsius nervo medio basi supra densius fulvo-hispidula, petiolis 2-6 cm. longis sparse hispidulis. Inflorescentia racemiformis rarius corym- biformis, pedunculo 2-4 ст. longo saepissime simplici rare 2- fido minute papillato, pedicellis 2.5-3.0 em. longis sparse pilosulis; calycis lobis oblongo- vel ovato-lanceolatis acumi- natis ca. 1.5 em. longis ea. 0.5-0.7 em. latis foliaceis glabris vel inconspieue papillatis; corollae rotatae dilute viridis glabrae lobis late oblongo-elliptieis adscendentibus late acutis mar- ginibus revolutis ca. 1.2 em. longis ca. 0.5 em. latis, faucibus annulatis ca. 0.4 em. altis ostio dense pilosis caeterumque gla- bris albidis; corona annulata margine minute crenulata co- rollae faucibus adnata et eis aequilonga per saepta 5 radialia 5-partita; gynostegio longiuscule (ca. 0.45 cm.) stipitato; antheris sub stigmate positis, appendieulis dorsalibus an- [Vor. 28 284 ANNALS OF THE MISSOURI BOTANICAL GARDEN guste 2-lobatis, lobis angustis divaricatis ca. 0.1 em. longis, polliniis excavato-pyriformibus eum caudiculis latis pellucidis ea. 0.15 em. longis, corpusculo medioeri, stigmate anguste 5- gono ca. 0.4 em. diam. inconspicue umbonato. Folliculi ignoti. —GUATEMALA: DEPT. SAN MARCOS: along Rio Negro near Finca La Patria, Volcán Tajumulco, alt. 1300-1400 m., March 13, 1940, J. A. Steyermark 37661 (Herb. Missouri Bot. Gard., ТУРЕ; Herb. Field Mus., ISOTYPE); DEPT. QUEZALTENANGO: slopes and ridges between Quebrada Chicharro and Montana Chicharro, on southeast-facing slopes of Voleán Santa Maria, alt. 1300-1400 m., Jan. 18, 1940, J. A. Steyermark 34342 (MBG; FM); DEPT. sUCHITEPEQUEZ: southwestern lower slopes of Vol- cán Zunil, vicinity of Finca Asturias, northeast of Pueblo Nuevo, alt. 1200-1300 m., Febr. 1, 1940, J. A. Steyermark 35317 (MBG, FM).—A species of the Trichostelma complex, differ- ing from G. stenosepala (Fimbristemma stenosepala) in the nearly glabrous corolline faucal annulus, and from G. caly- cosus (Trichostelma ciliatum; Fimbristemma calycosa) in the narrower calyx lobes. Our species differs from both in the very scanty indument. Although the three species undoubtedly are very closely related and may subsequently be shown to represent mere varieties of a single species, the constancy of Steyermark’s three collections casts some doubt upon that view at this time. ManspEN1A pseudo-edulis Woodson, spec. nov. Frutex volu- bilis; ramulis graciliusculis rimosis juventate minute pu- berulo-papillatis. Folia latiuscule elliptica acute acuminata basi late obtusa 4.5-11.0 em. longa 1.5-5.0 em. lata membra- nacea glabra, petiolis 1.0-2.5 em. longis puberulo-papillatis. Inflorescentia umbelliformis pauciflora, pedunculo ca. 0.5 em. longo sparse pilosulo, pedicellis similibus; calycis laciniis ovatis obtusis 0.2 em. longis pilosulis, corollae campanulatae albae extus sparse intus dense pilosulae tubo ca. 0.2 em. longo, ostio ea. 0.25 em. diam. ibique inter lobos umbonibus 2 munito ; lobis ovato-ellipticis obtusis 0.3 em. longis patulis; gynostegio corollae tubum aequante, stigmate plano vel paululo depresso; antheris ca. 0.1 em. longis simplicibus, polliniis pyriformibus 1941] WOODSON—ASCLEPIADACEAE AND APOCYNACEAE 285 vix 0.05 em. longis caudiculo aequilongo ; coronae squamis reni- formibus obtusis са. 0.05 em. longis. Folliculi ignoti.—GvaATE- MALA: DEPT. QUEZALTENANGO: lower south-facing slopes of Vol- can Santa Maria, between Santa Maria de Jesus and Cala- huaché, along great barranco between Finca Pirineos and San Juan Patzulin, alt. 1300—1500 m., Jan. 6, 1940, J. A. Steyermark 33633 (Herb. Field Mus., ryPE).—Apparently somewhat inter- mediate between the sections Pseudomarsdenia and Ruehssia. ManspENIA Steyermarkii Woodson, spec. nov. Frutex volu- bilis ut dicitur epiphyticus fere omnino glaber; ramulis eras- siuseulis glabris, internodiis sat elongatis. Folia oblongo-el- liptiea apice abrupte acuminata basi rotundata 6-10 em. longa 2.0—3.5 em. lata subeoriacea glabra, petiolis 1-2 em. longis. In- florescentia modo dichasialis pauciflora, pedunculo 2-fido ca. 2.5 em. longo glabro, pedicellis ca. 0.4 em. longis obscure puberulo-papillatis ; calyeis laciniis ovatis acutis 0.2 em. longis minutissime puberulo-papillatis. Corolla campanulata alba extus glabra intus pilosula tubo ca. 0.3 em. longo ca. 0.4 em. diam. intus dense pilosulo; lobis ovatis obtusis 0.25 ст. longis sparse pilosulis patulis; gynostegio ca. 0.3 em. longo; antheris ca. 0.2 em. longis, polliniis eum caudiculo ea. 0.02 em. longis corpusculum superantibus ; coronae squamis ca. 0.2 em. longis stigma latum conicum superantibus dimidia inferiore late trigonis deinde in ligulam sublinearem productis. Folliculi ignoti.—GUATEMALA: DEPT. SAN MARCOS: along Quebrada Can- julá, between Sibinal and Canjulá, Voleán Тасапа, alt. 2200— 2500 m., common on eut-over slopes, epiphyte, Febr. 18, 1940, J. A. Steyermark 36019 (Herb. Field Museum, rvp£).— This species apparently is the only known Central American mem- ber of the section V erlotia. AscLEPIAS Scheryi Woodson, spec. nov. Herba perennis ro- busta ca. 2 m. alta; caule superiore valido sparsiuscule to- mentello. Folia opposita sessilia late trigono-ovata basi late rotundata vel subtruneata apice subrotundata abrupteque mucronulata superiora 13-15 em. longa 8-12 em. lata mem- branacea supra sparsius subtus densius tomentella nervo medio valido venis secondariis multis late arcuatis. Inflores- ; [Vor. 28, 1941] 286 ANNALS OF THE MISSOURI BOTANICAL GARDEN centiae ut videntur saepissime binis laterales multiflorae, peduneulis 6-8 cm. longis minute tomentellis, pedicellis vix 2 em. longis filiformibus similiter vestitis. Flores lactei pur- pureo-tineti. Calycis lobi ovato-oblongi acuminati ea. 0.4 em. longi extus minute pilosuli. Corollae rotatae lobi ovato- elliptiei acuti vel minutissime emarginati ca. 0.7 em. longi patenti extus densius intus sparsius puberulo-papillati. Gyno- stegii columna ca. 0.08 em. longa ca. 0.25 em. crassa; foliolis paliformibus brevissime substipitatis ca. 0.5 em. longis dorso subearinatis, margine superiore late rotundato, lobis laterali- bus rotundatis dimidia brevioribus, corniculo juxta basim ай хо са. 0.5 em. longo super stigmate abrupte inflexo ; antheris ea. 0.25 em. longis, alis basi prominentibus; polliniis rhombo- obpyriformibus eum caudiculis ca. 0.175 em. longis, cor- puseulo medioeri. Folliculi maturi non visi immaturi late fusi- formi laeves tomentelliimMexico: MICHOACAN: pine woodland near Uruapan, alt. 1850 m., July 16, 1941, R. W. Schery 167 (Herb. Missouri Bot. Gard., түрк). The leaves of this species are very distinetive, and are comparable only to those of A. lanuginosa, although less densely tomentose beneath. The flowers of A. Scheryi, however, are quite different from those of A. lanuginosa, in which the hoods are gradually acute and more than twice as long as the anther head. * THE TECHNIQUE AND USE OF MASS COLLECTIONS IN PLANT TAXONOMY EDGAR ANDERSON Geneticist to the Missouri Botanical Garden Engelmann Professor in the Henry Shaw School of Botany of Washington University A slight extension of ordinary herbarium techniques prom- ises to inerease the accuracy of herbarium studies and greatly to widen their scope. Attention was called to the possibility in 1935.! Since that time its potentialities have become increas- ingly evident and details of its technique have been consider- ably improved. The method consists in supplementing ordi- nary specimens by mass colleetions made as deseribed in de- tail below. For small plants these mass collections may well consist of the entire plant. For larger plants considerations of space require the selection of some critical portion, as the leaves, the infloresence, or the fruits. For instance, in studying sugar maples (Acer saccharum and its relatives) a mass col- lection will consist of one leaf per tree (carefully selected from the same kind of non-fruiting branch) from thirty to fifty trees, and complete specimens of the usual sort, from two or three of the trees. Properly made and filed (see below) mass collections require only a little more space than ordinary herbarium speci- mens. Such a collection is a record of a population as well as of the individuals which make up that population, and it therefore gives the faets about variation which can be obtained from pop- ulations but not from individuals. In other words, it would bring into the herbarium information which now we can get only in the field. This information ean be grouped under three different heads. (1) Frequency of the variation.—The most important de- fect of the ordinary herbarium material, in biometrical termi- nology, is that, while it may give a reliable estimate of the 1 Anderson, Edgar, and W. B. Turrill. Biometrical studies on herbarium material. Nature 186: 986. (287) [Vor. 28 288 ANNALS OF THE MISSOURI BOTANICAL GARDEN range of variation, it does not allow an estimate of the fre- queneies within that range. That is to say, in non-mathemat- ical language, that it may give a good idea of extremes but it does not indicate the relative prevalence of the extremes or of any partieular intermediate. It is not enough to know that a variant exists; for its complete interpretation one needs to know how often it occurs in the places where it has been re- ported. As Dr. Fassett has put it in a homely analogy: ‘‘there are Demoerats and Republicans in both Mississippi and Ver- mont but their comparative frequency varies significantly be- tween these two regions.” (2) Discontinuity of variation.—This is potentially one of the greatest sources of error with present-day techniques. The chief eriterion for separating taxonomie entities is the degree of morphological discontinuity between them. At the present time it takes good judgment and often field experience to deeide if the discontinuity shown by a group of herbarium specimens is real or only apparent. This is particularly true for cate- gories smaller than the species. (3) Correlation between variables.—While an estimate of this correlation can be obtained from ordinary herbarium spec- imens, it can be derived much more precisely from mass col- lections and can be perceived more readily and its perception requires less biologieal judgment. Those who have undertaken monographie work will have encountered complexes in which variation was so extreme and involved so many different char- acters that it was difficult to comprehend. Mass collections make it possible to study such complexes analytically and to determine precisely the extent to which the variation in differ- ent characters is correlated. Anderson and Turrill? for in- stance, by using mass collections, were able to resolve the vari- ation in the Fraxinus Pallisae complex into two elements and to relate these elements to species of Fraxinus from southeast- ern Europe. There are two problems in making mass collections: what * Statistical studies on two populations of Fraxinus. New Phytol. 37: 160—172. 1936. 1941] ANDERSON—TECHNIQUE OF MASS COLLECTIONS 289 part of the plant to collect and the selection of a random sam- ple of the population. The first is not as difficult as it may seem to anyone who has not tried it. Taxonomie studies on the cus- tomary herbarium material are a necessary foundation for the making of mass collections intelligently, and the study of a taxonomie revision will tell what parts are signifieant and should be collected in quantity. The portion of the plant chosen for intensive collection should provide good morphological criteria; it should be easy to press and store in quantity; and its selection should be definable in precise terms. The follow- ing examples may make these points clear: Tripsacum, the terminal inflorescence of each plant; Monarda, an average flower-head from each plant, with its subtending bracts; Ver- onica peregrina, the entire plant. Wherever possible mass col- lections should be a series of one sample from each plant. Oc- casionally a single vegetatively reproduced individual (tech- nically known as a clone) may cover a very large area, and it may be difficult or impossible to know where one individual leaves off and another begins. This is particularly true in such plants as Sanguinaria canadensis whose rhizomes grow and branch vigorously and the organic connection between two branches usually rots away after a few years leaving them physically separate. In most cases a careful study of flower and leaf variation will reveal the probable extent of each clone, and a careful collector can minimize the chances of gathering a disproportionate number of samples from a single clone. The problem of a truly random sample is difficult, and bristles with difficulties which are not even suspected by the uninitiated. One should bear in mind that he is trying to make a record of a population of individuals and that the record will have the greatest significance if it is chosen at random from an actual inter-breeding population. Lacking the precise informa- tion as to what an ‘‘actual inter-breeding population’’ may be, one ean only use his biological judgment in selecting for each case an area which presents uniform conditions for that species and make his collections from that area. Even for those species whieh grow in definite, more or less isolated colonies, the [ Vor. 28 290 ANNALS OF THE MISSOURI BOTANICAL GARDEN trained eye can often find evidence for distinctive neighbor- hoods within such communities, and it will be a matter of in- dividual judgment whether these neighborhoods should be rec- ognized or ignored in making a sample of the population. When the area has been chosen one may either collect a speci- men from each individual within that area, if there be not too many, or make a random selection of thirty to one hundred or more individuals. One can walk across it in a straight line, making collections at every second or third step, or can use Figure 1. strings and collect every individual which is touched by the string or is close to it. If properly made, mass collections do not take up more space than a few herbarium sheets, and they give information which cannot be obtained from hundreds of ordinary specimens. If supplemented by complete specimens there should be no objec- tion to their fragmentary nature. Technical improvements in storing the collections have been made by my assistant, Mr. Leslie Hubricht, and are illustrated in the accompanying fig- ure. Since my collections are subject to intensive study but are not distributed in an ordinary publie herbarium they are kept 1941] ANDERSON—TECHNIQUE OF MASS COLLECTIONS 291 unmounted. To prevent the specimens from scattering, the genus covers are folded so that the edges meet down the middle rather than the side. One label is made for each collection and is pasted on the genus cover. The ordinary specimens made at the same time and place are mounted and labeled and are kept in the same genus cover. Each collection is given a geograph- ical name and all are assembled alphabetically under the genus or the species, depending upon the nature of the particular problem. The information derived from a study of mass collections is useful in two ways. It will, in the first place, aid the sys- tematist in cataloguing the various entities involved, species, varieties, forms, etc. While it may raise more new questions than it may solve old ones, it will aid in the production of mono- graphs whose categories are more accurately adjusted to the variation patterns of their particular genera. Mass collections have for some time been customary in avian taxonomy (see, for instance, Mayr?), and Kinsey, in a series of brilliant mono- graphs,‘ has shown their superiority in insect systematics. If taxonomy were to be nothing more than cataloguing and if taxonomists were to confine themselves to the problems raised by their herbaria, mass collections would still be a useful ad- junct to herbarium technique and in many critical groups would provide more efficient working material, even when their special difficulties of collecting and filing are considered. There is no reason, however, why taxonomy should be con- tent to cultivate such a narrow field. If collectors and her- barium administrators could be persuaded to encourage mass collections, critically made and carefully assembled, a second kind of problem could be investigated in herbarium material. The description and analysis of geographical trends in varia- tion, the delimitation and interpretation of centers of varia- tion, the establishment and analysis of variation patterns in 3 Mayr, Ernst. Speciation phenomena in birds. Amer. Nat. 74: 249-278. 1940. ‘Kinsey, Alfred C. The gall wasp genus Cynips. A study in the origin of species. Indiana Univ. Studies. 84—86: 1-577. 1930; The origin of higher categories in Cynips. Indiana Univ. Publ. Sci. Ser. 4: 1-334. 1936. [Vor. 28, 1941] 292 ANNALS OF THE MISSOURI BOTANICAL GARDEN different genera and families, are only a few of the problems that might well be investigated. It is already possible to cor- relate information from the field of taxonomy with that from cytogenetics. The time is not far distant when the biochemist of the germplasm will also turn to the taxonomist for morpho- logical evidence derived from studying the products of the germplasms. To speak with authority on such questions tax- onomists will need to refine their biological as well as their bibliographical techniques. MASS COLLECTIONS: CAMASSIA SCILLOIDES RALPH O. ERICKSON University Fellow in the Shaw School of Botany of Washington University In the St. Louis vicinity, Camassia scilloides (Raf.) Cory exhibits a considerable amount of variability, which is ap- parent on looking at a number of plants even casually. This fact has been given some recognition by the description of C. scilloides forma Petersenii Steyermark.! In the spring of 1940 an attempt was made to learn something about the sta- tistical nature of the variation. “Mass collections” (Ander- son,” Fassett*), each consisting of 25-50 or more inflorescences, and in some cases leaves, were made at the places which are cited in detail in the footnote,* and which will be referred to as New Athens, Allenton, Meramec Highlands, Gray Summit and Spring Creek, respectively. The plants of the first collection were first studied to decide what features of the variation could be most satisfactorily submitted to measurement. The dimensions of the inflores- cence seemed most promising. The “open” appearance of some of the inflorescences, as contrasted with the compactness of others, seemed obviously related to internode length and pedicel length, and these lengths were measured in plants of each of the collections. There is considerable difference in length between successive internodes, so that instead of meas- uring a particular one, the length of the lowest eight was used. 1 Rhodora 40: 178. 1938. ? Ann. Mo. Bot. Gard. 28: 287-292. 1941. * Ann. Mo. Bot. Gard. 28: 299-374. 1941. * ILLINOIS. ST. CLAIR CO.: in river-bottom woods, about 1 mile northwest of New Athens, May 7, m MISSOURL ST. LOUIS CO.: on top of a ridge, 2 miles south of Allenton, S. 10, 43N, R. 3E, April 28, 1940; along a ips: south slope below the Frisco тйлй кай at Meramec Highlands, S. 10, Т. 44N, В. 5E, May 18, 1940. FRANELIN CO.: at the еен >” Missouri Земи сзи Arboretum near Gray Summit, S. 17 & 20, T. 43N, R. 2E, April 21 and May 5, 1940; on a south slope at еей Creek, 4 miles UE of Stanton, May 11, 1940. [Vor. 28 294 ANNALS OF THE MISSOURI BOTANICAL GARDEN The lowest pedicel was measured in each inflorescence. Neither of these lengths changes appreciably after anthesis, and no in- florescences were measured in which at least half of the flowers had not bloomed. As a check, measurements of members of a few supposed clones were made, and were found to agree. When internode length was plotted against petiole length in a scatter diagram, a simple picture of the variability within each collection was obtained. Not only was there variability MERAMEC HIGHLANDS NEW ATHENS оор 100+ - E е $ 5 .. I е + + т = е m 9 Es ? "e “е 9 Z so Ф * z so e e 4 °° E Ae * $585 d T + ul е ul t 8 © Фо оо $ а 2° 9 ee oS 9 9 ЗА x "PI, + o” 24 ш ш ee ь = е & z T TJ 20 + PEDICEL LENGTH PEDICEL LENGTH Figure 1. within each of the colonies, but a striking difference was seen between colonies. The Meramee Highlands, Allenton, Gray Summit and Spring Creek collections, all of which were made within the Ozark region, produced essentially the same sort of scatter diagrams, while that for the New Athens material was quite different (see fig. 1). At New Athens the range of varia- tion in these two characters is quite restricted as compared with Meramee Highlands and the other Ozark stations. Although the Meramec Highlands collection includes practi- eally all the types represented at New Athens, most of the plants from Meramee Highlands lie completely outside the range of variation of the New Athens colony. 1941] ERICKSON— MASS COLLECTIONS: CAMASSIA 295 The same sort of difference in variability is shown by the seatter diagrams in fig. 2, where sepal length and width, as measured from camera-lucida drawings of fresh material, are plotted. (The difference, however, is not so pronounced; or is it exaggerated in the case of the internode and pedicel lengths by the use of an inappropriate scale?) Furthermore, the uniformity of the New Athens plants as compared with those collected in the Ozark localities is appar- ent in their general aspect, whether seen in the field or as dried MERAMEC HIGHLANDS p NEW ATHENS er ч е е 4 * 4 е „” А Ф H IE E. ое I . 4 " » = eto | è е ъ р е и = мр yas " z at ъъ ос" = г оо $ > е e o 4 e i s а & a ul е ш v v) 10 15 T 15 SEPAL LENGTH SEPAL LENGTH Figure 2. specimens. See the photographs of dried inflorescences in Many of the features of the variation which is so apparent in the Camassias of this region cannot be studied statistically, but some of them will be discussed in general terms here. The eolor of the flowers varies somewhat. While the prevailing color is a pure, pale blue, there is an appreciable range in the depth of the blue color, and in many plants a slight reddish tinge is perceptible. The petals (and sepals) vary both in size and shape, one of the details of the variation being the pres- * Wright in Jour. Amer. Stat. Assoc, 21: 162—178. 1926. [Vor. 28 296 ANNALS OF THE MISSOURI BOTANICAL GARDEN ence or absence of a sagittate base. It is also easy to see varia- tion in style length and form. The general aspect of the inflorescence presents easily ap- parent differences, as mentioned above. There are inflores- cences which are decidely cylindrical in shape, and others which appear more or less pyramidal, even when all allowances are made for the acropetal order of flowering. In some plants the inflorescence appears compact, with the flowers greatly crowded; in others it is open and spindly in appearance. The peduncle varies, the extremes being a stout and fleshy sort, and a quite slender and woody one. The bracts are usually inconspicuous and completely withered at anthesis, but plants can be found in which large green bracts, approaching in size the basal leaves, occur at the lowest few nodes of the inflores- cence. The largest of these bracts do not subtend flowers. Incidentally, no such large, leaf-like bracts were found among the New Athens plants. The leaves offer differences in size, and in the amount of bloom on the upper surface. There also appear to be differ- ences in shape of the bulb, which are not connected with its size. In all of the above-mentioned charaeters, the New Athens collection is less variable than are the other four. Although the two collections which have been discussed in detail were made less than fifty miles apart, they are from quite different habitats, and for that matter from different physiographic regions. Meramec Highlands (as well as Allen- ton, Gray Summit and Spring Creek) is on the northeastern edge of the Ozark Plateau, and at these Ozark localities Camassia grows for the most part on steep slopes. New Ath- ens is in the Till Plains section of the Central Lowland,’ and Camassia was there found growing in rich bottom-land soil. The former region is unglaciated, and has been occupied by plants continuously since preglacial times, while the latter was covered by the Illinoian ice sheet. While it is not the purpose of this paper to offer a complete * Fenneman, Physiography of the eastern United States. New York. 1938. 1941] ERICKSON—MASS COLLECTIONS: CAMASSIA 297 explanation of the facts presented above, one might suppose that the differences which have been demonstrated between Camassia of the Ozarks and of the Illinois bottom-lands are related to the different vegetational histories of the two regions. — А [VoL. 28, 1941] 298 ANNALS OF THE MISSOURI BOTANICAL GARDEN EXPLANATION OF PLATE PLATE 8 Fig. 1. Photograph of ten Sagre of Camassia scilloides from the mass col- leetion made near New A , Ill. Seale is in centimeters. Fi ig. 2. Photograph of M mtn of Camassia scilloides from the mass collection made at Meramee Highlands, M 8 PLATE Vor. 28, 1941 GARD., Bor. ANN. Mo. NEW ATHENS Mi = А P A NA S e “= Se MERAMEC HIGHLANDS 'AMASSIA ( MASS COLLECTIONS: ERICKSON MASS COLLECTIONS: RUBUS ODORATUS AND R. PARVIFLORUS NORMAN С. FASSETT Associate Professor of Botany and Curator of the Herbarium, University of Wisconsin CoNTENTS d: зи О о. a eS DO Н О о M . The kinds of vaste Е щик иа ee E T Taxonomy of intraspecifie — А © э 5.9 WM 3. Geographie occurrence of each form in mass шейш ер TAD 4. Correlation of variation with коны р. boh. Se э! ку 00 II. Variation in R. kien See ae И о T TTA я The kinds of varia puo mice А e ale The ** varieties?" a po bordbtohds Е, РАЛЛИ See hac a ee The forms of В. parviflorus . ee ee. co. а о това 4. Occurrence of forms in mass отмена ; . 827 5. Is В. parviflorus a d relie about the е Upper à Great Тама! . . 338 Summary . н . . 345 In 1938 а grant was made by the Wisconsin Alumni Research Foundation for the study of the so-called preglacial relies in the flora of the upper Great Lakes region. This study was eventually narrowed to one species, the Thimbleberry or Sal- monberry (Rubus parviflorus), whose range shows a gap be- tween Lake Superior and the Black Hills of South Dakota which has been interpreted as due to survival on nunataks about Lakes Superior and Huron. Funds from the estate of the late Dr. J. J. Davis, made available through the generosity of his daughter, Miss Marguerite Davis, made it possible for the writer to carry on this study in the summer of 1939, when the species was collected in South Dakota, Wyoming, Utah and Colorado. Friends in California, Alberta, Wisconsin, Oregon, Colorado, Indiana, New York, Montana and South Dakota have also contributed collections; their names are cited with their collections in tables 1 and v, and grateful acknowledge- ment is here made for their assistance, without which this study would not have been possible. The method used was different from ordinary taxonomic or Issued September 20, 1941. ANN. Mo. Bor. GARD., Vor. 28, 1941 (299) [Vor. 28 300 ANNALS OF THE MISSOURI BOTANICAL GARDEN ecologieal techniques: it consisted of collecting many indi- viduals from each colony of Thimbleberry and determining in the laboratory the percentage of plants showing each char- acter. By comparing figures from different regions it has been possible to judge their closeness of relationship; also, some taxonomie changes have been indieated. It became evident that results concerning a plant with a disrupted range like that of R. parviflorus were difficult to evaluate in absence of a similar study of a species with an essentially continuous range; ac- cordingly, mass collections were made, in 1938-40, of the east- ern А. odoratus. Findings in this species will be considered first. In assembling the distributional data in Maps 12, 24 and 27, the writer is indebted to many friends for assistance ; these in- clude Mr. H. D. House, Dr. Earl Core, Dr. T. M. C. Taylor, Dr. J. М. Fogg, Jr., Dr. Е. T. McFarland, Dr. E. Lucy Braun, Dr. R. M. Harper, Dr. A. J. Sharp, Dr. F. K. Butters and Dr. Hugh Raup. Maps 10, 12, 19, 20, 21, 24, 25, 26, and 29-34 were pre- pared from base maps of Hall’s ‘Outline Maps and Graphs,’ published by permission of the author and publisher, John Wiley & Sons, Ine. Maps 35 and 36 are (except for botanical data) from map by Erwin Raisz in Atwood's 'The Physio- graphic Provinces of North America,’ courtesy of Ginn & Co. Professor Walter Cottam, of the University of Utah, has most kindly supplied figs. 5, 6 and 7 of pls. 11 and 12. I. VARIATION IN В. odoratus 1. THE KINDS OF VARIATION Rubus odoratus, ranging from Nova Scotia and Quebec to Michigan, and southward to Georgia and Tennessee, shows variations which are to a great extent the exact counterparts of those which were described as varieties in R. parviflorus.’ The two pairs of characters (leaves soft-pubescent beneath vs. glabrous or glabrate beneath, and pedicels with long-stalked glands vs. short-stalked glands) which recombine to make the four ‘‘varieties’’ of В. parviflorus—hypomalacus, heterade- * Fernald, Rhodora 37: 276. 1935. 1941] FASSETT—MASS COLLECTIONS: RUBUS 301 nius, bifarius and grandiflorus—appear in just the same com- binations in В. odoratus. Each of these four combinations ranges essentially throughout the range of the species, so they seem better treated as forms than as varieties. Asin R. parviflorus, the distinetions within each pair of char- acters are not always perfectly marked. Clearly glabrous leaves occur, as do extremely velvety ones, but some inter- mediates are as difficult to place as they occasionally are in the western species. Many plants have leaves which are ap- pressed-pubescent and feel either harsh or velvety depending on how they happen to have been pressed. In some cases both surfaces are without doubt glabrate, the younger ones appear- ing densely velvety. In placing these questionable individuals recourse has been had to a character pointed out by Fernald?: truly velvety leaves ordinarily bear dark long-stalked glands on their upper surface. The correlation is good, but not per- fect, through much of the range of the species. It breaks down completely in two large colleetions from the unglaciated part of Indiana; here, isolated from the rest of the range for а long time, the plants often have a different appearance (lighter colored, more glabrous and lustrous leaves), frequently carry on the leaves a type of gland not generally found on the species in other parts of the country, and, as has just been stated, lack the usual correlation of glandular upper surfaces with velvety lower surfaces. The type of gland found mainly in Indiana is on the lower leaf-surfaces along the veins, and has a stalk 2-4 mm. long, which much exceeds those of other forms. Plants with long- stalked glands are not exclusively present in the collections from Indiana, but they outnumber plants with shorter glands. A plant of very rare occurrence has the glands on the pedicels nearly sessile; this parallels R. parviflorus var. scopulorum and var. parvifolius. Leaf shapes vary greatly in this species, but because of some variation even on single branches these characters are not as conveniently dealt with as are those of epidermal outgrowths. Examination of any good-sized collection will show singly- 2 Rhodora 24: 175. 1922. [VoL. 28 302 ANNALS OF THE MISSOURI BOTANICAL GARDEN toothed and doubly-toothed margins, triangular lobes and ob- long-triangular lobes, and depth of lobing varying from М to 1 the radius of the leaf. An individual with deeply lobed leaves has been described as В. odoratus var. columbianus, or R. columbianus.* Sometimes the calyx is covered with a white felt-like tomen- tum ; this proves to be due to the fungus infection Sphaerotheca humuli. 2. TAXONOMY OF INTRASPECIFIC VARIATION To facilitate discussion of these six phases of R. odoratus and of the geographic occurrence of the characters on which they are based, it seems advisable to give them names. That the parallelism of five of them with variants of R. parviflorus may appear clearly, they are given names identical with or phonetically similar to those used in the more western species. RUBUS oporatus L. f. hypomalacus, n.f., foliis subtus velutinis vel subvelutinis, supra fusco-glandulosis ; pedicellorum glandulis stipitatis 1-3 mm. longis.—Typr, in Herb. Univ. Wis.: roadeut, Ravine, Pa., July 11, 1940, Fassett 20812. (R. odoratus var. malachophyllus Fernald. ) В. оговатов f. heteradenius, n.f., foliis subtus glabris vel glabratis vel sparse appresso-pilosis, supra боше: pedieellorum glandulis stipitatis 1-3 mm. longis.—Typr, in Herb. Univ. Wis.: Wilmington Mountain, east of Bennington, Vt., Aug. 4, 1938, Fassett 20787. В. оровлтоз f. bifarius, n.f., foliis subtus velutinis vel subvelutinis, supra fusco-glandulosis; pedieellorum glandulis stipitatis 0.5-1 mm. longis.—TyPE, in Herb. Univ. Wis.: cuts and fills along U. S. 6 west of Bear Mountain Bridge, N. Y., July 10, 1940, Fassett 20806. В. oporatus f. glabrifolius, n.f., foliis subtus glabris vel glabratis vel sparse appresso-pilosis, supra Missionis: pedicellorum glandulis stipitatis 0.5-1 mm. 10п015.—ТҮРЕ, in Herb. Univ. Wis.: woods and roadside banks 3 miles north of Vassalboro, Me., Aug. 16, 1939, Fassett 20781 В. oporatus f. scopulorum n.f., foliis subtus glabris vel glabratis vel sparse appresso-pilosis, supra eglandulosis; pedicellorum glandulis subsessilibus vel stipitatis non quaterlongioribus quam glandulis.— Түр, in Herb. Univ. Wis.: roadeuts along U. 8. 30, Sideling Hill, Pa., * Millspaugh, Bull. W. Va. Exper. Stat. 2: 355. 1892. * Rydberg in Britton, Manual, p. 495. 1901. 1941] FASSETT—MASS COLLECTIONS: RUBUS 303 July 11, 1940, Fassett 20817. The illustration of В. parviflorus var. scopulorum, in Rhodora, vol. 37, pl. 365, fig. 5, might well serve also as an illustration for В. odoratus f. scopulorum. R. oporatus f. parahypomalacus, n.f., f. hypomalacum simulans, sed venis subtus glandulis 1-2 (-3) mm. longis.—Typr, in Herb. Univ. Wis.: woods and eut-over land, foot of Shenandoah Mt., east of Frank- lin, W. Va., July 12, 1940, Fassett 20823. В. oporatus f. paraheteradenius, n.f., f. heteradenium simulans, sed venis subtus glandulis 1-2(-3) mm. longis.—T Pr, in Herb. Univ. Wis.: along roadside, from halfway down to base of steep slope, as- sociated with Tsuga canadensis, Guthrie Creek, about 2 miles east of Leesville, Ind., July, 1939, В. M. Kreibel, C. Е. McGraw & Morris Reeves. These forms may be keyed as follows: а. Pedicels with glands on stalks mostly 0.5 mm. or more long, the stalks more than 4 times as long as the gland b. Stalks of glands on pedicels mostly more than 1 mm. lon с. Veins on lower leaf-surfaces with glands whose stalks are 0.2-1.0 mm. d. Leaves without dark-stalked glands on the upper surface, glabrous ath f. het or lightly appressed-pubescent Ъепеаёћ.................. eradenius d. Leaves with dark-stalked glands on the upper surface, more or less velvety.benesaih..-.- 2...5.» Ае TERRE esi ets ats f. hypomalacus с. Veins on lower leaf-surfaces with glands whose stalks are 1-2 (—3) mm. long е. Leaves glabrous or glabrate or lightly appressed- e beneath "c Pop LEO Tum OO ET araheteradenius д. Leaves velvety beneath?.... E Lee tus . parahypomalacus b. Stalks of glands on pedicels mostly 1 mm. or less long f. Leaves with dark-stalked glands on the upper surface, more or Leas: velvety beneath. susse ov KR еек bo nae ооо f. bifarius f. Leaves without dark-stalked glands on the upper surface, glabrous glabrate or T appressed-pubescent beneath ....f. м а. Pedicels with glands s subsessile or on stalks less than 0.5 mm. long an than 4 times as long as ie gland. Ligases ооо. {49 Г. та 3. GEOGRAPHIC OCCURRENCE OF EACH FORM IN MASS COLLECTIONS The object of this paper is not primarily to describe seven forms, or to point out the similarity of variation within R. odoratus to that within R. parviflorus. It is rather to study statistically the occurrence throughout the range of R. odoratus of each character used in the above key, and to determine what, if any, correlation there may be between the occurrence of these characters and the late geological history of the species. [Vor. 28, 1941] 304 ANNALS OF THE MISSOURI BOTANICAL GARDEN For this purpose specimens have been taken from as many in- dividuals as possible in each colony of the species which the writer has seen in the past three years; similar collections have also been made by several friends whose names appear in table r. The location of each mass collection is shown by a slanted figure on maps 1 and 2. It will be noted that in each state or province the numbers start anew. Table 1 gives the col- lection data for each collection, and table п the constituent forms of each. TABLE I MAINE: 1. Vassalboro, Aug. 16, 1939, 20781. 2. Augusta, Aug. 14, 1939, 20782. 3. Paris, Aug. 8, 1939, 20783. VERMONT: 1. Colchester, Aug. 7, 1939, 20785. 2.° Bur- ея Aug. 7, 1939, 20786. 3. Wilmington Mountain, Bennington, Aug. 4, 1938, 787. 4. Wilmington Mountain, Bennington, Aug. 4, 1938, 20788. MASSACHU- SETTS: 1. Near Cold River, east slope of Mohawk Trail, June 25, 1940, 20789. 2. West slope of Mohawk Trail below Hairpin Turn, June 25, 1940, 20790. ONTARIO: 1. West of Brockville, Aug. 7, 1939, 20791. New Уовк: 1. Westfall Road, Penfield Township, Monroe Co., June 30, 1940, Е. Е. Shanks. 2. Atlantie Avenue, Penfield Township, June 30, 1940, Shanks. 3. Eaton Road, ae Monroe Co., July 14, 1940, Shanks, 4. Pellett Road, Webster gee Monroe Co., June 30 > мы «> > M © 7. Sehe: e 24 go B ‚ ё Woodhull, June 23, 1940, 2 10. Greenwood, June 23, 1940, 20797. m Bolivar, June 23, 1940, 20798. т viris side of Allegheny River, 4 miles north of Limestone, June 23, 1940, 20799. 13. Three miles north of Limestone, June 23, 1940, 20800. 14. About a mile south а the Lake, Allegany State Park, June 21, 1940, 20801. 15. Quaker Bridge, June 21, 1940, 20802. 16. Allegany State Park, 5 miles west of Limestone, June 23, 1940, 20803. 17. Peekskill, July 10, 1940, 20804. 18. U. S. 6 east of Bear Mountain Bridge, July 10, Aci 20805. 19. U. S. 6 west of Bear Moun- tH 10, 1940, 20809. 2. Northwest of Mauch Chunk, July 10, 1940, 20810. 3. Easton, July 14, 1940, 20811. 4. Ravine, July 11, 1940, 20812. 5. Ravine, July 11, 1940, 20813. 6. Muir, July 11, 1940, 20814. ?. Water Street, Sept. 18, 1959, 20815. 8. Along U. S. 30, Tusearora Hill, July 11, 1940, 20816. 9. Along U. В. 30, Sideling конч July 11, 204, 20817. 10. DEN bee 28, 1940, 20818. 11. Ten miles of Erie 20, 1940, 20 YLAN „Ва awlings, July 12, 1940, 20820. ниф Sali gg ч. Romney, Е n pn Sh . Old Fields, July 12, 1940, 20822. 3. Franklin, July 12, 1940, 20823. INDIANA: 1. Jackson Co., about 2 miles east of Leesville, Fate, 39, В. M. Kreibel, С. Е. McGraw $ Morris Reeves. 2. Baek Creek, 1-115 sie west of Leesville, July, 1940, Kreibel, McGraw $ Reeves. * When a number is given without collector’s name the writer was the collector. * A. number not shown on maps 1 & 2 represents a station close to the preceding number. TABLE I OCCURRENCE IN EACH COLLECTION OF THE FORMS OF R. ODORATUS Š E "reu 3 S "S Е Е E E E S 5 с d > E S 1$ & | $1] mu E У = & fu > D Ф MAINE 1 1 2 20 8 10 5 6 4 $ 8j 5 4 VERMONT 1 7 Yi 1 2 4 9 4 4 7 11 1 9 4 3 5 T 5 MASSACHUSETTS 1 5 9 2 11 1 3 ONTARIO 1 1 5 8 New YorK 1 3 28 2 8 2 8 2 8 1 3 4 8 3 1 4 2 2 2 5 15 6 8 9 1! 5] 7 14 7 8 18 1 4 1 9 5 13 1 10 2 23 1 11 10 2 4 1 12 8 6 18 үз 1 14 4 10 15 10 11 16 4 4 17 28 35 1 1 1 18 7 1 1 1 19 23 4 3 11 1 20 8 4 3 6 21 6 7 5 2 PENNSYLVANIA 1 1 2 T 4 2 4 2 3 3 21 H 8 4 9 5 29 10 8 6 4 2 6 Я. "d 11 8 8 4 7 9 18 Па 5 6 1 10 18 8 1 4 2 12 25 14 а 3 2 MARYLAND 1 21 18 4 4 WEST VIRGINIA 1 19 7 3 2 4 2 8 79 5 3 10 7 INDIANA 1 22 30 2 7 11 [ Vor. 28 306 ANNALS OF THE MISSOURI BOTANICAL GARDEN 4. CORRELATION OF VARIATION WITH GEOGRAPHY Cursory examination of table п shows but one fact, namely, that 4 of the forms occur essentially throughout the range of the species (as far as represented in these collections—see map 12), and 3 are more loeal. But from this list certain data may be derived; some of these data are shown on maps 1 and 2. Map 1 shows the percentage in each collection of individuals with leaves glabrous or glabrate beneath (or lacking glands above). For example, in collection 2 from Maine, a total of 25 individuals shows 4 f. glabrifolius, and 5 f. heteradenius, a total of 9 glabrous individuals, or 36 per cent—accordingly, the figure 36 appears next to the 2 in Maine. Similarly, collec- tion 3 from New York has, in a total of 16 individuals, 1 f. gla- brifolius, 8 f. heteradenius, and 3 paraheteradenius, making a total of 12 glabrous individuals, or 75 per cent—the figure ap- pears near the 7 in New York, since collections 1—4 are too close together to be mapped separately. How significant are the percentages shown on map 1? Prob- ably not very significant individually, for the following rea- sons. First, the numbers in each collection are small, mostly below 50, sometimes less than 25. Second, R. odoratus spreads by underground stems, and more than one collection may be made from one individual (г.е., clone). This has been avoided as far as possible by taking specimens at some distance from one another or from isolated plants. When the figures from many collections are totaled, they be- come statistically more reliable. From Maine, 60 individuals have been collected, from Vermont 80, Massachusetts 29, On- tario 14, New York 448, Pennsylvania 256, Maryland 47, West Virginia 139, and Indiana 70. These represent many times the number of individuals ordinarily examined in the taxonomic treatment of a group, where a single sheet from a region is often taken to represent THE PLANT OF THAT REGION. In determining the significance of figures from one patch of Flowering Raspberry, and the significance of those from a general region, the ecology of the species must be taken into consideration. Its favorite habitat is recently cut soil; it is 1941] FASSETT—MASS COLLECTIONS: RUBUS 307 sometimes found on a river-cut bank, which may be more or less wooded, but its most frequent occurrence is along road- cuts. Nearly all of its stations are, therefore, more or less re- cent, the road-cuts being more recent than the river-cuts. But although any patch may be recent (and temporary) the species may be nevertheless of ancient occurrence in the region as a whole. There is another important point regarding individual patches of R. odoratus—the seeds are, presumably, spread by birds, and the syncarp may contain seeds of different genetic constitutions (due to heterozygosity in the mother plant, and the possibility of pollen coming from several sources), and a colony may perhaps be started from a number of seeds dropped at one time; the variation within any patch should then be de- termined partly by the number of seeds originally dropped. If a colony increases in size vegetatively its original genetic con- stitution will tend to be preserved, while expansion by seeds will increase the number of forms. Again, the constitution of the species throughout a region stands as more significant than in any single patch. With these facts in mind, map 1 may again be examined. It becomes evident that, in spite of lack of uniformity in many regions, the higher proportions of plants with glabrous leaves occur mostly toward the northern part of the range (omitting Indiana, for the present, as a more remote region). To empha- size this, we may divide the map into zones from north to south, by the broken lines running east and west. Within each of these zones the percentage of glabrous-leaved plants is de- termined (by totaling the figures from all the collections, not by averaging the percentages of each), with the result shown by the large figures along the right-hand margin of the map. There appears an unmistakable trend toward the appearance of velvety leaves southward and of glabrous leaves north- ward.’ With consideration of the line of farthest advance of ' Perhaps it should be restated at this point that by ‘‘ glabrous leaves?" is meant also the glabrate or somewhat pilose leaves lacking glands on the upper surface, and by ‘‘velvety leaves’’ is meant those which tend to be more pubescent beneath and have dark glands above; cf. the key (page 303) and the discussion on page 301. [Vor. 28 308 ANNALS OF THE MISSOURI BOTANICAL GARDEN Pleistocene glaciation (line of crosses on map 10) a definite correlation appears. If itis assumed that R. odoratus survived glaciation south of the glaciated regions and migrated north- ward after the disappearance of the ice? it becomes evident that the species became more glabrous as it spread northward. This has not involved the origin en route of a new character, for glabrous plants occur in the extreme southern part of the range (represented in the Gray Herbarium by sheets from Virginia, North Carolina, Kentueky and Tennessee) and vel- vety leaves occur in the extreme northern part of the range (specimens in the Gray Herbarium from Quebec and Nova Scotia), but rather a change in the proportion of occurrence of the character. Is this because (1) the glabrous plants are more suited, physiologically, to the northern regions, (2) the glabrous plants were for some reason more vigorous in spread- ing, or (3) simply that as the species migrated chance happened to favor partial elimination of pubescent individuals? The situation regarding pubescence of leaves in R. odoratus is just the reverse of that regarding the pubescence of the stem in the Red Raspberry ( Rubus idaeus, or R. strigosus). In the latter, the proportion of pubescent plants increases northward. Perhaps any purpose that may be ascribed to a hairy covering is of less effect than is the linkage of pubescence, or in other cases glabrousness, with other characters which have a closer relation to the environment. Similar progressive variation has been demonstrated for the snail, Partula suturalis, in the isolated valleys on the island of Moorea.’ Eastward, colonies are composed entirely of dex- tral individuals, and these gradually give way, westward and northwestward, to colonies composed entirely of sinistral in- dividuals (see map 11). On map 2 are plotted the percentages of individuals in each colony with short glands on the pedicels (f. bifarius and f. gla- brifolius). As оп map 1, there is lack of uniformity from local- * This does not preclude the possibility of interglacial migration also. * Crampton, H. E. Studies on the variation, distribution, and evolution of the genus Partula: the species inhabiting Moorea. Carnegie Institution of Washington. Publ. 410. 1932. 1941] FASSETT—MASS COLLECTIONS: RUBUS 309 ity to locality, but here, again, appear decided tendencies : the group of low figures in New York, for example, is striking. The series is not as definite from north to south as it was on map 1, but by grouping the localities as is shown by the broken lines on map 2 there can be demonstrated a fair uniformity within regions and great variation between regions. The ten- dency is not all in one direction as it was in pubescence of leaves. Starting with a low percentage (17) in the south, we find a rise as we go northeastward, a drop in the central region, and a sharp rise in the northern part. These facts favor the third hypothesis suggested above, namely, that as the species migrated there were slight changes in the frequencies of the characters due to chance or to unknown factors; with reference to pubescence of leaves all the changes were in the same direc- tion, and with reference to the glands of the pedicels they were in different directions in different regions. Two other characters appear in the key on page 303. One of these is the presence of subsessile glands instead of definitely stalked glands on the pedicels. Only two individuals have been seen with this charaeter, one from Chenango Bridge (New York no. 8) and the other from Sideling Hill (Pennsylvania no. 9). This small occurrence is of little significance except for com- parison with the situation in R. parviflorus, where in the plant of the Great Lakes such subsessile glands are found on only about 1 per cent of the plants (26 out of 2191 individuals col- lected), but in the western range of the species occur in much larger numbers (68 of 109 plants from the Black Hills of South Dakota, 98 of 242 plants from Montana, Wyoming and Utah, and 71 of 123 plants from Colorado). The remaining character used in the key lies in the stalked glands on the lower surface of the veins; ordinarily these do not exceed 1 mm., but on a majority of the plants from Indiana and from a very few else- where are hairs 1-2 mm. or more long. The occurrence of this character in large proportions only in the unglaciated part of Indiana is important, for it indicates a much more ancient mi- gration from the Alleghenian center to Indiana than to New York and New England. Previous to the more recent glacia- [Vor. 28 310 ANNALS OF THE MISSOURI BOTANICAL GARDEN tions, at least, R. odoratus migrated westward, perhaps chang- ing the frequency of some characters as it migrated. Either because of these changes or because of isolation as small popu- lations’? in the southern Alleghenies and in Indiana the plants of these two regions became different: the eastern plants be- came predominantly velvety-leaved and those of Indiana en- tirely glabrous-leaved; in the East the presence of glands on the upper leaf-surfaces became generally associated with vel- vety lower surfaces while in Indiana they appeared freely associating with nearly glabrous lower surfaces; the eastern plants retained short glands on the pedicels in many cases while those of Indiana eliminated this character and retained only long glands; and, finally, the majority of plants of Indiana retained long glands on the veins of the lower leaf-surfaces, while only a small minority of eastern plants showed this char- aeter. Available material is insufficient for further study of this phase of migration and isolation (see map 12 for an indi- cation of the proportion of the range of the species represented by mass collections), but collections from Virginia to Indiana and southward should yield interesting evidence. It is now possible to combine data concerning all the char- acters discussed above, to discover closeness of relationship of plants in different regions. For this purpose, localities whose populations show the closest resemblance, all characters con- sidered, have been grouped into regions. For example, соПес- tions 7-4 in New York, in the vieinity of Rochester (Station New York 1 on maps 1 & 2), agree in having a rather high per- centage of glabrous leaves (86, 100, 75, 66, on map 1) and a rather low percentage of short-glanded pedicels (20, 0, 6, 33, on map 2). 'These four collections are thus treated as one re- gional unit, labelled ** Rochester’’ on map 3. Localities 7 and 8 in Pennsylvania are in close agreement (both 0 for glabrous leaves, 42 and 64 respectively for short glands) and so are grouped together. The resulting regional groupings are shown on map 3, where each region is vertically shaded and labelled with a name. These regions appear to show little or no relation to physiographic features. " Compare Dobzhansky, Genetics and the origin of species, pp. 118-148. 1937. 1941] FASSETT—MASS COLLECTIONS: RUBUS ЕЕ The comparative values of figures from each region may be judged from the figure in parentheses just below each regional name on map 3, which indicates the number of individuals col- lected in that region. The numbers from Ontario and Mauch Chunk (Pennsylvania 1 and 2) are each only 14, so small that these regions are omitted from the map. On map 3 are also re- corded the percentages for each of the four characters in which the plants vary, obtained by totaling and averaging for all the plants in each region; these percentages are combined into a pie-diagram for each region. For example, among the 60 plants from Maine there were 24 of f. glabrifolius and 11 of f. heteradenius, a total of 35 plants with leaves not velvety be- neath, or 58 per cent ; this is indicated on map 3 by a 58 placed in the upper quadrant of the pie-diagram for the Maine region. There were 24 f. glabrifolius and 12 f. bifarius to total 36 indi- viduals with short glands on the pedicels, or 60 per cent; this is indicated by a 60 in the left quadrant in the pie-diagram. Subsessile glands on the pedicels, and long-stalked glands on the leaves below, are not represented in Maine, and this is indi- cated by zeros placed, respectively, in the lower and in the right quadrants. In the pie-diagrams on map 3, figures of 100 or 0 indicate uniformity in the characters concerned, while figures ap- proaching 50 indicate that the plants of the region are vari- able in the characters concerned. From the figures on map 3 may be calculated the average of percentage differences between any two regions. To compare, for example, Maine with northern Vermont: velvety and glandular leaves vs. non-velvety and non-glandular leaves, shows a difference of 5 (63 per cent minus 58 per cent), long glands on the pedicels vs. short glands shows a difference of 44 (60 per cent minus 16 per cent), and there is no difference in the other two characters. The average of percentage differ- enees (5 plus 44 plus zero plus zero, divided by 4) is 12. A com- parison of Maine with Indiana, however, gives a much higher figure, adding 42 (100 minus 58), 60 (60 minus 0), 0 (0 minus 0), and 59 (59 minus 0), to average 40. Table ш shows in tabular form the averages of percentage [Vor. 28 312 ANNALS OF THE MISSOURI BOTANICAL GARDEN differences between all regions. From it may be made a number of correlations with the assumed post-glacial history of the species; these become more obvious when the data are trans- ferred to maps. TABLE III AVERAGES OF DIFFERENCES OF PERCENTAGES BETWEEN REGIONS E E Е E | o $ = al Е А >, : TEE- яя = моа | а 2/2/2/2/8/8] аат Н НЕ 3181318 333488383 — = Ф ЕАН, — | 12 | 23 | 13 | 11 | 16 | 19 | 16 | 16 | 19 | 17 | 19 | 40 Northern Vermont 12 | — 10) 3| 41 6| 9| 8] 9 17 24110 Roches 23 | 10 | — | 11 | 13 | 15 | 18 | 17 | 7128 | 35 | 19 | 18 seca 18| 8|11|—| 7| 9/12] 9| 7182611 28 Western New York 11] 4/18] 7|—| 3] 7| 8/[13/19/|26,11|27 Allegany 16| 6/15) 9| 3|—| 9| 10| 15| 20127 | 12 | 25 rie 19| 91181121 71 9|—| 3118/13/20| 438 Bear Mt. 16| 8/17} 9| 8/10] 3—1 151 11| 18 4 on 16| 9| 7| 71 13| 15 | 18|15| —1 21| 28| 1812 Sehuylkill 19 | 17 | 28 | 18 | 19 | 20 | 13 | 11 | 21 | — | 7 45 Tuscaror 17 | 24 | 35 | 25 | 26 | 27 | 20 | 18 | 28 — | 16 | 52 Potomac 19 110 | 19 | 11| 111 12| 41 3181 91 16| — Indiana 40 | 28 | 18 | 28 | 27 | 25 | 33 | 34 | 25 | 45 | 52 | 37 | — In general, regions which are close to each other show closer relationship than do regions more distant from each other ; this is to be expected in a freely interbreeding population spread over a large area. But certain irregularities in this general pattern will appear in a species whose range has comparatively recently been altered by such an event as glaciation. This is brought out in map 4, which shows the close interrelationship of the regions across northwestern Pennsylvania, southern New York, and western New England (recently glaciated areas!!), fairly close interrelationship among the pre-Wiscon- sin regions from central Pennsylvania southward, but rather distant interrelationship between the pre-Wisconsin regions and the post-Wisconsin regions. This is brought out further in map 9 which shows only the more closely related regions. а The Allegany region was not actually glaciated, but was a tongue of unglaeiated territory only 10 miles across and was probably untenable for В. odoratus during glaciation. 1941] FASSETT—MASS COLLECTIONS: RUBUS 518 The Potomae region (map 6) and Ше Easton region (map 7) show relationship closer to most post-Wisconsin regions than to the much nearer pre-Wisconsin regions. This may be because the pre-Wisconsin stations have been isolated for a long time. Their close relationship with post-Wisconsin re- gions may indicate that much of the population of these regions was derived from the Potomae and the Easton regions. Furthermore, it seems to be more than a coincidence that in each general direction from the Potomac region, there is some correlation between geographic distance and distance of rela- tionship—this is shown on map 6 by the sets of parallel lines emerging from the Potomae region. The Easton region is similar to the Potomac region in this respect.!? The relation of each of these regions to other pre-Wisconsin regions is gen- erally more remote than to the post-Wisconsin regions, and shows no correlation with geographic distance. The relationships between the Tuscarora region and the other regions (map 8), and between the Schuylkill region and the other regions (map 9), show no correlation with distance. These regions are in general less closely related to the post- Wisconsin regions than are the Potomae region and the Easton region. Perhaps, then, there was greater contribution to the post- Wisconsin populations of the glaciated regions from the Potomae region and the Easton region than there was from Tuscarora region and the Schuylkill region. It must be borne in mind that other regions, not sampled in this study, might throw more light on this phase of the subject. The relation of the Indiana region to all other regions is distant, and shows little or no correlation between geographic distance and relationship (map 10). This is compatible with the assumption that the isolation of R. odoratus in unglaciated southern Indiana greatly antedates the postglacial migration of that species into New York and New England. It is conceivable that with full collections from every point in the range of the species, a map could be constructed along “The Easton region was covered by the older ice sheets but not by the Wisconsin. Sinee the Easton region is represented in this study by a eolleetion from but one loeality, its figures are probably of less value than those from a region like Potomae with its six collections. [Vor. 28 314 ANNALS OF THE MISSOURI BOTANICAL GARDEN the lines of map 5, which would actually show the course of postglacial migration. How far map 5 falls short of represent- ing the full faets may be judged from map 12, which shows the amount of coverage of the range of R. odoratus by the avail- able mass collections. П. Variation Іх В. parviflorus 1. THE KINDS OF VARIATION The more conspicuous variations within R. parviflorus are of the following ten types. On most of these characters, sub- divisions of the species have, by one taxonomist or another, been proposed. To judge from the literature, no attempt has previously been made to assemble from each region a large amount of material for the purpose of determining just how many subdivisions of the species might grow together. This discussion is based primarily on 104 such collections totaling 2137 individuals; the proportion of the range thus covered may be judged by comparing map 24, which shows the known range of R. parviflorus, with map 19, showing the locations of mass collections. Material of the species in the Gray Herbarium, the New York Botanical Garden, the University of California, Pomona College, and the University of Wisconsin has also been studied. 1. Calyx villous or not villous. Individuals with villous calyces ap- pear throughout a large part of the range of the species, always in company with a great majority of plants with the calyces not villous. Encountering villous ealyces only in plants from the Great Lakes region in what he called var. genuinus and in var. velutinus from the Coast Ranges of California, Professor Fernald stated (l. c., page 275) that these two varieties were closely related. When examined in the light of subsequent collections, which show villous calyces to oceur through- out the range of the species, they seem to represent rather a sporadic variation, present in about a third of the plants from the Coast Ranges and а much smaller proportion of individuals from the rest of the range. In maps 20 and 21 the percentage of plants in each region with villous calyces is shown by the figure in the lowermost sector of each pie- diagram. Map 13 shows the occurrence of plants with villous ealyees, except in California. 1941] FASSETT—MASS COLLECTIONS: RUBUS 315 2. Glands or setae of the pedicels. These are of four types. (1) The glands may be on stalks about 0.5 mm. long. (2) The glands may be on stalks reaching 1-2 mm. in length, or rarely even longer. There seems to be no correlation of gland color with length of the stalk. (3) The glands may be sessile or on stalks less than one-fourth as long as the diameter of the gland, or rarely the glands may be entirely absent. (4) There may be, instead of glands, broad-based setae; this type is very rare, having been observed in only two collections. Types 1-3 occur practically throughout the range of the species, and in combination with any of the other characters. The occurrence of glands about 0.5 mm. long is shown on maps 13, 16, and 17; the occurrence of glands 1 mm. or more long is shown on maps 14 and 15; the occurrence of sub- sessile glands is shown on map 18. In addition to the ranges shown on these maps, types 1-3 also are found in the Coast Ranges of Cali- fornia. Types 1-3 are not always clear-cut, and may grade into each other. 3. Leaves velvety beneath or not velvety beneath. Lower leaf- surfaces grade from perfeetly glabrous, through a condition where scattered appressed hairs are present, to densely velvety. Some of the intermediates make it difficult to maintain two categories. In the pres- ent study, those leaves which feel velvety to the touch are treated as sueh. Throughout most of the range of the species velvety leaves show little or no correlation with other characters; they are not common on plants with sessile glands on the pedicels, but such a combination oc- easionally occurs, particularly in California. In the Great Lakes re- gion 1093 out of 1346 individuals examined had velvety leaves; in the Coast Ranges velvety leaves are the rule almost without exception, being present on all but one of the 327 individuals examined ; in the intermediate region they are in the minority, occurring in 231 out of 660 plants examined. Maps 14 and 16 show the occurrence of velvety leaves, and maps 13, 15 and 17 show Ше occurrence of glabrous or sub- glabrous ones. 4. Petioles with minute puberulenee beneath the glands, or without such puberulenee. When studied in the Gray Herbarium by the writer, this character appeared to be of some importance, for, while common in the western states, puberulent petioles appeared only on one sheet from the Middle West, and this was from Keweenaw Point, Michigan, where a number of far western plants are known to occur. But exam- ination of the mass collections made about the Great Lakes showed plants with puberulent petioles from throughout the entire range. Its complete lack of taxonomie value in this region is shown by the fact (Vou. 28 316 ANNALS OF THE MISSOURI BOTANICAL GARDEN that the leaves on the upper part of a stem may have puberulent petioles while the leaves farther down may have the petioles glabrous beneath the glands. It was just coincidence that the only sheet in the Gray Her- barium from the Middle West, showing puberulent petioles, was from Keweenaw Point. One thing about this character may be significant, however; while the puberulenee of the petioles appears to be, for the most part, a response to exposure, nevertheless nearly all the plants from the Coast Ranges of California have all their petioles puberulent. The exeeptions are in a eolleetion from Eureka (Gillespie 15398, described in tables v, ут & уп, pages 329, 332, 333), in which there is great variation. Most of the individuals have the stem and petioles copiously villous (eharaeter no. 6), as var. velutinus, some do not, and a few are intermediate. Some have the petioles puberulent and others do not. That this is not entirely due to habitat was demonstrated by growing the offspring of a few of these plants in the greenhouse at the Univer- sity of Wisconsin. Plants grown from different seeds in one synearp, under identieal conditions in the greenhouse, showed the same varia- tions that were present in the original collection. 5. Pedieels villous or not villous. Of the 248 individuals in the mass collections from the Coast Ranges of California, all but 42 are char- acterized by the presence of spreading hairs exceeding the glands on the pedieels. Of nearly 2000 individuals from the rest of the range of Е. parviflorus only 3 have villous pedicels: two are from Wisconsin and one from southern California. A collection from the District of Renfrew, Vancouver Island, C. O. Rosendahl & Carl J. Brand, 3, also has long hairs on the pedicels. This collection has been seen in the Herbarium of the University of Cali- fornia. The same collection, as it appears in the Herbarium of Pomona College, shows this characteristic to a lesser degree, and is probably from a different clone. 6. Stem, stipules and petioles villous or not villous. Plants of the Coast Ranges nearly always have these organs more or less villous (rarely sparsely so), and so are separated as var. velutinus, as distin- guished from the rest of the species which lacks this character. For the correlation of this character with others see the discussion of var. velutinus, on page 318. 7. Cutting of leaves. В. nutkanus f. lacera Kuntze, or В. parvi- florus var. bifarius f. lacera Fernald,'* has leaves cleft 25-34 to the base, and А. parviflorus f. pedatifidus Hermann" has them cleft en- tirely to the base. ? Meth. Sp. 102. 1879. 5 Rhodora 37: 61, pl. 326, fig. 2. 1935. ^ Rhodora 37: 281. 1935. 1941] FASSETT—MASS COLLECTIONS: RUBUS 317 8. Cutting of sepals and petals. R. parviflorus var. Fraserianus J. К. Henry,'* or В. parviflorus var. bifarius f. Fraserianus Fernald,'* has the petals laciniate-dentate at summit, and a sheet from the Santa Cruz Mountains of California has the ealyx-lobes laciniate. 9. Color of petals. The petals, white in most plants, are recorded on the labels as turning pink with age on sheets from Del Norte County, California, and from Mt. Hamilton, Washington. 10. Size of flowers. This varies considerably from clone to elone, but there seems to be no justifieation for the division into a larger-flowered western variety and smaller-flowered variety about the Great Lakes, such as has been proposed by Torrey & Отау!" and by Farwell.* The smallest-flowered plants seen by the writer comprise a collection of 21 individuals from Tuolumne County, California (Wiggins 9245, de- seribed as California 1 in tables v and v1, pages 329 and 332) ; the flowers are about 2.5 em. in diameter, with a ealyx sometimes as little as 12 mm. in diameter. This small-flowered extreme may be identified, by the latest taxonomic treatment of the В. parviflorus group,” as ** var. gran- diflorus.’’ It is evident that here are variatiols of several series. Groups 7, 8, 9 and 10 include sporadie anomalies of rare occurrence, apparently occurring anywhere throughout the range of the species without rela- tion to each other or to the other series of variations. Groups 1-6 are concerned with pubescence. Groups 1, 2 and 3 have been used by Pro- fessor Fernald to distinguish varieties; an appraisal of these varieties in the light of new and more ample material follows. 2, THE **VARIETIES"' OF &. parviflorus 1. R. parviflorus var. genuinus Fernald, Rhodora 37: 277. 1935. This is distinguished by its villous calyx and is said to be confined to the region of the upper Great Lakes; the present writer has collected it from a much wider range (map 13). The case against var. genuinus, as defined by Fernald, lies in the fact that its main character, a villous calyx, may appear anywhere throughout the range of the species, in various combinations with other characters, for plants with villous calyx may have pedicels with long glands (like vars. hypomalacus and heteradenius), or short glands (like vars. bifarius and grandiflorus), or subsessile glands (like vars. scopulorum and parvifolius) ; the leaves may be glabrous or glabrate beneath (like vars. heteradenius and grandiflorus) or soft-pubescent (like vars. hypomalacus and bifarius). и Torreya 18: 54, fig. 1. 1918. 18 Am. М1а1. Nat. 11: 281. 1929. " Fl. М. Am. 1: 450. 1840. в Rhodora 37: 276. 1935. - [Vor. 28 318 ANNALS OF THE MISSOURI BOTANICAL GARDEN 2. R. parviflorus var. velutinus (Hook. & Arn.) Greene, as defined by Fernald, l. c., is distinguished from vars. hypomalacus and bifarius only by its villous calyx. But all individuals of these so-called varieties, as they oeeur in the Coast Ranges (to which region, as Fernald eor- rectly states, var. velutinus is confined), are characterized by having long spreading hairs on the nodal regions, often the internodes, and the petioles, and by strigose stipules; these characteristics do not occur elsewhere in the range of В. parviflorus. In other words, the plants of the Coast Ranges may or may not have the calyx villous and may have the glands of the pedicels long or short, but they are set off from the rest of the species by the pubescence of the nodes, petioles and stipules. Var. velutinus, as redefined on the basis of this character, also differs in tendencies from the rest of the species as follows: in var. velutinus 83 per cent of the individuals in the mass collections have long hairs on the pedicels, 33 per cent have villous calyces, and 100 per cent have leaves velvety beneath, whereas in the rest of the range of R. parviflorus only 0.0015 per cent of the individuals have long hairs on the pedicels, 0.019 per cent have villous calyces, and about 64 per cent have leaves velvety beneath. These facts are brought out visually in map 20, where the pattern of the pie-diagram for the Coast Ranges is conspicuously different from all the others. Var. velutinus is there- fore retained as a valid geographic variety, but on different characters, in part, from those relied on by Professor Fernald. 3. В. parviflorus var. hypomalacus Fernald, 1. c., p. 277, (4) var. heteradenius Fernald, 1. c., p. 279, (5) var. bifarius Fernald, 1. c., р. 280, and (6) var. grandiflorus Farwell, as redefined by Fernald, l. c., p. 281. The 2 characters of long glands vs. short glands on the pedicels, and leaves velvety beneath vs. leaves glabrous to glabrate beneath, oc- eur in 4 eombinations to make these 4 varieties. Their ranges as pub- lished by Professor Fernald (dots on maps 14—17 in this paper) are rather similar, the only significant departure from the general pattern being that var. grandiflorus alone oeeurs in South Dakota, southern Montana, Wyoming, Utah and Idaho. Even this amount of geographic segregation seems to break down with other collections, as shown by the x's on these same maps. Moreover there is very little segregation of these '' varieties"! in the field. Of 8 mass collections from Minnesota, 2 collections had 4 of these ‘‘varieties,’’ 2 had З ‘‘varieties,’’ 3 had 2 '* varieties," and 1 had 1 ‘‘variety.’’ This statement refers only to the 4 ‘‘varieties’’ whose names head this paragraph; in many cases others were also present. Of 6 mass collections from northern Wiseon- sin, 1 had 4 ''varieties," З had 3 ‘‘varieties,’’ 1 had 2 ‘‘varieties,’’ 1941] FASSETT—MASS COLLECTIONS: RUBUS 319 and 1 had 1 ‘‘variety.’’ Such cases might be cited indefinitely, but the mass collections are described in tables ут & уп. As stated in the discussion of that variety, the var. genuinus of Fernald breaks up on just the characters used by him to define vars. hypomalacus, heteradenius, bifarius and grandiflorus. Var. velutinus, as defined by Fernald, has subdivisions equivalent to vars. hypoma- lacus and bifarius. Var. scopulorum has a phase with velvety leaves and a phase with glabrous or glabrate leaves. It is eoneluded, then, that in the absenee of segregation of vars. hypomalacus, heteradenius, bifarius and grandiflorus on basis of geo- graphie distribution, habitat, or loeality, they represent forms only. 7. R. parviflorus var. scopulorum (Greene) Fernald, and (8) var. parvifolius (Gray) Fernald. These are distinguished from each other by the height of the plant, the width and pubescence of the leaves, the number of flowers in the inflorescence, and the length of the lowest pedicel. These distinctions do not hold in the field. Two collections in the gorge of Fish Creek Falls, Steamboat Springs, Colorado (pl. 10, fig. 4), consisted of some plants with sessile glands and others with stalked glands on the pedicels. The characters of those with sessile glands may be seen, in table iv, to make very diffieult any separation into two groups such as have been named scopulorum and parvifolius. These two so-called varieties with glands of the pedicel sessile or nearly so are recorded by Professor Fernald as occurring only from southern British Columbia, Oregon and Montana, southward to Ari- zona, New Mexico and northern Chihuahua. However, there are, in the herbaria of the University of California and of Pomona College, numerous specimens from California having the glands of the pedicels sessile or nearly so; they have leaves velvety beneath and cannot be placed by Fernald’s key. The present writer has collected plants with sessile glands (always accompanied by others with stalked glands) in the Black Hills of South Dakota and in the three Great Lakes states where R. parviflorus is known. Map 18 shows the occurrence of plants with sessile or subsessile glands, as originally mapped by Fernald, and as demonstrated by other collections not seen by him. The sessile or subsessile gland is not clear-cut from the stipitate gland; a distinction between them forms the primary division in Pro- fessor Fernald’s key. Var. scopulorum (including var. parvifolius) with glabrous leaves grades into var. grandiflorus, and its unnamed relative with pubescent leaves grades into var. bifarius. The case against vars. scopulorum and parvifolius, then, sums up as follows: the two are not distinct from each other in the field; they 320 CHARACTERS OF VAR. SCOPULORU ND TWO ке COLLECTIONS FROM FISH 7 AMBOAT SPRINGS, COLORAD [ Vor. 28 ANNALS OF THE MISSOURI BOTANICAL GARDEN VAR. PARVIFOLIUS, AND OF K FALLS, Я Fish Creek Fish Creek scopulorum parvifolius Falls I Falls II Height of plant 1-2 m. 1.5—6.0 dm. 0.4—1.5 m Width of leaves 1-3 dm. 0.5—1.3 dm. 1.2-2.4 dm 1.0—1.9 dm. Lower со Glabrous Minutely and | Glabrous to Glabrous to of lea sparsely distinetly distinetly pubeseent velvety velvety to glabrate Number of 3-7 1-2 (-4) 4 with 1 fl. flowers in 2 with 2 fl 5 with 2 fl. inflorescence 8 with 3 fl 7 with 3 fl. 4 with 4 fl 3 with 4 fl. 1 with 5 fl Length of 1.0-3.5 em. 2.0—6.0 em. 1.0 em. on 1 lowest 1.5 em. on 1 pedunele 2.0 em. on 3 2.0 em. on 6 2.5 em. on 3 2.5 em. on 1 3.0 em. on 2 3.0 em. on 5 3.5 em. on 2 3.5 em. on 3 4.0 em. on 6 4.0 em. on 1 4.5 em. on 1 5.0 em. on 1 5.5 em. on 6.0 em. on 1 6.0 em, on 1 occur with and grade into vars. grandiflorus and bifarius; the sub- sessile gland is not eorrelated with a glabrous or glabrate leaf; they are not confined to any limited part of the range of the species, but may rather be found in the same patch with the other ‘‘varieties’’ almost (or perhaps quite) everywhere that R. parviflorus grow Mr. Joseph Ewan, of the University of Colorado, who has kindly read the manuseript of this paper, comments as follows on the disposi- tion of var. parvifolius: “Та my opinion there is something valid about the thing isolated by Fernald, following earlier segregations, centering about New Mexico and Arizona, I have personally discovered, without any particular search in herbaria for them, sev- eral ‘good’ sheets of that entity. It is a zed by the concomitance of defi- nitely smaller leaves than any other variety or subspecies of the species, with ewer flowers to an inflorescence, and a slender habit. I do not feel that scopulorum is more than a form, of indefinite lines at best, but this southern phase of the species deserves, I believe, recognition. . Your collection from Fish Creek Falls, Steam- boat Springs, dips illustrates a — of the zone of overlap between var. parvifolius and f. scopulorum, if such a form is to be recognized, even demon- strating е. а hare swarm dating ao 2 long time ago in the movements of is Rubus.’’ 1941] FASSETT—MASS COLLECTIONS: RUBUS 321 These comments are introduced so that botanists in the Southwest may have opportunity to observe and perhaps decide upon the validity of this small-leaved phase of the species. For the present, var. parvi- folius is put into synonymy under f. scopulorum (page 323), and its revival would cause no nomenelatorial complication and require no new eombination. Here, then, is a set of characters appearing in different combinations, with any combination of characters being likely to occur anywhere in the range of the species. Such combina- tions of characters, lacking geographic ranges, are not usu- ally treated as varieties, but as forms, if they indeed receive any nomenclatorial recognition. Professor Fernald emphasizes?? the ‘‘disconcerting but in- disputable fact that these very differences in the distribution of glands in the inflorescence and of pilosity on the calices, branches and leaves or the absence of glands and pilosity from these areas, which mark the eight geographic segregates of Rubus parviflorus, are precisely the characters which are shuffled and reshuffled to add to the ever increasing score of ‘species’ of Blackberry (Rubus $ Eubatus)! In R. parviflorus not even our most ardent advocates of specific segregation, who have felt competent to make generic segregates, have noticed them; nevertheless, they are quite as conspicuous in В. parviflorus as in segregates of Rubus $ Eubatus, and if their phylogenetic importance is of equal value in the two sections, the Blackberries are eventually due for a pretty drastic realign- ment.’’ The ‘‘if’’ in the last statement was commented upon by Professor Fernald, a few years ago,”! in these very perti- nent words: **As pure logic, wholly dissociated from the actual vagaries of Nature, this may be conceded; but, surely, when applied in classification, the logie often fails ; charaeters which in one group are of great taxonomie importance in another may prove wholly unimportant and to be a series of nonconcomitant and unresolvable variables." The taxonomic value of characters lies less in what they do in other groups, or in how conspicuous they may be, than in how they act. Pubescence of stem and peduncles is used as a 2 Rhodora 37: 274. 1935. 7 Rhodora 35: 165. 1933. [Vor. 28 322 ANNALS OF THE MISSOURI BOTANICAL GARDEN primary character in separating species of Stellaria,?? is of varietal rank in Ranunculus abortivus, *? and serves only to separate forms in Oxalis europaea.** Again in О. europaea, while differences in pubescence of stem and pedicels are used only to separate forms, the presence or absence of scattered hairs on the upper surface of leaflets differentiates varieties ; this is not because pubescence of leaflet surface is more con- spicuous or perhaps of greater importance to the plant, but be- cause plants with such pubescence are found in different re- gions from those which lack it. This principle is summarized by Anderson?* when he writes: ‘‘Particularly significant is the fact that the difference between I. virginica and I. virginica var. Shrevei is of about the same order of magnitude as the differences between colonies of I. virginica var. Shrevei. It would indeed be possible to find two swamps in the same town- ship in southern Michigan whose iris populations have as great an average difference as that between Iris virginica of the At- lantic Coastal plain and Iris virginica var. Shrevei. But in this latter case the difference, slight though it is, characterizes a whole region and has superimposed upon it the varying pat- tern of colony differences in each region." Summary: With the exception of var. velutinus, the vari- eties of Rubus parviflorus described by Professor Fernald ap- pear invalid to the present writer, for the following reasons. (1) The entities are not confined to the ranges described by him, but each occurs nearly throughout the range of the species. (2) Rare is the colony that is composed of but one entity, and many colonies are composed of a mixture of as many as five entities. (3) There seems to be little or no association of char- acters one with another, since they may appear in almost any combination, many of these combinations being impossible to place in Fernald's key to varieties. 3. THE FORMS OF R. parviflorus The advisability of giving a formal name to each recombina- tion of certain pubescence characters may be debatable, and a Gray's Man., ed. 7, p. 381. 1908. ^ Wiegand, Rhodora 27: 135. 1925. ? Fernald, Rhodora 40: 417. 1938. ? Ann. Mo. Bot. Gard. 23: 494. 1936. 1941] FASSETT—MASS COLLECTIONS: RUBUS 323 a list of nearly a score of new names within one species is apt to make the taxonomist gasp.?* The present writer is of the opin- ion that forms should be named if the naming of them will serve a purpose. A group is named so that ‘‘we may be understood when we wish to speak of it.'?* There is to be occasion to speak of these combinations of characters, to discuss the sig- nificance of the occurrence of each, and eventually to distrib- ute material illustrative of them, and these procedures will be facilitated by the assignment of names to them. The following were treated as varieties by Professor Fernald R. PARVIFLORUS f. hypomalacus (Fernald) n. comb. R. parviflorus var. hypomalacus Fernald, Rhodora 37: 277. 1935. В. PARVIFLORUS f. heteradenius (Fernald) n. comb. В. parviflorus var. heteradenius Fernald, l. c., 279 В. PARVIFLORUS f. bifarius (Fernald) n. comb. В. parviflorus var. bifarius Fernald, 1. c., 280. В. PARVIFLORUS f. glabrifolius, n.f., foliis subtus glabris vel glabratis vel sparse appresso-pilosis ; реда glandulis stipitatis 0.5-1.0 mm. longis; calycibus non у111081$.—ТҮРЕ, in Herb. Univ. Wis.: small pateh along roadside, 3.6 miles west of Meldrum Bay, Ontario, Aug. 5, 1939, Fassett 20567. В. parviflorus var. grandiflorus Fernald, 1. с., 281; perhaps В. parviflorus var. grandiflora Farwell, Am. Midl. Nat. 11: 263. 1929. В. PARVIFLORUS f. scopulorum (Greene) n. comb. В. nutkanus var. scopulorum Greene ex Focke, Bibl. Bot. 17, pt. 72: 124. 1911. В. parvi- florus var. scopulorum Fernald, 1. c., 283. В. nutkanus var. parvifolius Gray, Pl. Fendl., Mem. Am. Acad. II, 4: 42. 1849. R. parviflorus var. parvifolius Fernald, 1. с., 284. In uniting these two varieties as one form, it seems advisable to use the name scopulorum rather than the one antedating it (in the varietal rank), which not only emphasizes a character not necessarily applicable to the plant but whose resemblance to the specific name is liable to invite confusion. В. PARVIFLORUS f. Nuttallii (Torr. & Gr.) n. comb. В. parviflorus Nutt., Gen. 1: 308. 1818. В. nutkanus В Nuttallii Torr. & Gr., Fl. N. 2 «Tf one were to attempt to name formae in Rubus the result might be appalling. One may note minor variations but there is no necessity to give them Latin names or to restrict them to formal categories.’’ L. H. Bailey, Gentes Herbarum 5, Fase. I. 19. 1941. ? The Vienna Rules of Nomenclature. Rhodora 9: 36. 1907. [ Vor. 28 324 ANNALS OF THE MISSOURI BOTANICAL GARDEN Am. 1: 450. 1840. R. parviflorus var. genuinus Fernald, l. c., 277. The phrase in Nuttall’s description, ‘‘segments of the calix villous,’’ would seem to indicate the plant treated by Professor Fernald as var. genuinus, but the Nuttall specimen from Michilimackinak Island is the plant ealled by Fernald var. grandiflorus, and in this treatment f. glabrifolius. The name on the label is not preceded by an asterisk so its position as type is not seeure, and the label bears the note **lost all but this fragment." Perhaps the villous calyx was observed by Nuttall on material subsequently lost. On the sheet with the Nuttall specimen are three others: one was collected by R. H. Kern in New Mexico; the second by Dr. Tiling in Sitka; and the third, without label, is definitely the var. velutinus of the Coast Ranges, having strongly villous calyces, and is mounted so that it partly overlies and partly underlies the Nuttall plant. Less confusion will be eaused by following Nuttall’s description, as Professor Fernald did, than by trying to redistribute names on the basis of this sheet. The following forms are the results of other combinations of the characters used in the determination of varieties by Pro- fessor Fernald: R. PARVIFLORUS f. trichophorus, n.f., f. bifarium simulans sed ealy- cibus villosis. Туре, in Herb. Univ. Wis.: Bark Point, Bayfield Co., Wis., July 10, 1938, N. C. Fassett & J. T. Curtis 20545. R. PARVIFLORUS f. villosus, n.f., f. hypomalacum simulans sed ealy- cibus villosis. Typr, in Herb. Univ. Wis.: near head of Nigger Grade, Palomar Mt., San Diego Co., Calif., Aug. 4, 1938, Frank F. Gander 6239. R. PARVIFLORUS f. allocalyx, n.f., f. scopulorum simulans, sed ealy- cibus villosis. ТҮРЕ, in Herb. Univ. Wis. : 7 miles north of Savoy, S. D., June 26, 1939, Fassett 20201. В. PARVIFLORUS f. micradenius, n.f., f. scopulorum simulans, sed foliis subtus velutinus. TYPE, in Herb. Univ. Wis.: gorge below Fish Creek Falls, Steamboat Springs, Colo., July 3, 1939, Fassett 20193. A rare type of epidermal outgrowth on the pedicels, not accounted for in Professor Fernald’s treatment, is a glandless flattened trichome 0.5-1.0 mm. long. This is associated with glabrous leaves in one case, and with velvety leaves in another. R. PARVIFLORUS f. adenius, n.f., pedicellorum setis 0.5—1.0 mm. longis eglandulosis; foliis subtus subglabratis. Туре, in Herb. Univ. Wis.: sand back of dunes, Whitefish Bay, Door Co. Wis., July 31, 1938, Fassett 20211. 1941] FASSETT— MASS COLLECTIONS: RUBUS 325 R. PARVIFLORUS f. acephalus, n.f., f. adenium simulans sed foliis sub- tus subvelutinis. Typr, in Herb. Univ. Wis.: Port Orford, Ore., Aug. 12, 1938, Doris K. Gillespie 15399. В. parviflorus var. velutinus, confined to the Coast Ranges of California, shows variations just paralleling those in the rest of the species. In studying the fluctuations of var. velu- tinus the writer has had the advantage of the loan of the ma- terial from the herbaria of the University of California and of Pomona College, as well as the most generous mass collections made by Drs. Gillespie, Mathias, Schreiber, Constance and Wiggins. To keep clear the parallelism of forms, each one in var. velutinus is here given a name based on that of the corresponding form in the widespread phase of the species. R. PARVIFLORUS var. VELUTINUS f. parbifarius, n.f., caulibus, petiolis, pedicellis peduneulisque villosis; foliis subtus velutinis; pedicellorum glandulis stipitatis 0.5-1.0 mm. longis. Туре, in Herb. Univ. Wis.: north slope of Strawberry Creek Canyon, 0.7 miles above its mouth, 1100 ft. alt., Berkeley, Calif., Aug. 4, 1938, L. Constance 2397. R. PARVIFLORUS var. VELUTINUS f. praebifarius, n.f., f. parbifarium simulans sed pedicellis non villosis. ТУРЕ, in Herb. Univ. Wis.: Eureka, Calif., Aug. 6, 1938, Doris K. Gillespie 15397. R. PARVIFLORUS var. VELUTINUS f. paratrichophorus, n.f., f. parbifa- rium simulans sed calycibus villosis. ТүрЕ, in Herb. Univ. Wis.: under Sequoia sempervirens, 5-1 mile from the ocean, Palo Colorado Canyon, Santa Lueia Mts., about 12 miles south of Carmel, Calif., Aug. 7,1938, Mildred E. Mathias 1389. В. PARVIFLORUS var. VELUTINUS f. parahypomalacus, n.f., caulibus, petiolis, pedieellis, peduneulisque villosis; ealycibus non villosis vel subvillosis ad basem; pedieellorum pedunculorumque glandulis stip- itatis 1-2 mm. longis. Туре, in Herb. Univ. Wis.: 1.5 miles southeast of Abbot's Lagoon, Marin Co., Calif., Aug. 6, 1938, Beryl O. Schreiber 2538. R. PARVIFLORUS var. VELUTINUS f. isohypomalacus, n.f., nodis stip- ulisque villosis; pedicellis ealyeibusque non villosis; pedicellorum pe- не glandulis stipitatis 1.0-1.5 mm. longis. Туре, in Herb. Univ. Wis.: Elk River 6 miles south of Eureka, Calif., Aug. 8, 1938, Doris K. Gillespie 15398. В. PARVIFLORUS var. VELUTINUS f. parvillosus, n.f., caulibus, petiolis, peduneulis, pedieellis, ealyeibusque villosis; pedicellorum peduneu- lorumque glandulis stipitatis 1-2 mm. longis. ТуРЕ, in Herb. Univ. [Vor. 28 326 ANNALS OF THE MISSOURI BOTANICAL GARDEN Wis.: 1.5 miles southeast of Abbot's Lagoon, Marin Co., Calif., Aug. 6, 1938, Beryl O. Schreiber 2538. R. PARVIFLORUS var. VELUTINUS f. paramicradenius, n.f., caulibus stipulisque sparse villosis; pedicellorum peduneulorumque glandulis subsessilibus. Туре, in Herb. Univ. Wis.: Eureka, Calif., Aug. 6, 1938, Doris K. Gillespie 15397. KEY TO VARIETIES AND FORMS OF Z. PARVIFLORUS a. Stem, stipules and petioles without long spreading hairs; leaves of thin tex- ture, not rugose above, their lower surfaces glabrous, glabrate, appressed- pubescent, or moderately velvety; pedicels only very exceptionally with long hairs exceeding the glands; petioles with or without a minute puberulence b. Pedicels with setae which are usually gland-tipped, the stalks more than 4 times as long as the diameter of the gland c. Setae of the pedicels gland-tipped m. long d. Setae 1-2 m a Leaves eod emi beneath LO CAV oot адът аба еее Оа ойе f. hypomalacus f. Calyx with villous hairs hiding the glands................ f. villosus е. Leaves not soft-pubescent Бепеаёћ.................... f. heteradenius d. Setae mostly about 0.5 mm. lon g. Leaves soft-pubescent beneath bu, о РУ ИР 11 тү f. bifarius h. Calyx with villous hairs hiding the glands.......... f. trichophorus g. Leaves not soft-pubescent beneath © БОН 1 Пе Р.А f. glabrifolius i. Calyx with villous hairs hiding the glands.............. f. Nuttallii c. Setae broad-based and not gland-tipped j. Leaves soft-pubescent Ъепеаїһ........................... f. acephalus j. Leaves not soft-pubeseent Ъепеа{һ.......................... f. adenius b. Pedicels биб or with sessile "m or with glands on stalks less than 4 times as long as the diameter of the k. Leaves soft-pubescent Ъепеаїһ........................... f. micradenius k. Leaves not soft-pubeseent beneath PIDEN ME о О МР in exe 0640504066 RAE eds f. scopulorum 1. Calyx with villous hairs hiding the glands.................. f. allocalyz a. rato ast and petioles with long spreading hairs; leaves of thickish tex- ose above, their lower surfaces always densely velvety; pedicels seh tae with long hairs exceeding the — petioles always with a minute афу) much shorter than the рЈапӣз....................... var. velutinus . Pedieels with copious villous hairs Зн the glands п. Gland-tipped setae of pedicels 1-2 mm. long о. Calyx not villows.......c0s0cseeseee’ var. velutinus f. parah о. Calyx with villous hairs hiding the glands... .var. velutinus f. poseen п. Gland-tipped setae of pedicels about 0.5 mm. lon ОО ETE Ne var. velutinus f. parbifarius 1941] FASSETT—MASS COLLECTIONS: RUBUS 527 о О" var. velutinus f. paratrichophorus m. Pedicels without villous hairs exceeding the glan q. Glands of pedicels on stalks mm. or more dt r. Glands of pedieels on пан 0.5-1.0 mm. long........ ern и. var. velutinus f. praebifarius г. Glands of pedicels on stalks 1.0-1.5 mm. 10п0.................... аке AMO жа veinte Ma ol оозе ated Is Союз var. velutimus f. isohypomalacus q. Glands of pedicels sessile or nearly so....var. velutinus f. paramicradenius 4. OCCURRENCE OF FORMS IN MASS COLLECTIONS Table v is a list of localities where mass collections have been made. Each is numbered, and the numbers refer to local- ities shown by slanted figures on map 19. Table ут lists the number of individuals of each form represented in each col- lection. TABLE V ONTARIO: 1. Scattered in woods along the road to Cameron Lake, 6 miles south of Tobemory, Aug. 3, 1938, 20563. 2. Small patch in clearing, 8 miles east of Mel- drum Bay, Aug. 4, 1939, 20566. 3. Large patch along the road and a few in the woods, 7.2 miles east of Meldrum Bay, Aug. 4, 1939, 20565. 4. Small pateh along the road, 1.7 miles east of Meldrum Bay, Aug. 4, 1939, 20567. 5. Roadsides and openings in woods, 1 mile southwest of Meldrum Bay, Aug. 5, is 20568. 6. Small pateh along roadside, 3.6 miles west of Meldrum Bay, Aug. 5, 1939, 20569. 7. Widely scattered patches in spruce-fir and alder woods, 0.7 miles west of Meldrum Bay, Aug. 5, 1939, 20570. 8. Scattered in woods and on raised beach, a little along roadside, Meldrum Bay, Aug. 5, 1939, 20571. 9. Thicket and openings near Ster- birch woods on raised cobblestone (limestone) beach, Gross Cap, 7 miles northwest of St. Ignace, Aug. 1, 1938, 20602. 2. Woods at the head of a sand beach, Pointe aux Barques, Aug. 1, 1938, 20601. ү Roadside and pastured woods (aspen, birch, white cedar) 2 miles southwest of Garden City, Aug. 1, 1938, 20600. 4. pe along roadside and in ; Woods Die cedar, white spruce, birch), Fairpor ч > 1938, 20603. 5. Roadside and along path in woods (fir, white spruce, birch), 5 iles south of Ford River, July 31 ET 20580. 6. Abundant along the abandoned high- way and rare along A highway, north of Cedar River, Aug. 1, 1938, 20604. 7. Sparse among bracken in aspen woods, 6 miles north of Faithorn, J uly 30, 1938, 20598. 8. Small patch in aspen and Norway pine, 1 mile north of Faithorn, J My 30, 1938, 20599. 9. Seattered, woods and talus of West Bluff, Keweenaw Peninsula, July 29, 1938, #0404. 10. Seattered in fir-white cedar woods near Lake Upson, b Mohawk, July 29, 1938, 20556. 13. Scattered in woods at summit of Wheel Kate, South Range, July 29, ж, 20561. 14. Pathway up Wheel Kate, ст 29, 1938, 20577. 15. Roadside 2 miles east of Painesdale, July 29, 1938, 20560. 16. Woods and roadsides, Toivola, x 29, 1938, 20495. 17. Big Limestone Mountain, near [Vor. 28 328 ANNALS OF THE MISSOURI BOTANICAL GARDEN L'Anse, July 29, 1938, 20493. 18. Old embankment (highway or R.R.!), Lake Mine, July 29, 1938, 20578. 19. Road fill 2 miles southeast of Ontonagon, July 29, 1938, 20574. 20. Mouth of Iron River west of Ontonagon, July 29, 1938, 20562. 21. Near Ontonagon River, June 28, 1939, J. T. Curtis. 22. Woods and sunny bank, west of Carp Lake Mine, Poreupine Mts., July 29, 1938, 20579. 27. Woods and roadsides, south end of Lake Gogebic, July 28, 1938, 20575. Wisconsin: 1. Seat- tered along edge of woods, Lake Michigan shore, North Landing, at northern tip of Door Co., July 31, 1938, 20513. 2. Roadside, North Landing, July 31, 1938, 20514. 8. Коби solid along Lake Michigan shore, Rowleys Bay, July 31, 1938, 20508. 4. Roadside and woods (beech and white pine), North Bay, July 31, 1938, 20526. 5. Among white pine, hemlock, mountain maple, Baileys Harbor, Aug. 5, 8, Chester Cook, C92. 6. Woods and clearing near Kangaroo Lake, Baileys Harbor, July 31, 1938, 20515. 7. Heavy woods along roadside, Jacksonport, July 20, 1938, 7. M. Gilbert & N. C. Fassett 20509. 8. Same data, 20510. 9. Abundant in sand back of dunes, among oak, white pine, hemlock, birch and poison ivy, Whitefish Bay, July 31, 1938, 20211. 10. Road through woods, Lily Bay, July 31, 1938, E 11. Gills Roek, July 31, 1938, 20530. 12. Scattered along roadside among beech, birch, white cedar, Sister Bay, July 31, 1938, 13. Field and roadside, hm ula State Park, near Fish Creek, July 31, ds je 14. Peninsula State Park, July 22, 1939, 20512. 15. Wooded bank near Wildeat Lake, north of Boulder Junetion, July 28, 1938, 20543. 16. Jute Lake fire tower, northeast of Boulder Junetion, July 5, 1939, J. T. Curtis. 17. Barksdale, i^ 8, 1938, 20541. 18. Woods and path- ways, Oak Island, off Bayfield Co., July 9, 1938, N. C. Fassett, J. T. Curtis $ Louis Knowlton 20544. 19. Roadside and red e gully, Bark Point, July 10, 1938, Fas- sett ф Curtis 20545. 20. Steep banks below Manitou Falls, Pattison State Park, July 14, 1938, 20542. MINNESOTA: 1. Apparently adventive, roadside 4 miles south- east of Gull Lake, Gunflint Trail north of Grand Marais, July 12, 1938, Fassett $ Curtis 20588. 2. Near s Lake, 10.8 miles south of Gull Lake, July 12, 1938, Fassett $ Curtis 20533. 3. Woods and roadside 13.7 miles south of Gull Lake, July 12, 1938, Fassett $ Ait 20532. 4. Roadsides and beaches of Lake Superior near mouth of Kadunee River 8 miles east of Grand Marais, July 12, 1938, Fassett $ Curtis 20537. 5. Abundant in old = road 1.5 miles from shore, Lutsen, July 13, 1938, Fassett $ Curtis 20531. 6. Manitou Falls, July 13, 1938, Fassett $ Curtis 20534. 7. Швеп City, July 13, 1938, Fassett $ Curtis 20535. 8. Gooseberry River, July 13, 1938, Fassett $ Curtis 20536. Soura DaAkOTA: 1. Shady places in canyon south of Pluma, June 26, 1939, 20199. 2. Но» near Pluma, June 26, 1939, 20200. 3. North-facing bluff, eov near Lead, June 26, 1939, 20864. 4. Frequent in damp pine woods at alt. 2500 ft., Terry, pr! 2, 1939, H. C. Cutler 2622. 5. Woods 3 miles south of Savoy, June 26, 1939, 20198. 6. Open ground, 7 miles north of Savoy, June 26, 1939, 20201. Tu 1. Bordering an aspen grove, near Beaver Mines, about 12 miles west of Pincher — foothills region, southeast of Crow's Nest Pass, June 30, 1940, E. H. Mos . Under aspen and spruce, 10 miles west of Pineher Creek, July 1, 1940, Moss. 3. In nd ncn pine woods near Waterton, Waterton Lakes Park, June 29, 1940, Moss. 4. Edge of woods 1941] FASSETT—MASS COLLECTIONS: RUBUS 329 and close to shore, same region and date, Moss. Montana: 1. North slope, alt. 5500 ft., Bear Canyon east of Bozeman, Aug. 11, 1938, F. B. Cotner. WYOMING: 1. Sunny bank near Steamboat Point, Yellowstone National Park,* June 28, 1939, 20209. 2. Near Sylvan Pass, Yellowstone National Park, June 28, 1939, 20205. 3. Near Tower Falls Junction, Yellowstone National Park, June 28, 1939, 20208. 4. Thicket on east side of Jenny Lake, Grand Teton National Park, June 29, 1939, 20206. 5. Camp ground in Hoback Canyon, June 29, 1939, 20207. UTAH: 1. Wooded eanyon on U. S. 40 east of Salt Lake City, June 30, 1939, 20204. 2. Roadside in Little Cottonwood Canyon near its mouth, June 30, 19 20202. 3. Moist woods along stream, Tanners Flat, Little Cottonwood Canyon, Suis 30, 1939, 20203. 4. Moist woods along stream, American Fork Canyon 5 miles from its mouth, July 1, 1939, 20210. CoLorapo: 1. Scattered in spruce woods, alt. about 8000 ft., Rabbit Ear Pass, 14 miles southeast of Steamboat Springs, July 3, 1939, 20195. 2. Roadside and wooded bank, 9 miles south of Steamboat Springs, July 3, 1939, 20196. 3. Gorge "gin Fish Creek Falls, Steamboat Springs, July 3, 1939, 20193. 4. Open woods amp ground, Fish Creek Falls, July 3, 1939, 20197. 5. In woods along road to bn Creek Falls, July 3, 1939, 20194. 6. Moist shady hillside, alt. 10,000 ft., Conejos River Canyon, July 24, 1938, Francis Ramaley 15672. 7. Open aspen PE well- drained soil above the Mei bod, Papoose Creek, Yampa River drainage, Rio Blanco Co., Aug. 11, 1938, F. Ramaley $ J. Ewan 16478. OREGON: 1. McKenzie Pass, July 15, 1939, Garland M. Powell. 2. Shaded roadside іп ‘аек sand’’ (stabilized dune area), 1.5 miles northwest of Warrenton, Aug. 9, 1938, W. T. McLaughlin. 3. Port Orford, Aug. 12, 1938, Doris K. Gillespie 15399. CALIFORNIA: 1. On beach above Meadow Brook OR camp, alt. 5600 ft., Brightman Flats, Tuolumne Co., July 16, 1939, Ira L. Wiggins 9245. 2. Low dense, more or less ground-cover under trees (Pinus contorta Murrayana, Abies magnifica, Lonicera, Epilobium, Saliz), 4 mile north-northeast of Silver Lake, El Dorado National Forest, Amador Co., Aug. 28, 1938, Beryl O. Schreiber 2546. 3. North Fork San Jaeinto River, San Jacinto Mts., Aug. 14, 1938, Carl Epling. 4. Fir-spruce association, alt. 1675 m., near head of Nig- ger Grade, Palomar Mt., San Diego Co., Aug. 4, 1938, Frank F. Gander 6239. 5. 6 А Dense shade, il moist habitat, woodland, alt. 100 ft., First Valley, Inverness, Marin Co., Aug. 7 38, Schreiber 2539. 9. Among scattered dense clumps of ачи ое in predominantly grassland areas, alt. 25 ft., 2.5 miles north- ast of Point Reyes Lighthouse, Marin Co., Aug. 6, 1938, his bi 2537. 10. Thisketa on north slope of usd Creek Canyon, 0.7 miles above its mouth, 1100 ft. alt., Berkeley, Aug. 4, 1938, L. Constance 2397. 11. Along banks of Gazos Creek, about 5 miles inland, Ben Matas Co., Aug. 12, 1939, Ira L. Wiggins 92964. 12. Under Sequoia sempervirens, Y5—1 mile P the ocean, Palo Colorado Canyon, Santa Lucia Mts., about 12 miles south of Carmel, Aug. 7, 1938, Mildred E. Mathias 1589. * For aid in obtaining a permit to collect specimens in Yellowstone National Park, the writer is indebted to Dr. C. Max Bauer, Park Naturalist, and Mr. C Skinner, Asst. Chief Ranger. [Vor. 28 330 ANNALS OF THE MISSOURI BOTANICAL GARDEN TABLE VI OCCURRENCE OF FORMS OF TYPICAL R. PARVIFLORUS s| È E 315 2/sS|2/2|8 si св S Ir e 2) 3] 3 1141219 ЕЕ ИЕН Е oe ЖЕ оса» ONTARIO 1 18 14 1 2 2 1 3 6 40 4 5 2 5 3 4 6 8 7 1 3 8| 1110 1| 8 9 4 37 MICHIGAN т" 20 2 20 3 2/1/16] 1 4 3 19 5 17 6 2 23 1 7 1/13] 2 8 6 1 9 6 23 5 10 3 97 4 11 11 29 12 4| 11| 15 13 93 14 1 14 15 19 13 2 16 2 28 2 17 1 46 1 2 18 5 15 19 2 13 20* 12 21 5 4 22* 19 23 11 111 24 10 74 1 251 13 26 2| 1| 95 1 27 3 14 WISCONSIN 1 2 4 2 2 3 5 8 9 19 4 18 4 9 5 l 15 1 6 9 10 7 2 7| 1 1 * This collection appears to consist of one clone, and does not enter into the data for maps 20-23. t This collection was made all from one clone, and does not enter into the data for maps 20—23. 1941] FASSETT—MASS COLLECTIONS: RUBUS 331 TABLE VI (Continued) o $|E E 5 . ЗЕ ааа: R P © Р S = Ss = в 8513 БЕ 2141512015191 015138 WISCONSIN 8 19 $ 9 1 22 ak 1 1 10 3 6 Ti 19 18 8 18 14 2 14 13 15* 10 16* 13 17 3 7 18 6 41| 33 1 1 3 1 19 3 3 15 2 1 20 7 1 MINNESOTA 1* 21 2 2 12 s 14 4 1 1 4 5 1 30 1 1 ó 2 30 1 1 6 2 10: 1 7 16 1 1 8 9 8 1 SouTH DAKOTA 1 5 20 2 8 3 8 2 5 5 4 20 8 Л 6 ди 21 6 9 1 ALBERTA vi 5 15 2 1 10 13 3 2 3 17 19 4 10 2 T 11 MONTANA 15 47 WYOMING 1 8 2 18 2 3 4 4 5 | 32 6 5 10 UTAH "T в. | 2 22 | 3 2 30 | 4 2 | 14 4 | 28 3 | * This collection appears to consist of one clone, and does not enter into the data for maps 20-23. { Two of these have villous pedicels, which have been found elsewhere only in var. velutinus. [Vor. 28 332 ANNALS OF THE MISSOURI BOTANICAL GARDEN TABLE VI (Continued) % % зЁ E 5 ы S 3 Е ә © S 3 $ Е 3 1519151915111 8179 sie 3 Е o г S $ E "> 2 " S $|s|' 5 S$ | $ & E S| Б à 5 ао: #5: Ж 2/818 2 S| | os lam а Ы Е $ z COLORADO 1 3 1 2 1 | 20 3 3 10 1 21 4 14 12 5 2 | 19 6 1 1 7 3 OREGON 1 T 2 4 20 8 10 5 10 CALIFORNIA 1 91 2 14 3 20 4 е“ 4 4 9 51 9 61 10 9 4 * This collection also contains one individual of var. velutinus. t This collection also contains 25 individuals of var. velutinus. 1 This collection also contains 12 individuals of var. velutinus. 5. IS R. parviflorus ^ PREGLACIAL RELIC ABOUT THE UPPER GREAT LAKES! The initiation of the present study was an attempt to com- bine two ideas. The first, published by Professor Fernald,?? was expressed as follows: ‘‘The high bluffs of Keweenaw evi- dently were not denuded by Wisconsin ice and, consequently, they served as centers on which many species survived... , some later to spread slightly to lower levels and, in case of readily dispersed species, like Rubus parviflorus . . . , to ex- tend over much of the Upper Lakes region... ? The sec- ond, elaborated by Dr. Sewall Wright and others, has been expressed by Dobzhansky* in the words: “А finite popula- tion left to its own devices must, therefore, suffer a progres- sive decay of its hereditary variability and sooner or later must reach a complete genetic uniformity." Again,?! “Те 9 i0: pi216 3 Genetics id the origin of speeies, p. 130. 1937. м Thid., 132. 1941] FASSETT—MASS COLLECTIONS: RUBUS 333 TABLE VII OCCURRENCE OF FORMS OF R. PARVIFLORUS VAR. VELUTINUS с 3 З " 5 Е EX Е % Е + з P * г 2 S. 3 E 5 f ©, = > > & + Bad * ag > > > > 3 > > в Q Ex 3 5 б р | 8 n & & > & = 2 > CALIFORNIA 4* 1 51 6 12 61 6 5 3 7 4 5 3 14 8 25 34 9 5 3 10 42 11$ 32 12 36 16 * This collection also ples ae 17 individuals of typical В. parviflorus. f This collection also ins 2 individuals of typical R. parviflorus. i This collection also sont ns 16 individuals of typical R. parviflorus Some pedicels bear only short-stalked glands; others in the same inflorescence bear long-stalked glands. § Inflorescences 1-few-flowered with long pedicels, resembling but more extreme than those illustrated for var. parviflorus (Fernald, l.c., plate 365, fig. 1). smaller the population size, the more rapid is the scattering of the variability and the eventual attainment of genetie uni- formity."' If, then, Rubus parviflorus was once confined for a long time on one or few nunataks about Lake Superior it should show less variation in that region than in some other parts of its range. Actually this proved to be only partially the ease. In so far as variation is expressed in the number of forms in each region, as listed in tables ут and уп, it is just as variable in Wisconsin as in California, and in Michigan as in Colorado. But while the number of forms is about the same in each re- gion, their proportions differ from region to region. A study of the proportions of individuals displaying each charaeter, in each region, shows that the loss of variability by the suppres- sion of minority characters has gone a little farther in the Great Lakes region than elsewhere. This is expressed mathe- matieally in table 1x, page 344. [ Vor. 28 334 ANNALS OF THE MISSOURI BOTANICAL GARDEN While the range and amount of variation are about the same in all regions, the number of forms varies greatly from colony to colony within each region. This may be seen in table vr under Ontario, where collections 2-8 are all from the vicinity of Meldrum Bay, and the number of forms in each collection varies from one to four. The number of forms in any colony is not entirely a factor of the age of the colony, in the sense discussed by Dobzhansky, but depends, as in R. odoratus (see page 307), on how many seeds started the colony and whether the eolony has inereased in size by vegetative spread or by seed reproduction. While R. odoratus grows most commonly on recently eut soil, R. parviflorus does not have this soil preference. In re- gions where the climate is favorable (this factor will be dis- cussed later) it forms a solid growth along the roadside (beau- tifully pictured by Fernald, l. c., pl. 363) and also occurs as scattered stems, connected by rootstocks, in thick woods. In plaees where a road euts through woods, the writer has tried unsuccessfully to determine whether the plant had established itself along the roadway and spread into the woods, or was originally seattered in the woods and had expanded along the roadway (pl. 9, fig. 1). In either case, and whatever the pe- riod of occupancy of the Upper Great Lakes area, the entire R. parviflorus population of that area appears to be function- ing as one breeding unit rather than as many isolated units. In map 22 the collections about the Great Lakes have been grouped into regions and a pie-diagram made for each region, just as was done for R. odoratus, except that now 6 characters are used instead of 4 as in the eastern species; the diagrams for the six regions in map 22 show close similarity, as do those in the southern part of the glaciated territory in map 3, rather than the dissimilarity shown in the unglaciated parts of map 3. In map 20 the stations listed in table v, shown by solid dots, have been grouped into seven regions; each region has been demarked by a line and given a name. From the data in tables v1 and уп a pie-diagram has been made for each region. These pie-diagrams show the plants of the Coast Ranges of Cali- fornia to be very distantly related to those of all the other 1941] FASSETT—MASS COLLECTIONS: RUBUS 335 regions: this is consistent with the taxonomic treatment of the plants of the Coast Ranges as a separate variety. Also evident is the rather close resemblance among the four central re- gions of Alberta, Wyoming, Black Hills and Colorado. Less to be expected is the resemblance of the Great Lakes region to the Sierra region. The significance of this will be pointed out below. From the figures in the pie-diagrams the averages of differ- ences of percentages between regions have been computed in the way described on page 311. Table vim shows the averages of TABLE VIII AVERAGES OF DIFFERENCES OF PERCENTAGES BETWEEN REGIONS $ 2 E “4 тЫ м а 3 Щщ A 8 = & g М Е E Е Е - Ф E Ps = = © E: 5 ea Е < © 7 о Great Lakes — 33 19 25 30 1 Black Hills 33 - 16 8 3 25 68 Wyoming 19 15 - 8 12 14 53 Alberta 25 8 8 - 5 17 61 Colorado 30 3 12 5 - 28 66 Sierra 8 25 14 17 23 - 43 Coast Ranges 41 68 53 61 66 43 - percentages of differences between all regions, and many of these are also shown on maps 20 and 21 by large figures super- posed on heavy lines connecting regions. Lower numbers indi- cate closer relationship, and these figures bring out more posi- tively the observations of the preceding paragraph. On map 21 are shown only the average percentage differ- ences of less than 10. Again the close relationships are empha- sized, one of the Rockies and the Black Hills, and the other of the Great Lakes and the Sierra region. [Vor. 28 336 ANNALS OF THE MISSOURI BOTANICAL GARDEN It is this latter fact which appears to the present writer as of great importance in the consideration of the claim of Rubus parviflorus to the title of preglacial relic in the Middle West. If the great gap in its range (see map 24) has been due to a former continuous range having been bisected by ice-sheets between Lake Superior and the Black Hills, these two regions could be expected to show a close relationship. On the con- trary, the relationship between these two regions appears on map 20 as remarkably remote. The Great Lakes region shows much closer connection with the region along the Pacific coast (Oregon) and inland California; this connection, since it was not across Wyoming, Colorado, or the Black Hills, was pre- sumably across a more northern region. Before discussing this possibility it is necessary to determine what ecological factors influence the range of this plant. Map 25 shows the range of Astragalus caryocarpus, a char- acteristic plant of prairies and plains. On the same map are shown the maximum extent of Pleistocene glaciation, and the Fall Line, but there is no obvious relationship between either of these and the range of the plant. Comparison with map 26, however, shows the correlation of this range with provinces of low humidity. It may now be observed that the range of Astragalus caryocarpus almost exactly rrrs INTO THE GAP in the range of Rubus parviflorus (map 24). If both are controlled by one factor, the reaction of one plant is positive, that of the other negative. The range of the Astragalus is correlated with aridity; is that of the Rubus correlated with humidity? Definitely, it is. Figures for a suitable portrayal of regions on a basis of humidity are not available, and if they were they would not include the microclimates so important in plant distribution. But some facts are clear without figures, and the following personal observations have convinced the writer of the dependence of Rubus parviflorus on a relatively high hu- midity and low summer temperature. On the north shore of Lake Superior, in Minnesota, the Thimbleberry is an aggressive plant, occurring not only in woods and openings along the shore and in canyons entering 1941] FASSETT—MASS COLLECTIONS: RUBUS 337 the lake, but as a common weed in many places. Particularly vivid was the experience of Dr. J. T. Curtis and the writer at Lutsen, where we were able to follow the elusive overgrown old lumber roads, only by the still persistent line of Thimble- berry (pl. 9, fig. 2). It has similar weedy habits near the shore of Bayfield County, Wisconsin, on the Apostle Islands, and in the Porcupine Mountains and on Keweenaw Point in the Upper Peninsula of Michigan. On Keweenaw Point it grows in the towns, forming a hedge between houses (pl. 10, fig. 3). : Inland the aggressiveness decreases. While it is apparently adventive along the Gunflint Trail going north from Grand Marais, Minnesota, it seems to be almost absent north and west of Duluth. This seems to be due partly to the slightly warmer area north of Duluth (map 27) and even more to the dry west winds of this region (map 28). In northern Wisconsin the southern limit of Thimbleberry is near the 66° July isotherm (map 27). In the canyon of Manitou Falls about 15 miles south of Superior, Wisconsin, it barely exists; it was possible to find only eight scattered and dwarfed flowering individuals. Likewise in Vilas County, Wisconsin, it is uncommon and usually not aggressive, al- though Dr. Curtis reports a thrifty colony at Jute Lake. In the more southern stations in Wisconsin it is confined to the heavier soils, while nearer Lake Superior the soil preference is scarcely marked. In Door County, Wisconsin, the Thimbleberry is rare on the Green Bay side of the peninsula, but on the Lake Michigan side it makes an almost solid growth, as is shown on map 27. Its aggressiveness well south of the 66° July isotherm is due to local conditions. Mr. Eric R. Miller, Meteorologist in Charge of the United States Weather Bureau at Madison, to whom I am also indebted for assistance in procuring the data from which maps 27 and 28 were drawn, tells me that the west side of Door County peninsula is warmed by the southwest winds which follow Green Bay, whereas the east side of the peninsula is cooled by proximity to Lake Michigan; this difference in temperature is not shown by readings from one station at the [Vor. 28 338 ANNALS OF THE MISSOURI BOTANICAL GARDEN tip of the peninsula and another near its base, but is noticeable when one drives across the peninsula. In the local cool zone along the Lake Michigan shore the saturation deficit would be lower even if the relative humidity were the same on the two sides, which it probably is not, due to frequent fog on the lake and loeal breezes from over the water. Dr. R. F. Griggs has recently called attention to the concen- tration of rare plants, many of them considered by some bot- anists to be preglacial relies, on islands and peninsulas,?? and in telling of the weedy tendencies of some of these plants he specifically mentions Rubus parviflorus in Keweenaw.?? It is weedy in Keweenaw Point because on the peninsula it is fav- ored by low temperature (map 27) and by the prevailing west winds (map 28) blowing across the lake and carrying moisture. It is also weedy along the Minnesota shore of Lake Superior where the atmosphere is cool and moist, and this weedy tend- ency is carried inland in the Arrowhead of Minnesota to a limited degree. On the Apostle Islands at the tip of Bayfield County, Wisconsin, and to a slightly less extent on the ad- jacent mainland, the Thimbleberry is rather aggressive in fol- lowing roadsides and woodland trails. This is true also in the Poreupine Mountains and about Ontonagon, Michigan. In Door County, Wisconsin, as already mentioned, conditions on the two sides of the peninsula make the plant rare on one side and very abundant on the other. Elsewhere in its range in the Lake states the Thimbleberry tends to grow on roadsides and along margins of woods, but it is scarcely aggressive or abundant ; its normal weedy propensi- ties are curbed by the absence of completely favorable atmos- pherie conditions. I do not understand the distribution north of Lake Huron. On the peninsula of Bruce County, Ontario, the Thimbleberry is very rare, and I have seen it only in woods near Cameron Lake. It is abundant about Meldrum Bay at the west end of Manitoulin Island and occurs in open ground on St. Joseph Island southeast of Sault Ste. Marie. I have traversed much ? Bull. Torr. Bot. Club 67: 589. 1940. в Ibid., p. 583. 1941] FASSETT—MASS COLLECTIONS: RUBUS 339 of the shore line of Bruce Peninsula, of Manitoulin Island, of St. Joseph Island, and of Drummond Island, and while of eourse a statement that it is not present except as shown on map 27 would be too sweeping, it is certainly absent in many localities apparently as suitable as those in which it is abun- dant farther west, and seemingly just as favorable as the few spots where it is found. Perhaps it is just in the process of spreading into this region and grows where seeds have hap- pened to land. There appear to be no collections from the north side of the Upper Peninsula of Michigan between Sault Ste. Marie and Munising. I have not collected along this shore except at Brim- ley, 15 miles west of Sault Ste. Marie, where the red clay banks would seem favorable for Thimbleberry, but none was found. In the summer of 1939 I went as far west as Utah and Colo- rado for the specific purpose of collecting quantities of this plant. The eastern botanist traveling for the first time from southern Wisconsin across Minnesota and South Dakota to the Black Hills will soon abandon the idea that glaciation accounts for the absence in these regions of various species common in the eastern forests. Even if he has been familiar with rain- fall maps, ete., he will not fail to be impressed by driving hours on end through progressively treeless plains, and then for mile after mile in a monotony of cactus and sagebrush. But on reaching the Black Hills, he will find again a region of for- ests, streams and swamps, of brook trout and beaver, and of sueh familiar eastern mesophytes as ostrich fern, bloodroot, eolumbine, dwarf Solomon-seal, ete.?* In the cool and moist eanyons of the northern Black Hills the Thimbleberry grows; Dr. Hayward's paper brings out its ecological relationships in his illustrations. His Fig. 7 shows Spearfish Canyon, which supports a heavy stand of white spruce with quantities of Thimbleberry; this is characteristic of a small region sur- rounded by country pictured in his Figs. 8-11, where the ab- sence of Thimbleberry can by no means be attributed to glacia- tion. м Of, Hayward, Н. Е. Bot. Gaz. 85: 353-412. 1928. [Vor. 28 340 ANNALS OF THE MISSOURI BOTANICAL GARDEN То the easterner visiting the western states for the first time the eonditions about Salt Lake City are most striking. Ris- ing from the arid plains and bordering the Wasateh Range are the terraees of Lake Bonneville, impressive evidence of a once great body of water. On the parched sides of the Wasatch Range are sagebrush, shadscale, rabbitbrush and greasewood (pl. 11, fig. 5), but in the canyons are heavy forests and rushing streams of icy water from the snowfields above—one might almost be persuaded that he is back in Glen Ellis in the White Mountains. It is in these cool moist canyons that the Thimble- berry grows (pls. 11-12, figs. 6, 7). The occurrence of Thimble- berry in mountainous areas and its absence in intervening lowlands is shown in map 36. These observations convince the writer that the gap in the range of Rubus parviflorus is due to aridity and not to glacia- {101.35 It is easy to concur with Professor Fernald’s state- ment,** ‘‘it seems quite illogical to argue that such species . . . have been arriving in post- Wisconsin times from different re- mote centers outside the area of general Pleistocene glaciation . . „7; the eross-eontinental migration must have been ancient, even if postglacial. In the same paragraph he expresses doubt that they could have made this migration **without leaving in their long hypothetical cross-country journeys a somewhat continuous train of intermediate stations." Had he said ‘‘more or less continuous"! it would have been easier to agree, for the intermediate stations are ‘‘more’’ for some species and ''less"' for others. They are less continuous for Poly- stichum Lonchitis (207) 7 Ceanothus sanguineus (210), Vac- ciniwm membranaceum (211), Adenocaulon bicolor (213) and many others, and for some other species they are more. Some of these other species will be discussed in the next paragraph. “ This statement is deliberately worded to refer to one species and one alone. Professor Fernald points out (1. c., pp. 212-213, 217-218) that both mesophytes and xerophytes show this gap in their ranges. The present investigation concerns but one species and the conclusion is that that species is not a preglacial relic. % L, ©., р. 208. " Figures refer to the page numbers where these are mapped in Professor Fer- nald’s paper. 1941] FASSETT—-MASS COLLECTIONS: RUBUS 341 Hultén decries** the practice of basing conclusions exclusively on extreme types of range, as follows: ‘‘When tackling the problems offered by the geographical areas of plants, many authors have chosen to discuss peculiar or singular types hav- ing a distribution out of the common, in the hope that they will suddenly give a clue to the solution of the problem. . . . It is surely more rational to start the investigation with the simplest types, those that show the least possible peculiarities. When they have been interpreted, the complicated and often strongly interrupted areas of the singular or peculiar types are likely to be better understood.’’ Populus balsamifera (map 29) has a broad range across the northern part of the continent, chiefly in glaciated regions; this range seems obviously to have been attained in postglacial time. The same is true of Picea glauca (map 30), but its south- ward extension down the Rockies and into South Dakota has been lost, leaving outliers in the Black Hills and in Montana. The range of Arabis divaricarpa (map 31) is becoming defi- nitely pinched in the region northwest of Minnesota. This trend has resulted, in А. Drummondi (map 32), in a breach between the eastern and the western parts of the range, with a few scattered intermediate stations—this is one of the more species of the preceding paragraph. An even wider gap ap- pears in Botrychium Lunaria (map 33), but comparison with map 29 shows how a distribution like that of the Botrychium might have been derived from one like that of the Populus. If the derivation was of this nature, and the distribution of the Populus is clearly postglacial, then the range of the Botry- chium is postglacial also. It is a short step from a range like that of Botrychium Lunaria to one like Rubus parviflorus (map 34), whose occurrence east of the Rockies has been re- duced to the Black Hills and the vicinity of Lakes Superior, Michigan and Huron. Perhaps it is dangerous to imply that the various gaps in the ranges shown in maps 29-34 are due to the varying toler- 38 Outline of the history of Arctic and boreal biota during the Quarternary period, pp. 9-10. 1937. [Vor. 28 342 ANNALS OF THE MISSOURI BOTANICAL GARDEN ance of the low humidity in the region between the Great Lakes and the moist ravines of the western mountains, since a de- creasing tolerance of low humidity shown by these species is not easily demonstrated experimentally. However, it is not necessary to explain to a botanist familiar with the northern states and southern Canada that balsam poplar and white spruce are found in a greater variety of habitats than is Thim- bleberry. While the former is by no means rigid in its require- ments, it does need a comparatively cool, moist climate. In the Middle West, wherever Thimbleberry grows, white spruce ordinarily grows also, but where white spruce grows Thimble- berry does not necessarily accompany it. The limiting factors of the Thimbleberry are the same as those of the white spruce, but to a greater degree. When, then, did the Thimbleberry spread from the western regions to the Great Lakes? At some time when the climate was sufficiently cool and moist for white spruce to spread from coast to coast, and south to the Black Hills. Perhaps this was while glacial influence was still felt®®; perhaps it was during some later humid period. It may be significant that in spite of the large number of eastern woodland gpecies isolated in the Black Hills there has not been described a single endemic variety of an eastern species from that region. These species must, then, have entered the Black Hills fairly recently. And conditions which would permit plants like bloodroot to spread westward to the Black Hills might also permit a migration of Thimbleberry eastward along a more northern route. Rubus parviflorus did not populate the Black Hills from the same stock as that which reached the shores of the Great Lakes. ? « = MAP 2 Slanted we locations of mass collections. Ereet figures to the right of, or above or below, the slanted figures: "m cent of individuals | in each НЕ свои with glands on the pedicels 0.5—1.0 m figures: per cent of uro in each region (enclosed in broken 0 mm. long. Tas lines) ‘vith glands on the pedice rratum: Vt. 4—change 50 106 1941] FASSETT—MASS COLLECTIONS: RUBUS 349 “NORTHERN VERMONT (32) ~ p- ------—1--. ——— 4% r I і МАР 3 Stations where mass collections have been made (dots) grouped into regions (vertieally lined). The figure in parentheses below the name of gi 8 его dividuals collected each region is the num f in in that region. In the pie-diagram for each region, each sector represents a character, and figure in the sector shows the percentage of total individuals having this character, as follow yi 8 upper sector, per cent of ipid iduals with — phun s above and glabrous or "gla abrate ake neath; left sector gone М f individuals with glan the pedi бів 0.5—1.0 mm. long; lower о per eent of individuals with gl: рзд e pe veda лот, от subsessile ; right sector, p t of individuals with glands 1.0-2.5 m ong on 3s lower sides of veins of leaves. To с i space, Indiana is inserted in the loe eH cor [Vor. 28 350 ANNALS OF THE MISSOURI BOTANICAL GARDEN MAP 8 MAP 9 P 4. Averages of differences и adjacent гене derived wes map 3; itus between neighboring regions in the more a: n post- -Wiseonsin areas (southern New Yo rk State) are aha, those between pre- Wisconsin aaa (central Pennsylvania and southward) are mostly fairly "close, and those between pre- Map 6. Differences between the Potomac region and other regions, derived from map 3. Mar 7. Differenees between the Eastern region and other regions, derived from map 3. Map 8. айныр between the Tusearora region and other regions, derived from map 3 Map 9. Differences between the Sehuylkill region and other regions, derived from map 3. 1941] FASSETT—MASS COLLECTIONS: RUBUS 391 * lade f и "3 "E Lm y UM га х МАР 10 Differences between the Indiana region and other regions. пате the cma E = x glaciation; wher part delimits Wiscon n and the n The base map is Hall "sc ое ‘Map, 801M The ms с: crosses e this line splits, the southern thes part the earlier Нот. [Vor. 28 352 ANNALS OF THE MISSOURI BOTANICAL GARDEN . "mo "^ PEDE я . + a” aw n ^ € 4 е OESTE. à ui xy ts м о > 1.29 / Фе i 2, 1 "P 455 NO RY - Б ` И к\ч, D нъ v G SUN IEL ЗЕ ра Та + S ` QUU u ` AN < 2 2% МАР 11 Western part of the island of Moorea. The figures indicate the proportion of dextral individuals in each isolated valley, for the snail Partula suturalis. Data from Crampton, l.c., tables 10 & 11. 1941] FASSETT—MASS COLLECTIONS: RUBUS 353 MAP 12 Cross- bitis areas: regions where mass collections of Е. odoratus have bee e. Dots: range (excepting Nova Scotia), based on collections in the Gray ра barium, New York Botanical а New nd Botanical Club, New York State niv ur of West Virginia, Toronto, Pennsylvania, Kentueky, Cin- cinnati, Tennessee, a Wise Pus] ther of Dr. E. y B and Dr Harper, Sehaffner ‘Ohi o Vascular Plants,’ and Deam’s ‘Flora of Indiana.’ The base map is Hall's ‘Outline Map, 801M. 354 ANNALS OF THE MISSOURI BOTANICAL GARDEN leetio MAP 13 MAP 14 parviflorus var. hypomalacus. Dots: range aecording to Fernald r4 colleetions. MAP 15 [Vor. 28 Е. err var. genuinus. Dots: range according to Fernald. X’s: other col- . X's: other R. parviflorus var. heteradenius. Dots: range according to Fernald. X’s: other collections. lections R. collections. Dots: R. Я eye var. ording to Fernald. MAP 16 R. pee var. bifarius. Dots: range according to Fernald. X’s MAP 17 MAP 18 : other eol- parviflorus var. grandiflorus. Dots: range aecording to Fernald. X's: other m, and triangles, var. parvifolius, ranges bec ni dn Я pice like them but with leaves velvety beneath. X’s: ща реге но of these two varieties, Crosses: thas E 355 RUBUS FASSETT—-MASS COLLECTIONS 1941] d № X - ar ^ 3а í ж, ? ЮЕ +? “= Ak. Н 1 e Н i 1 се 1 M^ : i > \ ч i" Ni 1 J ay H M ^t $ = | ПО 4 р. i № iN | Р м T > , . > . | k EB | Es a^ „— be — БЕ T E i — 1 H 1 м : H C | F xpi r7 cdi г $ i ju Py a | | . J EN a 3 * JN Г X г. Н шежа "mm ae v uc d | | A раз 4] T rM i } a) SY | LA. Id dM — i mi і Xam з 7 c F4 т^ за ГАРД S ie | H 4 t : zd i : = г oo = JC a i rt er] e La | eer] xod ui И | = PT НЕ t < P i r^i Ps e лет ВЕ ir Fi Н ; | ? + і ! | f НЕ i fi rS MEM + fae be Se aS M ME ES на ; Ij | а Ау CBE | | C MEAE ГА ; ! : tT] 1 i 1 j £ } i | H d H i { i 1 Д ya А е: ; Do Py 35-4 Л]! { ; ! а : ps eve Ц 1 ; ге 1 3-3. 5. maf Sf a 74! ! } i f Т ] 1 4 bans — ай Ыы: 4 eu rig ? aN i efi E H ] RE ы. / "— E ку E AE РД Minas we E 4 l - | ч — = ОДИ кан i E ы n 4 j : Los 7 - bere | Б г Ги i } А i j so sd Dos С als : PT AS + : LI T—— lal Ч MAP t alg See i T i 1 D 1 t chen a - „А... 1 | А A \ E H Н A үү | y | ANAL i / Н / ANM = x : Y P 5 + Н + m z - ? } : А > ( aa i = | F qu RTA QAR pte $NL— м i 4 | i ч f Н T 4 еее т С | ee: : E ERG / i 4 t A Up e ; dE | ; P ща Fs a H Н ST T Н xi С Ху Wa DAE i IN nS Fi i H Н ® ч А НТА | 2 ОА: EV FA | о 21: SAA 3 + i T р 7 — D < = TH S. i e xc uad | a А. @& | Н H ! i pg- e 4) T i I 2 ki i : + d CT Td e ы eT аа е Ж-Н ДА | НА. Е Li [o NE] "1; H Ыы i + E 1 Mem Ж КЕ чн s Е Да жд TE De 1 j { j i н Н Н i H Н i 1 Д Н Н H ; \ i H H H f Н Н i : j i Ц i i TE. H H iN] i ; ; i \ Ta. А у T 1-y-3- jJ 1] Ss зын ee na ща зА à | УЫТ j ef} tN „Уе ENS AN - T клы Е к 2. Н / ~ 4 : a 7 = } -—— 7 A 6 jose › = Је Е --- А = H и -- < i =i, / ме < e: 4 Е a , 7 Pa а м и, - d # Ц : j - у 1 r Н 7 j i ; > | : ; n LI I Я $ ; e * Ang - E AA КТУ I 2 = 7 y Pe 7 ge \ Є D [Vor. 28 356 ANNALS OF THE MISSOURI BOTANICAL GARDEN MAP 19 Slanted figures show locations of mass collections of В. parviflorus. A new series starts in each state or provinee, and figures missing indieat Lm close to the locality bearing a lower number. The base map is Hall's Outline Map, 801M MAP 20 tations where mass collections of R. parviflorus have been made Дио grouped into regions тее osed by lines). The figure He the name of eaeh region in indieates the number of individuals olleeted from that region. A pie-diagram for each ей indicate in each seetor the percentage of occurrence of a character as follows (refer to key): r A, per cent of individuals with leaves glabrous or glabrate — eetor B, per и po glands 0.5—1.0 mm, long on Yon pedicels; sector C, per cent with glands sessile or subsessile on the pedicels; sector D, per cent with villous calyx; sec e E, = — th villous pedie cels; — r F, per cent with stem, stipules and pd villous. In the с of each diagram (position X) is an initial referring to the name of the region. The a axe diferentes Drs o e TE table упт) are shown by large figures on the heavy lines praan efi region MAP 21 Differences less than 10. The base map for maps 20 and 21 is Hall’s Outline Map, 801M. 1941] FASSETT—MASS COLLECTIONS: RUBUS 357 в г GREAT LAKES Н 1223 Ы) [Vor. 28 358 ANNALS OF THE MISSOURI BOTANICAL GARDEN MAP 22 Stations about the Great Lakes where mass eolleetions have been made (dots) pos into in, жыр (enclosed by ellipses). For explanation of pie-diagrams see caption under m Th e below each pie-diagram indicates the number of individuals collected from that Ж —- = вт — o Гал МАР 23 Differences between regions about the Great Lakes, derived from map 22. 1941] FASSETT—MASS COLLECTIONS: RUBUS 359 MAXIMUM EXTENT OF TOTAL No” PLEISTOCENE GLACIATIO М MAXIMUM EXTENT ome WISCONSIN STAGE OF GLACIATION AI face UE MAP 24 Range of Е. parviflorus, and var. velutinus (in ellipse), from Fernald, 1.с., the Herbaria of the New York Botanical Garden, Pomona College, and the Universities of California and of Wisconsin, and letters from Drs. F. K. Butters and Hugh Raup. The base map is Hall’s Outline Map, 205C, for maps 24, 25 & 26. [Vor. 28 360 ANNALS OF THE MISSOURI BOTANICAL GARDEN MAP 25 Range of Astragalus caryocarpus, from Fassett, *Leguminous Plants of Wis- consin. ? MAP 26 The heavy рома етти some of the elimatie provinees described by miri eee Geogr. xd easy Pl. ПІ; fer бает deseribe - provinces as follows: C indi- cates sub mid, ce semiari rid; B' indie mesothermal and “yy micro- iari thermal; r REL precipitation ny Ree at all А and d indieates preeipita- tion deficient at all season 1941] FASSETT— MASS COLLECTIONS: RUBUS 361 / i B J ` 4 ГА j i y M ` : P = f H E л м.-77 | | я = X / | A A Н \ j | f : E> | d i SUL я | i } р t 7 { ua \ i ка і mia : \ EU t ва b! jJ. { x 2, pes И c Г x re^ X д == EN н ES v = 7 Не > 2 М | hes omen * Ж pU S. Ree № SS ae ; E 1 ЗА ВС са LE. rU. ee а ү = MAN IMUM EXTENT OF TOTAL T MES BERT ла WM! peersrocene GuACIATIOU у Ux fe ewe & MAXIMUM EXTENT © 1NNC Y Dai, у E : ма. WISCONSIN STAGE OF GLACIATION 9 : NI tatt LINE ; = IN MAP 25 MAXIMUM EXTENT OF TOTAL A 27 PLEISTOCENE GLACIATION \~\ SNI лла LIRE VA MAP 26 [Vor. 28 362 ANNALS OF THE MISSOURI BOTANICAL GARDEN MAP 27 Dots: range of В. parviflorus in the прие Great Lakes region, Dotted lines: isotherms for average temperatures for July, compiled from Sections 44, 45, 46, 47, 48, 49, 63, 64 & 65, ** Climatic Summaries of the United States," published by the U.S. Wea ther Bureau. MAP 28 Arrows ноар prevailing wind direction for July, compiled from EMEN Sum- maries of the United States.’’ Letters indicate d" referred to in the text, as follows: A Aposte Islands; B, езй BC (in € — , Bruce ни: вс "(їп Wisconsin), Bayfield County; D, Dulu жЕ DC, Door Coun При hirer Island ; GB, Green Bay ; and Marais; P, Kew weenaw Poit ; L, Lutsen; M, Munis- sing; MB, Mirum. Bayi MF, Mento Falls; MI, Manis "Island; O, On- tonagan orcupine Mountains; 8, Superior; SJ, St. Joseph Island; SSM, PM, Por Sault Ste. Marie; VC, Vilas County 1941] FASSETT—MASS COLLECTIONS: RUBUS 363 MAP 27 ONTARIO LAKE SUPERIOR ws ~ xy of 4 f »X caf ae ane) „М ЗУБ МВ 080—4 UPPER PENINSULA C s ре т 5 ‘ р 2n / BC МБ 5 ^ м ме” Е A [- o" Kor M > ^ элә T a2 з % %, ( ^ я » DC © А» 4 * © ^ > к ADLI M fe > AY | Ка к ы От т IA я яя м я ^ i O | я ы ^" is м * E А = КЫК. г LA > iu к x ^ Ж - 4 ^ я `A \ NG WISCONSIN» , | аа с А я. лай + Ка я L ча к я „47 K \> + x -J Ха от а я Карат P ё ^ ш ул > я z я z \ A X ^ ME. id ал 264 [Vor. 28 ANNALS OF THE MISSOURI BOTANICAL GARDEN MAP 29 Range of Populus balsamifera, generalized from Munns, U. S. Dept. Agr. Mise. Publ. 287: 76. 1938. The base map is Hall's Outline Map, 2050, for maps 29-34. MAP 30 Range of Picea glauca, generalized from Munns, l.c., 35. MAP 31 Range of Arabis divaricarpa, generalized from Hopkins, Rhodora 39: 131. 1937. MAP 32 Range of Arabis Drummondi, generalized from Hopkins, l.a., 138 MAP 33 North American --— of Botrychium Lunaria, generalized from Clausen, Mem. Torr. Bot. Club 19: 63. 1938. MAP 34 Range of Rubus parviflorus, generalized from map 24 of this paper. 1941] FASSETT—MASS COLLECTIONS: RUBUS 365 (Vor. 28 ANNALS OF THE MISSOURI BOTANICAL GARDEN 366 E geen рт | 1 Sii ^ ui Np 4 j 755 9^9 coe d с Е 4: - е * ка mE : р | үч” WB. ENAM e "e omn 2 LEM io vods T eT Mt d Spe 4 % Nr? е ж м Ba 271 z^. 141 е OS > ИЕ? r У SKE R d 2 25 „а ме aoe d Ph es Epe- Pay "i£: > e "m 56 SB. m 335 5 "А „не dus FE MSU cose E > Ес a * o ty 5 (сие Der " въз 12 y" i odo Sh I cs he ae та! RJ 30 S^ у 2 $ MS mc ode e e RA Jo зо * д WO 5 i RH я дай b $34 ВЕ Зов + uum 3 S538 ave ве“ -ва о5 а m + ——ÉH ЗЕЕ E Е] ед 33 Peg ES * Suto & o go ET + Я Я ofa 5890 ч « aygo BALL $ оисори?" 2 Bad Relation of the range arl3an * atat of Ginn & Co. FASSETT—MASS COLLECTIONS: RUBUS г » аб ES MS д NS pu да Cus к; us, пето Ж ! SE pe e Y bon | EE fO Cw Qua: ix ки — t. PANE trans. Й 9; Е ^ NA AA an 7? e LV Ap Y MAP 36 of Е. parviflorus (dots) to topography in Wyoming, Colorado, and parts e map from Erwin Raisz in Atwood's ‘ Physiographic Provinces,’ courtesy [Vor. 28, 1941] 368 ANNALS OF THE MISSOURI BOTANICAL GARDEN EXPLANATION OF PLATE PLATE 9 Fig. 1. Thimbleberry occurring in solid growth along a roadway, and as scattered plants in adjacent woods, near Toivola, Michigan. This patch contains at least three of the so-called varieties of this species. Fig. 2. Thimbleberry growing in an old lumbering road near Lutsen, Minnesota, At least three of the so-called varieties of R. parviflorus occur in this patch. дет” a озүнө” СТУ ANN. Mo. Вот. GARD., Vor. 28, 1941 * FASSETT—MASS COLLECTIONS: RUBUS EXPLANATION OF PLATE PLATE 10 imbleberry as a fence-row plant in Mohawk, Keweenaw County, imbleberry in the gorge below Fish Creek Falls, Steamboat Springs, s patch contains at least four of the so-called varieties of R. parviflorus. ANN. Mo. Bor. GARD., VOL. 28, 1941 PLATE 10 4 FASSETT—MASS COLLECTIONS: RUBUS 372 [Vor. 28, 1941] ANNALS OF THE MISSOURI BOTANICAL GARDEN EXPLANATION OF PLATE PLATE 11 Fig. 5. Mouth of Little Cottonwood Canyon, Salt Lake City, Utah. Photograph by Prof. Walter Cottam, University of Utah. City Creek Canyon, Salt Lake City, Utah. Photograph by Prof. Walter Fig. 6 Cottam, University of Utah. PLATE 11 TARD., Vor. 28, 1941 ANN. Mo. Bor. ( ICTIONS: RUBUS MASS COLLE FASSETT- EXPLANATION OF PLATE PLATE 12 le Cottonwood Canyon, Salt Lake City, Utah. Photograph by Prof. - , University of Utah. | ANN. Mo. Bor. GARD., Vor. 28, 1941 PLATE 12 ( га 75 ий wx ARTS FASSETT—MASS COLLECTIONS: RUBUS Annals of the Missouri Botanical Garden Vol. 28 NOVEMBER, 1941 No. 4 DICHOPHYLLUM MOOREI AND CERTAIN ASSOCIATED SEEDS HENRY N. ANDREWS Instructor, Henry Shaw School of Botany of Washington University A few years ago Elias (Moore et al., 36) described a flora from the Upper Carboniferous of east-central Kansas which contained certain elements such as Walchia and Taentopteris, plants generally accepted as more typical of higher horizons. In many ways the most interesting and certainly the most novel member of the flora is the plant described as Dichophyllum Moorei (fig. 1). In the hope that a more detailed knowledge of it might shed light on the general problem of leaf morphology in the seed plants we visited the type locality in the summer of 1939 and made a small collection. Excavations carried on during the following summer yielded some excellent speci- mens, and although much remains to be known about this in- teresting fossil sufficient information has been obtained to war- rant a short note on its gross morphology. Occurrence of the Fossils.— The locality from which both Elias’ and our own specimens were obtained lies about six miles northwest of Garnett, Kan- sas (Section 32, T. 19S, R. 19E), which is in the Victory Junc- tion member of the Stanton limestone and of Upper Carbonif- erous age. Its stratigraphy has been carefully worked out (Moore et al., '36), and despite the Permian aspect of the flora there seems to be no doubt that its age as given is correct. Issued November 27, 1941. ANN. Mo. Bor. GARD., Vor. 28, 1941 (375) [Vor. 28 316 ANNALS OF THE MISSOURI BOTANICAL GARDEN Some difficulty attended collecting due to the indisposition of the owner on whose property the original quarry is located. Tt was found, however, that excellent material could be obtained only a few inches below the surface of the adjoining county road less than 100 yards from the quarry. For permission to dig in the road I wish to express my appreciation to Mr. Ray Hardin, Anderson County Road Commissioner. Dichophyllum Moorei Elias.— The appearance of this plant may best be gained from the restoration (fig. 1), as well as figs. 2 and 3 which served as the basis for fig. 1. In the most complete specimens the main axis is observed to break up near the distal end into two or three main branches which in turn divide, ultimately resulting in two or three subdivisions or telomes. Since there is no reason to believe that we are dealing with a fern the term pinnule will not be applied to these final subdivisions. It is significant to note that the plant possesses a well-developed cuticle, a point that supports the supposed xerophytic nature of the flora and sug- gests gymnospermous rather than filicinean affinities. Un- fortunately the cellular details of the cuticle are poorly pre- served, and macerations have proven of little value. Lateral branches may be observed in fig. 3. These start to divide almost immediately in the same dichotomous fashion as the distal branches. It has been suggested by Jongmans (Moore et al., '36, p. 16) that these specimens are referable to Callipteris flabellifera (Weiss) Zeiller, but Elias has pointed out that ‘‘they differ distinctly chiefly by the character of the lateral pinna; in the species from Kansas they are palmate, while in the European form they are pinnate.” I am entirely in agreement with Elias in making a generic distinction be- tween Callipteris and Dichophyllum, but I believe that our specimens of the latter clearly reveal that they are not strictly palmate but rather present a combination of pinnate and di- chotomous branching. In some specimens (fig. 6) the second- ary branching at first glance appears to be palmate but it is evident from figs. 3, 4 and 7 that it is more in the nature of a uni- form dichotomy, with never more than two or three terminal subdivisions arranged in a strietly palmate fashion. 1941] ANDREWS—DICHOPHYLLUM MOOREI 311 A comparison of the specimens illustrated here with Cal- lipteris flabellifera (Gothan, in Potonié, '07) reveals a rather sharp distinction in the mode of branching of the two, the frond of the latter being pinnately divided in contrast to the pinnate- dichotomous branch system of Dichophyllum. The indirect evidence available shows that there is no rea- son to refer this plant to the Filicineae as its supposed inclu- sion in the genus Callipteris might suggest. 'The rather heavy cuticle and associated seeds (p. 379) point toward a gymno- spermous plant. Although a further consideration of the affin- ities of Dichophyllum must be speculative certain comparisons are not without significance. The fossil record has supplied a wealth of evidence which in- dicates that the leaf of modern ferns has been derived from a branch system which became confined to a single plane and progressively webbed. It is, moreover, highly probable that such a phylogenetie trend has resulted in the typical bilobed leaf of the modern Ginkgo. Although it is true that a single Ginkgo tree may harbor much foliar variation, chiefly in the degree of dissection of the lamina, the farther back we go in geologie time the more finely divided the leaves become until they pass over imperceptibly (at least as far as this charaeter is concerned) into typieal Baiera species. In Baiera specta- bilis and B. Lindleyana, for example, there is little left that one can call а lamina. It seems likely that here, as with the ferns, a branch system has given rise to the lamina, and it is possible that we may partially bridge the gap between the ‘‘leaves’’ of the above-mentioned species of Baiera and a branch system proper through a form such as Dichophyllum. It must be em- phasized that the latter is not postulated as a ‘‘missing Пик’? in a direct line of development but rather as a representative stage in the transition of side branches to a leaf-like structure composed of petiole and blade. The morphology of Dichophyllum is particularly interesting in the light of Mrs. Arber’s recent (241) interpretation of leaf and stem in the angiosperms. The mode of branching that is found in Dichophyllum and Psygmophyllum, as well as cer- tain of the better-known Coenopterid ferns, adds weight to her [Vor. 28 318 ANNALS OF THE MISSOURI BOTANICAL GARDEN conception of the shoot as the basic unit of plant structure. In so far as we know these forms, it is not possible to relegate their branch system to the classical categories of stem or leaf, and it is probable that the term shoot will be a generally acceptable one. Mrs. Arber’s concepts are not concerned with phylogeny. She specifically states: ‘‘This view has no phylogenetic impli- cations; it does not commit us to any opinion as to the origin of the leaf as a matter of history, but is concerned with what the leaf actually is, here and now." It seems legitimate, however, to expand this concept to include certain apparent phylogenetic possibilities. A comparison of a long shoot of Ginkgo biloba with the shoot of Dichophyllum as restored in fig. 1 reveals a striking similarity in basic structure. The dichotomous side branches of the latter may well be the forerunners of the deeply dissected Baieras which in all probability are ancestral forms of the modern Ginkgo. The closest affinities of Dichophyllum seem to lie with certain species of the genus Psygmophyllum, especially P. cuneifolium. There is, however, considerable variation in Psygmophyllum, and it is perhaps doubtful whether it constitutes a natural as- semblage of species. Whether this be so or not the species in- cluded in that genus strongly suggest that leaves as they ap- pear in P. Kidstoni may have had their origin from such forms as P. cuneifolium, where we find a branch system not unlike Dichophyllum, through P. Grasserti with its deeply dissected “leaves.” (Fora more detailed description of Psygmophyllum see Seward, '19, p. 79-90). Associated Seeds.— At least four or five species of seeds occur with the other plant remains at Garnett.’ Of these one is of particular in- terest because of its abundance, its frequent association with 1 The large number of clearly defined and apparently new seed species found here is of considerable interest. Some certainly belong to the coniferous remains, while others are probably referable to the pteridosperms or other gymnospermous groups. It is proposed to include a detailed consideration of these seeds at a later date in a general revision of American Carboniferous seed impressions and compressions. 1941] ANDREWS—DICHOPHYLLUM MOOREI 319 Dichophyllum and its various features which set it apart as generically distinct from anything previously described. The seed in question is characterized by two prominent horn- like projections at the micropylar end (figs. 8, 9, 10) and ap- pears to be comparable with that figured by Elias as Samarop- sis n. sp. B. However, according to the accepted concept of Samaropsis (Seward, '17, pp. 348-354), that genus is dis- tinctly winged while our specimens show no semblance of a wing. Because of this disagreement and since they do not conform to any described genus it is proposed to assign to them a new binomial. | Diceratosperma Carpenteriana gen. et sp. nov.— Samaropsis n. sp. B. Elias, in Moore, Elias and Newell, A ‘‘Permian’’ flora from the Pennsylvanian rocks of Kan- sas. Jour. Geol. 44: 12, fig. 7(5). 1936. Seeds, presumably platyspermie, 6.0-8.0 mm. long, 3.54.0 mm. broad. Two prominent horns, approximately one third as long as body of seed, at mieropylar end. Presence of pollen chamber indicated by papilla-like cast between horns. Locality: six miles northwest of Garnett, Kansas: Section 32, T. 19S, R. 19E. Horizon: Victory Junction member of the Stanton Limestone, Missouri Series. Age: Upper Carbon- iferous. The species is named in honor of Mr. A. C. Carpenter of Ottawa, Kansas, whose knowledge of the local geology and willing cooperation greatly facilitated my collecting. Acknowledgment.— I am indebted to Dr. M. K. Elias for placing unpublished photographs of Dichophyllum at my disposal and for helpful suggestions concerning the Garnett flora. Literature cited.— Arber, A. (1941). oe interpretation of leaf and root in the angiosperms. Biol. Rev. 16: 81-105 Moore, R. C., M. K. Elias, and N. D. Newell (1936). A a | flora from the Pennsylvanian rocks of Kansas. Jour. Geol. 44 Potonié, H. (1907). Abbildungen und Beschreibungen qued Pflanzen-Reste. Lief. : 64. Herausg. Konig. Preuss. Geol. Landes. u. qup пана, А. С. (1917). Fossil Plants. ПТ. Cambridge . (1919). Fossil Plants. IV. Cambridge. [Vor. 28, 1941] 382 ANNALS OF THE MISSOURI BOTANICAL GARDEN EXXPLANATION OF PLATE PLATE 14 Dichophyllum Moorei Elias Fig. 2, No. 1432, x .5; fig. 3, No. 1433, x .5; fig. 4, No. 993, x .6; fig. 5, No. 1227, x .6; fig. 6, No. 1428, x 1. Ann. Mo. Вот. Garb., Vor. 28, 1941 PLATE 14 ANDREWS- — DICHOPHYLLUM MOOREI EXPLANATION OF PLATE PLATE 15 Moorei, No, 1425, x 1. Carpenteriana, Nos. 1434, 1435, 1436 теврес- PrarE 15 Bor. Garb., Vor. 28, 1941 ANN. Mo EI DICHOPHYLLUM MOOR ANDREWS — A REVISION OF THE NORTH AMERICAN SPECIES OF THE GENUS ANISACANTHUS! STANLEY HARLAN HAGEN Formerly Assistant in the Henry Shaw School of Botany of Washington University INTRODUCTION Anisacanthus is a small genus of the Acanthaceae, which, be- cause of the relatively few collections and the obscure specific characters, has been poorly understood. The present study was undertaken to correlate the morphologieal characters of the taxonomie entities of the genus as a whole, and to deter- mine the relative value of the different characters in specific and varietal delimitation. Anisacanthus is one of several genera segregated from Justicia, having been deseribed by Nees? in the year 1842. He based it on Justicia quadrifida Vahl? which in turn was founded on J. coccinea Cavanilles,* a plant grown in the gar- dens of Madrid from stock brought from Mexico probably in the latter part of the eighteenth century. Vahl had changed the species name from coccinea to quadrifida because of the earlier J. coccinea Aublet. Under the name Justicia quadrifida Vahl, it was not infrequently cultivated in European gardens; and fortunately several authentic specimens from horticulture have been preserved in herbaria. Nees, however, soon changed the name of the type species to Anisacanthus virgularis (Salis- bury) Nees,* and аз such it was known until Standley revived the specific name quadrifidus.® 1 An investigation carried out in the Graduate Laboratory of the Henry Shaw School of Botany of Washington University and submitted as a thesis in partial ey of the requirements for the degree of master of science in the Henry w School of Botany of Washington University. poda ab Esenbeck, C. G. Linnaea 16: 307. 1842. Vahl, M. Enum. Pl. 1: 124. 1805. * Cavanilles, A. J. Icon. et Deser. Pl. 2: 77, pl. 199. 1793. * Nees in ОС. Prodr. 11: 445. 1847 * Standley, P. C., Contr. U. В. Nat. Herb. 23: 1343. 1926. Issued November 27, 1941. (385) [Vor. 28 386 ANNALS OF THE MISSOURI BOTANICAL GARDEN Three of the species of Anisacanthus have been segregated from other genera. Nees transferred A. pumilus from Justicia, and Gray removed A. Wrightii and A. Thurberi from Drejera. A. insignis, A. abditus, A. Gonzalezii, and A. tulensis have been deseribed from collections of later explorations. Grateful acknowledgment is made to Dr. George T. Moore, Director, for placing at the author's disposal the splendid fa- cilities for study offered by the Missouri Botanical Garden, and Dr. J. M. Greenman, Curator of the Herbarium, for his guidance during the study and for securing loans of materials needed. The author also appreciates the helpful cooperation of the librarians, and the constructive criticism offered by other members of the staff. GENERAL MORPHOLOGY Stems.—The genus Anisacanthus consists of suffruticose plants, varying from erect to prostrate in habit, and mostly less than a meter in height or length. The wood is tan or brown in color, evenly grained, soft and brittle to relatively hard. The bark of the upper branches is thin, fibrous, rather brittle, ex- foliating in longitudinal strips; on the lower branches it is gray, and relatively fine-grained. Stem pubescence is mostly confined to two opposite longitudinal strips continuous with the base of the petioles. The stem is terete, usually minutely fluted or striate. Leaves.—The leaves vary considerably on a single plant, but within fairly consistent ranges throughout the group. The mature leaves are usually lanceolate-acuminate with acute or obtuse bases, but sometimes with broader and cordate bases. In one species, A. insignis, the linear and sessile leaf-characters serve as a basis for segregating the variety linearis. Dietrich, in describing Justicia pumila, mentioned the fact that the base of the leaf-blade was somewhat oblique. This character is not confined to that one species, however, but is infrequently found in others. Cystoliths are usually noticeable on the upper surface, especially if the leaf is glabrous. Punctate 1941] HAGEN—THE GENUS ANISACANTHUS 381 glands on the lower surface are also characteristic of the group. Pubescence, when present, is mostly of short hairs thinly scattered over the surface or limited to few hairs along the veins, midrib, and petiole. Inflorescence.—The inflorescence is of the indeterminate class. It may be a spike, raceme, or a panicle; in some cases the spike may be secund. Ii panieulate, the axis is usually shortened so that the cluster of flowers appears more like a head than a panicle. Considerable variation may be observed in the nature of the inflorescence, such as a secund spike with a single flower at a node, two opposite axillary flowers at a node, flowers 2 or 3 in each axil, an open or a shortened pan- icle. While a particular form of inflorescence is characteristic for certain species, there is usually some variation, even on the same plant. The bracts are usually sessile, more or less triangular, acute, with the same general sort of pubescence as that of the calyx and the pedicel. Bracteoles are smaller, but similar in form. Both bracts and bracteoles are usually caducous, but may persist until the flower is fairly mature. Calyz.—The calyx furnishes the more important characters in specific delimitation. These characters are quite consistent and easily distinguished with a lens. One of the striking fea- tures is the type of pubescence; with the exception of A. Gon- zalezti, which is essentially glabrous, the surface is covered either with pilose or glandular hairs, or even with sessile glands. There may be some hairs present with the glands, but one type or the other is dominant. The glandular hairs are usually stipitate, the stipe cells being hyaline and the secretory cell at the tip brown. In one or two species the glands are re- duced, sessile, and merge into puberulence. In 4. quadrifidus and A. Wrightü, the calyx is sessile or only very shortly pedieellate. The calyx-lobes range in size from shallow ones 1-2 mm. long, as in A. Wrightii, to the tentacular and subulate ones, 2 em. long, as in A. Thurberi. All measurements were taken from ealyces in fruit or full flower, as the calyx is accrescent. [Vor. 28 388 ANNALS OF THE MISSOURI BOTANICAL GARDEN Corolla.—When Nees segregated Anisacanthus from Jus- ticia he described the corolla as having an entire posterior Пр. However, earlier workers were correct in describing the lip as slightly emarginate; both types occur and variations may be found on the same plant. The corolla is tubular or funnel- shaped, more or less curved, with spreading, recurved, ligulate to elliptical lobes. The mature corolla is from 315 em. in length, with the lobes usually about as long as the tube. The total length, the length of the tube and lobes, and relative amplia- tion at the throat are of importance in specific delimitation. The corolla is usually bright red and thinly covered with short, pilose, red hairs. In only one species, A. abditus, is the corolla glabrous. Stamens.—The stamens are inserted on the corolla-tube, alternate with the lobes and usually near the base of the cen- tral lobe of the 3-parted anterior lip. The filaments are highly colored and vary in length with the corolla-lobes, seeming to extend to the tips of the lobes. The anther-sacs of Anisacan- thus are essentially parallel, equal, and inserted at the same height on the filament; these characters are key characters in generic delimitation. The filament is attached at the base of the connective which joins the sacs from about half to two- thirds of the length from the top. Pistil.—The ovary is partly surrounded at the base by a disc. The style is filiform, as long as the corolla, and terminates in a slightly enlarged 2-lobed stigma. The characters of the ovary, style, and stigma are essentially the same for all species, and thus of little taxonomic importance. Frwit.—The capsule is generally uniform throughout the genus, but in a few cases it presents characters of taxonomic value. It is typically a subglobose, 2-celled body and a rela- tively distinct stipe of more or less equal length. In A. Gon- zalezii the body is not separated from the stipe by a noticeable constriction, but gradually tapers to the base. The seeds are 2 to 4 in number and are usually nearly parallel and opposite in the capsule. They vary only slightly within the genus, being 1941] HAGEN—THE GENUS ANISACANTHUS 389 more or less discoid, tuberculate and shiny, and held upright at maturity by curved retinacula. GENERIC RELATIONSHIPS Anisacanthus is a member of the tribe Justicieae and the sub- tribe Eujusticieae which is usually placed near the end of the family. The tribe includes plants with a 2-lipped corolla, 2-4 stamens with 1-2-locular anthers. As regards other members of the tribe, the genus appears to be closest to Carlowrightia. However, Anisacanthus has a much larger calyx, corolla, and fruit; and the corolla is 2-lipped (3 plus 1 lobes), whereas in Carlowrightia it is about equally 4-parted. Amisacanthus is a much larger plant than Carlowrightia and, so far as known, eontains no herbaceous species as does Carlowrightia. It ap- pears to be fairly closely related to Chileranthemum and Odontonema, but is probably more advanced than these genera, since it lacks staminodia which they possess. Mucro- nate appendages at the base of the anther sacs are wanting in Anisacanthus; by this character alone it may be distinguished from Justicia. Also the anther-sacs are less equal and more oblique in Justicia. Anisacanthus also has been confused with Beloperone, but this genus has mucronate appendages on the anther-saes and is more closely allied to Justicia. Moreover, Amsacanthus is quite closely related to Jacobinia with which it has been frequently confused. They both lack the ap- pendages on the base of the anther-saes, but the equal and parallel anther-saes of Amisacanthus readily distinguish it from Jacobinia. RELATIONSHIP OF THE SPECIES While the author has not divided the genus Anisacanthus into subgenera or sections, there are certain groups of species which are more closely related to each other than to others. From the calyx alone it is evident that 4. pumilus shows closest relationship to A. tulensis, the main difference being in size (pl. 17, 1 and m). The relationship is also indicated by the (VoL, 28 390 ANNALS OF THE MISSOURI BOTANICAL GARDEN slender, ligulate-lobed corolla, and by the consistent number of 4 seeds in the capsules of both species. Similarly, the calyx of A. Thurberi appears to be an exag- gerated form of that of А. insignis (pl. 17,1 and h). However, these two species are not as closely related as are A. tulensis and A. pumila. The size and proportions of the corolla parts differ, and the seed number is not consistent in A. insignis. The type species, A. quadrifidus, seems closest to A. Wrightu. The calyx is quite similar, although differing in proportions (pl. 17, f, g, and j). As a matter of fact, the more deeply cleft calyx form in А. Wrightti, which is found gen- erally in plants of the area between the two species, was first described as A. junceus and later placed in synonymy under A. quadrifidus. The size and form of the corolla and the char- acteristic secund spicate inflorescence give additional evidence of the relationship. Although the calyx of Anisacanthus abditus somewhat re- sembles that of A. quadrifidus (pl. 17, k), several characters segregate it from all the other species. The large bracts, the nature and extent of the glandular pubescence, and the pro- portions of the corolla are all peculiar to this species. Also, the capsule is smaller than is usual in the genus. Another distinct species is A. Gonzalezii, its cylindrical gla- brous calyx-tube and ciliate calyx-lobes being different from those of every other species (pl. 17, n). The capsule char- acters are also unique in the genus (pl. 17, b), although the seed arrangement somewhat resembles that in A. abditus. GEOGRAPHICAL DISTRIBUTION The species of Anisacanthus are semi-xerophytic, usually growing on exposed, rocky slopes within a range confined almost entirely to Mexico. The center of distribution appears to be in southern Mexico, in the general region of Puebla, Oaxaca, Morelos, and the state of Mexico. Migration seems to have been mainly northward. Only three entities, A. insignis var. linearis, А. Thurberi, and A. Wrightii, extend into the United States. Representatives of the genus are found 1941] HAGEN—THE GENUS ANISACANTHUS 391 throughout Mexico, however, and it may be anticipated that further collections will extend the present known ranges. The ranges of the species in the southern part of Mexico around the main eenter of distribution overlap somewhat, whereas the northern species are more distinct. There seems to be a small secondary center of distribution in the Sonoran region Thurberi @ Wrightii + var. brevilobus x |* Pir rrr FE > s H P Я Е m insignis $ var. linearis + tulensis abditus Fig. 1. The geographical distribution of the species of Anisacanthus in North America. which shows a northerly migration with one species extend- ing into Arizona and southwestern New Mexico. It is interest- ing to note the regions where Anisacanthus is absent; with the exception of the two Sonoran species one might say that it is not found on the coastal slopes or lowlands. In general, the plants of this genus are characteristic of the mountainous region of southern Mexico and the high central plateau ex- tending northward (fig. 1). [Vor. 28 392 ANNALS OF THE MISSOURI BOTANICAL GARDEN ABBREVIATIONS The herbaria from which material was obtained for study and from which specimens have been cited are indicated by the following abbreviations: С = University of California, Berkeley. F = Field Museum of Natural History, Chicago. M - Missouri Botanical Garden, St. Louis. NY = New York Botanical Garden. TAXONOMY Anisacanthus Nees in Linnaea 16: 307. 1842; in DC. Prodr. 11: 445. 1847; Bentham. & Hooker, Gen. Pl. 2: 1117. 1873; Gray, Syn. Fl. №. Am. 21; 326. 1878, and ed. 2, 21: 457. 1886; Engler & Prantl, Die Natürl. Pflanzenf. 48: 327-328. 1895; Standley in Contr. U. S. Nat. Herb. [Trees and Shrubs of Mexico] 23: 1342-1343. 1926. Branched shrubs, mostly about 1 m. high; stems terete, more or less minutely striate, younger branches usually pubescent in two opposite lines continuous with the base of the petioles, older stems covered with brown or gray bark exfoliating in thin strips; leaves opposite, petiolate or sessile, lanceolate to linear, punctate, eystoliths common on the upper surface; in- florescence spicate, racemose or paniculate, flowers borne singly or several at a node, secund or opposite; bracts and bracteoles mostly triangular to linear, usually caducous, agree- ing with the calyx in pubescence; calyx subequally 5-lobed, lobes triangular to linear-acuminate, pubescent to glabrous; corolla usually red, mostly pilose, tubular to funnel-shaped, more or less arcuate, somewhat inflated at the base around the ovary, 2-lipped, usually recurved, the posterior lip entire or slightly emarginate, the anterior lip 3-lobed; stamens 2, epipetalous anteriorally, usually at the base of the central lobe and alternate with the corolla-lobes, filaments more or less thick and fleshy, colored, glabrous, anthers 2-celled, subequal, inserted at the same level or nearly so, parallel or slightly di- vergent at the base, connected to the middle or slightly below, 1941] HAGEN—THE GENUS ANISACANTHUS 393 non-mueronate or appendaged ; style filiform, glabrous, about as long as the corolla, stigma simple, 2-lobed ; fruit a subpyri- form eapsule, slightly beaked, usually narrowed at the base to form a relatively distinct stipe, glabrous and shiny; seeds 2-4, discoid, more or less tuberculate, each supported by a eurved retinaeulum usually at about the same height in the body of the capsule. Type species: Anisacanthus quadrifidus (Vahl) Nees in Linnaea 16: 307. 1842. KEY TO THE SPECIES A. Braets of the inflorescence foliaceous, obscuring the calyx; entire surface of the plant pee at least in the younger stages............ 1. A. abditus AA. Bracts of the inflorescence neither foliaceous nor obscuring the calyx; glands, s popu) confined to the inflorescence. B. Calyx к or nearly so except for the hirsute-ciliate lobes, not AE Te а? A ets ore Fo ole Meee Ls TROC IDE NOE 2. A. Gonzalezii BB. Calyx alsin Se а the lobes not hirsute-ciliate. C. Cor D. ds dex UE laneeolate to laneeolate, pilose to hispid И ee eee NOR TUUM ID UL T 3. A. insignis DD. Leaves sessile, linear to linear-lanceolate, essentially glabrous PTT ee eee See ee За. A. insignis var. linearis CC. d 3—4 em. long Calyx 9-20 mm. long; stamens epipetalous below sinuses of PNVETIOL Пр. Gn ls vov о ои 4. A. Thurberi DD. Calyx 5-10 mm. long; stamens epipetalous at sinuses of anterior lip. E. Calyx mostly 7-10 mm. long; lobes 5-6 mm. long, attenuate e Е е ОК КГК: ТАЕКЕ QE vie ГУ 5. A. quadrifidus EE. Calyx mostly 5 mm. long; lobes 1-3 mm. long, more or less aeute, not alternate F. Calyx-tube equal ia or shorter than the 1оЪез.............. PDF Wu ceu n LE IMP 6. A. Wrightüi FF. Calyx-tube much longer than the "in RON EE не РРР И de ба. A. Wright var. brevilobus BBB. Calyx pubescent, not glandular; the lobes not hirsute-ciliate. C. Ultimate branches pubescent in lines; ealyx 7-12 mm. long........ ICE IU ERE Ани" . A. pwmilus CC. Ultimate branches evenly pubescent over entire ненадо ealyx long... Ул жуз RENIBUS ee ар 8. A. tulensis 1. A. abditus Brandg. in Zoe 3: 348-349. 1893; Standl. in Contr. U. S. Nat. Herb. 28: 1342. 1926. Shrubs about 1-2 m. high, more or less glabrous-pubescent [Vor. 28 394 ANNALS OF THE MISSOURI BOTANICAL GARDEN throughout; leaves ovate to lanceolate, 1-6 em. long, 0.5-3 cm. broad, petioles slender, 0.5-2.5 em. long; bracts of the inflores- cence quite distinct from the upper leaves, sessile, ovate to ob- spatulate, obtuse, about 1 cm. long, 4-5 mm. broad; inflores- cence a terminal or lateral spike; calyx obscured by the sub- tending bracts, 5-7 mm. long, rather thin and delicate, deeply 5-cleft, lobes lance-attenuate; corolla 3-4 em. long, glabrous, tubular, the 4 lobes subequal, shorter than the tube, 5 mm. broad, ovate; filaments alternate, attached near the base of the central anterior lobe, 7-9 mm. long, anthers 3 mm. long, saes connected to below the middle; ovary extending about 1 mm. above the disc; capsule 1 em. long, about 3 mm. in diam- eter, the body more or less oblong, slightly longer than the stipe; seeds 4, 2-3 mm. in diameter, tuberculate, one pair borne obliquely about 2 mm. above the other. DISTRIBUTION: mountainous region of central Sonora, Mexico. Mexico: SONORA: Las Durasnillas, Мау 14, 1892, Brandegee s.n. (C TYPE, Е); vicinity of Alamos, March, 1910, Rose, Standley $ Russell s.n. (Е); Caramechi, Rio Mayo, Dee. 8—10, 1934, Gentry 1178 (Е, NY, M). 2. A. Gonzalezii Greenm. in Proc. Am. Acad. 39: 89. 1903; Standl. in Contr. U. S. Nat. Herb. 23: 1342. 1926. Suffruticose; stems uniformly pubescent; leaves ovate- lanceolate to linear-lanceolate, 1.5—4 ст. long, 4-17 mm. broad, puberulent or often slightly pubescent above, pilose below, petioles 1-5 mm. long, pubescent ; inflorescence spicate, flowers 1-2 in each axil, sessile; bracts linear to linear-lanceolate, 0.5-2 em. long, 1-3 mm. broad, finely pubescent, ciliate; calyx 5-11 mm. long, tube cylindrical, glabrous, lobes lance-attenuate, hirsute-ciliate, about as long as the tube; corolla 3.5—4 em. long, slightly infundibuliform, posterior lip emarginate, lobes about one-third total length of the corolla, ligulate, slightly recurved ; stamens attached near the base of the anterior lip and alternate with the lobes, filaments 11-13 mm. long, anther-sacs 4 mm. long, parallel, connected to below the middle; ovary extending about 1.5 mm. above the dise; capsule 16-19 mm. long, 3—4 mm. in diameter, gradually tapering to the base; seeds 4, 3-4 mm. in diameter, one pair borne obliquely above the other. 1941] HAGEN—THE GENUS ANISACANTHUS 395 DISTRIBUTION: central and northern Oaxaca, Mexico. МЕХІСО: OAXACA: Las Naranjas, Aug., 1908, Purpus 3018 (C), and $041 (Е, M, NY) 3. A. insignis Gray, Syn. Fl. N. Am. ed. 2, 21: 457. 1886; Standl. in Contr. U. S. Nat. Herb. 23: 1343. 1926. Р]. 18, fig. A. Drejera puberula Torr. in U. S. & Mex. Bound. Surv. Bot. 123. 1859, in part. Anisacanthus pumilus Wats. in Proc. Am. Acad. 18: 133. 1883, in part, not Nees in DC. Prodr. 11: 445. 1847. Slender, vine-like shrub, 1.5-2.5 m. high; stems pubescent in lines; lower leaves lanceolate to ovate, 2.5-7 em. long, 1-3 em. broad, slightly acuminate, puberulent, pilose to tomentose, petioles 3-20 mm. long, tomentose; flowers borne in short race- mose clusters in the axils of fallen leaves; bracts and bracteoles puberulent, sometimes glandular, obovate, elliptic or ovate, re- duced from about 10 mm. to about 2 mm. in length; pedicels 3-8 mm. long, glandular ; calyx 6-10 mm. long, stipitate-glandu- lar, lobes separate almost to the base, linear-ligulate, somewhat attenuate; corolla mostly 5 em. long, curved, slightly infundi- buliform, lobes linear, recurved, longer than the tube; stamens alternate, inserted at the base of the central anterior lobe, anthers 4 mm. long, anther-sacs parallel, slightly oblique; disc at base of ovary about 1 mm. high; capsule 2 em. long, stipe as long as or longer than the body; seeds 4, sometimes 1 or 2 aborted, about 5 mm. in diameter, slightly obliquely discoid, brown, more or less tubereulate. DisTRIBUTION: central Mexican plateau region from Durango to Coahuila and Chihuahua. MEXICO: CHIHUAHUA: gravelly banks along the Cibolo of the Rio Grande, May- June, cig s.n. (NY); valley of the Rio Conchos below Santa Rosalia, April 21, 1847, Gregg 504 (M, NY); vicinity of Chihuahua, alt. 1300 m., April 8-27, 1908, Palmer d ye M, NY); rocky hills near Chihuahua, March 23 & April 17, 1885, 2 M, NY); Santa ee south of in 1 + April 30, 1847, Wislizenus 267 (M COAHUILA: Val of Parras, April 11, 1847, Gregg 40? (M); Parras, June 8-28, 1880, Palmer on (F, NY), and Oct. " 11, 1898, 450 (C, F, M, NY); Parras, Feb.- March, 1905, iia 1046 (C, F, M, NY); Sierra de Parras, Oct., 1910, банн 4752 (C, Е, М). DURANGO: City of Durango and vicinity, Apsil- Mot, 1896, Palmer 25 (С, Е, М, NY); ie April 11, 1900, Trelease 75 (M). [Vor. 28 396 ANNALS OF THE MISSOURI BOTANICAL GARDEN За. A. insignis Gray var. linearis Hagen, var. nov.” Pl. 18, fig. B. Lower leaves and those of the flowering branches sessile, linear to linear-lanceolate, 1-4 em. long, 1-5 mm. broad, es- sentially glabrous. DISTRIBUTION: southwestern Texas and northern Mexico. TEXAS: CHISOS MOUNTAINS, BREWSTER СО.: lower end of Juniper Canyon, alt. 1158-1219 m., July 15-18, 1921, Ferris § Duncan 2990 (M, NY); dry creek bed, lower Blue Creek Canyon, alt. 1370 m., June 23, 1931, Moore ф Steyermark 3202 (C, М ТУРЕ, NY); Aug. 2, 1931, Mueller 8166 (Е, М); rocky cliffs and ledges, Oak Canyon, May 24, 1928, Palmer 34148 (M, NY). MEXICO: COAHUILA: near the northern entranee of El Puerto de San Lazaro, June 16, 1936, Wynd $ Mueller 106 (M). 4. A. Thurberi (Torr.) Gray, Syn. Fl. N. Am. 2': 328. 1878, and ed. 2, 2': 457. 1886; Hemsl. Biol. Centr.-Am. Bot. 2: 522. 1882; Woot. & Standl. in Contr. U. S. Nat. Herb. 19: 597. 1915; Standl. in Contr. U. S. Nat. Herb. 23: 1342. 1920. Drejera Thurberi Torr. in U. S. & Mex. Bound. Surv. Bot. 124. 1859. Drejera puberula 'Torr. Ibid. 123. Low shrub, 0.5-1.5 m. high; stem pubescence reduced to a few seattered hairs in two opposite lines; leaves lanceolate, mostly 3-5 em. long, 7-15 mm. broad, hispid to glabrate, petioles 1-5 mm. long; flowers borne singly in the axils of the bract-like upper leaves, or in 2—5-flowered greatly fore- shortened axillary racemose clusters; bracts and bracteoles lanceolate to linear, acute, mostly 2-10 mm. long, 1-3 mm. broad, puberulent, often slightly pilose along the midrib and margin; pedicels 3-7 mm. long, glandular; calyx .9-2 em. long, deeply 5-cleft, stipitate-glandular, puberulent, with few pilose hairs scattered throughout, lobes linear-attenuate, more or less tentacular; corolla dull red, 3.5-4 em. long, tube funnel- form, lobes shorter than the tube, ovate, obtuse, divergent but scarcely recurved, upper lip slightly emarginate; stamens т Anisacanthus insignis Gray var. linearis, var. nov., A. insigni similis, sed foliis inferioribus sessilibus linearibus vel lineari-lanceolatis, 1-4 em. longis, 1-5 mm. latis, glabris differt. 1941] HAGEN—THE GENUS ANISACANTHUS 397 epipetalous, inserted below the sinuses of the anterior lip, fila- ments about 2 em. long, anther-sacs 3-4 mm. long, equal, par- allel, connected to below the middle; capsule about 15-18 mm. long, about 6 mm. in diameter, stipe shorter than the subovoid body; seeds 2, slightly tuberculate, about 5 mm. in diameter. DISTRIBUTION: from southern Sonora in Mexico northward along the western mountainous ‘Slopes into central Arizona, and eastward into the southwestern part UNITED STATES: ARIZONA: near creek, Rhoda Riggs’ Ranch, Chiricahua Mts., alt. 1676 m., Oct. 22, 1906, Blumer 1304 (F, M, NY); Box Canyon, Chiricahua Mts., alt. 1676 m., July 9, 1907, Blumer 1248 (F, M, NY); Nogales, May, 1892, Brandegee s.n. (C); Fort Whipple, on gravelly hillsides, May 3, 1865, ан + Palmer 190-с (М); 40 miles south of Fort Whipple, Aug. 5, T Coues $ P 130 = ); 5 miles west of Tucson, in Tucson Mts., March 7, 1934, ш. pd F); Douglas, rocky draws and washes, May, 1907, Goodding per (С, М); dry, Iis hillsides 5 miles northeast of Rodeo, June 16, 1930, Goodman 4 Hitchcock 1153 (C, F, M, NY); Rincon Pass, Oct. 19, 1900, Griffiths e (NY); Sabenio Cafion, March 30, 1901, Griffiths 2595 (NY); Santa Ritas, foot of Old Baldy, E 1901, Grifiths 2652 (NY); Roadside Mine, Pima County, April 21, 1932, rison $ Kearney 8528 (Е); hills along the Rio San Pedro, Sept. 9, 1858, еи 20 (NY); Congress Junction, alt. 91 m., May 4, 1903, Jones s.n. (M); Cave Creek Canyon, Chirieahua Mts., 182 ads: m., June 26-29, 1927, Kusche s.n. (Е); Santa Catalina Mts., April, 1881, Lemmon s.n. (C) ; south of Bisbee, Mexican boundary line, Oet. 3, 1899, Mearns 1024 (С, M, NY); Patagonia, May 6, d an May 25, 1884, 195 (F); E M ui Bs: ain 25, 1884, Parish s.n. (NY); dn April, 1884, Parish $ Parish s.n. (С); acho Mts. . March 23, 1930, Peebles 6473 (NY); Santa Catalina n pul 18, e Pringle s.n. (F), and June, 1882, s.n. (M, NY); Sierra Tueson, April 25, 1884, Pringle s.n. (F), and June 1, 1884, 4832 (F, NY); Tueson, in arroyo, April 15, 1901, Shear 4231 (NY); Tucson, date lack- ing, Spring s.n. (C) ; west of Sonoita, alt. 1433 m., April 18, 1934, Stone 39 (NY); ж tone Cabin Canyon, Santa Rita Mts., alt. 1524 m., May 23, 1903, Thornber 294 C, М); foothills Ut Tueson Mts., alt. 762 m., May 9, 1903, Thornber 474 (C) ; Ft. чакащи: A m Wilcox s.n. d and E 1. NEW MEXICO: ong the Gila River and m liff and upper eanyon of Gila, "ami Co. ES 1300—1350 m. ptm 25, a anas 16794 (M); Mangas Canyon, May and Jue. 1880, Greene s.n. (NY), June 1, and дек: 3, 1880, s.n. (Е), and June and July, 1880, 12502 (M) ; Dog Spring, Grant Co., May 26, 1892, Mearns 113 (NY); Mangas T 18 miles northwest of Silver City, alt. 1453 m., June 9, 1903, Metcalfe 113 (C, M, NY); on the upper Rio Grande, below El Pa 1851- 1852, Wright 1456 (M, N Y). MEXICO: SONORA: San Miguel de Horcasitas, May, 1892, Eisen s.n. (С); San Bernardo, Rio Mayo, Feb. 16, 1935, Gentry 1309 (F, M, NY); El Alamo, Magda- lena, May 25, 1925, Kennedy 7107 (C); Torres, Jan.-Mareh, 1902, Purpus 409 (С, M). [Vor. 28 398 ANNALS OF THE MISSOURI BOTANICAL GARDEN 5. A. quadrifidus (Vahl) Nees in Linnaea 16: 307. 1842; Standl. in Contr. U. S. Nat. Herb. 23: 1343. 1926. Justicia coccinea Cav. Те. Pl. 2: 77, pl. 199. 1793, not Aublet. Justicia quadrifida Vahl, Enum. Pl. 1: 124. 1805. Justicia virgularis Salisb. Parad. Lond. pl. 50. 1806. (?) Justicia superba Hort. ex Nees in DC. Prodr. 11: 445. Anisacanthus virgularis (Salisb.) Nees in DC. Prodr. 11: 445. 1847; Hemsl. Biol. Centr.-Am. Bot. 2: 522. 1882; Gray, Syn. Fl. М. Am., ed. 2, 21: 457. 1886 Younger branches green, somewhat purplish, pubescent in two opposite lines; leaves short-petiolate to subsessile, linear to lanceolate, long-acuminate, 2-5 em. long, 3-13 mm. broad, glabrous except for minute pubescence at base and along mid- rib above, punctate, the upper leaves much reduced; inflores- cence spicate, secund, flowers sessile to short-pedicillate, al- most always solitary, rarely 2 in the axils of the upper bracts; bracts and bracteoles lanceolate, glandular, ciliate, 3-4 mm. long, caducous; calyx glandular, 6-10 mm. long, deeply 5-cleft, lobes ciliate, lance-acuminate, 5-6 mm. long at anthesis; corolla scarlet, 3.5—4 em. long, slightly curved, scarcely dilated at the throat, lobes recurved, ligulate, about as long as the tube; stamens inserted at the sinuses of the anterior lip, anthers 3-4 mm. long, anther-saes equal or nearly so, parallel; capsule 15- 18 mm. long, stipe as long as the body; seeds 4, subdiscoid, about 5 mm. in diameter, more or less tuberculate. DisTRIBUTION: central Oaxaca, Mexico, northward into Puebla, Hidalgo and Queretaro, and westward into San Luis Potosi and Zacatecas MEXICO: HIDALGO: Ixmiquilpan, July, 1905, Rose, Painter $ Rose 8944 (Е, M, NY); oaxaca: Vallé de Oaxaca, alt. 1600 m., Nov. 8, 1906, Conzatti 1519 (F); Las Naranjas, Aug., 1908, Purpus 3018 (F, M, NY) ; Oaxaea Valley, alt. 1524 m., Nov. 7, 1894, Smith 731 (M, NY); PUEBLA: vieinity of Puebla, Ix Distriet Pos gs, alt. 2110 m., July, -— Де $ Amable 3567 (M, NY), and Aug., ie 8 (M); Tehuaean, Dec., 1892, зно 10600 (Е); entes. Nov., 1909, M 35 (Е); valley near чне alt. 1524 m., Aug. 5, 1901, Pringle 9390 (F, M); d dur 1911, Purpus 5669 (C); near El Riego, Tehuacan, Sept., 1905, Rose, P er $ Rose 100383 (Е, М); QUERETARO: near San Juan del Rio, Aug. 17, 1905, pons Painter $ Rose 9515 (NY); same locality, Nov., 1827, Berlandier 1237 (NY); SAN LUIS POTOSI: Dec. 27, 1848, Gregg 578 (M); 1878, rry $ Palmer 706 (Е); gravel washes, Bocas, Aug. 17, 1891, Pringle 3820 (С, 1941] HAGEN—THE GENUS ANISACANTHUS 399 F, NY); region of San Luis Potosi, 1879, Schaffner 367 (F, NY); ZACATECAS: city of Zacatecas, 1908, Lloyd 10 (F). Besides the above, several authentic collections from Euro- pean gardens have been examined. Although data accompany- ing them are usually quite incomplete, they have been of im- portance in establishing the identity of this species. 6. A. Wrightii (Torr.) Gray, Syn. Fl. N. Am. 21: 238. 1878, and ed. 2, 21: 457. 1886; Hemsl. Biol. Centr.-Am. Bot. 2: 522. 1882; Standl. in Contr. U. S. Nat. Herb. 23: 1343. 1926. Drejera Wrightii Torr. in U. S. & Mex. Bound. Surv. Bot. 123. 1859. Drejera juncea Torr. Ibid. 124. Anisacanthus junceus (Torr.) Hemsl. Biol. Centr.-Am. Bot. 2: 522. 1882. Young branches dull green, pubescent in two lines; lower leaves broadly lanceolate, 1—5 em. long, 0.5-2 em. broad, hispid to glabrate, petioles 3-10 mm. long, pilose; flowers usually borne singly or in pairs in secund, terminal, spicate inflores- cences, subsessile to short-pedicellate; bracts and bracteoles lanceolate-acuminate, 2-5 mm. long, puberulent; calyx about 0.5 em. long, puberulent, glandular, lobes ovate to lanceolate, 2-3 mm. long, acute, few, scattered hairs along margin and at tip; corolla 3-4 em. long, lobes narrowly ovate, obtuse, pos- terior lobe slightly emarginate, tube slender, nearly straight, scarcely dilated at the throat, longer than the lobes; stamens epipetalous, inserted near the sinuses of the central anterior lobe, filament about 1 em. long, anthers 3 mm. long, anther-sacs connected nearly half their length from the tips; ovary extend- ing about 2 mm. above the disc ; capsule about 15 mm. long, body about 6 mm. thick, shorter than the stipe; seeds 2, rarely 4, about 5 mm. in diameter. DisTRIBUTION: mostly along the eastern part of the central plateau region of Mexieo, in the states of Tamaulipas, Nuevo Leon, and Coahuila. It has been found at one station in northern Michoacan, and from several localities in Bexar, Uvalde, and Kinney counties in Texas. STATES: TEXAS: Fort Clark, June-July, 1857, Blake s.n. (NY); San Antonio, Bexar Co., tropical life zone, June 27, 1911, Clemens $ Clemens s.n. (F, M); Conean, Uvalde Co., along rocky creeks, June 15, 1916, Palmer 10212 [Vor. 28 400 ANNALS OF THE MISSOURI BOTANICAL GARDEN bead: valley of the Rio Grande below Donana, Parry, Bigelow, Wright 4: Schott п. (NY); rocky banks along the Rio а June, 1884, Reverchon 1580 (Е, M, NY); j same TP Schott s.n. (Е), and Oct. 19, pe 738 pedi Uvalde Co., MEXICO: COAHUILA: La Pena, Nov., 1852, Thurber 849 (NY TYPE of Drejera juncea Torr.); MICHOACAN: Morelia "Dos Tetecos, March 7, 1911, Arséne 36 (F); NUEVO LEON: Monterrey, Gtindatens, alt. 540 m., June, 1911, Arsène $ Abbot 6228 (M); Monterrey, damp rich soil in thickets, Edwards $ Eaton s.m. (NY); by streams, El Carrizo, alt, 427 m., June 16, 1906, Lozano 10256 (C, F, M, NY); arroyos of me ear Мойту, July 23, 1888, Pringle 1891 (C, F, NY); valley of Monterrey, phn 17, 1889, Pringle 2710 (C, F, M), and ria 19, 1903, 11663 (F); TAMAULIPAS: La Vegonia, vieinity of San Jose, alt. 823 m., July 20, 1930, Palin 10526 (Е); La Tamaulipeca, vicinity of San Mi sa July 27, 1930, pgs 10670 (Е); vicinity of Victoria, alt. 320 m., Feb. 1-April 9, 1907, Pale? 120 (C, F, M, NY). The specimens of Lozano and Bartlett are atypical, having са]усез very nearly like those of A. quadrifidus. Thurber’s specimen, which Torrey called Drejera juncea, is somewhat like the two above, but judging mainly by the glandular characters, general size of the calyx and of the anthers, its relationship is evidently with A. Wrightii rather than A. quadrifidus. 6a. A. Wrightii (Torr.) Gray var. brevilobus Hagen, var. nov.’ Lower leaves subsessile, lanceolate to linear-lanceolate, 1.5—4 em. long, 3-8 mm. broad, glabrate; calyx 4-5 mm. long, the tube several times longer than the triangular, acute lobes. DISTRIBUTION: northern Mexico, known only from the type locality. MEXICO: COAHUILA: Chojo Grande, 27 miles southeast of Saltillo, July 16, 1905, Palmer 719 (C, F, M TYPE, NY). This variety is quite distinct on account of the short, triangu- lar, acute calyx-lobes, and the narrower, subsessile leaves. 7. A. pumilus (Dietr.) Nees in DC. Prodr. 11: 445. 1847; Gray, Syn. Fl. №. Am., ed. 2, 21: 457. 1886; Hemsl. Biol. Centr.- Am. Bot. 2: 522. 1882; Standl. in Contr. U. S. Nat. Herb. 23: 1342. 1926. *Anisacanthus Wrightii (Torr.) Gray var. brevilobus, var. nov., foliis in- ferioribus subsessilibus lanceolatis vel lineari-lanceolatis, 1.5-4 em. longis, 3-8 atis, glabris; petiolis 1-2 mm. longis; calyce 4-5 mm. longo, lobis tubo nis aed triangularibus, acutis. 1941? HAGEN—THE GENUS ANISACANTHUS 401 Justicia pumila Dietr. in Vollst. Lex. Gartn. Nachtr. 4: 197. 1818. Drejera Greggu Torr. in U. S. & Mex. Bound. Surv. Bot. 124. 1859. Anisacanthus Greggü (Torr.) Gray, Syn. Fl. N. Am. 21: 328. 1878. | Shrub, 2.5—3 m. high ; stem dark brown, pubescent in two op- posite decussating lines; leaves ovate-lanceolate to linear- lanceolate, 2-3 em. long, 3-10 mm. broad, hispid, subsessile to short-petiolate; flowers in compact racemose clusters of from one to several borne in the axils of fallen leaves; bracts and braeteoles about 2 mm. long, puberulent, ciliate, acute; calyx short-pedicellate, about 7-12 mm. long, puberulent, pilose, deeply and subequally 5-lobed, the lobes about 6-9 mm. long, acute; corolla 4—5 em. long, bright red, thinly pilose, slender, eurved, slightly dilated at the throat, the linear-ligulate lobes longer than the tube, recurved ; stamens inserted near the base of the central lobe of the anterior lip, anther-sacs 4 mm. long, nearly equal, parallel, connected about half way to the base; eapsule about 2 em. long, body 5-6 mm. thick, as long as the stipe; seeds 4, about 5 mm. in diameter, slightly tuberculate. DisTRIBUTION: northern Miehoaean to San Luis Potosi, and northwestward along the plateau region to southern Chihuahua. Mexico: Gonzales Junction, April, 1910, Rusby s.n. (NY); MICHOACAN: vicinity of Morelia, Quinceo, alt. 1900 m., Mareh 11, May 25, 1909, Arséne 2778 (M), and July 18, 1909, s.n. (NY); hills east of Zipimeo, May 11, 1849, Gregg 810 (M); GUANAJUATO: Silao, July, 1903, Purpus 140 (C); SAN LUIS POTOSI: in the region of San Luis Potosi, 22? М. Lat., alt. 1829-2434 m., 1878, Parry $ Palmer 706% (M); CHIHUAHUA: battleground of Paso del Gallinero, near Dolores, Dee. 29, 1848, Gregg 587 (NY ТУРЕ of Drejera Gregg Torr., M). 8. A. tulensis Greenm. in Field Mus. Publ. Bot. Ser. 2: 343. 1912. Justicia superba Sessé & Mocino, Pl. Nov. Hisp., ed. 2, p. 3. 1893, in part; not J. superba Hort. ex Nees in DC. Prodr. 11: 445. 1847, in synonymy. Ultimate branches tomentose to pilose over entire surface; leaves lanceolate to linear-lanceolate, 1.5-5 ст. long, 4-20 mm. broad, puberulent above, sparsely pubescent below, mostly (VoL. 28 402 ANNALS OF THE MISSOURI BOTANICAL GARDEN along the veins and margins, petioles 1-7 mm. long, tomentose to pilose; inflorescence mostly secund, racemose, usually not more than 2 flowers maturing in a cluster; bracts linear- lanceolate, 5-10 mm. long, puberulent, ciliate, bracteoles similar but smaller, both bracts and bracteoles caducous; pedicels 1-5 mm. long, tomentose to pilose, puberulent; calyx puberulent, tomentose to sparsely pilose, 10-14 mm. long, lobes lanceolate, as long as the tube or slightly longer; corolla 5—5.5 em. long, finely pilose, tube slender, arcuate, about 2 em. long, lobes linear-ligulate, recurved, distinctly longer than the tube, pos- terior lip slightly emarginate ; stamens alternate, inserted near the base of the central lobe of the anterior lip, filaments 18-25 mm. loug, anthers 4 mm. long, anther-sacs subequal, slightly oblique, connected almost half their length from the tip; cap- sule 17-25 mm. long, about 5-8 mm. thick, stipe longer than the body; seeds 4, subdiseoid, 5-6 mm. in diameter, tuberculate. DISTRIBUTION: southern Mexico. MEXICO: OAXACA: Santa Maria del Tule, alt. 1600 m., March 31, 1907, Conzatti 1778 (Е TYPE); MORELOS: Ayacapixtla, 1787-1795-1804, Sesse 4 Мосто 293 in part, and 373 (F). EXCLUDED SPECIES Anisacanthus glaberrimus Jones in Contr. West. Bot. No. 15: 151. 1929 = (?) Odontonema. List or ExstccATAE Arséne, G. 36 (6), —, 2778 (7). Detwiler, S. B. 4? (4). Arsène, С. & Bro. Abbot. 6228 (6). Edwards, Dr., & Maj. Eaton. — (6). — G. & Bro. Amable. 1568, 3567 Eggleston, Ww. W. 16794 (4). isen, G. — (4). аны. Н.Н. 10526, 10670 (6). Ferris, Roxana 8. & C. D. Dunean. 2990 Berlandier, J. 1237 (5). (За). Bigelow, 7. M. — (3). Gentry, Howard Seott. 1178 (1), 1809 Blake, S. F. — (6). (4). Ийме, J. C. 1248, 1304 (4). Goodding, Leslie N. 2232 (4). Brandegee, T. S. — (1), — (4). Goodman, George J. & C. Leo Hiteheoek. Clemens, Mr. & Mrs. Joseph. 1065 (6). 1153 (4). Conzatti, C. 1519 (5), 1778 (8). Greene, Edward Lee. —, —, 12502 (4). Coues, Elliott & Edward Palmer. 130, Gregg, J. 407, 504 (3), 578 (5), 587, 190-c (4). (7). 1941] HAGEN—THE GENUS ANISACANTHUS Griffiths, David. 2018, 2595, 2652 (4). Harrison, G. H. & T. H. Kearney. 8528 4 (4). E Sutton. 598 (4). Jones, Mareus E. — (4 Kennedy, P. в. 210? (4). Kusche, J. Aug. — (4). Lemmon, J 4). Liebmann, 10600 (5). Lloyd, Franeis E. 10 (5). L F.L 56 (6). Mearns, Edgar A. 113, 1024 (4). Metealfe, O. B. 113 (4). Moore, John Adam & Julian Steyer- mark. 3202 (3a). Mueller, Cornelius H. 8166 (3a). 4 Palmer, Edward. 25, 51, 430, 1017 (3), — (4), 120 oe 719 (6a). Palmer, Ernest 10212 (6), 34148 a). Parish, W. F. 195 (4). "eis Samuel B. & W. F. (4). rry, C. J. M. Bigelow, Charles pius & А Sehott. — (6). Parry, C. C. & Edward Palmer. 706 (5), 706% Peebles, В. H. 6473 (4). а. м р 403 Pringle, C. G. 268, 862 (3), —, 4832 (4), 1891, 2710, 11663 53 (6), $820, 9390 (5). Purpus, C. A. 140 (7). 409 (4), 1046, 4752 (3), 3018, 3041 (2), 3018, 5669 (9). Reverchon, J. i (6). Rose, J. N., Joseph H. Painter & J. 8. 8944, Me 10033 (5) Paul C. Standley & P. G. Rusby, H. H. — (7). Mini а. G. 367 (5). Sch 733 (6). MA pus D. 3057 (6). Sessé, Моето, Castillo & Maldofiado. 293, 873 (8). Shear, C. L. — (4). Smith, Charles L. 781 (5). Spring, — — (4 Stone, Mrs. Predariek M. 89 (4). Thornber, J. J. 294, 474 (4). ). Thurber, George. 849 (6 Trelease, William. 75 (3). Wileox, T. E. —, — (4). Wislizenus, F. 26? (3). Wright, Charles. 1456 (4), 435 (6). d, yle & Cornelius H. Mueller. (3a 106 INDEX ro GENERA AND SPECIES Previously published and accepted names are cip by ordinary type; new names by bold face type; and synonyms by it Page Ааа Goce а ас ose E 392 ВОО ee 393 glüberrwius eres 402 Gonzales о. 394 LOL TH е ее е: 401 ПОНУН Е 395 а РЕНО Т 396 be bu ОРНА E JE 399 оО ero Е. 400 ULT e PET алати 395 спра шн. 2:.5........- 398, 398 Page д Coss ева кор co is 396 TulGnBiB- oos ТУТА 401 ТР Svo eub vei 398 УМЕ .,....... tnm n 399 var Drevilobus оо, 400 Beloperone .......... eer rn 389 ODarlownghtia ....2... 4 rn 389 Omleranthémum ало 44. cues 389 (bbs ОЧА OI IO 386 В. 401 Lc) Seay a E EE IR E 399 РЪЦЕ: о. 395, 396 [Vor. 28, 1941] 404 ANNALS OF THE MISSOURI BOTANICAL GARDEN Page Page а ни OA РЕЕСТР СРТ Рр 401 DAG das va ТУА PE Y 399 ПиН TA. oo rra a nist a a T ИРИНА 389 с dd М 398, 401 ПИ а И 385 OFFIC ciu rod 398 si ir, MP о ae eae 398 . ОВОМ „лето 389, 402 EXPLANATION OF PLATE PLATE 16 A. From Salisbury, ‘Paradisus Londinensis,’ pl. 50. 1806, illustrating Justicia virgularis Salisb., which equals Anisacanthus quadrifidus (Vahl) Nees. B. From Cavanilles, “Теопев et Deseriptiones Plantarum,’ pl. 199. 1793, illustrating Justicia coccinea Cav., which equals Anisacanthus quadrifidus (Vahl) Nees, the type species of the genus. NUWHOVH AHL SOHINVOVSINV SANNAÐ ANN. Mo. Bor. GARD., Vor. 28, 1941 № INS à “ « NU N Š. У MOLI AL AY POM Hn Py 90D WIDILS PLATE 16 VANS [Vor. 28, 1941] 406 ANNALS OF THE MISSOURI BOTANICAL GARDEN ExPLANATION OF PLATE PLATE 17 = . Anther of Anisacanthus quadrifidus x 5. Dehisced capsule of Anisacanthus Gonzalezii showing position of seeds, x 1.4. Single valve of the capsule of Anisacanthus abditus inner-face view, x 1.4. Dehiseed capsule of Anisacanthus quadrifidus with one seed in plaee, x 1.4. Single flower of Anisacanthus quadrifidus, x 2.1 alyx of Anisacanthus Wrightii x 3.5. Calyx of Anisacanthus Wrightii var. brevilobus x 3.5. Calyx of Anisacanthus insignis x 3.5. Calyx of Anisacanthus Thurberi x 3.5. . Calyx of Anisacanthus quadrifidus x 3.5. . Calyx of Anisacanthus abditus x 3.5. Calyx of Anisacanthus pumilus x 3.5. m. Calyx of Anisacanthus tulensis x 3.5 n. Calyx of Anisacanthus Gonzalezii x 3.5. тз њоро т o ге pp с. p ANN. Mo. Вот. GARD., Vor. 28, 1941 PLATE 17 HAGEN—THE GENUS ANISACANTHUS EXPLANATION OF PLATE PLATE 18 A. Anisacanthus insignis Gray. From a typical specimen, Palmer 430, in the Missouri Botanical Garden ageing B. Anisacanthus insignis Gray v ar. Иначе Hagen. From the type collection, Moore § Steyermark 3202, in the Missouri Botanieal Garden Herbarium. PLATE 18 8, 1941 9 Mo. Bor. GARD., VOL. ANN. ко, 430 m Anisacanthus insignis Graz $ 146460 7 EARUM — HAGEN—THE GENUS ANISACANTHUS 3“ BOTANIC А 2 1007515 мат $ PLANTS OF TEXAS him Mosnterns Baronne Cts м 9202 Anisacanthus insigni н <= Lower е CONTRIBUTIONS TOWARD A FLORA OF PANAMA! V. COLLECTIONS CHIEFLY BY PauL Н. ALLEN, AND ву Ковевт Е. Woopson, Jr. anD Ковквт W. ScHERY ROBERT E. WOODSON, JR. Assistant Curator of the Herbarium, Missouri Botanical Garden Assistant Professor in the Henry Shaw School of Botany of Washington University | AND ROBERT W. SCHERY Assistant in the Henry Shaw School of Botany of Washington University LYCOPODIACEAE (W. Е. Maxon, Washington, D. C.) Lycopoptum Horrmannt Maxon—cuirigui: Casita Alta to Cerro Copete, alt. 2300-3300 m., July 10, 1940, Woodson & Schery 364. New to Panama. Known previously only from Volcan Barba, Costa Rica, at 2500-2800 m. LYCOPODIUM ORIZABAE Underw. & Lloyd—cutiriqui: Casita Alta to Cerro Copete, alt. 2300-3300 m., July 10, 1940, Wood- son & Schery 365. New to Panama. Heretofore known only from Orizaba, Mexico, and a single collection from Estrella, Costa Rica (Stork 1959). POLYPODIACEAE (W. Е. Maxon, Washington, D. C.) ELAPHOGLOSSUM REVOLUTUM (Liebm.) Мооге—снтвтотт: vi- cinity of Boquete, alt. 1200-1500 m., July 24—26, 1940. New to Panama. The specimens agree closely with type material col- lected in Mexico by Liebmann and with Cuban and Hispaniola specimens, of which a good series is at hand. Elaphoglossum rampans (Baker) Christ, founded on Wright’s no. 3959, from Cuba, is synonymous. TAXACEAE (P. C. Standley, Chicago) Popocarpus Allenii Standl., sp. nov. Arbor excelsa 35 m. alta et ultra, trunco basi 1.5 m. diam., ramulis brevibus crassis sat 1 Issued November 27, 1941. (409) [Vor. 28 410 ANNALS OF THE MISSOURI BOTANICAL GARDEN dense foliatis; gemmae parvae vix ad 2 mm. longae, perulis crasse coriaceis apice rotundatis; folia alterna rigida coriacea anguste lanceolata, adulta 3.5-4.5 ст. longa 7-8 mm. lata, apicem rigidum fere spiniformem versus sensim attenuata, basi acuta et sessilia, costa supra prominente vel prominula, apicem versus minus prominente, subtus vix prominula saepius fere plana; caetera ignota.—PANAMÁ: moist upper region of Cerro Campana, alt. about 1000 m., April 21, 1941, Allen 2437 (Herb. Field Mus., TYPE), 2424 (seedling) ; сос: hills north of El Valle de Antón, vicinity of La Mesa, alt. about 1000 m., Jan. 21, 1941, Allen 2298. The collector's notes regarding the tree are as follows: ‘‘For nearly a year I have been picking up seedlings of this tree, any- where from 45 cm. to 3.5 meters tall. They are everywhere in the hills above Campana, and north of El Valle, at elevations of 600 to perhaps 960 meters. Finally the mature trees have been located on Cerro Campana. The trees are about 33-36 meters tall, without branches for 18-21 meters. The huge columnar trunk is fully 1.5 meters in diameter, and covered with reddish brown, fissured bark somewhat resembling that of the California redwood.”’ It is with considerable diffidence that the writer proposes this new species of Podocarpus, although it seems to be quite as distinct as a good many other species that have been proposed in the genus. One might expect the Panama tree to be identical with P. oleifolius Don, common in the nearby mountains of Costa Rica, and extending southward to Peru. In Costa Rica the genus is confined, however, to the higher mountains, at 1400 to 3000 meters, and mostly at 2000 meters or more. P. oleifolius may be distinguished at once from P. Alleni by the fact that the costa is impressed rather than elevated on the upper leaf surface. So far as may be judged from leaf char- acters, P. АЦепи is more closely related to P. guatemalensis Standl., of Guatemala and British Honduras, which also has the costa prominent upon the upper surface. P. guate- malensis, likewise, occurs at low elevations, descending to sea- level. 1941] WOODSON & SCHERY—FLORA OF PANAMA. V 411 It is possible that ultimately P. Allenw will have to be com- bined with P. guatemalensis, but it is to be expected that flowers and fruit may furnish better characters for separating it. The adult leaves of the Panama tree are so much smaller than those of the northern one that it is believed by the writer that two distinct species are involved. The seedling leaves of P. Allen are lance-linear, 9-12 em. long, and 9-14 mm. wide. Mr. Allen states that the local name of this tree is Cuahau. ERIOCAULACEAE (H. N. Moldenke, New York) Tonina FLUVIATILIS Aubl.—cuririqui: vicinity of Boquete, alt. 1200-1500 m., July 24—26, 1940, Woodson & Schery 752. Known to occur from Cuba and British Honduras to Colombia, Trinidad and Brazil, but has not previously been collected in Panama. BROMELIACEAE (L. B. Smith, Cambridge, Mass.) ArcHMEA Allenii L. B. Smith, spec. nov. (pl. 19), acaulis, stolonibus robustissimis procreans; foliis subfasciculatis, erectis, 6-7 dm. longis, inflorescentiam multo superantibus, utrinque lepidibus minutis peradpressis obtectis, vaginis mag- nis, anguste ellipticis, integris, quam laminis haud atrioribus, laminis ligulatis, late acutis et mucronatis, basi nullo modo angustatis, ad 59 mm. latis, dense serrulatis, viridibus, con- coloribus ; seapo gracili, ad 35 em. alto, albido-flocculoso ; scapi bracteis magnis, erectis, dense imbricatis, oblanceolatis, acutis, membranaceis, pulchre roseis, ad apicem versus denticulatis, basi flocculosis ; inflorescentia erecta, simplicissima, densissime ellipsoidea, 8-10 cm. longa, 4 em. diametro; bracteis florigeris erectis, inferioribus eis scapi similibus et flores multo super- antibus, supremis anguste lanceolatis, acuminatis, integris, sepala aequantibus vel quam eis paulo brevioribus; floribus sessilibus ; sepalis liberis, valde asymmetricis, mucro erecto ad 6 mm. longo incluso ad 23 mm. longis, eum ovario dense albido- lepidotis; petalis delapsis solum cognitis, ca. 3 em. longis, basi ligulis binis dentatis auctis, albis vel pallide lilacinis |! Allen]; ovario orbieulato, valde compresso et triangulato, ad 17 mm. [Vor. 28 412 ANNALS OF THE MISSOURI BOTANICAL GARDEN longo, tubo epigyno crateriformi, placentis interno loculorum angulo prope apicem affixis, ovulis caudatis.—cocrÉ: epi- phytie, vicinity of La Mesa, north of El Valle de Antón, alt. 1000 m., May 12, 1941, P. H. Allen 2378 (U. S. Nat. Herb., TYPE; Gray Herb. and Herb. Missouri Bot. Garden, photograph); trail to Las Minas, north of El Valle de Antón, alt. 1000 m., May 10, 1941, P. H. Allen 2467 (U. S. Nat. Herb.). The very showy rose scape-bracts of Aechmea Allenu are suggestive of 4e. Mariae-Reginae, but the large long-muero- nate sepals and the large serrulate lower floral braets immedi- ately set it apart from this and all related species. ArcHMEA Tonpuzit Мег & Pittier—panrEN: epiphytic, rain forest, Cana-Cuasi Trail, near Camp 2, Chepigana District, alt. 1650 m., March 12, 1940, M. E. € R. A. Terry 1529 (Herb. Field Mus.); epiphytie, rain forest, Cana-Cuasi Trail near crest, Chepigana District, alt. 1500 m., March 15, 1940, M. E. € R. A. Terry 1592 (Herb. Field Mus.). Previously known from Costa Rica; identified by description. АвснмЕА Увштсни Baker—panrEN : rain forest, north slope of Cana-Cuasi Trail, Real District, alt. 1650 m., March 13, 1940, M. Е. € В. A. Terry 1545 (Herb. Field Mus.). Previously known from Costa Riea and Colombia. Catopsis \УлмаЕвтхт Mez & Werceklé—cmimiquí: Finca Lérida to Pena Blanca, alt. 1750—2000 m., July 9, 1940, Wood- son Ё Schery 319. Previously known from Costa Rica. Guzmania DowwNELLsMITHII Mez ех Donn. Smith—cocrí: epiphytic, vicinity of La Mesa, north of El Valle de Antón, alt. 1000 m., May 12, 1941, P. H. Allen 2382 (U. S. Nat. Herb.). Previously known from Costa Rica. GuzMANIA GLOMERATA Mez & Werceklé—cocréí: epiphytic, north rim of El Valle de Antón near Cerro Turega, alt. 650— 700 m., June 30, 1940, Woodson & Schery 199; DARIEN: epi- phytic, Cana-Cuasi Trail, Chepigana District, alt. 1200 m., March 12, 1940, M. E. € В. A. Terry 1526 (Herb. Field Mus.) ; same (Camp 2), alt. 600 m., M. E. Ф В. A. Terry 1528 (Herb. Field Mus.). Previously known from Costa Rica ; identified by deseription. 1941] WOODSON & SCHERY—FLORA OF PANAMA. V 413 GUZMANIA GUATEMALENSIS L. B. Smith—cocrí: epiphytic, vicinity of La Mesa, north of El Valle de Antón, alt. 1000 m., May 12, 1941, Р.Н. Allen 2368 (U.S. Nat. Herb.) ; DARIEN : epi- phytic, rain forest, Cana-Cuasi Trail near crest, Chepigana District, alt. 1500 m., March 15, 1940, M. E. € R. A. Terry 1589 (Herb. Field Mus.). Previously known from Guatemala and Colombia. Guzmania MUSAICA (Linden) Mez—parien: epiphytic, rain forest, crest, Cana-Cuasi Trail, Real District, alt. 1650 m., March 16, 1940, M. E. & В. A. Terry 1608 (Herb. Field Mus.). Previously known from Colombia. РттсатвмтА VarEnu Standley—panami: terrestrial, Cerro Campana, alt. 1000 m., April 21, 1941, P. H. Allen 2414 (U.S. Nat. Herb.). Previously known from Costa Rica. THECOPHYLLUM ACUMINATUM L. B. Smith—panamA: epi- phytie, summit of Cerro Campana, alt. 800-1000 m., Sept. 1, 1940, P. H. Allen 2215 (Herb. Field Mus.). Previously known from Costa Rica. THECOPHYLLUM CRASSIFLORUM Мех & Werceklé—onrimiqví: terrestrial, Potrero Muleto to summit, Volcán de Chiriquí, alt. 3500-4000 m., July 13-15, 1940, Woodson & Schery 453. Previ- ously known from Costa Rica. THECOPHYLLUM ORORIENSE (Mez) Mez—ocumiqví: Finca Lérida to Peña Blanca, alt. 1750-2000 m., July 9, 1940, Wood- son & Schery 321. Previously known from Costa Rica. Bracts green with orange tips instead of yellow and scarlet, but the plant otherwise close to the type. TiLLANDsIA ÁcosrAE Mez & Tonduz—cuiiqví: epiphytic, vicinity of Puerto Armuelles, alt. 0-75 m., July 28-31, 1940, Woodson & Schery 823. Previously known from Costa Rica. TILLANDSIA COMPLANATA Benth.—cutrigui: epiphytic, vi- cinity of Bajo Mona and Quebrada Chiquero, alt. 1500 m., July 18, 1940, Woodson & Schery 602. Previously known from Costa Rica, the West Indies and northwestern South America. TiLLANDSIA INCURVA Griseb.—cHIRIQUi: epiphytic, vicinity of Finca Lérida, alt. 1750 m., July 7-11, 1940, Woodson d [Vor. 28 414 ANNALS OF THE MISSOURI BOTANICAL GARDEN Schery 223. Previously known from Costa Rica, southern Florida, Cuba, Jamaica, Venezuela, Colombia and Bolivia. VnrESIA РАСНУЗРАТНА Mez & Werceklé—cnmiqví: epiphytic, vicinity of Bajo Chorro, alt. 1900 m., July 20-22, 1940, Wood- som & Schery 679. Previously known from Costa Rica; identified from description. LILIACEAE ANTHERICUM APODASTANTHUM Donn. Sm.—cuiriqgui: Po- trero Muleto, Volean de Chiriqui, alt. 3500 m., July 13, 1940, Woodson & Schery 379. Previously known from Guatemala and Costa Rica. ANTHERICUM MACROPHYLLUM Baker—cocrÉ: north rim of El Valle de Antón, near Cerro Turega, alt. 650—700 m., June 30, 1940, Woodson, Allen & Schery 178. Previously known from Costa Rica, This is probably the same plant recently described as Anthericum panamense Standl. (Field Mus. Publ. Bot. Ser. 22: 327. 1940), based on a collection (Allen 2195) from the same locality. Our plants agree satisfactorily with Baker’s original description. AMARYLLIDACEAE ZEPHYRANTHES CITRINA Baker—socas DEL TORO: Isla Colon, alt. 0-120 m., Aug. 25, 1940, H. von Wedel 543. Previously known from Tropical America according to Baker, with no lo- са Шу given. Type supposedly came from British Guiana. The species was described from a living specimen sent to England and cultivated there. It has since been introduced to Florida gardens. ZEPHYRANTHES ROSEA Lindl.—n0cas DEL TORO: Isla Colon, alt. 0-120 m., Aug. 25, 1940, Н. von Wedel 544. Previously known from Cuba and described from a horticultural specimen brought to England. The species has since been introduced to Florida. ZEPHYRANTHES TUBISPATHA Herb.—socas DEL токо: Isla Colon, alt. 0-120 m., Aug. 25, 1940, H. von Wedel 542. Previ- ously known from Argentina? and Jamaica according to the Bot. Mag. A specimen, supposedly this species, from Tobago 1941] WOODSON & SCHERY—FLORA OF PANAMA. V 415 is in the herbarium of the Missouri Botanical Garden. The species has been introduced into Florida where it is cultivated. DIOSCOREACEAE (C. V. Morton, Washington, D. C.) DroscongA SrANpLEYyri Morton—cuirigui: vicinity of Bajo Chorro, alt. 1900 m., July 21, 1940, Woodson & Schery 664, 652. Previously known from Costa Riea. MARANTACEAE CrENANTHE DASYCARPA (Donn. Sm.) К. Sch.—socas DEL toro: Isla Colon, Aug. 14, 1940, H. von Wedel 433. Previously known only from Costa Rica. ORCHIDACEAE (Louis O. Williams, Cambridge, Mass.) Paumorcuis trilobulata L. О. Williams, sp. nov. (pl. 20). Herbae terrestres, caespitosae, usque ad 35 cm. altae. Folia subchartacea, plicata, elliptiea, acuta vel acuminata. Inflores- centia terminalis, pauciflora ; bracteae ellipticae vel lanceolato- ovatae, acutae vel acuminatae. Sepalum dorsale spathulato- oblaneeolatum, obtusum, tri- vel quinquenervium. Sepala lateralia spathulato-oblanceolata, obtusa, arcuata, tri- vel quinquenervia. Petala lineari-oblanceolata, obtusa, leviter arcuata, trinervia. Labellum oblanceolato-flabellatum, tri- lobulatum; lobi laterales rotundati, obtusi; lobus medius transverse oblongus, lobos laterales paulo excedens. Columna generis. Caespitose, terrestrial herbs up to 35 cm. tall. Stems up to about 15 em. long and 4 mm. in diameter, bearing 3—4 leaves near the apex. Leaves subchartaceous, plicate; lamina of the mature leaves 10-22 em. long and 3-6 em. broad, elliptie, acute or acuminate, with 7 prominent nerves; petiole up to 7 em. long, plieate, nervose. Inflorescence up to 7 cm. long, terminal, much exceeded by the leaves, few-flowered; rachis about 34 em. long; bracts 1-1.5 em. long and 1.5-5 mm. broad, elliptic to lanceolate-ovate, acute or acuminate. Flowers about 13 mm. long, white. Dorsal sepal about 12 mm. long and 3 mm. broad, spatulate-oblanceolate, obtuse, 3—5-nerved. Lateral sepals 10- [Vor. 28 416 ANNALS OF THE MISSOURI BOTANICAL GARDEN 12 mm. long and 3-4 mm. broad, areuate, spatulate-oblanceo- late, obtuse, 3-5-nerved, somewhat carinate along the mid- nerve dorsally. Petals 11-12 mm. long and about 2 mm. broad, linear-oblanceolate, obtuse, somewhat arcuate, 3-nerved. Lip 10-12 mm. long and about 5-6 mm. broad near the apex, ob- lanceolate-flabellate, trilobulate at the apex, the base or claw adnate to the column along its median line for about 2-3 mm., more or less enfolding the column; dise of the lip pubescent, especially so toward the base and less so toward the apex, fleshy-thickened toward the apex; lateral lobes about 1 mm. long (free portion), rounded, obtuse; mid-lobe about 1.5 mm. long, transversely oblong, blunt, slightly exceeding the lateral lobes. Column slender, arcuate, about 10 mm. long; anther operculate, incumbent; pollinia 4, exappendiculate, ceraceous or appearing so, pyriform.—cocLÉ: terrestrial, flowers pure white; region north of El Valle de Antón, trail to Las Minas, alt. about 1000 m., May 10, 1941, Allen 2463 (Herb. Ames, TYPE) ; terrestrial, flowers white; El Valle de Antón and vicin- ity, alt. 500—700 m., July 23-27, 1935, Seibert 459 (Herb. Mis- souri Bot. Gard., Herb. Field Museum, Herb. Ames). Palmorchis trilobulata is not easily confused with any previ- ously described species of the genus. It is, perhaps, most closely allied to P. Powellii (Ames) Schweinf. & Correll, from which it is easily distinguished both vegetatively and by the size of the flowers, and especially by the very different lip. Pevexia Funckiana (Rich. & Gal.) Sehltr.—cocrí: terres- trial, flowers green; vicinity of La Mesa, hills north of El Valle de Antón, alt. about 1000 m., Jan. 21, 1941, Allen 2315. Previ- ously known from Mexico, British Honduras, Guatemala and Costa Rica, possibly also in South America. ACOSTAEA COSTARICENSIS Schltr.—coc.#: epiphytic, flowers tan; hills north of El Valle de Antón, alt. about 1000 m., Jan. 21, 1941, Allen 2352. The genus is new to Panama, previously known from Costa Rica. PLEUROTHALLIS CoNCAVIFLORA С. Sehweinf.—cocrí: epi- phytic ; vicinity of El Valle de Antón, alt. about 600 m., Dec. 10, 1939, Allen 2058; epiphytic; vicinity of La Mesa, hills north of 1941] WOODSON & SCHERY—FLORA OF PANAMA. V 417 El Valle de Antón, alt. about 1000 m., Jan. 21, 1941, Allen 2317. Previously known from Costa Rica. PLEUROTHALLIS macrantha L. О. Williams, sp. nov. (pl. 21, figs. 1-6). Herbae caespitosae, epiphyticae vel terrestres, usque ad 23 em. altae. Folia elliptica, obtusa, coriacea. Inflo- rescentia racema pluriflora quam folia duplo longior. Sepalum dorsale ellipticum vel oblanceolato-elliptieum, acutum. Sepala lateralia plusminusve usque ad apiees connata, oblongo- oblanceolata, acuta. Petala oblongo-obovata, leviter obliqua, truneata vel paulo retusa. Labellum subteres et supra sul- catum, prope apicem expansum et basi biauriculatum et tri- eallosum. Columna generis. Caespitose, epiphytie or terrestrial herbs up to 23 cm. tall. Secondary stems 6-8 em. long, subequal to the leaves in length on mature plants, almost completely invested with 1 or 2 loose, scarious sheaths, unifoliate. Leaves 4-7 ст. long and 1.5-2.8 em. broad, elliptic, obtuse, coriaceous. Inflorescence a several- flowered raceme, about twice exceeding the subtending leaves; sheath at the base of the peduncle short, about 9-11 mm. long; bracts 6-7 mm. long, ovate-lanceolate, scarious, semiamplexi- caul. Dorsal sepal 16-20 mm. long and 6-8 mm. broad, elliptic to oblanceolate-elliptic, acute, 3-nerved, keeled dorsally along the mid-nerve. Lateral sepals 16-18 mm. long and together 6-8 mm. broad, connate almost to their apices, oblong-oblance- olate, acute, together 4-nerved, somewhat gibbous at the base. Petals about 9-11 mm. long and 5-6 mm. broad near the apex, oblong-obovate, slightly oblique, truncate or slightly retuse, 3-nerved, with an inconspicuous cushion-like callus near the base. Lip 10-12 mm. long and 3-4 mm. broad toward the apex when expanded, about 2 mm. broad near the middle and 3 mm. broad across the auriculate base, subterete and suleate above, somewhat laterally compressed, expanding toward the apex and into 2 auricles at the base, with 3 small calluses between the basal auricles. Column of Ше genus.—cniniQuí: terrestrial, flowers pale greenish-yellow, spotted chocolate; vieinity of Bajo Chorro, alt. 1900 m., July 20-22, 1940, Woodson & Schery 666 (Herb. Ames, TYPE). [Vor. 28 418 ANNALS OF THE MISSOURI BOTANICAL GARDEN Pleurothallis macrantha seems to be most closely allied to P. pachyglossa Lindl., with which it perhaps shares the distine- tion of having the largest flowers of the genus. Pleurothallis macrantha is distinguished from P. pachyglossa by the trun- cate petals, the laterally compressed lip, which differs in sev- eral characters, and by the details of the sepals and clin- andrium. PLEUROTHALLIS MURICATA Schltr. (P. diuturna Schltr.)— cocLÉ: epiphytic on Calabash tree; El Valle de Antón, alt. about 800 m., Jan. 20, 1941, Allen 2294. New to Panama under this name. Maraxis SIMILLIMA (Reichb. f.) О. Kuntze—currieui: ter- restrial ; vicinity of Bajo Chorro, alt. 1900 m., July 20-22, 1940, Woodson & Schery 685. New to Panama, previously known from Costa Rica. Epipenprum ($ Evepmenprum) Allenii L. О. Williams, sp. nov. (pl. 22, figs. 3-5). Herbae epiphyticae, erectae, usque ad 3 dm. altae. Caules ancipiti. Folia elliptica vel lanceolata, acuta vel breviter acuminata. Sepalum dorsale lanceolatum, obtusum. Sepala lateralia sepalo dorsali similia sed leviter arcuata. Petala linearia, uninervia. Labellum unguiculatum; lamina subreniformis, subintegra vel obscure trilobata. Col- umna sectionis. Erect epiphytic herbs up to about 3 dm. tall. Stems ancipi- tous, covered with the bases of the 3 to 6 leaves. Leaves 4-12 em. long, 1.2-3 em. broad, elliptic to lanceolate, acute or short- acuminate. Peduncle about 10-15 em. long, ancipitous, bialate, the wings prominent at the base but disappearing at about the middle of the peduncle. Inflorescence short, the rachis covered with the scarious, imbricated bracts; bracts up to about 1 em. long. Dorsal sepal about 10 mm. long and 4 mm. broad, lanceolate, obtuse, 5—7-nerved, with a short dorsal apicule near the apex. Lateral sepals similar to the dorsal sepal except somewhat arcuate. Petals about 10 mm. long and 0.75 mm. broad, linear, 1-nerved. Lip unguiculate; the lamina about 10 mm. long and 12 mm. broad, subreniform, subentire or obscurely 3-lobed, the terminal lobe inconspicuous and slightly 1941] WOODSON в SCHERY—FLORA OF PANAMA. У 419 retuse, with two approximate submammillar calluses at the base. Column of the section, about 7 mm. long.—cocrÉ: epi- phytie, leaves and stem red, sepals and petals pale pink, labellum white; hills north of El Valle de Antón, about 1000 m. alt., July 14, 1940, Allen 2203 (Herb. Ames, 58635, TYPE). Epidendrum Allenii seems not to be closely allied to any Central American species of Epidendrum, perhaps nearest to E. panamense Schltr. and to E. Lankesteri Ames (which may be synonyms), but differs from both of these by the very large differently shaped lip, by the wing on the peduncle and by other details. EPIDENDRUM cRASsILABIUM Poepp. & Endl.—cocré: Allen 2205; Woodson & Schery 204. A not uncommon species but ap- parently new to Panama. Previously known from Costa Rica, the West Indies, Brazil, Duteh Guiana, British Guiana, Vene- zuela, Colombia, Ecuador and Peru. EPIDENDRUM DIFFORME Jacq. var. FIRMUM (Reichb. f.) AHS. —cuirigui: epiphyte; Finca Lérida to Peña Blanca, alt. 1750- 2000 m., July 9, 1940, Woodson Ё Schery 298. This variety has been reported previously from Guatemala and Costa Rica. Eprenprum ($ Еоврпхъвом) ellipsophyllum Г. О. Wil- liams, sp. now. (pl. 22, figs. 1-2). Herbae epiphyticae usque ad 25 em. altae. Caules leviter fusiformes et indurati. Folia el- liptica, breviter acuminata. Sepala elliptica vel oblanceolata, obtusa. Petala oblanceolata, obtusa. Labellum unguiculatum ; unguis columnae adnata; lamina orbiculari-ovata, trilobata, lamellata, lobi laterales parvi. Columna sectionis. Erect epiphytic herbs up to about 25 cm. tall. Stems slightly fusiform and indurated, with only the annual leaves persisting and the sheaths of the older leaves becoming scarious and dis- integrating. Leaves 11-12 em. long and 2-2.8 cm. broad, elliptie, short-acuminate, usually only 2 or 3 persisting. Pedun- cles up to about 4 em. long, ancipitous, generally covered with imbricated bracts. Inflorescence short, 1-, or at most, few- flowered; bracts up to about 2 em. long, cucullate, lanceolate- triangular, acute. Flowers green, segments stiff and eori- aceous. Sepals 10-12 mm. long and 3-4 mm. broad, elliptic to [Vor. 28 420 ANNALS OF THE MISSOURI BOTANICAL GARDEN oblanceolate, obtuse, 5-nerved. Petals about 10-11 mm. long and 3-4 mm. broad, oblanceolate, obtuse, with 3 principal nerves. Lip unguiculate, the claw mostly adnate to the column ; lamina 5-6 mm. long and 5-6 mm. broad, orbicular-ovate in out- line, 3-lobed, the lateral lobes small and situated toward the base of the lip, with a 2-ridged lamellate callus plate at the junc- tion of the claw and the lamina. Column slender, about 8 mm. long; clinandrium entire and obtuse; pollinia not seen.— cocLÉ: epiphytic, flowers green; hills north of El Valle de Antón, 1000 m. alt., June 23, 1940, Allen 2178 (Herb. Ames, 98640, TYPE). Epidendrum ellipsophyllum does not seem to be closely allied to any other Central American species but is perhaps nearest to E. Lankesteri Ames. Epwenprum ($ Коврпвховом) gibbosum L. О. Williams, sp. nov. (pl. 21, figs. 7-10). Herbae epiphyticae vel terrestres usque ad 4 dm. altae. Folia lineari-oblonga vel lineari-lance- olata, obtusa, eoriacea. Inflorescentia brevis, racemosa, pauci- flora. Sepalum dorsale elliptico-oblongum, obtusum. Sepala lateralia oblongo-ovata, obtusa. Petala elliptiea vel oblongo- oblanceolata, obtusa. Labellum late oblongum vel suborbicu- lare, obtusum et leviter emarginatum, basi cordatum, car- nosum, leviter gibbosum. Columna sectionis. Kpiphytic or terrestrial herbs up to 4 dm. tall. Stems simple or branched below, covered with the ample leaf-sheaths. Leaves 9-14 em. long and 15-18 mm. broad, linear-oblong to linear-lanceolate, obtuse and unequally bilobed, articulate, coriaceous. Leaf-sheaths ample, somewhat inflated, disinte- grating with age. Inflorescence short, racemose, few-flowered ; braets 10-18 mm. long and 12-22 mm. broad, cochleate, acute, scarious. Dorsal вера! 10-12 mm. long and 5-6 mm. broad, el- liptic-oblong, obtuse, apiculate, 5—7-nerved. Lateral sepals 10-12 mm. long and 6-7 mm. broad, oblong-ovate, obtuse, apicu- late dorsally, 7-nerved. Petals about 10 mm. long and 5 mm. broad, elliptie to oblong-oblanceolate, obtuse, 5-7-nerved. Lip about 11-12 mm. long and 9-10 mm. broad, broadly oblong or suborbieular, obtuse or shallowly emarginate, cordate at the 1941] WOODSON & SCHERY—FLORA OF PANAMA. V 421 base, fleshy, especially tumid along the median nerve toward the apex, with 2 inconspieuous calluses at the base, basal por- tion somewhat gibbous and partly enfolding the column. Column of the section, about 4 mm. long.—cnuiniQví: terres- trial, lowers maroon; Casita Alta to Cerro Copete, alt. 2300- 3300 m., July 10, 1940, Woodson Ф Schery 367 (Herb. Missouri Bot. Gard., TYPE). Epidendrum gibbosum is most closely allied to E. barbae Reichb. f., a species native of Costa Riea, from which it may be distinguished by the shape and base of the lip and the dif- ferent callus thickenings. The leaves of E. gibbosum are quite obtuse and unequally bilobed, while those of E. barbae are more or less acute and not unequally bilobed. The column of E. gibbosum is situated well down within the gibbous base of thelip while that of E. barbaeisnot. Although the specimen of this species was terrestrial, it is a potential epiphyte. Epmenprum ($ Evepmenprum) pendens L. О. Williams, вр. nov. (pl. 23). Herbae epiphyticae, pendentes, usque ad 45 em. longae. Folia elliptica vel elliptico-lanceolata, coriacea, disticha, marginata. Inflorescentia uni-pauciflora. Sepalum dorsale oblongo-oblaneeolatum, acutum, carnosum. Sepala lateralia elliptiea, acuta, carnosa. Petala linearia, obtusa vel acuta, carnosa. Lamina labelli trilobulata, suborbicularis, nervo medio erassato et nervis radiantibus crassatis ornata; lobi laterales rotundati; lobus medius subquadratus, emargi- natus. Columna generis. Pendent, epiphytic herbs about 45 em. long. Leaves 9-15 em. long and 2-3.5 em. broad, elliptic to elliptic-lanceolate, coriaceous, distichous, marginate, gray, deciduous; leaf- sheaths loose, completely investing the stem, pergameneous, ancipitous, the abscission line prominent. Inflorescence of 1 or few large, terminal flowers subtended by a leaf-like sheath up to 3 em. long. Flowers large for the genus, green. Dorsal sepal about 5 em. long and 1.2 em. broad, oblong-oblanceolate, acute, fleshy, with 5—7 main nerves. Lateral sepals 4.3—4.8 em. long and 1.5 em. broad, elliptic, acute, fleshy, with about 7 main nerves. Petals about 4.5-5 cm. long and 2 mm. broad, [Vor. 28 422 ANNALS OF THE MISSOURI BOTANICAL GARDEN linear, somewhat narrowed to the base, obtuse or acute, fleshy, l-nerved. Lip with the elaw adnate with the column to the apex of the column; lamina about 2.5-3 em. long and 3 em. broad, 3-lobed, suborbicular in general outline, with a thick- ened central nerve and with thickened nerves or ridges radiat- ing from it and covering the remainder of the lip, lateral lobes rounded, about 1.5 em. long, mid-lobe 1-1.5 em. long and about 1.5 em. broad, subquadrate, emarginate. Column 1.5-2 em. long and 6-8 mm. in diameter at the apex, fleshy.—cocrf: epiphytic, pendent plants, foliage gray, flower terminal, green; vicinity of La Mesa, El Valle de Antón, alt. about 1000 m., June 22, 1941, Allen 2570 (Herb. Ames, TYPE). Epidendrum pendens is most closely allied to E. Medusae (Reichb. f.) Pfitz., a rare plant of the Ecuadorian Andes, from which it is easily distinguished by the 3-lobed, nonfimbriated lip. If the genus Nanodes Lindl. were maintained this species would be the third one known. Mr. Allen's note (in lit.) is of considerable interest: “Ері- phytic, entire plant directly pendent, foliage distichous, stiff, gray, flowers terminal, green. This is one of the most remark- able orchids I have ever seen. The foliage is a dead, almost lead-gray, although the plant was the picture of perfect health. The flower is the sort of thing that you would expect a child to construet, after being told in a general way what the floral pat- tern of an orchid should be." The description is from a dried plant and a single flower pre- served in spirits. Крпкховом ramosum Jacq. var. angustifolium (Cogn.) L. О. Williams, comb. nov. (E. imbricatum Lindl. var. angusti- folium Cogn. in Mart. Fl. Bras. 3*: 171. 1898, as var. angusti- folia; E. imbricatum Lindl. Gen. & Sp. Orch. Pl. 110. 1831, non Lam. ; E. ramosum Jacq. var. imbricatum (Lindl.) Ames, Hub- bard & Schweinf. in Bot. Mus. Leafl. Harv. Univ. 2: 47. 1934). —BOCAS DEL TORO: epiphytic, Río Crieamola, between Finca St. Louis and Konkintoé, alt. 10-15 m., Aug. 12-16, 1938, Wood- son, Allen & Seibert 1898. Cogniaux’ varietal name is correct 1941] WOODSON & SCHERY—FLORA OF PANAMA. V 423 for this unit when it is maintained as a variety. The specimen cited above is the only one from Panama. ScarHvaLorris Lindeniana (Rich. & Gal.) L. О. Williams, comb. nov. (Hexadesmia fasciculata Brongn. in Ann. Sci. Nat. Bot. П, 17: 44. 1842, non Scaphyglottis fasciculata Hook. ; H. Lindeniana Rich. & Gal. in Ann. Sci. Nat. Bot. IIT, 3: 23. 1845; H. rhodoglossa Reichb. f. in Bonplandia 4: 326. 1856; H. pachybulbon Schltr. in Fedde Repert. Beih. 17: 26. 1922; H. falcata C. Schweinf. in Bot. Mus. Leafl. Harv. Univ. 8: 45, pl. 1940) --сосъй: epiphyte, flowers light green; hills north of El Valle de Antón, alt. 800-1000 m., Jan. 2, 1941, Allen 2291. An extremely variable species as to flower size, the present specimen having very small flowers. Previously known from Mexico to Panama. ©$сАрРнүаготтїв ($ HEgxapEsMiA) tenella L. О. Williams, sp. nov. (pl. 24). Herbae epiphyticae, tenellae, caespitosae, usque ad 5 dm. altae. Caules elongati, tenelli, sub apicibus fusi- formes. Folia linearia vel lineari-lanceolata. Sepala lance- olata, acuta. Petala oblongo-lanceolata, acuta vel obtusa. Labellum rhombico-lanceolatum. Columna sectionis. Slender, caespitose, epiphytic herbs up to about 5 dm. tall. Stems elongated, slender, becoming naked and in age swollen and fusiform below the summit, bifoliate. Leaves 20-23 cm. long and 7-16 mm. broad, linear to linear-lanceolate, acute or the apex minutely bilobulate. Inflorescence terminal in the axis of leaves, a fascicle of 1 or few flowers. Dorsal sepal 8-9 mm. long and 3-3.5 mm. broad, lanceolate, acute, 5-nerved. Lateral sepals 8-9 mm. long and 3-4 mm. broad, lanceolate, acute, 5—-7-nerved. Petals 8-9 mm. long, 3-3.5 mm. broad, ob- long-lanceolate, acute or obtuse, 3- or 5-nerved. Lip about 8-9 mm. long and 4—5 mm. broad, rhombic-lanceolate but somewhat eonstrieted above the middle, fleshy. Column of the section, about 5 mm. long; pollinia 6, жаху. —сосіё: epiphytic; hills north of El Valle de Antón, alt. 800-1000 m., Nov. 21, 1940, Allen 2276 (Herb. Ames, TYPE). Scaphyglottis tenella is allied to S. pulchella (Schltr.) L. [Vor. 28 424 ANNALS OF THE MISSOURI BOTANICAL GARDEN Wms.! but is easily distinguished by the smaller flowers and differently shaped lip. ZYGOPETALUM parviflorum L. О. Williams, sp. nov. (pl. 25). Herbae robustae, terrestres vel epiphyticae, usque ad 1 m. altae. Folia lineari-elliptica vel elliptica, acuta vel acuminata, plieata. Seapus bracteosus; inflorescentia racemosa, pluri- multi-flora; bracteae lineares vel lineari-lanceolatae, acutae. Sepalum dorsale oblongo-ovatum, acutum, earnosum. Sepala lateralia oblongo-ovata, obtusa, carnosa. Petala oblonga, ob- tusa, leviter obliqua. Labellum oblongum, panduratum, apice rotundatum, prope basim callo lunato ornatum. Columna generis. Robust, terrestrial or epiphytie herbs up to 1 m. tall. Stems leafy, short, becoming thickened into pseudobulbs, up to 9 em. long and 1.5 em. thick. Leaves 30-70 em. long and 2.5-7 cm. broad, linear-elliptie to elliptic, acute or acuminate, plicate, at- tenuated at the base into a petiole. Scape up to 1 m. long, with several braets ; inflorescence racemose, several-many-flowered, braets 5—14 mm. long, linear to linear-lanceolate, acute. Dorsal sepal about 10 mm. long and 5 mm. broad, oblong-ovate, acute, fleshy. Lateral sepals 10-11 mm. long and 5-6 mm. broad, ob- long-ovate, obtuse, 5-(7-)nerved, fleshy. Petals about 7-8 mm. long and 4—4.5 mm. broad, oblong, obtuse, slightly oblique, 5- nerved. Lip 7-8 mm. long and 6.5-7.5 mm. broad, oblong, pan- durate, about 2 mm. broader above the middle than below the middle, the apex broadly rounded, with a transverse, lunate callus at the base, fleshy. Column of the genus.—cniiQvÍ: terrestrial, flowers purple, lip violet, column white; vicinity of Bajo Chorro, alt. 1900 m., July 20-22, 1940, Woodson & Schery 605 (Herb. Ames, түрк) ; in rain forest, Bajo Chorro, Boquete District, alt. 6000 ft., Feb. 28, 1938, Davidson 345. Zygopetalum parviflorum is not closely allied to any other species of the genus which is known to me. The flowers are much the smallest of the racemose species of the genus. The 'ScaPHYGLOTTIS pulchella (Sehltr.) L. О. Williams, comb nov. (Ramonia pul- chella Schltr. in Fedde Repert. Beih. 19: 295. 1923). 1941] WOODSON & SCHERY—FLORA OF PANAMA. V 425 present species resembles a Еигудоре ит in all respects ex- cept the small size of the flowers. MaxiLLARiA Biolleyi (Schitr.) L. О. Williams, comb. nov. (Ornithidium Biolleyi Schltr. in Fedde's Repert. 9: 29. 1910). —CHIRIQUI: epiphyte in rain forest; Bajo Chorro, Boquete District, alt. about 2000 m., Jan. 14, 1938, Davidson 125. Known in Costa Rica and Panama. MaxiLLaniA Bradeorum (Schitr.) L. О. Williams, comb. nov. (Camaridium Bradeorum Schltr. in Fedde's Repert. Beih. 19: 141. 1923).—cnigiQví: epiphyte in rain forest; Bajo Chorro, Boquete District, alt. about 2000 m., Jan. 13, 1938, Davidson 118. Known in Costa Rica and Panama. МахплавтА fulgens (Reichb. f.) L. О. Williams, comb. nov. (Ornithidium fulgens Reichb. f. Beitr. Orch. Cent. Am., p. 76. 1866).—cocrÉ: epiphytic; hills north of El Valle de Antón, alt. 800-1000 m., Nov. 21, 1940, Allen 2256. Previously recorded from Costa Rica. Maxi.LARiA umbratilis L. О. Williams, nom. nov. (Cama- ridium nutantiflorum Schltr. in Beih. Bot. Centralbl. 36, Abt. 2: 417. 1918, non Mazillaria nutantiflora Schltr.).—cureiqui: epiphyte in rain forest; Bajo Chorro, Boquete District, alt. about 2000 m., Feb. 18, 1938, Davidson 308. Known in Costa Rica and Panama. ONCIDIUM PANDURIFORME Ames & Schweinf.—cnriRiQUÍ: epi- phyte; vicinity of Bajo Chorro, alt. 1900 m., July 20-22, 1940, Woodson & Schery 704. Previously recorded from Costa Rica. Гвоснил$ овлсплѕ Sehltr.—cocrÉ: epiphytic, growing on Calabash, sepals and petals yellow, striped reddish-brown, lip yellow with spot of red at base, pseudobulbs and leaves dark bronzy red; El Valle de Anton, alt. about 800 m., Jan. 20, 1941, Allen 2295. Previously recorded from Honduras and Costa Rica. CaMPYLOCENTRUM Ввкхеви Schltr.—cuirigui: epiphyte; vi- cinity of Bajo Mona and Quebrada Chiquero, alt. 1500 m., July 18, 1940, Woodson Ф Schery 579. New to Panama, previously known from Costa Rica. [Vor. 28 426 ANNALS OF THE MISSOURI BOTANICAL GARDEN PIPERACEAE (William Trelease, Urbana, Ш.) РЕРЕЕОМтА Scheryi Trel., spec. nov. Herba terrestris sat magna glabra; eaulibus inferne crassiusculis superne gracil- ibus; foliis alternatis peltatis rotundato-ovatis breviter acute- que acuminatis basi aliquantulum concavis са. 13 em. longis 8.5 em. latis, petiolis gracilibus 5-7 em. longis; spicis solitariis terminalibus ca. 100 mm. longis 5 mm. erassis, pedunculo gra- cili ca. 2 em. longo ad apicem in ramulo 2-bracteato ca. 3 em. longo gesto; baccis elongato-ovoideis rostratis, stigmate ad basim in rostro anteriori gesto.—cHrrigui: Quebrada Velo, vicinity of Finca Lérida, alt. ca. 1800 m., July 8, 1940, Woodson & Schery 247 (Herb. Univ. Illinois, TYPE). Ртрев fagopyricarpum Trel., spec. nov. Frutex parvus glaber; internodiis elongatis gracillimis; foliis lanceolato- ellipticis apice acute caudatis basi acutis ca. 12 em. longis 4 cm. latis omnino pinnate nervatis, nervis utraque latere ca. 10, post exsiccationem firme chartaceis, petiolis 0.5 em. longis; spicis ca. 40 mm. longis 5 mm. crassis, pedunculo gracili 10 mm. longo; baccis suleate ovoideis apiculatis, stigmatibus parvis.— CHIRIQUÍ: vicinity of Puerto Armuelles, alt. 0-75 m., July 28- 31, 1940, Woodson & Schery 852 (Herb. Univ. Illinois, TYPE). LORANTHACEAE Psirracantuus Scheryi Woodson, spec. nov. ( Eupsittacan- thus Ligulati). Frutex parasiticus omnino glaber ; ramis ramu- lisque erassis teretibus ad nodos haud incrassatis continuis. Folia opposita vel subopposita breviter petiolata inaequilat- eraliter rhombo-ovata apice late obtusa vel subrotundata basi latissime cuneata petiolo brevissimo incluso 9-11 em. longa 6.0-9.5 em. lata valde coriacea olivacea nervo medio immerso nervis lateralibus penninerviis et vix sensis. Inflorescentiae laterales fasciculatae congestae; pedunculis foliorum petiolos subaequantibus; pedicellis ea. 0.5 em. longis. Flores gemini inferne aurantiaei superne lutei; eupula prorae simili sub- membranacea margine integro ca. 0.2 еш. longa 0.1 еш. lata; ealyeulo urceolato са. 0.25 em. alto ostio 0.2 em. diam. obconico 1941] WOODSON & SCHERY—FLORA OF PANAMA. V 497 margine integro; perigonio gracili maturitate 3.2 ст. longo prope basim conspicue (ca. 0.2 em. diam.) dilatato; lobis 6 basi inconspicue incrassato-appendiculatis; antheris basifixis late oblongis 0.2 em. longis, filamentis invieem inaequilongis 2.0— 2.3 em. longis paulo supra medium perigonii adnatis; stigmate late fusiformi, stylo perigonium subaequante.—cHIRIQUI: vi- cinity of Bajo Mona and Quebrada Chiquero, alt. 1500 m., July 18, 1940, Woodson & Schery 581 (Herb. Missouri Bot. Garden, TYPE). This striking species is closely related to Ps. dilatatus A. C. Smith, of Colombia, but has much broader leaves and a smaller calyculus with entire margin. The plant from which the type specimen was obtained was a magnificent epiphytie shrub of many stout branches, several of which were nearly 6 feet in height. BALANOPHORACEAE ConvNAEA crassa Hook f.—cutrigui: thick bamboo thicket, below Potrero Muleto, Volean de Chiriqui, alt. ca. 3500 m., July 13-15, 1940, Woodson Ё Schery 470. Previously known from Costa Rica and Colombia. Apparently parasitic on the roots of bamboo. The tuberous rhizomes attain a diameter of nearly a foot. ANNONACEAE CvwBoPETALUM lanugipetalum Schery, spec. nov. Arbor; ramis novellis brunneo-pubescentibus, internodiis са. 2 em. longis; foliis elliptico-lanceolatis subsessilibus, petiolis са. 0.1-0.2 em. longis, laminis coriaceis utrinque glabris apice angustis cuspidatisque basi fere obtusis inaequilateraliterque subrotundis 15-27 em. longis 5-7 em. latis nervis subtus promi- nentibus secundariis sat distantibus inter se laxe reticulatis prope marginem late areuatim confluentibus deinde usque marginem reticulatis. Flores solitarii subterminales; pedicello pendulo ea. 20 em. longo; sepalis minutis ca. 0.3 cm. longis; petalis interioribus earnosis cucullatis minute velutinis ses- silibus triangulari-ovatis ca. 3 em. longis 1.5 em. latis venis [Vor. 28 428 ANNALS OF THE MISSOURI BOTANICAL GARDEN profunde immersis; petalis exterioribus coriaceis subtriangu- laribus sparse puberulis 1.8-2.0 em. longis 1.0-1.6 em. latis; staminibus multis capitatis са. 0.5 em. longis.—cocr£: north rim Е Valle de Antón near Cerro Turega, alt. 650-700 m., June 30, 1940, Woodson & Schery 191a (Herb. Missouri Bot. Gard., TYPE). This species is characterized by its moderately long pendu- lous pedicels, glabrous leaves, large fleshy velvety-pubescent inner petals, and very small calyx lobes. The leaves are ellip- tic-lanceolate with very prominent mid-veins and prominent confluent lateral veins which unite near the leaf margin, giving a lobed appearance to the reticulum. The outer petals are not fleshy and are only slightly more than half as large as the inner ones. The flowers are green. Desmopsis glabrata Schery, spec. nov. Arbuscula cire. З m. alta; ramulis gracilibus novellis lentissimis pubescentibus vel glabratis, cortice rufo-bruneo, internodiis 3-6 cm. longis; foliorum laminis subcoriaceis, utrinque glabratis, ellipticis, basi cuneatis ibique in petiolum decurrentibus, apice gradatim acuminatis, acumine 5-15 mm. longo, 10-20 em. longis, 3—6.5 em. latis, costa supra leviter impresso, subtus valde prominente, nervis lateralibus utrinque 6-8 areuatim adscendentibus, circ. 5 mm. infra marginem conjunetis; petiolis glabratis asperis, supra canaliculatis, circ. 5 mm. longis ; inflorescentiis foliis op- positis vel suboppositis, 1-floris (in typo) ; pedunculis circiter 1 em. longis, prope apicem bracteam rotundo-cordatam folia- ceam gerentibus, circ. 2 ст. longis; florum pedicellis cire. 6— 8 em. longis, gracillimis, leviter pubescentibus, ad medium quorum est bractea squamiformis circ. 1 mm. longa; sepalis 3, ovato-triangularibus, 3 mm. longis latisque, crassis, extus pubescentibus, intus lepidote glabratis; petalis 6, linearibus vel lineari-lanceolatis, 13 mm. longis, basi 3—4 mm. latis, apice inflexis et obtuse acuminatis, crassis, valvatis, extus pubescent- ibus, intus lepidote glabratis, basi marginibus vulgo revolutis, staminibus subpeltatis, 1-1.2 em. longis; ovariis pilosissimis, 7-carpellatis; stylis brevibus; stigmatibus subclavato-capitatis, setaceo-pubescentibus, 2 mm. latis; fructibus ignotis.—socas 1941] WOODSON & SCHERY—FLORA OF PANAMA. V 429 DEL TORO: vicinity of Bocas del Toro, alt. near sea-level, Aug. 1, 1940, Н. Wedel 279 (Herb. Missouri Bot. Gard., TYPE). This species is evidently closely related to D. microcarpa R. W. Fries, from which it ean be distinguished by the gradu- ally acuminate apex of the leaves, the larger foliaceous bracts, the longer pedicels, and the narrow bluntly acuminate petals, densely pubescent without. The leaves are almost perfectly elliptic, entirely glabrous on both surfaces. The short peduncle bears a large foliaceous bract, from which extends the slender pedicel. Midway up the pedicel is found a small squamiferous bract. The flowers are thick and leathery, densely pubescent without and lepidote-glabrous within. The ovary, style and stigma are densely hairy. The torus is cylindric, truncate. CUNONIACEAE WziNMANNIA pseudolaurina Woodson, spec. nov. Arbor ut dieitur ca. 8 m. alta; ramulis juventate conspicue compressis minute sparseque pilosulis tandem teretibus glabris. Folia simplieia breviter petiolata late elliptica apice obtusa vel late acuta basi obtusa margine levissime crenata petiolo ea. 0.2— 0.4 em. longo incluso 5-10 em. longa coriacea nitida subtus nervo medio rare inconspicue pilosulo eaeterumque glaber- rima. Inflorescentiae fasciculatae ca. 8 em. longae, axibus mi- nute puberulis. Flores in fasciculis inter se 0.2-0.5 em. dis- tantibus aggregati; pedicellis 0.2-0.3 em. longis puberulis. Sepala 5 ovato-lanceolata acuminata са. 0.1 ст. longa apice minute barbata. Petalaignota. Stamina 10; filamentis anguste subuliformibus са. 0.2 em. longis glabris. Ovarium ovoideum ea. 0.1 em. longum pilosum. Folliculi ca. 0.3 em. longi dense pilosuli; seminibus late oblongoideis rufis ea. 0.1 em. longis pilosis.—PANAMÁ: Cerro Campana, trail from Campana to Chica, alt. 600-800 m., Aug. 10, 1941, very common along edges of grassland, P. H. Allen 2656 (Herb. Missouri Bot. Gard., TYPE). W einmannia laurina H.B.K., to which our species is closely related, bears glabrous ovaries and follicles, more lax, elongate inflorescences, and more sharply crenate-serrulate leaves. (VoL. 28 430 ANNALS OF THE MISSOURI BOTANICAL GARDEN CONNARACEAE (J. A. Steyermark, Chicago) Connarus Allenii Steyermark, sp. nov. Folia 3-5-ЕоПо аа, foliolis oblanceolatis subito breviter acuminatis, basi an- gustato-cuneatis, chartaceis, utrinque glabris, nervis later- alibus utroque latere 10-15 fere rectis; inflorescentia later- alis subracemosa, rhachidi 1.8-2 em. longa, floribus breviter pedieellatis; sepala 2.5 mm. longa subobtusa extus in statu fruetifero glabra epunctata ; petala 4—4.5 mm. longa lanceolata utrinque glabra epunctata; folliculi 2.3-2.5 em. longi lateraliter compressi oblique mucronati, basi 5 mm. longe stipitati, peri- earpio intus tomentoso. Tree 6 m. tall, bark of branches light gray, glabrous; leaves 3-5-foliolate, rachis 7-12 em. long, glabrous; leaflets 8-18 em. long, 2.2-3.5 em. broad, oblanceolate, abruptly short-acuminate at apex, narrowed-cuneate at base, chartaceous, dull above, glabrous on both sides; petiolule 2-2.5 mm. long; secondary (lateral) nerves 10-15 on each side, divergent, more or less straight; inflorescence lateral, subracemose, rachis smooth, short, 1.8-2 ст. long; flowers short-pedicellate; sepals 2.5 mm. long, 1-1.25 mm. broad, subobtuse, glabrous (in fruit) on out- side, not punctate; petals 4—4.5 mm. long, 1 mm. broad, lance- olate, glabrous on both sides, epunctate; follicle 2.3-2.5 cm. long, 1.5 em. broad, laterally compressed, obliquely mucronate, stipitate at the base, 5 mm. long; pericarp subligneous, dull orange-brown, glabrous outside, finely lineolate, tomentose within; seed 1.3 em. long, 0.8 em. thick, laterally compressed, testa black, shining.—parren: tree 6 m. tall, fruit orange; trail between Pinogana and Yavisa, ca. 15 m., March 17, 1937, Allen 249 (U. S. Nat. Herb., түре, fragment of type in Herb. Field Mus.). Connarus Allenii is distinct in its narrowly oblanceolate 5- foliolate elongated leaflets and short lateral rachises of the inflorescence. It is related to C. Turzaninowi, which differs in its smaller oval leaflets, and to C. punctatus, which has the fruit subglabrous instead of tomentose within and broader leaflets. 1941] WOODSON & SCHERY—FLORA OF PANAMA. V 431 OXALIDACEAE Охатлв coccinea Woodson & Schery, spec. nov. Suffruticosa са. 4-5 dm. alta. Caulis basi 2-3 mm. crassus, lignosus brun- neus, uno vel duobus ramis magnis basi ex quibus rami minores brevissimi pubescentibus cum pilis adscendentibus oriuntur. Folia inaequilateraliter alternata trifoliolata petiolata. Folio- lae facie superiore subglabrae (pilosae juventute), facie infe- riore plus minusve dense adpresse pilosae et subglaucoideae; foliola media elliptico-lanceolata, ovato-lanceolata vel ellip- tica, 15-28 mm. longa, 7-13 mm. lata, apice emarginata vel ob- tusa, margine paulo undulato, petiolulo 4-5 mm. longo; foliolis lateralibus parvis subsessilibus inaequalibus apice emargi- natis, 3-6 mm. longis, 2-4 mm. latis, petiolis plus minusve pu- bescentibus, 10-15 mm. longis, stipulis destitutis. Pedunculi axillares petiolis longiores, 15-20 mm. longi, pubescentes apice subumbellati; pedicellis brevibus 1-2 mm. longis coccineis pilosis. Sepala coccinea, 4 mm. longa, basi 1 mm. lata, exteriora lanceolata pubescentia. Petala aureo-lutea, ca. 8 mm. longa. Filamenta maiora infra medium appendiculata supra ap- pendieem pubescentia. Styli glabri. Capsula ignota.—socas DEL TORO: near Nievecita, alt. 50 m., Aug. 1940, Woodson Ф Schery 1029 (Herb. Missouri Bot. Gard., TYPE). This species is close to O. Neaei DC. but has a more tenuous, less robust habit. The buds and calyces are slightly smaller and tinged a deep red. The young stems and leaflets are also tinged deep red, this being apparent even in the dry plant. The older leaflets are glabrous on the upper surface. The fibrous root system is stout. RUTACEAE (C. L. Lundell, Ann Arbor, Mich.) ZANTHOXYLUM Scheryi Lundell, sp. nov. Arbor parva, ramulis crassis, angulatis. Folia imparipinnata; foliola 7 vel 9, petiolulis 5-9 mm. longis; lamina subcoriacea, integra, ob- longo-elliptica vel elliptica, 3-7.5 em. longa, 1.9-4 em. lata, apice obtusa et emarginata, basi acuta. Inflorescentiae parce puberulae, cymoso-paniculatae, axillares, usque ad 15 cm. [Vor. 28 432 ANNALS OF THE MISSOURI BOTANICAL GARDEN longae. Pedicelli 2.5-3 mm. longi. Sepala 5. Petala 5, glabra, lanceolato-oblonga, 3 mm. longa. Filamenta 2-2.5 mm. longa. A tree, 6 m. high, branchlets stout, angled, very minutely and sparsely puberulent, essentially glabrous. Leaves odd-pinnate, the petiole up to 6 em. long, the rachis up to 13 em. long, the petiole and rachis canaliculate. Leaflets 7 or 9; petiolules canaliculate, 5-9 mm. long; leaf blades subcoriaceous, oblong- elliptic or elliptic, 3-7.5 em. long, 1.9-4 em. wide, apex abruptly narrowed, obtuse and emarginate, base acute, decurrent, sparsely puberulent above along the impressed costa, gla- brous otherwise, margin entire, paler beneath, costa prominent on under surface, primary lateral veins slender and inconspic- uous, 12-14 on each side. Inflorescences cymose-paniculate, axillary, solitary, up to 15 em. long, including peduncle 9 cm. long, minutely and sparsely puberulent; bracts small, decid- uous. Pedicels 2.5-3 mm. long, puberulent. Flowers green- yellow, usually 5-parted. Sepals triangular-ovate, about 0.7 mm. long, acute. Petals glabrous, lanceolate-oblong, 3 mm. long, obtuse and usually emarginate. Filaments 2-2.5 mm. long. Rudimentary pistils usually 5.—cutrieuf: Potrero Mu- leto to summit, Volean de Chiriqui, alt. 3500-4000 m., July 13- 15, 1940, Woodson € Schery 451 (Herb. Univ. Michigan, түре). This remarkable species has affinity to Z. chiriquinum Standl. The entire elliptic or oblong-elliptic leaflets and the puberulent panicles up to 15 em. long are characteristics by which Z. Scheryi may be readily separated. POLYGALACEAE (S. F. Blake, Washington, D. C.) PorvaaArna ЁкхргЕвт Chod. var. HETEROTHRIX Ваке. —сосіЁ: between Las Margaritas and El Valle, 1938, Woodson, Allen & Seibert 1261 (U. S. Nat. Herb.) ;снтвтөсї: common weed at 800 m., vicinity of El Valle de Antón, alt. 600 m., Allen 1933 (U. S. Nat. Herb.). Previously known only from Colombia. PoLyGaLa GALIOIDES Poir.—cocrLÉ: on open llanos, Llano Bonito, north of Las Margaritas, alt. 400-500 m., July 26, 1935, Seibert 523 (U. S. Nat. Herb.) ; vicinity of El Valle, alt. 100- 800 m., Sept. 5, 1938, Allen 773 (U.S. Nat. Herb.) ; cumiqví: vi- 1941] WOODSON & SCHERY— FLORA OF PANAMA. V 433 cinity of El Valle de Antón, alt. 600 m., Allen 1932 (U. S. Nat. Herb.). Not previously known north of Colombia, although er- roneously recorded by Hemsley from Mexico and Guatemala. CELASTRACEAE (C. L. Lundell, Ann Arbor, Michigan) MavrENUS Моорѕох Lundell, Ann. Missouri Bot. Gard. 26: 291. pl. 22. 1939 (Salacia blepharodes Pittier, Contr. U. S. Nat. Herb. 18: 159. 1916, not Maytenus blepharodes Lundell; Maytenus vulcanicola Standl., Field Mus. Publ. Bot. 22: 153. 1940).—cnm:iiQuí: a tree, 2-4 m. high, around El Potrero Camp, Volean de Chiriquí, at edge of forest, alt. 2800-3000 m., Mareh 10-13, 1911, H. Pittier 3086 (U. S. Nat. Herb., No. 677476, TYPE of Salacia blepharodes); a tree, З m. high, Loma Larga to summit, Volean de Chiriquí, alt. 2500-3380 m., July 4—6, 1938, Woodson, Allen & Seibert 1065 (Herb. Univ. Michi- gan, TYPE of M. Woodsoni); a tree, 3 m. high, flowers maroon, same locality and date, Woodson, Allen & Seibert 1088; near summit, Volean de Chiriquí, alt. about 3700 m., Feb. 12, 1940, M. E. Terry 1309 (Field Mus. Herb., түре of Maytenus vul- canicola); a tree, 4 m. high, flowers maroon, Potrero Muleto to summit, Volean de Chiriquí, alt. 3500—4000 m., July 13-15, 1940, Woodson & Schery 398; a small tree, 6 m. high; same locality and date, Woodson & Schery 440. The types of Salacia blepharodes and Maytenus vulcanicola do not differ from the type of M. Woodsom. VITACEAE Cissus BREVIPES Morton & Standl.—cocré: north rim of El Valle de Antón, near Cerro Turega, alt. 650-700 m., June 30, 1940, Woodson & Schery 182. Previously known from the type locality in Costa Rica. LOASACEAE KLAPROTHIA MENTZELIOIDES НВК.—сн1в190!: vicinity of Bajo Chorro, alt. 1900 m., July 20-22, 1940, Woodson & Schery 650. New to Panama, known previously from South America and Costa Rica. [Vor. 28 434 ANNALS OF THE MISSOURI BOTANICAL GARDEN BEGONIACEAE (1. B. Smith ф В. С. Schubert, Cambridge, Mass.) ВЕвомтА vestita C. DC.—cnumiqvuí: epiphytic, vicinity of Bajo Chorro, alt. 1900 m., July 20-22, 1940, Woodson & Schery 621. Previously known from Costa Rica. MELASTOMACEAE (H. A. Gleason, New York) CENTRONIA PHLOMOIDES Тнапа—снтвтоо!: vicinity of Вајо Chorro, alt. 1900 m., July 21, 1940, Woodson & Schery 657. Previously known only from Costa Rica. CLIDEMIA sETOSA (Triana) Gleason—cocrí: north rim of El Valle, near Cerro Turega, alt. 650-700 m., June 30, 1940, Woodson € Schery 200. Previously known from Vera Cruz to Costa Rica. Мтсомта orNo YLLA Donn. Sm.—ococré: north rim of El Valle, near Cerro Turega, alt. 650-700 m., June 30, 1940, Woodson & Schery 174. Previously known from Guatemala, British Honduras, Honduras, and Colombia. Мтсохта Pirriert Cogn.—cuirigui: vicinity of Bajo Chorro, alt. 1900 m., July 20, 1940, Woodson & Schery 630. Previously known from Costa Riea. OssaEA МІСВАХТНА (Sw.) Maef.—cnuiqví: Quebrada Velo, alt. ca. 1700 m., July 8, 1940, Woodson & Schery 267. Guate- mala to Peru and the West Indies, but apparently new to Panama. There are five species of Blakea in Costa Rica and Panama which are apparently closely related to each other while dis- tinetly different from the other species of the genus from the same regions. They belong to the section Pyxidanthus, in which the thick leathery bracts are connate for nearly or quite their whole length into two cups, which are scarcely different in length and from which the calyx barely protrudes ; the leaves are coriaceous and the flowers unusually large. Another fea- ture of all five is the character of the pubescence. This may be abundantly developed or confined to the leaf-buds alone; it consists of conic many-celled hairs closely aggregated into a 1941] WOODSON & SCHERY—FLORA OF PANAMA. V 435 velvety indument, from which project other similar hairs of much larger size, up to 2 mm. long. The differences between them may be summarized in a brief key: Petals eoarsely tubereulate; anthers short-spurred at the bas Sepals tipped with a reflexed foliaceous appendage; leaf- ‘blade about twice йй long №: uere ешге шш 9a CE Yon NO SACO EE calycosa Gl. Sepals unappendaged; leaf-blades cordate at base, about as “& as long В АВ tS E PI T ERIT B. tuberculata Donn, Sm. Petals not tuberculate; anthers rounded at the bas Younger parts, except the leaf-buds, glabrous or nearly so; leaf-blades 5-nerved, with an additional marginal pai Leaf-blades rounded at the base, the ous veins spreading at right angles; hypanthium and calyx together 15-18 mm. long.............. ОВОО JEU QUUD E NS QU OT. B. intercepta sp. nov. Leaf-blades euneate at the base, the secondary veins spreading at an angle of 70°; hypanthium and calyx together 20-25 mm. long..... B. Woodsoni Gl. Young stems, petioles, leaf-veins beneath, peduncles, bracts, and sepals densely pubescent; leaves 5-nerved with an additional marginal pair...... ое В. brunnea sp. nov. BraxzA brunnea Gleason, sp. nov. Arbor; rami subteretes, petioli, venae subtus, pedunculi, bracteae, et sepala dense brun- neo-velutini. Petioli validi 3-6 em. longi. Laminae subcori- aceae, ellipticae, usque 20 em. longae 12 ст. latae, apice brev- iter apiculatae, basi late rotundatae, supra glabrae opacae venis impressis, subtus brunneae, 5-пегуіае jugo marginali neglecto, venis dense pubescentibus valde elevatis, venulis tenuiter pubescentibus sub angulo fere recto divergentibus. Pedunculi solitarii erassissimi 15-20 mm. longi. Bracteae toto connatae, involucram subcylindricam 25 mm. longam for- mantes. Flores 6-meri. Hypanthium glabrum. Calyx breviter 6-lobatus, extra pubescens. Petala staminaque matura non visa. Antherae dolabriformes 7 mm. longae. Stylus 3 em. ! BLAKEA intercepta Gleason, sp. nov. wie inan rami vix cogniti ut videtur ancipites fere glabri. Petioli subglabri em. longi. Laminae elliptieae, usque 20 cm. longae 11 em. latae, obtusae, E. die late acutae, 5-nerviae, jugo marginali neglecto, kragas fere glabrae, venis lateralibus sub angulo fere recto loc ru ed Inflorescentia floresque ignoti. Braeteae connatae hypanthium amplecten E glabrum circa 10 mm. longum. Calyx glaber circa 6 mm. longus, ao obatus. Type collec s in Costa Rica by Otto Kuntze, E э, 1874, and labeled in Kuntze's bà с Blakea multiflora Don. Topobea Trian [Vor. 28 436 ANNALS OF THE MISSOURI BOTANICAL GARDEN longus glaber attenuatus.—cnruqví: vicinity of Bajo Chorro, alt. 1900 m., July 20-22, 1940, Woodson & Schery 615 (Herb. New York Bot. Gard., TYPE). CoNosrEcrA chiriquensis Gleason, sp. nov. Arbor glaberrima, ut videtur parva, ramis crassis rotunde 4-angulatis cicatricibus foliorum rotundis notatis. Petioli gracili, usque 3 em. longi. Laminae subcoriaceae, ellipticae, usque 12 em. longae 6 сш. latae, abrupte breviterque caudato-acuminatae, remote denti- culatae, basi abrupte cuneatae, 3-nerviae vel vix 3-pli-nerviae, jugo marginali neglecto, venis secundariis sub angulo 70? divergentibus supra obsoletis. Panicula circa 1 dm. longa, a basi ramosa, pedicellis 3-5 mm. longis. Alabastra 9-10 mm. longa in specimine typico, in altero usque 14 mm., abrupte acuminata in rostrum crasse subulatum 2 mm. longum. Hypan- thium late poculiforme, 4 mm. longum. Petala late rhom- boidea, alba, 12 mm. longa, valde asymmetrica, latere uno in lobum triangularem produeto, altero rotundato. Antherae ob- longae, obtusae, 2-loculares, 2.8 mm. longae; connectivum dorse proper basin tumidum. Ovarium 6-loeulare, fere toto inferum, truncato-conicum; stylus crassus, 4.4 mm. longus; stigma hemisphaerieum, 1.5 mm. latum.—currigui: vicinity of Finca Lérida, alt. 1750 m., July 11, 1940, Woodson & Schery 376 (Herb. New York Bot. Gard., түрк); vicinity of Callejon Seco, Volean de Chiriquí, alt. 1700 m., July 17, 1940, Woodson & Schery 479 (Herb. New York Bot. Gard., сотурв). The closest relatives of C. chiriquensis are two Costa Rican species, C. rhodopetala Donn. Sm., with more sharply angled branches, longer shining leaves with the secondary veins spreading at right angles, and much smaller petals, and C. Pit- пеги Cogn., with smaller, thinner, proportionately wider, dis- tinetly 3-pli-nerved leaves. Мевтамта ($ Davra) panamensis Gleason, sp. nov. Arbor glaberrima 20 m. alta, ramis junioribus subteretibus ad nodos paullo inerassatis. Petioli atropurpurei 25-40 mm. longi. Laminae chartaceae, ovato-ellipticae, usque 13 em. longae 7 cm. latae, in euspem brevem obtusam rotundatae, integrae, 1941] WOODSON & SCHERY—FLORA OF PANAMA. V 437 basi late cuneatae, 5-nerviae, venulis creberrime reticulatis. Panicula magna ramosa multiflora; flores 5-meri breviter (5 mm.) pedicellati ad apieem ramulorum. Hypanthium late poculiforme 6.5 mm. latum. Calyx patulus, 3.5-4 mm. latus, limbo vix 5-lobato. Petala late et inaequilateraliter cuneato- obovata, 3 em. longa. Stamina episepala: filamenta 17 cm. longa; antherae crasse subulatae, arcuatae, 7 mm. longae; con- nectivum non appendieulatum, supra basin tumidum. Sta- mina epipetala: filamenta 12 mm. longa; antherae graciliter subulatae, valde arcuatae, 12—13 mm. longae; connectivum ad basin productum in appendicem dorsalem linearem apice dila- tatam 4 mm. longam. Ovarium superum; stylus gracilis, 13 mm. longus; stigma punctiforme.—cuirigui: Casita Alta to Cerro Copete, Volcan de Chiriquí, alt. 2300-3300 m., July 10, 1940, Woodson € Schery 361 (Herb. New York Bot. Gard., TYPE). ARALIACEAE (A. C. Smith, Jamaica Plain) ScHEFFLERA epiphytica A. C. Smith, sp. nov. Frutex epi- phytieus circiter 2 m. altus ubique inflorescentia excepta glaber, ramulis erassis subteretibus pallidis striatis; foliis sim- plicibus, petiolis crassis ad 15 mm. longis, inferne late alatis, superne subteretibus et leviter canalieulatis, medium versus ligula coriacea oblongo-lineari ad 20 mm. longa persistente ornatis; laminis chartaceis siecitate viridi-olivaceis elliptico- oblongis, 18-23 em. longis, 9-11 em. latis, basi leviter cordatis vel rotundatis, apice abrupte acuminatis (acumine ad 8 em. longo acuto), margine integris et saepe leviter revolutis, costa utrinque prominente, nervis secundariis utrinsecus 12-16 sub- patentibus marginem versus anastomosantibus supra paullo subtus valde prominulis, venulis saepe utrinque prominulis; inflorescentia composito-paniculata, rhachi ramulisque gracil- ibus striatis mox glabris, pedunculo subnullo, rhachi 2—4 em. longa, ramulis paucis ad 35 em. longis (stipitibus ad 10 ет. longis inclusis) bracteis chartaceis oblongis acutis 22-30 mm. longis glabris subtentis; pedunculis numerosissimis patentibus gracilibus sub anthesi ad 9 mm. sub fructu ad 15 mm. longis [Vor. 28 438 ANNALS OF THE MISSOURI BOTANICAL GARDEN minute puberulis mox glabris; floribus 5-8 per umbellam, pedi- cellis gracilibus primo puberulis sub anthesi ad 3 mm. sub fruetu ad 6 mm. longis, bracteis basi pedicellorum membrana- ceis acutis deltoideis circiter 0.7 mm. longis; calyce levi ob- conico sub anthesi 1-1.5 mm. longo et circiter 2 mm. diametro, limbo minuto membranaceo, lobis nullis vel obscure apiculatis; petalis membranaceis glabris in calyptram mox caducam con- natis, calyptra semiglobosa circiter 2 mm. longa et 2.5 mm. diametro, apice obtuse umbonata, conspicue reticulato-ner- vata; staminibus 5, filamentis gracilibus 0.4—0.7 mm. longis, antheris deltoideo-ovoideis cireiter 1.5 mm. longis, basi sub- cordatis, apice obtusis, stylis 5 in columnam carnosam conicam cohaerentibus, apice minute liberis, loculis et ovulis 5; fructibus subglobosis ad 3.5 mm. diametro conspicue 5-costatis disco et stylorum columna eoronatis.—cocrÉ: epiphytic shrub 2 m. tall, flowers white; hills north of El Valle de Antón, vicinity of La Mesa, alt. 1000 m., Jan. 21, 1941, Allen 2344 (U. S. Nat. Herb., Nos. 1,806,122 and 1,806,123, түрк) ; same locality, Allen 2723 (Arn. Arb.). This remarkable plant is at once distinguished from other American species of Schefflera by its simple leaves; to the best of my knowledge only two Old World species of the genus with simple leaves are known. The plant is certainly a member of the genus Schefflera (nomen conservandum propositum; see A. C. Smith, in Trop. Woods 66: 5. 1941), agreeing in such un- mistakable characters as the conspicuous petiolar ligules, the panieulately umbelliferous inflorescences, and the calyptrate petals. Since both simple and compound leaves are found in such genera as Oreopanax and Didymopanaz, the discovery of а simple-leaved American species of Schefflera is not entirely surprising. VACCINIACEAE (A. C. Smith, Jamaica Plain, Mass.) Recent collectors in Panama have obtained several plants of the tribe Thibaudieae of unusual interest; in the present paper eight species and one genus are proposed as new, while five others are reported from Panama for the first time. The 1941] WOODSON & SCHERY—FLORA OF PANAMA. V 439 eolleetors whose plants are cited are Paul H. Allen, M. E. Davidson, C. and W. von Hagen, and R. E. Woodson and R. W. Schery. Their plants have been seen in the following herbaria: Arnold Arboretum (A); Missouri Botanieal Garden (Mo); New York Botanieal Garden (NY) ; U. S. National Herbarium (US); the parenthetical letters indicate the place of deposit. MacLEANIA epiphytica sp. nov. Frutex epiphyticus 1 m. altus ubique praeter corollae lobas glaber, ramulis cinereis teretibus gracilibus; petiolis rugosis crassis 4—6 mm. longis; laminis coriaceis siccitate olivaceis oblongo-ellipticis, 5-8 еш. longis, 2-3 em. latis, basi rotundatis vel late obtusis, apice ob- tusis, margine integris et incrassatis, 5-pli-nerviis, nervis prope basim orientibus supra subplanis vel prominulis subtus leviter elevatis, costa subtus subprominente, rete venularum utrinque ineonspieue prominulo; inflorescentiis apicem ramu- lorum versus axillaribus fasciculatis vel obscure racemosis 2— 4-floris, floribus bracteis minutis subtentis; pedicellis crassis 6-11 mm. longis, apicem versus leviter incrassatis, paullo supra medium braeteolis papyraceis ovato-deltoideis acutis circiter 1.5 mm. longis et latis bibracteolatis; calycis tubo obprismatico sub anthesi eireiter 3.5 mm. longo obscure 5-angulato, limbo subereeto lobis late deltoideis acutis circiter 1 mm. longis et 2.5 mm. latis inclusis circiter 2 mm. longo, sinibus rotundatis; corolla tenuiter carnosa cylindrica sub anthesi 17—18 mm. longa et basim versus circiter 5 mm. diametro, apicem versus con- tracta, lobis 5 deltoideis subacutis circiter 1.5 mm. longis intus albo-puberulis; staminibus 10, filamentis membranaceis liberis ligulatis cireiter 3 mm. longis, antherarum loeulis 4—4.5 mm. longis, tubulo unico rigido circiter 3 mm. longo rima ovali 2— 2.5 mm. longa dehiscente; stylo corollam subaequante, stig- mate inconspieuo.—cocLÉ: epiphytic shrub 1 m. high; corolla scarlet, tipped with green; hills north of El Valle de Antón, alt. 1000 m., June 23, 1940, Allen 2171 (A, TYPE). Macleania epiphytica, related to those species of the genus with a single anther-tubule and glabrous corollas, sharply dif- fers from them in having its filaments separate rather than connate in a tube. In foliage, the new species suggests M. com- [Vor. 28 440 ANNALS OF THE MISSOURI BOTANICAL GARDEN pacta A. C. Smith and M. insignis Mart. & Gal., but it has larger flowers than the former and smaller flowers than the latter. Psammısra panamensis sp. nov. Frutex epiphyticus ad 50 em. altus ubique praeter florum filamentas glaber, ramulis сгаз- sis fuscis apicem versus conspicue complanatis; petiolis ru- gosis erassis (circiter 3 mm. diametro) subteretibus 8-10 mm. longis; laminis tenuiter coriaceis siccitate fuscis oblongo- lanceolatis, 26-34 em. longis, 5.5-7.5 em. latis, basi subatten- uatis, apice gradatim acuminatis, margine integris et anguste revolutis, pinnatinerviis, costa supra elevata subtus promi- nente, nervis secundariis utrinsecus 8-10 supra prominulis subtus aeute elevatis marginem versus conspicue anastomo- santibus, inferioribus adscendentibus superioribus patentibus, rete venularum utrinque prominulo; inflorescentiis axillaribus breviter racemosis sub anthesi ut videtur 6-8-floris, pedunculo subnullo, rhachi 6-10 mm. longa angulata, pedicellis crassis 8- 12 mm. longis bracteis papyraceis ovatis subacutis circiter 1.5 mm. longis subtentis, basim versus bracteolis ut bracteis simil- ibus bibraeteolatis, apicem versus paullo incrassatis; calyce late campanulato sub anthesi 7-8 mm. longo, tubo 4-5 mm. longo et summo 5-7 mm. diametro, limbo tenuiter coriaceo sub- erecto circiter 3 mm. longo, lobis 5 deltoideis acutis circiter 2 mm. longis et 3 mm. latis; corolla tenuiter carnosa (basim versus tenuiore) truncato-conica sub anthesi 6-7 mm. longa et basi circiter 5 mm. diametro, apicem versus contracta, lobis 5 deltoideis subacutis circiter 1 mm. longis et latis; staminibus 10, filamentis submembranaceis ligulatis 1.5-2 mm. longis, margine puberulente-ciliolatis, apicem versus angustatis, con- neetivis angustis alternatis obscure bicalcaratis vel interdum ecalearatis, antheris rigidis 3.5—4 mm. longis, loculis basi con- spieue incurvatis, tubulis cylindrico-conicis circiter 1.5 mm. longis rimis elongatis ovalibus dehiscentibus; stylo corollam subaequante, stigmate inconspieuo.—cocrí: epiphytic shrub 90 em. high; hills north of El Valle de Antón, alt. 1000 m., June 23, 1940, Allen 2152 (A, TYPE). Psammisia panamensis is the second species of the genus thus far known from North America, but it is not of the rela- 1941] WOODSON & SCHERY—FLORA OF PANAMA. V 441 tionship of P. ramiflora Kl. of Costa Rica and Panama. The new species appears to be closely related only to the Peruvian P. globosa A. C. Smith, the two being essentially identical in foliage. Psammisia panamensis, however, in comparison with its ally, has a substantially larger calyx with conspicuous lobes, a larger and conical (rather than subglobose) corolla, and ob- viously larger stamens with less conspicuous and essentially unspurred connectives. THIBAUDIA COSTARICENSIS Hoer. Bot. Jahrb. 42: 311. 1909; А. C. Smith, Contr. U. S. Nat. Herb. 28: 425. 1932.—cutriqvi: tree to 10 m. high, common in rain-forest; corollas pink and white; Boquete Region, Cerro Horquete, alt. about 2000 m., von Hagen 2023 (A, NY); epiphytic shrub in rain-forest; calyx rose; corolla white; Boquete Region, Bajo Chorro, alt. about 1800 m., Davidson 380 (A). The species has previously been known only from Costa Rica, elevations up to 1600 m. only having been reported. ANTHOPTEROPSIS A. C. Smith, gen. nov. Frutices, foliis breviter petiolatis, laminis subcoriaceis; in- florescentiis terminalibus vel axillaribus breviter racemosis, floribus pedicellatis bracteis subtentis; calyce cum pedicello articulato alis 5 conspicuis e basi ad apicem adscendentibus lobis alternatis praedito, dentibus 5 inconspicuis; corolla саг- поза cylindrica 5-lobata ; staminibus 10 aequilongis, filamentis membranaceis, antheris basim versus dorsifixis, loculis leviter granulatis basi paullo incurvatis, connectivis angustatis alter- natim summo loculorum calcare mediano unicalearatis, tubulis 2 ad basim distinctis flexibilibus per rimas ovales introrses elongatas dehiscentibus; ovario in calyce immerso disco an- nulari complanato coronato, loculis 5, placentis magnis, ovulis numerosis, stylo filiformi corollam subaequante, stigmate mi- nute subeapitato. ANTHOPTEROPSIS insignis sp. nov. Frutex ad 4 m. altus ubique praeter inflorescentiae partes et ramulos juveniles in- conspicue puberulos glaber, ramis ramulisque teretibus cin- ereis gracilibus leviter rimosis juventute fuscis angulatis; [Vor. 28 442 ANNALS OF THE MISSOURI BOTANICAL GARDEN petiolis paullo incrassatis supra complanatis 2-3 mm. longis; laminis elliptico-oblongis, (3-)4-6 em. longis, (1.5-)2-3 em. latis, basi anguste rotundatis, apice obtusis vel obtuse cuspi- datis, margine integris et anguste recurvatis, 5-pli-nerviis, nervis paullo supra basim orientibus adscendentibus utrinque valde prominulis, costa subtus valde elevata, rete venularum utrinque leviter prominulo; inflorescentiis terminalibus vel raro apicem ramulorum versus axillaribus, 5-12-floris, floribus braeteis papyraceis deltoideis circiter 1.5 mm. longis sub- acutis inconspicue ciliolatis subtentis, rhachi subtereti ad 1 em. longa; pedicellis subteretibus rugulosis crassis 15-20 mm. longis superne gradatim incrassatis et saepe inconspicue pu- berulis, medium versus bracteolis ut bracteis sed circiter 1 mm. longis bibracteolatis; calyce carnoso subgloboso sub anthesi alis inelusis 8-10 mm. diametro, tubo limbum longitudine sub- aequante, limbo erecto, dentibus minute apieulatis, alis car- nosis (vel siccitate submembranaceis) 2-3 mm. latis, apicibus alarum dentes 1 mm. excedentibus; corolla sub anthesi 16-19 mm. longa et 4—5 mm. diametro, inconspicue 5-angulata, su- perne pilis fuscis debilibus circiter 0.3 mm. longis parce pilosa demum glabrescente, lobis deltoideis apiculatis circiter 1 mm. longis et 2-2.5 mm. latis, sinibus acutis; staminibus circiter 9 mm. longis, filamentis distinetis basim corollae saepe cohae- rentibus pallidis ligulatis circiter 3 mm. longis et basi 1.3 mm. latis, superne dorso pallide hirsutis, antheris circiter 7 mm. longis, loculis 4-5 mm. longis, connectivis pallidis, calcare circiter 0.5 mm. longo, tubulis quam loculis paullo angustior- ibus acutis.—cocrÉ: large straggling shrub 4 m. high, the stem 15 em. in diameter, calyx fleshy, pink, corolla yellow; hills south of El Valle de Antón, alt. about 700 m., May 9, 1941, Allen 2500 (А, TYPE, Мо). This specimen, one of many remarkable plants collected in Coclé by Mr. Allen, is not readily placed in any existing genus. At first glance (as indicated by my generic name) it suggests Anthopterus Hook., but closer examination indicates that it differs in many essential features. Other genera showing cer- tain points of relationship are Thibaudia R. & P., Macleania 1941] 443 WOODSON & SCHERY—FLORA OF PANAMA. V "e[q1xog So[nqnj f uerpour ‘Tesiop 'soArjoouuoo эзви -леув uo 3uosoid sands {1800818 f[oug so[noor] 'e[qixog sepnqnj < (194 NH sisuajpdno ‘ұр ul [91314вәл) Surqov[ sands {181918 f[oug so[noor] "'e[qixog вәгпд03 Зотов sands + теүпп®1@ Á[oug samor "раби вота *(uvrpeu you “әлт}әәп -под Jo xəde 3*9 [81938] mq) yueseid sands ‘rej -nuvid 41810148 so[noor "рез вота -nj :2urqov[ sands $18] -nueid 4800138 вәүпәот :вләцутү "remqng, “герат, "Seqo[ sno -nordsuoour əy} SZurpooo 04 рә8птм &үвпопәтавпоә/ ‹{тәәтрәа цум әўвүпәт}гу| ‘SOSNUIS 03 PosuTM ATMOLIvU Кола! ‹тәәтрәа Чата озере Кепе | ‘pedutm 4151019148000) 'eje[ooo1n 10 тетапт, ‘sgua 94} Zurpoooxo S9qO[ SNOTAGO YIM 304 ‘sasnuts 03 родшм ‹әә -трэ4 цул впопит поо "0je[oooin ло гета :20200 "рэЗитм jou f[eorped цум өјегпәтуіу |‹ *(seqo[ sno -"nordsuoou; 94$ Surpooo ‹тәәтрә цата oje[nory1y :х во sisdo4241doqyup Tipnoqui T, snaa?doqyyup- DisMAUDSq D1wD210D0J [Vor. 28 444 ANNALS OF THE MISSOURI BOTANICAL GARDEN Hook., and Psammisia Kl. The accompanying table summa- rizes the more important points of difference among these genera, as regards their relationship to Anthopteropsis. From this summary it is perceived that Anthopteropsis has some essential characters in common with each of the men- tioned genera. Its ealyx is faintly suggestive of that of Mac- leania pentaptera Hoer., but staminal characters definitely re- move it from that genus. The spurs of the connectives of Psam- misia are of quite a different type from those of the new genus, and in many other respects Psammisia may be excluded as a close relative. The conspicuously winged corollas of An- thopterus amply differentiate that genus. The species of Thibaudia have no suggestion of accentuated calyx-wings, and the vestigial connective-spurs of T. cupatensis Huber are hardly of the same type as the spurs of Anthopteropsis. Nevertheless, Thibaudia may be the closest ally of the new genus. I have already discussed (in Bull. Torrey Bot. Club 63: 316. 1936) the several elements which have in the past been re- ferred to Thibaudia, and it hardly seems advisable to expand that generic concept further by dubious inclusion of the pres- ent plant. CavENDiIsHIA gaultherioides sp. nov. Frutex epiphyticus ad 1 m. altus ubique glaber, ramulis gracilibus subteretibus cin- ereis vel purpurascentibus; foliorum laminis subsessilibus (petiolis erassis ad 3 mm. longis et diametro) chartaceis vel tenuiter eoriaeeis siccitate fusco-olivaceis bullatis ovatis vel suborbicularibus, 5-12 em. longis, 4-10 em. latis, basi rotunda- tis vel leviter cordatis, apice obtusis vel rotundatis, margine integris et anguste revolutis, pinnatinerviis, costa supra basim versus elevata subtus prominente, nervis secundariis utrinse- eus 5-10 inferioribus adscendentibus superioribus brevibus patentibus, omnino supra in sulcis prominulis subtus promi- nentibus marginem versus nervo marginali conjunctis, rete venularum utrinque leviter prominulo; inflorescentiis axillar- ibus eongestis breviter racemosis basi bracteis deciduis suf- fultis, rhachi erassa (1.5-2 mm. diametro) sub anthesi ad 2 em. longa 5-12 flora; bracteis floriferis papyraceis concavis ellip- 1941] WOODSON & SCHERY—FLORA OF PANAMA. V 445 tieo-oblongis vel suborbieularibus 6-10 mm. longis et latis, basi paullo angustatis, apice acutis vel rotundatis, conspicue et copiose flabellato-nerviis (nervis utrinque prominulis), mar- gine saepe minute glanduloso-ciliolatis; pedicellis inconspicuis crassis circiter 1 mm. longis bibracteolatis, bracteolis ut brac- teis sed angustioribus vel interdum reductis ; calyce campanu- lato circiter 6 mm. longo, tubo circiter 2 mm. longo et 3-4 mm. diametro, limbo erecto-patente summo 4—5 mm. diametro con- spicue nervato, lobis 5 ovato-deltoideis, circiter 1.5 mm. longis et 2-3 mm. latis, apice apiculatis, margine glanduloso-ciliolatis, basi saepe minute imbricatis, sinibus aeutis; corolla tenuiter carnosa cylindrica 7-10 mm. longa, 3-4.5 mm. diametro, supra medium interdum ineonspieue glandulosa, lobis 5 oblongo- deltoideis acutis; staminibus 10 subaequalibus, filamentis mem- branaceis ligulatis alternatim 2-2.5 mm. et 3-4 mm. longis, antheris alternatim circiter 6 mm. et 5 mm. longis, tubulis quam loculis circiter duplo longioribus; stylo gracili corollam subaequante, stigmate irregulariter peltato.—cocLÉ: epi- phytie shrub 50 em. high, the braets pink, the corollas white; hills north of El Valle de Antón, alt. 1000 m., Allen 2170 (A); epiphytic shrub 1 m. high, the bracts shell-pink, the corollas greenish yellow; vicinity of La Mesa, north of El Valle de Antón, alt. 1000 m., April 12, 1941, Allen 2375 (А, түре, Mo). Cavendishia gaultherioides is a remarkably distinct. species, characterized by its subsessile pinnate-nerved leaf-blades and its compact inflorescences with conspicuously nerved and often glandular-ciliolate bracts and bracteoles, subsessile flowers, and narrowly imbricate or at least contiguous calyx-lobes. Its only close relatives are C. complectens Hemsl. and C. com- pacta A. C. Smith; the former differs in its cordate-based amplexicaul leaf-blades, longer inflorescences, more numerous flowers, obvious pedicels, and its larger eglandular bracts; the latter differs in its larger pli-nerved caudate-acuminate peti- olate leaf-blades, elongate inflorescences, large bracts, obvi- ously pedicellate and larger flowers, etc. CavenpisH1A Allenii sp. nov. Frutex epiphyticus ad 1 m. altus ubique (corolla non visa forsan excepta) glaber, ramulis [Vor. 28 446 ANNALS OF THE MISSOURI BOTANICAL GARDEN subteretibus fusco-cinereis; foliorum laminis subsessilibus (petiolis inerassatis rugulosis 2-3 mm. longis) coriaceis sicci- tate fusco-olivaceis ovatis, 6.5-10 em. longis, 4-7 cm. latis, basi rotundatis vel inconspicue subcordatis, apice obtusis vel subrotundatis, margine integris et valde recurvatis, 5-pli- nerviis, costa supra paullo elevata vel subplana subtus promi- nente nervis lateralibus utrinsecus 3-5 inconspicuis superne praedita, nervis secundariis e basi orientibus adscendent- ibus utrinque elevatis, venulis immersis; inflorescentiis axil- laribus racemosis, rhachi angulata crassa (1.5-3 mm. dia- metro) 2.5-6 em. longa multiflora, floribus congestis; bracteis floriferis papyraceis concavis ellipticis, 17-20 mm. longis, 10- 15 mm. latis, apice aeutis, margine integris, conspicue et cop- iose flabellato-nervatis (nervis utrinque prominulis); floribus subsessilibus, pedicellis inconspicuis ad 1 mm. longis bibracte- olatis, braeteolis ut braeteis sed angustioribus et paullo brev- ioribus; calycis tubo post anthesin subgloboso levi 3-5 mm. diametro, limbo erecto papyraceo tubuloso lobis inclusis 7-8 mm. longo basi 2.5-3 mm. apice circiter 4 mm. diametro con- spicue striato-nervato, lobis 5 oblongo-deltoideis 2-2.5 mm. longis et latis, apice apiculatis, margine interdum inconspicue glanduloso-ciliolatis, sinibus acutis; fructibus subglobosis ad 15 mm. diametro calycis limbo conspicuo coronatis, pericarpio tenui ut videtur submembranaceo, seminibus numerosis ovoi- deis circiter 1 mm. longis conspicue serobieulatis.—cocrÉ: epi- phytie shrub 1 m. high; fruit blue; along trail to Las Minas, north of El Valle de Antón, alt. 1000 m., May 10, 1941, Allen 2471 (A, TYPE). Although no corollas are available, the cited collection so ob- viously represents a new species that I venture to describe it. Cavendishia Allenii is at once distinguished by its subsessile leaves with rounded bases and obtuse apices, its subsessile flowers, and its papyraceous and conspicuously nerved bracts, bracteoles, and calyx limb, the latter being tubular and elongated. The mature fruit has a remarkably thin pericarp for the genus. The new species is suggestive of the preceding (C. gaultherioides ) in its subsessile flowers and conspicuously 1941] WOODSON & SCHERY—FLORA OF PANAMA. V 447 nerved bracts and bracteoles, but its calyces and foliage are so different that the close alliance of the two is dubious. CavENDISHIA Calycina sp. nov. Frutex epiphyticus ubique glaber, ramulis gracilibus subteretibus cinereis; petiolis ru- gosis subteretibus crassis (ad 2 mm. diametro) 4-7 mm. longis; laminis coriaceis elliptico-oblongis, 7-11 ст. longis, 3-5 em. latis, basi rotundatis vel late obtusis, apice acuminatis (acu- mine gracili ad 2 em. longo acuto), margine integris et leviter recurvatis, subtus minute nigro-punetatis, 5- vel 7-pli-nerviis, nervis adscendentibus prope basim orientibus cum costa supra leviter impressis subtus prominentibus, exterioribus debilibus, rete venularum supra prominulo subtus subimmerso; inflo- rescentiis apicem ramulorum versus axillaribus breviter race- mosis ut videtur 8-12-floris, rhachi circiter 3 mm. crassa et 10-12 mm. longa; bracteis floriferis papyraceis vel submem- branaceis oblongo-elliptieis, 28-30 mm. longis, 16-18 mm. latis, apice rotundatis, margine integris; pedicellis crassis 4—5 mm. longis basim versus bracteolas 2 circiter 4 mm. longas gerenti- bus; calyce campanulato 7-8 mm. longo, tubo brevissimo cir- citer 4 mm. diametro, limbo erecto papyraceo lobis inclusis circiter 6 mm. longo, lobis 5 oblongis, 3-3.5 mm. longis, 2.5- 3 mm. latis, apicem versus scariosis, apice calloso-apiculatis, margine leviter undulatis et saepe contiguis, sinibus acutis vel anguste rotundatis; corolla tenuiter carnosa cylindrica, sub anthesi 16-18 mm. longa et 3-4 mm. diametro, lobis parvis deltoideis subacutis; staminibus 10 subaequalibus quam corolla paullo brevioribus, filamentis chartaceis nigrescentibus ligu- latis alternatim circiter 3 mm. et 5 mm. longis apicem versus obscure cinereo-pilosis, connectivis angustis nigrescentibus, antheris alternatim circiter 14 mm. et 12 mm. longis, tubulis quam loculis circiter duplo longioribus; stylo gracili corol- lam subaequante, stigmate peltato.—cnminiQví: epiphyte, in rain-forest, corolla red, tipped with white; Boquete District, Bajo Chorro, alt. about 1800 m., Jan. 18, 1938, Davidson 157 (A, TYPE). Although the cited specimen bears only one disintegrated in- florescence and two flowers, the materialis quite adequate to in- [Vor. 28 448 ANNALS OF THE MISSOURI BOTANICAL GARDEN dicate that a very distinct new species is represented. While the calyx-lobes are not exactly ‘‘imbricate,’’ as indicated in my key (Contr. U. S. Nat. Herb. 28: 463. 1932) for C. compacta A. C. Smith, this Colombian species nevertheless appears to be the closest ally of the new species. Cavendishia calycina differs from C. compacta in its smaller leaf-blades with the nerves oriented nearer the base, its more compact inflorescences, thin- ner and larger bracts, and smaller calyx, especially as regards the lobes, which are erect but not imbrieate nor thickened at the margin. CAVENDISHIA CAPITULATA Donn. Sm. Bot. Gaz. 25: 147. 1898; А. C. Smith, Contr. U. S. Nat. Herb. 28: 450. 1932. —6nriiqví: alt. 1500-1900 m., Allen 1434 (Mo), Davidson 92 (A), 120 (A), Woodson € Schery 574 (NY), 576 (Mo), 632 (NY); сосьй: alt. 1000 m., Allen 2721 (A). Although known from abundant material collected in Costa Riea, the above-cited specimens are apparently the first of the species from Panama. CAVENDISHIA LONGIFLORA Donn. Sm. Bot. Gaz. 37: 420. 1904; А. C. Smith, Contr. U. S. Nat. Herb. 28: 452. 1932.—cocr£: epiphytic shrub, the bracts and flowers bright rose; hills north of El Valle de Antón, alt. 650-1000 m., Woodson Ф Schery 193 (NY); epiphytic shrub, the branches 2 m. long; bracts rose- pink, the corollas white; Allen 2207 (A, Mo). This species has previously been known only from the type locality near La Palma, San José, Costa Rica. A recently col- lected specimen from the Province of Chiriqui in Panama, Seibert 176 (US), also doubtless represents the species, but it is a form with very narrow leaves. CAVENDISHIA STENOPHYLLA А. C. Smith, Ann. Missouri Bot. Gard. 27: 327. 1940.—cocré: epiphytic shrub 1 m. high; flowers purple; hills north of El Valle de Antón, alt. 1000 m., Allen 2162 (A); shrub 1.5 m. high, growing on steep banks along trails; flowers deep purple; vicinity of La Mesa, north of Е] Valle de Antón, alt. 1000 m., Allen 2373 (A, Mo). The cited specimens are the second and third collections of 1941] WOODSON & SCHERY—FLORA OF PANAMA. V 449 the species and precisely match the type, Allen 1880, from the Province of Panama. The following description may now be added to the original: Rhachis inflorescentiae ad 10 cm. longa, corolla tenuiter earnosa cylindrica sub anthesi 16-19 mm. longa et 4-5 mm. diametro, basi paullo contracta, extus pilis circiter 0.3 mm. longis laxe cinereo-puberula, lobis 5 deltoideis subacutis cir- citer 1 mm. longis et 2 mm. latis; staminibus subaequalibus, filamentis submembranaceis ligulatis distaliter parce puberulis alternatim circiter 2 mm. et 5 mm. longis, antheris gracilibus alternatim circiter 16 mm. et 13 mm. longis, tubulis quam loculis 1.5-2-plo longioribus; stylo gracili corollam sub- aequante, stigmate obscuro. CavENDISHIA chiriquiensis sp. nov. Frutex epiphyticus gra- eilis, ramulis glabris fuscis subteretibus vel apicem versus saepe leviter complanatis; petiolis gracilibus subteretibus 4— 7 mm. longis, juventute cinereo-puberulis mox glabris; laminis glabris chartaceis ovatis, 4.5-6 em. longis, 1.5-2 cm. latis, basi obtusis, apice longe caudato-acuminatis (acumine gracili 1.5— 2 em. longo acuto), margine integris et saepe leviter recurvatis, 3(obseure 5)-pli-nerviis, costa supra paullo impressa subtus elevata, nervis secundariis prope basim orientibus adscendent- ibus supra subplanis vel leviter impressis subtus prominulis, rete venularum supra saepe prominulo subtus immerso; in- florescentiis praeter corollas glabris axillaribus vel subter- minalibus racemosis 6-8-Нот1з, basi decidue bracteatis, pedun- culo brevi, rhachi gracili leviter angulata 1.2-3.5 em. longa; bracteis floriferis conspicuis papyraceis vel submembranaceis obovato-oblongis, 16-25 mm. longis, 4—10 mm. latis, basi an- gustatis, apice obtusis, utrinsecus glandulas circiter 10 mar- gine gerentibus; pedicellis rectis 4-6 mm. longis basim versus bibraeteolatis (bracteolis papyraceis oblongo-lanceolatis 1.5— 2 mm. longis saepe glandulosis) apicem versus glandulas pau- eas magnas (0.2-0.3 mm. diametro) sessiles gerentibus; calyce 3.5-4 mm. longo, tubo cylindrico 2-3 mm. diametro basi obscure apophysato et velut pedicellis glanduloso, limbo papyraceo suberecto 2-3 mm. longo, lobis 5 deltoideis subacutis 0.7-1 mm. [Vor. 28 450 ANNALS OF THE MISSOURI BOTANICAL GARDEN longis textura omnino glandulosis, sinibus complanatis; cor- olla (unica immatura visa) cylindrica albo-puberulente; stylo persistente gracili 20-22 mm. longo, stigmate obscure peltato. —cHIRIQUI: epiphytic shrub, the bracts rose, the flowers pale lavender-blue; vicinity of Bajo Chorro, alt. 1900 m., July 20- 22, 1940, Woodson € Schery 688 (NY, түрк); Boquete Region, von Hagen 2033 (A, NY). Although the cited specimens lack mature corollas, I have no doubt that a new species is represented, as the essential characters of the inflorescence are found in the bracts, pedicels, and calyces. From the persistent styles, one may assume that the corolla is about 2 em. long. Cavendishia chiriquiensis is a very distinct species, char- acterized by its small caudate-acuminate, few-nerved leaf- blades, its glandular-margined bracts, and its glandular calyces and pedicels. Among Central American species, it is suggestive only of C. Endresii Hemsl., from which it differs in the above-mentioned characters, its shorter pedicels, more conspicuous bracteoles, and presumably longer corollas. Other relatives are the Colombian C. rhynchophylla A. C. Smith and C. Purdiei A. C. Smith, but these differ in obvious features of foliage and inflorescence. CAVENDISHIA MICONIOIDES A. C. Smith, Contr. U. S. Nat. Herb. 28: 503. 1932. (Thibaudia melastomoides H. B. K. Nov. Gen. & Sp. 3: 272. 1818; Cavendishia melastomoides Hoer. Bot. Jahrb. 42: 279. 1909; non Hemsl. 1881).—onurmuiqví: tree 5 m. high ; corolla red, white-tipped ; between Casita Alta and Cerro Copete, alt. 2300-3300 m., Woodson & Schery 357 (NY). The species has previously been known from numerous col- lections in Colombia and Ecuador. The cited specimen from Panama has the leaves slightly smaller than the average of those from South America, but I find no consequential separat- ing characters. CaVENDISHIA PUBESCENS (Н. B. К.) Hemsl. Biol. Centr.-Amer. Bot. 2: 273. 1881; A. C. Smith, Contr. U. S. Nat. Herb. 28: 485. 1932.—cuirigvi: shrub 2-3 m. high; bracts pale pink; 1941] WOODSON & SCHERY—FLORA OF PANAMA. V 451 corolla white, faintly pink-tinged; Boquete Region, alt. about 1150 m., Davidson 649 (A). In making the above-cited combination, Hemsley cited See- mann 1613 from Boquete, but, since I had not seen any ma- terial of the species from Panama in 1932, I did not include Panama in the range of this common South American species. Saryria Allenii sp. nov. Frutex epiphyticus multiramosus ubique praeter inflorescentiam glaber, ramis ad 2.5 m. longis, ramulis erassis subteretibus (juventute angulatis) demum cinereis; petiolis rugosis erassis 6-10 mm. longis; laminis coriaceis siccitate fuscis elliptico-oblongis, 6-9 em. longis, 2— 3 em. latis, basi acutis vel attenuatis, apice cuspidatis (acumine ad 4 mm. longo obtuso), margine integris et saepe conspicue recurvatis, 5-pli-nerviis, nervis interioribus 1-2 em. supra basim orientibus cum costa supra impressis subtus prominent- ibus, nervis exterioribus paullo debilioribus, venulis immer- sis; inflorescentiis e ramulis defoliatis orientibus breviter race- mosis ut videtur 2-6-floris, rhachi gracili glabra ad 8 mm. longa, pedicellis gracilibus (apicem versus conspicue gradatim incrassatis) 9-11 mm. longis distaliter albo-puberulis, basi bracteis oblongis circiter 1.5 mm. longis subtentis, prope me- dium inconspicue bibraeteolatis; calyce ubique extus pilis patentibus circiter 0.2 mm. longis persistenter albo-puberulo, tubo circiter 2 mm. longo et 3 mm. diametro, limbo erecto- patente circiter 2 mm. longo tenuiter carnoso, lobis 5 deltoideis acutis circiter 1 mm. longis et 2.5 mm. latis; corolla carnosa urceolata, 10-13 mm. longa, basim versus 4-6 mm. diametro, apicem versus ad 2-3 mm. diametro contracta, distaliter ut calyce albo-puberula, lobis 5 acutis circiter 0.7 mm. longis et 1.3 mm. latis; staminibus 10 alternatim circiter 4 mm. et 4.5 mm. longis, filamentis glabris in tubo membranaceo pallido eireiter 2.5 mm. longo connatis, antheris alternatim circiter 2 mm. et 2.3 mm. longis, tubulis latis poris circiter 0.6 mm. longis dehiscentibus; stylo gracili corollam subaequante, stigmate minuto subpeltato.—cocLÉ: much-branched epiphytic shrub, the branches to 2.5 m. long, the flowers arising from old branch- lets ; pedicel and calyx coral-pink; corolla coral-pink, the distal [Vor. 28 452 ANNALS OF THE MISSOURI BOTANICAL GARDEN portion pure white; vicinity of La Mesa, north of El Valle de Antón, alt. 1000 m., April 12, 1941, Allen 2367 (A, түрк). Satyria Alleni is closely related only to S. meiantha Donn. Sm. of Guatemala and British Honduras, from which it differs in its smaller leaf-blades with more highly connate secondary nerves and completely immersed veinlets, its persistently puberulent calyx, and its smaller anthers. The common Cen- tral American S. Warszewiczw Kl. differs from S. Allenii in its long pedicels, glabrous flowers, and much longer corollas and anthers. Satyria ovata A. C. Smith differs from the new species in its proportionately broader and obtuse leaf-blades, apparent veinlets, elongate corollas, ete. DisrERIGMA Номвоготи (Kl. Nied. Bot. Jahrb. 11: 224. 1889; А. C. Smith, Brittonia 1: 222. 1933.—curriqui: epi- phytie shrubs with pale pink flowers; vicinity of Bajo Chorro, alt. 1900 m., Woodson & Schery 633 (NY), 684 (NY). Although not otherwise reported from Panama, the species occurs in some abundance in the adjacent parts of Costa Rica and also in Colombia; it will doubtless be found in other parts of Panama. THEOPHRASTACEAE (C. L. Lundell, Ann Arbor, Mich.) Cuavisa Alleni Lundell, sp. nov. Arbor, ramulis crassis. Folia petiolata, petiolo ad 2 em. longo, lamina glabra, sub- eoriacea, anguste cuneato-oblanceolata, 30-43 em. longa, 6— 7.5 em. lata, apice acuta vel subacuminata, basi anguste cune- ata, margine remote serrulata. Inflorescentiae з pendulae, sub- pauciflorae, ad 36 cm. longae, parce minute puberulae. Pedi- celli ad 2 mm. longi. Flores 4-meri. Sepala basi coalita, sub- orbieularia, 1-1.2 mm. longa, fimbrillata. Corolla ea. 1 em. diam., 6 mm. longa, aurantiaca, lobis suborbieularibus. Ova- rium floris в substipitiforme. A tree, 10 m. tall, branchlets thick, about 1.2 em. in diam., with the leaves crowded at the apex, obscurely and minutely furfuraceous. Petioles thick, up to 2 cm. long, the basal half drying blackish. Leaf blades glabrous, subcoriaceous, slightly paler beneath, narrowly cuneate-oblanceolate, 30-43 ст. long, 1941] WOODSON & SCHERY—FLORA OF PANAMA. V 453 6-7.5 em. wide, apex acute or subacuminate, base narrowly cuneate, decurrent, margin remotely serrulate, the teeth not over 1 mm. long, veinlets reticulate on both surfaces. Stami- nate racemes borne chiefly below the leaves on old wood, slen- der, pendulous, up to 36 em. long, sparsely puberulent, with few scattered flowers. Bractlets minute, triangular, about 0.5 mm. long, minutely puberulent. Pedicels reflexed, minutely puberulent, up to 2 mm. long, usually about 1.4 mm. long. Flowers orange-colored, 4-merous. Sepals suborbicular, 1-1.2 mm. long, fimbrillate. Corolla about 1 em. in diam., 6 mm. long, the basal third connate, the lobes suborbicular. Staminodia alternate with the petals, small. Stamens united into a tube, conglutinate, 8-dentate from above. Abortive ovary substipiti- form.—cANAL ZONE: vicinity of Madden Dam, alt. 100 m., Aug. 1, 1941, P. H. Allen 2671 (Herb. Missouri Bot. Gard., ТУРЕ; Herb. Univ. Michigan, fragment of TYPE). The narrowly cuneate-oblanceolate serrulate leaf blades, long pendulous puberulent staminate racemes, and short puberulent reflexed pedicels characterize the species. The other members of the genus known from Central America all have entire leaves. MYRSINACEAE (C. L. Lundell, Ann Arbor, Mich.) Arpis1a coclensis Lundell, sp. nov. Arbor parva, ramulis erassis, furfuraceis. Folia petiolata, petiolo ad 1.5 em. longo, lamina subcoriacea, integra, cuneato-oblanceolata vel elliptico- oblonga, 15-22 ст. longa, 5.6-8.2 ст. lata, apice acuta, basi euneata. Inflorescentiae terminales, pyramidales, corymbosae, lepidotae. Pedicelli ad 1.8 em. longi. Sepala 5, late ovata, 3-3.5 mm. longa, ciliata, punctata. Petala 5, lanceolato-oblonga, 10— 11 mm. longa, acuta, reflexa. Stamina ca. 9 mm. longa. A. small tree, 3 m. high; branchlets thick, furfuraceous, the scales brownish, compact. Petioles stout, up to 1.5 em. long. Leaves elustered at apex of branchlet, the blades pallid, sub- coriaceous, entire, cuneate-oblanceolate or elliptic-oblong, 15- 22 em. long, 5.6-8.2 ст. wide, apex acute, base cuneate, decur- rent, sparsely lepidote beneath, costa slightly impressed [Vor. 28 454 ANNALS OF THE MISSOURI BOTANICAL GARDEN above, prominent beneath, the primary lateral veins fine but evident on both surfaces, obscurely reticulate. Inflorescence brown-lepidote, terminal, pyramidal, weak with slender rachis and primary branches, 22 cm. long, 18 cm. wide, the flowers corymbose, comparatively few, large, rose-pink. Pedicels lepidote, slender, up to 1.8 em. long. Sepals 5, free almost to base, broadly ovate, 3-3.5 mm. long, apex rounded and obliquely emarginate, lepidote and punctate, ciliate. Petals 5, cohering at base only, lanceolate-oblong, 10-11 mm. long, acute, incon- spicuously punctate, reflexed at anthesis. Stamens about 9 mm. long; filaments united into a tube about 2 mm. long; anthers sessile, erect, lanceolate, acuminate, cohering at base. Ovary subglobose, glabrous; style equaling stamens.—cocr£: vicinity of El Valle de Antón, alt. about 600 m., P. H. Allen 2056 (Herb. Univ. Michigan, TYPE). A. coclensis, apparently referable to the subgenus Icacorea, has peculiar stamen characteristics somewhat suggestive of the subgenus Walleniopsis. The filaments coalesce into a tube and the erect sessile anthers adhere at base. The species may have affinity to A. Pittieri Mez, a plant known to the writer only from Standley’s brief description (Field Mus. Publ. Bot. 18: 893. 1938). Акшал geniculata Lundell, sp. nov. Arbor parva, ramulis furfuraceis. Folia petiolata, petiolo 3-6 mm. longo, lamina membranacea, anguste elliptiea vel oblanceolato-elliptica, 5.8— 11 em. longa, 2.5—4.3 em. lata, apice subabrupte acuminata, acumine obtuso, basi attenuata, parce lepidota. Inflorescentiae geniculatae, terminales, lepidotae, subumbellatae. Pedicelli ad 10 mm. longi. Sepala 5, late ovata, 1 mm. longa, subintegra. Petala 5, lineari-oblonga, 5 mm. longa. Stamina 3.5 mm. longa. A tree, 4 to 5 m. high; branchlets slender, furfuraceous, fer- ruginous. Petioles furfuraceous, canaliculate, 3-6 mm. long. Leaf blades membranaceous, paler beneath, narrowly elliptic or oblanceolate-elliptie, 5.8-11 em. long, 2.54.3 em. wide, apex subabruptly acuminate, the acumen obtusish, base attenuate, decurrent, sparsely lepidote, costa plane above, prominent be- neath, primary veins 12-16 on each side, conspicuous beneath. 1941] WOODSON & SCHERY—FLORA OF PANAMA. V 455 Inflorescence terminal, abruptly geniculate at base, lepidote, bipinnate, few-flowered, less than 4 cm. long, the flowers sub- umbellate, white. Pedicels slender, up to 10 mm. long. Sepals 9, broadly ovate, 1 mm. long, prominently punctate, the margin scarious, subentire, apex rounded, obscurely emarginate laterally. Petals 5, linear-oblong, slightly wider above middle, o mm. long, united (1 mm.) at base, acutish, orange-punctate. Stamens about 3.5 mm. long; filaments stout, less than 1 mm. long; anthers linear-lanceolate, concolorous, abruptly apicu- late, dehiscent apically. Ovary glabrous; style slender, 4 mm. long.—cuHiriqgui: vicinity of San Bartolomé, Peninsula de Buriea, alt. up to 5 m., July 28-Aug. 1, 1940, Woodson Ё Schery 944 (Herb. Univ. Michigan, TYPE). A. geniculata, referable to the subgenus Icacorea, is note- worthy for its geniculate inflorescence. Arpisia rigidifolia Lundell, sp. nov. Arbor parva, ramulis erassiusculis. Folia petiolata, petiolo crassiusculo, 3-5 mm. longo, lamina rigide coriacea, pallida, elliptica, 3.5-6.5 cm. longa, 2-3.2 ст. lata, apice obtusa, basi late cuneata. Inflores- centiae terminales, pyramidales, ad 7 cm. longae et latae, fur- furaceae, umbellatae. Pedicelli 4—5 mm. longi. Sepala 5, ovata, 1-1.3 mm. longa, minute erosa. Petala 5, oblonga, 5 mm. longa, apice acutiuscula. Stamina са. 3.5 mm. longa. Stylus 5.2 mm. longus. A small tree, 2.5 m. high; buds and tips of branchlets dark reddish brown, furfuraceous ; branchlets rather thick, gnarled, with very short internodes. Petioles stout, 3-5 mm. long. Leaves clustered at ends of branchlets, the blades rigidly cori- aceous, ‘pallid, paler on under surface, minutely punctate, el- liptie, 3.5—6.5 cm. long, 2-3.2 em. wide, apex obtuse or obtusely subacuminate, base broadly cuneate, decurrent, costa plane above, prominent beneath, primary veins slender, prominulous beneath. Inflorescence terminal, tripinnately paniculate, py- ramidal, up to 7 cm. long and wide, furfuraceous, many- flowered, the flowers white, umbellate; bractlets up to 1 cm. long, lepidote. Pedicels 4-5 mm. long. Sepals 5, ovate, 1-1.3 mm. long, punctate with few conspicuous orange-red glands, the [Vor. 28 456 ANNALS OF THE MISSOURI BOTANICAL GARDEN margin scarious, minutely erose. Petals 5, oblong, 5 mm. long, apex acutish, asymmetrical, punctate with few inconspicuous glands. Stamens about 3.5 mm. long; filaments about 1 mm. long, thick; anthers lanceolate, concolorous, apically dehiscent, not apiculate. Ovary glabrous; style 5.2 mm. long.—cocr£: vi- cinity of El Valle, alt. 800-1000 m., Dec. 22, 1936, P. H. Allen 71 (Herb. Univ. Michigan, TYPE). A, rigidifolia is a none too well marked species related to А. minor Standl., and referable to the subgenus Icacorea. From A. minor it differs in having gnarled branchlets, small rigidly coriaceous leaves, comparatively large tripinnately paniculate furfuraceous inflorescences, smaller sepals, and a longer style. Arpisia Scheryi Lundell, sp. nov. Arbor glabra, ramulis crassiusculis. Folia petiolata, petiolo ad 8 mm. longo, lamina glabra, integra, coriacea, lanceolata vel lanceolato-elliptica, 4.5-7.5 em. longa, 1.6-3.2 em. lata, apice acuminata, basi acutiuscula. Inflorescentiae terminales, pyramidales, multi- florae, umbellatae. Pedicelli 2-3 mm. longi. Sepala fere 1 mm. longa, subintegra. Corolla glabra, 4 mm. longa. Stamina 3 mm. longa. Antherae latae, oblongo-elliptieae, apice rotundatae et minute emarginatae. Ovarium glabrum. A tree, 20 m. high; branchlets reddish, rather stout, glabrous. Petioles narrowly winged, up to 8 mm. long. Leaf blades en- tire, glabrous, coriaceous, drying reddish-brown, paler be- neath, lanceolate or lanceolate-elliptie, 4.5-7.5 em. long, 1.6-3.2 cm. wide, apex attenuate, acuminate, base acutish, decurrent, primary veins conspicuous on both surfaces. Inflorescence terminal, pyramidal, up to 14.5 em. long, 13.5 em. wide, gla- brous, reddish-brown ; the flowers pale pink, numerous, umbel- late. Pedicels slender, 2-3 mm. long, glabrous. Flower buds about 3 mm. long. Sepals scarcely 1 mm. long, ovate, subentire, scarious, prominently orange-punctate. Corolla 4 mm. long, the petals short connate at base, lanceolate-oblong, acutish, orange-punctate, glabrous. Stamens 3 mm. long; anthers ob- long-elliptic, nearly twice as long as filaments, rounded and shallowly notched at apex. Ovary globose, glabrous.—cnurni- 1941] WOODSON & SCHERY— FLORA OF PANAMA. V 451 901: vicinity of Bajo Chorro, alt. 1900 m., July 20-22, 1940, Woodson & Schery 686 (Herb. Univ. Michigan, TYPE). A. Scheryi, which belongs to Ше subgenus Icacorea, is allied to A. multiflora Griseb. and A. pallidiflora Standl. From the two, A. Scheryi may be recognized immediately by its short pedicels. The tree is remarkably large for the genus. AnnisIA SEIBERTI Standl., Ann. Missouri Bot. Gard. 24: 198. 1937.—cocrÉ: flowers white; El Valle de Antón and vicinity, alt. 500—700 m., July 23-27, 1935, R. J. Seibert 456 (Herb. Field Mus., TYPE) ; a small tree, 3 m. high, fleshy calyx pink ; vicinity of El Valle, alt. 800-1000 m., Dec. 22, 1936, P. H. Allen 72; а small tree, 2 m., flowers pale pink; vicinity of El Valle, alt. 800— 1000 m., Sept. 5, 1938, Allen 786; a tree, 8 m. high, calyx and pedicels white, corolla pink; between Las Margaritas and El Valle, July 15-Aug. 8, 1938, В. E. Woodson, Jr., Р. H. Allen £ R. J. Seibert 1239; a shrub, 3 m. high, petals white, faint laven- der-tinged ; same locality and date, Woodson, Allen & Seibert 1746; a small tree, 4 m. high, flowers pale lavender ; north rim of El Valle, June 4, 1939, Р. H. Allen & А. H. G. Alston 1846; PANAMA: a small tree, 3 m. high, flowers pink; summit of Cerro Campana, alt. 800-1000 m., Sept. 1, 1940, Allen 2226. A. Seibertii is very close to A. opegrapha Oerst. of Ше sub- genus Graphardisia. A. opegrapha has larger narrower ob- lanceolate rather than elliptic or oblanceolate-elliptie leaves, somewhat less ample inflorescences, shorter fruiting pedicels, and narrower smaller sepals. These are differences of perhaps minor importance. For the loan of the type of A. Seibertii, an isotype of A. opegrapha, and other material of these in the Field Museum Herbarium the writer is indebted to the Director, Mr. C. C. Gregg. Arpista Woodsoni Lundell, sp. nov. Arbor parva. Folia petiolata, petiolo crassiusculo, ad 6 mm. longo, lamina coriacea, integra, 5-7.5 em. longa, 1.7-2.8 em. lata, apice obtusa, basi late cuneata. Inflorescentiae terminales, са. 5 ст. longae, glabrae, subeorymbosae. Pedicelli 1-1.3 ст. longi. Sepala 5, [Vor. 28 458 ANNALS OF THE MISSOURI BOTANICAL GARDEN late ovata, ca. 2 mm. longa, subintegra. Bacca globosa, 6 mm. diam. A tree, 8 m. high, buds and tips of branchlets furfuraceous, glabrous otherwise; branchlets slender, terete and brown. Petioles comparatively stout, marginate to base, up to 6 mm. long. Leaf blades coriaceous, paler beneath, entire, 5-7.5 em. long, 1.7-2.8 em. wide, apex narrowed to an obtuse point, base broadly cuneate, decurrent, costa plane above, slightly elevated on lower surface, primary veins slender, evident but not con- spicuous. Inflorescence terminal, about 5 ет. long, few- flowered, the flowers subeorymbose. Pedicels slender, 1-1.3 em. long, slightly curved. Sepals 5, broadly ovate, about 2 mm. long, rounded at apex, punctate with conspicuous orange-red dots, the margin scarious, minutely erose, essentially entire. Berries globose, 6 mm. in diam., black-purple when ripe.— CHIRIQUÍ: vicinity of Finca Lérida, alt. 1750 m., July 7-11, 1940, Woodson & Schery 230 (Herb. Univ. Michigan, TYPE). А. Woodson belongs to the subgenus Icacorea where its nearest relative appears to be A. minor Standl. A. minor is de- scribed as having stout pedicels 2-5 mm. long, while the pedi- cels of A. Woodson are slender and usually exceed 1 cm. in length. PanaTHESIS Woodsoni Lundell, sp. nov. Arbor, ramulis fur- furaceis. Folia petiolata, petiolo 7-12 mm. longo, lamina chartacea, oblanceolata, 5-9 em. longa, 1.5-2.6 em. lata, apice acuminata, acumine obtusiusculo, basi acuta, subintegra, supra glabra, subtus parce lepidota. Inflorescentiae axillares, ad 12 em. longae, parce lepidotae. Pedicelli 7-11 mm. longi. Se- pala 5, anguste triangularia, 1-1.2 mm. longa, punctata. Petala 5, extus puberula, intus tomentosa, punctata, lanceolato- attenuata, 6 mm. longa. Stamina 3-3.5 mm. longa, filamentis glabris, antheris ca. 2.3 mm. longis, haud punctatis. Ovarium ad apicem tomentellum; stylus basi tomentellus. Ваеса de- presso-globosa. A tree; branchlets furfuraceous, ferruginous, rather slender, with short internodes and conspicuous leaf scars. Petioles fur- furaceous, canalieulate, 7-12 mm. long. Leaf blades thinly 1941] WOODSON 4 SCHERY—FLORA OF PANAMA. V 459 chartaceous, oblanceolate, 5-9 cm. long, 1.5-2.6 ст. wide, apex short-acuminate, the acumen obtusish, base acute, slightly de- current, margin somewhat repand, essentially entire, rather sparsely lepidote beneath, minutely but conspicuously black- punctate, costa slightly impressed above, prominent beneath, the primary veins slender, 11-14 on each side, prominulous. Inflorescence axillary, up to 12 cm. long, with long slender peduncle, sparsely lepidote, ferruginous. Pedicels slender, 7-11 mm. long. Sepals puberulent, punctate, narrowly tri- angular, 1-1.2 mm. long. Corolla pink, the petals puberulent outside, tomentose within, conspicuously linear-punctate, lanceolate-attenuate, widest at base, 6 mm. long. Stamens 3-3.5 mm. long; filaments glabrous, subequaling anthers; anthers about 2.3 mm. long, base sagittate, attenuate to an acutish apex, concolorous, not punctate. Apical half of ovary and base of style tomentellous, the style about 5 mm. long, linear- punctate. Berries deep purple, depressed-globose, about 7 mm. in diam.—cuirigui: Finca Lérida to Peña Blanca, 1750-2000 m., July 9, 1940, Woodson 6 Schery 331 (Herb. Univ. Michigan, TYPE). P. Woodsoni, closely allied to P. Seibert&i Lundell, differs in its smaller leaves, considerably longer pedicels, larger flowers, and concolorous anthers. The anthers are not punctate, and this is probably the most distinctive characteristic of the plant. GENTIANACEAE (Caroline K. Allen, Jamaica Plain, Mass.) Harena euryphylla, n. sp. Herba perennis (?), caule sim- pliei ramoso procumbente, inferiori reliquiis foliorum in- structo, ad 50 em. alto, paullo alato ; internodiis inferioribus ad 2 cm., superioribus ad 7 em. longis; foliis sessilibus, superiori- bus caulinis basi petiolum 0.5-1.5 em. longum foliaceum simu- lantibus, longe attenuatis, lamina elliptica conspicue apiculata, 3-nerviis, nervio medio prominente, ad 7.5 cm. longis et 2.7 em. latis; inflorescentia terminali raro axillari cymosa aliquid laxa et pauciflora, ad 6.5 em. longa; calyce ad 1 cm. longo et corollae longitudinem ca. ?4 aequante; lobis 3-nerviis oblance- olatis longe apiculatis minute papillosis ; corolla (calcaribus in- (Vor. 28 460 ANNALS OF THE MISSOURI BOTANICAL GARDEN clusis) ad 1.5 em. longa, ad 1 em. lata, viridescenti-flava (fide collectoris), tubo totae corollae longitudinem 14 vel minus aequante; corollae lobis elliptieis, margine eroso; calcaribus corollae longitudinem fere aequantibus divaricatis leviter pendulis; eapsula ovata ad 17 em. longa, stylo apice longe apieulato; seminibus ellipsoideis 1.25 mm. longis.—cCHIRIQUÍ: Casita Alta to Cerro Copete, ca. 2300-3300 m., July 10, 1940, Woodson € Schery 362 (Gray Herb., туре, Herb. Missouri Bot. Gard., ISOTYPE). The foliose stem and the general habit of this new species recall the widespread North American species H. deflexa. Actually H. euryphylla can claim close relationship with H. guatemalensis Loesener, but differs in its broad leaves, looser inflorescence and its flowers with divaricate, slightly pendu- lous, scarcely incurved spurs. (J. A. Steyermark, Chicago) Voyria Allenii Steyermark, sp. nov. Caulis crassiusculus 9-13 em. altus, internodiis superioribus 4-5 mm. longis, in- ferioribus 13-17 mm. longis; flores vulgo solitarii et ramos terminantes, pedunculis 5-9 mm. longis; calycis tubus cam- panulatus 4.5-5 mm. longus, lobis 5 late ovatis obtusis minute ciliolatis 1.5-3 mm. longis; corolla laete rosea hypocrateri- formis extus ubique minute puberula, 3.5-4 cm. longa, lobis elliptieo-ovatis obtusis 1.4—1.5 em. longis 0.8-0.9 em. latis; fila- menta brevissima, antheris subgloboso-rhomboideis 1.2-1.3 mm. longis ; stylus gracilis 2.7 em. longus. Stem stout, 9-13 em. high, 2-2.5 mm. thick, once-branched in the upper half; scales opposite, dull red, those on the upper floriferous stems 4-5 mm. apart, the ones in the lower portion of the stems 13-17 mm. apart, broadly ovate, 4—5 mm. long, 3—4 mm. broad, the sinus subacute, obtuse to rounded, minutely ciliolate, connate in lower half to third, glabrous; usually a solitary bright pink flower terminating each branch of the stem; peduncles 5-9 mm. long, stout; calyx-tube campanulate, 4.5-5 mm. long, 3-6 mm. wide at the throat, lobes 5, minutely ciliolate, broadly ovate, obtuse, 1.5-3 mm. long, 1.5-8 mm. broad 1941] WOODSON & SCHERY—FLORA OF PANAMA. V 461 at the base, subequal; corolla bright pink, large and showy, hypocrateriform, broadened at the base around the ovary and slightly constricted in the upper portion below the insertion of the stamens, minutely puberulous over the entire outer sur- face, 3.5—4 em. long, broadened to 5-6 mm. around the ovary, slightly narrower in upper half, lobes elliptic-ovate, obtuse, the margins round-eurved and narrowed at the base, 1.4—1.5 em. long, 0.8-0.9 em. broad, minutely papillate without, minutely ciliolate; filaments inserted in the upper half of the tube 6-7 mm. below the throat, very short, 0.5 mm. long; anthers sub- globose-rhomboid, 1.2-1.8 mm. long; style slender, 2.7 em. long, glabrous; stigma dome-shaped, sinuate-peltate; ovary glabrous, ellipsoid, 8-9 mm. long, in fruit 1.5 em. long, 5 mm. in diameter.—cocrÉ: hills north of El Valle de Antón, alt. 800 m., Sept. 8, 1940, Allen 2240 (Herb. Field Mus., TYPE). The genus Voyria, as delimited by Aublet and excluding species of Leiphaimos which have sometimes been merged with it, comprises a few rare and local species confined to Colombia and British Guiana. The genus has not hitherto been reported for Central America. Voyria Allenii is most closely related to V. macrantha Killip! described from Co- lombia, from which it differs in its shorter calyx-tube, shorter corolla-lobes, pink instead of red-purple corolla and corolla- tube puberulent throughout on the outside. From V. rho- dochroa Sandwith,? it differs in the flowers borne solitary and terminal on the one or two branches, not fasciculate-corymbose, and in the larger corolla-lobes. From V. rosea Aubl. it differs in its obtuse calyx- and corolla-lobes and solitary terminally borne flowers, while from V. caerulea Aubl. it differs in its scales and ealyx-lobes rounded and not acute, in its pink instead of blue or violet corollas, and fewer-branched stems. APOCYNACEAE STEMMADENIA Allenii Woodson, spec. nov. Arbuscula са. 3 m. alta; ramulis gracilibus dichotome compositis omnino glabris. Folia opposita elliptica apice acute subeaudato- 1 Journ. Wash. Acad. Sci. 26: 361. 1936. з Kew Bull. 1931: 55. 1931. [Vor. 28 462 ANNALS OF THE MISSOURI BOTANICAL GARDEN acuminata basi in petiolo brevissimo ca. 0.5 em. longo gradatim attenuata petiolo incluso 6-15 em. longa 2.0—5.5 em. lata mem- branacea glabra. Inflorescentia 1—4-flora; pedunculo са. 1 em. longo; pedicellis са. 0.5 em. longis omnino glabris. Calycis lobi oblongi obtusi 0.6-0.9 em. longi subfoliacei glabri. Corolla *'fere alba, centro dilute flava’’; tubo proprio са. 1.25 em. longo basi ea. 0.3 em. diam. apicem versus attenuato ibique et stam- inigero et sinistrorse contorto ; faucibus conicis ca. 2 em. longis ostio ea. 1 em. diam.; lobis oblique obovatis ea. 1.5 em. longis patulis. Folliculi није gibbosi angustissime attenuati ca. 3.5 em. longi medio са. 0.7 em. erassi.—cocré: hills north of El Valle de Antón, alt. ca. 1000 m., July 14, 1940, P. H. Allen 2187 (Herb. Miésoürt Bot. Gard., et ; same locality, alt. 800 m., Sept. 8, 1940, Allen 2239 (Herb. Missouri Bot. Gard., co- TYPE). At first glanee, this species was mistaken for S. Alfari (Donn. Sm.) Woodson, which, however, has much smaller calyx lobes, and smaller corollas with narrower throat and shorter lobes. Mr. Allen has taken considerable interest in this species, and finds it to be fairly abundant in the neighborhood of the type locality. ASCLEPIADACEAE Сокоговов chiriquensis Woodson, comb. nov. (Vince- toxicum chiriquense Woodson, in Woodson & Seibert, Ann. Missouri Bot. Gard. 24: 199. 1937). This combination will be made necessary by the impending conservation of Vincetoxicum Moench over Vincetoxicum Walt. (see Perry, Rhodora 40: 281—282. 1938). Сохоговов fuscoviolaceus Woodson, spec. nov. Frutex vol- ubilis ; ramulis gracilibus glabris, internodiis 7-12 em. longis. Folia petiolata ovato-oblonga basi late rotundata fere cordata apice subcaudato-acuminata 4.5-11.0 ст. longa 1.5-4.5 cm. lata firme membranacea glabra subtus pallidiora, petiolo 1.5— 2.5 em. longo. Inflorescentia bostrycino-umbellata pluriflora; peduneulo ca. 1 em. longo glabro; pedicellis ca. 0.5 em. longis glabris; braeteis vix manifestis. Flores viridi-fusci; calycis 1941] WOODSON & SCHERY—FLORA OF PANAMA. V 463 laciniis ovato-oblongis acutis glabris 0.25 em. longis; corolla rotata ca. 0.9 em. diam. faucibus minute pilosulis caeterumque glabra, lobis late ovatis obtusis minute emarginatis patulis; gynostegio subsessili ca. 0.15 ст. diam., stigmate depresso, polliniis cochleariformibus caudiculis inclusis ea. 0.05 em. longis; corona rotata margine leviter crenulata gynostegium vix aequante. Folliculi immaturi fusiformes laeves glabri.— CHIRIQUÍ: vicinity of Bajo Chorro, alt. 1900 m., July 20-22, 1940, Woodson & Schery 695 (Herb. Missouri Bot. Gard., TYPE). Most closely related to the preceding, but differing in the conspicuously smaller leaves and flowers. These two species differ from those of typical Gonolobus in the fact that the anthers are placed about the margin of the stigma head and not beneath it, together with other rather slight characters which probably should be viewed as subgeneric rather than as generic. Similar characters have not been found to be generic in other Gonolobeae. CONVOLVULACEAE IPoMoEA DEMERARIANA Choisy (determination from herbar- ium comparisons, but type not available). Probably the same as Ipomoea cardiosepala Meisn. of Colombia, a later name published in Mart. Fl. Bras. Also specimens of this species can be found in herbaria evidently incorrectly labeled as Ipomoea phillomega (Vell. House.—socas DEL Toro: vicinity of Isla Colon, alt. 0-120 m., Aug. 23, 1940, H. von Wedel 514. Previ- ously known from South America and the West Indies. ТрРомоЕА mucronata Schery, spec. nov. Suffrutex volubilis in fruticibus; caulibus, pedunculis petiolisque summe pubes- centibus; laminis foliorum integris, lanceolatis, profunde cordatis, 4-9 em. longis, prope basim 4-9 cm. latis, utrinque pubeseentibus praesertim in venulis, petiolis quam laminis longioribus, 4-14 em. longis; pedunculis quam petiolis longi- oribus, 10-25 em. longis; inflorescentiis subeorymbiformibus, floribus multis vel paucis, terminalibus 3.5-4.5 cm. longis, puniceo-purpureis, pedicellis pubescentibus floribus breviori- bus (plerumque са. 1 ст. longis); sepalis subaequalibus, [Vor. 28 464 ANNALS OF THE MISSOURI BOTANICAL GARDEN pilosis, oblongo-ellipticis, apice subulatis vel mucronatis, cori- aceis, 7-11 mm. longis, 2-4 mm. latis; stylis persistentibus ; capsulis dehiscentibus, bicubieulis, apice leviter pilosis; sem- inibus glabris, suborbicularibus, ca. 4 mm. diametro cum locis figitis orbicularibus.—Panama: cmuiuqví: near Peña Blanca, alt. 1750-2000 m., July 9, 1940, Woodson & Schery 283, 323 (Herb. Missouri Bot. Gard., түре) ; Cosra Rica: san José: vi- cinity of El General, alt. 850 m., Feb. 1936, А. Е. Skutch 2570. This species is close to I. trichocarpa Ell. and I. longipes Garcke, but differs chiefly in having dense pubescence, laven- der or rose flowers and pilose sepals. The outer mucronate sepals are slightly smaller than the inner ones. The 2-celled capsule dehisces longitudinally, usually into three parts. In the specimen examined, only one seed developed in each сар- sule. The seed has a circular, raised, tire-like area at the point of attachment. VERBENACEAE (H. N. Moldenke, New York) CITHAREXYLUM HIRTELLUM Standl.—curriguf: vicinity of Boquete, alt. 1200-1500 m., July 24-26, 1940, Woodson d Schery 755. LABIATAE (C. C. Epling, Los Angeles) SALVIA IRAZUENSIS Fern.—currigui: Potrero Muleto, Volcan de Chiriquí, alt. 10500 m., July 13, 1940, Woodson & Schery 386. Previously considered to be an endemie of Costa Riea. SOLANACEAE (C. V. Morton, Washington, D. C.) SoLANUM SEAFORTHIANUM Andrews.—cniRIQUÍ: vicinity of Finca Lérida, alt. 1750 m., July 7, 1940, Woodson & Schery 239. New to Panama. Probably escaped from cultivation. SOLANUM QUINQUANGULARE Willd.—cuiriqui: Finca Lérida to Репа Blanca, alt. 1750-2000 m., July 9, 1940, Woodson d Schery 290. Previously known from northern South America. SOLANUM ROBLENSE Bitt.—cnmuquí: vicinity of Callejon Seco, Volean de Chiriquí, alt. 1700 m., July 17, 1940, Woodson & Schery 487. Previously known from Costa Rica. 1941] WOODSON & SCHERY—FLORA OF PANAMA. V 465 SOLANUM EVOLVULIFOLIUM (Greenm.—cHiRIiQUÍ: vicinity of Bajo Chorro, alt. 1900 m., July 20, 1940, Woodson & Schery 649. Previously known from Costa Rica. GESNERIACEAE (C. V. Morton, Washington, D. C.) DrYMONIA ALLOPLECTOIDES Hanst.—cocrLÉ: north rim of El Valle de Antón, near Cerro Turega, alt. 650-700 m., June 30, 1940, Woodson & Schery 162. New to Panama. Previously known from Costa Rica. ALLOPLECTUS ICHTHYODUMA var. PALLIDUS Morton.—cHIRI- 971: Casita Alta to Cerro Copete, Volcan de Chiriquí, alt. 2300- 3300 m., July 10, 1940, Woodson & Schery 359. Previously known from Costa Rica. CoLUMNEA CONSANGUINEA Hanst.—cuiriQut: vicinity of Bajo Chorro, alt. 1900 m., July 20, 1940, Woodson Ф Schery 651. Previously known from Costa Rica. САМРАМЕА Номвоготи (Kl.) Oerst.—currigui: vicinity of Bajo Chorro, alt. 1900 m., July 20, 1940, Woodson & Schery 656. Previously known from Costa Rica. BESLERIA PYCNOSUZYGIA Donn. Sm.—cnuriniQuí: vicinity of Bajo Chorro, alt. 1900 m., July 21, 1940, Woodson Ф Schery 683. Previously known from Costa Rica. ACHIMENES CANDIDA Lindl.—cutirigui: llanos, vicinity of Boquete, alt. 1200-1500 m., July 26, 1940, Woodson & Schery 764, Previously known from Guatemala and Costa Rica. ACANTHACEAE (E. C. Leonard, Washington) GrockERi1A reflexiflora Leonard, sp. nov. (fig. 1). Herba epi- phytica, caulibus subteretibus glabris vel ad nodos puberulis; lamina foliorum oblonga vel oblongo-lanceolata, acuminata, basi angustata, membranacea, undulata, glabra, in costa et venis parce strigosa; petioli breves; panicula grandis, puber- ula, ramis tenuibus ; bracteae subulatae ; pedicelli reflexi ; calyx glaber, segmentis anguste triangulatis ; corolla glabra, auranti- aca, curvata, bilabiata, labio superiore oblongo apice minute [Vor. 28 466 ANNALS OF THE MISSOURI BOTANICAL GARDEN bidentato et serrato, inferiore trilobo, lobis brevibus, latis; stamina exserta; ovarium glabrum. An epiphytic herb up to 1 meter high; stems subterete, gla- brous or puberulent at the nodes; leaf blades oblong or oblong- lanceolate (one leaf of the uppermost pair excepted, this ovate, а Glockeria reflexiflora Leonard: a, pair of leaves sub- tending p^ panicle to show asymmetry; b, portion of panicle; ealyx; d, corolla; e, upper lip; f, lower lip (a, b, half natural dei а. в. 7, twice natural size. subsessile, cordate at base), up to 13 em. long and 4.5 cm. wide, acuminate (the tip blunt), narrowed at base, thin, undulate, glabrous, or the costa and lateral veins (6 or 7 pairs) sparingly strigillose; petioles up to 5 mm. long; flowers numerous, re- flexed, borne in a puberulent terminal panicle 30 em. long and about 10 em. in diameter, the ultimate branches very slender; 1941] WOODSON & SCHERY—FLORA OF PANAMA. V 467 bracts subulate, up to 3 mm. long; pedicels slender, 2 to З mm. long, reflexed; calyx 4 mm. long, glabrous, the segments nar- rowly triangular, 3 mm. long and 0.5 mm. wide at base, the tips blunt; corolla about 2 em. long, orange, glabrous, slightly curved, 1 mm. in diameter at base, enlarged from the tip of the calyx to a tubular throat about 4 mm. in diameter, the upper lip erect, oblong, 5 mm. long, 3 mm. wide, rounded, minutely bidentate and serrate at tip, the lower lip short, truncate, with 3 low broad lobes; stamens exserted, the filaments flat, gla- brous; ovary glabrous; capsule not seen.—cuHIRIQui: vicinity of Bajo Mona and Quebrada Chiquero, alt. 1500 m., July 18, 1940, В. E. Woodson, Jr. € В. W. Schery 547 (U. S. Nat. Herb. No. 1,791,983, TYPE). This species is easily recognized by its large puberulous panicle of reflexed orange flowers. Dicuretera gracilis Leonard, sp. nov. (fig. 2). Herba, cauli- bus breviter striatis, glabris vel ad nodos tomentulosis ; lamina foliorum lanceolato-ovata, acuta vel acuminata, basi rotundata, membranacea, integra, glabra, costa et venis obscure strigosis ; petioli graciles; cymae tripartitae; pedunculi axillares graciles; bracteae capituli inaequales, coriaceae, ciliatae, intus puberulae, extus glabrae vel costa parce strigosa, bractea pos- terior linearis, obtusa, anterior lanceolata, apice rotundata; braeteolae lanceolatae, acutae, ciliatae, minute pubescentes; calycis segmenta anguste triangulata, minute pubescentia, leviter striata; corolla incarnata, pubescens, bilabiata, labio superiore acuminato, inferiore trilobo, lobis rotundatis; ovar- ium glabrum. Herbaceous; stems up to 30 em. high or more, ascending, 2 mm. in diameter or less, faintly striate, glabrous, or puberulous at the nodes; leaf blades lance-ovate, up to 3.5 em. long and 1.6 em. wide, acuminate at apex (the tip blunt), rounded at base and slightly decurrent on the petiole, thin, entire, glabrous ex- cept the costa and lateral veins (6 pairs), these inconspicuously and minutely strigose, more densely so above than beneath, sparingly ciliate, the cystoliths minute; petioles slender, up to 5 mm. long; cymes of 3 capitula on slender peduncles up to [Vor. 28 ANNALS OF THE MISSOURI BOTANICAL GARDEN 468 9 em. long, glabrous; pedicels 1 to 3 mm. long; bracts subtend- ing the cymes similar to the leaves but smaller, those subtend- mz "us э. Fig. 2. Dicliptera gracilis Leonard: a, portion of plant, natural size; b, ealyx, spread to show segments; c and d, braets; e, braetlet (b, c, d, e, twiee natural size). ing the capitula unequal, firm, ciliate, the inner surface ap- pressed-puberulent, the outer surface glabrous or the costa sparingly strigose, the larger bract linear, 13 mm. long, 2.5 1941] WOODSON & SCHERY—FLORA OF PANAMA. V 469 mm. wide near tip and base, 2 mm. wide at middle, obtuse, 3- nerved, the lateral nerves inconspicuous, the margins subhya- line near the base, the smaller bract of the pair lanceolate, 7 mm. long, 2.5 mm. wide at base, rounded at apex, 1-nerved, the margins subhyaline; bractlets lanceolate, 5.5 mm. long, 1.5 mm. wide at middle, acute, 1-nerved, finely pubescent toward the tip, ciliate; calyx 7 mm. long, finely pubescent toward tip, ciliate, the hairs of the inner surface appressed, the segments narrowly triangular, 4 mm. long, 1 mm. wide at base, gradually narrowed to a slender tip, faintly striate-nerved ; corolla 3 em. long, bright pink, pubescent (except the lower portion), 2.5 mm. in diameter at base, narrowed to 1.5 mm. at tip of calyx, thence enlarged to about 7 mm. and narrowed again to 4 mm. at mouth, the posterior lip 8 mm. long, acuminate, the lower lip about 5 mm. long, with 3 rounded lobes; ovary glabrous; capsule not seen.—cHIRIQUÍ: vicinity of Bajo Mona and Que- brada Chiquero, alt. 1500 m., July 18, 1940, R. E. Woodson, Jr. € R. W. Schery 569 (Herb. Missouri Bot. Gard., TYPE). Related to D. trifurca Oerst., but easily distinguished from that by its small thin leaf blades with rounded bases and by the much smaller braets. CUCURBITACEAE FRANTZIA PITTIERI (Cogn.) Pittier—socas DEL Toro: Nieve- cita, alt. 100 m., Aug. 3-20, 1940, Woodson & Schery 1062. Pre- viously known from Costa Rica. FEVILLEA CORDIFOLIA 1. —СнІВІ901: vicinity of San Barto- lomé, Peninsula de Burica, alt. 0-50 m., July 28-Aug. 1, 1940, Woodson & Schery 872. Previously reported from Costa Rica and evidently widely distributed in tropical America. Also known from the West Indies and South America. RUBIACEAE (P. C. Standley, Chicago) CEPHAELIS chiriquensis Standl., sp. nov. Frutex metralis omnino glaber, ramis crassiusculis teretibus, internodiis su- perioribus 3-7.5 cm. longis; stipulae ca. 1 cm. longae vel paullo breviores crassae fere ad basin bilobae, lobis approximatis late [Vor. 28 470 ANNALS OF THE MISSOURI BOTANICAL GARDEN oblongis apice anguste rotundatis erectis; folia longiuscule petiolata chartacea, petiolo gracili 2.5-3 em. longo; lamina el- liptico-ovalis 9-14.5 em. longa 4.5-7 em. lata, apice abrupte breviter acuminata, acumine anguste triangulari, basi acuta, concolor, costa gracili subtus bene elevata, nervis lateralibus utroque latere са. 20 teneris prominentibus arcuatis regulari- bus et parallelis, venulis fere obsoletis; inflorescentia ter- minalis sessilis e capitulis 3 breviter pedunculatis composita, peduneulis erassissimis 7-10 mm. longis; capitula inaequalia, lateralibus paullo minoribus, centrali fere 2 ст. alta et aequi- lata; bracteae exteriores coriaceae fere 2 em. longae late ovatae acutae vel acuminatae, interioribus paullo brevioribus ovatis vel oblongo-ovatis, intimis lanceolatis vel lineari-lanceolatis ; flores bene evoluti non visi.—cnriniQví: vicinity of Bajo Chorro, alt. 1900 m., July 20-22, 1940, Woodson & Schery 610 (Herb. Field Mus., түрк). The bracts are said to be purple. The species is not espe- cially close to any other known from Panama or elsewhere in Central America. The inflorescence is distinctive, consisting of three large, terminal, short-pedunculate heads. Faramea Woodsonii Standl., sp. nov. Frutex 1-2-metralis omnino glaber, ramis gracillimis teretibus viridibus, inter- nodiis plerumque brevibus; stipulae 6-7 mm. longae virides breviter connatae sublaxae et non adpressae, basi late ovatae, sensim in setam rigidam attenuatae, persistentes; folia inter minora breviter petiolata laete viridia firme erasseque mem- branacea, petiolo erassiuseulo 3-4 mm. tantum longo; lamina ovali-elliptica vel elliptico-oblonga 5-7.5 em. longa 1.5-3 em. lata subito breviter acuminata, basi acuta vel subacuta, supra sublucida, costa tenera prominente, subtus paullo pallidior, costa prominente, nervis lateralibus utroque latere ca. 8 tener- rimis vix elevatis angulo lato fere recto abeuntibus remote a margine laxe junctis, venulis fere obsoletis; inflorescentia terminalis umbelliformis pauciflora erecta 1.5-2 ст. longe pedunculata, basi foliis 2 lanceolato-oblongis viridibus acumi- natis 1.5-2.5 em. longis bracteiformibus fulcrata; pedicelli graciles sed rigidi usque 8 mm. longi inaequales; hypanthium 1941] WOODSON & SCHERY—FLORA OF PANAMA. V 471 1.2 mm. altum late obovoideum in sicco cyaneum, calyce vix 0.6 mm. longo remote denticulato, disco crasso calyce fere duplo longiore; corolla caerulea, tubo ca. 8 mm. longo sursum paullo dilatato et vix ultra 2 mm. lato, lobis 4 ut videtur patentibus lineari-lanceolatis longe sensim attenuatis ca. 14 mm. longis; antherae inclusae; stylus gracilis З mm. longus.—cocrÉ: north rim of El Valle de Antón, near Cerro Turega, alt. 650-700 m., June 30, 1940, Woodson & Schery 164 (Herb. Field Mus., TYPE). Not easily confused with any other Central American species, being well marked by the small umbels of blue flowers, each umbel subtended by two green, bract-like leaves. The gen- eral appearance of the plant somewhat suggests that of some species of Cestrum, a resemblance observable also in other species of Faramea, particularly certain South American ones. HorrMANIA AREOLATA Standl.—cuiriqui: vicinity of Bajo Chorro, alt. 1900 m., July, 1940, Woodson & Schery 661. A small shrub with yellow flowers. Known previously only from the type, which was obtained in the same region. Horrmania DavipsoNiaE Standl.—Known previously only from the type, Davidson 216, from Bajo Chorro, Boquete Dis- trict, Provincia de Chiriquí, at 1800 meters. An additional col- lection has been made recently: cmimiquí: vicinity of Bajo Mona and Quebrada Chiquero, alt. 1500 m., July, 1940, Wood- son Ё Schery 595. The flowers are described as brownish-yel- low; Mrs. Davidson described them as pale pink. Well- developed eorollas, not described when the species was pub- lished, are present with the later collection. The corolla is glabrous outside, 5 to 6 mm. long, with a very short tube, scarcely one-third as long as the lobes. Horrmannia Woodsonii Standl., sp. nov. Frutex vel arbus- eula 2-4-metralis omnino glabra ramosa, ramis subteretibus sordide in sieco brunneis gracilibus, internodiis brevibus; stipulae parvae latissime triangulares acutiusculae ; folia inter minora opposita graciliter petiolata membranacea, petiolo 1- 2.2 em. longo; lamina lanceolata usque oblongo-lanceolata, anguste elliptico-lanceolata, vel interdum lanceolato-ovata, (Vor. 28 412 ANNALS OF THE MISSOURI BOTANICAL GARDEN vulgo prope medium latissima, concolor, longe angusteque acuminata, basi cuneato-attenuata et decurrens, costa tenui vix subtus elevata, nervis lateralibus utroque latere 6-7 tener- rimis leviter areuatis angulo semirecto adscendentibus ; flores axillares brevissime racemosi, racemis paucifloris petiolo vulgo paullo brevioribus brevissime pedunculatis, pedicellis gra- cilibus vix ad 3 mm. longis; hypanthium turbinatum in sicco angulatum 2 mm. longum basi angustatum, lobis 4, ea. 0.6 mm. longis latissime ovato-triangularibus acutiusculis; corolla lutea in alabastro acutiuscula 4-5 mm. longa, tubo erasso, lobis subpatentibus oblongis acutiusculis tubo bene longioribus; antherae lobis corollae paullo breviores; bacca ut videtur fere matura late oblongo-obovoidea vix ultra 2.5 mm. longa.— CHIRIQUÍ: vicinity of Callejón Seco, Volcán de Chiriquí, alt. 1700 m., July 17, 1940, Woodson & Schery 503 (Herb. Field Mus., TYPE); Quebrada Velo, alt. 1800 m., July 8, 1940, Wood- son Ё Schery 260. In the key to species of Hoffmannia in ‘North American Flora,’ the present plant would run either to H. chiapensis Standl., a species of Guatemala and southern Mexico, with which it has no close relationship, or to H. Pittieri Standl., of the Voleán de Chiriquí, Panama. It is related to the latter, which, however, is larger in all parts and has a much more ample inflorescence, often copiously branched and always with more numerous flowers. COMPOSITAE (S. F. Blake, Washington, D. C.) ARCHIBACCHARIS panamensis Blake, sp. nov. Herba prope apicem ramosa 1.5 m. alta; caulis dense et sordide pilosius- eulus; folia oblongo-elliptiea v. elliptico-obovata brevissime petiolata acuta basi euneata tenuia remote calloso-denticulata penninervia supra puberula subtus densius pilosiuscula; capi- tula feminea numerosa laxe cymoso-paniculata 5 mm. alta 20- 21-flora, floribus hermaph. 1-2; phyllaria anguste linearia acuminata dorso puberula; corollae fem. ligulatae, ligula erecta 1 mm. alta ramis styli breviore; corollae hermaph. usque ad apicem tubae 5-partitae. 1941] WOODSON & SCHERY—FLORA OF PANAMA. V 413 ‘‘Tall herb, 1.5 m. high;" stem simple below the inflores- cence, bearing branches in the upper axils, subterete, somewhat striate, pithy, З mm. thick, straight, densely pilosiusculous with more or less crisped, many-celled, brownish hairs about 0.2 mm. long, not glandular; petioles broad, naked, sordid-puberulous, 1-1.5 mm. long; blades 7-9 ст. long, 2.3-3 em. wide, callous- apiculate (apiculus 1 mm. long), remotely denticulate above the entire lower 25 of their length (teeth 5—8 pairs, slender, cal- lous-tipped, about 0.5 mm. high, 4-11 mm. apart), above dark dull green, evenly puberulous on surface and veins with sev- eral-celled spreading hairs, more densely so on costa, beneath somewhat grayish-green, rather densely spreading-short- pilose with lax several-celled hairs (0.2-0.4 mm. long), more densely so along costa and veins, feather-veined (the principal veins about 6 pairs, prominulous beneath, the veinlets translu- cent, not prominulous, the costa prominent and whitish at base); leaves of the branches smaller; individual panicles terminal and in the upper axils, loose, many-headed, pubescent like the stem, on peduncles 6 em. long or less, 1.5-6 em. wide, together forming a large leafy panicle, the pedicels mostly 6-10 mm. long, filiform, naked or with a filiform-subulate braet ; involucre about 4-seriate, graduated, 4-5 mm. high, the phyl- laries narrowly linear or the outer linear-subulate, 0.2-0.5 mm. wide, acuminate, with greenish midline and whitish subscari- ous margin, the outer puberulous throughout, the middle puberulous on back above, the inmost glabrous except for the ciliolate tip; pistillate heads 5 mm. high, thick-cylindric (when moistened), 20-21-flowered, the pistillate flowers 18- 20, the hermaphrodite 1-2; pistillate corollas about 2.7 mm. long, the tube slender, densely puberulous at apex, other- wise essentially glabrous, about 1.7 mm. long, the ligule erect, elliptic or oblong, concave, emarginate or 3-denticu- late, 0.8-1 mm. long, white; hermaphrodite (staminate) co- rollas 3.2 mm. long, whitish, the tube 1.5 mm. long, sparsely puberulous, the limb 1.7 mm. long, campanulate, divided to the very base into 5 teeth, these oblong, acute, obscurely puberu- lous dorsally, about 0.4 mm. wide; achenes (scarcely mature) [Vor. 28 414 ANNALS OF THE MISSOURI BOTANICAL GARDEN of pistillate flowers oblong, strongly compressed, 4-nerved, rather densely hirsutulous, 1.4 mm. long, those of the hermaph- rodite flowers inane, linear-oblong, densely hirsutulous, 1.2 mm. long; pappus of slender whitish bristles, about 3 mm. long; style branches of hermaphrodite flowers lance-oblong, very acute, hispidulous, about 0.7 mm. long; staminate plant unknown.—cocrÉ: vicinity of El Valle, alt. 100-800 m., Sept. 5, 1938, P. H. Allen 751 (U.S. Nat. Herb., No. 1,793,562, TYPE). Nearest Archibaccharis irazuensis Blake, in which the leaves are lanceolate and acuminate, the phyllaries broader, firmer, and less pubescent, the stem-pubescence longer, and the floral details different. In A. irazuensis the heads are 31-39-flowered, with 3-6 hermaphrodite flowers; the pistillate corollas are tubular-filiform, without ligule; and the hermaphrodite co- rollas have a definite throat (about 0.7 mm.), half as long as the teeth or more. A. panamensis appears to inhabit a very different life zone, since the recorded altitudes for specimens of А. wrazuensis examined run from 1500-1900 to 2700-3000 meters in Costa Rica, and 3500-4000 meters in Panama (see below). The genus Archibaccharis has not hitherto been known south of Costa Rica. Two additional species can also be recorded from Panama: ARCHIBACCHARIS IRAZUENSIS Blake.—cuiriqui: Volcán de Chiriqui, between Potrero Muleto and summit, alt. 3500-4000 m., July 13-15, 1940, Woodson € Schery 463. Previously known from Volcan Irazü and its vicinity, and from the Cerro de las Vueltas in the Province of San José, Costa Rica. ARCHIBACCHARIS TORQUIS Blake.—cutrigui: vicinity of Cal- lejón Seco, Voleán de Chiriqui, alt. 1700 m., July 17, 1940, Woodson & Schery 481. Previously known from Vera Cruz to Costa Rica. PTEROCAULON ALOPECUROIDES (Lam.) DC.—panami: hills above Campana, alt. 600-800 m., Dec. 23, 1938, Allen 1325 (U.S. Nat. Herb.). Previously known from the West Indies and northern South America. The genus, apparently, has not been 1941] WOODSON & SCHERY— FLORA OF PANAMA. V 479 recorded heretofore from the Mexican-Central American region. Curpapium sessile Blake, sp. nov. Frutex; caulis (supra) 6-angulatus striatus parce strigillosus; folia elliptico-lance- olata acuminata basi longe cuneata sessilia crenato-serrata pergamentacea longe supra basin triplinervia supra sparse tuberculato-strigillosa non aspera aetate bullata subtus aequa- biliter strigillosa ca. 14 em. longa 4 em. lata; panieula parva 4 ст. lata terminalis adscendenti-pilosa ; capitula parviuscula 4-5 mm. alta sessilia per 3-6 glomerata, glomerulis 5-7 mm. diam. ca. 12 in pedunculis partialibus 4-6 mm. longis ; phyllaria 2 ovata obtusa appressa 5-9-nervia breviter ciliata et prope apicem strigillosa ; flor. fem. 3, hermaph. 6; receptaculum intra flor. fem. solum 1-paleatum ; ovarium fem. glaberrimum. Shrub 2 m. high; stem sparsely strigillose with subtubercu- late-based hairs, whitish-pithy, 3 mm. thick above; upper inter- nodes 5-8 ст. long; upper leaves 10-14.5 cm. long, 2.5-4 em. wide, broadest near or slightly above the middle, acuminate to an obtusish callous tip and somewhat falcate, cuneate at base (the cuneate base entire, 3-5 em. long), crenate-serrate with about 9-24 pairs of depressed callous-tipped teeth (these 0.5-1 mm. high, 3-9 mm. apart), above brownish green, evenly but not densely tuberculate-strigillose with mostly deciduous hairs with small persistent lepidote bases, more densely antrorse-strigose along the chief nerves, essentially smooth to the touch, in age slightly shining and somewhat bullate with impressed veins and veinlets, beneath lighter brownish green, uniformly but not densely strigillose on surface and veins, triplinerved 3-4 em. above the base, 4—5 mm. wide at base, the bases of each pair not connate; panicle small, much surpassed by the upper leaves, more or less ternately divided, the peduncle 2 em. long, the panicle as long, the bracts small, linear, 7 mm. long or less, subsearious ; phyllaries (in flower) brown- ish, 3.2-3.7 mm. long, 2-2.3 mm. wide; pistillate flowers all paleate, their pales similar to the phyllaries, their ovaries (im- mature) obovoid, obeompressed, perfectly glabrous, 1.2 mm. long, with a distinct inflexed beak, their corollas cylindrie, 2- [Vor. 28, 1941] 416 ANNALS OF THE MISSOURI BOTANICAL GARDEN toothed, white, glabrous, 1.6 mm. long; single pale of the disk oblong, obtuse, ciliate, 2.5 mm. long; ovaries of the hermaph- rodite flowers villous at apex, their corollas (immature) white, 2.5 mm. long, hispidulous on the teeth.—cutrigvi: vicin- ity of Bajo Chorro, alt. 1900 m., July 20-22, 1940, Woodson & Schery 658 (U. S. Nat. Arb. No. 124226, түрк). The only close ally of this species is C. subsessilifolium Hieron., of Ecuador, fragments from the type of which are in the U. S. National Herbarium. In that species the branches are described as ‘‘setose-hirtous with erect hairs’’ [Hierony- mus] or ‘‘villous with erect hairs’’ [O. E. Schulz], the leaves are very rough above, antrorse-hirsute beneath, the larger sub- cordate at base, and the phyllaries are narrowly ovate, acumi- nate, 5 mm. long, with recurved reddish apex. RuMFORDIA POLYMNIOIDES Greenm.—ocHiRIQUÍ: vicinity of “Кем Switzerland,’’ central valley of Rio Chiriquí Viejo, alt. 1800-2000 m., Jan. 6-14, 1939, Allen 1413 (U. S. Nat. Herb.). Previously known only from Costa Riea. LIABUM РОГХУАМТНОМ Klatt.—cuirigui: Chiriquí Viejo Val- ley, April 26, 1938, G. White 93 (U. S. Nat. Herb.) ; vine climb- ing to 8 m. on stumps, vicinity of “Хеу Switzerland," central valley of Río Chiriqui Viejo, alt. 1800-2000 m., Jan. 6-14, 1939, Allen 1415 (U. S. Nat. Herb.). Previously known only from Costa Rica. Liasum sacirratum Sch. Bip.—curirigut: trail from Cerro Punta to head-waters of Rio Caldera, alt. 2250-2500 m., Jan. 14, 1939, Allen 1458 (U. S. Nat. Herb.). Not previously known north of Colombia. JUNGIA FERRUGINEA L.f.—cnumiqví: vicinity of “Мех Switzerland,’’ central valley of Rio Chiriquí Viejo, alt. 1800— 2000 m., Jan. 6-14, 1939, Allen 1367 (U. S. Nat. Herb.) ; liana climbing 30 m. into tops of largest trees, vicinity of Cerro Punta, alt. 2000 m., Jan. 21-24, 1939, Allen 1520 (U. S. Nat. Herb.). Previously known from Costa Rica and northern South America. EXPLANATION OF PLATE PLATE 19 Li Aechmea Allenii L, В. Smith. From type specimen, P. H. Allem 2378, in the United States National Herbarium. ANN. Mo. Bor. GARD., Vor. 28, 1941 PLATE 19 WOODSON & SCHERY—FLORA OF PANAMA [Vor. 28, 1941] 480 ANNALS OF THE MISSOURI BOTANICAL GARDEN EXPLANATION OF PLATE PLATE 20 Palmorchis trilobulata L. О. Williams: fig. 1, plant, x 15; fig. 2, flower opened out, x 2; fig. 3, lip and eolumn from the side, natural position, x 3; fig. 4, column id lip with the lip expanded, x 4 ANN. Mo. Bor. GARD., Vor. 28, 1941 PLATE 20 PALMORCHIS | Да А душа ui 7 Le Я ЙН 77 ДЕ J 77777 РАЯ ed x MEOS axe HP n nae “н [2 7774 j J (ИЙИ / JL 7. D WOODSON & SCHERY—FLORA OF PANAMA [Vor. 28, 1941] 482 ANNALS OF THE MISSOURI BOTANICAL GARDEN EXPLANATION OF PLATE PLATE 21 Figs. 1-6. Pleurothallis macrantha L. О. Williams: fig. 1, plant, x%; fig. 2, flower opened out, x ip m 3, cei from the side in natural position, x 1%; fig. 4, lip from the side, x 3; fig. 5, lip from above «s one basal lobe spread out, x 3; fig. 6, cross-section бе « тр of the lip, х Figs. 7-10. Epidendrum gibbosw . O. Williams: pt 13 p X 0; fig. 8, flower expanded, x 2; fig. 9, lip from ps side i in natural position, x 2; fig. 10, floral braet, x 1. , ANN. Mo. Bor. Garp., Vor. 28, 1941 PLATE 21 PLEUROTHALLIS macrantha <£. yns /4 EPIDENDRUM _9266озит PP pf s. spi = <-#:; A GORDA W^ DIL ON WOODSON & SCHERY—FLORA OF PANAMA [VoL. 28, 1941] 484 ANNALS OF THE MISSOURI BOTANICAL GARDEN EXPLANATION OF PLATE PLATE 22 Figs. 1-2. Epidendrum ellipsophyllum L. О. Williams: fig. 1, plant, x №; fig. 2, flower opened out, x 115. Figs. 3-5. Epidendrum Allenii L. O. Williams: fig. 3, plant, x №; fig. 4, flower opened out, x 1%; fig. 5, column (and base of the lip) from the side, x 2%. ANN. Mo. Bor. GARD., Vor. 28, 1941 PLATE 22 w E „Анн L W2 WOODSON & SCHERY—FLORA OF PANAMA [Vor. 28, 1941] 486 ANNALS OF THE MISSOURI BOTANICAL GARDEN EXPLANATION OF PLATE PLATE 23 Epidendrum pendens L. О. Williams: fig. 1, plant x №; fig. 2, flower opened out, х1. PLATE 23 ANN. Mo. Bor. GARD., Vor. 28, 1941 WOODSON & SCHERY—FLORA OF PANAMA [Vor. 28, 1941] 488 ANNALS OF THE MISSOURI BOTANICAL GARDEN EXPLANATION OF PLATE PLATE 24 Scaphyglottis tenella L. O. Williams: fig. 1, plant, x %4; fig. 2, flower opened out, x 244; fig. 3, column and lip from the side, x 5. ANN. Mo. Вот. GARD., Vor. 28, 1941 PLATE 24 „ SCAPHYGLOT T 15 Хара у tenella L Наз BERT adf 4 WOODSON & SCHERY—FLORA OF PANAMA EXPLANATION OF PLATE Zygopetalum parviflorum L. О, Williams: fig. 1, plant, x 14; fig. 2, flower opened | + еі | я E : Ё ANN. Mo. Вот. GARD., Vor. 28, 1941 PLATE 25 ZYGOPETALUM роли L ylas AN А 297 NW ИА : NN / MAN | ) OWN d N р / N / T UN i АУ 2, WN f \ 4 \ Y Ха) / D < 1 WOODSON & SCHERY—FLORA OF PANAMA THE CYTOGENETICS OF POA PRATENSIS! WILLIAM L. BROWN Cytogeneticist, United States Golf Association, Green Section, Washington, D. C. ormerly Research Fellow, Henry Shaw School of Botany of Washington University INTRODUCTION Poa pratensis L., the common Kentucky Blue grass, is uni- versally recognized by taxonomists as an extremely variable species. Many different strains are included within the species, and although American botanists do not generally recognise them as being sufficiently distinct to warrant their segregation from pratensis they are conspicuously present. Basic informa- tion on their variability, inheritance, and methods of repro- duetion would not only make it possible to classify the various strains more efficiently but it would allow us to use the species to best advantage for pasturage, turf, ete. It would also indi- cate what improvements might be expected by breeding for su- perior strains and the methods to be followed in making such improvements. In this investigation an attempt has been made to correlate data from morphological, cytological, and genetical studies in an effort to arrive at a clearer understanding of the behavior of the species. MATERIALS AND METHODS Materials for this study have, for the most part, come from the Missouri Botanieal Garden Arboretum, Gray Summit, Missouri. In the spring and summer of 1941 additional data have been obtained from plants growing in the nurseries of the United States Golf Association, Green Section, at Arling- ton, Virginia. Most of these plants came originally from estab- 1 An investigation carried out at the Missouri Botanical Garden in the graduate laboratory of the Henry Shaw School of Botany of Washington University and submitted as a thesis in partial fulfillment of the requirements for the degree of Doctor of Philosophy. (493) [Vor. 28 494 ANNALS OF THE MISSOURI BOTANICAL GARDEN lished turf selected at various places throughout the United States. At the Missouri Botanical Garden Arboretum eleven lots of seed from well-established clones were harvested in 1939, all being taken from open-pollinated panicles. The parent clones varied in width from 1 to 3 feet and were growing in an area approximately 50 feet square. There was therefore consider- able opportunity for cross-pollination between clones. Seed was sown in the greenhouse the first week in January 1940. In February individual plants were transferred to 2-inch pots and removed to a cool greenhouse. They remained here until April, at which time they were moved to the nursery. Both permanent and smear methods have been employed in cytological studies. La Cour’s 2 BD fixative was used in all cases where permanent preparations were desired. The Feul- gen staining technic, supplemented by crystal violet and aceto- carmine, was used to advantage. PROGENY ANALYSES G.S. Lot 1.—Seeds from open-flowering panicles were col- lected from a single clone in 1939. The resulting progeny was analyzed after two year’s growth, i. e., after the plants had reached full maturity. Of 133 plants in the family, 131 were uniform and morphologically closely resembled the mother parent. These were of tall upright habit, possessing relatively narrow leaves and average-size panicles. The two aberrant plants were easily distinguished by their low habit of growth, darker green color, and wider leaves. G.S. Lot 2.—In this lot 103 plants were grown to maturity, 102 of which were morphologically constant and indistinguish- able from the female parent. They resembled those of Lot 1 very closely and were definitely of the **hay"' type. They were tall, with narrow leaves and panieles of medium size. The one aberrant plant was a low, spreading type with wide leaves and a heavy inflorescence. It resembled somewhat the ‘‘off types" in Lot 1, but deviated from the maternal parent even more than the aberrants of the first lot. 1941] BROWN—CYTOGENETICS OF POA PRATENSIS 495 G.S. Lot 3—Of the 100 plants composing this progeny all were absolutely uniform, with growth habit and morphology resembling closely those of the first two lots. G.S. Lot 4—These plants represent a type very different from those of the previous lots. They are characterized by low growth, light green leaves, and small purple inflorescences. Fifty-five seedlings were raised to maturity, all of which were absolutely uniform and indistinguishable from the maternal parent. G.S. Lot 5.—The parent of this lot of plants was of tall up- right growth habit. The leaves were slightly more than me- dium width and dark green in color. Panicles were of average size. Progeny of this parent, consisting of 55 plants, may be segregated into two distinct classes: (1) those identical with the parent, to which class belong 45 of the plants; (2) 10 plants with lax, glossy, light green leaves almost twice as wide as those of the parent, and inflorescenses extremely long and heavy. By observation alone one could easily place each of the plants into one of these two classes. No intermediates oe- curred. G.S. Lot 6.—Of the 60 plants comprising this selection, all were uniform with the exception of some slight differences in time of flowering. However, this variation can hardly be re- ferred to as of a genetie nature since it might easily be the re- sult of soil heterogeneity, differences in soil moisture or dis- turbances from cultivation. G.S. Lot 7.—This progeny, consisting of 60 plants, was of a type almost identical with Lot 1. They were strikingly uniform throughout and showed no morphological deviation from the parents. G.S. Lot 8.—As a whole, these plants resembled the aber- rants of Lot 3. They are characterized by tall culms, wide, lax, light green leaves, and very long panicles. The progenies were surprisingly uniform and, with one exception, quite similar morphologically to the maternal parents. The one aberrant in a progeny of 60 plants possessed stiff, narrow, dark green leaves and a small, almost dwarfed inflorescence. [Vor. 28 496 ANNALS OF THE MISSOURI BOTANICAL GARDEN G.S. Lot 9.—The 60 plants grown to maturity were uniform throughout and indistinguishable from the maternal parent. G.S. Lot 10.—'The plants from which seed for this progeny was eolleeted were characterized by low, spreading growth and very vigorous habit. The leaves were wide (5-6 mm.) and dark green in color. Panicles were large, heavy, and late- flowering. Fifty of the 60 seedlings closely resembled the fe- male parents morphologically, while the other 10 were vari- ants. However, as has been shown for Lot 5, the variants were quite uniform among themselves. All 10 plants fell into a class characterized by narrow leaves and small panicles. G.S. Lot 11.—These plants were typical ‘‘hay’’ type with tall upright growth, moderately narrow leaves and average- size panicles. The parent clone was of this general type and 53 of the 55 plants grown from seed were almost identical with the parent. The two aberrants differed markedly from the typical plants. They were a low, spreading, wide-leaved type with very long heavy panicles. TABLE I SUMMARY OF MORPHOLOGICAL VARIATION IN GRAY SUMMIT PROGENIES Number | Number | Per cent plants | aberrants aberrants | G.S.- 1 133 | 2 1.50 G.S.- 2 103 | 1 0.97 G.S.- 3 100 | - - G.S.- 4 55 | - - G.S.- 5 55 10 18.18 G.S.— 6 60 - - G.S.- 7 60 | - - G.S.- 8 60 | 1 1.66 G.S.- 9 60 - - G.S.-10 60 | 10 16.66 G.S.-11 55 | 2 3.63 All the plants described in the foregoing paragraphs were obtained from seed collected from open-pollinated panicles. With the exception of Lots 5 and 10, these progenies have been quite uniform and very similar to the maternal parents. The Arlington plants were subject to a somewhat different 1941] BROWN—CYTOGENETICS OF POA PRATENSIS 497 method of study. The seed had been harvested from single isolated plants in nursery rows. Collection was made not only from open-pollinated panicles but from inflorescences that had been isolated from foreign pollen by bagging prior to anthesis. In some instances seed-setting in parchment isola- tion bags was very poor, but enough seed was always produced for progeny analyses. Seedlings were started in the green- house and then transferred to nursery rows. Those from free- flowering and isolated panicles of the same plant were placed in adjacent rows for comparison. At the height of their flower- ing season they were analyzed morphologically in the same manner as were the G.S. progenies. The contrast between progenies from isolation and open- pollination was sometimes very marked (pls. 28-29). For ex- ample, B-2 after open-pollination produced a uniform and ma- ternal-like progeny typical of apomictic strains. The selfed progeny of this line, however, showed a degree of variation that might be expected only from sexually reproducing plants. Even before flowering, distinct differences between plants could be easily ascertained from vegetative characters, the variability in leaf width being especially noticeable (figs. 1 and 2). At the time of flowering distinct differences in the inflores- cences, including size of panicles, number of florets per spike- let, etc., were evident. There also occurred in Ше selfed line of this progeny a number of weak plants, many of which did not survive beyond the seedling stage. Other weak plants which were grown to maturity exhibited characters entirely foreign to P. pratensis. Another example of very distinct differences in open and selfed lines of the same strain is provided by selection B-6. Here also the progenies grown from open-pollinated panicles were uniform and exhibited a tall upright habit of growth (pl. 28). The progeny produced after isolation, although not quite so variable as the selfs in B-2, was of an entirely differ- ent growth habit from that produced after open-pollination, almost without exception being low-growing, spreading, semi- decumbent types (pl. 29). [Vor. 28 498 ANNALS OF THE MISSOURI BOTANICAL GARDEN In eontrast to the B-2 and B-6 seedlings, those of B-17 and B-1 showed absolutely no differences between open and selfed lines, and all were essentially maternal-like in appear- ance (pls. 26 and 27). In addition to the three strains mentioned above, progenies of open and selfed lines of eleven other selections were grown LEAF WIDTH (mm) B-2 OPEN Fig. 1. Variation in leaf width in B-2 progeny after open-pollination, to maturity and analyzed. Six of these progenies showed very marked differences in the number of aberrants occurring after selfing and after open-pollination, the selfed line always pro- ducing more variable types than the open-pollinated one. Strains which had previously been concluded to be apomictie behaved as sexually reproducing plants when they were iso- lated from foreign pollen. When segregation of a degree ex- 1941] BROWN—CYTOGENETICS OF POA PRATENSIS 499 hibited by some of these selfed lines does occur it must be as- sumed that reproduction has taken place by gametie union. If this is true, what is there to prevent the open-pollinated lines of the same plant from functioning in a like manner? Assum- ing that the plant is potentially sexual, one might logieally have expected the open-pollinated lines whose parents are sub- LEAF WIDTH (mm) әл > 3 B-2 SELFED Fig. 2. Variation in leaf width in B-2 progeny after selfing. ject to eross-pollination with other plants to exhibit more vari- ation than the selfs. This, however, did not occur. As a result of these analyses it is concluded that there are within Poa pratensis both sexual and asexual strains. А]- though the asexual strains are apparently predominant, the sexual ones occur frequently in most populations. Aside from purely sexual and purely asexual strains, there are those [Vor. 28 500 ANNALS OF THE MISSOURI BOTANICAL GARDEN plants which apparently reproduce asexually when open-polli- nated but which reproduce sexually when selfed. CHROMOSOME STUDIES The polyploid nature of the genus Poa has long been estab- lished with 7 as the accepted base number. Within Poa pra- tensis somatic chromosome numbers range from 28, recorded by Avdulow (231), to 110. This extensive polyploid series is not eonfined to euploidy, but a great many aneuploid numbers have been reported. The modal chromosome number for the species has been established as 56, Brown (239). These 56- chromosome types are, for the most part, repr of average Blue grass, that is, they are of tall, upright habit, pos- sess moderately narrow leaves, and average approximately three florets per spikelet. Unfortunately, few correlated cytological and morphologi- cal data are available on P. pratensis. Müntzing ('40) makes some mention both of the cytology and morphology of some of the biotypes with which he has worked. Over a period of years Ihave observed that those low-growing plants with wide leaves, large spikelets, and heavy inflorescences have, with few excep- tions, lower chromosome numbers than do those with narrow leaves, small spikelets, and average-size inflorescences. In faet, there is considerable evidence that approximately 6n is the optimum for chromosome inerease beyond which there is a marked decline in vigor. In the present investigation no effort has been made to ob- tain even random chromosome numbers from a great many lots of plants. Instead, one lot was selected from which it was felt that chromosome data would be of particular significance. The great majority of progenies, after open-pollination, exhibited little, if any, variation and were eliminated as possible mate- rial for chromosome studies. After studying the morphologi- cal variability in G.S. Lot 5 it was immediately apparent that chromosome data from this progeny would be of special interest. It will be remembered that the progeny of Lot 5, after open- dad 1941] BROWN—CYTOGENETICS OF POA PRATENSIS 501 pollination, segregated into two distinct classes. Forty-five of the 55 plants were typical of the maternal parent while the re- maining 10 were of a very different morphological type. Cyto- logical examination of 35 of these plants has yielded some very interesting facts on the morphological effects of chromosome elimination from the original complement. The typical plants of the progeny, for the most part, possessed 49 somatic chro- mosomes (table п). The ‘‘off types," ог aberrants, all of which belonged to a single morphological type, were, with one exception, shown to have 42 chromosomes, or one less genom than the typical plants. Thus the change from a moderately narrow-leaved, upright ‘‘hay’’ plant to a wide-leaved, low- growing one with increased vigor is the result of the elimina- tion of one genom from the germ-plasm. The fact that all but one of the aberrant types contained one less genom instead of from 1 to 7 fewer chromosomes is of considerable interest. A number of 2n plus 1-, 2-, or 3- or 2n minus 1-, 2-, or 3-chromosome types have been investigated in other plants (Zea, Datura, ete.), and in most such cases the elimination or addition of even one chromosome was usually morphologically apparent. This, however, is not true in those plants of P. pratensis that have been studied. One of the typical plants (G.S. 5-5) possessed 47 chromosomes, yet it could not be morphologically distin- guished from those with 49 chromosomes. A few scattered chromesome counts have been made on plants selected from lots other than G.S. 5. Although these are so few that no conclusions can be drawn concerning the cytology of the lots as a whole, the following bit of information is of some interest. Lot 1, it will be recalled, produced a very low percentage of aberrants and is apparently an apomictic strain. Chromosome counts were made from six plants of this lot, and in all the 2n number was found to be 56. Four out of five plants from Lot 2, another apomictic strain, were found also to possess 56 chromosomes. The number for the fifth plant could not be exactly determined but it was between 55 and 58. Lot 10, which reproduces at least partially by gametic union, yielded plants with 42, 41, and 42--1 chromosomes. [Vor. 28 502 ANNALS OF THE MISSOURI BOTANICAL GARDEN These data suggest that apomixis may be more prevalent in the higher chromosome types and that polyploids of approxi- mately 6n reproduce sexually. TABLE II SUMMARY OF CHROMOSOME NUMBERS IN G.S. LOT 5 Plant 2n Type of plant 5- 1 49 Typieal 5- 3 42 Aberrant 5- 3 49 Турка 5-4 49 Typiea 5- 5 47 Тура 5-6 Туріса 5-7 48-49 Typica 5-8 48+1 Туріса 5- 9 42 Aberrant 5-10 49-50 Typical 5-11 49 Typical 5-12 49 Typieal 5-13 49 Typieal 5-14 49 T 5-15 49 Typical 5-16 42 Aberrant 5-17 49 Typi 5-18 49 Typical 5-19 49 Typical 5-20 48—49 Typical 5-21 43+1 Aberrant 5-22 42 Aberrant 5-23 49+1 Typical 5-24 49 Typical 5-25 49-50 Typical 5-26 49*1 Typical 5-27 49 Typical 5-28 41-42 Aberrant 5-29 49 Typical 5-30 49 Typical 5-31 49 Typical 5-32 50 Typie 5-33 42 Aberrant 5-34 49-50 T 5-35 49 Typieal Meiosis: Osbervations on meiosis were made on a number of plants whose chromosome numbers had been previously de- termined somatieally. Although meiosis in P. pratensis is not regular, the irregularity is not of a degree that might be ex- pected in a highly polyploid species (figs. 3-13). Pairing is, for the most part, by bivalents, although univalents, trivalents, and quadrivalents have been observed in all cells examined. 1941] BROWN—-CYTOGENETICS OF POA PRATENSIS 503 Figs. 3-8. Camera-lucida юра of meiotie chromosomes Poa pratensis, x 2000: fig. 3. ри, s from an apomietie plant отша 4 poi hic oig 3 tri- valents, and approx. 23 biv dieit univalents in outline; 3: Anaphase I ie apo илиги plant чал өш univalents di Pug all polar chromoso mes 8 fig. 5. Gene a: II EU lagging; fig. 6. Un Dons dividing at pod odd T sexual plant; taphase I showing non-conjugation in ria ai pw: fig. 8. Anaphase т dui unequal distribution with a dieentrie ehromoso [Vor. 28 504 ANNALS OF THE MISSOURI BOTANICAL GARDEN The majority of the bivalents are ring-shaped, averaging two chiasmata. Figure 11 represents a meiotic configuration which is rather typical of 42-chromosome plants. At Metaphase I the configuration consists of 17 II’s, 4 I's, and 1 IV. At first ana- phase most of the univalents divide after lagging in the region of the equatorial plate (fig. 12). The 56-chromosome plants show no more meiotic irregularity than do the hexaploids (fig. 3). The configuration at Metaphase I consists usually of 1-2 ТУ в, 1-2 III's, approximately 4 I's, with the remaining chro- mosomes paired as bivalents. Univalents usually lag at Ana- phase I, but they always reach the poles in time to be included in one or the other of the polar groups. Distribution is, with very few exceptions, fairly equal at the anaphases. The tet- rads and microspores appear to be normal, and micronuclei have not been observed in any of the cells examined. Some plants are obviously heterozygous for an inversion or dupli- cation as indicated by the occasional occurrence of dicentric chromatids (fig. 8). А POMIXIS The occurrence of apomixis in Poa pratensis was first shown by Müntzing (33). He studied a number of plants represent- ing eight biotypes. A large percentage of his material was found to have aneuploid chromosome numbers, and for the most part the numbers within biotypes remained constant. Since the individual plants composing any one biotype exhib- ited little morphological variation, Miintzing assumed them to be apomictic. In a study of twin seedlings in Poa pratensis he (737) found a rather high occurrence of polyembryony, which he considered to be in some way correlated with apomic- tie propagation. He states further that . . . “formation of ‘triploid’ twin plants, however, resulting from fertilization of unreduced ovules, demonstrates that apomixis in Poa praten- sis is not absolute.”’ Akerberg (’36) investigated seed production in eighty bio- types of P. pratensis. After employing both emasculation and free-flowering methods he concluded that pollination was nec- 505 1941] BROW N—CYTOGENETICS OF POA PRATENSIS со 9200$ 9900153 0323$ 13 gs. 9-13. Camera-lucida drawings of meiotic chromosomes in Poa pratensis, x 2000: figs. 9 & 10, Diakinesis in SETE | ком drawn in 2 planes—3 univalen 16 een P nd 1 qua adrivalent; fig. phase I from sexual plant showing 4 univalents, 17 bivalents, and 1 quadrialent fig. 12. Anapha plant showing lagging con ера all polar ehromosomes shown; fig. Metaphase chromosomes of fig. 11 drawn pios: ately. [Vor. 28 506 ANNALS OF THE MISSOURI BOTANICAL GARDEN essary for seed production, but that pollen from Poa alpina had about the same effect as that from P. pratensis. For bio- types with aneuploid chromosome numbers and partial pollen sterility, Akerberg postulated apomietie seed production. In a later investigation of progenies of P. pratensis, he (239) attempts to determine the percentage of apomixis on the basis of morphological constancy. Of 703 plants investigated a little more than 90 per cent had the appearance of the maternal parents. These were assumed to be primarily apomietie and the 10 per cent of aberrants to be sexual. When 58 of the 703 plants were examined cytologically, 43 were found to have the same chromosome number as the mother plant; the re- maining 15 were aberrants. Material was also collected from nature, and ‘‘of these 44 families, 37 were practically speak- ing morphologically constant."' Armstrong (’37) interprets the embryology of P. pratensis quite differently from the authors previously mentioned. Although he admits the occurrence of aneuploidy he still be- lieves the species to be sexual and assumes that only gametes with certain chromosome numbers are capable of function- ing. He has observed, in the embryo sae, a variation in the position of the functioning megaspore. This, he says, ‘‘pro- vides a mechanism for the elimination of megaspores with an abnormal chromosome complement and for the choice of the megaspore containing the normal chromosome comple- ment." On the basis of meiotic pairing he has assumed that 6/16 of the pollen would contain the normal chromosome сош- plement, and he further states, ‘‘Assuming the same fre- quency of irregularity in megasporogenesis as in microsporo- genesis, 6/16 of the megaspores would be normal, and with a choice of four megaspores the chance of obtaining one with the normal number is quite good.’’ Tinney (’40) found that the young ovules of P. pratensis always show a single, elongated, very conspicuous macro- spore mother cell. The nucleus of this cell undergoes meiosis, and usually three haploid macrospores are formed, all of which subsequently disintegrate. The embryo sac arises from a cell % 1941] BROWN—CYTOGENETICS OF POA PRATENSIS 507 of the nucellus and when mature consists of three antipodals, a primary endosperm cell containing two nuclei and the egg apparatus. The diploid egg, according to Tinney, develops into a proembryo by parthenogenesis, the development begin- ning frequently before pollination. He did not observe endo- sperm development until after pollination and therefore con- cluded that pollination or the growth of pollen tubes in stylar tissue may be necessary for endosperm development and con- sequently seed development. Tinney and Aamodt (’40) have analyzed 102 progenies of Poa pratensis in an attempt to arrive at the type of seed de- velopment in various biotypes. In 48 of them no variant types occurred, and these were thought to be apomietie. In one progeny, however, 21.9 per cent of the types were variant. These, it was thought, might have arisen through mutation or as a result of sexual reproduction. Since all investigators of the mode of reproduction in Poa pratensis have shown that apomixis is one of the predominant factors involved, many have assumed that this phenomenon is of general occurrence within the species and can be taken ad- vantage of in selection, etc. This assumption is to some extent undoubtedly true, but in view of the results obtained from this study it seems that we must distinguish between apomixis after selfing and apomixis after crossing or open-pollination Apomiais after Open-Pollination: It has been shown earlier that in lots B-2 and B-6 two very different types of progeny result when the same plants are selfed and allowed to produce seed without isolation. In plants reproducing by open-pollina- tion the resulting progeny is essentially maternal-like morpho- logically and exhibit little, if any, variation within lines. Since Poa pratensis is an extremely diverse species it must, it seems, be assumed that constant morphological types of the nature just mentioned are being produced apomictically. The fact that foreign pollen must be present for the production of uni- form progenies, however, is somewhat unusual and especially significant in the behavior of the species. Apomizis after Selfing: If plants of strains B-2 and B-6 and [Vor. 28 508 ANNALS OF THE MISSOURI BOTANICAL GARDEN some other strains are isolated by bagging, progenies in direct morphological contrast to those obtained after open-pollina- tion result. A wide range of variation both in vegetative and inflorescence characters usually results in these lines. Most of these strains segregate in a manner that would be expected only if the plant were reproducing sexually. These prelimi- nary results indieate that in some strains of Kentucky Blue grass apomixis is dependent upon the presence of foreign pollen (pollen from another strain). The most logical hypo- thesis explaining the phenomenon is that the presence of for- eign pollen on a stigma stimulates the development of some somatie cell of the embryo sae which, after the initial stimula- tion, goes ahead and functions as an embryo. Should the inflo- rescence of the same plant be bagged, thereby eliminating all foreign pollen, reproduction then takes place by gametie union (with the egg cell perhaps funetioning normally). Besides ex- plaining the differences in progeny obtained from the same plant after isolation and free-flowering, this hypothesis ex- plains the varying but usually small percentage of aberrant forms present in most progenies of open-pollinated plants; i. е., they are the result of selfing, whereas the majority of the florets have received a stimulation from foreign pollen (as might be expected to occur in natural populations or in nur- sery rows). The recent work of Englebert ('41) is very interesting in light of the results reported in this paper. Miss Englebert reports that in several species of Poa, including Poa pratensis, no endosperm was developed in emasculated, non-pollinated florets although embryos were found. She concluded, there- fore, that the ‘‘germination of pollen on the stigmas stimu- lated the development of the aposporous polar cell to an endo- sperm (pseudogamy).’’ Although no embryological investi- gations have been attempted in the present study, the evidence from progeny analyses strongly indicates that in some strains of P. pratensis pollen plays a much more important role than the mere stimulation of endosperm development. Several objections may be raised to the hypothesis just pre- 1941] BROWN—CYTOGENETICS OF POA PRATENSIS 509 sented. For example, the increased temperature within a parch- ment bag might result in mutation and therefore in aberrant types. If this were true, however, one would expect comparable results in all selfed lines since the same methods of isolation were used in all plants investigated. As was noted earlier, this is not the case. Some strains, for example, B-6, produce noth- ing but uniform progenies regardless of whether or not the progenies are the product of isolation or open-pollination. As a further check on this point, three strains of P. pratensis which were thought to be apomictic were selected in 1940. Progenies of these were grown after (1) isolation by bagging, (2) isolation by caging, and (3) open-pollination. At maturity all three showed some slight random variation, but within any one strain there were no significant differences between plants produced under the different conditions. Yet in other selec- tions which morphologically were almost identical with some of these lots, marked differences occurred in the number of variants produced after isolation and after open-pollination Behavior of Apomicts in Generations beyond the Fi: In most instances the demonstration of apomixis in Blue grass has been the result of progeny analyses based on one genera- tion of plants. As a result, few data are available on the be- havior of the supposed apomicts beyond the first generation. For example, it is not definitely known whether or not a selec- tion of Poa pratensis which produces a uniform progeny in one generation will continue to do so in following generations, yet, for practical purposes in seed production and strain improve- ment, this information is necessary. Records obtained from three selections of Poa pratensis, each of which has been main- tained for three plant generations, shed some light on this problem. In each of these selections progenies have been studied from both selfed and open-pollinated lines. B-37.—' The plant from which this line was started is typieal of ‘‘average’’ Blue grass, i. e., it is upright in habit, has narrow dark green leaves, average-size panicles, and produces an abundance of seed. The first-generation seedlings obtained after free-flowering were uniform and very similar to the ma- [Vor. 28 510 ANNALS OF THE MISSOURI BOTANICAL GARDEN ternal parent. However, approximately 15 per cent aberrant types occurred in the selfed lines, and these exhibited a rather wide range of variability. The second and third generations behaved essentially in the same manner. The open-pollinated lines were again constant morphologically while the selfs showed considerable variation. B-2.—Although a different type of plant morphologically, the B-2 progeny for three generations behaved very like that of B-37. The plants resulting from open-pollination showed no significant variation, although a number of ‘‘off types" oc- curred after selfing. These were, as a group, less vigorous and of a less desirable type than their parent. There were, never- theless, a few exceptional cases in which promising segregates were obtained from selfed lines. B-30-19.—In this strain, as in the two previous ones, asexual reproduction, resulting in uniform progenies, seems to func- tion without change for at least three generations. In each generation, however, aberrant forms resulted from selfing. TABLE III VARIATION IN PROGENY OF OPEN-POLLINATED LINES Selection Per cent aberrants number 1939 1940 seal B-37 4.32 3.81 4.00 B- 2 1.63 3.21 1.71 B-30-19 0.09 0.09 1.04 These results indicate that in apomictic or otherwise asex- ually reproducing strains of Poa pratensis progenies may be grown or strains may be reproduced by seed without the inter- ference of segregation in later generations. Deviation from type amounting to from 1 to 5 per cent of the total population would not likely have an undesirable effect on these selections when used in pasturage or turf. However, it cannot be as- sumed that all Poa pratensis will behave in a manner exhibited by selections B-2, B-37, and B-30-19. As was shown earlier, С. S. Lot 5 and С. S. Lot 10 produce approximately 20 per cent 1941] BROWN—CYTOGENETICS OF POA PRATENSIS 511 aberrants in progenies resulting from open-pollination. Devi- ations of this magnitude would soon result in a heterogeneous mixture if attempts were made to reproduce the types from seed. Discussion Amphimixis results in many and varied gene combinations and hence in an increased genetic morphological variability, but, due to a variety of both internal and external causes, only a small per cent of the gene combinations produced in any sex- ually reproducing species survives. As a result the majority of plant species remain more or less constant from generation to generation. There are, however, in the plant world a great many groups whose variability makes the delimiting of spe- cies exceedingly ‘‘difficult.’’ It is interesting to note that the majority of these groups are those in which asexual reprodue- tion is the predominant mode of propagation (Dobzhansky, 37). Under such a system any genetic change occurring with- in a species is not usually eliminated as it might well be if the species were sexual, but instead it is maintained and propa- gated vegetatively as a new clone. This soon develops into a variety of forms, a so-called ‘‘species complex." In Poa pra- tensis this condition is further accentuated by the operation of both amphimixis and apomixis within a single strain. Although evidence from progeny analyses indicates that the great ma- jority of Poa pratensis strains reproduce normally by apo- mixis, in each of the progenies studied from 0.9 to 18 per cent of the offspring were found to be the product of sexual repro- duction. In nature a great many of these “пе?” strains would be eliminated due to their inability to compete successfully with other strains, but a number of them might be expected to survive and these will, in many instances, further propagate their kind through apomixis. Thus it is evident that Poa pra- tensis is equipped not only for producing new or aberrant types but is especially well equipped for propagating these types vegetatively through seeds once they do occur. Since both sexual and asexual reproduction have been dem- onstrated in Poa pratensis, information on the factors affect- [Vor. 28 512 ANNALS OF THE MISSOURI BOTANICAL GARDEN ing apomixis is a prime prerequisite to any strain-improve- ment program. However, little information is available on the causes of a breakdown in the sexual reproducing system in plants. Darlington (237) has shown that when such a break- down does occur it is usually replaced by some form of asexual reproduetion. It has been further suggested that asexual re- produetion may be the result of hybridization since many apomiets are known to be of hybrid origin. This is a possibil- ity in P. pratensis which might well have originated as an allo- ploid hybrid. Brown (740) has shown that, despite the extreme variation exhibited by P. pratensis, there is a more or less constant grouping of all variables into two main complexes. One in- cludes plants with wide leaves, many florets per spikelet, long and wide panicles, and many panicle branches. Into the other complex fall those plants with narrow leaves, few florets, short and narrow panieles, and few panicle branches. Regardless of what characters are considered or to what degree the vari- ability may oceur, the varying units always tend to group them- selves into one of these two complexes. This suggests strongly that in the germ plasm of P. pratensis there are at least two entirely different elements. While it might be suggested that the differences between strains are a result of changes within the germ plasm, crossing-over, translocations, inversions, in- terchanges, etc., such differences would tend to occur at ran- dom and not in the distinct groups which have been demon- strated. It is impossible to prove conclusively the alloploid nature of P. pratensis without reproducing it artificially. Nevertheless, all its peculiarities point to a probable hybrid origin. Whether or not hybridization can be shown to be a ba- sis for apomixis remains to be seen, but at least P. pratensis is another example of an apomietie species whose behavior is essentially that of a hybrid. Chromosome Numbers and Apomixis: The correlation be- tween per cent apomicts and chromosome number as exhibited in the G.S. plants is very interesting and perhaps significant. It will be recalled that in G.S. Lot 5 the chromosome numbers 1941] BROWN—CYTOGENETICS OF POA PRATENSIS 513 were found to be either approximately 42 or approximately 49. This progeny contained 18.18 per cent aberrants. Like- wise, G.S. Lot 10, containing 16.66 per cent aberrants, exhibited chromosome numbers of 42 and approximately 42. In con- trast, six typical plants from Lot 1 and five from Lot 2 were found to have 2n numbers of 56. These lots yielded 1.50 and 0.97 per cent aberrants respectively. Although this evidence is not conclusive it does indicate that apomixis tends to in- crease as chromosome number increases. The evidence for a hybrid origin in the 42- and 49-chromosome plants is just as pronounced as it is in the octoploids. Therefore, according to these data, it must be assumed that chromosome duplication (perhaps following hybridization) is at least one of the fac- tors affecting apomixis. On the other hand, it is not likely that any one or any few factors govern the operation of apo- mixis in Poa. Indeed, the isolation experiments suggest that the presence or absence of foreign pollen has a very marked effect on the functioning of asexual reproduction. Genetics of the Various Genoms: Perhaps the most impor- tant contribution contained in this paper concerns the morpho- logical effect of a duplication or elimination of chromosomes within a single strain. It is only logical to expect differences in morphological characters to be correlated with differences in chromosome numbers. Since, however, Poa pratensis is exceedingly variable both morphologically and cytologically and since it is not a plant which lends itself well to cytological study, in the past it has been difficult to arrive at any conclusion regarding the morphological effects of the addition or elim- ination of one or more genoms of the germ plasm. Two lots of G.S. plants, G.S. 5 and G.S. 10, have afforded an excellent opportunity for a study of this sort. These progenies, it will be recalled, segregated into two quite clear-cut types. The tall, upright one possessed narrow leaves and closely resem- bled the parent, while the second type was in direct contrast to the parent morphologically. The 2n number of the first type was 49, that of the second 42. From these data two facts are evident. First, at least two [Vor. 28 514 ANNALS OF THE MISSOURI BOTANICAL GARDEN types of P. pratensis can be transformed into entirely differ- ent types morphologically by the addition or elimination of one set of chromosomes. From the practical standpoint this knowledge should be of considerable importance to those in- terested in producing Blue grass for specific purposes. If one could predict with reasonable accuracy the types of Blue grass expected to segregate from certain selected strains, a great many of the initial difficulties encountered in strain improve- ment would be eliminated. Second, at least some unstable chromosomal types may revert to more stable and potentially sexual forms. Thus, the odd-numbered 49-chromosome plants which would be quite unstable upon cross-fertilization have produced 42-chromosome plants whose chromosomal stability could be maintained regardless of whether reproduction takes place by sexual or asexual methods. On the basis of these observations the extreme diversity of Poa pratensis can be fairly well explained, and our knowledge of the factors affecting the behavior of the species, although still not complete, suggests a number of new approaches to the problem. The presence of a long euploid series within P. pra- tensis results in considerable variation. The various odd- chromosome forms, most of which are being maintained asex- ually, more than double this original variation. The complex- ity is further accentuated by the simultaneous operation of apomixis and sexual reproduction within single strains and the ability of unstable chromosomal types to revert to poten- tially sexual forms. SUMMARY In eleven lots of plants of Poa pratensis grown from seed produced by open-pollination, aberrant types amounting to from 0.09 to 18.18 per cent of the populations have been ob- served within strains. Although the majority of progenies contained from 1 to 5 per cent aberrants, two partially sexual strains produced 16.66 and 18.18 per cent off types. In one lot which was investigated in detail the chromosome number of the maternal-type seedlings was 49, that of the aberrants 42. 1941] BROWN—CYTOGENETICS OF РОА PRATENSIS 515 In some strains of P. pratensis the behavior of progenies after bagging and open-pollination is very different. Certain strains which after open-pollination produce uniform, mater- nal-like progenies may after bagging produce highly variable progenies which are apparently the products of gametic union. To explain this phenomenon, an hypothesis is offered based on an apomietie stimulating effect of foreign pollen. In certain of these cultures the transition from a typical "hay" type Poa pratensis to a low-growing, wide-leaved spreading type of plant was brought about by the loss of one genom from the germ plasm. It is further shown that chromo- some complements composed of odd-numbered chromosome sets may segregate into stable even-numbered chromosomal types. Apomixis and the factors affecting this type of reproduction are discussed in the light of the results of this investigation. Evidence is presented which shows that apomixis is more prev- alent in high polyploids than in those strains characterized by lower chromosome numbers. It is suggested that Poa praten- sis may have originated as an alloploid, partially apomictic hybrid. LITERATURE CITED i ec = Е а über die Samenbildung bei Роа pratensis L. Bot. 9: 213-2 ; pests pd and sexual seed formation in Poa pratensis. Hered- itas 25: s ко J. M. (1937). A cytological study of the genus Poa L. Canadian Jour OD. 281-997. тт b. E (1931). Karyo-systematisehe Untersuchung der Familie Gramineen. Appl. Bot., Genet. & Pl. Breed. 43 Ж. ж pow а 939). Gheombabine complements of five species of bio with an analysis of variation in Poa pratensis. Amer. Jour. Bot. 26: 717—723. , (1940) Variation in Poa pratensis. Master's thesis РИ Жайып University, St. Louis. Darlington, C. D. (1937). Recent advances in eytology, 2nd. ed. 671 pp. London. жо Хе (1939). Genetics and the origin of species. Columbia Univ. p. ант, v. (1941). The development of twin ns in sacs, embryos, and endo- n Poa arctica. R. Br. Canadian Jour. Res. C 19: 135—144. Mining pa (1933). Apomictic and sexual seed ео іп Роа. Hereditas D [Vor. 28, 1941] 516 ANNALS OF THE MISSOURI BOTANICAL GARDEN , (1937). Polyploidy from twin seedings. Cytologia, Fujii Jubilaei Volumen Part 1:211-227. 1937. 1940). Further studies on apomixis and sexuality in Poa, Ibid. 26: 115-190. Tinney, Е. W. (1940). Cytology of parthenogenesis in Poa pratensis. Jour. Agr. Res. 60: 351-360 and О. 8. ed (1940). The progeny test as a measure of types of seed- аи: in Poa pratensis L. Jour. Hered. 31: 457—464. EXPLANATION OF PLATE PLATE 26 ар tion В-1. Typical representatives of progeny resulting from open-pollina- n. Compare with plate 27. ANN. Mo. Вот. GARD., Vor. 28, 1941 PLATE 26 BROWN—CYTOGENETICS OF POA PRATENSIS EXPLANATION OF PLATE PLATE 27 Selection В-1. Typical representatives of progeny resulting from selfing. Com- ак qa All plants shown in plates 26 and 27 are progeny of the same maternal t. ANN. Mo. Bor. GARD., Vor. 28, 1941 PLATE 27 BROWN—CYTOGENETICS OF POA PRATENSIS EXPLANATION OF PLATE PLATE 28 Typical representatives of progeny resulting from open-pollina- | with plate 29. ANN. Mo. Bor. GARD., Vor. 28, 1941 PLATE 28 BROWN-—CYTOGENETICS OF POA PRATENSIS (Vou. 28, 1941] 522 ANNALS OF THE MISSOURI BOTANICAL GARDEN EXXPLANATION OF PLATE PLATE 29 Typical representatives of progeny resulting from selfing. Note the variability and low, spreading habit of these plants as compared with those shown in All plants shown in plates 28 and 29 are progeny of the same maternal parent. ANN. Mo. Вот. GARD., Vor. 28, 1941 PLATE 29 t. x SOR. NE. ч мг BROWN—CYTOGENETICS OF POA PRATENSIS GENERAL INDEX TO VOLUME XXVIII 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 italics; and previously published names and all other matter, in ordi- nary type A Acanthaceae, Panamanian, 465 r er saccharum, 162, 287, var. Schneck- oriai, 6; eei 207; circi- omphocarpoides, 207; hirtella, 207; кы ilis, 206; Pringlei, 206 Acerotis, 204 Aechmea ie 411, 478 All en, РЕ К, OE Е Gen- Woodson Jr. and Robert я г 409 Amaryllidaceae, Panamanian, 414 Ampelamus, 208, 211; albidus, 211, 212; ligulatus, 210 Amphimixis in ae pratensis, 511 Amphistelma, pM ran из; castanea, 232 Ananther Binary variation in Th in plant pee 287 ler an A preliminary ind of the genus Tripsaeum, 2 E Henry Dichophyllum Moorei and certain associated seeds, 375; апа Cortland S. Pearsall. On the flora of the үү formatio on of southwestern Wyoming, 165 imia, 172; ae 172 nemia 'elongat 172; Fremonti, 165, 188, 190, forma fertilis, 168, 180, 182, 188, 190; hesperia, oe occidentalis, 173; va er ini Anisacanthus, A revision of the North American species of the gen us, 385 Anisacanthus, 385, 392; 406; glaberrimus, 402; 394, 406 ; Gre 406, 408, "n junceus, 399; 400, pumilus, 406 ; pumilus, 395; quadrifidus, 393, 398, PA 406 ; Tiarberi, 3 396, 406 ; tulensis, rTP 406; virgular 398; "Wr rightii, 300° 406, var, brevilc id 400, 406 Annonaeeae, Panamanian, 427 Anthanotis, 204 Anthephora hermaphrodita, 268; pubes- Anthistiria gigantea, 268 СЕ E 443; insignis, 441 Anthopteru Apocyn aceae, "Miscellaneous New Ascle- piadaceae 'and, from tropico] Amer- €— 507; after selfing, 507; chro- e num abers а nd, 512 Кыт sly Panamanian, 437 Archibaeeharis panamensis, 472 Ardisia coclensis, 453; geniculata, 454; а" 455; Scheryi, 456; Wood- ет б. The North мен I, Perspective of the genera, 193; при sound Miscellaneous ap 275 naman Asclepias: foral gies ЗА of, 193; southwestern Un ited Ба, E. Aselepias, 204, 216; subg. Acerates, 207 ; subg. clepi odolus, 207; subg. Ascle- pete 207; auriculat a, 207; bra- chystephana 246; circinalis, dd ; Cutleri, 246; Engelmanniana, 207; ubg ү эе ат, 06; E 207; hirtella, 207; humilis, 206; о- leuca, 206; insi a E 286; maerotis, 246; Phenax 6; subg. Polyotus, 207; Pringlei, 206; Ruthiae, 245; S subg. Solanoa, 207; Solanoana, 207; Sper- ‚ 24 7; uncialis, 246; villosa, 232; ithoda ryon, 207 Asclepiodella 205 dora, 204; pre ае 207; gom- р ocarpoides, 207; nsignis, 207; zan- thodacryon, 207 (523) 024 Де: оре 208, ped m 214; pubescens, 215; tahe ensis, Astragalus ая range ef, 336, B "m Tul Amd 178, 377; spectabi- j| 278, sp. 178, 184, 186, 377 Balanophoraeeae Panamanian, 427 la, 178 Begoniaceae, Panamanian, 434 an iit ‚ 389 pyrene, effect of, on yeast cultures, 2,4 Berry, Edward Cain. A monograph of the genus s Parmelia i in North America, radeseantia brae- teata, 147; eri advantages of, €: probable — of, in other ganisms, 162 anamanian plants deter- ined y: е 472; Polyga- em" 432 Blakea brunnea, 435; intercepta, 435 "en 215; ' neriifolium, 215, 244 Borreri debet: 136 reg ‚ Panamani ian, 411 Brow Willia m L. The cytogenetics of Poa ciated, 493 C Callaeolepium, 217; Warscewiczii, 994 Callipteris flabellifera, 376 Camassia scilloides: Mass — = 293; at ан Highlands, 29 at New Athens, 293, 298; не of, 293, pe inflorescences, of leaves, 293, 296, of pedicel and inter- node length, "oi of sepal length and width, 295; f. Petersenii, — Carbon measurement " ev y volumes plotted ика , 17, 18 Carboniferous, plants of the, 171, 375 TEM Som Де mentation 9; on — morphology, 2 Ca wu ВН 389 —— tropicalis, 2; effect of car- inogen 13 Caveniishia Allenii, 445; calycina, 447; Я Сенче 449; gaulthe rioides, [Vor. 28, 1941] ANNALS OF THE MISSOURI BOTANICAL GARDEN a rag ig 433 Cenchrus pu ubes 269 Cephaelis chiriquensis, 469 Cer Chile ме баво Chloris radiata, 268 Cholanthrene, effect of, on Saecharomy- ces ellipsoideus Chromosome o el of Poa pratenn, 500, 503, 505; of Tradescantia brac E 160 Chthamalia, 217; Nummularia, 229; pedun culata, 230; pubiflora, 230 Clavij ] 52 e, 475 Pag — collections from a, 289 angulatus, 254; ан, 254 Collema d tum, Compositae, Panamanian, 472 C wii e i иаа óc 177 Connaraceae, namanian, Con зем Allen ii 43 Conostegia chiriquensis, 436 Contributions toward a flora of Pan- a V. Collections chiefly by Pau H. Allen, and by Robert E. Woodson Jr. and Robert Schery, 409 Convolvulaceae, Pana ian, 46 orni abu. t of the, 165, 172 Cross- несы туку їп Роа pratensis, 494 Cucurbitaceae, Panamanian, 469 oniaceae, ндн, p r, Hug and Edga r Anderson. A preliminary survey of the genus Tripsa Cyatheaceae: Dicksoniaceae, 176 Cyclodo C вора «uu 427 Cynanchum ubg. Ampelamus, 222; subg. Melichampia, 209; ers "Metalepis, und, subg. Me- tastelma, 214; obliquum, 229; Pal- ri, 2 ; prostratum, 230; Re ў rotate 24; saepimento sinaloense, 210; 275; subg. Tylodontia, unifarium, 210; utahense, 215; vil- losum, e viridiflorum, 235; Wat- sonian 210 Pers ostemma, 216 C wii os ‘of Poa pratensis, The, 490 Cytology of yeast cultures, 25 1941] INDEX D Dactylodes angulatum, 254; Dactylodes, 254 Daetylotheea plumosa, 171; Sturi, 171 Decastelma, 2 1 Desmopsis ‘glabrata, Di-benz- ее oct of, on yeast eultur P unis с Carpenteriana, 379, 384 Diehophyllum Moorei: and certain asso- eiated seeds, "E кы of, 375; restoration of, 3. Dieliptera zu 0466, Dictyanthus, 217, 218; aeneus, 237; achistanthus, 236; eratopetalus, 236; parviflorus, 237; Раси. 287; prostratus, 236; retioulatus, 236; stapeliaeflora, 237; tigrin i dt. berosus, 237; SR Qoid Digitaria, 252 Dioscoreaceae, ОЕ 415 Ditassa, 208, Dodge, Carli ‘William, ees Sanford Dodge, and George Thom s Johnson. Some affects of На дебне on yeasts, Dodge, ry И ‘pane Wirata odge, and George s Johns Ви effects of aum on EN EUN ^ a Greggii, 401; juncea, 399, ; puberula, 395, 396; Thur- beri, 29. Wrightii, 399 Dryopteris color adensis, 174 E Ecliptostelma, 243; _ molle, 243 nsis, 210; ligulata, 210 Enslinia, 208 Epicion, 208 Epidendrum NUS 418, 484; ellipso- phyllum, 419 4; gibbos osum, 420, 482; Ae Set an, poo: ramosum var. angu ustifo Epling, C. determined па Equisetaceae, 178 Equisetum sp., 178, 1 Eriekson, Ralph О. ES collections: Camassia чис жи Eriocaulacea гера С, Panamanian Labiatae ‚ 137; furfur cea . Cladonia, 139; fama Аира 13 Ezolobus, 238; albomarginatus, 242 525 F Faramea оо 470 n C. Mass eolleetions Fassett, Norm Rubus спона and R. parvi огоз, 299 iren р of Saecharomyees ellip- V s, measurement of, 11, 16, 17, Fernald в **varieties"" of Rubus parvi- 317 D 165, Fimbristemma, 268; hee 243, 284; stenosepola, 243, 284, Warscewiczii, Fischeria, 237; aristolochiaefolia, 242; viridis, 238 of en SOM formation of southwester ing, On се 165 Floral лан of Дө, 1 Fossil flora: of the Frontier E us of cake жый Wyoming, 165; of Kansas, 375 Frontier. ' formation of southwestern ing: On the flora of the, 165; аа s to Prid ру den forma- MO. 216; Torreyi, 217 G Gentianaceae, Panamanian, 459 Geography: correlation of variations in Rubus кь чаг with, 303, in Е. par- viflorus with, 332 Geology: correlation between characters of Rubus — and R. parviflorus n г the e, б Кын 166 Gesneriaeeae, Panamanian, 465 Ginkgo, 377; biloba, 378 Glaciation in relation to occurrence of ubus odoratus, 308, of R. parviflorus, 332 Gleason » Жо: Жу Panamanian Melasto- mac e determined by, 4 Gleicheniaceae, a Gleichenia pectin 74 Gleichenites = уну А 174, 184, 186, ; Gieseckiana, 175 Pa reflexiflora, 465, 466 TEMO - of, а: growth of yeast eultures, 12, Phi o sd а Pro purpurascens, Gonolobus, 238; Ри 4 231; albo- 242; hypoleucus, 206; a 236; araneosus, 'aristolochiaefolius, 242; arizonicus, 243; atratus, 233; Baldwynianus, 9 28; bayatensis, 226; 526 ANNALS OF THE bicolor, 236; m 230; biflo rus, 228; calcic caudatus, 233; chiapensis, 242; ad huahuensis, 232; chiriquensis, 462; chrysanthus 295; congestus, 224; cteniophorus, 243, '983; cynanchoides, 228; das stephanus, 243; deci j 228; diadematus, 225; Ekmanii, 226; su Eugonolobus, 242; flavidulus, rus, ; Las longipetiolatus, 282 macrophyllus, 212; ; 233; Shortii, 231; oblongifolius, 29, var. brevicoro- порттан, Ds productus peendodarbatus, 243; и 4; unifarius, variifoliu us, 226; vel tinus, 234, var. calycinus, 235; viridiflorus, 235 Gonolobus, 218 : Kentueky blue, 493; Tripsacum, ditat Plains environment, relation of, to variability in n Tradescantia, 158 Growth studies of yeast cultures, 3 H Hackelochloa SIN, 268 acytome f, in counting yeast cells, w Hagen, Stanle ey Harlan. A revision of the ч rth American species of the genus Anisacanthus, 3 жш а епгурһуПа, Heliostemma, 218 Hemarthria compressa, 268 —€— к use of mass collec- tions in, 287, 29 Phu: Акче 817; Pringlei, 223 [ Vor. 28 MISSOURI BOTANICAL GARDEN Hoffmannia Woodsonii, 471 Hostea, [epar ry in Tradeseantia, 153 nia enteromorpha, 50 I Ibatia, 217 ; maritima, 223; mollis, 223; muricata Illinois bottom- „ж? ng in, 29 3, 296, 2 а: Camassias grow- Дел РРР 68; con- spersa isidiata, 76; crinita, 127; diatrypa, 43, 44; dubia, 90; encaw sta, 53; enteromo ; incurvus, 67; olivacea, 60; physodes, 46; quercina, 91; stygia, 57 Ipomoea mucronata, 463 Irmischia, 208 Ischaemum aristatum, 268; glabrum, 254; muticum, 269; rugosum, 268 J Jacobinia, 389 Johnson, George Thomas, Carroll Wil- liam г Dodge, Bertha Sanford Do odge, effects of carcinogens on Jurassic rocks, plants of, 172 Justicia, 089 ; coccinea, 398, 404 ila, 401 quadri da, 398; 398, 401; virgularis, 398, 404 ; pum- superba, K Klukia, 172 L Labiatae, Panamanian, 464 Labidostelma, 218; guatemalense, 224, arizonicum, 243; uk asc 281 Lachnostoma, 238; Balbisii, 232; 3 r is, 50; laciniatus - physodes, E etu. 109 natus, 56; ivaceus, 59; M UE 80; per- амн дај 130; 'perlatus, 122; pertusus, 1941] INDEX 43, 44; physodes, 45; pubescens, 56; quercinus, 91, 92; зала, зсотеиз, 92; stygius, Di: tiliaceus, Di Lichenoides furfuraceum, 137, 138 Liliaceae, Panamanian, 414 Loasaceae, Panamanian, 433 Lobaria furfuracea, 137, 188; incurva, ata, 122, var 4; , 46; sazatilis, 82; stygia, 57; da 43; tiliacea, овы > = е 8 © S ма, e о $ a's Q с Loranthaeea Panamanian, 4 26 Lundell, C. L., Panamanian plants de- by: et ДА d lastraceae, 433; мтч 453; Rutaceae, 431; eophrastaceae, 452 ТУ астма е, 409 Lygonium, 1 Lyonia, 208 Macleania, 443; epiphytica, 439 congestiflora, Macroscepis, 217, 218; 224; vata, 295 б heterom era, 179 pei үле gs Robert E. Wood- son ew A аре from the xa aio United States, 245 о 265 ‚ Nort American, 249 Malovetia Cuatrecasatis, 274; lata, 275 Mandevilla jasminiflora, 27 73; lanci- folia, 274; nerioides, 274; ’ Syrinx, 274 Manisuris cylindrica, 252, p Marantaceae, Panamanian Marsdenia, 201, 243; ыа 244; bicolor, 236; Gilgiana, 244; neri- sicci 244; ‘pseudo- -edulis, 284; ’ Stey- rma 5 2 Mass collections: The technique and use of, in plant taxonomy, 287; Camassia iere E Rubus odoratus an R viflorus, 299; Tradescantia acuminata, 225, arane osus, atrocoronata, 222; Balbisii, 231; 227; bayatensis, 226; belizensis 232; bicolor, 236; biflor ra, 228; brevicoronata, 228; calcarata, 232; calcicola, 224; campechiana, 234; um, - _catoliniensis, 28; castanea, 232; caudata, 233, 277; E 236; yc ahuen- si chry tha, 222; subg. Chthamalia, 227; conge esta, "224; 021 congestiflora, 224; cordata, 236; cordifolia, 222; ‘crassifolia, 236; crenata, 233: cyclophylla, 233; cyn- anchiflora, 278; cynanchoides, 228; Decasnei, 232; decipiens, 228; dia- demata, 225; dictyantha, 236; "subg. . Microdactylon, 236; mollis, 223; nigrescens, 233; msis, 226; Numm ularia, 229; obliqua, 229; ’ oblongata, 226; obo- ciflora, 226; Р; vonii, 237; крек ata, 230; tiolaris, 223; subg. Pherotrichis, 231; picturata, 233, 277; pilosa, 233; pinguifolia, 235; Pittieri, 234, 277; subg. Poicilla, 225; subg. о st mm 232; po rphy rantha, 223; Pringlei, 223; product 230; Prosthecidisons, 223; prostrata, 230; pseudobar 235 subg. Ptycanthera, 205: ubiflora, pueblensis, 223; Purpusii, 223; Qui- , rosii, 224, 281; reticulata, 234; rupestris, 232; ffneri, 230; scopul , 232; Shortii, 231; Sin- tenisii, lag реч nd 237; stape- liaeflor 23 as etal a, 231; кааш, brs: тей. 225; tenuis, 279; subg. ‘Tiarastemma, 232; tigrina, 296; tinctoria, 277; вор Trichosacme, 225; tristiflora, 223; tuberosa, 237; Tuerckheimii, A 280; , 235 ewiczii M Wootonii, 231; yucatanensis, M exillaria Biolleyi, 425; Bradeorum, "M 425; umbratilis, Ee Panamania n plants de- Lye эн эту Ае "A00; 409 Media used in growing yeasts, 3 528 ANNALS Meiosis in еже pratensis, 502 Melastomaceae, Panamanian, 434 Mellichampia, 208; ligulata, 210; rubes- Meramee Highlands, Mo., mass collec- tions of Camassia scilloides from, 293, 298 Meriania en nsis, 436 Меер. are rie 213, 214, 275 анас Methyl cholanthrene, effect of, on yeast cultures, 2, 4, 8 Microdactylon, ^r) cordatum, 236 measurements of Mieroorganisms, of, 5 Mierostelma, 204 puo eee „1176; рег 165, 176, в: variabilis, 118, 117; 1 P oni sen new Foe ера A and Apocynaceae from tropieal America, 27 Moldenke, H. N., Panamanian plants determined и Eriocaulaceae, 411; erbenaceae, 464 Monograph of the genus Parmelia in North Ameriea, north of Mexico, 31 Morton, C. V., Panamanian plants de- termine : Dioscoreaceae, 415; Gesneriaceae, 465; Solanaceae, 464 Mycocandida onyehophila 2; effect of carcinogens Myrsinaceae, E 453 Nanaturis, 208 Nephrodes 243; fruticosa, 244; neri- m Lag" Ti., 88 — of ge cries scilloides from, 293, 2 eo Nun iim. survi ival of thimbleberry on, 299, 332, 345 о Odontonema, 389, 402 Odontoste канен e j 228; carolinensis cipiens, 298; оных , 229; po m 229; obliqua, 229; Shortii, 231 Oligoron, 204 Onistis, 204 pice pollination in Poa pratensis, 497; after, 507 Iur. vv pria а 415 чр 208; acuminata, 225; oblon- a, 226 [Vor. 28 OF THE MISSOURI BOTANICAL GARDEN Otanema, 204 Otaria, 204 Oxalida aceae, Panamanian, 431 деи ите 431 um, 208; eordifolia, 208 205 Ozarks: Camassias growing in, 293, 296, 298 P Pachystelma, 218; cordatum, 236 Palmorchis A Pai 480 Panama, Contribut ward a flora of, V. Сайн за Mut by Paul H. Allen, Robert E. Woodson, Jr. and Parmelia, A m hune of the genus, а North America, north of Mexico, chemical reactions, 38; eco logy, 32; history, 31; morphology, 34; tax 41 "—— 79; alei, 1 € na, 90; hypom ela, 88, esas. ui nr t. in i Canalicularia (sect.), 135; caperata, 113, var, incorrupta, 118 a fuga, 68; ceratophylla var. phyllodes, 46; cetrarioides, — var. rubescens, 127; cetrata, 100, wr ven. 101; о 70; та a, 124; cir- rhata, 136; Cladonia, 189; olpodes, 141; conspersa, 72, var. isidiata, 76; 65; cribellata, 52, 53; =, 127; Со обреда 12 99; cubensis, 108; dubia, 0; usta, a endoleuea, 67; еп- ыш enteromorpha, 50; parmelia (subgen. ), А n.), 42; Hver melia sect.), 135, nii formes (subgen e сё.), 135; ех- "erii erecta, 1 55; Fuparmetia (cube niaeform “а 136, 141; 135; Everniformes (se 66; Finkii gy жюре, 79; potr chyna (sect.), 79; ELLE са 1941] INDEX "Cn 67; leio- 2; margarita a Хатои melia (Euparmelia seet.), бо; lano- рагтейа — (seet.), Menegazzia (subgen.), 42; Mi ichauziana, 124; molliuscula, 71; multispora, ý Rakad lensis, А ; olivacea, уа aspidota, 63, var. glabra, 64; о beo conspurcata, 65; 133; E erla : va obseura, 49, var. platyphylla, 49, var vittata, Е ysodes var. епсаиз а, , var. enteromorpha, 50, var. vul garis, 46; praesignis, 119; praeter- visa, 133; proboscidea, 132; prolixa, 58; Pseudevernia (subgen.), 135; 56; quereina, 91; reticu udia Е ; rudeeta, 96; rudiata, 96; saxatilis, 82, var. Anzi, saxatilis ar. leucochroa, 82, var. omphalodes, ў 8 sa r. laevigata, 109; sorediea, 120; sphaerosporella, 71; stuppea, 141; stygia, 57; subglauca, 119; sublaevigata, 107; sublinearis, 79; submarginalis, 124; subolivacea 60; suleata, 86; sulphurosa, 112; tex- ana tilt y 95; ili ea, 91, var. sublaevi- gata, 107, var. sulphurosa, yu, tine- toria, 133; vagans, 71; villo 135; Xanthoparmelia бл. ы). нн Deas бури (sect.), 67 armotrema Part la sutura, em variation E map oo 352 Patt talias, 208; P i ‚210 iai Cortland в, e y N. Andrews and. On the flora of the Frontier for- mation of southwestern Wyoming, 165 3 , 232; lepto- 32 Philibertia, 216; tomentella, 217; Tor- reyi, 217 Philibertella, 216; Torreyi, 217 Physcia, 4 Piper fagopyricarpum, 426 529 Piperaceae, Panamanian, 426 ете caperatum, 113, 114; perfora- 13 Pleurothallis macrantha, 417, 4 Poa pratensis, The cytogenetics а 493: ing from open-pollination, 497, 516 520, from selfing, 497 518, 522; vari- ation in, 496, Podocarpus A lenti 409 Podostemma, 205 Podostigma, 204 Pogon Mi dei saccharoideum, 269 Poicilla, 217; acuminata, 296; oblon- gata, 226 ; ovatifolia, 225; tamnifolia, Poicillopsis, 218; acuminata, 226; ob- 6 Във inP a pratensis, 497 а ае, нча апіап, 432 s, 204 B еИ 174; Panamanian, 409 So emma, 217; Ui = > 232; scopu- n, 233; viridiflora, 2 Po semiteres, ge а. environment, on of varia- tio n Тездедеъбна be to, 158 Er glacial relic, Rubus parviflorus as a, 299, 3 раны, 218; Berterii, 225; uatemalensis, 223 Psammisia, 443, panamensis, 440 135, ei urfuracn 138 Psygm tru ity 377. ъс XEM 378; Kidstoni, 378; Grosserti, 378 Ptycanthera, 217 nthera Berterii, 225, 226; mollis, Ptycha 223; oblongata, 226; ovatifolia, 225 Q Quadricasaea, 271; caquetensis, 272; inaequilateralis, 2 72 R Ra uwolfia qo 271 Revision of t orth American species of the genus н са s, 38 мера 217; cor "iota '222; fruti- osa, 222 umbellata, 2 Rottboellia "hirsuta, 268, обо; Myurus, Roulinia, 208; ligulata, 210; Palmeri, 210; Rensoni, 210; sinaloensis, 210; unifaria, 210 530 — yor ~~ 211; jaliscana, 209; 210; lignosa, 210; Pal- m 910; fari, 210 Rome M Panaman n, 469 jou 302; $ Eu- lorum, rice. 314; че eden] Rubus odoratus: geographic occurrence of forms in mass collections, 303; vari- ation in, 300, sentation of, with geography, 306, kinds 300; tax- onomy of intraspecific, 302; f. bi- farius, 302; var. columbianus, 302; f. glabrifolius, 302; f. heteradenius, 302; f. hypomalacus, ; var. mala- chophyllus, 302; f. paraheteradenius, 303; f. par ahypomalacus, 303; f. scopulorum, 302 Rubus odoratus and В. Laien uen mass collections of, 299; maps showing lo- eations of, 347-367; habitats of, 368- 374 Rubus parviflorus: един of, 336; forms of, 3 mass collec- tions, 327; kinds of pel oed on, 314; с varieties”? ог, бї]; fe acephalus 24; 2 allocalyx, grandiflorus, glabrifolts 338; 323 . het erad , 323; f. micraden 324; tai 323; var. патио ^ 319, 225 f. scopulo rum, 323; var. scopulor 319, 323; f. trichophorus, 324; var velutinus, 318, 324, 325, f. Ioh ypo- malacus, 325 rah malacus, 325, f. paramicradenius, 2 Г. оре - f. parbifarius, 325, f. parvillosus, 325, f. pra farius, 325; f. silos, 324 сеи Goepperti, 173 ceae, Panamanian, 431 ended 217; dais. 236 S Saeaton, 265 Sae eharomyces ee effect of тк eareinogens оп, 1; cell d of, 36: fermentation of, 11, growth o of, 8, 18; morphology of, г population studies, 1 Samaropsis, 379 Sarcostemma, Torreyi, 217 216; tomentella, 217; [Vor. 28 ANNALS OF THE MISSOURI BOTANICAL GARDEN Satyria Allenii, 451 —— Lindeniana, 423; tenella, вебе а age Schery, Robert W., 72-94 E. Woodson, r. and. Couteiivations toward a flora of и V. Oolleetions — by Paul H. Allen and by Robert E. Woodson, Jr. pelt Robert W. Sehery, 409 Schizaea, 171; elegans, 172 Schizaeaceae, 168; fossil history of the, 171 ———— expansa, 172 Schizaeopteris mezozoica, 172 аон , 205; pur y derer; 207 Schubert, B. G, L. B. Smith and. Pana- manian Begoniaceae scel by, 4 Seeds, Diehoplyllum Moorei and eertain associated, 375, 379 Selfing in e pratensis, 497, 518, 522; apomixis a 0 Senftenbergia ы, 171; Sturi, Seutera, 208 Smith, A. C., Panama plants determined by: Araliaceae, 437; Vacciniaceae, 438 Smith, L. B., Panamanian Bromeliaceae determined by, 411; and B Schu- bert, Panamanian ' Begoniaceae de- termined by, 411 Solanaceae, Panamanian, 464 Solanoa, 205; purpurascens, 207 , p » es, Two new anian plants de- mined by: Rubiaceae, 469; Tax- aceae, на Stelmagonu Stemm m Ае, 4 Steyermark, J. A., ib anian plants determined by: Connaraceae, 430; ie unge, 460 Stylandr a, 204 Sugar, loss of, during growth of yeast cultures, 12 T | bine 208 Tassad 08 Tax ори Panamanian, "> Taxonomy: plant, The echnique = use of mass collections 287; of i traspecifie variation in Ru ratus and R. parviflorus, 302 1941] INDEX 581 нк 217; prostrata, 230 heophrastaceae, Panamanian, 452 Нети 443 Th ximbleberry : habitats of, 299, 365— т; mass einig of, 290 ; occur- nunataks, 299, 332, 345; am pri 303, 347 367; variation in, PEE 306, 314 chys mucronata, 269 а tia bracteata: Binary varia- tion in, 147; pared of, 160; mass eolleetions of, 150; type A, 147, 148; type B, 148, 149; hirsutiflora, 151; virginiana, 147, 151 Trelease, William, Panamanian Piper- aceae determined by, 426 Trichosacme, 217; lanata, 22 Trichostelma, 238; ciliatum, 242, 284; oblongifolium, 243 Tripsacum, A preliminary survey of the genus 249 Tripsacum, 252; utiftlorum, 260; aegilopoides, 208; aristatum, 268; australe, 253, 259; avenacea, 268; ciliare, 268; compressum, 954, 268; cylindricum, 268; dactyloides, 253, lewis DN vias laxa, 269 ropic ca, cellaneous new Asclepiadaceae ka. gere ed from, 27 о 208; cubensis, 214 Tympananthe, 217; suberosa, 237 U Urostephanus, 217; gonoloboides, 222 У Vacciniaceae, Panamanian, 438 Variation: barat E n Tra descantia brac- Siena 147; 0a pratensis, 493; use of mass Ari in study of, 150, 287; in Camassia seilloides, 293; in Rubus odoratus, 300, and R. par mius 314, correlation of, with geog. raphy, 306, relation of, to environ- ment, 158, 296; taxonomy of intraspe- cific ‚309 е бус, Panamanian, 464 Vincetoxicum, 218, 238; acuminatum, 93 " . г 233 eniophorum, 243; cyclophyllum, 233; cynanchoides, 228; dasystepha- num, 243; diadematum, 295; discolor, 234; сараа 229; ; floridanum, 229; Gen tle 4, 234; andiflorum, 285; Gr 230; LeSueurii, E Lundellii, n ‚ 22 gacarphum, ped nigrescens, 233 3; pem "-; на = =: ex "zh 2 = 3 го ws > su = су e LE = > "T r2 r2 "ез Кока: 231; Sintenisii, 296; stenan- um, 243; stenopetalum, 231; steno- pho Да 234; suberiferum, 234; tor- florum, 223; variifolium, 226; velu- tinum, 235; viridiflorum, 235; Woo- n „231 Vit Panamanian, 433 Vost Allenii, 460 W а pseudolaurina, 4 bes anamanian 1 ганда. сае аа by, 415 у Robert Е., Jr. Miscellaneous new Asclepiadaceae and Apocynaceae from j^ aep America, 271; The North Am : f ados © of the gene ra, 193; and 245; and Robert W. Schery. тиви tions toward a flora of Panama. V, —- оч by Paul H atlas and by Robert E. Woodson, Jr. and Robert w. piles 409 532 Wyoming, southwestern, On the flora of the frontier formation of, 165 Y Yeast cultures: cell differentiation in, ‚ 14, 16; ie al en studies, 19, 21, 22 ANNALS OF THE MISSOURI BOTANICAL GARDE [Vor. 28, 1941] N Yeasts, Some effects of careinogens on, 1 Z Zanthoxylum Scheryi, 431 Zea, ^ 9 Zosima, 216 emis md ee эй 424, 4 Zymonema capsulatus, 2; demain 2, effect of careinogens on Vome XXVIIE 2. о: Annals of the Missouri Botanical Garden FEBRUARY, 1941 Some Effects of Carcinogens on Үеавїз.................... Lue Carroll W. Dodge, Bertha S. Dodge and George Thomas ова 1- 30 А se e i ge of the Genus Parmelia in North America, North of МОО аа ааа ра сай Re NAE irse Edward C. Berry 31-146 PUBLISHED QUARTERLY AT FULTON, MISSOURI, BY THE BOARD OF TRUSTEES OF THE MISSOURI BOTANICAL GARDEN, ST. LOUIS, MISSOURI. Entered as second-class matter at the атое ы dtr: Missouri, under the Act of March 3 Annals of the Missouri Botanical байса A Quarterly Journal Containing Seientifie Contributions from the Missouri Botanieal Garden and the Henry Shaw Sehool of Botany of Washington University in affiliation with the Missouri Botanieal Garden. Information The ANNALS OF THE MISSOURI BOTANICAL GARDEN appears four times dur- ing the ealendar year: February, April, Se pets and November. Four numbers constitute a volume. Subscription Price........ $6.00 per volume Single Numbers.......... 1.50 each* Contents of previous issues of the ANNALS OF THE MISSOURI BOTANICAL GARDEN are listed in the Agricultural Index, published by ће Н, W. Wilson ompany. *Except No. 3 of Vol. 22, price $5.00, and No. 2 of Vol. 25, price $2.50. STAFF OF THE MISSOURI BOTANICAL GARDEN Director GEORGE T. Moore HERMANN VON SCHRENK, : EDGAR ANDERSON, . Pathologist Geneticist JESSE M. GREENMAN, Вовевт E. Woopson, JR. Curator of the Herbarium Assistant Curator of à the Herbarium CARROLL W. DODGE, NELL C. HORNER { : logist Mycologis Librarian and Editor -33 of Publications BOARD OF TRUSTEES OF THE MISSOURI BOTANICAL GARDEN President GEORGE C. HITCHCOCK Vice-President DANIEL K. CATLIN Second Vice-President EUGENE PETTUS — L. RAY CARTER GEORGE T, MOORE DUDLEY FRENCH A. WESSEL puram. JOHN 8. LEHMANN ETHAN A. H. SHEPLEY EX-OFFICIO MEMBERS GEORGE В. THROOP, BERNARD F, DICKMANN, Apre adea of Washington Mayor of the City of University St. Louis WILLIAM SCARLETT, A. H. TIMMERMAN, Bishop of the Diocese of President of the Academy of Missouri Science of St. Louis Max G. BARON President of the Board of Education of St. Louis GERALD ULRICI, Secretary