an aie JOURNAL. sah i OF THE pea ARNOLD ARBORETUM HARVARD UNIVERSITY B. G. SCHUBERT EDITOR L. I. NEVLING, JR. C. E. WOOD, JR. LAZELLA SCHWARTEN CIRCULATION VOLUME 46 PUBLISHED BY THE ARNOLD ARBORETUM OF HARVARD UNIVERSITY CAMBRIDGE, MASSACHUSETTS 1965 Mi@sOURL BoTANICAL Garogn LIBRARY DATES OF ISSUE No. 1 (pp. 1-90) issued January 15, 1965. No. 2 (pp. 91-242) issued April 15, 1965. No. 3 (pp. 243-368) issued July 15, 1965. No. 4 (pp. 369-503) issued October 15, 1965. TABLE OF CONTENTS LETTERS FROM CHARLES SPRAGUE SARGENT TO REGINALD SOMERS Cocks, 1908-1926. Edited by Joseph Ewan ..................... Susan DeLano McKe vey, 1883-1964. Richard A. Howard... THe GENERA OF VITACEAE IN THE SOUTHEASTERN UNITED StaTEs. eorge K. Brisicky 5... 6 aa a Dates oF PUBLICATION OF THE JOURNAL LINNAFA: FurTHER EyI- DENCE... 7. FE. Moore, Jr., and C. 8, Wood, J... ComPaRATIVE ANATOMY OF THE LEAF-BEARING CacTacear, XIII. THE OccURRENCE OF WATER-SOLUBLE ANISOTROPIC BopIEs IN Arr-Driep AND ALCOHOL-DEHYDRATED LEAVES OF PERESKIA AND Prresxiopsis. J. W. Bailey A New Cuscuta rrom Nepau. M. L. Banerji and Sitesh Das .... THE GENERA OF POLYGONACEAE IN THE SOUTHEASTERN UNITED States. Shirley A. Graham and C. E. Wood, Jr. .................... Lerrers FROM CHARLES SPRAGUE SARGENT TO REGINALD SOMERS Cocks, 1908-1926 (Continued). Edited by Joseph Ewan .... ANATOMY OF THE PaLM RuaPIS EXCELSA, I. Mature VEGETATIVE Axis. Martin H. Zimmermann and P. B. Tomlinson ............ Stupies in Anyssum: Near EasTerRN REPRESENTATIVES AND THEIR Auuiss, II. Section MENIocUS AND SECTION PSILO- NEMA. T’. R. Dudley THE GENERA OF VALERIANACEAE AND DIPSACACEAE IN THE SOUTH- EASTERN Unirep States. I. K. Ferguson A PRELIMINARY Report ON Funirers. Lorin I. Nevling, Jr. ........ GRISEBACH’s FLorA oF THE BritisH West INpIAN IsLaNnps: A BIoGRAPHICAL AND BIBLIOGRAPHICAL INTRODUCTION. William T. Stearn THE GENERA OF TILIACEAE AND ELAEOCARPACEAE IN THE SOUTH- EASTERN Unirep States. George K. Brizicky POLLEN oF DecENeria viTIENSIS. A. Orville Dahl and John R. Rowley LETTERS FROM CHARLES SPRAGUE SARGENT TO REGINALD SOMERS Cocks, 1908-1926 (Continued). Edited by Joseph Ewan. .... NoMENcLATURAL Notes on Drpsacus FULLONUM AND Dipsacus sativus. I. K. Ferguson and George K. Brizicky .................... THE Rusiacrous Genus MussaenpA: A NEW SPECIES FROM Nortu Burma. Don M. A. Jayaweera Potten MorpHonocy in THE JUGLANDACEAE, II: SURVEY OF THE Famity. Donald R. Whitehead Foe ay LETTERS FROM CHARLES SPRAGUE SARGENT TO REGINALD SoMERS Cocks, 1908-1926 (Concluded). Edited by Joseph Ewan... 411 q CoMPARATIVE ANATOMY OF THE LEAF-BEARING CacTackak, XIV. ° | PRELIMINARY OBSERVATIONS ON THE VASCULATURE OF Cory- { vase cyte Uk Auk... |, RN fora Oe Nteeea ici OEEe Lee eaa eee 445 “PROTEIN Bopres.” I. W. Bai OY seis Sane cesar SAE. eee cate 453 : FS SR CTOn Be: Reais .shesacdccaksdsctiih.. 006 i OOR> mein. 465 SRIEE, DY VORNADO. i001 SRR) ag oR aks nck? 493 VoLUME 46 NuMBER 1 JOURNAL OF THE ARNOLD ARBORETUM ‘HARVARD UNIVERSITY B. G. SCHUBERT EDITOR L. I. NEVLING, JR. C. E. WOOD, JR. LAZELLA SCHWARTEN CIRCULATION JANUARY, 1965 PUBLISHED BY THE ARNOLD ARBORETUM OF HARVARD UNIVERSITY al : TSOUFI SoTaNIcal CAMBRIDGE, MASSACHUS: ee ge Re: “aa a PEs THE JOURNAL OF THE ARNOLD ARBORETUM Published quarterly by the Arnold Arboretum of Harvard University. Subscription price $10.00 per year. During 1965 volumes I-XX will be reprinted and available from ey Reprint Corporation, 16 East 467TH Srreet, New Yorn, N. Y. 1 Subscriptions and remittances should be addressed to Mrs. Lazenia ScuwartTen, ARNOLD ArBorETUM, 22 Diviniry AVENUE, CaMBrRiIDGE, Massa- cHussEtTTs, 02138. CONTENTS OF NO. 1 Lerrers FROM CHARLES SpracuE SARGENT TO REGINALD SOMERS Cocks, 1908-1926. Edited by Joseph Ewan 1 Susan Detano McKeEtvey, 1883-1964. Richard A. Howard Tue GENERA OF VITACEAE IN THE SOUTHEASTERN UNtTep STATES. George K. Brizicky Dates OF PUBLICATION OF THE JoURNAL LinNAEA: FuRTHER Evi- pence. H. HE. Moore, Jr. and C. EZ. Wood, Jr. Comparative Anatomy or THE LzAF-BEARING CacTacnan, XIII. | Tue OccurRRENCE or WateR-SoLuBLe Antsorropic Bopres In Am-Driep anp AtcoHot-Drenypratep Leaves or PERESKIA AND Pereskiopsis. I. W. Bailey A New Cuscura rnom Nurat. M. L. Banerji and Sitesh Das “Volume 4, No. 4, sectudiny pages 401-505, i spas oe et come &. sti JOURNAL OF THE ARNOLD ARBORETUM VoL. 46 JANUARY 1965 NUMBER 1 LETTERS FROM CHARLES SPRAGUE SARGENT TO REGINALD SOMERS COCKS, 1908-1926 * Edited by JosEpH Ewan ! WHEN CHARLES SPRAGUE SARGENT DIED IN 1927, after serving 54 years as Director of the Arnold Arboretum, he left a varied and significant legacy of botanical accomplishment. The Arnold Arboretum which he established is a monument to his vision, energy, and ability. His published works are a record of his interest in and knowledge of woody plants. After comple- tion of his magnum opus, the Silva of North America (1891-1900), Sar- gent wrote the Manual of the Trees of North America which was, in fact, a condensed version of the Silva, available to many more interested people than had access to the original work of 14 folio volumes. e production of a manual with its necessarily concise statements of characterization, distribution, uses, and so forth, entails an enormous effort and a refinement of style not easy to achieve. To obtain the greatest pos- * In response - a request by the editors of the boom of the Arnold Arboretum, well known, has Beier for publication this series of salt from Professor Sargent of Volume 46, history of a floristically critical area, the southeastern United States, now being studied intensively. Eds. * The 302 letters of Charles Sprague Sargent reproduced here have been faithfully Tulane University. They were pldagenn misspelled place names and a few obvious rected, All the letters carried the heading “Jamaica Plain, Mass.,” althoug Arboretum during Sargent’ s excursions. Salu- e Personnel” (Rochester Chapbooks, Publisher, 1961, 51 pp.), chiefly to Mr. JoHN - BAR (1859-1927), with footnotes identifying most of the persons ations 2 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 sible precision in statements of range and other details, Sargent encouraged correspondents in various parts of the country to provide specimens and all available information about the trees in their particular localities. From the time of publication of the first edition of the Manual Sargent worked for approximately fifteen years toward the completion of the second edi- tion. Among the many correspondents who contributed to the effort, Sargent especially mentioned, in the preface to the second edition, Alice Eastwood who collected in Alaska and New Mexico, and Professor Cocks who explored the forests of Louisiana. Sargent apparently first met Professor Cocks early in 1908. Unfor- tunately we do not know the content of the first letter Sargent received, for incredibly Cocks’s incoming letters were not kept together, and only occasional letters or paragraphs from them may be found pocketed with herbarium specimens in the Arnold Arboretum. Professor Sargent quickly detected the aptitudes and values of his newfound correspondent, for his 302 letters to Cocks, reproduced here, were written sometimes daily, once two in a day, and with few long intervals, until twenty days before Profes- sor Cocks’s death on November 21, 1926. Sargent himself died four months later, on March 22, 1927. Reginald Wodehouse Somers Cocks was born August 31, 1863, at Pap- worth, Worcestershire, England. He attended Trinity College, Cambridge, where he took the M. A. degree, in 1889, with first honors in classics. His interest in Latin persisted through his lifetime. There is a letter from Ben- jamin Lincoln Robinson, Curator of the Gray Herbarium, dated February 13, 1917, thanking Professor Cocks for his estimate of a paper which con- tained some Latin with which he disagreed. In this Dr. Robinson wrote, “T have no doubt that you are entirely right that my Latin contains various slips. Although there are Cocks specimens in the Tulane Herbarium labelled “Feliciana, March, 1892,” and “Feliciana, April, 1895,” as well as some collections made in 1897, Professor Cocks’s active botanizing began in March, 1898, and continued through that season. On May 27, 1898, he presented his first paper at the meeting of the Louisiana Society of Natural- ists on Selaginella apus |S. apoda| growing in profusion at Mandeville, Louisiana. The Society had been organized July 3, 1897, and ape iba May 1, 1899, with Professor J. H. Dillard as its first president. “Mr S. 0c ks, M. A.”’ is listed as a member of the executive committee for sis year 1898-99, at the time he was an instructor at Boys’ High School in New Orleans. Cocks and Professor George E. Beyer of Tulane completed a series of “excellent lectures on botany and zoology” given under the auspices of New Orleans Free Lecture System, according to a report in the Gulf Fauna and Flora Bulletin for June, 1899. June, 1899, included trips as far afield as Lake Charles prairies and Selma, Alabama, the home of Mrs. Cocks, and August, excursions to Alexandria and vicinity. From 1900 on, year after year, regular field trips, with special attention given to the y plants, were made into all sections of the state. Cocks, like another eles, Philip Henry Gosse, first arrived in Canada from England 1965 | EWAN, LETTERS FROM SARGENT 13) COCKS REGINALD WODEHOUSE SOMERS CocKS JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 CHARLES SPRAGUE SARGENT 1965 | EWAN, LETTERS FROM SARGENT TO COCKS 5 but evidently did not stay in the north. A few herbarium labels suggest that he may have arrived in New Orleans as early as 1890 but this evidence is not wholly trustworthy. On other occasions specimens from opposite sides of the state carry the same month; yet it is highly unlikely that in those days he could have moved about so rapidly. Cocks generally recorded only the month and the year on his labels, and did not number his collec- tions. Professor Sargent subsequently numbered some sheets for easy ref- erence. For example, a Carya collection from Clear Lake, Natchitoches Parish, May, 1909, later became number 1765; another from Bayou La- combe, St. Tammany Parish, across the state, also May, 1909, is number 1766. After a year as Professor of Botany at Louisiana State University, 1906-07, Professor Cocks moved to Tulane to occupy the newly created Ida A. Richardson Chair of Botany. Mrs. Richardson took an active interest both in living plants and in books about them (our Tulane set of Sargent’s Silva was her gift). It is notable that, on several occasions, Sargent mentions sending seeds or plants of foreign introductions, recently grown at the Arnold Arboretum, as possible additions to New Orleans gardens. This was in the spirit of Ida Richardson who is credited, and with good evidence, for the introduction of Cocculus laurifolius, Siphonan- thus indicus, and other exotics into local gardens and patios. Through the years Professor Cocks supported the activities of the New Orleans Garden Society, and a volume of his collected papers was presented to the Society by one of its members, Mrs. J. Leo Burthe. Over and beyond his lecturing to garden clubs and his teaching of botany classes (his lecture notes have survived) was his enthusiasm for searching out new plant records for Louisiana. Here Sargent found a devoted colleague. It is abundantly clear from his letters that Professor Cocks conscientiously ran the errands Sargent put to him. And the errands kept coming, in extremis with Sar- gent’s advancing concern that the revised Manual might not be finished before his life was ended. The first edition of the Manual had appeared in 1905 and immediately corrections and additions began to accumulate in the desk copy. When another collector provided a range extension in Louisiana, Sargent wrote Cocks to verify it. Mattoon found Pinus car- ibaea [P. elliottii var. elliottii]! in the Tangipahoa region, and Sargent wrote for Cocks to hunt for it in other parts of the state where it might have been overlooked. From these collecting trips Sargent recognized five new species or varieties of hickory, a new hybrid oak (Quercus « cock- sit) from near Alexandria, a hawthorn (Crataegus cocksii) from near Winnfield, and a linden (Tilia cocksii) from near Lake Charles. Sargent accompanied Cocks on several field trips over the state, testing revised keys in the field, collecting specimens in flower and fruit, and following up old records made by nineteenth century botanists. The two men marked trees to return to for successive vouchers, especially the oaks where the yields of fruit vary from year to year. The zeal of the plant collector and the comradeship of the botanists por- trayed in the letters which follow may be gleaned from a letter by Ernest ‘ 6 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 Jesse Palmer (often mentioned in Sargent’s letters) to Professor Cocks dated March 5th, 1926, the year of his death. Palmer wrote ‘With refer- ence to the Yucca louisianica at Ruston L[ouisian|a, of which I sent you a small snap, I regret to say that I have no duplicate material. When I saw it and made the photographs I was travelling with Dr. Small in his truck (the famous ‘weed wagon’). I took some specimens, but as you know, flowers of these plants are very hard to dry, and as we didn’t have much chance for changing dryers on the trip practically all of the material was spoiled, and I only managed to save enough for one meager specimen for our herbarium. “T was sorry to learn that you have been laid up. Have had a bad cold myself the past month, but have managed to keep going. | Palmer lived thirty-six years after this bout.] Professor Sargent, I am glad to report, has been in much better health and has been at the Ashoxctum almost every day during the winter. He is truly a remarkable man.’ ACKNOWLEDGMENTS My thanks first, to Professor William T. Penfound, of the University of Oklahoma, formerly at Tulane, for insuring the preservation of the Sargent letters during his years at the University. To Charlotte Ponder for her careful typing of the manuscript, and to Nesta Dunn Ewan for her critical editorial eye and her assistance with the indices, I express my very real appreciation. April 7, 1908. My Cear Professor Cocks: I want to tell you what a very pleasant day I had in Shreveport, thanks to your letter of introduction. I saw the Mesquite and the Gleditsia, although they were not yet in flower. The Gleditsia is certainly very close to Gleditsia aquatica. There is only one plant of it. I should say now, without having seen the flower, that it would be best to consider it a variety of that species, although the longer pods and greater number of seeds is remarkable. Possibly other individuals will be foun I noticed in Jackson Square, New Orleans, three or four small trees of what I thought to be an eastern Asiatic Acanthopanax which I did not recognize. Do you know anything about these trees, and can you dry some specimens of the foliage for me? Is it possible to find out where these plants came from? I am afraid they are not old enough yet to flower. This short visit to Louisiana impressed me more than ever with the beauty and interest of the state and I feel very keen to get back and to visit those parts of the state which I have not seen, namely the region east of Baton Rouge and west of Vicksburg. I hope another year we may be able to manage a trip together. It is a great pleasure to me to make your acquaintance and I hope that I shall hear from you again. 1965 | EWAN, LETTERS FROM SARGENT TO COCKS 7 z April 21, 1909. You remember that you sent me a specimen of a Tilia from Shreveport which I told you was probably Tilia americana. I am satisfied now that it is not Tilia americana, and that it is perhaps Tilia leptophylla which you will find mentioned in Small’s Flora. I very much want to get flowers of this Louisiana Tilia and to know something of its distribution in the state. The same plant appears to grow on the Red River in Arkansas. Can’t you get your correspondents to look out for this tree in Louisiana and especially to obtain good flowering specimens as these have not, so far as I know, been collected? I wish we were in the woods together now. 3 December 8, 1909. I now expect to be in Boston at the end of the month and, if you are coming here, I hope you will stay at my house. You can reach Boston very easily from there and also the Arboretum. It would be a great pleasure to me to see you again and to have you as my guest. There are many sub- jects we ought to discuss about southern botanical explorations. 4 January 17, 1910. I am very glad to hear that you got home safely and are enjoying warm weather again. Here we have had bitterly cold weather and a good deal more snow, and I am longing for Louisiana or some other country of a oa climate than New England. Many thanks for your package of plan pee is certainly Aesculus austrind. The Bumelia from Slidell without fruit is B. lanuginosa. The others seem to be correctly named. It was a great pleasure having you in Brookline. 5 April 8, 1910. I write to say that I got home safely and to thank you again for all your kindness to me during my visit in Louisiana. I never had a better week or saw finer or more interesting trees, and we certainly must have another trip together before long. I understand that you will obtain flowers of Aesculus Pavia for us, flowers of Crataegus and Persimmon from Lucknow, and fruit of Salix from Shreveport and Bayou Sara. Any other specimens will be most acceptable. I am curious to hear about the Willow at New Orleans. The Shreveport Willow grows at Fulton, Arkansas, and also at Allenton which is about 8 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 thirty miles from St. Louis and apparently is a common and widely dis- tributed tree. I cannot believe that it is only Salix nigra. A little more study of Louisiana Willows is going to lead to all sorts of surprises. The trouble is we have all been taking too much for granted. I find that we have n’t here a copy of Dodson’s Bulletin on Northern Louisiana Shrubs. If you can get a copy of it for the library here, I shall be much obliged. I hope you got home comfortably and found everything all right at home. The mud at Fulton was about three feet deep. I never saw such roads before, but it did not rain and I was fortunate in finding in full flower the Oak I was looking for. I had a good day at Allenton and found there at least five species of Prunus, all of which had been passing for Prunus americana, and so it goes. 6 April 21, 1910. Glad to get your letter of the 18th. I am much interested in what you say about Hamamelis. Have you perhaps found a third species? The Mis- souri plant flowers in February and March in a region which must be several weeks earlier than Louisiana. The habit you describe, too, is dif- ferent. The Missouri plant grows always in the sandy and rocky borders of small streams and spreads into large thickets by underground roots. I should judge from what you say that your plant has not this habit. How about the Willows at New Orleans and also the Diospyros at New Orleans? Did you get the fruit of the Willow at Bayou Sara? I am anxious to hear fuller accounts of your trips since we separated. It is a very early spring here indeed. 7 May 16, 1910. I have yours of the 10th and will write you about the plants when they arrive. Pearl Island must be directly at the mouth of Pearl River. I think I have seen the name on old maps. If you have access to a copy of Bar- tram’s Travels, notice what he says about it. I suppose you know that he went as far as Baton Rouge. I cannot think that there is any trouble in finding this island, and I have no doubt the Plum-tree is still growing on it. I very much hope that you intend to pass August in this part of the world. We will try to give you as good a time as possible. 8 June 7, 1910. Very sorry to hear that you had such a bad time on Campbell Island,? but the fact that you found no Plum tree there may save us a lot of trouble in the future. I have been very busy lately and have only just now been able to look at your plants. We shall send you soon the names of several of them. ? Lake Pontchartrain. 1965] EWAN, LETTERS FROM SARGENT TO COCKS 9 With regard to the Hamamelis, your plant from Richlands appears identical with the spring-flowering species from west of the Missouri River. Probably it is the same thing which grows at Alexandria, Louisiana, I do not feel so sure of the Covington plant. I collected it there myself and have always felt doubtful about it. We must try and get flowers or find out just when it blooms. I find that Mohr in 1885 collected your Lindera at Opelousas. We have what looks like the same thing, too, from Georgia, and in Missouri there are plants with less pubescence than yours which seems to connect it with the northern form, J am inclined to think now that your plant will have to be considered a variety. Your fruiting specimen which we have was collected June 15th and of course — the fruit was ripe. You ought to try and get mature fruiting specim The Richlands Diospyros panes looks very different both in the character of its tomentum and in the shape of the leaves. I presume that the trees in the New Orleans swamps are the same thing, although the leaves are certainly less cordate and more like those of more northern forms but the tomentum seems identical. The thing now is to get the fruit and the mature leaves. I hope you will be able to manage this. We ought, too, to have mature leaves of the Willows which you collected, and also mature leaves and fruit of the different plums. This I am afraid will prove more difficult, but perhaps you will be able to manage some of them. Any hope from our friends at Shreveport in this matter? The plum business is very difficult owing to the difficulty of getting material and finding good characters by which to distinguish the different forms. I hope you are arranging to be in this part of the world this summer, June 13, 1910. Your Vacciniums are, — 72, V. Elliottit, 79, V. Elliottit? 80, V. stamineum, 81, V. vacillans, 89, V. stamineum, 90, V. vacillans, 99, V. glaucescens, 101, V. vacillans. This I believe disposes of your unnamed plants except the Willows and Persimmons, and it will take some time to work these up 10 October 12, 1910. Many thanks for the three lots of plums which arrived today in capital condition. I have to thank you, too, for the Persimmon fruit which came sometime ago. So far as the shape of the fruit of the two kinds is con- 10 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 cerned I see nothing very remarkable about it for the fruit of the common Persimmon shows as much or more difference in variation. The pubescence on the Persimmon, too, seems to be a very variable character. The fact that one form grows in water continuously, or nearly continuously during the year, and the other only on dry land is remarkable and we ought to find some other characters beyond the shape of the fruit to separate these. I think there is a difference perhaps in the time of blooming and possibly in the size of the flowers. Is there any thing new in the Salix situation and have you foliage speci- mens for us? I hope you are well and enjoying life, and have no University troubles on your hands. 11 December 14, 1910. I am much obliged for your letter of November 28th (I am ashamed to see how long it has remained unanswered) and for the package of plants. 3£8, Pearl River, is certainly Asimina parviflora. I do not know any A. pubescens. #9, C vii. of Shreveport, is Crataegus spathulata. #11, C iii. of Shreveport, is Crataegus arborescens. 413, Amelanchier canadensis |var.| tomentula Sargent is the best I can do for this. It seems the same as the Covington tree. +22 and #23, ii. and iii., Shreveport, Crataegus spathulata. I cannot say anything about the plums yet; indeed I am more ignorant about them than I was a year ago, still we will do something with them sooner or later. As you know, it is an extremely difficult genus to collect, because they flower when there is nothing else to collect and the fruit ripens early when it is hot and botanizing is disagreeable. I have never seen flowers or fruit of Cephalotaxus japonica, which seems to be an abnormal garden form with erect branches of some other species. If you have specimens of the flowers and fruit, we shall be very glad to get them. Have n’t you got a lot of Willow material for me? That southern Black Willow question is not yet settled. Indeed there is so much work to do in the south that I do not know which way to turn. I certainly hope to be down in your part of the world again this spring. We have received through the Department of Agriculture a specimen of Tilia collected at Lake Charles which is perfectly glabrous and looks like Tilia floridana. We should investigate it. We have had some cold weather and some snow, but it is not as bad as it was when you were here last year. 12 December 31, 1910. I hope you are going to be able this winter to further investigate the Louisiana Hamamelis. I am describing the Missouri and Arkansas spring- 1965 | EWAN, LETTERS FROM SARGENT TO COCKS 11 blooming plants as Hamamelis vernalis. The specimens from Covington collected by myself and by you and your specimens from Alto, Richland Parish are certainly different in the pubescence from the Missouri plant. Of course we have n’t got any flowers of any Louisiana specimen. In the Gray Herbarium there is also a Hamamelis collected at Alexandria which I suppose is the Richland Parish plant. Why can’t you take up this mat- ter this winter and get flowers if possible from all these stations, or at least find out whether they do or do not flower in the winter or early spring? 13 January 9, 1911. Very sorry indeed to hear that you are under the weather. Take care of yourself, pray, for there is lots of work ahead of you if we are ever going to solve all the mysteries of Louisiana trees. It is my hope and intention to pay you a visit this spring either before or after I go to Texas, or at both times. I have not formulated my plans yet. Mr. Rehder here is working over the North American species of Malus and is describing a number of new species and forms. On April 3rd, 1885, I collected at Pinnyville [Pineville], Louisiana, which is just across Red River from Alexandria, specimens of an Apple growing in a glade in the pine woods which we used to refer to Malus angustifolia. I do not believe it is that species and probably it represents an undescribed one. It ought to be looked up and fruit should be collected. Of the true Malus angusti- folia we have from La. two specimens, one collected by me March 26, 1886, at Opulusas [Opelousas], and one collected by you at Bayou Lacombe, No. 1779. I do not remember where Bayou Lacombe is. Both these Louis- jana specimens differ from the specimens of the southeast, Georgia, Florida, etc., by the pubescence on the leaves and petioles. I should suppose Malus angustifolia would be a common Louisiana plant, especially in eastern Louisiana as we have specimens from Pearl River Valley in Mississippi. This is a thing which ought to be looked up. I have just finished describing seven species of Prunus from Arkansas and Texas, and I do not believe that these descriptions cover any of your Louisiana trees. Prunus is a genus which certainly has got to be more carefully studied in Louisiana. I think we have pretty good material of that pubescent Willow but I hope you will be able to collect Willows wherever you see them in Louisiana as we want more material and more information about these different forms of the Black Willow. I am suggesting calling the Bayou Sara tree Salix nigra variety altissima, for as it grows at Fulton, Arkansas, this is the tallest of all the Willows I have ever known anything about. The amount of material we have representing Louisiana woody plants is still unsatisfactory and my knowledge about them is even more so. I am counting on you for a good deal of help. With kind regards and best wishes for the New Year, I am iz JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 14 February 10, 1911. I have today your letter of February 6th and the specimen of Oak. This looks to us like the tree which they call in England the Lucombe Oak and which is supposed to be a hybrid between Quercus Cerris and Quercus Suber. Hope you will get plenty of Willow material and the Witch Hazel flowers if possible. To show how poor we are in Louisiana plants I could not find this morning a Louisiana specimen of Liquidambar in our herbarium. I am going to Florida on the 18th for a few days and I shall return home before going to the southwest. I want to go to Texas before Louis- iana, so it will be April, I suppose, before I reach New Orleans. I do not care much where I go in the state but we ought to look up that Tilia at Lake Charles; and I should like to go, too, to some more places east of the Mississippi. Opulusas [Opelousas] is always very interesting to me, so I shall be entirely at your disposal. 15 February 15, 1911. I write another line to remind you that there is a lot to do on Prunus in Louisiana in the hope that you may be able to get good flowering material from trees so located that we may get fruit from them later in the season. This is one of the most difficult groups to work up as the fruit is hard to preserve, and it flowers so early that it is necessary to make special trips to get them. 16 March 6, 1911. The two specimens have arrived. Certainly this is not the same as my spring-blooming Missouri and Arkansas species which has the inner sur- face of the sepals bright red. This plant of yours is remarkable in the small size of the flowers and of the fruit which is not more than half the size of the ordinary Hamamelis fruit. This may be the variety parvifolia of Nuttall, a variety afterwards taken up by Torrey & Gray but dropped by recent authors. Judging from the time it flowers, your plant may repre- sent a third species; at any rate we have got to know more about it. Will you be on the lookout for a yellow-twigged Tree Willow which is common in eastern Texas and which you ought to find in the Shreveport and western Louisiana region? This is either a species or a good variety of Salix nigra and we want to know more about it and to get more material. I am hoping now to go to Texas on March 15th and to meet you at Lake Charles on my way back, but I will write again before I leave home. 17 March 14, 1911. I am glad, or perhaps really sorry, to hear that Plum trees are so abun- dant in western Louisiana. I am leaving here on Thursday, the 16th, and 1965 | EWAN, LETTERS FROM SARGENT TO COCKS 13 go via St. Louis to Texas. I cannot exactly fix now the date of our meeting at Lake Charles but I will telegraph you from Texas as soon as I know. may not be able, however, to send word very long in advance, so I hope you will be ready to start at short notice as it will be horrid to be tied up alone at Lake Charles. Will you send me a line to the post-office at San Antonio in about a week from today and tell me if everything will be all right? I think it would be an excellent plan to pay Tracey a visit as I have never met him and possibly something might be gained by doing so. 18 April 4, 1911. I got home safely and very comfortably as that was an excellent train, to find here the coldest April Ist on record. The ground is still deeply frozen, new ice has formed on the ponds and there is not a sign of leaf or flower of any kind. I wish I were back with you in New Orleans. Our Lake Charles Tilia differs from Tilia leiophylla in the absence, or almost absence, of pubescence and in the serration of the leaves. Tilia leiophylla has generally rather coarsely serrate leaves, but in all the Lake Charles specimens the teeth are reduced to mucros. Still I think we must consider our plant a form of Tilia leiophylia, especially as we have speci- mens from Fulton with practically glabrous leaves. It will be interesting and important, however, to get flowers and fruit from Lake Charles. I have specimens of all the Louisiana Crataegus described by Beadle and I feel quite sure that the Lake Charles Crus-galli tree which we saw is undescribed. I hope you won’t fail to get the fruit, and I greatly regret that we did not get flowers or make specimens. I am quite impatient for the specimens of the Crataegus from Covington to arrive for I cannot be- lieve that this is C. brachyacantha. The bark is different and the peduncles and pedicels are much shorter than those of that species. If it turns out to be new, this is one of the most interesting Crataegus discoveries that I have made in years. C. brachyacantha is common at Minden and Munro[e], and I see that Bush collected it at Natchitoches. As you perhaps know four Crus-galli species were collected by Beadle near Opelousas. My C. edita appears also to grow there, and of course C. berberifolia. About Shreveport there are several Crus-galli species, but I think you will find the entire western border of the state full of species of this group. I am very much interested in that Crataegus with white bark, cordate leaves and partly grown fruit which we found on our drive from Covington near the schoolhouse. I thought it might be C. silvicola but the leaves as I remember them are much too cordate for that species I feel quite sure that the Covington Viburnum is V. pubescens. Vibur- num dentatum seems to be confined entirely to the Appalachian region. T ought to see what you have taken for V. molle. I hope you won’t forget good specimens from any localities of the Louisiana Sambucus as this genus is pretty puzzling. 14 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 I have ventured to order and send to you a telescope for your driers after the pattern of mine, which has proved very satisfactory in strength and lightness. Please accept it in memory of the pleasant days we have passed together in Louisiana. Notice, too, that I have had it made much deeper than my own and draw from this fact your own conclusions. I hope to hear from you very soon about your work since I left. If Prunus and Crataegus are to be finished within three years there is not a moment to lose. 19 April 6, 1911. The Vitis which we collected at Lake Charles is V. canescens. You col- lected the same thing two or three years ago at Richland Parish and called it V. labrusca. I do not see that the books credit Louisiana with V. canescens. The Highbush Blackberry of which we saw so much at Lake Charles and Covington is Rubus Andrewsianus of Blanchard. 20 April 19, 1911. The package of plants has arrived and I am very glad to see them. The Malus from Covington appears to be Malus angustifolia, but we want to get good fruiting specimens of this and a handful or two of fruit. It is an entirely different plant from the one at Pineville. This I hope you have looked up for I very much want to get the fruit, having good flowering specimens already. The Viburnum from Glen Gordon is not Viburnum pubescens but V. scabrellum. The Crataegus I cannot distinguish from C. i aaa ae and I be- lieve we shall find it has blue fruit. The Euonymus from Covington is E. americanus. What you call Vitis canescens from Glen Gordon is Vitis cordifolia and different from the Lake Charles plant which I told you by mistake was canescens. There is no Vitis canescens; it should be V. cinerea. The Cov- ington plant seems different from that from Lake Charles. The shrubby Vaccinium from Covington is V. stamineum. You have also collected this from Shreveport. It is desirable to get good fruiting specimens of this The Rubus from Claybourne is the same as the tall-growing species from Lake Charles. The oldest name for this is R. argutus of Link, which must supersede R. Andrewsianus. The little Rubus from Glen Gordon with prostrate stems is R. trivialis. The Dewberry from Lake site with the large flowers and fruit is, I feel satisfied, an undescribed s I hope you found the Malus from Opelousas which, if it is pipet ae is a very tomentose form of it. I hope you are going to be able to clear up many doubts and mysteries id SU Neier es 1965 } EWAN, LETTERS FROM SARGENT TO COCKS 15 this season and that you will make good headway with Prunus, Malus and Crata 21 June 10, 1911. Do you know anything about the writer of the enclosed? Possibly he may be made useful as a collector or in connection with your proposed garden. I shall hope to hear from you soon again about the various plants in which we are interested. 22 July 22, 1911. The enclosed may interest you, although I daresay you are in com- munication with this lady.* I suppose it was her uncle who was associated with Riddell. I know of no other publication by Carpenter. How are you getting on and what is the Plum situation? I hope good, for in Texas and Arkansas there is no fruit whatever. And how about Malus, Crataegus, etc.? I hope that you are not worn out by the heat and that I shall soon hear from you again. I cannot think that it has been hotter in Louisiana than it has in Boston this year. We have never known anything like it or ex- perienced such a drought. Let me hear from you when opportunity offers. 23 August 26, 1911. How are you and what is the botanical situation? I hope that you won’t be too late for the fruit of Aesculus Pavia from Mandeville as we want it for a subject for one of Mr. Faxon’s plates. I think it would be well to gather it before it is ripe, although it should be fully grown. Perhaps the best way would be to wrap up a fruiting branch in paper and send it by mail. This might give better results so far as the drawing goes than a dried specimen. Any word of the Malus from Pineville which we are very keen to get hold of? Prunus in Texas and Arkansas is behaving very badly this year and there is no fruit on any of the plants. I hope it is doing better in Louisiana. I am just back from a short trip to Colorado where I had a few good botanical days and from whence I returned via the Great Lakes, which I found a most interesting trip. 24 December 19, 1911. It is a long time since I have heard from you. I hope you are well and that the botanical situation is not too much for you. * Mrs. StaucHTer of Ruston, La., who provided information on WM. M. Carpen- TER for Mee Cocks (cf. Tulane Graduates Mag. 3: 122. 1914). 16 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 I was looking at Louisiana Crataegus in the herbarium the other day and it is astonishing what a small amount of material we have from that state and how little I know about it. The Plums and Apples, too, are weighing rather heavily on me. Have you got any more plants to send us from this year’s collecting? I hope you will have a very pleasant Christmas and a happy and pros- perous New Year. 25 February 3, 1912. It is a long time since I have heard from you but I hope this does not mean that you are not well. I hope to be in Louisiana coming from Texas early in April and I want to meet you somewhere and make another trip with you. I hope you will be able to manage this. The Louisiana Prunus situation is in bad shape as only fragmentary material has so far been collected. I think it is desirable that you make a Plum journey early and before I get to Louisiana, otherwise we shall miss them again. For example, numbers 1 and 2 of the Butler plantation were well out of bloom on March 28, 1910; from Laurel Hill to Covington we have leaves only. +1 from Clear Lake was out of bloom March 30th. Shreveport plants were all out of bloom April Ist, and specimens collected at Opelousas were out of bloom March 20th. Bush’s collections at Natch- itoches April Ist, 1909, were long out of bloom. I should think it would be desirable to look over the state for Plums as early as March 10th, or perhaps even earlier in the south. The trees should be carefully marked to insure mature leaves and fruit later in the season. I want very much to go with you to that place north of Alexandria to look for the Malus which you have n’t been able to find. I am very sure it is an undescribed species. As for Louisiana Crataegus the situation, I hope, is not hopeless but I think we must confess that we are making very little headway with it. Please let me hear from you. 26 March 28, 1912. I am still pretty shaky but the doctor tells me that I can probably get off in a week or ten days for Louisiana. I do not know about the trains yet, but would n’t it be feasible for us to meet say at Munroe and drive from there to Alexandria? How long would this take and would it be a feasible trip? It would at least show me a new country. Or I could per- haps get to Alexandria quicker than I could to New Orleans, look up the Apple across the river, then go on to Shreveport and Natchitoches, and return to New Orleans. Let me have an answer to this if you can before I start. If not, I will telegraph you when I find out if I can really go and tell you when and where I will be. 1965 | EWAN, LETTERS FROM SARGENT TO COCKS 17 27 October 16, 1912. Returning from a short journey in Missouri I find your letter of Sep- tember 26th and the package of specimens which you have been good enough to send us. I cannot do much with the Crataegus unless you have the notes on the color of the anthers. This is essential to a correct determination of these troublesome things. Have you also fruit? I can say, however, that No. 1 N.O. is C. viridis and xi. and xviii. of Natchitoches are C. brachyacantha. C. brachyacantha was the species which we found so abundant a year ago last spring near Covington. The Malus and Vacciniums I will write about as soon as the specimens are mounted. 28 February 18, 1913. We have a specimen of a Maple collected by you at Natchitoches April 27, 1911, with fully grown fruit. It is a very distinct form of the Sugar Maple and perhaps an undescribed species. There is nothing like it in our herbarium. I am writing to you now to suggest that as this tree had fully grown fruit on the 27th of April it must flower very early, probably some time before we get to Natchitoches. It is desirable if possible to get flowers this spring as we may want to publish a plate of it if it turns out to be new. Do you remember the tree? Is it common, etc.? The nearest Sugar Maple to this station which we know is at Fulton, Arkansas, and there is another at Boerne, Texas, which must be four or five hundred miles away. What is the prospect of the season, is it going to be early or late? I will let you know a few days before I start for New Orleans. 29 February 25, 1913. I am much obliged for the specimen of Maple which arrived today. I think it is safe to refer it to Acer leucoderme Small. Please write me again about the season and how vegetation is coming on. 30 April 12, 1913. We arrived in New Orleans safely but Tracey failed to turn up there. From New Orleans we went to Natchez, staying there about thirty-six hours and then came home from Chattanooga, arriving here last night. I have been quite anxious about you for I am afraid the rain which started in after we left Natchez has made a great deal of trouble in the Red River Valley and that you are prevented from doing some of the work you were going to do. I have n’t had time to do much with plants since I got home but the Opelousas Malus seems to be very near Malus ioensis var. Palmeri, and 18 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 the species from Win[n]field to be a form of Malus angustifolia with re- markably acuminate leaves. It can hardly, however, be specifically distinct. At Natchez we found Hydrangea quercifolia very common. Has this ever been found in Louisiana where it ought to grow I should think, but we have no specimens from anywhere west of Alabama? I had supposed that it was an Appalachian species but finding it at Natchez makes me think it may grow south of Natchez and possibly across the river, I also found at Natchez what I believe is my Aesculus splendens although I have n’t seen the fruit yet and I think that we found the same species at Winnfield. If so this is rather a fine plant for the Louisiana flora. That little Viburnum acerifolium, if it is that species, is certainly interesting for this appears to be an Appalachian species, at least we have no specimens taken west of Alabama and none from west of the Mississippi River. These ought to be looked up. Although Quercus breviloba is credited in my Manual to Louisiana I do not find that we have any specimens of it from that state. I should suppose that if it occurs there at all it would be about Shreveport as it is common at Fulton. I have n’t taken time yet to find out the things which you have n’t sent us, but I will write you again next week. Please let me hear what hap- pened to you after we parted. I have n’t seen anything so fine anywhere on this journey as the woods at Opelousas. 31 April 15, 1913. A few of the Louisiana plants which we do not have in the herbarium are Mimosa strigulosa, Parkinsonia aculeata, Desmodium if there are any woody species, and Erythrina herbacea. Have you ever seen this last assume the habit of a shrub or small tree, in southern Florida perhaps a tree of considerable size? Of Quercus laurifolia we have only the specimen with young leaves col- lected by us at Lake Charles two years ago. I should be glad to have fruit- ing specimens from Lake Charles and from other parts of the state. I will send you a list of other Louisiana desiderata as soon as I have time to look them up. You are O.K. with Prunus Munsoniana. 32 April 15, 1913. As the copy of your paper, for which many thanks, appeared this morn- ing I am afraid you had to shorten your trip and were unable to get to several of the places you hoped to visit. Too bad! It looks as if we were never going to get at the bottom of the Louisiana Crataegus. Mr. Rehder, who is working up Malus, has made a new variety of M. ioensis which he calls creniserrata. The type for this is the Pineville plant. He refers to it also the Crowley and Opelousas plants. In regard to Crowley I find that you are right and that we have fruit from there and 1965 | EWAN, LETTERS FROM SARGENT TO COCKS 19 some fragmentary leaves. I shall be very glad to get leaves collected now with the young fruit, especially leaves of vigorous shoots as it is always important to collect these. Your Malus angustifolia from Covington, etc., is behaving very badly and does not seem to belong to the typical form of that species as it turns out to have hairy pedicels, etc. I wish you would collect some more material of this plant from that region with good vege- tative leaves, etc. If Malus angustifolia, which we have always believed to be perfectly glabrous, is going to behave in this way it will make addi- tional trouble for us. 33 April 19, 1913. As you know, Louisiana was not credited in my Manual for Avicennia nitida. We have a specimen from you gathered at Grand Isle. Can you give me the range of the species in Louisiana, that is the range east and west on the coast? Is it always shrubby in Louisiana or sometimes a tree? I do not know where Grand Isle is, but perhaps you will tell me. 34 April 21, 1913. As I wrote you, we consider the Pineville plant the type of Malus ioen- sis variety creniserrata, nov. var. Rehder. The Natchitoches plant which we have from you in flower only is referred to this variety although it looks a little different. The Opelousas and Crowley plants we now refer to Malus ioensis var. Palmeri Rehder. The type of this variety is in south- western Missouri, but this form is common in southeastern Missouri and in Arkansas. You understand that what we have always called Malus angustifolia now has to be called Malus coronaria. Rehder makes a new variety of this, variety puberula, the type of which is the Winnfield plant. I found exactly the same thing at Natchez and we are referring to this variety all your specimens from eastern Louisiana. So far as our specimens show we have no typical M. coronaria or angustifolia from Louisiana, all your speci- mens having more or less pointed leaves and hairy pedicels, while in the real coronaria the leaves are usually rounded at the apex and the entire plant is glabrous. As I wrote you the other day, I hope you will get more material from eastern Louisiana. Now in regard to the Crataegus which we collected: Opelousas, #1, a Viridis species, by roadside, distinguished from C. viridis by the bark. Fruit needed. C #2 agrees well with C. tersa of Beadle. C +3, the Crus-galli species with small corymbs, does not fit in with any described species. Of course we want fruit of this. C +4 we were badly fooled by; it is not a Crus-galli at all but C. brachyacantha. C #5, a Viridis species. No flowers. Winnfield +1. I don’t know this. 20 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 +2. This was the small slender tree near the quarry, certainly a distinct new species. Fruit certainly should be obtained. Lake Charles. The thornless Crus-galli in yard on roadside to English Bayou. I don’t know this. These are all we collected this year. I do not think there ought to be much trouble with the Louisiana species if we can only get material for, with the exception of C. brachyacantha, apiifolia and spathulata, they all belong to the Viridis and Crus-galli, with the exception also of that plant we found in western Louisiana which belongs to some other group. It looks something like C. silvicola of Beadle. I collected a number of years ago specimens of a good many Crus-galli forms from Mendum | Minden] which I think is a field which should be investigated. I will write you again about Crataegus in the course of a few days. I am very anxious to hear from you and what happened to you after we separated. 35 April 30, 1913. Too bad that the weather is so much against us; still if we keep on I think we shall get to the bottom of our Louisiana problems. I understand that Robin’s Flora is a supplement to his general Travels in three volumes, and if the librarian of the Howard Library can get us a copy of the whole work I shall be greatly indebted to him and to you. We have a wonderful season here this year and have never had so many flowers before. Better luck next time you start out! 36 May 20, 1913. I have this morning the package of your plants, for which many thanks. Carya Nos. 1 and 3 from Bernice is C. alba. No. 2 from Bernice I don’t know. The Carya from West Feliciana, Nos. 1, 2 and 3, I don’t know. No. 4, C. alba. No. 5, C. cordiformis. When I say I don’t know I mean that I don’t feel sure about this with- out more material. Of course we ought to have fruiting specimens in the autumn. Acer appears to be A. floridanum, and all the specimens of Malus can be referred, we think, to M. coronaria var. puberula. The Selma Ilex is I. longipes. Have you ever found the Hydrangea west of the Mississippi River? There is still, I think, two Maples and another Malus to come from Natch- itoch. I never heard from that banker whom I asked to send the Malus speci- men. Can’t you stir him up on the subject? Mr. Harbison, who collects for the Arboretum in the southeast, has been to Selma this spring to look for the Molles species of Crataegus you told 1965 | EWAN, LETTERS FROM SARGENT TO COCKS 21 me about and for the //ex. I told him that you had a plantation near Selma and he apparently lost a good deal of time in trying to locate it. He even went to the court-house to look for a Cocks owning property in that part of the world. I suppose the trouble was that the plantation does not stand under your name but under that of some member of your wife’s family. Harbison will be back there sometime this year and I wish you would write me how he can find your plantation and also the exact local- ities for the Molles species of Crataegus and for the /lex. I hope the world is treating you kindly and that you are coming north this summer. There is a lot of work which ought to be done in Louisiana in the autumn. 37 June 18, 1913. I am glad to get your letter. The Holly from Selma is /lex longipes of Chapman. The last Maple you sent from Natchitoches appears to be Acer leuco- derme, but we also have had from the same region from you in previous years what must be Acer floridanum; in fact these Natchitoches Maples do not behave very well and show a tendency to too much variation. Acer nigrum from there we have never seen yet but I hope you will be able to get it this autumn. I am going to write you in a few days about the Louisiana things which especially need looking up this autumn. I am glad in the meantime that you are in such a good field as you report Selma to be. By all means send the Plums and collect everything woody you can lay your hands on, and please pay especial attention to the Hickories. Harbison has been to Selma two or three times in the interest of the Arboretum and will be back there probably this autumn but I am afraid not until after you have left. I gather from your letter that you don’t expect to come north this sum- mer. I wish you would let me know about this and, if you are coming, when you will be here and how long you expect to stay. I hope in any case nothing is going to interfere with work in Louisiana this autumn. 38 June 25, 1913. I have your note of June 21st and the box of Plums. I do not know this plant if it is not a late-ripening variety of the Chickasaw. I suppose you are making specimens. I have made out the following memorandum of plants in Louisiana which need looking after this autumn: Opelousas. Crataegus #1, a Viridis species on the bank by roadside east of Opel- ousas. Crataegus +3, low rich woods fifteen miles west of Opelousas, a tomen- tose Crus-galli species. Crataegus +£5, a Viridis species in low wet woods, fifteen miles west of Opelousas. 22 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 The Quercus, like the Post Oak, but growing in low wet woods. Mature leaves and fruit of this should be collected. Quercus ludoviciana, new hybrid, Sargent. Our hybrid tree fifteen miles west of Opelousas. I have described the tree with the above name in the new part of Trees and Shrubs. Fruit and mature leaves needed. The red-flowered Aesculus in woods east of Opelousas, perhaps Aesculus Pavia but the leaflets are too pubescent below. us. Mature leaf specimens of the species in the woods fifteen miles west of Opelousas, probably M. ioensis var. Palmeri. Hickory #1, large tree with smooth bark in woods east of Opelousas. The trees marked 3 and 4 are probably this form. Hickory #2, large tree in woods east of Opelousas, with rather rough ark. Winnfield. Crataegus £1, the Crus-galli species by roadside close to town. Crataegus 3¢2, the very tall slender tree near quarry, no doubt a new Crus-galli species. Malus coronaria, var. puberula. The type of this variety is from Winn- field. Fruit and leafy shoots are needed The red-flowered Aesculus which I believe is my Aesculus splendens. It is important to get the fruit of this for if it proves to be Aesculus splendens this is a new species for Louisiana. cores Two species in woods near qua A large Hickory in yard in Winnfield, This is probably an undescribed species. Lake Charles. Crataegus #1. On road to English — Irish? — bayou. The tree is in a yard, about fifteen feet high, with stem a foot in diameter and wide- spreading branches. Hickory. The tree with pale close bark in the low woods along the river, common. The young leaves are bright red. This is probably an unde- scribed species and perhaps the same as one of the Hickories in the woods east of Opelousas. We should have autumn leaves and fruit of the following Hickories: Bernice II, West Feliciana; Nos, 1, 2 and 3, Bayou Lacombe, 1785, May 1908; and #1766, May 1909. We need good material of Malus from Crowley and of the two forms at Natchitoches, with fruit and leaf shoots. Also from Natchitoches more specimens of the three forms of Acer, A. floridanum, A. leucoderme and A. nigrum; the last I have never seen. I have a great deal of incomplete Crataegus material from Shreveport, Munroe, Richmond, Natchitoches and Crowley, but it seems to me that it would be better to concentrate this autumn on the things which we saw this spring and which I have enumerated above, and try and get these finished up. If we don’t begin pretty soon and finish some of these doubtful Louis- sicecaeatiate amen tec 1965 | EWAN, LETTERS FROM SARGENT TO COCKS 23 iana trees we shall never get anywhere I am afraid. There are apparently more species of Crataegus at Natchitoches than anywhere else in Louisiana and of course it is desirable to collect fruiting specimens and number the trees this autumn if they are in good condition with the view of completing them in the spring, but as I have already said it is most important to finish up the things which we got this spring. Wistaria frutescens and macrostachya appear to run together and we are inclined to consider the latter a variety of the former. I certainly col- lected the two plants at Covington. We have no fruiting specimens of Wistaria from Louisiana and if you come across any this autumn I wish you would make specimens from the different localities. 39 July 9, 1913. Many thanks for the second lot of Plums which have arrived safely. I have n’t seen Harper’s Alabama book yet but I believe I have ordered it. It is not perhaps surprising that it is n’t complete as his botanical ob- servations are chiefly made from the windows of moving trains. Have you seen Small’s new edition of The Flora of the Southeastern States? I cannot see that he says much about Louisiana plants in it, and he has certainly missed all the interesting new things in the way of trees and shrubs that have been found in the south in late years. It does n’t seem to me that this book, in spite of its great size, adds much to our knowledge of the subject. I do not believe it is any hotter in Alabama than it has been here during the past week and I hope it is n’t as dry there as it is here. 40 July 18, 1913. I have yours of the 18th and the lots of Plum fruit, for which many thanks. Sardis must be a great Plum country, and I hope you are drying leaves and making notes of the different forms. I am off tonight for Puget Sound, but shall be back at the end of three or four weeks. Anything you send in my absence will be taken care of. I return Hedrick’s letter. His statement confirms my belief that it is impossible to tell much about Plum trees from herbarium specimens alone. I cannot believe that Prunus Watsoni grows in Alabama. 41 August 15, 1913. I am just back from the Pacific coast and find here your letter of July 17th and the leaf specimens of the Sardis Plums. We feel quite sure that your No. 7 is not P[runus] pennsylvanica. I do not know what it is but it looks like a new species of Cherry. You better send some notes about the size, habitat, etc., of this tree. No. 8 is not P. mitis and also looks as if it might be a new Cherry. No. 5 is, as you say, 24 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 P. umbellata, and I think you are right in referring No. 6 to P. munson- tana. | am afraid I cannot throw any light on the other specimens The Selma region seems to be a terror for Plums and I hope we shall get to the end of them sometime. I hope you are all right for the autumn campaign in Louisiana for it is important to clear up some of these Louisiana mysteries. By all means collect Crabapples from the Selma region. There is probably one new species there with leaves very pubescent on the lower surface. I’ve just written Harbison to suggest that he go at once to Selma & look over the Plum field with you. 42 August 27, 1913. W. H. Lamb of the Forest Service of the United States pronounces an Ash collected by Dr. J. S. Schenck May 1901 in a Palm Swamp, New Or- leans, and previously referred to F[raxinus| profunda, to Britton’s F. Michauxti. Do you know anything about this? Specimens of F. Mich- auxit which I have seen look distinct. Leaflets on the under surface are slightly pubescent and the wing of the seeds is short and broad. This may be a good species. 43 September 24, 1913. I find that from southern Missouri I can best reach Louisiana at Mun- roe and I suggest that we meet there instead of at Shreveport, then from Munroe we can go in either direction. It looks as if I could be there on the 8th or 9th of October, but I will telegraph you two or three days in advance of my arrival. I think, however, you had best be there on the evening of the 8th. 44 September 30, 1913. In looking over the plant you recently sent we find that the J/ex, a small shrub from the banks of the Alabama River, is /lex monticola. There is a second specimen of this species from Sardis. The Bumelia is B. lycioides. Vitis No. 1 is V. cordifolia. Vitis No. 2 is V. aestivalis. I am leaving on Thursday and shall be able to telegraph you probably from St. Louis two or three days in advance of my reaching Munroe. 45 October 20, 1913. We had a couple of good days with Harbison on Mobile Bay and finally got home on Friday. Hickories are very abundant on the shores of the Bay and there is but one species there, a Pignut, which shows remarkable variation. 1965 | EWAN, LETTERS FROM SARGENT TO COCKS 25 The Winnfield Sambucus is S. canadensis variety submollis Rehder, which appears to be the common form west of the Mississippi from Mis- souri to Texas. The Sambucus which we collected at Opelousas and which appears to be common in the neighborhood of New Orleans and along Mississippi Sound is Sambucus Simpsonii Rehder, a species first distin- guished in western Florida south of Tampa Bay. It is sometimes a small tree and appears to be very common in all of southern Louisiana, Missis- sippi and Alabama. The Florida plant, as I remember it, did not have the bright red petioles, and these were not noticeable in plants which I saw about Mobile. The very shining leaves are very unlike those of any form of Sambucus canadensis and this I think is a good addition to your Louis- iana flora. I wish you would dry me a specimen of what they cultivate in New Or- leans as Rose of Sharon as I do not seem to be able to place this plant from memory. I hope, too, you will send specimens of the different forms of Ligustrum which they cultivate i in New Orleans with deciduous leaves, and the Araucarias. Indeed we should be glad of herbarium specimens of any of the uncommon trees and shrubs cultivated in New Orleans as there ought to be representatives in this herbarium of every woody plant cul- tivated in the United States I will write again soon about some of your other plants. I hope you have got over the hardships of Opelousas 46 October 25, 1913. The //ex with solitary fruit on short pedicels which we collected at Winnfield is //ex monticola. This is the first time that we have had any indication that this species grows west of the Mississippi River, and Small gives its western locality as Alabama. Perhaps this therefore is a new addition to the Louisiana flora. I think we shall have to call the Viburnum which we supposed might be V. prunifolium V. rufidulum as the buds are covered with the rufous to- mentum of this species which varies on the petioles. I find that there is n’t a specimen of Persea borbonia from Louisiana in our herbarium. I suppose it is a common enough tree in the state and I wish you would let us have a specimen. I should be much obliged, too, if you will sometime take a look at Acer Drummondii as it grows in the swamps and give me some idea of the size this tree attains under favorable conditions, both in height and trunk diameter, and also some notes on the character of the bark. I don’t remember that I have ever been close to the tree and if it is to be considered a species I must have more facts about it than I have at present. 47 November 10, 1913. We have sent the plants named on the enclosed list to Mr. Marshall, Superintendent of the Audubon Park, to whom I also send a list. 26 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 Cupressus funebris is an important Chinese tree which ought to do well in New Orleans, and I hope a good place will be selected for it. I don’t know as there is much to say about the other plants. I will write a line also, as you suggested, to Mr. Morgan, President of the Audubon Park Commission. 48 November 18, 1913. I have your letter of November 13th as well as the Louisiana specimens of Hickory. Of Nos. 2, 3 and 7 from West Feliciana there are no nuts, No. 2 of course being C. cordiformis. Of No. 3 of Grand Ecore there are no nuts. No. 6 from Natchitoches looks like C. pallida, and I think No. 2 from Natchitoches is the same although there are no nuts of this. No. 7 from Natchitoches is C. ovata approaching variety Nuttallii, No. 8 Natchi- toches looks like C. ovata but there are no nuts. No. 9 Natchitoches is C. myristicaeformis but there are no nuts and we should be glad to have them. You speak of having found C. —— at Natchitoches but I don’t find either nuts or leaves in the packa I feel helplessly confused about oe southern Pignut Group. The more material I see the worse the confusion becomes. You certainly did not send the specimen of the Oak which you took for Q. breviloba. I think you had best send me a specimen of your arborescent Alder from Selma that I may feel sure of the species. When are you going to send me the lists of trees from your Alabama place? Such a list certainly should be published. P.S. — Please send us some of your printed labels for these last speci- mens. 49 November 26, 1913. I have your note of November 21st and the nuts, for which many thanks. We had to use some of your labels with the nuts and I wish you would send me about fifty of your blank labels for future use. I will write you in a day or two about the Selma Apples and Hickories My correspondent in Natchez, who is not a botanist,* has found there what I feel quite sure is Carya texana which is called Pignut. My cor- respondent writes, — “One who travels in the adjacent Louisiana country tells me that the Bitter Pecan or Pignut is thick in the swamps and that representatives of German dealers have taken the output of the sawmills for use in the making of carriages. The wood is said to be very tenacious.” This may be Carya texana and it would be interesting if you could get into touch with some one in that part of the state and find out about it. She also writes, —‘‘I hear of a nut somewhat larger, thin-shelled and very sweet in the West Feliciana Parish.” * Miss C. C. Compton. 1965 | EWAN, LETTERS FROM SARGENT TO COCKS 27 You once told me that you had heard of a flat Pecan somewhere near Baton Rouge and it is possible that this may also be texana which is evi- dently much more widely distributed than we had supposed. If it grows in Natchez there is no reason why it should not be in West Feliciana. 50 November 28, 1913. In regard to your Alabama Malus Nos. 1 and 2 appear to be the same and look very much like M. ioensis, var. Palmeri, Before we can be ab- solutely sure of this, however, we should see the flowers. No. 3 is most like M. fragrans, 1 think must be a new species; certainly judging by the leaves it is a very distinct form. 51 December 3, 1913. I am glad to get your letter of November 27th and to hear that the plants arrived safely. Although I wrote both to the Superintendent of the Park and to the Chairman of the Park Commission I have never heard a word from either of them. They seem to have curious ways in your town. I understand from your letter that the express bill has been settled by the Park people and I will interview our expressman on the subject. I do not think the plants which you said you sent on “Monday” have yet turned up. ae December 6, 1913. I have your package of plants. The Hibiscus from New Orleans is H. mutabilis from southern China. The Oak from Natchitoches, called by you Ouercus brevifolia, is Quercus stellata Margaretta, The Ilex Amelanchier is all right. This is the first time we have ever had any wild specimen of this plant, having known it only from a plant in cultivation. This we have lost and we very much want to get it again either by seeds or roots. Pos- sibly this can be managed next year. We shall be glad, of course, to get flowering specimens. Can you make Harbison understand just where it grows? i The Viburnum from Natchitoches is V. scabrellum. Your Vacciniums Nos. 1, 2, 3 and 5 are V. Elliottii. No. 7 is V. stamineum or one of the allied species; and Nos. 4 and 6 are virgatum which has not before been in our herbarium from Louisiana. Rosa humilis is all right. The Rubus from Winnfield is, so far as we know, an undescribed species. We collected the same thing at Washington. In regard to your three Acers from Natchitoches No. 1 is floridanum, No. 2 leucoderme var., certainly a very curious looking form, and No. 3 leucoderme like the typical plant. I think this will account for all your specimens. I hope you will soon send us another supply. 28 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 53 i December 20, 1913. Your last package of plants has arrived. In regard to Rubus, Nos. 2, 4 and 7 are R. trivialis; the others all seem to be the common Highbush Blackberry which we have collected together. It is nearest R. Andrews- ianus, at least it should be referred to that species until the southern Black- berries are more critically studied than they have been. I find that the older name for Zizyphus vulgaris is Z. sativa Gaertner. 54 February 24, 1914. Your Cornuses have arrived and seem to be correctly named, except what you call C. sericea is C. obliqua or Purpusi, the western and appar- ently the southern form of C. amomum, the correct name for C. sericea. We have had no specimen of this form so far south and I am glad to have it. In regard to the Sassafras business, Fernald in the January 1913 number of Rhodora has taken up a White Sassafras, first distinguished by Nuttall under the name of S. variifolium var. albidum. This variety is distin- guished by its glabrous branchlets and leaves which are glabrous when young. According to Fernald, it grows from western Massachusetts to the Carolina mountains. According to Nuttall, the inhabitants of the Caro- linas distinguish the two forms by the names of the White and Red Sassa- fras, the root of the White Sassafras being much more strongly camphor- ated than the ordinary sort and nearly white. No one says anything about the wood of this variety but if the roots are white the wood might be. I think it would be well to look at Sassafras in Louisiana and see if you can find it with glabrous branchlets and leaves. By the way, we have no speci- men of Sassafras from Louisiana. I am sorry about the flowers of Acer Drummondii. Have you no speci- mens that you can spare us? Does not Acer Negundo grow in eastern Louisiana? I have it from the state only from Shreveport and Opelousas? I should like to see the eastern Louisiana specimens. There are no speci- mens either from Mississippi or Alabama here. po) February 26, 1914. I have been looking over the Louisiana Hickories and I do not get very much comfort from them but send you the following notes: Opelousas, East. #5: this is the tree with close smooth bark but with the buds, fruit and foliage of the Shagbark, C. ovata. This is the most interesting perhaps of the Louisiana Hickories, for if it is a species or even a variety it changes all our ideas about the Shagbark Group of the genus. If we should consider it a hybrid what would be its parentage? We saw but one tree and it is most desirable to locate others in that Group and try 1965 | EWAN, LETTERS FROM SARGENT TO COCKS 29 to get flowering specimens and young leaves this spring. 1 and 3; on the bank of the bayou. We thought these were the same, but notice that the rachis and midribs of the young leaves of #1 are very pubescent, and that of #3 are perfectly glabrous. #4 appears to be the same as #3. +3 and #4 have very large buds and I cannot place them, or #1 either, although #1 may possibly be a tree which is common on Mobile Bay. Opelousas, West. #1 and 2, the two trees standing near each other in an open wood not far from the swamp which you crossed in the spring to find the large tree of Crataegus brachyacantha. 31 has pear-shaped and +2 subglobose fruit; they are probably the same and I believe an unde- scribed species or a form of the Winnfield Yard tree, distinguished by hairy branchlets and the yellow scales on the fruit. I have seen what seems to be the same thing from Florida and Georgia. It is very important to collect these two numbers in their spring condition that we may see if the young leaves are covered with yellow scales like those of the Winnfield yard tree. +4, “rich woods west of Opelousas; close pale bark; young leaves fragrant and red like the Lake Charles tree.” This we did not relocate in the autumn. Winnfield. The yard tree; this so far as I can tell now is an unde- scribed species. We have sufficient material. Like so many other species it has both globose and pear-shaped fruit, for I take for granted that the trees in the yard are of one species. #2, ‘‘tall tree in woods”. This may be the same as the yard tree but I am not sure. The leaves look all right but I can find little trace on the buds of the yellow scales which are so con- spicuous on those of the yard tree. The fruit of #2 is depressed-globose. I should like to see the young leaves of this tree. I do not remember where it is. Was it at the last place we went to in the afternoon, in open ground on a slope north (?) of the town where we found also a large Plum tree? On April 6th at Winnfield, ‘‘in dry woods”, we collected a specimen with young leaves white tomentose on the lower surface and only sparingly pubescent above. The branchlets are red, slender and glabrous. This I suppose is one of the trees in the low ground before we got to the quarry. In the autumn we collected in this region C[arya] alba but the tree we collected from in the spring is not C. alba. What is it? We missed it in the autumn. Munroe. In dry woods near Munroe we collected specimens of C. alba and of a tree of the Pignut Group with very slender pubescent branchlets and small buds; leaves with broad leaflets and pubescent petioles and rachis. Old nuts from the ground of what was probably this tree are very small. This appears to be a common tree in dry woods near Munroe. I do not know it. Bernice #2. Only young leaves and flower-buds. Can this be the same as the yard tree at Winnfield? The region for it would, of course, be all right. It ought to be looked up. West Feliciana. #1 looks like C[arya| porcina. The buds are much larger than those of the northern tree. #2, no nuts. Can this be the Winnfield yard tree? #3, C. ovata, no nuts. #4; the nuts look like those 30 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 of C. megacarpa. I find no foliage of this. #6, C. porcina? #7 and 8: I do not know if these may not be the same as the Munroe Pignut. They are pubescent like that tree, but the nuts are much larger than the old ones we picked up from the ground. Bayou Lacombe. I can say nothing about two very fragmentary speci- mens, Nos. 1766 and 1785. Lake Charles. I think that we have to begin all over again. You turned to the right at the river after passing Irish Bayou and I to the left. Along the river I found trees with bright red, fragrant leaves on April 2d, and pale close bark. Is it possible to say if the tree you collected from is the same? This Lake Charles tree must be undescribed. I believe the same thing is west of Opelousas. Lucknow.® Of the tree with small, rather thick-shelled pear-shaped fruit er : have sent to you I have not seen foliage and know nothing bou ane. +1 and 2. I have no idea what these are. #5 seems to be a pear-shaped C. alba with rather small buds. #6 looks like C. pallida, but it is the only station west of Selma I know for it and it should occur at intermediate stations. Does this tree have pale bark and does it grow in low ground or on high ridges? I should like to see young leaves. +7, C. ovalis with small fruit. #10: I think this is a new species which grows also at Fulton, Arkansas. Young leaves and flowers are important as well as notes on size, habitat, etc. ¢xi, a pear-shaped ovata! What is the bark like. 43, Grandcone.® The fruit looks like that of one of the forms of C. ovalis but the buds are covered with yellow scales. Unless it is an- other C. pallida it is too much for me. Natchitoches seems to be the storm centre for everything that ought and ought not to grow in Louisiana. I do not want to describe any new species or varieties of Carya until next winter, so that we may have the benefit of another season’s field work. It is important that the trees of Lake Charles, east and west of Opelousas, Lucknow, Bernice and West Feliciana, should be visited this spring. Winnfield is less important and there is still much to be done there. If you are able to get good spring material it may be desirable for me to go to Louisiana again in the autumn and take another look at some of these trees, for I feel that we are not going to get any right conception of them except by constant field observation. West of Opelousas there is the new apple (#2) of which flowers are needed, and it would be well to get apples from Crowley as your correspondent seems to have given out. I expect to go to southern Florida this spring, leaving here on the 16th of March, and to stop in Georgia and South Carolina probably on my way north. You may think that Louisiana Hickories are bad, but I am inclined to think that those in south-western Missouri and in the southeastern states are even more troublesome. Let me hear what the prospects are of your being able to do a good deal of field work this spring. Harbison is to go to Selma for your Plum flowers *>Sw. of Rayville, Richland Parish. ®Error in original letter for Grand Ecore? 1965 | EWAN, LETTERS FROM SARGENT TO COCKS 31 as soon as he hears from a friend there that the season is opened. I hope you do not feel that we know anything yet about the Louisiana Plums. 56 April 16, 1914. I am back from the south where, although the spring was very backward and the weather rather cold, I saw much to interest me and was able to extend southward the range of a number of trees. Just before I left I had a note from you saying you were sending me flowers of Acer Drummondii, but the package has not yet arrived. I suppose it will turn up in time. I am inclined to think you are right in supposing there are two quite distinct Lindens in Louisiana, one quite glabrous and the other covered on the lower surface of the leaves with stellate pubescence. The former is well represented at Lake Charles; the latter we found at Winnfield and Opelousas, and you found it at Natchitoches. This appears to be the com- mon species of eastern and southern Texas, except on the coast where Tilia pubescens occurs. It is desirable to get flowers of the pubescent form from Natchitoches or some other convenient station. I am more than ever convinced how little we really know about our southern trees and how much there is still to do. I think I found on the Caloos[a}hatchee River in Florida what must be a new Water Hickory or, if not a species, a very distinct variety. I see Small credits Louisiana with Cyrilla parvifolia but we have no specimens from the state in our herbarium. I am anxious to hear what success you have had with the Hickories this spring. 57 May 20, 1914. I have taken a look at your Hickory specimens. No. 6 at Natchitoches, if the flowers and fruit are from the same tree, is something I do not know at all. It has slender branchlets and small buds of a Pignut but the young leaves are as pubescent as those of C[arya| alba. The nut is very much compressed and the husk splits readily. If there are many trees like this, I do not see why it is not a new species. No. 10, Natchitoches. Is there not a mixture here? The flowering speci- men is C, alba but the fruiting specimen is one of the Pignut group. There appear to be no flowers of xi. Natchitoches and 3 Grand Ecore, and no fruiting specimens of 12, 15 and 16 Natchitoches and 20 Grand Ecore. Note that 12 Natchitoches and 20 Grand Ecore have the yellow scales on the young leaves of the yard tree at Winnfield and of the trees with the round and pear-shaped nuts of West Opelousas. I should suppose there was something in the scales of the young leaves of specific value i help us. reer B, 9, 13, 14 and xi. seem to be new numbers of which there are no fruiting specimens. 32 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 West Feliciana 10 and 13 are, I suppose, new numbers but we have no fruit to match them. I believe that the round and pear-shaped fruits of 1 and 2, West Opelousas, belong to one species and that there is no charac- ter in the shape of the fruit in Hickories. I hope you are well and not too hot. 58 June 17, 1914. I have read with much interest your Notes on the Flora of Louisiana, No. 1,‘ but why not follow the rules of the Vienna Congress, as you do in some cases, and call Quercus acuminata Q. Muehlenbergii and Magnolia foetida M. grandiflora? Also why not write “canadensis” in Cercis cana- densis with a lower case c when you write caroliniensis with one? I do not seem to remember Pinus glabra from West Feliciana. Did I have a specimen from there? Of course I know it is in Covington and in Jackson, Mississippi. I am looking a little into Tilias and I find two specimens of yours which I think with our present lights will have to be called T. americana. One was collected at Alexandria in June 1905 and the other at Wakefield, West Feliciana, June 1907. Unfortunately neither specimen is numbered. Do you suppose it would be possible to locate these trees? You once insisted that the St. Charles Tilia was different from the one at Natchitoches. There are certainly two species at Lake Charles, one like the Natchitoches and Shreveport tree with very thin leaves with scurfy pubescence on the lower surface which soon rubs off or on some individuals is persistent, and the other which is entirely glabrous. The former is T. leptophylla; the latter seems undescribed. You collected this last in flower May 12, 1911, but not in fruit. We have, however, a poor fruit specimen collected by Andrew Allison in the vicinity of Lake Charles (No. 322, no date) which came to us from the National Museum at Washington. Do you remember the tree from which you collected the flowers and can you locate it? We collected on March 26, 1911, Tilia leptophylla at Lake Charles and no doubt you were looking for our tree when you made a specimen on May 12th of that year. This May 12th tree should certainly be looked up and investigated further. Tilia leptophylla appears to be the species of Opel- ousas and Winnfield, and I presume is the common Louisiana Linden. There is a Tilia collected at Selma by Harbison with large leaves oblique at the base and covered below with silvery white tomentum, which occurs also on the petioles and peduncles, and the young branchlets are stellate- pubescent. I do not think that this can be T. heterophylla. The specimens were collected May 5th in young bud. This tree should be investigated and collected. Is it too late for the flowers now? Fruit, of course, is also needed. Judging by Harbison’s material, this tree can hardly be referred to any of the described species. It is evident that there are several more Tilias in the southern states than were formerly recognized. I am working on Hickories and am going to write you about them very soon. * Published in Plant World 17: 186-191. 1914. 1965 | EWAN, LETTERS FROM SARGENT TO COCKS 33 If your wife has been in Washington I presume she has entirely recov- ered. I certainly hope so. I hope you are going to have a good summer in Alabama. I am going over to England for a few weeks on the 9th of July but shall be back in time for the south in the autumn. 59 une I have yours of the 20th and the specimens. The ee eh Poe looks like Tilia americana. I suppose for the present, at least, it will have to be so-called. Your specimen from Avery Island is not like either of the Lake Charles species, and if it is common west of the Mississippi River I have not seen it there before. I have seen specimens of what I believe to be the same thing from River Junction, Florida. I hope you will gather and press a number of specimens with flowers or young fruit of Tilia which you say is common on your place, for we have by no means got to the bottom of this Linden business yet. Your Amelanchier looks most like A. sanguinea which is a northern spe- cies. You better send me specimens of your Azaleas or anything else interesting. Can you not arrange to stop here on your way south from Maine? I expect to be back on the 25th of August and it will be a great pleasure to see you and your wife here, and it would give us an opportunity to talk over many things about the autumn campaign. I suppose you could stop in Boston as well on the way south as on the way north. I hope I shall hear from you again before I leave on the 8th of July. 60 : June 29, 1914, Many thanks for yours of the 24th with the Lindens. No. 2 certainly looks like | Tilia] leptophylla, and No. 3 like Harbison’s plant. This seems to me the same as your Avery Island species. If I am right in this it makes its distribution from western Louisiana to western Florida. The general outline of the leaf looks like that of heterophylla, but the tomentum on the lower surface is less dense, and this new tree has more or less pubescent branchlets, peduncles, pedicels, etc. The range, too, is quite different and I do not see why it is not a good species. I hope you will keep watch of it. I want to describe as new a Hickory from Starkville, Mississippi, a large tree with dark rough bark and ovate fruit with a husk of medium thickness, and a slightly angled nut. The Starkville tree is peculiar in the ovalis group, to which it belongs, in the tomentum on the lower surface of the young leaflets, and in the dark rough, not scaly, bark. Your No. 10 of Natchitoches, although I have no young foliage, can probably be referred to this species. Do you remember anything about it? I do not remember if I told you that I think we must call the Winnfield yard tree and 1 and 2 from West Opelousas C. arkansana. They all vary among themselves and from the type, but unless we are going to make a 34 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 species of practically every Hickory tree a good deal of allowance has got to be made for variation. I hope you are keeping a sharp lookout for Selma Hickories for I feel practically as much in the dark about the southern species as I did two years ago. 61 July 7, 1914. I have yours of the 2d and the package of Lindens. The perfectly glab- rous and very coarsely toothed leaves are those of what we have considered Tilia americana. 1 think it can be so considered until we find out some- thing to the contrary. The species which you have not sent before with leaves brownish below and covered with very fine pubescence is Tilia Michauxii, and the white-leaved species is certainly the same as that from Avery Island and River Junction, Florida. I have n’t seen this tree from any part of Louisiana but Avery Island. I expect to be back here on the 24th of August, and if by any chance your return has been delayed please let me find a letter from you then telling me where you are and what your plans are. 62 November 6, 1914. We got home safely a couple of weeks ago and saw more Hickories in Texas and in Arkansas than we did in Louisiana this time. About the Louisiana species I shall write you later. I have arranged to send Palmer to Natchitoches early next spring and to let him make his headquarters there through the season, or long enough to solve the botanical problems of that mysterious place. I think we can depend on him to make a good collection. I wish you would send me a specimen of the arborescent form of Rhus glabra either from Alabama or Louisiana. I should like, too, Louisiana specimens of the variety of Taxodium which you first reported from the state. The further west the specimen comes from the better. I should also like to have for purposes of future reference a specimen of Pinus glabra from Bayou Sara, as this so far as I know is the extreme western station reported for this species. We are having wonderful weather here but it is still dry. Do not work too hard over your class. Real botany is much more important. 63 November 25, 1914. Many thanks for the western specimens of Prunus. I hope to live long enough to get these things straightened out but the es at present does not seem to me very good. It is a tough propositi It is not surprising that Carya texana should be Ata near Shreveport. I believe it is a very widely distributed tree. 1965 | EWAN, LETTERS FROM SARGENT TO COCKS 35 am very sorry to hear about the losses in your family. It is an awful business altogether and I am afraid the end is still far distant. 64 December 21, 1914. I am taking another look at your Prunuses from Alabama and Louisiana and report as follows: Sardis 1 and 3, P. angustifolia, the latter with unusually broad leaves, but I suppose it is a vigorous shoot. Umbellata, 2, 4, 5, 8, and 11 from River Road. VI., munsoniana. 15 and 16, leaves only, may be the same. VII., the new Cherry, no flowers and the tree destroyed. 9, evidently the Big Tree Plum. Natchitoches 1, munsoniana. Natchitoches 6 and 7, only flowers. Natchitoches 3, 4 and 5, P. tarda. Shreveport 17 and 20, P. tarda. Ruston 2, fruit but no leaves. 10, munsoniana. 4 may be P. americana. This number is worth looking up. 9 may be a small-leaved Big Tree but there are no flowers or fruit. The common Plum of Louisiana is what we have called for the present the Big Tree Plum. I refer the following numbers to it: 20, 21, West Fel- iciana. 1, the Butler place. 4, Laurel Hill. Ventree’s Plantation, April 27, 1910. Laurel Hill, ii., iii. 1, open woods, Hammond. ? Butler place, 2. Covington, March 28, 1911. Natchitoches, 11. Ruston, 6 and 8. Ven- tree’s plantation, i., ii. Pineville, September 20, 1912. Archibald, 18. Shreveport. k., iii. Alta, 93, 14. Clear Lake, 1. Mhoon Plantation, i., ii. This is a common tree from Missouri to San Antonio, Texas, and, judging from your #9 from Sardis, extends into Alabama, although this is the only Alabama specimen I have seen. It appears to vary greatly in the size of the fruit, in the time of flowering, and in the time of the ripening of the fruit. I tried to make two or three species of it but I am afraid they are not very good, with the exception perhaps of P. reticulata which has dis- tinctly thicker leaves, smaller flowers and very late-ripening fruit. It ought to be possible to find some characters in the bark, and #2 from the Butler place may be distinct. I do not think the color of the fruit makes much difference in Plums as most of the species have red and yellow fruits. Prunus munsoniana is probably only introduced into Louisiana and Alabama as it is a more western species which has been cultivated for a great many years and is beginning to establish itself in many different parts of the country. I am still in doubt about the right name for the Big Tree Plum. It may be the Prunus mexicana of Watson, based on a very poor specimen of northeastern Mexico, or it may be that my Prunus arkansana is the name 36 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 It is unfortunate that the red-fruited Plum which Bartram found on one of the islands off the Louisiana coast has disappeared.* This was described by Rafinesque as P. coccinea which, if it is the Big Tree Plum, would be the oldest name. It may, of course, have been a red-fruited form of Prunus umbellata which appears to be common in the east Gulf States. With the compliments of the season and all good wishes for the New Year, I am, 65 Janu an ine [1915!] I can report about your West Feliciana Hickories as follo 2, 3, 6, 10, 13 and 20 I refer to Carya megacarpa Sevitl although of #3 we have no fruit. Next to Carya alba this is the commonest coast Hickory from South Carolina to Louisiana, extending into western Mis- sissippi and probably into western Louisiana. It appears to run into C. porcina and I am now thinking of making both, that is porcina and mega- carpa, forms of C. ovalis. West Feliciana +1, April and October 1914, which is very pubescent, is either another species, or, judging by the nuts, a variety of megacarpa. Notice that #1, April 20, 1913, is perfectly glabrous. I have n’t kept this specimen. #7 and 8 are the pubescent tree like #1 of 1914. 1785 of 1908 is evidently the same, but as there is no fruit I have not kept it. 1766 of 1907 is evidently megacarpa. I have n’t kept this. I believe that C. arkansana or C. texana get into Louisiana at Natch- itoches but I am not ready to say much yet about the western Louisiana Hickories. I hope that you are well and that 1915 is going to treat us all better than 1914 has, although present appearances for the happiness and pros- perity of the world do not seem particularly good. P.S.— I suspect that numbers 3, 4, 9, 13 and 14 of East Opelousas are the same as #1 of West Feliciana. 66 January 7, 1915. I have yours of the 14th. Am very sorry to hear about the grippe. Prunus is very difficult, especially when one does not see the plants growing. I have no doubt from what you tell me that your 1 and 3 of Sardis must be separated from angustifolia, for the time of the ripening of the fruit certainly ought to have some specific significance. I will look at the specimens again. It is my plan to let Mr. Palmer, who is collecting for the Missouri Bo- tanical Garden and the Arboretum, make his headquarters at Natchitoches ® Travels 423. 1791. Francis Harper (Naturalist’s ed. Travels 598. 1958) suggests Prunus hortulana Bailey. Rafinesque (FI. Ludoviciana 435. 1817) quotes Bartram al- most word for word, even to the “subulated point” of the leaf! 1965] EWAN, LETTERS FROM SARGENT TO COCKS 37 this year in the hope that he will be able to make a pretty complete collec- tion of the flora of that region. As he can reach Shreveport easily from Natchitoches I understand it ought to be a good place of operations for work about Shreveport and eastern Texas. I think if he can make a good collection in that part of the state it ought to simplify work for us in the other parishes. I wish you would tell me how early you think he ought to get to Natchitoches to get the earliest things, which I suppose will be Prunus and perhaps Salix. I should like to have him there early enough to get all the Plums as they come into flower. 67 January 11, 1915. Thanks for your letter received on Saturday. I went to look at those Plums again before finally deciding about reticulata, arkansana, etc. The type of P. mexicana is in the Gray Herbarium; it is a miserable scrap. The leaves look a good deal like what I call arkansana. Carya #1 from Lake Charles, the tree with stellate pubescent petioles and rachis, young leaves very red and fragrant, growing in low wet ground. The branchlets are glabrous but pubescent in your specimen of October 1913. Is this the same? Flowers have never been collected and are needed. Your specimen from Lake Charles, October 1913, is probably this #1. I do not know anything like this but we ought to find it in other places if it is a species, for Hickories are generally pretty widely distrib- uted. We ought to see more of this tree. My only other Lake Charles specimen from you is #2, September 1914. I have also two unnumbered specimens of fruit collected at Lake Charles by you in 1913; they are both pear-shaped and one is much larger than the other. Do you know to what they belong? +4, east of Opelousas, with red and fragrant young leaves, may be this #1 of Lake Charles but the pubescence is quite different. The fruit is pear-shaped but then the shape of the fruit does not count for much ap- parently. Is n’t #5 from Natchitoches the same as +1 from Lake Charles? +1 of East Opelousas, the tree with globose fruit, between the road and bayou and close to the road, may represent a new species. I do not find flowers and young leaves of this. Did you ever collect them? This is only a slight dose of what I will give you soon on the Hickory situation. If you think the trees in one state are difficult, how about the whole country? It is a tough proposition but I hope we are making a little headway. 68 January 14, 1915. I have had to give up the idea of our #5 of East Opelousas as a form of C. ovata with smooth bark. The nut is reddish, not white, and not suf- ficiently angled for ovata. The leaflets are constantly seven, while it is rare 38 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 to find seven leaflets in C. ovata and of a different shape; and the branch- lets are too slender. I understand that your numbers 9, 13 and 14 from the same region are trees with smooth pale bark. and 9 from Natchitoches which you think the same as 5 from Opel- ousas I believe are C. megacarpa, although the leaves are somewhat pubes- cent. The fruit of these numbers is distinctly pear-shaped and shows the yellow scales of megacarpa. I hope that in these scales we have found a good specific character. I have drawn up a description of Opelousas #5, for which I shall propose the name of C. leiodermis. I do not see why it has not got to go into a new group but I want to know more of this tree. +2 of Natchitoches seems like a tree which is growing at Starkville, Mississippi, which I am calling megacarpa stellipila. The fruit of your specimen is smaller but the pubescence on the young leaves is the same, and your plant is certainly some kind of a megacarpa. #¢1, Natchitoches, C. megacarpa. +7, Natchitoches, C. ovata. #13, Natchitoches, C. megacarpa. +10, Natchitoches, the leaf specimen is C. alba, but the fruit under this number seems to be C. megacarpa. £1, West Feliciana, seems the same as #2 Natchitoches, that is var. stellipila. +7, West Feliciana, C. megacarpa. +8, West Feliciana, seems the same as 1, West Feliciana. +4, West Feliciana, of which there is no leaf specimen, is probably a megacarpa. Have you leaves of this? If #4 is megacarpa, this leaves none of the West Feliciana numbers unaccounted for, but there are still many problems west of the Mississippi. 69 January 20, 1915. I feel sure that there is some mixture in Natchitoches #6. The flower- ing specimen seems to be Carya alba, while the mature leaves and fruit under this number belong to an entirely different group. I do not suppose you remember anything about this. 70 January 22, 1915. Have you any evidence that Carya cordiformis grows anywhere on the Louisiana coast? I have specimens from West Feliciana and from the neighborhood of Opelousas. I have nothing from Lake Charles or the Covington region. 71 January 27, 1915. I have yours of the 22d. I have about finished now all I can do on Hickories until I get more information and then shall take up the Lindens, about which I will write you later. Re) ee Pe SLT | et ey eee, eee eI 1965 | EWAN, LETTERS FROM SARGENT TO COCKS 39 I should not think there would be any necessity for your going to Natchitoches again if Palmer does his work as well as I hope he will, and in a week or so I will make suggestions for spring work. I very much hope to get to Louisiana myself again this spring for a few days but I do not know if I can manage it. If I do, of course we must meet somewhere. I am sorry to hear that you are likely to find new Plums in Louisiana, the worst proposition that there is. 72 January 29, 1915. I have yours of the 26th and write to say that I shall be very glad to have a specimen of C[arya| cordiformis from the Lake Charles region as it apparently is not a common coast tree. I have placed the Winnfield yard tree and 1 and 2 from West Opelousas with C. arkansana. The Winnfield tree may be all right but I am not very well satisfied with 1 and 2. C. arkansana usually grows in high dry ground and has dark deeply furrowed bark. These Opelousas trees were in low ground and had pale bark. I do not know what else to do with them unless to make a new species of them, and I do not like to do this until we can find more individuals. This is a thing still to attend to. 73 February 6, 1915. I have been through the Louisiana specimens of Crataegus again and I do not think that the situation is entirely hopeless. The trouble is that the plants of which we have insufficient material are pretty widely scattered, making it difficult to get what is necessary for their determination. Pal- mer will look after Natchitoches and I suppose can get to Shreveport where there are still a number of species we do not know. I have placed Crus-galli 33 from West Opelousas with C. tersa of Beadle which we found in the same locality, and your Crataegus 39 from Natchitoches with C. edura Beadle. In 1901 I collected at Shreveport a Viridis species which I now think is C. velutina Sarg., the type of which grows on the Red River at Fulton; and Bush has apparently collected it at Natchitoches “by iron bridge.” This is a new species for Louisiana. I collected at Shreveport another Viridis species with villose corymbs, color of the anthers not given and no fruit. It was in bloom on April 20th and seems distinct. And also on the same date another Viridis species with very broad, ovate leaves which had been out of bloom, : erie suppose, for two weeks. These two appear to be undescribed s At Munroe in 1901 I collected sterile branches of a shrub two to three feet tall with three-lobed leaves on the young shoots. It looks like a Viridis species. What is it? From you I have #16 & 23, a Viridis species collected June 1908 at Richlands, with half-grown fruit only. Do you know anything of these? 40 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 Natchitoches Crus-galli species. You do not give the color of the anthers of xii., xiii., xvi. and xvii. Can you? Minden. I have several fragmentary Crus-galli specimens from Minden, collected by me in April 1901 and by Bush in August of that year. I be- lieve you have never been there. In August Bush collected there “on right hand side of road to prairie west of town, just before getting to spring,” a Viridis species “with bark and branches rough, dark, scaly and flaky, ex- actly as in C. aestivalis which is beside it.” This sounds like the Viridis species we saw at Munroe growing with C. aestivalis. The specimens look alike. Flowers are much needed. Pineville. I collected on April 3, 1885, in bud a Crus-galli specimen with small, broadly obovate leaves pale and glabrous below and villose few-flowered corymbs, no spines and red branches. Is it not time that this specimen should be cleared up? You do not give the color of the anthers of your #19 Pineville, April 24, no year. No number and no fruit from a Crus-galli specimen collected by you out of bloom tae 5, 1912. This must have flowered very early for a Crus- galli spec I collected in 1901 at Shreveport a very narrow-leaved and distinct Crus-galli, twenty stamens, color of the anthers not given. All the undetermined and undescribed Louisiana species belong to the Crus-galli or Viridis groups and there are not many of them. The only exception is the plant from Holtonville which had pretty well grown fruit on March 20, 1911, and is therefore a very early flowered species. It may be a pruinosa but I am not sure. If it is, it is the earliest flowered of the whole group and hundreds of miles from any other species of the group. This certainly ought to be investigated this year. Munroe, where there are also Crataeguses to investigate, Minden and Pineville seem to me to be the Crataegus regions for you to visit this spring if you can manage it. At Munroe there are Hickories to look for and these I will write you about soon. These three places are not very far apart and might, I should think, be conveniently taken in on one trip. I hope to be able to meet you at Lake Charles, or some other convenient place, about the middle of April. 74 February 19, 1915. I believe now that we have the Hickories from east of the Mississippi pretty well in hand with the exception of one or two Alabama and Missis- sippi species or varieties, but there is still a bad lot of them in western Louisiana, southern Arkansas and Texas. I cannot believe now that 1 and 2 of western Louisiana can be referred to C. arkansana, variable as that species appears to be. The bark and place of growth won’t do. Notice the prominent tufts of white hairs on the margins of the leaves of +2 col- lected by you in April. These are absent from the young leaves of the Winnfield yard tree which seems to be a good arkansana. It is important i eee ee 1965 | EWAN, LETTERS FROM SARGENT TO COCKS 41 to find more of these West Opelousas trees and to learn something about them. Can you undertake this? +4 of East Opelousas seems the same as West Feliciana 1 and 8, mega- carpa stellipila. #1 of East Opelousas I am making a new variety of C. ovalis, var. sub- globosa, distinguished by its large fruit and by the stellate hairs on the rachis and midribs of the leaves. I am sorry we have no winter-buds of this. #3 of East Opelousas is a pretty good megacarpa and I am so consider- ing it. The buds, however, are very large. This seems to clear up East Opelousas, at least for the present. One new species and a new variety from there are pretty good I think. One of the interesting things to follow up is the tree “in dry woods”’ at Winnfield (I suppose this means near the quarry), with leaves hoary- tomentose below and slender branchlets. We thought when we collected it that it was some form of alba but of course it is not that. It is nearest stellipila but that so far as I know never has the dense hoary tomentum on the under surface of the leaflets. This is certainly an important tree to follow up. I will write again on the Lake Charles and Munroe Hickories. 75 February 20, 1915. I cannot make much out of the St. Charles’ ® Caryas partly because the specimens are not all properly numbered. This is my fault, however, at least very largely. Your #1 of October 1913 is certainly a different species from your #1 of September 1914. The fruit +1 of October 1914 belongs, I judge, with the September specimen although it is marked 1 October. Why cannot this be the same as East Opelousas #1, C. ovalis subglobosa? Perhaps my specimen collected April 2d, 1913, by river, with bright red unfolding leaves may be the same, and I am not at all sure that your specimen without number, October 1913, may be the same as your #1 of that date. #1 of October 1913 has very pubescent branchlets and buds. I have two lots of Lake Charles pyriform fruit from you marked Lake Charles Eng- lish Bayou, and West Lake Charles. These are not numbered but perhaps you can tell me to which specimen they belong. I fear, however, that the whole Lake Charles situation is so mixed up that it can only be straight- ened out in the field. On April 2, 1913, the Hickory leaves of Lake Charles were nearly fully grown. To be really useful they should be collected in an ordinary season by the middle of March I should think. Can you run over there for a day about that time and get very young leaves of all the Hickories and then meet me there about the middle of April? It seems to me that we should be there together. ® Surely a lapse for Lake Charles? 42 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 Carya from Bernice ii. April 9, 1913, is no doubt C. arkansana. I can tell nothing about the Munroe material which is too fragmentary. Pos- sibly the tree in the field, the first one we collected from last autumn, is an arkansana but I am not sure. In these woods there is a small-fruited spe- cies with pale bark which I do not know at all. If we meet at Lake Charles we might go to Munroe together and then I could get into Mississippi without going to New Orleans. Have you ever been to Delphi which is east of Munroe, and have you ever collected about Fort Hudson? I was there fifty-two years ago but of plants have no recol- lection except of the Magnolia which was very fine If such curious Hickories are found at Natchitoches, Lake Charles, Opelousas and Winnfield, why are there not others in other places in the state, and why is it not up to you to look out for them? I hope Palmer will pretty thoroughly cover Natchitoches and Shreveport and the regions be- tween Shreveport and Texarkana, in Caddo Parish, where I believe that some of the Fulton, Arkansas, trees will be found. I believe you have not been in that corner of the state and I doubt if this has ever been visited by a botanist. So far as I now know the Hickories which need more investiga- tion are 1 and 2 of West Opelousas, that is other stations where this tree should be found, the Lake Charles and Munroe species, and the tomentose Winnfield tree. The situation is bad, but not as bad as it was. I hope you will be able to make the early visit to Lake Charles. 76 February 23, 1915. I have yours of February 17th about the earliness of the season and am writing Palmer to advance the time of his arrival in Natchitoches. I am surprised to find that there is no specimen in the herbarium here of Castanea pumila from Louisiana where it must be a common tree. 77 February 25, 1915. Our large Opelousas Malus +2 seems to be the same as the Pineville tree which we have called M. ioensis creniserrata. It is certainly a very distinct looking form and perhaps when these things are better known than they are now would be considered a species. The trouble is that of these so-called species of Malus there seem to be many intermediates. 78 February 25, 1915. We are still in doubt for want of material about the common Malus at Selma. Judging by the leaves, it is the same as ioensis var. Palmeri, which is the first species which we collected west of Opelousas, and west of the Mississippi appears to be a common tree. I am afraid Harbison cannot get to Selma this spring. Is there any one 1965 | EWAN, LETTERS FROM SARGENT TO COCKS 43 there who could send you the flowers? There might be some one there who would know enough to gather them fresh and send them to you for pressing. 79 February 26, 1915. In writing you about Castanea pumila I overlooked the fact that we have a leaf specimen collected by you at Covington March 1911. The leaves are remarkably glabrous and very green on the under surface of that species. Sometime perhaps you will be able to get better material. 80 February 27, 1915. I have this morning your letter of February 23rd about Lake Charles, etc. I think now in Hickories that we can put a good deal of reliance on characters which have usually been passed over, like the scales on the unfolding leaves and on the winter-buds, therefore I think it desirable that we should get very young foliage of the Lake Charles Hickories this season if possible. That is why I suggested the middle of March but I daresay that would be too early. You can judge of that, however, in New Orleans much better than I can here. I should like very much to go there with you in April for I am particularly interested in those trees, and I think we ought to see them together. I do not know if it is really necessary to go to Winnfield in the spring as we have material of that tomentose plant, but if we are in Alexandria and going to Munroe it would not be too much trouble to stop off there for a few hours if the trains are running conveniently. I should hate the idea of passing a night there. The Hickory difficulties are narrowing down now unless some one is un- fortunate enough to unearth a new lot of them. 81 April 23, 1915. We got home safely last night direct from Selma via Montgomery. We had a couple of days in Selma and of course went out to your plantation. I do not think we saw many of the Plums because Harbison did not know apparently how to locate them. As he went about with you at Sardis he did not pay especial attention to exact localities. We saw your caretaker who was much pleased to know that I had seen you so recently and was very anxious for news of your wife. We did not get out to the bluffs on your place but went to Hatch’s Bluff, which I suppose has about the same vegetation as yours. There does n't seem to be but one Tilia there which looked very much like No. 2 at Lake Charles, but on the other side of the river we found another species with leaves very silvery white and tomen- tose on the lower surface, and like one of the Florida and Georgia species. 44 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 What most interested me at Hatch’s Bluff was a tree which looks like a Scarlet Oak but has fruit unlike that of any other Oak I have seen. If it is a Scarlet Oak it is certainly a very distinct variety and possibly it will have to be called another species. I will send you in a day or two a couple of the acorns and cups, and I especially invite your attention to the study of this tree. We only saw one specimen which was standing at the edge of the bluff about two or three hundred yards, I should think, from the point where we reached it across a plowed field close to a negro’s house. We saw the Oak which you called Quercus Durandii. It is certainly not the Texas tree of that name or one that grows near Columbus, Mississippi. I suppose, however, that it may be the Quercus austrina of Ash[e|. There was only one Plum in bloom and that was a good sized tree on the left- hand side of the road which leads past your place, which we reached by turning to the right at your gate. This tree was in the bottom, and in the bark looked a good deal like Prunus americana. I will send you in a few days a bit of my specimen, also the acorns of the peculiar Oak. Another Plum which interested me was a very large tree with dark, slightly scaly bark, on the right-hand side of Marion Junction Road, just after the crossing of the second bridge on the Orville Road, one mile west of the city limits going out by the Orville Road. This is a large tree at the foot of a rather steep bank and is rather inclining towards the road; it is quite new to me. It was out of flower but I will send you a piece of the specimen. Perhaps you will be able to refer these two specimens to some of your numbers. We found Mississippi and Alabama very hot, dry and dusty, and in our minds Lake Charles is still the pleasantest place we saw on this trip. I will write you again when I get things somewhat straightened out. | To be continued | 1965] SUSAN DELANO MCKELVEY, 1883-1964 45 SUSAN DELANO McKELVEY, 1883-1964 RicHarp A. Howarp In JUNE oF 1964, we received from Mrs. Susan Delano McKelvey a letter requesting her retirement from the Board of Overseers’ Committee to Visit the Arnold Arboretum and from her appointment as Research Associate. Mrs. McKelvey stated she could no longer do the things she used to do and wished to make way for someone more active in both of the roles she cherished. Her wish was a difficult one to approve and so these notes were originally prepared in appreciation for service at a friend’s retirement. We did not know that time was to be so short, for Mrs. McKelvey died on July the eleventh. Susan Adams Delano McKelvey was born in Philadelphia, Pennsyl- vania, March 13, 1883. She was graduated from Bryn Mawr, married Charles W. McKelvey in 1907, and made her home for a number of years at Oyster Bay, Long Island. About 1920, she moved to Boston. De- termined to study landscape gardening, she consulted Professor Charles Sprague Sargent at the Arnold Arboretum, volunteering her services in return for instruction in the identification of plant materials. Professor Sargent believed in testing volunteers, often with menial tasks such as pulling weeds or, in the case of Mrs. McKelvey, washing clay pots in the greenhouses. Under the guidance of William H. Judd, who was in charge of the greenhouses, Mrs. McKelvey also assisted in plant propagation; but she carefully reserved some time for reading, which led to research in the library. The famous lilac collection of the Arnold Arboretum was being developed, and, at Sargent’s suggestion, she undertook the task of compiling information on these plants. The manuscript she submitted was of such value that E. H. Wilson added a brief history of the lilac and Alfred Rehder a key and a generic description. The horticultural informa- tion, the descriptions, and the documentation, however, were the patient effort of Mrs. McKelvey. The manuscript was published in 1928 as The Lilac: A Monograph. It was dedicated simply to “The Professor.” This book, though long out of print, is still the only monograph of a favorite among flowering shrubs. In the following years Mrs. McKelvey spent some time in the south- western United States and became interested in the Yuccas and their relatives, often the most conspicuous plants of the drier areas. She sought help from various people in the identification of this difficult group and in the process learned at first hand how little information was avail- able. At the suggestion of Dr. Alice Eastwood, of the California Academy of Sciences, she began a diligent study of these difficult plants, photo- graphing and collecting specimens for the herbarium and for cytological 46 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 work. With Karl Sax, then a young cytologist, later to become director of the Arboretum, she published a paper basic to the understanding of the genus Yucca and its frequent associate Agave. In the study of the explorations that led to the discovery of the earlier published species of Yucca, Mrs. McKelvey became an authority on the routes of the early travellers and the explorers of the great Southwest, thereby uncon- sciously laying the foundation for still further research. The first part of her work on Yucca was published by the Arboretum in 1938. Its acceptance as a scientific study of high quality was soon evident, but the years of World War II delayed the issuance of part two until 1947. The last of the three books written by Mrs. McKelvey was probably the most demanding of her abilities to extract historical, particularly botanical, data from general travel accounts. In 1955, the Arnold Arbo- retum published The Botanical Exploration of the Trans-Mississippi West, 1790-1850, by Susan Delano McKelvey. Although a volume of over eleven hundred pages, each seemingly with one or more footnotes of documentation, the book is an eminently readable account both of history and of geographical exploration from the botanical point of view. The reviews of this book which appeared in history journals, as well as in bo- tanical ones, attest the value of the work with its interweaving of data from so many sources. For her achievements Mrs. McKelvey was acknowledged and honored by horticultural societies and organizations. In 1929, she received the Centennial Gold Medal from the Massachusetts Horticultural Society and the Schaeffer Medal from the Pennsylvania Horticultural Society. In 1957, the New York Botanical Garden honored her with the Sarah Gildersleeve Fife award given to individuals for work in the field of botanical literature. Mrs. McKelvey received her first appointment as a research assistant to the staff of the Arnold Arboretum in 1931, when Professor Oakes Ames was supervisor. It was the privilege of the succeeding directors to con- tinue her appointment as a member of the research staff. In 1928, Mrs. McKelvey accepted from the Board of Overseers of Harvard University an appointment to the Committee to Visit the Arnold Arboretum. Rarely did she miss a meeting or an opportunity to lend support and to express her interest in the work or the collections. She valued the living collections as plants to be enjoyed, as well as to be studied. She knew the value of herbarium specimens, and the difficulty of preparing them, and was appreciative of good herbarium management. Further, she realized the contributory value of cytological and anatomical work in her own monographic studies. Thus she fulfilled her duties on the Visiting Committee, interested and knowledgeable in the many aspects of a modern arboretum The association of Susan McKelvey with the Arnold Arboretum was a generous one. From her early efforts as a volunteer in the greenhouse to her bequest of research materials, her contributions were many. With her brother she devoted her energies toward raising a memorial fund to 1965 | SUSAN DELANO MCKELVEY, 1883-1964 47 Professor Sargent which remains a significant part of the endowment of the Arnold Arboretum. Through the years, her many gifts of letters and records, of specimens and photographs, of books and equipment, have all enriched our organizatio We regret the passing of a valued friend and supporter. BIBLIOGRAPHY * — 1925 — Syringa rugulosa, a new species from western China. Jour. Arnold Arb, 6: 153, 154. — 1928 — The Lilac: a Monograph. i-xvi + 581 pp. front., 171 pls. Macmillan Company, New York. — 1932 — A white pine blister rust demonstration. Horticulture 10: 331. 1933 Taxonomic and cytological ecletionehien of Y. ucca and Agave. (With Karl Sax). Jour. Arnold Arb. 14: 76-81. pl. 55. Arctomecon californicum. Natl. Hort. aha 3: 50. A verification of the occurrence of Yucca Bevel in Arizona. Jour. Arnold Arb. 15: 350-352. pls. 114-117. — 1935 — Notes on Yucca. Jour. Arnold Arb. 16: 268-271. pls. 138, 139. — 1936 — The Arnold Arboretum. Harvard Alumni Bull. 38: 464-472. 938 — Yuccas of the southwestern United States. Part one. 1-150. pls. 1-80, maps. Arnold Arboretum, Jamaica Plain. Yuccas of the southwestern United § cree Part two. 1-192. pls. 1-65, maps. Arnold Arboretum, Jamaica Pla — 1949 — A new Agave from Arizona. Jour. Arnold Arb. 30: 227-230. pls. 1, 2 om 195 Botanical exploration of the trans- Mississippi V West, 1790-1850. i-xl + 1144 pp. maps. Arnold Arboretum, Jamaica Plain — 1959 — A discussion of the Pacific Railroad report as issued in the quarto edition. Jour. Arnold Arb. 40: 38-67. * Prepared by Lazella Schwarten, 48 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 THE GENERA OF VITACEAE IN THE SOUTHEASTERN UNITED STATES ? GEORGE K. BrIzIcky ViraceseE A. L. de Jussieu, Gen. Pl. 267. 1789, “Vites,” nom. cons. (GRAPE FAMILy) Unarmed [rarely armed] woody, sometimes + fleshy vines climbing by stem-tendrils opposite the leaves [or rarely tendrilless perennial herbs, shrubs, or small sometimes succulent trees|; branches often swollen {or articulated] at the 3—7-lacunar nodes. Leaves alternate [very rarely the upper ones opposite|, usually 2- ranked, simple or palmately [rarely times also terminal [rarely only terminal], sometimes provided with a tendril, usually manifestly peduncled, bracteate. Flowers small, regular, usually 4- or 5-merous (except the usually 2-carpellate gynoe- cium), hypogynous, pediceled, bi- and/or unisexual ba least function- ally), the plants often polygamo-dioecious or -monoecious. Sepals connate, the calyx tube small, saucer- to cuplike. Petals pret [very rarely connate at base] and spreading or recurved at anthesis, or apically co- herent and dropping off as a “cup” at anthesis. Stamens opposite the petals, distinct [very rarely connate into a “staminal tube”, sterile [or very rarely wanting] in @ flowers; filaments filiform; anthers introrse [very rarely extrorse], dorsifixed above the base or near the middle, 2- locular at anthesis. Pollen medium sized or rarely small, usually 3-colp- wanting]. Gynoecium usually 2-carpellate [very rarely 3—8-carpellate], ‘Prepared for a generic flora of the ser aoai sna United States, a joint project of the Arnold Arboretum and the Gray Herbarium of Harvard University made pos- sible through the support of George R. Cooley and a ‘Nationa gna Foundation and under the direction of Carroll E v2 Bats fon eed C. Rollins. This treat- ment follows the pattern established in es ant sg in Qe series (Jour. Arnold Arb. 39: 296-346. 1958) and continued through those in volumes 40-45 (1959-1964). The area covered is bounded by and includes North Carolina, Tennessee, Arkansas, and Louisiana. The descriptions are based primarily on the plants of this area, with any supplementary ao in brackets. References which the author has not seen are marked by an a The author is cine to Carroll E. Wood, Jr., for his many valuable suggestions ; to Mrs. Julia Morton, of the Morton Collectanea, University of Miami, for fresh ma- terial of Cissus infected with fungus; and to Mrs. Gordon W. Dillon for = help in the preparation of the typescript. The figure was drawn by R. P. Monro ee eh a i a cia” i rca mmm 1965 | BRIZICKY, GENERA OF VITACEAE 49 syncarpous, usually rudimentary in ¢ flowers; stigma usually small, simple [rarely lobed to parted]; style single, short to long; ovary superior, often variously adnate to disc from base to almost the top, 2{-8]-locular often incompletely so; ovules 2, collateral [rarely ovule solitary] in each locule, ascendent from carpellary margin at base of locule, anatropous, apotropous, 2-integumented, with a thick nucellus. Fruit a 1-4[-6]-seeded berry, usually + fleshy and juicy. Seed coat usually hard, bony or crustaceous, with an adaxial raphe and abaxial chalaza; endosperm copious, hard-fleshy, rich in protein (aleurone grains) and oil, usually adaxially 3-lobed [rarely unlobed and/or ruminate]; embryo “spatulate,” small, straight, + axial; cotyledons small; radicle elongated, inferior. (Including Leeaceae Dumortier, Anal. Fam. Pl. 21, 27. 1829, nom. cons.) TYPE GENUS: Vitis A primarily pantropical family of about 13 genera with some 700 species, a few genera primarily of the North Temperate Zone; four genera in our area.* The genus Leea L., of the Old World Tropics, which differs from the rest of the family mainly in its nonclimbing habit, sympetalous corolla, staminal tube, extrorse anthers, lack of a nectariferous disc, and ovary with 3-8 one-ovulate locules, is considered to represent either a subfamily of Vitaceae (Gilg) or a family of its own (Suessenguth, Schultze-Motel). the corolla (petals free and expanding, or apically connate and forming a “cup”), the inflorescence type, and rarely the character of the tendrils, also are considered to be of generic significance. Since most of the indi- vidual generic characters may vary considerably within and form transi- tions between the genera, generic limits are in some cases somewhat un- certain. Although the vast majority of Vitaceae are sympodial, tendril- -bearing vines, there are some tropical tendrilless species with the stems either monopodial or sympodial.2 The occurrence of intermediates between in- * After this article was in press, L. H. Shinners recorded (Sida 1: 384. 1964) a m of Tulane Uni versity, Covington, St. Tammany Parish, Louisiana. This collection 7 ivation.” possibly represents an escape from cultiv: * The Braun- “Eichler theory (Troll, pp. 30-32; Suessenguth, pp. 183-190) which re- gards the shoots of tendril-bearing members of Vitaceae as congenitally formed sym- i of ontogeny, should be men- dener (Das Mikroskop see aay based ois as rare tioned. According to this theory, supported recently by Shah pesiden on the same basis, the tendrils in Vitaceae are extra- lateral oa of a nopodially branched shoot, destitute of subtending le lea It seems appropriate os nant Troll’s comments (loc. cit.) on Nageli & cenit theory. ae cn cae the ontogeny 50 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 florescences and tendrils (e.g., inflorescences sensitive to a contact stimulus and performing slight twining movements; those divided into a floriferous part and a tendril; or tendrils bearing a few flowers at the tips of their branches) is ample evidence for regarding the tendrils as modified in- florescences. The tendrils (and inflorescences) are either ‘“continuous” (at each of 3—7 or more successive nodes) or “intermittent” (lacking on every third node) and are either monochasially 2—12-branched, or some- times once or twice forked, or unbranched. The tendrils attach them- selves to the support by twining (coiling) or by adhesive discs sad eine at the apices of their branches in response to a contact stimulus (e.g. most of the species of Parthenocissus; the monotypic Pterocissus Urb. & Ekm.; and some species of Cissus and Tetrastigma (Miq.) Planch.). Cross-pollination by insects seems to be the rule, although at least in some monoclinous or polygamo-monoecious species or varieties self- pollination (including geitonogamy) has also ron recorded. Floral anatomical data known for a few genera and species indicate incomplete septation of the ovary and “anatomically parietal’ placenta- tion (Kashyap, 1957; Nair & Mani). Chromosome counts made for eight genera (about 57 species) are 2n = 22, 24, 26, 28, 30, 32, 38, 40, 44, ca. 45, 48, 50, 52, 60, 72, 80, 96, 98, Suggesting that both aneuploidy and polyploidy have been of importance in the evolutionary development of Vitaceae. very characteristic seeds of VirompEAE make possible the safe de- termination of fossils of this subfamily. The seeds are usually more or less oblong-obovate or obovate in outline (or sometimes pyriform and distinctly beaked at the micropylar region); the shape in cross section depends upon the number of seeds in a fruit. The seed coat is composed of a thin, outer, parenchymatous membrane and a hard, usually bony, inner layer (sclerotesta). The adaxial (ventral) surface of the sclerotesta exhibits a median cordlike to threadlike raphe, extending from the hilum to the seed apex and onto the convex abaxial (dorsal) side, where it joins a round to linear, depressed to somewhat elevated “chalazal knot.” On the adaxial surface the raphe is flanked on either side by a deep groove (an infolding of the sclerotesta) varying in shape and length; the endosperm is consequently deeply three-lobed on the adaxial side. (See Fic. 1.) The family is closely related to Rhamnaceae, from which it differs mainly in the predominance of sympodial vines, the hypogynous flowers, the predominantly two-ovulate locules, the apotropous ovules, the some- what different pollen (similar to Araliaceae rather than to Rhamnaceae), leads to the improbable assumption of “extra-axillary’ branches destitute of subtend- that, on the contrary, the latter itself requires explanation through comparison with the fully doses [mature] state.” (Translation supplied.) 1965 | BRIZICKY, GENERA OF VITACEAE 51 the type of fruit (berry), the usually lobed or ruminate endosperm, and the “spatulate” (vs. “investing”) embryo. Together these families con- stitute the order Rhamnales. It is noteworthy that species of Tetrastigma, and perhaps of some other genera of Vitaceae, are the exclusive host plants of the species of Rafflesia in Malesia. Economically the family is most important for the species of Vitis (q.v.). b he Q . 1. Diagrams to show characteristic seeds of Vitaceae a ly Vitoideae wax poe membranaceous layer of seed coat removed; relative sizes of seeds modified for ease of comparison. a—c, Vitis (subg. Vitis) Labrusca: a, convex abaxial surface, showing “chalazal knot” with wrt extending to see za to show variation in raphe a grooves: d, Cissus incisa; e, Ampelopsis arborea; ft, pa ese pony Bs g, Vitis (subg. Muscadinia) rotundifolia. REFERENCES: Apatia, R. D., B. N. Mutay, & G. R. Hrncorant. A contribution to the embryology of Vitis trifolia L. Jour. Univ. Bombay II. 19(3B): 1-10. 1950: IL. ibid. 21(3B): 51-60. 1953. [= Cayratia trifolia (L.) ee ApKINSON, J. Some factors of the anatomy of the Vitaceae. Ann. 27: 133-139. pl. 15. 1913. [Cissus, Leea, Parthenocissus, and Vitis, nash anatomy. | ARBAUMONT, J. D’. La tige des Ampélidées. Ann. Sci. Nat. Bot. VI. 11: 186-255. pls. 11-14. 1881. [Anatomy.] ArsHap Att, M. Studies on the nodal anatomy of six species of the genus Vitis (Tourn.) L. Pakistan Jour. Sci. Res. 7: 140-152. 1955.* [Vitis apparently understood in the sense of Bentham & Hooker to include all genera of Vitaceae, except Pterisanthes Bl. and Leea.] sr pe Loranthacées. Hist. Pl. 11: 408-494. 1892. [Vitaceae, as a tribe e’”’) of Loranthaceae, 426-430, 472, Beek G., & J. D. Hooxer. Ampelideae. Gen. Pl. 1: 386-388. 1862. BERLESE, A. N. Studi sulla forma, struttura e sviluppo del seme nelle Ampelidee. 52 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 Malpighia 6: 293-324, 482-536. pls. 11-18bis. 1892. [Form, structure, and development of seeds in Vitaceae. | Berry, E. W. Seeds of a new species of Vitaceae from the Wilcox Eocene of Texas. Jour. Wash. Acad. Sci. 19: 39-41. 1929. [Ampbelocissites lytlensis, presumably intermediate between Ampelocissus and Vitis Bousats, D. Contribution a l’étude des causes de la resistance des Vitacées au mildiou de la vigne (Plasmopara viticola (B. et C.) Berl. et de T.) et leur mode de transmission héréditaire. (English summary.) Ann. Amél. Pl. 9: 5-233. 1959. [See also ibid. 11: 401-500. 1961.] Branpt, M. Untersuchungen iiber den Sprossbau der Vitaceen mit besonderer Beriicksichtigung der afrikanischen Arten. Bot. Jahrb. 45: 509-563. 1911. [An i i a work on morphology, especially on branching systems, of Vitace CANDOLLE, : P. pe. Ampelideae. Prodr. 1: 627-636. 1824. CHaApEFAUD, M. Sur la morphologie des Vitacées. Compt. Rend. Acad. Sci. Paris 228: 1660-1662. 1949. [Some data in support of the hypothesis of monopodial branching system in the tendril-bearing Vitaceae. See also BUNGNON under Vitis. ] Darwin, C. On the movements and habits of climbing plants. Jour. Linn. Soc. Bot. 9: 1-118. 1867. [Vitaceae, 79-87. ] Gitc, E. Vitaceae. Nat. Pflanzenfam. III. 5: apy 1896. [See also ibid. Nachtr. III. 211. 1908, and Nachtr. IV. 193. 1915.] Kasuyap, G. Studies in the family Vitaceae. I. ie al morphology of Vitis trifolia L. Agra Univ. Jour. Res. Sci. 4(Suppl.): 777-784. 1955.* [Cayratia trifolia. | Studies in the family Vitaceae. II. Floral anatomy of Vitis trifolia Linn., Vitis latifolia Roxb. and Vitis himalayana Brandis. Jour. Indian Bot. Soc. 36: 317-323. 1957. [Cayratia trifolia, Ampelocissus latifolia, and Parthenocissus himalayana; septation of the ovary incomplete; placentation anatomically parietal. | . Studies in the family Vitaceae. III. Floral morphology of Vitis latifolia Roxb., Vitis himalayana Brandis and Vitis trifolia Linn. Ibid. 37: 240-248. 1958. [Morphology and embryology KiRCHHEIMER, F. Rhamnales I: Vitaceae. Jn: W. Joncman, ed., Fossil. Catal. 4: 1-174. 1939. [Basic seed structure in living and fossil representa- tives, 1-9; history of Vitaceae, 118-129. LENGERKEN, A. von. Die Bildung der Haftballen an den Ranken einiger Arten der Gattung Ampelopsis. Bot. Zeit. 43: 337-346, 353-361, 369-379, 385- 393, 401-411. pl. 4. 1885. [Morphology, anatomy, and functioning of tendrils; Ampelopsis, 3 spp. (including “A. hederacea” apparently = A. brevipedunculata), 343-359; and Parthenocissus (as ““Ampelopsis”), 3 spp., 359-393. | Mrxr, S. Seed remains of Vitaceae in Japan. Jour. Inst. Polytech. Osaka Univ. D. 7: 247-273. 1956. [Fifteen spp. in 5 genera sige Pliocene to present; numerous illustrations of seeds of living and fossil spp.; in the keys, legends, — mon eaintie substitute “chalaza” and ‘ “chalazal” for ‘“‘charaza”’ and “charazal.’’| Mutay, B. N., N. Cc Narr, & M.S. R. Sastry. Contribution to the embryology of Vitaceae. Proc. Rajasthan Acad. Sci. 4: 16-28. 1953.* Narr, N. C., & K. V. Mant. Organography and floral anatomy of some species of Vitaceae. Phytomorphology 10: 138-144. 1960. [Cissus, 5 spp.; Ee ee ee ee a as a i ae aa ll ws 1965 | BRIZICKY, GENERA OF VITACEAE 53 Vitis, 1 sp.; Ampelocissus, 1 sp.; Cayratia, 1 sp.; placentation parietal, cae sap derived from axile condition & P. N. N. NAmMBIsAN. Contribution to the floral morphology and embryology of Leea sambucina Willd. Bot. Not. 110: 160-172. 1957. [Disc absent; placentation anatomically sean embryo sac of Polygonum- type, synergids egglike, antipodals epheme PLANCHON, J. E. Ampelideae. Jn: A. & C. DE “CANDOLLE, Monogr. Phaner. 5(2): 305-654. 1887. [Basic monograph of the family. ] Rarer, M. S. The anatomy of grape and related genera. Oregon State System Higher Educ. Grad. Theses 1932/1942: 84, 85. 1946.* [Six genera of Vitaceae. SCHNEIDER, C. K. IIlustriertes Handbuch der Laubholzkunde. vol. 2. v + 070 pp. Jena. 1907-1912. [Vitaceae, 300-323. 1909; 1032-1034. 1912.] SCHULTZE-MorTEL, W. Reihe Rhamnales. Jn: H. Metcutor, Engler’s Syllabus der Pflanzenfamilien. ed. 12. 2: 300-304. 1964. [Vitaceae, 302, 303; Leeaceae, 303, SHAH, J. J. Studies on 1 the stipules ae six ier of Vitaceae. Jour. Arnold Arb. 40: 398-412. 1959. [Cissus, ; Cayratia, 2 spp.] Origin, development and et of tendrils of Vitaceae. Delhi Univ. Proc. Summer School Bot. Darjeeling 1960: 430-444. 1962. [Cissus, 3 spp.; Cayratia, 2 spp.; “the tendril or an inflorescence represents an extra-axillary and ebracteate lateral branch . . . the shoot of the Vitaceae is obviously a monopode . . . the sympodial as well as the dichopodium theories are not supported by the present study.” SUESSENGUTH, K. Vitaceae. Nat. Pflanzenfam. ed. 2. 20d: 174-371. 1953. [Extensive bier rae | , . Leeaceae. Ibid. 372-390. 1953. [Vitaceae subfam. Leeoideae Gilg. ] Trott, W. Vergleichende ae der hoheren Pflanzen. Band 1(Teil 1). xii + 955 pp. Berlin. 1935-1937. [Vitaceae: tendril-bearing shoots (Vitis), 30-32; shoot system Vout 627, 628; tendril systems, 840-846. See also 7.] SUESSENGUTH, 180-1 VaTSALA, P. Chromosome studies in Ampelidaceae. Cellule 61: 191-206. 1960. [Vitis vinifera, 7 vars.; Cissus, 5 spp.; Cayratia, 2 spp.; Ampelocissus (“Ampelopsis”), 1 SP; ' Leea, 1 sp.; includes also a list of chromosome us] Vita, P., & V. VERMOREL. es, graphie. 7 vols. Paris. 1901-1910. [A monu- mental, although somewhat obsolete, work on the family, primarily on Vitis. | Key TO THE GENERA OF VITACEAE General characters: usually woody vines climbing by tendrils opposite the leaves; leaves alternate, simple or palmately 3-5-foliolate, rarely bipinnate ; in- florescences opposite the leaves; flowers small, regular, 2-carpellate gynoecium), hypogynous ; stamens opposite the petals ; nectariferous disc intrastaminal, conspicuous, rarely obscure; ovary 2-locular, with 2 subbasal, ascendent ovules in each locule ; fruits 1-4-seeded berries ; seed with a conspicuous abaxial chalazal knot and 2 deep adaxial grooves, one on each side of the raphe. A. Petals expanding at anthesis, distinct; inflorescence usually cymose, without a tendril; pith Neate B. Flowers 4-mero inflorescence an umbel-like compound ox ae chasium) ; mea 3- 5 faliclate, Sie AG, ee ee es SSUS. 54 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 B. Flowers 5-merous. C. Nectariferous disc conspicuous, cupular, distinct at least in the upper part; inflorescence a thyrse, usually repeatedly bifurcate and corymb- like, rarely raceme-like; leaves simple or bipinnate; ay bifurcate, without MEARE MRM. is alia Se enlists, teehee wee eee . Ampelopsis. C. Nectariferous disc obscure, reduced and fused with ie of ovary; inflorescence a pyramidal cymose panicle with monochasially branched rachis; leaves palmately (3)5-foliolate; tendrils monochasi- ally 3-12-branched, developing adhesive discs at tips of their PIED Ser ene A etiam dining as 3. Parthenocissus. A. Petals cohering at apex and falling off as a unit at anthesis; inflorescence a panicle, sometimes with a tendril; pith brown; leaves simple. .... 4. Vitis. 1. Cissus Linnaeus, Sp. Pl. 1: 117. 1753; Gen. Pl. ed. 5. 53. 1754. Deciduous |or evergreen] + woody, sometimes + fleshy [or herba- ceous|] vines [rarely tendrilless perennial herbs, shrubs, or small + succulent trees]; tendrils continuous, simple or bifurcate, without [or very rarely with] adhesive discs; roots often tuberous. Leaves simple, unlobed or lobed, and/or palmately 3[—5]-foliolate, membranaceous | to leathery] or + fleshy, petioled [rarely sessile |. Inflorescences compound umbel-like cymes (pleiochasia) without tendrils, each opposite a leaf [rarely terminal], continuous, the secondary branches sometimes bifur- cate. Flowers 4-merous [exceptionally 5-merous|, bisexual [or bi- and unisexual, the plants then polygamo-monoecious], greenish- or creamy- yellow, white, or purplish. Calyx cuplike, indistinctly shallowly 4-lobed to subentire. Petals 4, usually expanding, spreading at anthesis. Stamens 4. Nectariferous disc cuplike, adnate high up on the ovary [or distinct and 4-parted, or replaced by 4 distinct bowl-like glands], the free upper margin + 4-lobed or crenate to subentire, spreading or upright. Stigma small; style rather long, + filiform; ovary 2-locular, apparently incom- pletely so (in our species).* Berries subglobular-ovoid to -obovoid, 5—12 mm, in diameter, dark purple to black, usually 1(2)-seeded, inedible. Seed obovoid, indistinctly beaked at base; chalazal knot linear, on lower part of abaxial face, imperceptibly merging with the narrow, linear, salient raphe running toward seed apex on abaxial and down to hilum on adaxial face; 2 adaxial grooves short, oblong, near base of seed (Fic. 1, d); sides of seed indistinctly very coarsely reticulate. Typr spEcIEs: C. vitiginea L., 2n = 26. (Name from Greek, Aissos, an ancient name for ivy, applied by Linnaeus to this genus.) A nearly pantropical sae (exclusive of Hawaii and Polynesia) of nearly 350 species, a few in the warmer parts of the temperate zones. Subgenera Cissus (subg. ie Phin), of pantropical distribution, * Flowers of a few random herbarium specimens of Cissus incisa and C. trifoliata, boiled and dissected, invariably showed ovaries with only a narrow central strip of the anes which bore two collateral ovules on each ied at base. Additional material uld be investigated before any conclusions are draw Sic a 1965 | BRIZICKY, GENERA OF VITACEAE 55 and CyPHOSTEMMA Planch. (sometimes segregated as the genus Cypho- stemma (Planch.) Alston), of the Old World Tropics, have been recognized. Three species of subg. Cissus occur in our area. Cissus incisa (Torr. & Gray) Desmoul., marine-vine, a stout vine with fleshy, three-foliolate or deeply three-parted leaves, usually occurs on limestone in hammocks or in rocky and sandy open woods and on bluffs, from central peninsular Florida to Texas, Arizona, and Mexico, north to Arkansas, Missouri, and Kansas. Cissus trifoliata L., of the West Indies, Mexico (?), and northern South America (?), differing in the always three-foliolate leaves with smaller, often flabellate, more regularly toothed leaflets, and smaller fruits, occurs in hammocks in southern peninsular Florida and on the Florida Keys. The very polymorphic, tropical American Cissus sicyoides L., with more or less membranaceous, simple, unlobed, medium-sized leaves, and often with long aérial roots, occurs in hammocks, from the Florida Keys northward in peninsular Florida to Polk County. The inflorescences of this species (sometimes also of C. trifoliata) frequently are greatly expanded and deformed by a smut, Mycosyrinx Cissi (DC.) Beck, the flowers being transformed into elon- gated subcylindrical formations containing spores of this fungus. The genus Spondylantha Presi was based on this monstrosity. The few published data indicate an “anatomically parietal” placenta- tion in a few investigated Asiatic species (Nair & Mani). Chromosome counts for 12 species are 2m = 22, 24, 26, 28, 32, 44, ca. 45, 48, 50, and 96, Suggesting that both aneuploidy and polyploidy have been of im- portance in the evolution of the genus. The genus is closely related to the monotypic Laotian Acareosperma Gagnep., and probably to Cayratia Juss. The genus is in need of a modern revision The tuberous roots of Cissus trifoliata are reputedly poisonous. The aérial roots of C. sicyoides are locally used in Central America for wattling baskets, and leaves of this species seem to contain saponines and are locally used for washing linen in tropical America. REFERENCES: Under family references see BRANDT, DARWIN (pp. 83, 84), Nam & MAant1, SHAH, SUESSENGUTH (pp. 237-277), VaTSALA, and Vata & VeERMOREL (1: 80-108. BANERJEE, K. G. Some observations on the anatomy and biology of the aérial adventitious roots of Vitis guadrangularis Wall. Proc. Indian Sci. Congr Assoc. 16: 229. 1929.* [C. quadrangularis L.} Cuavan, A. R., & J. aH. Origin and development of the vegetative axillary bud in Vitis repens W. & A. Ibid. 42(3, abs.): 225. 1955.* [C. repens Lam. | ; . Septate epidermis and stomata in the tendril es ee pallida W. & A. Curr. Sci. Bangalore 24: 84, 85. 1955.* [C. repanda Vahl. } Descorncs, B. Un genre méconnu de V itacées: iemlind et distinction des genres Cissus L. et Cyphostemma (Planch.) Alston. Not. Syst. Paris 56 JOURNAL OF THE ARNOLD ARBORETUM [ VoL. 46 16: 113-125. 1960. [Data in support of segregation of Cyphostemma; many new combinations. ] Lawrence, G. H. M. Cissus and Rhoicissus in cultivation. Baileya 7: 45-54. 1959. [See also ibid. 8: "S2, 53. 1960.] Lyncu, R. I. On branch tubers and tendrils of Vitis gongylodes. Jour. Linn. Soc. Bot. 17: 306-310. pl. 15. 1879. [C. gongylodes (Bak.) Burch. ex Planch. | Mapan, C. L., & S. L. Nayar. A ig ayy ina po of the stem of Cissus quadrangularis Linn. Jour. Sci. Indus. Res. 18C(12): 253-255. 1959.* Narr, N. C., & V. PARASURAMAN. amas of Vitis pallida. Curr. Sci. Bangalore 23: 163, 164. 1954.* [C. repanda.]| Suau, J. J. On the “sensory epidermis” of stipules . Ries repens W. and A. Curr. oy aor en 23: 302, 303. 1954.* [C. repens. ] Nucleated vessel elements in tendrils of Vie ene W. & A. Ibid. 65. 1954.* [C. repens. | . On the nature of primary vascular — in the tendril of Vitis pallida Ww. & A. Ibid. 24: 165. 1955.* [C. repan —. Phloem fibre in the tendril of Vitis ie W. and A. Sci. Cult. 20: 390, 391. 1955.* [C. repens. | . Developmental pattern of the primary vascular system in the stem of a Cissus sp. Jour. Indian Bot. Soc. 39: 443-454. 1960.* [Cissus sp., close to C. quadrangularis, the node quadrilacunar, 453. | SrivasTAvA, J. G. On aérial roots in Vitis quadrangularis Wall. Curr. Sci. Bangalore 20: 133. 1951.* [C. quadrangularis. | 2 Ampelopsis Michaux, Fl. Bor.-Am. 1: 159. 1803; on Planchon in A. & C. de Candolle, Monogr. Phaner. 5(2): 45 3 Deciduous woody vines; tendrils usually few and scattered [or many and continuous], bifurcate, devoid of adhesive discs. Leaves membrana- ceous, simple or bipinnate [pinnate or palmate], long-petioled. Inflor- escences few- to many-flowered thyrses, each opposite a leaf, continuous, often repeatedly bifurcate and corymb-like, more rarely elongated, raceme- like, the ultimate branches ending in usually umbel-like, crowded, sessile, 3-flowered dichasia. Flowers 5-merous [4-merous in 1 species], bisexual, greenish. Calyx very small, + saucer-like. Petals 5, spreading at anthesis. Stamens 5. Nectariferous disc cupular, distinct except the lower part adnate to the ovary, shallowly lobed or crenulate to subentire on the margin. Stigma small, simple; style elongated, slender; ovary 2-locular, apparently incompletely so in our species. Berries subglobular to obovoid, 8 mm. in diameter, green, blue, or black, 1—4-seeded. Seed obovate to broadly obovate in outline, convex on abaxial, + angular on adaxial side; chalazal knot + spatulate, at or below center of abaxial side; raphe narrowly linear, salient on abaxial, threadlike on adaxial surface; 2 adaxial grooves oblanceolate, usually somewhat curved, in the lower half of seed (Fic. 1, e). Lectotype species: A. cordata Michx.; see Britton & Brown, Illus. Fl. No. U. S. ed. 2. 2: 509. 1913. (Name from Greek, ampelos, grapevine, and opsis, likeness, appearance; i.e., having the ap- pearance of the vine, Vitis vinifera.) 1965 ] BRIZICKY, GENERA OF VITACEAE 57 A genus of about 22 species, of tropical and warm-temperate regions of Asia (19 species, centered in China) and North America (3 species) ; two indigenous and one naturalized species in our area. Ampelopsis cor- data Michx., 2n = 40, a high-climbing, nearly glabrous vine, the leaves simple, ovate, cordate or truncate at base, unlobed or slightly 3-lobed, coarsely and sharply toothed, the berries greenish or bluish, occurs in rich woods and bottomlands from northern Florida to Texas and Mexico, north to southeastern Nebraska, southern Illinois, southern Indiana, southern Ohio, and Virginia. It seems to be closely related to the eastern Asiatic A. brevipedunculata (Maxim.) Trautv. (A. heterophylla Sieb. & Zucc.), 2n = 40, a more or less pubescent vine with ovate-cordate, 3-lobed, simple leaves and bright-blue berries, which is cultivated for ornament and is sporadically naturalized in at least North Carolina and Georgia (Small), northward to Ohio and New England. Am~pelopsis arborea (L.) Koehne (A. bipinnata Michx.), pepper-vine, 2n = 40, generally a high-climbing vine (or sometimes bushy), with bipinnate (or -ternate) leaves, rather small, coarsely toothed or incised leaflets, and dark-purple to black berries, occurs in swampy woods from southern Florida to Texas, north to Okla- homa, Missouri, southern Illinois, and to Virginia and Maryland. It is mexicana Rose, of western Mexico (Sinaloa to Guerrero), remain un- certain Inflorescences showing transitions to tendrils, with a branch or the peduncle spirally coiled, have been frequently observed’ in Ampelopsis cordata. In at least our species, the septa of the ovary are not connate at the center. The genus appears to be homoploid, 2m = 40 having been found in the six species investigated. A close relationship to Ampelocissus Planch., Vitis, and especially to Parthenocissus has been assumed by various authors. The genus is in need of a modern revision. REFERENCES: Under family references see BRANDT, LENGERKEN (pp. 343 3-359), SCHNEIDER (pp. 318-323, 1033, pen! SUESSENGUTH (pp. 313-315), VATSALA, and VraLa & VERMOREL (1: 68-75. —_— Back pas heterophylla. Addisonia 16(3): 35, 36. pl. 530. [A. brevipedunculata Deneck P. I. Seeds of Ampelopsis from the tertiary deposits of the territory of USSR. (In Russian.) Bot. Zhur. 42: 643-648. 1957. MEEHAN, T. The native flowers and ferns of the United States. vol. 2. v 200 pp. oe 1-48. 1879. [Vitis indivisa Willd. (= Ampelopsis a: 21-24, pl. 6.] 3. enep ots Planchon in A. & C. de Candolle, Monogr. Phaner. 5(2): 447. 1887. pbctoue ai evergreen] high-climbing vines; tendrils aagbige monochasially 3-12-branched, with [rarely without] adhesive discs a 58 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 tips of branches. Leaves palmately [very rarely pedately| (3)5(7)-folio- late, long petioled, leaflets medium sized to large, coarsely toothed, sub- sessile to long petiolulate. Inflorescences usually many-flowered cymose panicles with prolonged monochasially branched rachises [or corymb-like bifurcately branched thyrses], lacking tendrils, opposite the leaves, inter- mittent, and sometimes also terminal. Flowers 5-merous, bisexual |some- times also functionally ¢ |, greenish. Calyx small, cupular, shallowly irregularly 5-lobed. Petals 5, concave, thickish, expanding at anthesis. tamens 5. Nectariferous disc usually obscure, fused with the ovary base, distinguishable from the ovary by darker color, secretion of nectar, and natomical structure [or rarely disc evident]. Stigma simple, small; style narrowly conical, short; ovary 2-locular, the septa meeting in the center and apparently not or weakly connate. Berries subglobular, 5-10 mm. in diameter, dark blue to black, often glaucous, 1—4-seeded, inedible. Seed obovate to obovate-suborbicular in outline, convex on the abaxial, angular, + keeled on the adaxial side; chalazal knot round to short- spatulate, situated in or above the center of abaxial surface, extended to the seed apex into a linear + salient raphe inconspicuous and threadlike on the adaxial side; 2 adaxial grooves narrow, cracklike, slightly curved, extending nearly from apex to base of seed (Fic. 1, f). LrcTotyPr species: P. quinquefolia (L.) Planch. (Hedera quinquefolia L.); Britton & Brown, Illus. Fl. No. U. S. ed. 2. 2: 511. 1913. (Name from Greek, parthenos, virgin, and kissos, ivy; an equivalent of vigne-viérge, the French name for the type species.) A genus of about 15 species, of the temperate and tropical regions of eastern Asia and North America, centered in Asia (12 species); one species in our area. Parthenocissus quinquefolia (Ampelopsis quinquefolia (L.) Michx., A. hederacea (Ehrh.) DC.),* Virginia creeper, woodbine, 2” = 40, occurs *Rehder (1905) presented strong arguments for regarding Hedera quinquefolia L. as a species of Parthenocissus developing adhesive discs on its tendrils. Consequently he applied P. quinquefolia (L.) Planch. to this species. Suringar supported Rehder’s iscs, known at present as P. vitacea (Knerr) Hitchc. Rehder’s has generally been a American and European botanists, but a few botanists have adopted bner’s interpretation. Thus, in his Catalogus Florae Austriae (1 1957), Janchen treats P. quinquefolia species without adhesive discs on the tendrils, citing in i cea (Knerr) Hitchc. and Hedera quinquefolia L. T species with adhesive rece crane known as P. quinquefolia (L.) Planch., Janchen c f of escens (Schldl.) Graebner (amplif. Janchen)” with “P. quinquefolia Rehder, Schneider, Fritsch, Hegi, Mansfeld, non Graebner” and “non Hedera quinque- folia L.” in synonymy. Made in good faith, but hardly well founded, this action un- fortunately returns confusion to dc already stabilized nomenclature of the American species of Parthenocissus Gleason (1947) showed that Vitis inserta Kerner, on which Parthenocissus inserta (Kerner) Fritsch is based, “is merely a synonym of P. quinquefolia.’ Hence, the correct name for the species known under that name is P. vitacea (Knerr) Hitchc. i. 1965 | BRIZICKY, GENERA OF VITACEAE 59 in hammocks and woods and on rocky banks from Cuba, the Bahamas, and southernmost Florida (including the Keys) north to Maine, Vermont, southwestern Quebec, New York, Ohio, Indiana, Illinois, Wisconsin, Minnesota, and southern Manitoba, west to Texas, and in Mexico and Guatemala (Huehuetenango). The species is variable, and several forms and varieties, mostly horticultural, have been distinguished. The most frequently encountered wild form of var. guinquefolia is f. hirsuta (Donn) Fern. (P. hirsuta (Donn) Small, P. pubescens (Schlecht.) Graebn.), with soft-pubescent lower surface of leaflets, young branchlets, and inflo- rescences. Parthenocissus quinquefolia var. murorum (Focke) Rehd., differing from the typical variety mainly in tendrils with 8-12 short branches (as against 3~8 elongated branches) and somewhat broader and thicker leaflets, has been recorded at least from Florida, the Bahamas, Cuba, and Mexico. The species is closely related to P. laetivirens Rehd., of central China. Parthenocissus vitacea (Knerr) Hitchc.,4 2m = 40, which differs from P. quin quefolia hi in the bifurcately branched corymb-like in- florescences (with a tendency toward racemose branching), tendrils usually lacking rahe discs, and somewhat larger berries and seeds, occupies a range north and west of our area, occurring from Nova Scotia and Quebec to southern Manitoba and Montana, south to New England, Pennsylvania, Ohio, Indiana, Illinois, Missouri, Kansas, New Mexico, western Texas, and Arizona. The closely related P. heptaphyila (Buckl.) Britt. ex Small, perhaps a variety or subspecies of the preceding, with (6—)7-foliolate leaves and smaller leaflets, seems to be restricted to central Texas. The flowers in at least our species are proterandrous, and cross-pollina- tion appears to be the rule. Bumblebees, honeybees, and some other Hymenoptera, Diptera, and Coleoptera have been recorded as pollinators. Chromosome numbers, known in four species, are invariably 2m = 40. Hybridization (in cultivation, as well as in the wild) between P. quinque- folia and P. vitacea, assumed by Schneider (p. 316), can not be sub- stantiated at present, since no authentic records of such hybrids have been available. The genus seems to be closely related to Ampelopsis, as well as to Ampelocissus and Vitis. A modern revision of the North American species, Suessenguth (p. 312) adopted P. vitacea but made a misleading remark. Having questionably placed Ampelopsis quinquefolia DC. var. vitacea Knerr, Bot. Gaz. 18: 93, in the synonymy of P. vitacea Hitchc., he noted, “fraglich ob hierher gehorig, for it to climb a wall or even a tree unless the bark be very ieee owing to the structure of its tendrils. It climbs more like the grape and the clem ... On ex- amination the tendrils will be found ¥ baad ae like grape tendrils, ait curling and grasping by recurved tips, rather than t, digitate and clinging by disk-like expan- sions as in the case of the typical species rena variety ].” 60 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 based on field studies, with special regard to the morphology of in- florescences and tendrils, is desirable. Both eastern North American species and the eastern Asiatic Partheno- cissus tricuspidata (Sieb. & Zucc.) Planch. (Ampelopsis tricuspidata Sieb. & Zucc., A. Veitchii Hort.), Boston ivy, 2n = 40, are valued ornamentals, used especially for covering arbors, rocks, bushes, and walls. Fruits of at least P. quinquefolia are said to be poisonous. REFERENCES: Under family references see BRANDT, DARWIN (pp. 84-87), LENGERKEN (pp. 359-393, 401-403), KasHyap (1957; 1958), SCHNEIDER (pp. 313-318, 1033), SUESSENGUTH (pp. 309-313), VaTSALA, and ViALA & VERMOREL (1: 62-67. 1910). Carter, A. Notes on pollination. Bot. Gaz. 17: 19-22. 1892. [Parthenocissus — (“Ampelopsis’”), 19, 20 »} Des Moutins, C. Note sur une propriété singuliére des vrilles de la vigne-vierge (Ampelopsi Paes Mich.). Actes Soc. Linn. Bordeaux 24: 105-113. Ln i O. Anpassung-Erscheinungen bei einiger Kletterpflanzen. Abh. aturw. Ver. Bremen 4: 558-560. 1875. [Ampelopsis hederacea var. dumetorum Focke and var. murorum Focke, distinguished by tendril char- acters and climbing habits, 559, 560.] Guieason, H. A. The preservation of well known binomials. Phytologia 2: 201- 212. 1957. [P. vitacea, 204, 205.] Gorr, E. S. The Virginia creeper. Garden Forest 3: 392-394. 1890. [Indicates existence of 2 vars. differing in their tendrils and san habits. See also W.R. Lazensy, Bot. Gaz. 13: 233. 1888, and 15: 233. GRAEBNER, P. Die Benennung der Wilden Wein- Arten unserer ‘Gatien, Garten- flora 57: 59-67. 1908. [See also ibid. 49: 215, 216, etc. : . Die Parthenocissus-Arten. Mitt. Deutsch. Dendrol. Ges. 40: 1-10. 1928. [Seven spp. in N. Am. ] JANCHEN, E. Parthenocissus quinquefolia Eengpan geri unser gewOhnlicher Wilder Wein. Phyton Austria 1: 170- KNerR, E. B. Notes on a variety of FAs quinquefolia. Bot. Gaz. 18: 70, 71. 1893. [A. quinquefolia var. vitacea, with discless tendrils. | MansFELp, R. Zur Nomenklatur der Farn- und Bliitenpflanzen Deutschlands. V. Repert. Sp. Nov. 46: 286-309. 1939. [Parthenocissus, 303.] Moens, P. Ontogenése des vrilles et différenciation des ampoules adhésives chez quelques végétaux (Ampelopsis, Bignonia, Glaziovia). Cellule 57: 371-401. 19 P. tricuspidata (as ““Ampelopsis Veitchii Michx.”), ontogeny of tendrils and differentiation of adhesive discs. Reuper, A. Die amerikanischen Arten der Gattung Parthenocissus. Mitt. Deutsch. Dendrol. Ges. 14: 129-136. 1905. Scoccan, H. J. Flora of Manitoba. Natl. Mus. Canada Bull. 140: 1-619. pls. 1-8. 1957. [Both - quinquefolia and P. vitacea (“P. inserta”) occur in southern Manitoba, 391. ] SovEcEes, R. Embryogénie des Ampelidacées (Vitacées). Développement de Vembryon chez l’Ampelopsis hederacea DC. Compt. Rend. Acad. Sci. Paris 244: 2446-2450. 1957. [Same type as Geum urbanum; embryogenetically both Celastrales and Rhamnales approach Rosales. | 1965 | BRIZICKY, GENERA OF VITACEAE 61 SURINGAR, J. V. Nomenklaturalia. Mitt. Deutsch. Dendrol. Ges. 43: 199-217. 1931. | Parthenocissus, 203-207.] 4. Vitis Linnaeus, Sp. Pl. 1: 202. 1753; Gen. Pl. ed. 5. 95. 1754. Deciduous, rarely evergreen, woody vines; bark of main stem and branches usually shreddy and exfoliating, without distinct lenticels, or rarely close, not exfoliating, with numerous lenticels; pith brown, usually interrupted by diaphragms within the nodes, rarely continuous through the nodes; tendrils usually bifurcate, rarely simple, not forming adhesive discs, intermittent, rarely continuous, Leaves simple [rarely palmately compound], often lobed, dentate. Inflorescences panicles (perhaps thyrses), each opposite a leaf, sometimes with a tendril at the apex of peduncle. Flowers 5-merous, fragrant, both bisexual and (at least functionally) uni- sexual [rarely bisexual only], the plants polygamo-dioecious to tricecious. alyx very small, saucer-like, or obsolete, slightly 5-dentate to entire. Petals 5, coherent at apex, separating only at base, falling off as a cup at anthesis. Stamens 5, straight in ¢ and bisexual flowers, reflexed and with sterile pollen [or exceptionally wanting] in @. Nectariferous disc of 5 flat, + distinct or coherent glands alternate with the stamens. Gynoe- cium rudimentary in ¢ flowers; stigma small; style conical, short; ovary 2-locular. Berries globular [to ellipsoid], 5-25 mm. diameter [or more in cultivars|, purple to black, more rarely whitish, pink, bronze, or green- ish, glaucous or not, 2-locular, 2-4-seeded. Seeds mostly pyriform, + dis- tinctly beaked at base, more rarely obovate to obovate-oblong and beak- less; chalazal knot round to elliptic, mostly in the center of the abaxial side; 2 adaxial grooves deep, rather long, narrow, + curved hag the upper part (Fic. 1, a-c, g). LecrorypE species: V. vinifera L., = 38; see Britton & Brown, Illus. Fl. No. U. S. ed. 2. 2: 505. 1913. sie Latin name of the vine, V. vinifera.) A genus of nearly 60 species, of temperate and tropical regions of the - Northern Hemisphere, centered in China and in eastern North America. Two very distinct subgenera have been recgonized, subg. Vitis, 2m = 38, comprising the vast majority of species, and subg. MuscapINIA, 2n = 40?, including three American species. Morphological differences between the subgenera (see below), coupled with different chromosome numbers and a rather strong genetic barrier, seem to favor the generic segregation of subg. MuscaprntA. Since, however, cytological data are known only in one of the three species of the latter, and the morphological differences can be evaluated only after a thorough study of the related genera, Muscadinia is retained here as a subgenus of Vitis. . Subgenus Vitis (§ Euvitis Planch.). Bark of the main stem and bran- ches shreddy and exfoliating, without distinct lenticels; pith interrupted by diaphragms within the nodes; tendrils forked; infructescences usually elongated, many fruited; seeds usually pyriform, + distinctly beaked at base, smooth on the abaxial side. About 17 species (or perhaps fewer) in 62 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 our area. There is no agreement as to the interrelationships of these; the sequence in which they are mentioned below reflects only increasing area of occurrence, not taxonomic relationship. The species of subg. Vitis often intergrade, and their delimitation appears to be rather difficult; hence the species concepts of various authors differ, and the specific status of some species is still uncertain. A thorough taxonomic and nomenclatural re- vision is needed. Vitis illex Bailey (V. cordifolia var. sempervirens Munson), V. Smalliana Bailey, V. sola Bailey (apparently closely related to and perhaps conspe- cific with V. tiliifolia HBK., of the West Indies, Central America, and northern South America), and V. Shuttleworthii House (V. coriacea Shut- tlew. ex Planch. 1887, not Mig. 1863) are restricted to Florida. Vitis araneosa Le Conte (V. rufotomentosa Small), the specific status of which is uncertain, ranges from Florida to Louisiana and southeastern Virginia. The Mexican-Texan V. Berlandieri Planch., 2m = 38, and V. candicans Engelm., mustang grape, 2” = 38, seem to occur in our area at least in western Arkansas. Vitis Lincecumii Buckley, 2n = 38, of Texas, Okla- homa, Arkansas, Louisiana, and southeastern Missouri, recently has been recorded from Scott County, Tennessee. Vitis Baileyana Munson is found in upland and mountain areas in Alabama and northwestern Georgia, west- ern South and North Carolina, and eastern Tennessee, north to southeast- ern Virginia, West Virginia, and southeastern Kentucky. The taxonomic status of V. Lincecumii and V. Baileyana needs further clarification. While Gleason (New Britt. Brown Illus. Fl. NE. U. S. 2: 517. 1952) suggested that both might be not yet fully stabilized species of hybrid origin, Steyermark (Fl. Missouri 1036, 1037, 1040. 1963) reduced them to the synonymy of V. aestivalis var. aestivalis and V. vulpina, respectively. Vitis palmata Vahl (V. rubra Michx.), 2n = 38, ranges from northeastern Texas, northern Louisiana, and central Mississippi north to western Oklahoma, central Missouri, southern Illinois and southern Indiana, west to southern Kentucky and western Tennessee. Vitis rupestris Scheele, sand-grape, 2m = 38, has a similar range, but does not seem to occur in Mississippi and Louisiana: reports of this species from Pennsylvania, the District of Columbia, and North Carolina appear to be based on intro- ductions and escapes.° Five additional species of subg. Vitis are of wide distribution in eastern North America: Vitis vulpina L. (V. cordifolia Lam.), frost-grape, chicken-grape, 2m = 38; V. Labrusca L., fox-grape, 2n = 38; V. aesti- valis Michx., summer-grape, pigeon-grape; V. cinerea Engelm. ex Millard (including var. cinerea, graybark-grape, 2m = 38, and var. floridana Munson, sometimes regarded as a distinct species, V. Simpsonit Munson) ; and the most widely distributed species of North America, V. riparia Michx. (V. vulpina of authors, not L.), frost-grape or river-bank-grape, °In his Spring Flora of the Dallas-Fort Worth Area, Texas, Shinners stated, “The name of this species is misapplied by Fernald and Rehder to plants extending north through and far beyond our area.” Unfortunately, the basis for this statement was not explaine 1965 | BRIZICKY, GENERA OF VITACEAE 63 which occurs from New Brunswick and Quebec to Manitoba and Mon- tana, south to Virginia, Tennessee, Arkansas, Texas, and New Mexico. Vitis argentifolia Munson (V. aestivalis var. argentifolia (Munson) Fern.: V. bicolor, of authors) has a range similar to that of V. aestivalis and is usually regarded as a variety of it. Since both seem to occur together within nearly the same range, it appears to be more appropriate to regard V. argentifolia either as a form of V. aestivalis or as a distinct species. The latter solution seems preferable until further study clarifies the problem. Subgenus Muscaprnia (Planch.) Rehd. (Muscadinia (Planch.) Small). Bark of main stem and branches close, lenticellate, not exfoliating; pith continuous through the nodes; tendrils simple; infructescences very short, few-fruited; seed obovate to oblong-obovate, scarcely beaked, usually transversely wrinkled on the abaxial side. A strictly American subgenus of two or three species, in the southeastern United States, the West Indies, and Mexico; two species in our area. Vitis rotundifolia Michx. (M. rotundifolia (Michx.) Small), muscadine, southern fox-grape, scuppernong, 2m = 40, occurs in hammocks, swamps, open woods, thickets, and riverbanks, from southern Florida to Texas, northward, mostly on the Coastal Plain, to Virginia, Delaware, south- western West Virginia, southeastern Kentucky, Tennessee, extreme south- eastern Missouri, and Oklahoma. Leaves in this species are distinctive but rather variable; berries are subglobular, 12-25 mm. in diameter, with thick, tough, purple-black to bronze (in the cultivar ‘Scuppernong’) skin and tough, sweet pulp with a characteristic musky flavor. Long aérial roots are often present. This species is highly resistant to virus and fungal diseases and insect pests, sisi to phylloxera and nematodes. Vitis Munsoniana Simpson ex Munson (M. Munsoniana (Simpson ex Munson) Small), closely related to and differing from the preceding mainly in the sour, not musky-flavored, black, persistent berries, 8-15 mm. in diameter, with a tender skin and pulp and smaller seeds, is found in hammocks and scrub, in peninsular Florida and Georgia, on the Florida Keys, and in the Bahamas. The leaves, very similar to those of Vz rotundifolia, are said to be persistent, at least in southern Florida. Vitis Popenoei Fennell (V. vulpina var. yzabalana S. Wats.), of Mexico (Vera- cruz) and Guatemala (Izabal), seems to be a distinctive species eyed related to V. rotundifolia. However, fruits and seeds, cytology, and dis tribution of this species need to be studied. — of V. souendibalin from Mexico apparently are referable to V. Popeno Although the flowers in the genus usually are 5-merous, the number of floral parts (petals and stamens) may vary from three to six (nine), the Regarding the distribution of sexes, the plants may be polygamo-dioecious (androdioecious; rarely gynodioecious in some cultivars) or trioecious, 64 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 more rarely dioecious (with only staminate and functionally pistillate flowers). Cross-pollination seems to be the rule, although in many culti- vated varieties with bisexual flowers self-pollination also occurs. Data on pollinating agents of our species are few, but bees seem to be most important. Chromosome numbers, known in 21 species and numerous diate are invariably 2n = 38 in the species of subg. Vitis, and 2 40 in V rotundifolia. Polyploidy within the genus is rare, and ce = 57 and 76 have been found in a few cultivars of V. vinifera. ‘So it is to be considered that speciation in this genus is achieved by intrachromosomal variations brought about by accumulation of genic differences” (Vatsala). Although wild hybrids have rarely been recorded, the species of subg. Vitis hybridize freely, and numerous, sometimes complex, artificial hybrids between Vitis vinifera and some North American species, as well as be- tween some of the latter, have been produced. There seems to be, how- ever, a genetic barrier between the species of the subgenera. While the cross V. (Muscadinia) rotundifolia 2 V. vinifera 8 has always been unsuccessful, the reciprocal cross resulted in a highly sterile F; (2n = 39). Two such hybrids have recently been made fully fertile by chromosome doubling with colchicine, and a small population of tetraploids has been raised from the artificial amphiploids (cf. Dermen, 1964). A complex triploid (2n = 59) hybrid, (V. Lincecumii « Labrusca vinifera) @ x V. rotundifolia 4, which bears only staminate flowers, has been produced. A close relationship with Ampelocissus and a more distant one with Ampelopsis and Parthenocissus have been assumed by various authors. According to Levadoux, Boubals, and Rives, “In the family of Vitaceae, it is possible to distinguish a comparium containing the genera with 2” = 40 chromosomes [Ampelocissus, Ampelopsis, Parthenocissus, and Vitis]. In this comparium there appears a remarkable exception, the cenospecies Euvitis | Vitis subg. Vitis| to which nearly all the cultivated vines belong, and which is characterized by a chromosome number of 2” = 38. The linnean species, traditionally differentiated within this cenospecies, are ecospecies between which no genetical barrier can be observed.” The genus is of great economic importance for the edible fruits from which raisins, grape juices, and wines are made. The horticultural varie- ties of economic importance have originated from Vitis vinifera (including extremely numerous cultivars), V. Labrusca (e.g., cultivars ‘Concord’, ‘Champion’, ‘Chautauqua’, and others, collectively known as V. labrus- cana Bailey), V. aestivalis, and V. riparia, and from crosses between certain of these species. In the United States, cultivars of V. vinifera are restricted to the Mediterranean-type climate of California; but numerous cultivars derived from hybrids of V. vinifera with cultivars of the hardy American V. Labrusca are widely grown under the climatic conditions of the eastern United States. All of the latter have a characteristic aroma and produce wines quite different from those from V. vinifera. Within the range of V. rotundifolia its cultivars are widely grown for the fresh 1965 | BRIZICKY, GENERA OF VITACEAE 65 fruit used for home and local consumption and in the production of a distinctive wine. REFERENCES: The vast number of references has been reduced here primarily to thos either of general interest or dealing specifically with the southeastern United States. Under family references see especially ADKINSON, ARBAUMONT, BEILLE (pp. 105, 107-109), Bousats (1959; 1961), BRANDT, DarwIN (pp. 79-83), KIRCHHEIMER (pp. 1-9, 118-129), Narr & Mant, SCHNEIDER (pp. 301-312, 1032, 1033), SUESSENGUTH (pp. 283-299; bibliography, pp. 174-179, 283, 284, 296, 297- 299, and 334-337 [cultivars]), TROLL (pp. pote i 628), VATSALA, and VraLa & VERMOREL (especially 1: 109-718. pls. 1-70. Battey, L. H. The species of grapes peculiar to North America. Gent. Herb. 3: 149-244. 1934 Barrett, H. C. Vins cinerea as a source of Spry characters in grape breed- ing. Proc. Am. Soc. Hort. Sci. 70: 165-168. Buncnon, F. Sur la phyllotaxie des grappes . aa vrilles de la vigne (Vitis vinifera L.). Compt. Rend. Acad. Sci. Paris 228: 770, 771. 1949. . Sur la valeur morphologique des grappes et des vrilles de la vigne (Vitis vinifera L.). Ibid. 937-939. [Data from phyllotaxy in support of the thesis of monopodial branching in Vitis. Sur les caractéres distinctifs des rameaux végétatifs, des grappes et des vrilles de la vigne (Vitis vinifera L.). Ibid. 1967-1969. DERMEN, H. Colchiploidy in grapes. Jour. Hered. 45: 159-172. 1954. [Poly- ploidy successfully induced in 10 vars. and selections of bunch grapes, in one of V. vinifera, and 16 muscadine grapes. . Cytogenetics in hybridization of bunch- and muscadine- -type grapes. Econ. Bot. 18: 137-148. 1964. [“Two such hybrids known as ‘N. C. 6-15 and ‘N. C. 6-16’ were made fully fertile by chromosome doubling with colchicine.’’] DetjeN, L. R. Pollination of the rotundifolia grapes. Jour. Elisha Mitchell Sci. Soc. 33: 120-127. 1917 Dorsey, M. J. Variation in the floral structures of Vitis. Bull. Torrey Bot. Club 39: 37-52. pls. 1-3. 1912. [Bisexual flowers occur in the cultivars of many spp. of Vitis; variations in perianth, androecium, and gynoecium. | Ecucut, T., T. Kato, & M. Kome. Flower bud differentiation and pacar of grapes. og Japanese; English summary.) Jour. Hort. Assoc. Japan 2 46-52. Esau, K. iat — in the grapevine, and its seasonal changes. Hilgardia 18: 217-296. 1948 FENNELL, J. L. Two new North American species of Vitis. Jour. Wash. Acad. Sci. 30: 15-19. 1940. [V. Popenoei, 17-19.] ee fe Heprick, U. P. The grapes of New York. Rep. N. Y. Agr. Exp. Sta. Part 2. [N. Y. Dep. Agr. 15 Annual Rep. 3(2).] xv + 564 pp. 101 pls. Albany. 1908. [Beautiful colored plates of some wild spp. and very humerous cultivars. | HeEcepus, A. Conclusions aig Ca apt relatives a la cealhoap histologique de % vigne. Acta Bot. Acad. Sci. Hungar. 6: 257-266. 1960. HENKE, O. Biochemische und ssenpeeielacdia “sana an Vitis-Art- bastarden. Ziichter 30: 213-219. 1960. KRAVCHENKo, L. K. Some data on floral morphology, the flower type, and 66 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 biology of flowering of wild species of the genus Vitis L. (In Russian.) Dokl. Akad. Nauk Uzbek. SSR 7. 1961.* [Twenty-two North American spp. investigated; derivative nature of unisexual and antiquity of bisexual flowers shown (see Vassilczenko, p. 492).] KrocuMat, A., & W. Grierson. Brief history of grape growing in the United States. Econ, Bot. 15: 114-118. 1961. Lattin, G. bE. Zur Vererbung der Anthocyanbildung bei Vitis. Proc. Int. Congr. Genet. 9: 823-827. 1956.* LEVADOUX, L. Les populations sauvages et cultivées de Vitis vinifera L. (Eng- lish summary.) Ann. Amél. Pl. 6: 59-118. 1956. [Morphological and taxonomic data; extensive bibliography. | , D. Boupats, & M. Rives. Le genre Vitis et ses espéces. (English sum- mary.) Ann, Amél. Pl. 12: 19-44. 1962. Loomis, N. H., & C. F. WrittaMs. A new genetic flower type of the muscadine grape. Jour. Hered. 48: 294, 304. 1957. Witiiams, & M. M. Murpny. Inheritance of flower types in muscadine grapes. Proc. Am. Soc. Hort. Sci. 64: 279-283. 1954. [See also N. H. Loomis, ibid. 52: 276-279. 1948; and C. F. Wititams, ibid. 64: 274-278. 1954.] Lutz, H. J. Injuries to trees caused by Celastrus and Vitis. Bull. Torrey Bot. Club 70: 436-439. 1943. McNet1, E. M. New plant finds in southern West Virginia. Castanea 25: 135. 1960. [V. rotundifolia, near Mullens, Wyoming Co. Massey, A. B. Native grapes and their wildlife value. Bull. Va. Polytech. Inst. 38(2): 1-20. 1945. [Includes the apparently erroneous record of the northernmost locality of V. Munsoniana: Smith Island, Cape Fear River, C., 20 A... 20. MeryaniAn, A. S. Uber die Dorsiventralitit der Weinrebe. (Zur Morphologie und Biologie der Gattung Vitis.) Angew. Bot. 12: 470-502. 1930 NEcRUL, A. M. Evolution of the dimensions of grape seeds and berries (based USSR). (In Russian; French summary.) Izv. Timiryazeva Sel’skokhoz. Akad. 33: 167-176. 1960.* OBeRLE, G. D. A genetic study of variations in floral morphology and function in cultivated forms of Vitis. Tech. Bull. N. Y. Agr. Exp. Sta. 250: 1-63. 1939. [Includes review of literature on floral morphology, pollen develop- ment, and inheritance of flower type and sex. Otmo, H. P. The use of seed characters in hee identification of grape varieties. Proc. Am. Soc. Hort. Sci. 40: 305-309. a: Correlations between seed and as development in some seeded varieties of Vitis vinifera. Ibid. 48: 291-297. 1946. . Breeding tetraploid grapes. /bid. 59: ws 290. 1952. [See also ibid. 41: 225-227. 1942.] PaTeL, G. I., & H. P. OLtmo, Cytogenetics of Vitis: I. The hybrid V. vinifera X V. rotundifoha. Am. Jour. Bot. 42: 141-159. 1955. [Includes many references to earlier works on hybridization of V. rotundifolia. | Causes of difference in success in reciprocal crosses between Vitis vi- nifera Linn. and V. rotundifolia Michx. Cytologia 21: 411-416. 1956. [Failure in the cross V. rotundifolia 2 X V. vinifera & “must reside in the cytoplasm of either embryo-sack or/and maternal tissue surrounding the embryo-sack.”’] 1965 | BRIZICKY, GENERA OF VITACEAE 67 . Induction of polyploidy in the sterile F, hybrids of Vitis vinifera Linn. and Vitis rotundifolia Michx. X. Phyton Buenos Aires 7: 63-68. 1956. [ Colchicine treatment. ] Interspecific triploid hybrid in grape. Caryologia 9: 340-352. 1957. Prrovano, A. L’hérédité des caractéres dans " ee entre Vitis vinifera et Vitis Bibewien. Viticult. Nouv. 6(68): 4-6 Popova, N. A. On the first phases of embryogenesis in grapes. (In Russian.) Sadov. Vinograd. Vinodel. Moldavii 11(5): 25~27. 1956.* RrpereEAu-Gayon, P. Les anthocyannes du genre Vise application a la dif- férenciation des vins. Compt. Rend. Acad. Sci. Paris 250: 591-593. 1960. R. J. ANDeRSon, Jour. Biol. Chem, 57: 795-814. 1923, and 61: 685-694. 1924: ANDERSON & NABENHAUR, Jour. Biol. Chem. 61: 97-107. 1924, and Jour. Am. Chem. Soc. 48: 2994. 1926; SCHRINER & ANDERSON, Jour. Biol. sii 80: 743-752. 1928.] ae AuD. Les anthocyannes de la baie dans le genre Vitis. Compt. cae rar “Sci Paris 244: 233-235. 1957. [V. vinifera and 11 N. Am. eae "We Prospection préliminaire des espéces Américaines du genre Vitis. Ann. Amél. PI. 13: 51-82. 1963. [Includes maps by Loomis of distribution of 12 spp. of e. N. A.] Sax, K. Chromosome counts in Vitis and related genera. Proc. Am. Soc. Hort. Sci. 26(1929): 32, 33. 1930. [Undocumented; see also A. M. NEGRUL, Ziichter 2: 33-43. 1930; * and RANDHAWA & Iver, Hort. Advance 4: SMITH, M. B., & H. P. Otmo. The pantothenic acid and riboflavin in the fresh juice of diploid and tetraploid grapes. Am. Jour. Bot. 31: 240, 241. 1944. SNYDER, J. C. Primordial development of the inforescenc of the Concord grape. Proc. Am. Soc. Hort. Sci. 30: 247-252. TKACHENKO, G. V. The role of the substances nee by the stigma in the pollination of the grape (Vitis vinifera L.). (In Russian.) Bot. Zhur. 44: 963-967. TURNER, L. M. Anatomy of aerial roots of Vitis rotundifolia. Bot. Gaz. 96: 367-371. 1934. VaSSILCZENKO, I. T, The recent status of the problem of the origin of cultivated grape (Vitis vinifera L.). (In Russian.) Bot. Zhur. 49: 487-502. 1964. Wacner, E. Uberdauernde Nucleoli und tripolare Spindeln in der Pollenmeiose eines as a Naturwissenschaften 40: 488. 1953. [V. vinifera x V. ripari Weaver, R. J. Pr olonging dormancy in Vitis vinifera with gibberellin. Nature 183: 1198, 1199, 1959. Wittiams, C. F. Hybridization of Vitis degen Inheritance of anatomical stem characteristics. N. C. Agr. Exp. Sta. Tec 1923.* . Relation of berry size to ng type of seedlings in muscadine grape crosses. Proc. Am. Soc. Hort. Sci. 69: 254-260. 1957. 68 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 DATES OF PUBLICATION OF THE JOURNAL LINNAEA: FURTHER EVIDENCE H. E. Moore, Jr., AND C. E. Woon, Jr. FOSTER HAS RECENTLY REVIEWED the problem of determining the dates of publication of the nineteenth-century periodical Linnaea in which a very large number of plants were described and named between 1826 and 1882. His list brought together the partial data published by six previous authors and other evidence adduced from articles in Linnaea itself, from reviews or comments in contemporary publications, and from a few other sources. In spite of these detective efforts, accurate dates have been lacking for many parts and volumes. One of the difficulties in dating this, and a number of other botanical periodicals in which the title-page date of a volume may differ greatly from that of the actual publication of the individual parts, stems from the formerly widespread practice of discarding the covers of the parts when the volumes were bound. Thus, the dates of publication of the parts of only a few volumes of Linnaea have been taken previously from their original covers, these being missing in most sets which have been con- sulted. We were therefore happily surprised in August, 1964, when, having occasion to check a reference to Linnaea in the library of the Royal Botanic Garden, Edinburgh, we found that of the complete run of forty-three volumes, twenty-one (20-28, 32-43) have the original covers of the six or seven component parts (Hefte) bound in at the back. Each cover is dated with month and year and includes a table of contents and sometimes notes which provide additional data about the date of publication of the plates and the index (Register). Since throughout most of its history Linnaea seems to have been behind schedule, these cover dates may reasonably be considered the earliest possible for each Heft and should be taken as the date of publication, unless other evidence is available, as in the case of parts of volumes 42 and 43, the dates of which are given by Van Steenis-Kruseman and Stearn as one to six months later than the cover date. All other cover dates in the list which follows are in general accord with Foster’s list, which in many instances provides the latest possible publication date. To avoid duplication of effort, it should also be mentioned that we examined an incomplete run of Linnaea in the library of the University of Edinburgh and that Moore examined two sets at Oxford University (Rad- cliff Science Library and library of the Botany Department), but in all volumes the original covers are missin For convenience, the earlier references to dates of publication of parts of Linnaea are repeated here: 1965 | MOORE AND WOOD, THE JOURNAL LINNAEA 69 BLAKE, S. F., Proc. Biol. Soc. Washington 46: 192. 1933. [Vol. 28 only; see pt Fl. ‘Melee Bull. 4(17): 907. 1962.] Foster, R. C., Jour. Arnold Arb. 43: 400-409. 1962. Kun7ze, ae Rev Gen, 3(2): 158, 159. 1898. ooreE, H. E., Jr., Gent. Herb. 8: 375. 1954. STEENIS- care M. J. vAN, & W. T. STEARN, FI. Males. I. 4: cxcvm, 1 cxcIx. 1954. SVENSON, H. K., Rhodora 41: 313. 1939. URBAN, ‘ Bot. Jahrb. 19: 562. 1894. Volume 22 TITLE-PAGE DATE: 1849. Heft 1(Bogen 1-8, Taf. I): VOLUME DaTE Volume 20. TITLE-PAGE DATE: 1847. Heft 1(Bogen 1-8, Taf. I*): 1-128 May, 1847 ““) Diese Tafel wird beim nichsten Hefte nachgeliefert.” Heft 2(Bogen 9-16*) : 129-256 June, 1847 ‘“*) Tafel I. wird hiernachgeliefert.”’ Heft 3(Bogen 17-24, Taf. II, III): 257-384 ibe 1847 Heft 4(Bogen 25-32): 385-512 Aug. 1847 Heft 5(Bogen 33-40) : 513-640 Oct. 1847 Heft 6(Bogen 41-50) : 641-781 Dec. 1847 Volume 21. TITLE-PAGE DATE: 1848. Heft 1(Bogen 1-8, Taf. I-III*): 1-128 Feb. 1848 ‘“) Diese drei Tafeln werden mit dem nachsten Hefte ausgegeben werden Heft 2(Bogen 9-16, Taf. IV, V*): 129-256 Apr. 1848 ‘*) Die drei ersten Tafeln kénnen erst mit dem dritten Hefte pocigieegs werden.” Heft 3(Bogen 17—24*) : 257-38 July, 1848 “*’) Tafel I-III ras ae ache” Heft 4(Bogen 25-32): 3 Aug. 1848 Heft 5( Bogen 33-40, oe oe 513-640 Oct. 1848 Heft 6(Bogen 41-49) : 641-780 Jan. 1849 March, 1849 Heft 2(Bogen 9-16, Taf. oy nee e, 1849 Heft 3(Bogen 17-24) : 257-384 July, 1849 Heft 4(Bogen 25-32, Taf. 7 ‘ale 385-512 Aug. 1849 Heft 5(Bogen 33-40): 513 Heft 6(Bogen 41-48, Taf. 5: ne 768 Heft 7( Bogen 49-57) : 769-898 Not ‘dated, en possibly also Dec. 1 Volume 23. TITLE-PAGE DATE: 1850. Heft 1(Bogen 1-8): 1-128 Feb. 1850 Heft 2(Bogen 9-16) : 129-256 May, 1850 70 JOURNAL OF THE ARNOLD ARBORETUM VOLUME Heft 3(Bogen 17-24, Taf. I, II): 257-384 Heft 4(Bogen 25-32): 385-51 2 Heft 5(Bogen 33-40, Taf. III, IV): 513-640 Heft 6(Bogen 41-49) : 641-770 Volume 24. TITLE-PAGE DATE: 1851. Heft 1(Bogen 1-8): 1-128 Heft 2(Bogen 9-16): 129-256 Heft 3(Bogen 17-24): 257-384 Heft 4(Bogen 25-32): 385-512 Heft 5(Bogen 33-40): 513-640 Heft 6( Bogen 41-51): 641-804 Volume 25. TITLE-PAGE DATE: 1852. Heft 1(Bogen 1-8, Taf. 1): 1-128 Heft 2(Bogen 9-16): 129-256 Heft 3(Bogen 17-24): eos Heft 4(Bogen 25-32): 385-5 Heft 5(Bogen 33-40, Taf. D: 513-640 Heft 6(Bogen 41-49) : 641-772 Volume 26 TITLE-PAGE DATE: 1853. Heft 1(Bogen 1-8): 1-128 Heft 2(Bogen 9-16, Taf. I, II): 129-256 Heft 3(Bogen 17-24, Taf. III): 257-384 Heft 4( Bogen 25-32): 385-512 Heft 5(Bogen 33-40) : 513-640 Heft 6(Bogen 41-52): 641-807 Volume 27. TITLE-PAGE DATE: 1854. Heft 1(Bogen 1-8): 1-128 Hefte 2, 3(Bogen 9-24): 129-384 Heft 4(Bogen 25-32): 385-512 Heft 5(Bogen 33-40): 513-640 Heft 6(Bogen 41-51): 641-799 Volume 28. TITLE-PAGE DATE: 1856. Heft 1(Bogen 1-8, Taf. I, II*): 1-128 “*) Die zu diesen: Hefte gehorigen beiden Tafeln werden spater nachgeliefert Heft 2(Bogen 9-16): 129-256 “Die beiden Tafeln zum ersten Hefte werden mit dem dritten ausgegeben werden.” DATE July, 1850 Aug. 1850 Oct. 1850 Jan. 1851 May, 1851 July, 1851 Sept. 1851 Nov. 1851 Jan. 1852 June, 1852 June, 1852 Dec. 1852 Feb. 1853 Apr. 1853 June, 1853 Dec. 1853 Feb. 1854 Apr. 1854 Aug. 1854 Feb. 1855 May, 1855 Sept. 1855 Nov. 1855 Jan. 1856 Feb. 1856 Apr. 1856 Aug. 1856 Aug. 1856 Sept. 1856 [voL. 46 aa cea 2 1965 | MOORE AND WOOD, THE JOURNAL LINNAEA VOLUME Date Heft 3(Bogen 17-24*) : 257-384 Jan. 1857 ‘“*’) Mit diesem Heft wird Taf. I. u. II. nt Heft 4(Bogen 25-32): 385-512 June, 1857 Heft 5(Bogen 33-40) : 513-640 Aug. 1857 Heft 6( Bogen 41-49) : 641-767* Feb. 1858 *Cover of Heft 6 notes: ‘“‘Das erste Heft des 29sten Bandes der Linnaea ist unter der Presse.” Volumes 29-31. Cover-pages lacking. Volume 32. TITLE-PAGE DATE: par Heft 1 (Bogen 1-8): Heft 2( Bogen 9-16, aa. : II): 129-256 Heft 3(Bogen 17-24, Taf. III): 257-384 Hefte 4, 5( Bogen 25-40): 385-640 Heft 6(Bogen 41-49) : 641-784 Register pages 786-801 published with Volume 33, Heft I Volume 33. ITLE-PAGE DATE: 1864 u. 1865. Heft 1(Bogen 1-8): 1-128 Heft 2( Bogen 9-16): 129-256 Hefte 3, 4(Bogen 17-32, Taf. I, II): 257-512 Heft 5(Bogen 33-40) : 513-640 Heft 6(Bogen 41-49, Taf. III) : 641-770 Volume 34. TITLE-PAGE DATE: 1865 Ay 1866. Heft 1(Bogen 1-8): 1-12 Heft 2( Bogen 9-16): aoe Heft 3( Bogen 17-24): 257-384 Heft 4(Bogen 25-32) : 385-512 Heft 5(Bogen 33-40): 513-640 Heft 6(Bogen 41-47): 641-752 Volume 35. TITLE-PAGE DATE: 1867 u. 1868. Heft 1(Bogen 1-8, Taf. I, I): 1-128 Heft 2(Bogen 9-16): 129-256 Hefte 3, 4(Bogen 17-24, Taf. III-XXII*) : 257-384 *There are 2 plates marked XXII, one a habit of Acacia tortilis, the other con- taining diagnostic features of Acacia verek and A. glaucophylla. Heft 5( Bogen 25-32): 385-512 Heft 6( Bogen 33-40) : 513-637 March, 1863 Feb. 1864 May, 1864 May, 1864 Aug. 1864 Dec. 1864 March, 1865 June, 1865 March, 1865 July, 1865 Sept. 1865 Dec. 1865 Feb. 1866 Dec. 1866 Apr. 1867 July, 1867 Dec. 1867 March, 1868 Nov. 1868 72 JOURNAL OF THE ARNOLD ARBORETUM VOLUME Volume 36. TITLE-PAGE DATE: 1869 u. 1870. Heft 1(Bogen 1-8): 1-128 Heft 2(Bogen 9-16): 129-256 Heft 3(Bogen 17-24): 257-384 Heft 4(Bogen 25-32): 385-512 Heft 5(Bogen 33-40) : 513-640 Heft 6(Bogen 41-50) : 641-790 Volume 37. TITLE-PAGE DATE: 1871 bis 1873. Heft 1(Bogen 1-8): 1-128 Heft 2(Bogen 9-16): 129-256 Heft 3(Bogen 17-25): 257-416 Hefte 4, 5(Bogen 26-34, Taf. 1): 417-544 Heft 6 (Bogen 35-42): 545-663 Volume 38. TITLE-PAGE DATE: 1874. Heft 1(Bogen 1-8): 1-128 Heft 2(Bogen 9-16) : 129-256 Heft 3(Bogen 17-24): 257-384 Heft 4(Bogen 25-32): 385-512 Heft 5(Bogen 33-40) : 513-640 Heft 6(Bogen 41-48) : 641-753 Volume 39. TITLE-PAGE DATE: 187 2 Heft 1(Bogen 1-8): 1-12 Hefte 2, 3(Bogen 9-16, Tat i 129-256 Heft 4 (Bogen 17-22): 25 Heft 5(Bogen 23-28): rele Heft 6(Bogen 29-33): 449-526 Volume 40. TITLE-PAGE DATE: cap Heft 1(Bogen 1-6): 1-9 Hefte 2, 3(Bogen 7-12, ad a re 97-192 Heft 4(Bogen 13-18): 193 Heft 5(Bogen 19-24): oe Heft 6(Bogen 25-30): 385-468 Volume 41. TITLE-PAGE DATE: 1877. Heft 1(Bogen 1-7): 1-112 Heft 2(Bogen 8-12): 113-192 Heft 3(Bogen 13-18): 193-288 Heft 4(Bogen 19-24): 289-384 Hefte 5, 6(Bogen 25-36) : 385-576 Heft 7(Bogen 37-41): 577-655 DATE May, 1869 Sept. 1869 Jan. 1870 Apr. 1870 Sept. 1870 Dec. 1870 Feb. 1871 Sept. 1873 Dec. 1873 Jan. 1874 May, 1874 July, 1874 Aug. 1874 Dec. 1874 Feb. 1875 June, 1875 July, 1875 Aug. 1875 Oct. 1875 Jan. 1876 Dec. 1876 Dec. 1876 Feb. 1877 Apr. 1877 Aug. 1877 Oct. 1877 Feb. 1878 [voL. 46 1965 | MOORE AND WOOD, THE JOURNAL LINNAEA VOLUME Volume 42. TITLE-PAGE DATE: 1878 u. 1879. Heft 1(Bogen 1-7): 1-112 Heft 2(Bogen 8-12): 113-192 Heft 3(Bogen 13-18): 193-288 Heft 4( Bogen 19-24): 289-384 Heft 5(Bogen 25-30) : 385-480 Hefte 6, 7( Bogen 31-42) : 481-667 Volume 43. TITLE-PAGE DATE: 1880 bis 1882. Heft 1(Bogen 1-4, Taf. . 1-66 Heft 2(Bogen 5- 9): 67-13 Hefte 3, 4(Bogen 10-16, ae er 139-252 Hefte 5, 6(Bogen 17-31): 253 Heft 7(Bogen 32-36): ee i-xxiii BarLey Hortorrum, CORNELL UNIVERSITY N ARNOLD ARBORETUM, HARVARD UNIVERSITY DATE Feb. 1878 Oct. 1879 July, 1880 74 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 COMPARATIVE ANATOMY OF THE LEAF-BEARING CACTACEAE, XIII THE OCCURRENCE OF WATER-SOLUBLE ANISOTROPIC BODIES IN AIR-DRIED AND ALCOHOL-DEHYDRATED LEAVES OF PERESKIA AND PERESKIOPSIS I. W. Batrey ! IN A PRELIMINARY SURVEY of the occurrence of crystals in the leaf- bearing Cactaceae (Bailey 1961), I noted the common presence in various species of Pereskia of brown isotropic bodies during earlier stages of the clearing of leaves in three per cent sodium hydroxide. Although isotropic in polarized light, the radial striations and occasional concentricities of these bodies superficially resembled those that occur in certain ‘‘sphero- crystals’ of calcium oxalate reported upon in the Cactaceae by Mobius (1885). They differed fundamentally, however, from the abundantly occurring druses and other crystalline forms of calcium oxalate in the family which are insoluble and retain their anisotropy in dilute solutions of sodium hydroxide. I concluded that the chemical composition and the factors involved in the formation of the isotropic bodies merited detailed investigation. Having obtained more extensive collections of Pereskia, I now find that of 14 putative species of the genus all exhibit a tendency to form more or less numerous bodies comparable to those previously reported upon (Bailey 1961). This is true in the case of leaves from herbarium speci- mens of 11 species collected from plants growing in their native habitats and, likewise, in the case of leaves of 14 species preserved in formalin acetic alcohol (FAA).? Furthermore, the normal living leaves of eight species, although they have formed abundant druses of calcium oxalate, do not contain such bodies, but tend to develop them when the leaves are air dried or are transferred to alcohol varying in dilutions of from 95 to 50 per cent. In the ordinary white light of a microscope the crystalline contents of the bodies are colorless. In polarized light they are as strikingly bire- fringent as the druses of calcium oxalate (Fic. 1). When dehydrated rapidly lose their birefringence and commonly turn brown (Fic. 2). Upon prolonged treatment in alkali, the isotropic residues of the aniso- * This investigation was supported by a grant from the National Science Founda- tion. I am indebted to the American Philosophical Society for the loan of a Wild ga soy * Forty per cent pao hn five parts: glacial acetic acid, five parts: 50 per cent ethyl Lobe ninety part 1965 | BAILEY, LEAF-BEARING CACTACEAE, XIII 75 tropic bodies ultimately lose their striated and other internal structural features and dissolve (Bailey 1961, figs. 30-33). This behavior is in marked contrast to that of druses and small individual units of calcium oxalate which, in the same leaves, remain insoluble and retain their anisotropy (Fic. 2). That the birefringent bodies are not composed of calcium oxalate is indicated, not only by their behavior when transferred to dilute sodium hydroxide, but also by their solubility in water at ordinary room temperatures. In Pereskia, the size, form, and abundance of the water-soluble bire- fringent bodies varies more or less extensively in different collections of a taxon, in different leaves of the same plant, and in different layers of the same leaf. The larger more conspicuous bodies commonly are composed of a group of adjacent cells filled with compact birefringent contents which frequently exhibit fine striations oriented radially toward the center of the body (Fic. 5). Smaller bodies of spherical or angular form have a similar compact crystalline composition, but fewer adjacent cells are con- cerned in their development. In some cases, single large independently occurring cells of the mesophyll are not completely filled with crystalline contents, but may contain one or more small, spherical, radially striated bodies or hemispherical striated ones attached to the walls of the cell. When many diversely dehydrated leaves of Pereskia are examined, some of them contain, in addition to larger bodies, minute water-soluble, bire- fringent units diffused throughout parts of the mesophyll (Fic. 4). These units vary markedly in crystalline form from rotund to rectangular, dia- mond-shaped, slender boat-shaped, and thin angular platelets. Some of the larger ones appear to arise by fusion of smaller units forming rings or spheres with or without isotropic interiors. These birefringent units, not only are water soluble, but also lose their anisotropy and turn brown in dilute sodium hydroxide. At times, certain crystalline forms of the smaller units tend to be aggre- gated in varying degrees of compactness. For example, as illustrated in Fic. 3, single diamond-shaped crystals may be surrounded by a layer of aggregated crystals of different form. In other cases, groups of adjacent cells contain aggregations of small birefringent units in varying degrees of compactness. There appear to be transitional forms of crystallization be- tween such structures and the commonly occurring extreme form illus- trated in Fic. 5. At times, large bodies of the latter general form, having compact radially striated crystalline interiors, are jacketed by flat, more or less rectangular, birefringent units which are loosely aggregated and hap- hazardly oriented. I have not encountered slender, much elongated, needle-like crystals occurring independently in Pereskia. However, particularly when living leaves are dehydrated in 95% alcohol, aggregations of such crystals may be formed in a few adjacent cells subtending stomata (Fic. 6). Where the elongated crystals are oriented parallel to one another, the aggregation may superficially resemble raphides of calcium oxalate, but the fascicles are not confined to single enlarged cells (i.e., idioblasts), are not embedded in 76 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 mucilage, and are water soluble. Very infrequently one encounters a radially oriented aggregation of such needle-like crystals in a single un- usually large cell of the mesophyll (Fic. 7). It is evident from observations recorded in preceding paragraphs that the birefringent structures induced during dehydration of the leaves of Pereskia exhibit more extensive variability in crystalline form than do druses of calcium oxalate persisting from the living condition in the same leaves. However, it is significant that, in spite of the diversity in the induced forms of crystallization, all of them are characterized by their solubility in water and by their similar behavior when transferred to dilute sodium hydroxide. It is not obvious at present why some dehydrated col- lections of a taxon or clone form the bodies, whereas others do not; why some leaves of a shoot contain them, when adjacent leaves are without them; and why some parts of a leaf have them, when other parts are devoid of them. Nor is it evident what factors are concerned in producing the variability in forms of crystallization. More detailed and extensive investigations, particularly adequate chemical analyses, starting with nor- mal living leaves are highly desirable for elucidating such uncertainties. In 36 collections of Pereskiopsis of varying taxonomic affinities, kindly collected and preserved for me in FAA by Boke, Moran, Kimnach, and others, a majority of 20 has leaves and occasionally young stems containing water-soluble birefringent bodies of varying abundance, sizes, and diver- sified crystalline forms. The largest of these bodies, up to and exceeding 600 microns in diameter, are much larger than the largest of those that occur in Pereskia (compare the contrasting sizes at the same magnification in Fics. 1, 8, & 9). In addition, the larger birefringent bodies of Pereskiop- sis differ markedly in their internal crystalline composition from those of Pereskia. In Pereskiopsis, the larger bodies of multicellular origin tend to be rings or disks of varying thickness oriented periclinally in relation to the external surfaces of leaves and young stems. They are readily detectable without magnification, appearing superficially as embossed white pustules in the surfaces of leaves and stems which shrink during dehydration in FAA. In some of these bodies, all of the cells in the interior of the body are devoid of crystalline content, except for cells which contained druses of calcium oxalate in the living leaf or stem. In such a body, the isotropic core is jacketed by an anisotropic layer of varying thickness which appears as a birefringent ring in polarized light. More frequently the bodies have a cell with water-soluble birefringent contents at its center surrounded by an isotropic zone (which may contain water-insoluble druses of calcium oxalate) and, in turn, by an outer birefringent one having more or less numerous outwardly projecting bulges (Fic. 9). Less frequently most cells in the interior of a body may contain loosely aggregated water-soluble crystals (Fic. 8). The individual crystals in these large birefringent bodies vary markedly in size, form, degree of aggregation, and orientation. Particularly in the outer birefringent zone of the bodies, many of the constituent cells exhibit 1965 | BAILEY, LEAF-BEARING CACTACEAE, XIII 77 a conspicuous tendency to contain slender, elongated, more or less needle- like crystals oriented radially toward the center of the body. Similar forms of crystallization and orientation may be present in the central cell of a body. In the case of the large outwardly bulging cells illustrated in Fic. 9, some of them may have needle-like, radially oriented crystals in their inner or outer parts, whereas remaining parts of the cell contain minute widely diffused birefringent specks resembling “crystal sand.” : Si . Text-Fics. A-E. Crystalline bodies in large cells of mesophy ll of nccinge: ar as seen in nonpolarized light, 300. A-D, Pereskiopsis aft. rotundifolia (B B-1): A, radially oriented needle- like crystals i in plasmolyzed cell; B, c ead form at the ge concealed by black specks; C, concentric layering of needle- like crystals; four centers of cryst aie in a large cell. E, Pereskiopsis porteri (Kicanach 76): eens brwetine needle-like serials extending outwardly into adjacent cells of the mesophyll. Drawn by Elmer W. Smith. Some dehydrated collections of Pereskiopsis contain smaller, more nearly spherical bodies. Some of these, when viewed in optical section (Fic. 10), resemble the larger body illustrated in Fic. 9. A single cell with birefringent contents at the center of the body is surrounded by an iso- tropic layer and, in turn, by an outer anisotropic one. In other cases, the bodies are birefringent throughout, their central part exhibiting very narrow anisotropic concentricities partly concealed in Fic. 11 by four positions of extinction when viewed in polarized light. The crystalline contents of the cell in the center of such bodies may be deposited at times in similar con- 78 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 centricities (Fic. 10), whereas in other cases the central cell contains larger elongated units resembling those that occur in cells of the external parts of the bodies (Fic. 11). It is significant in this connection, however, that the longer axis of the crystalline units — regardless of extreme varia- tions in size and form — tend to be radially oriented toward the center of the bodies. It should be noted, in addition, that cells, particularly in the interior of the larger bodies, not infrequently contain black granular con- tents which tend to obscure internal crystalline configuration when the bodies are viewed in polarized light. Some of Dr. Boke’s collections of Pereskiopsis preserved in FAA in their native habitats (particularly in his B-1, B-18, B-20, B-22, and B-31, with putative affinities to P. aquosa, P. chapistle, P. rotundifolia, or P. spathu- lata) contain, in addition to multicellular bodies, large independently dis- tributed cells with water-soluble crystalline contents. Some of these cells (Trext-ric. A) have slender, elongated, radially oriented units resembling the configuration that occurs so infrequently in Pereskia (Fic. 7). Occa- sionally a large cell may contain several bodies of similar crystalline form (Text-Fic. D). The contents of other cells of the same leaf (TEXT-FIG. C) commonly exhibit conspicuous concentricities, due apparently to varia- tions in the size or form of the crystalline units in successive zones of their deposition. In some cases the longer needle-like crystals of the external zone extend outwardly into parts of adjacent cells forming expanded bodies of more or less conspicuously lobed or angular forms (Text-Fic. E). At times, the inner parts of some of the cells contain numerous black isotropic specks (TextT-Fic. B) which tend to obscure the anisotropy in the interior of concentric configurations. DISCUSSION AND CONCLUSIONS The water soluble birefringent bodies of Pereskia and Pereskiopsis are formed during dehydration of the “sap” of living cells. In Pereskia, the larger bodies are remarkably similar regardless of whether leaves are air dried, as in the case of herbarium specimens, immersed in 95%, 70%, and 50% ethyl alcohol, or preserved in FAA. In none of my numerous collec- tions of different species of the genus do larger crystalline bodies compara- ble to those of Pereskiopsis occur. Although highly variable, the forms of crystallization differ so consistently in Pereskia and Pereskiopsis as to provide, even in herbarium specimens, an additional criterion for separating the genera taxonomically. Young green stems in some of my collections of Opuntia (preserved in FAA) contain water soluble birefringent bodies of sizes and crystalline forms resembling those that occur in Pereskiopsis, suggestive of possible taxonomic relationship. Furthermore, the ‘“spherocrystals” in Echino- cactus, Mamillaria, and Anhalonium observed and briefly discussed by Lauterbach (1889) and Michaélis (1896) were found in tissues treated with alcohol. It now appears that these anisotropic bodies, unlike normally occurring druses and other forms of calcium oxalate, belong in a special 1965 | BAILEY, LEAF-BEARING CACTACEAE, XIII 79 water soluble category. The occurrence of such bodies in all three sub- families or tribes of the Cactaceae at least suggests that a tendency for their formation during dehydration of the “sap” of living cells may be widely distributed throughout this characteristically xerophytic family. Since the publication of Schleiden’s (1845) classical paper, the Cactaceae have proved to be a family having extraordinarily abundant druses and other water insoluble crystalline forms of calcium oxalate in their tissues, indicative, in turn, of the production of excessively large amounts of oxalic acid in their metabolism. It is significant in this connection that in living cacti, crystals of calcium oxalate occur in the mesophyll of leaves (where present) and in parenchymatous cells of the pith, cortex, xylem, and phloem of both stems and roots. In contrast to this, the water soluble crystals formed during dehydration of “sap” in Pereskia and Pereskiopsis occur primarily in the chlorenchyma of leaves and in the cortex of green stems which function in photosynthesis. Unfortunately, owing to numerous complexities and uncertainties, it does not appear possible at present to determine the exact chemical composition of the water soluble bodies merely by their forms of crystallization. For example, upon the basis of superficial resemblances in forms of crystalliza- tion, it has been suggested by Lauterbach (1889) and others that the bire- fringent bodies in dehydrated tissue of the Cactaceae are composed of inulin or related substances. Such is not the case in collections available to me where tests for such composition of the crystalline bodies are negative. That calcium is involved in the chemical composition of the larger birefringent bodies of Pereskiopsis may be demonstrated by treatment in sulphuric acid. When transferred to this acid the birefringence of the bodies is rapidly reduced, but is quickly restored by the formation of crystals of calcium sulphate. The transition to calcium sulphate tends to occur more rapidly than in druses of calcium oxalate present in the same tissue. It is of interest in these connections that similar forms of crystallization in alcohol have been encountered in certain representatives of other fami- lies of angiosperms, viz. cactus-like euphorbias, Mesembryanthemum (Aizoaceae), Basella (Basellaceae), Ceropegia and Stapelia (Asclepia- daceae), Galtonia (Liliaceae), and Nolana (Nolanaceae). Belzung (1893), Belzung and Poirault (1892) and Mirande (1898), following the earlier investigations of Hansen (1888), Leitgeb (1888) and others, concluded upon the basis of extensive microchemical and other tests that calcium malate and calcium “malophosphate” (?) are concerned in such plants during the formation of water soluble crystals when their tissues are de- hydrated in alcohol. A similar conclusion was reached subsequently by Kean (1931) in his investigation of Mesembryanthemum. ; Although these investigations are suggestive rather than conclusive from modern chemical points of view, they do indicate that the water soluble crystals are in all probability calcium salts of organic acids, occurring in close association with calcium phosphate. It is unlikely that the crystals are composed solely of calcium phosphate, as suggested by earlier investi- 80 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 gators, since such crystals are almost insoluble in water as are the crystals of calcium oxalate. That dehydration in alcohol is not essential for their formation is clearly indicated by their typical occurrence in the dehydrated cellular ‘‘sap” of air dried leaves, for example in herbarium specimens. At present, cumulative circumstantial evidence from these phylogeneti- cally highly specialized genera, in correlation with the behavior in the characteristically xerophytic Cactaceae, raises a fundamentally significant question. Are there evolutionary changes in the metabolism of angiosperms (viz. those having excessive succulence) which lead to the formation of unusually large amounts of such organic acids as oxalic, malic, etc.? The phenomena are highly complex and variable and involve various uncertain- ties. An adequate answer to the question will involve active and sustained cooperation between botanists and chemists. Preliminary observations on the only normal living leaves of Pereskiop- sis available to me thus far (from Dr. Boke’s culture of his B-78) suggest that this genus may provide more favorable material for studying succes- sive stages in the formation of birefringent bodies during dehydration than does Pereskia. A more detailed discussion of the phenomena will be under- taken when living leaves and young stems of other species of Pereskiopsis become available. LITERATURE CITED BartLey, I. W. 1961. Comparative anatomy of the leaf-bearing Cactaceae, III. orm and distribution of crystals in Pereskia, Pereskiopsis and Quiabentia. Jour. Arnold Arb. 42: 334-346. BetzunG, E. 1893. Nature des sphérocristaux des Euphorbes cactiformes. Jour. Bot. Morot 7: 221-229, 261-267. . Porrautt. 1892. Sur les sels de l’Angiopteris evecta et en parti- culier is malate neutre de calcium. Jour. Bot. Morot 6: 286-298. Hansen, A. 1888. Uber Spharokrystalle. Arbeit. Bot. Inst. Wiirzburg 3; 92-122. Kean, C. I. 1931. Artificial Shier Me feat tissues of Mesembryanthe- mum. Trans. Bot. Soc. Edinb. 3 LAUTERBACH, H. 1889. ecunaeuse ibe Bau und Entwicklung der Sekret- behalter bei den Cacteen. Bot. Centralbl. 37: 257-264, 289-297, 329-336, 369-375, 409-413. Leitces, H. 1888. Uber sphirite. Mitt. Bot. Inst. Graz 1: 257-260. MicwHaéE.is, P. 1896. Beitrage zur vergleichenden Anatomie der Gattungen Echinocactus, Mamillaria und Anhalonium. Doctoral Thesis. Halle. Mrranpe, M. 1898. Contribution 4 l’étude du malate de neutre calcium et du ome garg de calcium dans les végétaux. Jour. Bot. Morot 12: 6-12, 32-42, 58- Mostus, M. eae. “se as aera von Kalkoxalate bei Cacteen. Ber. Deutsch. ScCHLEIDEN, M. J. 18 - ‘Beitrige zur Anatomie der Cacteen. Mém. Acad. Imp. Sci. St. '-Pétersbourg 4: 335-380. eee a eS ae 1965] BAILEY, LEAF-BEARING CACTACEAE, XIII 81 EXPLANATION OF PLATES PLATE I Fics. 1-4. Parts of leaves of hone photographed in polarized light. 1, Pereskia aculeata, showing birefringence of large water-soluble yer and anisotropy of small druses of sain "ouillate in cytes ted leaf, aculeata, showing loss = birefringence by the water-soluble bodies au retention of anisotropy by druses of calcium oxalate after treatment in dilute sodium by aggregations of small crystals of different form, x 180; 4, P. corrugata, show- ing diffusely distributed minute water-soluble crystals, x 114 PLATE II Fics. 5-7. ga ee birefringent bodies of Pereskia photographed i in non polarized light, scone 7, P. pititache showing radially oriented needle-like crystals in a single large c ell. PLATE III Fics. 8 & 9. Large page Benin birefringent bodies of ern (Boke B-22) pr cbc in FAA, photog aphed in polarized light, X 114; 8, crystals sei throughout the body; 9 ‘ body having a wen at the center with birefringent ontents, surrounded by an isotropic layer (except for druses of calcium oxalate ee from the living leaf) and jacketed eieunally by a birefringent zone. PLATE IV Smaller more nearly spherical water-soluble poate gira of Pe ae “photographed in optical section in polarized light. From Boke’s collection B-20 aii in FAA, X 330; 11, from Boke’s prise Be B-1 preserved in FAA. x 4 Jour. ARNOLD Arps. VOL. 46 PLaTE I * © i: BAILEY, LEAF-BEARING CACTACEAE, XIII : 8 Jour. ARNOLD Ars. VoL. 46 PraTeE II BarLey, LEAF-BEARING CACTACEAE, XIII Jour. ARNOLD ARB. VOL. 46 Prate III BaILEY, LEAF-BEARING CACTACEAE, XIII OUR. ARNOLD ARB : } Jour. ARNOLD ARE bi PLATE IV BAILEY, LEAF-BEARING CACTACEAE, XIII 86 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 A NEW CUSCUTA FROM NEPAL M. L. BANERJI AND SITESH Das DuRING THE MONTHS OF SEPTEMBER AND OcTOBER, 1960, while the senior author was carrying on botanical exploration in East Nepal, a sizeable collection was made which included a conspicuous Cuscuta with reddish stem and dark brown flowers, parasitic on Hypericum patulum, Banerji 1199. On examination in the herbarium this plant did not agree well with any of the known species of Cuscuta. For identification of the material constant reference to the monograph of the genus by Yuncker (1932) and to Santapau and Patel’s (1957) work on the Bombay Cuscutas has been made. Because of the presence of a single style on the gynoecium, very careful study has been made of Cuscuta reflexa Roxb., both of mate- rial in the Central National Herbarium, Calcutta, and of literature available in the Indian floras. By kindness of Dr. S. K. Mukerjee, Keeper of the Central National Herbarium, Calcutta, Wallich’s and Hooker’s sheets, mentioned by Yuncker (loc. cit.) under reference to C. reflexa Roxb., have also been examined, and flowers from these sheets have been dissected and studied. As the type of C. reflexa is not definitely known, Roxburgh’s Icones have been referred to, as well as Van Ooststroom’s treatment of the genus Cuscuta in Flora Malesiana (1955). Our conclusion is that the material collected from East Nepal is definitely not C. reflexa Roxb. Amongst the Convolvulaceae collected by Zimmerman in 1954 in Nepal, very kindly sent by the Director, Conservatoire et Jardin Botaniques, Genéve, to the senior author for determination, is a Cuscuta, Zimmerman 1302, which is exactly the same as Banerji 1199. In the Central National Herbarium, Calcutta, is a sheet, Burkill 35707, from Dibrugarh (Assam) which, although named Cuscuta reflexa Roxb., is quite different from other sheets so determined. Further, this specimen of Burkill’s superficially resembles the present materials from Nepal in characters of calyx, size of corolla, and position of the anther lobes. It is pertinent to mention that on this sheet (Burkill 35707) is a sketch of the characteristic globose-shaped gynoecium. A close examination of the flowers revealed a gynoecium with a prominent style 0.5—0.75 mm. long, in both the Nepal specimens and in Burkill’s material. The infrastaminal appendages have also been found to differ from those of C. reflexa Roxb. A study of the pollen grains, made by the junior author, showed marked differences, too. Pollen from Wallich’s and Hooker’s materials has been found to measure 27.2, while the pollen from the Nepal materials meas- ures 30.4y in diameter in polar view. With such marked differences from C. reflexa Roxb., it seems best to treat these materials collected from East Nepal and Assam as a distinct new species. 1965 | BANERJI AND DAS, CUSCUTA FROM NEPAL 87 Cuscuta santapaui Banerji & Das, sp. nov. Caules sat crassi et asperi, rubreoli. Flores 6-7 mm. longi, breviter pedicellati, bracteati, in fasciculos cymosos paucifloros aggregati; pedicelli 1.5—-1.75 mm. longi, bracteis minutis. Calyx cupulatus, plus minusve urceolatus, carnosus infra, membranaceus ad margines, laciniis alte divisis late deltoideis, 1.4—1.6 mm. latis, 1.25—1.5 mm. longis, altera alteri incum- benti, subaequalibus, externis quidem duabus minoribus. Corolla alba reticulationibus alte brunneis ornata, carnosa, ca. 5.5—6.5 mm. longa, cylin- drica; corollae tubus 3—3.5 mm. diametro ad anthesim, laciniis late lanceo- latis, 1.5-2 mm. longis, mucronatis ad apicem, patentibus. Appendices infrastaminales ad trientem tubi corollin longae, ex oblongis obovatae, in parte superiore incurvae, 2—2.5 mm. longae et 1 mm. latae; marginibus irregulariter fimbriatis, fimbriis etiam ad pontes visis. Stamina corollae lobis breviora; antherae integrae 1.25 mm. longae, 0.5 mm. latae; filamenta brevissima, dorsi xa. Ovarium globosum, vulgo 2.75-3 mm. diametro; stylus 1, crassus, 0.5-0.75 mm. longus, sulco uno indistincto per totam longitudinem insignitus, supportans stigmata bina linearia 1.5—1.75 mm. longa ad basin saccata. Fructus capsula circumscissilis; semina 4, paulum oblonga, singula ca. 0.5 mm. diametro. Hootyrus, Banerji 1199, lectus in via Nayapati ad Risingo in Nepalia ad altitudinem ca. 2500 m., positus in herbario Calcuttensi (caL); 1soTypr subeo- dem numero positi in herbario Arnold Arboretum (A), et in herbario universitatis Floridae australis (usF), in U. S. A., in herbario Bot. Surv. Nep. in Nepalia et in herbario universitatis Kalyanensis in India. Typi lecti sunt die 8 mensis octobris anni 1960 inficientes Hypericum patulum. Paratypus, Burkill 35707, ad Dibrugarh in Assamia lectus die 18 mensis novembris anni 1911 inficiens Callicarpam sp. positus in herbario Calcuttensi (cAL). Paratypus alter, immermann 1302, lectus in Nepalia “de Simigaon au col,” ad 2150 ee ol die 15 septembris 1954, et positus in herbario Hortus Botanici Genevensis (G) et in herbario universitatis Kalyanensis prope Calcuttam in India. This new species of Cuscuta resembles C. reflexa Roxb. in general ap- pearance, although differing markedly from it. The calyx is shallower and more or less urceolate; the corolla tube smaller; the infrastaminal appen- dages more deeply fimbriate and bearing fimbriae at the bridges. The ovary is globose with a clear or distinct style bearing two long and linear stigmas saccate at their bases; the seeds are minute. To contrast the characters of the new species with those of C. reflexa sei illustrations H. Santapau, now Director, Botanical Survey of India, who initiated the senior author in taxonomic studies and whose contributions in the field of taxonomy in India are enorm The authors thank Father a for the Latin translation of the 88 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 1965 | BANERJI AND DAS, CUSCUTA FROM NEPAL 89 diagnosis of the new taxon and Dr. S. K. Mukerjee for the facilities granted. The junior author is grateful to the University of Kalyani for the award of a scholarship during the tenure of which this work was car- ried out. REFERENCES CrarkeE, C. B. 1883. Cuscuta. In: Hooxer f., Fl. Brit. Ind. 4: 225-228. OoststRoom, S. J. vAN. 1938. Cuscuta. In: Blumea 3: 68-72. 1953. Cuscuta. In: C. G. G. J. van STeEnts, Fl. Males. I. 4(4): 391-394. SANTAPAU, H. & V. PaTteL, 1957. The genus Cuscuta in Bombay. Jour. Bombay Nat. Hist. Soc. 54: 707-713 Yuncker, T. G. 1932. The genus Cuscuta. Mem. Torrey Club 18: 109-331. INDIAN Arp Mission INDIAN Empassy, NEPAL AND UNIVERSITY OF KALYANI KALYANI, INDIA i. , C. santapaui: A, flower; B, calyx; C, interior of split te eos pun ‘intrstaminal scales have been ga in outline abe D, infrastaminal scales; E, stam and G, gynoecia; H, s a-g, C.r : a, ‘flower: b, calyx; c _ interior ae ee and flattened corolla “Gnfestaminal iio have been shown in outline only); d, infrastaminal scales; e, stamen; I and g w gynoecia. (All figures of C. panes drawn from Hooker’s material. VOLUME 46 NuMBER 2 JOURNAL OF THE ARNOLD ARBORETUM HARVARD UNIVERSITY B. G. SCHUBERT EDITOR L. I. NEVLING, JR. C. E. WOOD, JR. LAZELLA SCHWARTEN CIRCULATION APRIL, 1965 PUBLISHED BY THE ARNOLD ARBORETUM OF HARVARD UNIVERSITY CAMBRIDGE, MASSACHUSETTS PS O08! Boran jeune APR OT tons THE JOURNAL OF THE ARNOLD ARBORETUM Published quarterly by the Arnold Arboretum of Harvard University. Subscription price $10.00 per year. During 1965 volumes I-XX will be reprinted and available from the Kraus 10017 REPRINT CorporaTion, 16 East 467TH Srreet, New York, N.Y. 1 cryin and remittances should be addressed to Mrs. Lazenua WARTEN, ARNOLD ArBorETUM, 22 Diviniry AvENUE, CamBripcs, Massa- 138. engi 02 CONTENTS OF NO. 2 Tue GENERA OF POLYGONACEAE IN THE SOUTHEASTERN UNITED States. Shirley A. Graham and C. E. Wood, Jr. «0.0.0.0... Letrers From CuHaries Sprague Sarcent To Recinatp SoMERS Cocxs, 1908-1926 (Continued). Edited by Joseph Ewan .... ANATOMY OF THE Patm Ruapis excensa, I. Mature VEGETATIVE Axis. Martin H. Zimmermann and P. B. Tomlinson ............ Srupres tn Atyssum: Near Eastern REPRESENTATIVES AND THEIR Autres, II. Section Meniocus AnD Section PsiLoNEMA. T. R. Dudley Tue GENERA OF VALERIANACEAE AND DrpsSACACEAE IN THE SOUTH- EASTERN Unirep States. I. K. Ferguson A Pretimimary Report on Funirera. Lorin I. Nevling, Jr. ........ Volume 46, No. 1, including pages 1-90, was issued , eee 15, 1966. 160 181 * 218° 232» a Second-class aes paid at aie Massachusetts JOURNAL OF THE ARNOLD ARBORETUM VoL. 46 APRIL 1965 NUMBER 2 THE GENERA OF POLYGONACEAE IN THE SOUTHEASTERN UNITED STATES? SHIRLEY A. GRAHAM AND C. E. Woon, Jr. POLYGONACEAE A. L. de Jussieu, Gen. Pl. 82. 1789, “Polygoneae,” nom. co (BuCKWHEAT Famity) Annual or perennial herbs, shrubs, trees, or vines, often with conspicu- ously swollen nodes, the nodes completely or partly surrounded by a mem- branaceous or scarious sheath, the ocrea, or this reduced to a row of hairs or completely wanting in a few genera. Leaves entire [rarely pinnatifid or palmately cleft], alternate or whorled [or opposite], petiolate to sessile, membranaceous to fleshy or leathery, occasionally articulated at the base with the stem. Inflorescences diverse, the flowers seldom solitary, mostly clustered into few-flowered fascicles, each fascicle subtended by a bract and each flower by a persistent sheath (ocreola) and occasionally also by 2 scarious bractlets; rarely fascicles subtended by and partly included in an involucre, the flowers subtended at the base only by numerous bractlets. Flowers regular, with a variable number of parts, mostly 3- or 5-merous, bisexual or unisexual, the base (receptacle?) narrowed into a short or long stipe which articulates with the pedicel. Tepals (2—)5 or 6, in 2 species outside this area, and references marked by an asterisk have not been seen by the authors. ts. Richard A. Howard and Ben W. Smith have reviewed the treatments of Coccoloba and of Rumex subg. Acetosa, respectively; their suggestions and corrections are gratefully acknowledged. Dr. Howard has generously allowed us to report two chromosome counts made by him; and Dr. R. B. Channel] collected excellent flowering and fruiting material which has been indispensable in the study of Brunnichia. The illustrations were prepared by Arnold D. Clapman. 92 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 whorls, imbricate or quincuncial in bud, often petaloid, + united, forming a short to long floral tube, persistent and sometimes enlarging in fruit. Stamens (3—)5-8 (9), inserted perigynously in 2 whorls, variously ar- ranged but generally paired, occasionally alternating with nectariferous teeth and/or united into an annulus surrounding the ovary; filaments filiform, generally of two lengths, those opposite the internal whorl of tepals longest; anthers 2-locular, longitudinally dehiscent, versatile or basifixed; pollen of diverse types. Gynoecium (2—)3-carpellate, syncar- pous; stigmata punctate, capitate, fimbriate, or flabelliform; styles 1-3, included to exserted; ovary 1, superior, lenticular or pyramidal, 1-locular, rarely incompletely 3-locular at base, sessile or slightly stalked; ovule 1, basal, on a short and thick or Jong and thin funiculus, eclictropous, the inner integument longer than the outer at anthesis and forming the micro- pyle. Fruit a lenticular or pyramidal (3-angled) achene, scarcely to en- tirely inclosed by the membranaceous tepals or the leathery or fleshy, ac- crescent floral tube. Embryo developing in the micropylar end of the seed, straight or curved, lying in one of the angles of the seed or along the middle of one side, outside of and curved around the abundant mealy or horny, sometimes ruminate endosperm or centered in it; perisperm want- ing; radicle superior; cotyledons of diverse shapes. Embryo sac develop- ment normal (Polygonum type). TypE Genus: Polygonum L. Predominantly of the North Temperate Zone, consisting of about 30 genera and 700 species; represented in the southeastern United States by seven genera. The family is a very natural and easily recognized one, but infrafamilial categories are much debated. At least five major classifica- tions, none of which is entirely satisfactory, have been proposed (cf. Meissner, Bentham & Hooker, Dammer, Gross, and Jaretzky). The classification of Gross, which seems to be most in accord with the com- monly accepted relationships of the genera, is followed here, although the subfamilial sequence is not considered by the writers to be an expression of the order of evolution within the family. Two genera, Fagopyrum Mill. and Emex Campd., included in Small’s Manual of the Southeastern Flora, are not known to be naturalized in the Southeast. Fagopyrum esculentum Moench (F. sagittatum Gilib., Poly- gonum Fagopyrum L.), buckwheat, 2m = 16, grown extensively as a grain in Russia, is cultivated to a limited extent in the United States. Although escaping to roadsides and other disturbed areas, it probably does not persist for more than a season, at least in our area. The genus is included, however, in the key to genera because of its occurrence as a waif. The Mediterranean Emex spinosa (L.) Campd. is a waif evidently rarely introduced as seed in ballast but apparently not persistent in the Southeast. Various authors have related the Polygonaceae to several families, in- cluding the Chenopodiaceae, Phytolaccaceae, Basellaceae, and Caryophyl- laceae, mainly on the basis of the unilocular ovary and single basal ovule. pagers - <= Se cm oa ~ 1965] GRAHAM & WOOD, GENERA OF POLYGONACEAE 93 The family is now generally regarded as derived from and advanced over the Caryophyllaceae. Floral anatomy of the Polygonaceae has been studied intensively. The most recent workers, Laubengayer and Vautier, agree that the basic floral plan is three-merous, with whorled parts, as in Eriogonum and Rumex, and that the five-merous condition, found, for example, in Polygonum, is derived. In five-merous flowers one tepal is bivalent, as shown by its position partly in and partly out of the two whorls of tepals and by its two vascular traces. (In some cases, the second trace may be reduced or missing, but the tepal generally remains intermediate in position between the two whorls. ) According to Laubengayer, the unilocular, uniovulate ovary was derived from a multiovulate gynoecium with free-central placentation by suppres- sion of the lower portion of the ovary and by adnation of the lower part of the placenta to the lower ovary wall. The “funiculus” is, then, the remaining portion of the free-central placenta. Joshi, however, believes the ovule is borne on a true funiculus, with only the basal end of the stalk perhaps representing a vestige of a free-central placenta. (See Lauben- gayer and Vautier for summaries of the views of other authors on various aspects of the floral anatomy.) The characteristic ocrea of the Polygonaceae has been interpreted vari- ously as an outgrowth of the sheathing base of the petiole, as the product of the fusion of either a row of adjacent stipules or two lateral stipules, or as an expanded axillary stipule. The pollen morphology is extremely diverse. The basic type from which others are thought to have been elaborated is ellipsoid, heavy- and smooth-walled, and has three long furrows, each with a central germ pore. It is found in such entomophilous genera as Eriogonum, Fagopyrum, and Antigonon, and in some insect-pollinated species of Polygonum. Evolution of the pollen grain is thought to have proceeded from this in two direc- tions: in insect-pollinated species (such as Polygonum Persicaria), toward further elaboration of exine sculpture with concomitant reduction of fur- rows and increase in number of pores, and in wind-pollinated species toward reduction in number of pores and the thickness of exine, and loss of furrows. The family is of minor economic importance, including a few food plants (e.g., Fagopyrum, Rheum Rhaponticum L.), some tropical timber trees, a number of ornamentals, and some noxious weeds. Fagopyrum was formerly the major commercial source of a flavonol, rutin, which strengthens capillary blood-vessel walls, preventing hemorrhage in victims of high blood pressure (cf. Couch, Humphreys). REFERENCES: BarLton, H. Polygonacées. Hist. Pl. 11: 367-400. 1892. BENTHAM, G., & J. D. Hooker. Polygonaceae. Gen. Pl. 3: 88-105. 18 BUCHINGER, M. Nota sobre la subdivisién de la familia de las Poligonécess. Bol. Soc. Argent. Bot. 7: 42, 43. 1957. 94 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 Coucn, J. F. Rutin for the capillaries. ay U. S. Dep. Agr. 1943/47: 711- 715. 1947. [Obtained from Fago DamMMe_r, U. Polygonaceae. Nat. Pia III. la: 1-36. 1892. Duke, J. A. Polygonaceae. Jn: R. E. Woopson, Jr., & R. W. ScHEry, Flora of Panama. Ann. Missouri Bot. Gard. 47: 323-359. 1960. [Including general information on Rumex, Polygonum, Coccoloba, Antigonon; treatment of Coccoloba by R. A. How Epman, G. Zur Entwicklungsgeschichte der Gattung Oxyria Hill, nebst zyto- logischen, embryologischen und systematischen Bemerkungen iiber einige andere Polygonaceen. Acta Hort. Berg. 9: 165-291. 1929. [Includes much of general interest about the family; large bibliography. | EmBercer, L. La structure de la fleur des Polygonacées. Compt. Rend. Acad. - — 208: 370-372. 1939. [See also Rev. Gén. Bot. 51: 581-638. ane bs Zur ne der Bliithen von Polygonum. Osterr. Bot. Zeitschr. 78: 229-241. Gross, H. Beitrage zur atts der Polygonaceen. Bot. Jahrb. 49: 234-339. 1913. [Morphology, anatomy, taxonom Humpureys, F. R. The occurrence and industrial production of rutin in south- eastern Australia. Econ. Bot. 18: 195-253. 1964. [From leaves of Eucalyptus macrorrhyncha; includes review of use of Fagopyrum and flowers of Sophora japonica as major sources; large bibliography. | JARETZKY, R. Beitrage zur ee der Polygonaceae unter Beriicksichtigung des Oxymethylanthrachinon-Vorkommens. Repert. Sp. Nov. 22: 49-83. pl. 22. "1925. . Histologische und karyologische Studien an Polygonaceen. Jahrb. Wiss. Bot. 69: 357-490. 1928. Josu1, A. C. The nature of the ovular stalk in Polygonaceae and some related families. Ann. Bot. II. 2: 957-959. 1938. raidafin P. Handbook of flower pollination. vol. 3. (Transl. J. R. A. Davis.) v + 644 pp. Oxford. 1909. [Rumex, 337-340; Pisce 341-349. | patie ath R. A. Studies in the anatomy and morphology of the polygona- ceous flower. Am. Jour. Bot. 24: 329-343. 1937. [Includes all genera of our area except Antigonon. | sional me L. Morphological and cytological studies on Fagopyrum esculen- Am. Jour. Bot. 22: 460-473. pls. 1, 2. 1935; II. Embryogeny. bid. og 129-133. 1936. MEISSNER, C. F. Monographiae generis Polygoni prodromus. iv + 169 pp. 7 pls. Genéve. 1826. ——-. iulasrieeters [except Eriogoneae]. DC. Prodr. 14: 1-186. 1856. Mirra, G. C. The origin, development and morphology of the ochrea in Poly- gonum iicatake L. Jour. Indian Bot. Soc. 24: 191-199. 1945. PerpricEAT, C. A. Anatomie comparée des Polygonées et ses rapports avec la morphologie et la classification. Actes Soc. Linn. Bordeaux 55: 1-91. pls. aa Passes . , fil. Polygonaceae. Jn: G. Hecr, Illus. Fl. Mittel-Europa. ed. 2. 3: oo. 1958. [Rumex, Oxyria, Rheum, a oe Fagopyrum. | soar H. N. The dispersal of plants throughout the world. xx + 744 pp. 2 pls. Kent, England. 1930. [See index for scattered aiesin on dis- oa of Rumex, Polygonum, and Brunnichia. Roserty, G., & S. Vautier. Les genres de Polygonacées. Boissiera 10: 7-128. 1965] GRAHAM & WOOD, GENERA OF POLYGONACEAE 95 1964. [An unsatisfactory treatment based on gross morphology and the unique evolutionary ideas of Roberty; much “splitting” of categories be- tween family and species, much “lumping” at the specific level (e.g., Rumex [ca. 100 spp.] reduced to 10 spp. and Eriogonum [ca. 150 spp.] cet to 8 spp., with 72 names, mostly of well- este species of the western U. S., equated with the distinctive E. tomentosum of the south- tag U. S.); not all names included; inadequate ti cb studied; use- ful bibliography. | Stnnott, E. W., & I. W. Bartey. Investigations on the phylogeny of rie rms. 3. Nodal anatomy and the morphology of stipules. Am. Jou Bot. 1: 441-453. 19 SIRRINE, E. Structure th seed coats of Polygonaceae. Proc. Iowa Acad. Sci. 2: 128-135. pls. 7-9. 1895. SouEcEs, R. “asl og sur l’embryogénie des Polygonacées. Bull. Soc. Bot. France 66: 168-199. 1919. [Polygonum Persicaria.| Ibid. 75-85. [| Rhe TUTIN, T. C. ed. Polygonaceae. Jn: T. G. Tutin, V. H. Heywoon, e¢ al., eds., FL. Europaea 1: 75-89. 1964. [Koenigia, Polygonum, Bilderdykia, Rey- noutria, Fagopyrum, Oxyria, Rheum, Rumex, Emex, Muehlenbeckia, Calli- gonum, Atraphaxis; treatments of various genera prepared by T Tutin, D. A. Wess, K. H. RECHINGER, and Wess & A. O. CHATER VAUTIER, S. La vascularisation florale chez les Polygonacées. Candollea 12: 219-343. 1949. [Including all genera of our area Wess, D. A., & A. O. CHaATER. Polygonaceae. /n: V. H. Heywoop, ed., Flora Europaea. Notulae systematicae ad Floram Europaeam spectantes No. 2. Repert. Sp. Nov. 68: 187-189. 1963. [Includes discussion of generic limits in Polygonaceae, especially Polygonum, sensu lato Wi.iiams, B. C. The occurrence of intercellular canals in root tips of plants in the Polygonaceae and the Labiatae. (Abstr.) Am. Jour. Bot. 37: 668. 1 WobeHouse, R. P. Pollen grains in the identification and classification of plants. VI. Polygonaceae. Am. Jour. Bot. 18: 5z.- 293 [Includes descriptions of pollen of Eriogonum, Rumex, Polygonum, Anti- ono nt alacant E. F. Observations on the development and germination of the seed in certain Polygonaceae. Am. Jour. Bot. 1: 454-476. pls. 45-48. 1914. [Includes Rumex, Fagopyrum, Polygonum, Polygonella.| Key TO THE GENERA OF POLYGONACEAE General characteristics: Jeaves mostly alternate and entire; nodes usually swollen and surrounded by a membranaceous sheath ; flowers regular, the tepals the pedicel; ovary 1, superior, 1-locular, with a single basal orthotropous ovule ; fruit a lenticular or 3-angled achene. A. Plants herbaceous or menos occasionally (in Polygonum) twining climbing vines without tendri B. Flowers subtended by ‘at partly inclosed in an involucre; ocreae want- oy op a PE II Sie aire re ae aoa 1. Eriogonum. B. Flowers not involucrate, subtended by membranaceous ocreolae; ocreae present; stamens (3—)5-8 (9). 96 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 C. Tepals 6, the outer spreading or reflexed and remaining small in fruit, the inner erect, enlarging in fruit and often with conspicuous, raised venation or callosities. . Rumex. . Tepals (2—)4 or 5 (6), all erect (the outer sometimes veflexed in fruit in Polygonella), without raised venation or callosities. D. Tepals remaining small in fruit; achene pyramidal, exserted; em- bryo embedded in the endosperm; cotyledons convolute; flowers in small panicles, the upper appearing corymbiform. ............ Pi ed re oe ae Pea Lees See eee est [Fagopyrum. | . Tepals enlarging in fruit; achene lenticular or pyramidal, generally inclosed by the tepals; embryo in one angle of the achene outside the endosperm; cotyledons accumbent or incumbent, not convo- lute; flowers in panicles or spikelike racemes . Branches nodal, not adnate to the internode; flowers 2 to several at a node in fascicles, the ocreolae imbricate, small and inconspicuous, scarcely unilaterally flared. .............. RRO ate eee Pie tities, Dg is Sy ioce hc Nel Sa WE ps 3. Polygonum. . Branches appearing internodal due to adnation to the internode above the node of origin; flowers solitary at the nodes, in the axils of conspicuous, imbricated, unilaterally flared ocreolae. Pe eS Oar Ge toate a ad aad 4. Polygonella. C Oo ica) — A. Plants shrubs, small trees, or tendril-bearing vines is F. Plants shrubs or small trees; leaves thick, leathery, reniform to orbic- ular; ocreae present; fruit an achene surrounded by a fleshy floral Me eee ee oe gen deat « 5. Coccoloba. ; ails tendril-bearing vines; leaves membranaceous, ovate to deltoid; ocreae wanting or much reduced; fruit a 3-6-angled achene inclosed by membranaceous to leathery tepals. G. Tepals rose-red, membranaceous; stipe remaining narrow in fruit; stigmata reniform- RRS solos es cay ee gaa ee 6. Antigonon. G. Tepals yellow-green, leathery; stipe expanding in fruit to form a broad unilateral wing; stigmata irregularly flabelliform. kr Subfam. ERIOGONOIDEAE Meissn. Tribe ErtoconeaE [Dumort. |] Eriogonum Michaux, Fl. Bor.-Am. 1: 246. pl. 24. 1803. Herbaceous, suffrutescent, or occasionally shrubby, bi- or triennials or perennials [or annuals] with large taproots, inhabiting open, dry, sandy or rocky regions; densely woolly [rarely glabrous]; branching di- or tri- chotomously; nodes lacking ocreae. Leaves basal in rosettes and cauline [or basal only], petiolate to sessile, the blade often decurrent on the petiole, the lower (abaxial) surface densely silver or brown tomentose. Basal leaves diverse in shape and size, in ours narrowly to broadly elliptic to oblanceolate; cauline leaves alternate, whorled [or opposite], narrowly to broadly elliptic, becoming bracts toward the top of the plant. Bracts foliaceous [or stipuliform], connate at base of involucre. Inflorescences Nig | | dlmmanmaa natin ete naan aan ee 1965] GRAHAM & WOOD, GENERA OF POLYGONACEAE 97 diverse, basically cymose; floriferous branches axillary and/or terminal; flowers subtended by and partly included within a turbinate [to cylin- drical], 3—5(—8)-lobed involucre, the involucres solitary on the peduncle [or clustered]. Flowers 3-merous, the base narrowed into a long stipe [or this nearly wanting], bisexual, subtended at the base by numerous linear or lanceolate, scarious, setiferous to villous bractlets; flowers many [rarely 1] in an involucre, exserted at anthesis. Tepals 6, petaloid, white to yellow [or red], ovate to linear, erect or nearly so at anthesis, enlarging and remaining erect in fruit or the outer whorl becoming oriented at right angles to the pedicel and the inner remaining erect, covering the fruit. Stamens 9, two opposite each outer tepal and one opposite each inner tepal, persistent; filaments pilose at base, exserted [or included J ; anthers versa- tile; pollen ellipsoid, tricolpate with long, tapering furrows, the exine thick, coarsely granular with 3 meridional strips between the furrows. Gynoecium 3-carpellate; stigmata 3, capitate; styles 3, filiform, long and exserted [or short and included]; ovary 3-angled, glabrous or villous. Fruit a 3-angled (or -winged?) ovate- or oblong-pyramidal, beaked achene. Embryo straight [or curved], centered [or excentric] in mealy endo- sperm; cotyledons thick, orbicular; radicle shorter [or longer] than coty- ledons. Type species: E. tomentosum Michx. (Name from Greek, erion, 1, and gonu, knee, in reference to the tomentose pubescence and the geniculate nodes of some species.) — UMBRELLA-PLANT. A North American genus of about 150 species in four subgenera and 14 sections, best developed in the western United States and well known for the taxonomic difficulties it presents. The four species occurring in the Southeast belong to subg. ErtoGonuM (subg. Eueriogonum S. Wats.) ,? comprised of perennials with turbinate involucres without angles or nerves and two to five or more foliaceous bracts, and sect. ERIOoGONUM (§ Eriantha Benth.), having flowers with a stipelike base, achenes not winged, and embryo straight, centered in the endosperm. Most widespread in the Southeast is Eriogonum tomentosum Michx., occurring from South Carolina to Florida along the sand hills of the inner part of the Coastal Plain. It is easily recognized by the dense, brown tomentum on the undersurface of the leaves and by the whorled cauline leaves. The three other southeastern species, E. longifolium Nutt., E. Harperi Goodman, and E. floridanum Small, have white pubescence and alternate cauline leaves. They and the eastern Texas E. vespinum Shinners *It is not clear ge the names proposed by Watson (1877) were meant to be subgenera or sections. He refers to them as sections and indicates to which section the earlier names of Saehini (1856) and Torrey & Gray (1870) _belong. The rank of te) (Coville & Morton, Munz & Keck, and Anderson), oie n’s names are treated here as subgenera and Bentham’ s and Torrey & Gray’s as section. 98 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 are morphologically so similar that maintenance of four species is highly questionable, and a re-evaluation of their taxonomy is recommended. Eriogonum longifolium occurs in Texas, Oklahoma, Kansas, southern Missouri, and Arkansas, while E. Harperi is described as endemic to Col- bert and Franklin counties in northwestern Alabama. Specimens examined from the area of Hot Springs, Arkansas, are difficult to place in either. Eriogonum floridanum, endemic to central peninsular Florida, differs from the preceding two species in having generally larger flowers, 6-10 mm. long vs. 4-8 mm. long, with very long pedicels (exceeding 1 cm.) in pe- dunculate, rather than nearly sessile, involucres. Eriogonum vespinum is distinguished from EF. longifolium only in being less densely pubescent and having slightly longer peduncles and floral stipes. Chromosome numbers, mostly undocumented, of several species of the western United States have been reported (2m = 18, 22, 24, 32, 34, 40, and 80), the most common number being 2” = 40. At least two western species, Eriogonum fasciculatum Benth. and E. nudum Dougl., include lyploid races of 2n = 40 and 80. The basic gametic number is thought to be x = 10, with hybridization and aneuploid increases and decreases accounting for the wide range of numbers. Natural hybrids of several species are known or suspected, but production of artificial hybrids is hampered by the difficulty of emasculating the small flowers, each of which produces but a single seed ; The genus is considered to be closely related to the endemic American Chorizanthe R. Br. and Oxytheca Nutt., the only other members of the family with involucrate, nine-staminate flowers. REFERENCES: Under family references see LAUBENGAYER, VAUTIER, and WODEHOUSE. ANDERSON, F. W. “Indicative” Eriogonums. Bot. Gaz. 12: 250-252. 1887. [E. ovalifolium in Montana not indicative of silver in the soil. Anpverson, J. M. A revision of Eriogonum section Pedunculata. Diss. Abstr. 20: 2509. 1960.* BENTHAM, G. Eriogoneae. DC. Prodr. 14: 5-28. 1856. CovILLeE, F. V., & C. V. Morton. Eriogonum intrafractum, a new species and new subgenus from Death Valley, California. Jour. Wash. Acad. Sci. 26: 303-306. 1936. ep oe oe Ganvocer, M. Le genre Eriogonum (Polygonaceae). Bull. Soc. Bot. Belg. 42: 183-200. 1905. _Lielues "Aagelianon of 23 new spp. and many new vars., all from the U Goopman, G. J. A new = can from the Southeast. Bull. Torrey Bot. Club 74: 329-331. 1947. [E. Harperi.] Mounz, P. A., & D. D. Keck. A California flora. 1681 pp. Berkeley, Calif. 1959. [Eriogonum, 332-354.] SHINNERS, L. H. Eriogonum vespinum eh Laerreoreee a new species from eastern Texas. Field Lab. 22: 68, 69. 19 STEYERMARK, J. A. Rare Missouri ey a Umbrella plant (Eriogonum longifolium). Missouri Bot. Gard. Bull. 44: 106, 107. 1956 Stokes, S. G. The genus Eriogonum, a preliminary study based on geographic distribution. 132 pp. San Francisco. 1936. a ii i te 1965] GRAHAM & WOOD, GENERA OF POLYGONACEAE 99 G. L. Stepsins. Chromosome numbers in the genus Eriogonum. Leafl. West. Bot. 7: 228-233. 1955. Torrey, J., & A. Gray. A revision of the Eriogoneae. Proc. Am, Acad. Sci. 8: 145-200. 1870. [Eriogonum, 146-190.] Watson, S. Descriptions of new species of plants, with revisions of certain genera. Proc. Am. Acad. Sci. 12: 246-278. 1877. [Eriogonum, 254-269.] Contributions to American botany. II. Descriptions of some new species of North American plants. /bid. 14: 288-303. 1879. [Eriogonum, 295, 296.] Subfam. POLYGONOIDEAE | Meissn. | Tribe RuMIcEAE Dumort. 2. Rumex Linnaeus, Sp. Pl. 1: 333. 1753; Gen. Pl. ed. 5. 156. 1754. Herbaceous annuals or perennials of open, generally disturbed habitats. Plants glabrous, with long taproots; ocreae scarious, cylindrical, the margins often becoming shredded with age. Leaves basal and cauline, entire, sometimes with crisped or undulate margins, membranaceous to subcoriaceous, long-petiolate, lanceolate [linear] to ovate, the base acute to obtuse or hastate, the apex acute; cauline leaves alternate, the upper ones reduced. Inflorescence composed of [a single terminal or] several axillary panicles, the flowers mostly 3 to several in an axillary fascicle at each node. Flowers 3-merous, green or reddish, bisexual or unisexual, the plants monoecious, polygamomonoecious, or dioecious; when monoeci- ous the ¢ flowers borne above (distal to) the @ ones. Tepals 6, slightly united at base, forming a short floral tube which is narrowed into a long stipe shorter than to exceeding the pedicel; 3 outer tepals linear, reflexed and remaining small in fruit; 3 inner tepals (“valves”) broadly ovate, erect and accrescent, inclosing the fruit, in some species developing raised reticulations and/or thickened callosities (“grains” or tubercles) on 1 or all 3 tepals, or with a prominent midvein on the outer (abaxial) side of each, the margins entire or variously toothed. Stamens 6, in pairs Opposite the outer whorl of tepals; filaments short; anthers basifixed; pollen spheroidal, 18—32 » in diameter, 3—6-colporate, the furrows long and narrow, variously arranged depending on the number present, the exine finely to coarsely reticulate. Gynoecium 3-carpellate; stigmata 3, fimbri- ate; styles 3; ovary 3-angled. Fruit a 3-angled, pyramidal achene. Seed coat smooth, shining. Embryo nearly straight, lying along the middle of one side of the achene; cotyledons linear to lanceolate, generally parallel or rarely perpendicular to wall of achene. (Including Acetosa Mill., Aceto- sella (Meissn.) Fourr.) Lectotype species: R. Patientia L.; see Britton & Brown, Illus. Fl. No. U. S. ed. 2. 1: 653. 1913. (Name adopted by Linnaeus from ancient Latin.) — Dock, SORREL. A genus of over 100 species, world-wide in distribution, although poorly represented in tropical regions. About 50 species and hybrids are re- ported from North America, 11 of these introductions from Europe. 100 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 Thirteen to fifteen species in three of the four subgenera occur in the southeastern United States. Several species are widely distributed as noxious weeds of cultivated ground and roadsides. Subgenus RuMEx (subg. Lapathum Pers.) is comprised of annuals and perennials, with leaf bases round to cuneate, never sagittate or hastate, flowers bisexual and inner tepals with or without callosities. The chromo- somes are the smallest in the genus, x = 10, and 2m ranges from 20 to 200. Polyploidy and inter- and intraspecific hybridization have been re- ported in both sections of the subgenus. In sect. RuMEXx (§ Simplices Rech. f.), the single stem is unbranched with a terminal inflorescence. The six species of this section which oc- cur in the southeastern United States are of European origin and are highly polymorphic, mostly with several subspecies. Rumex crispus L., 2n = 60; R. pulcher L., 2n = 20; and R. obtusifolius L., 2n = 40 and 60, are widespread in the United States. Growing along roadsides and in waste places throughout our area, they are occasionally sympatric (€.g., in New Orleans, Louisiana; Tallahassee, Florida; and Granger and Hawk- ins counties, Tennessee). Putative natural hybrids of R. crispus X R. obtusifolius have been reported from Louisiana and of R. crispus R. pulcher from Virginia. Rumex crispus in Great Britain is composed of several races, all evidently interfertile (Rechinger, 1961). Although widely distributed in the northern and western states, R. conglomeratus Murr., 2m = 20, appears in our area only in the more easterly parts of North Carolina. Also common elsewhere in the United States but of limited distribution in the Southeast are R. Patientia L., 2n = 60, known from scattered localities in Tennessee and North Carolina, and R. mari- timus L. var. fueginus Dusén (confused by Small with R. persicarioides L.), 2m = 40, reported from Little Rock, Arkansas. According to Rech- inger (1937), R. sanguineus, 2n = 20, is probably not present in the Southeast but has been confused by American taxonomists with R conglomeratus. Section AxILLarEs Rech. f., characterized by plants with leafy axillary branches and secondary inflorescences, has its center of development in North America. Extra-American species are believed by Rechinger, on the basis of their isolated systematic position, to be relics of ancient lines of evolution. Seven species of this section have been collected in the Southeast, but R. mexicanus Meissn. and R. cuneifolius Campd., native to Mexico and South America, respectively, are known from limited ma- terial which suggests they are not established introductions. Rumex verticillatus L., 2n = 48, 60, R. floridanus Meissn., and R. fascicularis Small, restricted to the Atlantic and Gulf coastal plains, are distinguished from one another on the basis of slight morphological differences which scarcely support specific rank. Also occurring on the Coastal Plain in Louisiana is R. chrysocarpus Moris (R. Berlandieri Meissn., R. Langloisit Small). Rumex altissimus Wood, 2n = 20, widely distributed in the eastern and central United States, is known from most of the Southeastern States, generally occurring inland from the Coastal Plain. Several experi- 1965] GRAHAM & WOOD, GENERA OF POLYGONACEAE 101 mental crosses made among species in the section (see Sarkar) show a high degree of pairing and good seed-set in the F; generation of the dip- loids, but completely sterile triploid and tetraploid hybrids. Subgenus Acetosa (Mill.) Reichenb., morphologically and cytologically the most variable subgenus, includes about 40 species in six sections, with a predominantly Old World distribution (most abundant in northern and eastern Africa, southern Europe, and the Middle East). The plants are shrubs or perennial or annual herbs and are hermaphroditic or dioecious, commonly with hastate or sagittate leaves, and inner tepals much longer than the achene and mostly without callosities. Chromosome numbers vary from 2m = 8 to 40. The dioecious annual, Rumex hastatulus Baldw. ex Ell., 2n = 8 2,9 $,or2n = 10 2,10 4, is the only member of sect. ACETOSA subsect. AMERICANAE Live & Sarkar; its nearest relatives are the western American R. paucifolius Nutt. ex S. Wats. (§ Paucifoliae Live & Sarkar) and the Old World R. Acetosa L. and R. thyrsiflorus Fingerh. (§ Acetosa), both with 2n = 142, 154 and adventive in northeastern merica. Rumex hastatulus is widespread in sandy soils of the Coastal Plain from North Carolina to northern Florida, west to Texas and Okla- homa, and occurs in the adjacent Piedmont on disturbed, well-drained sites. Sex determination in Rumex hastatulus, R. Acetosa, and R. paucifolius, 2n = 14 and 28, has been studied by several workers. An X/A balance mechanism of sex determination similar to that originally discovered in Drosophila occurs in R. Acetosa, with sex expression determined by the ratio of autosomes to X chromosomes, the Y chromosomes having no effect. Sex expression in R. paucifolius appears, on the other hand, to follow the more primitive X/Y scheme, discovered in Silene (Melandrium) and de- scribed later in R. Acetosella, in which strong male factors are located in the Y chromosome and the effects of the autosomes are secondary to the influence of XX versus XY. Two karyologically distinct races differing in sex-chromosome mor- phology and chromosome number occur in Rumex hastatulus. The differ- ences in karyotype have involved loss of one autosomal pair and the appearance of multiple Y chromosomes, changes which Smith has postu- lated to be the result of two successive translocations. Populations from North Carolina to Florida and Mississippi have a modified X/A mech- anism (2n = 6 + XX 9 and 6 + XY,Yo2 4 ). Although sex is determined primarily by the ratio of X chromosomes to autosomes, the Y chromosomes are not entirely neutral but contain a region which enhances the expression of maleness. Populations from Louisiana and Texas to Oklahoma and Arkansas have ten chromosomes in both sexes (27 = 8 + XX @ and 8 + XY 4). The genetics of sex determination in this western race have not yet been analyzed. ; Staminate plants of Rumex hastatulus are “heterogametic,” producing male- and female-determining pollen in equal numbers; carpellate plants are “homogametic.” In North Carolina, carpellate plants (57%) are more abundant than staminate (43%). Smith has suggested that the disparity in sex ratio in R. hastatulus may be attributable to a differential 102 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 selection favoring either the female-determining pollen or the female- determined embryos prior to the seedling stage, a phenomenon which has been demonstrated in R. Acetosa and in other dioecious plant species. Apogamy occurs in a few species of the subgenus but accounts for only an estimated 1-2% of seed-set. Subgenus AcETOSELLA (Meissn.) Rech. f. consists of a polyploid series of perennial dioecious herbs with hastate or ‘sagittate leaf bases and callus-free tepals which generally do not exceed the achene. One to four species are recognized, depending mainly on the emphasis placed on chro- mosome number. Love (1944b), maintaining that “polyploidy alone seems to catapult a species into being,” has recognized four species: Rumex angiocarpus Mutb., 2n = 14; R. tenuifolius (Wallr.) A. Love, 2n = 28; R. Acetosella L., sensu sieteto: 2n = 42; and the arctic R. graminifolius Rudolph ex Lamb., 2” = 56. The morphological distinctions between these are few, mostly quantitative, and tend to disappear when popula- tions are studied over the entire range. Hexaploids (2n = 42) with the character of R. angiocarpus (tepals adnate to the fruit) have recently been found in populations from Europe, Canada, and Australia (Johnson & Briggs, p. 166). This complex, including Eurasian weeds of nearly cosmopolitan distribution, is represented in our area by diploid and hexaploid plants. Sex chromosomes are present, and the sex-determining mechanism is thought to be of the X/Y type. Rumex Acetosella is generally considered an indicator of acid, worn-out soils, but Artist, finding that the species grows on soils of pH 4.5—8.5, believes its distribution 1s probably dependent on a combination of several ecological factors, rather than on pH alone. Pollination in Rumex is almost exclusively anemophilous, although in- sects occasionally are attracted by the red fruiting inflorescences. The fimbriate stigmata are well adapted to receiving air-borne pollen. Rumex crispus and R. obtusifolius are distinctly proterandrous; several other species are reportedly self-pollinated. Experimental crosses between sub- genera have been unsuccessful. Floral structure in Rumex is believed to have been derived by reduction from the basic three-merous type possessed by the closely related Rheum. Internal vascular bundles and internal phloem have been described from several species of Rumex, including R. crispus, R. Patientia, and R. conglo- meratus. The perennial species without internal bundles are considered the oldest members of the genus The genus is of little economic importance. The leaves of some species are eaten as salad greens, the oxalic acid content imparting a slightly sour taste. Roots of Rumex hymenosepalus Torr., canaigre, contain up to 30% tannin and have been suggested as a commercial source of tannin. Some weedy species, such as R. crispus, are pests in cultivated areas, but can be controlled by cutting away the taproot to three inches below the soil surface, since regenerative capacity has been found lacking in the rest of the taproot (Healy). 1965 ] GRAHAM & WOOD, GENERA OF POLYGONACEAE 103 REFERENCES: Under family references see DUKE, KnuTH, LAUBENGAYER, RECHINGER, Ru- LEY, SOUEGES, TUTIN, VAUTIER, WODEHOUSE, and Woopcock AuLEs, H. E., C. R. Bett, & A. E. Raprorp. Species new to the flora of North or South Carolina. Rhodora 60: 10-32. 1958. [R. Patientia, 13, 14.] ALLEN, C. E. The genotypic basis of sex-expression in angiosperms. Bot. Rev. 6: 227-300. 1940. Anonymous. Canaigre. Bull. Misc. Inf. Kew 1890: 63~69. 1890. Jbid. 1894: 167, 168. 1894. [R. hymenosepalus. | Artist, R. C. The value of Rumex Acetosella as an acid indicator. Butler Univ. Bot. Stud. 2: 81-91. 1932. CHANCELLOR, A. P. Studies on the ecology of some species of the genus Rumex. Brit. Weed Control Conf. Proc. 3: 197-203. 1956.* Dupceon, W. Morphology of Rumex crispus. Bot. Gaz. 66: 393-420. pls. 17- 19. 1918, [Floral morphology. | Fink, B. Contribution to the life-history of Rumex. Minn. Bot. Stud. 2: 137-153. pls. 9-12. 1899. FREEMAN, O. M. Notes on the flora of Polk peg North Carolina. Castanea 20: 37-57. 1955. [R. hastatulus, R. pulcher, 45.] oo en N. W., & D. S. Brack. Canaigre: a potential domestic source of nin. U. S. Dep. Agr. Prod. Res. Rep. 28. 32 pp. 19 ey J. L., & A. P. CHANCELLOR. The comparative hicks? of closely re- lated species living in the same area. IV. Rumex: interference between individuals in populations of one and two species. Jour. Ecol. 47: 679- 695. 1959. [R. obtusifolius, R. crispus, R. conglomeratus, R. sanguineus, R. Hydrolapathum. Heaty, A. J. Control of docks. New Zealand Jour. Sci. Tech. A. 34: 473-475 1953. JACQUETY, = Etude de la couche protectrice sur les souches de deux Rumex. C Rend. Acad. Sci. Paris 243: 1437-1439. 1956. [R. crispus, R. oote? folius.] Description et histogenése d’une souche primaire de Rumex obtusifolius DC., dans sa phase en rosette. /bid. 245: 2528-2531. 1957. etait Mo orphologie de l’inflorescence de Rumex obtusifolius D. C. Ibid. 247: 1481-1484, 1958. Jensen, H. W. Meiosis in Rumex. I. Polyploidy and the origin of new species. Cytologia 7: 1-22. pl. 1. 1936; II. The origin and behavior of the so-called sex chromosomes in Rumex. Ibid. 23, 24. pls. 2, 3. [Incorrectly rejects the existence of i * ] Jounson, L. A. S., & B. G. Briccs. Taxonomic and cytological notes on Acetosa a Acetosella i in nena Contr. New S. Wales Natl. Herb. 3: 165-169. 196 Josut, ‘ C. The anatomy of Rumex with special reference to the morphology of the internal bundles and the origin of the internal phloem in the Poly- gonaceae. Am. Jour. Bot. 23: 362-369. 1936. Kitanr, Y. A new root-tip stain peer on Siamese Acetosa. (Abstr.) (In Japanese.) Jap. Jour. Genet. 31: 302. Krause, E. H. L. Lapathon und Patience. Taevitinees iiber die Geschichte von Rumex Patientia. Beih. Bot. Centralbl. 24(2): 6-52. 1908. [R. Patientia in early literature; its uses. ] 104 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 Love, A. Physiological differences within a natural polyploid series. Hereditas 28: 504-506. 1942. aie: subg. Acetosella. | The dioecious forms of Rumex subgenus Acetosa in reas Bot. Not. 1944: 237-254. 19442, [Establishes 4 sects. in subg. Ace Cytogenetic studies on Rumex, subg. Acetosella. ons 30: 1-136. 1944b. [Bibliography of 189 references covering topics of polyploidy and sex chromosomes in plants; distribution maps of ene Acetosella. | . Agamospermy in Acetosa. Ibid. 35: 390-393. 1949. Sex determination in Rumex. Proc. Genet. Soc. Canada 2: 31-36. 1957." ARKAR. Cytotaxonomy and sex determination of Rumex pauci- folius. Canad. Jour. Bot. 34: 261-268. 1956. McVaucu, R. The vegetation of the granitic flat-rocks of the southeastern United States. Ecol. Monogr. 13: 119-166. 1943. [R. hastatulus, 140; distribution map, 154.] MAHEsHwarI, P. Origin and development of internal bundles i the stem of Rumex crispus. Jour. Indian Bot. Soc. 8: 89-117. pls. 1, & B. StncH. On the internal bundles in the stem of och Patientia L. Proc. Indian Acad. Sci. 45: 153-157. 1942.* Marek, S. Morphological and anatomical features of the fruits of genera Polygonum L., Rumex L. and keys for their determination. (In Polish; Notte, M. Rhizomerfall mit vegetativer Vermehrung hei Kanes: Bot. Jahrb. 76: 224-250. 1954. Norpstept, O. Apogami hos Rumex. Bot. Not. 1907: 238. 1907. ReEcuHInGceER, K. H., fil. Vorarbeiten zu einer Monographie der Gattung Rumex. [I.] Beih. Bot. Centralbl. 49(2): 1-132. pls. 1-3. 1932; II. Die Arten der Subsektion Patientiae. Repert. Sp. Nov. 31: 225-283. pls. 135, 136. 1933; III. Die siid- und zentralamerikanischen Arten der Gattung Rumex. Ark. Bot. 26A(No. 3): 1-58. pls. 1-6. 1934; IV. Die australischen und neu- seelandischen Arten der Gattung Rumex. Osterr. Bot. Zeitschr. 84: 31-52. 1935; V. The North American species of Rumex. Publ. Field Mus. Bot. 17: 1-151. 1937; VI. Versuch einer natiirlichen Gliederung des Formenkreises von Rumex bucephalophorus L. Bot. Not. 1939: 485-504. 1939; VII. Rumices Asiatici. Candollea 12: 9-152. 1949; VIII. Monograph of the genus Rumex in Africa. Bot. Not. Suppl. 3(3): 1-114, 1954. Lines of evolution and geographical pee ah pl in Rumex subgen. Lapathum. Watsonia 1: 19-23. 1949. . Notes on Rumex Acetosa L. in the — Isles. Jbid. 5: 64-66. pls. 4,5. 1961. [Includes observations on R. c SarKAR, N. M. Cytotaxonomic studies on sn section Avxillares. Canad. Jour. Bot. 36: 947-996. 1958. SayRE, J. D. Physiology of stomata of Rumex Patientia. Science 57: 205, 206. 1923. SmitH, B. W. Sex chromosomes and natural polyploidy in dioecious Rumex. Jour. Hered. 46: 226-232. 1955. : ie mechanism of sex determination in Rumex hastatulus. Genetics 48: 1265-1288. 1963. . The evolving karyotype of Rumex hastatulus. Evolution 18: 93-104. 1964. 1965] GRAHAM & WOOD, GENERA OF POLYGONACEAE 105 WerLt, J., & M. Bournértas. X Polygonorumex Guineti J. Weill (Rumex obtusifolius L. X Polygonum Hydropiper L.). Bull. Soc. Bot. France 3: 32 1-326. 1946. Tribe Potyconeae [Persicarieae Dumort. ] 3. Polygonum Linnaeus, Sp. Pl. 1: 359. 1753; Gen. Pl. ed. 5. 170. 1754. Annual or perennial herbs, shrubs, or twining vines of diverse, com- monly moist habitats, often occurring as tenacious weeds in disturbed areas. Plants glabrous or pubescent; stems erect or prostrate, simple or branched, unarmed or with short recurved spines, the nodes often con- spicuously swollen; ocreae completely or partly surrounding the nodes, cylindrical, membranaceous or scarious, often becoming hyaline with age, the distal margin entire, shredded, or bearing 1 to many short awns. Leaves basal and/or cauline, membranaceous [to + fleshy], entire [rarely lyrate-pinnatifid or palmately cleft], alternate, sessile or petiolate, mostly linear to ovate, occasionally cordate, hastate, or sagittate, in some articu- lated with the stem, Inflorescence composed of axillary fascicles of 1 to several spirally arranged, pedicellate flowers; fascicles solitary in the axils of leaves or arranged into 1 or more terminal or axillary racemes (termed “spikes’”’), less commonly the racemes of fascicles paniculate [rarely the lower half of the inflorescence bulbil-bearing]. Flowers regular, green or white to pink or red, commonly 5-merous, bisexual or functionally unisexual, the stipe generally short, occasionally long; pedicel articulated with the flower, subtended by a scarious ocreola. Tepals (2—)4 or 5 (6), + distinct, petaloid, erect, accrescent, completely or partly inclosing the achene, the outer ones smooth or keeled, rarely developing into broad wings in fruit. Stamens (3—)5~8 (9), the number and arrangement variable within each species, generally included, alter- nating with short, oblong, flattened nectariferous teeth, or the teeth lack- ing; anthers versatile: pollen of diverse types, some characterizing sections of the genus. Gynoecium (2—-)3-carpellate; stigmata 2 or 3, punctate, pe Nag or fimbriate; style 1 and deeply cleft or 2 or 3, included to exserted, straight and deciduous, or hooked at the apex, indurated and persistent; ovary lenticular or 3-angled. Fruit a lenticular or 3-angled pyramidal achene, both types occasionally occurring on the same plant. eed coat smooth and shiny or rough and dull, light to dark brown. Embryo in one of the angles of the achene, outside of and curved around the abundant mealy or horny endosperm; cotyledons linear to lanceolate, accumbent or incumbent. (Including Antenoron Raf., Bilderdykia Du- mort., Bistorta Mill., Duravia (S. Wats.) Greene, Persicaria Mill., Pleur- opteropyrum Gross, Pleuropterus Turcz., Reynoutria Houtt., Tiniaria (Meissn.) Webb & Mog., and Tracaulon Raf.) LectoTyPE SPECIES: P. aviculare L.; see Britton & Brown, Illus. Fl. No. U. S. ed. 2. 1: 659. 106 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 1913. (Name from ancient Greek, polygonon, an herb, from poly, many, and gonu, knee, the swollen nodes producing the appearance of jointed stems.) — KNOTWEED, SMARTWEED. A highly variable and taxonomically difficult genus of about 150 species, nearly world wide in distribution and composed of several natural species- groups, the taxonomic ranks of which are much in dispute. In recent years data from cytology, palynology, floral anatomy, and chemistry have been used in an attempt to determine the relationships of the taxa and to devise a more natural classification. Unfortunately, emphasis often has been placed on one technique, without regard for existing information from other studies, thus leading to conflicting classifications. There is a need for correlation and careful weighing of all data in future studies of Polygonum, sensu lato. The system followed here is, in general, modified from those of Steward, Steyermark, and Gleason. The infrageneric divisions, being based almost solely on variable vegetative characters, are considered sections of the genus. Thus defined, Polygonum is divided into eight sections, five of which occur in our area. Section Tovara, distinguished from the rest of Polygonum mainly by its persistent, indurated, hooked styles, is considered by some taxonomists to be worthy of generic rank. However, to establish a genus in this family on the basis of a single character, even a floral one, is out of keeping with the magnitude of differences upon which the other genera of Polygonaceae are based. Section TINIARIA is considered to include sect. PLEUROPTERUS, since morphological distinctions between the two, though present in the species of our area, are lacking in several Asiatic species (e.g., P. baldschuanicum Regel, a woody vine with capitate stigmata but a much-branched paniculate inflorescence). Identical pollen morphology (see Hedberg), as well as chemical similarities (Jar- etzky), support this union. Fagopyrum Mill., which differs from Poly- gonum in having nonaccrescent tepals, the embryo embedded in the endo- sperm, cotyledons rolled around the radicle, and more or less corymbiform inflorescences, is maintained as a distinct genus. The pollen of Polygonum is of interest because of its unusual diversity ; sectional limits within the genus are marked in several cases by distinct spermae. In P. Persicaria, for example, the archesporial cell acts directly as the pollen mother-cell, dividing to form four microspores in each locule. Rarely it divides to form two pollen mother-cells, resulting in eight pollen grains. In P. lapathifolium and P. sagittatum there are eight pollen grains in each locule, and in P. aviculare, P. Convolvulus, and P. orientale, thirty-two. Other species have several pollen mother-cells, and the number of pollen grains may be 64, 128, or 256. Species with porate pollen have fewer grains per locule than those with colporate pollen. The number of pollen grains does not characterize sections. 1965] GRAHAM & WOOD, GENERA OF POLYGONACEAE 107 Numerous chromosome counts have been made, but, unfortunately, many are unreliable due to questionable identifications and lack of voucher specimens. On the basis of all reported counts there seems to be no corre- lation between chromosome numbers and recognized sectional limits, wide ranges being found in each section. Section Poryconum (§ Polygonum DC., § Avicularia Meissn., Poly- gonum L., sensu stricto; including Duravia (S. Wats.) Greene) is com- prised of generally prostrate, much-branched annuals, with small leaves usually less than 2 cm. long and articulated at the base with the stem and flowers in axillary clusters. The pollen is spheroid-prolate, 3- or 4- colporate with furrows 4-44 the length of the polar axis, elliptic pores, and a smooth or granulate exine. Chromosome numbers of 2 = 20, 22, 24, 40, 60, and 66 have been reported. Several widely distributed species of the section are present in our area, including the polymorphic P. aviculare L. (P. neglectum Bess., P. buxiforme Small), 2n = 20, 22, 40, 60; the coastal American endemic P. glaucum Nutt., 2m = 20, 40; P. erectum L., 2n = 40; and P. tenue Michx., 2n = 20. Equally widespread in the United States, but of more restricted occurrence in the Southeast (mainly in Louisiana, Arkansas, and Tennessee), are P. ramosissimum Michx., 2n = 20; P. prolificum (Small) Robins., 21 = 60; and P. cam- porum ‘Meissen: The Asiatic P. argyrocoleon Steud. ex Kuntze has been collected as a waif on the outer banks of North Carolina. The species of sect. PocycoNumM reportedly are inbreeding, the flowers being self-pollinated and perhaps also cleistogamous. In recent work on P. aviculare and related species, Styles has suggested that both hybridiza- tion and apomixis are rare, the extreme variability in morphology resulting from the influence of seasonal changes and of environmental factors in general. Section Prrsicarta (Mill.) DC.* (Persicaria Mill.) includes plants of low, wet habitats and waste ground, with flowers in axillary or terminal spikes or racemes. The leaves are mainly elliptic to lanceolate and cauline, the tepals are neither keeled nor winged, and the branched stems are glabrous or pubescent, without recurved prickles. The pollen is spheroid, polyporate, with scattered circular pores and a reticulate exine. Reported chromosome numbers include 2” = 20, 22, 24, 40, 44, 60, and 66. The section is the largest of the genus, with about 100 species. Of the 14 species in the Southeast, at least five are introductions from Europe or Asia: Polygonum lapathifolium L. (Persicaria lapathifolia (L.) S. F. Gray), 2n = 22; P. Persicaria L. (Pers. Persicaria (L.) Small), 2n = 40, 44; P. Hydropiper L. (Pers. Hydropiper (L.) Opiz), 2n = 20, 22; P. caes- pitosum var. longisetum (De Bruyn) Steward; and P. orientale L. (Pers. orientalis (L.) Spach), 2" = 22, 24. The last two occur only in the northern part of our area, while the other three are widespread. Species * Persicaria was used “ authors before De Candolle as an infrageneric category (e.g., see Linnaeus, Sp. P 60. 1753) but gree clear indication of rank. De Candolle (Lamarck & i. Candolle, Fl. Franc. ed. 3. 3: 365. 1805) appears to have been the first to apply the name without st to a section of Polygonum, 108 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 indigenous in our area are the variable, often aquatic P. coccineum Muhl. (Pers. Muhlenbergii (S. Wats.) Small); P. densiflorum Meissn. (Pers. portoricensis (Bert.) Small) and P. hirsutum Walt. (Pers. hirsuta (Walt.) Small), of the Atlantic coastal plain; and three highly variable species, P. pensylvanicum L. (Pers. pensylvanica (L.) Small), P. hydropiperoides Michx. (Pers. opelousana (Riddell) Small, Pers. setaceum (Baldwin) Small), and P. punctatum Ell. (Pers. punctata (Ell.) Small, P. robustius (Small) Fern.). Polygonum bicorne Raf. (Pers. longistyla (Small) Small) extends from the west into Louisiana, Arkansas, and Mississippi. Poly- gonum mississippiense Stanford (Pers. mississippiensis (Stanford) Small) has been described from Mississippi and Persicaria paludicola Small, 1933 (not Pers. paludicola (Makino) Nakai, 1926), from the Everglades, Florida. The northeastern Polygonum Careyi (Pers. Careyi (Olney) Greene), although included in Small’s Manual, does not appear to extend southward into our area. A number of species in sect. PerstcartA have both terrestrial and aquatic forms (sometimes remarkably unlike) which can be changed from one to the other by merely changing the habitat. Such amphibious species are often sterile, spreading by means of rhizomes. Cleistogamy is re- ported in the section, and a tendency in the flowers toward separation of the sexes is apparent in some species, e.g., P. hydropiperoides. Several putative hybrids have been recorded between species of sect. PERSICARIA, especially in Europe (for a list of hybrids see G. Hegi, Illus. Fl. Mittel- Europa, ed. 2. 3: 396-400. 1958), but almost no controlled experimental studies have been made. Careful biometrical, ecological, and cytological studies are needed for a better understanding of the species limits in this, as in the other sections of Polygonum. Section EcHINOcAULON Meissn. (Tracaulon Raf.) is closely related to and at times has been combined with sect. Persicarta, from which it differs mainly in the sagittate or hastate leaves, reflexed prickles on the angles of the stem, and reclining, creeping, or climbing habit. The pollen is indistinguishable from that of sect. PersicArtA. Chromosome numbers of 2” = 20, 22, 24, 40, 44, 60, and 66 have been reported for the section. Two species, P. sagittatum L. (Tracaulon sagittatum (L.) Small), with sagittate leaves, and P. arifolium (T. arifolium (L.) Raf.), with hastate leaves, occur in tidal marshes and other wet places in our area. The Brazilian P. Meisnerianum Cham. & Schlecht. var. Beyrich- ianum Meissn., easily recognized by its cordate, narrowly lanceolate leaves and dichotomously branched inflorescence with glandular hairs, was col- lected, probably as a waif, along the Tchefuncta (Chefuncte) River, St. Tammany Parish, Louisiana, in the late 1800’s. Section Tovara (Adans.) Benth. & Hook.* (Antenoron Raf.®), com- ‘Torrey (Fl. State N. Y. 2: 151. 1843), apparently, was the first to use Tovara as an infrageneric category of Polygonum, but without clear indication of rank. The first certain use of Tovara as - section of Polygonum appears to be that of Bentham & Hooker (Gen. Pl. 3: 98. 1880 °If sect. Tovara is recognized at ‘the generic level, the name Tovara Adans. (Fam. 1965] GRAHAM & WOOD, GENERA OF POLYGONACEAE 109 prised of one to three species disjunctly distributed in eastern America and eastern Asia, is distinguished by an elongate spikelike inflorescence, a 4-parted floral tube, and a persistent style, 2-parted to the base with recurved, hooked tips, which in fruit becomes very rigid and bent obliquely downward. According to Hedberg, the pollen is spheroid and 12- porate, with oblong pores arranged as though along the 12 edges of a cube; the exine is finely reticulate. The eastern American Polygonum virginianum L. var. virginianum (Antenoron virginianum (L.) Roberty & Vautier, Tovara virginiana (L.) Raf.) is widely distributed in shady bottom lands and thickets, from Quebec to northern Florida, west to eastern Texas and Minnesota, with disjunct populations in central Mexico (Hidalgo and Puebla). A glabrous variety, P. virginianum var. glaberrimum (Fern.) Steyerm., was described from our southeastern coastal states. The eastern Asiatic members of this section have been considered as either two distinct species (P. filiforme Thunb. and P. apoénse Elmer) or three varieties of P. virginianum, depending on the emphasis placed on density of pubes- cence, shape and size of leaves, shape of base of leaves, and color of tepals. Polygonum virginianum var. filiforme (Thunb.) Nakai, dis- tributed from southern Korea and Japan to southwestern China, has a reported chromosome number of 2” = 44. The dispersal mechanism of fruits in this group is unique in the family. During maturation of the fruit a separation layer forms across the pedicel, leaving only the vascular cylinder intact. When the fruits are mature, disturbance of the plant causes the vascular cylinder to break and the fruits to be catapulted forcibly as much as three or four meters. The origin of the catapulting force has not been satisfactorily explained in modern physiological terms (cf. Reed & Smoot). In addition to being thrown from the parent plant, the fruits, with their persistent hooked styles, may catch in the coats of animals, becoming dispersed over even greater areas. Section TrnrariaA Meissn. (including Reynoutria Houtt., 1777; Bilder- dykia Dumort., 1827; Tiniaria (Meissn.) Webb & Mogq., 1846; and Pleuropterus Turcz., 1848) is comprised of annual or perennial trailing or erect plants with axillary or terminal spikes or panicles. The tepals are keeled or winged, the stigmata capitate or fimbriate, and the pollen similar to that of sect. PotycoNum, but with furrows generally more than +, the length of the polar axis and a smooth to finely reticulate exine. Chromosome numbers of 2” = 20, 22, 40, 44, and ca. 88 have been re- ported for the section. Three species occur in our area. Polygonum Convolvulus L. (Bilderdy- Pl. -2: 276. 4763) cannot be used, since it has been rejected in favor of a later homonym, Tovaria Ruiz & Pav. (Prodr. 49. 1794; see Int. Code Bot. Nomencl. 263. 1961), of the Tovinibtins or Cruciferae. The later name is an orthographic variant of the peers both having been designated in honor of a 16th century Spa physicia Simon a Tovar. "Female is replaced by the next oldest legitimate generic name, pri ies Raf. (Florula Ludov. 28. 1817; type species: A. racemosum Raf. = A. virginianum (L.) — & Vautier). 110 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 kia Convolvulus (L.) Dumort.), 2” = 20 and 40; P. scandens L. var. scandens (B. scandens (L.) Greene); and P. scandens var. cristatum (Engelm. & Gray) Gleason (B. cristata (Engelm. & Gray) Greene and B. dumetorum sensu Small), 2m = 20, are widespread in the Southeast and beyond. Polygonum cilinode Michx. (B. cilinodis (Michx.) Greene), of the northeastern United States, reaches its southern limit in Tennessee and North Carolina. In addition, the Japanese P. cuspidatum Sieb. & Zucc. (Pleuropterus Zuccarinii Small, Reynoutria japonica Houtt.), 2n = 4 and ca, 88, and P. sachalinense F. Schmidt ex Maxim. (Reynoutria sachalinensis (Maxim.) Nakai), flowering bamboo, are often cultivated for their vigorous erect growth and showy panicles of flowers. Both species persist after cultivation and are very difficult to eradicate due to an ex- tensive, spreading rhizome system. In the Northeast, escapes are becom- ing obnoxious weeds. Both species have been collected in Tennessee and North Carolina and may be expected elsewhere in our area in the future, if not controlled. Extrafloral nectaries are found in several species of sect. TrNTARIA (€.2-; P. Convolvulus and P. sachalinense). They consist of a circular to oval depression at the base of the petiole on the abaxial side containing multi- cellular, short-stalked, glandular trichomes. Fully developed before the leaf unfolds, the nectaries are thought to function as hydathodes, pro- viding a release for excess turgidity developed during formation of the leaf. REFERENCES: Under family references see DUKE, GEITLER, JARETZKY (1925), KNUTH, LAUBENGAYER, Mitra, RECHINGER, RIDLEY, SHARMA, SOUEGES (1919), TUTIN, VAUTIER, WEBB, WobEHOUsE, and Woopcock. Under Rumex see MAREK an WEILL & Bournérus. Antes, H. E., C. R. Bett, & A. E. Raprorp. Species new to the flora of North or ‘South Carolina. Rhodora 60: 10-32. 1958. [P. hirsutum, 13.] A. ADFORD. Species new to the flora of North Carolina. Jour. Elisha Mitchell Sci. Soc. 75: 140-147. 1959. [P. sachalinense, 141.] BRENCKLE, J. F. Notes on Polygonum (Avicularia). [I.] Photoperiodism and taxonomy. Bull. Torrey Bot. Club 68: 491-495. 1941; Notes on the Avicularia. 11. Phytologia 2: 169-171. 1946; Notes on Polygonum. ats. Ibid. 402-406. 1948; [IV.] On some Asiatic Polygonums. /bid. 3: 300- 303. 1950; Notes on Polygonum. V. Ibid. 361-366. ; Burk, C. J. Distribution records and range extensions from the North Caro- lina outer banks. beeen 26: 138, 139. 1961. [P. sawatchense, 139, redetermined by H. E. Ahles as P. argyrocoleon CouttTer, S. atopy in the genus Polygonum. Bot. Gaz. 17: 91, 92. 1892. Doma, Y. On the pollen grain formation in genus Polygonum. (In Japanese; English summary.) Bot. Mag. Tokyo 70: 31-37. 1957. . Cytological studies in Polygonum and related genera. I. [bid. 73: 337-340. 1960. . Consideration on the ming ge differentiation in Polygonum. 1-2. (In Japanese; English summary.) Jour. Jap. Bot. 37: 3-12, 81-88. 1962. Fassett, N. C. The variations ey Polygonum punctatum. Brittonia 6: 369- 393. 1949, 1965] GRAHAM & WOOD, GENERA OF POLYGONACEAE 111 FERNALD, M. L. The variations of Polygonum pensylvanicum. Rhodora 19: 70-73. 1917. FLanacin, V. L. A report on starch storage in septate fibers of Polygonum coccineum var. pratincola. Trans. Kans. Acad. Sci. 64: 304-310. 1961. Fucus, C. Sur le développement des structures de l’appareil souterrain du Polygonum cuspidatum Sieb. et Zucc. Bull. Soc. Bot. France 104: 141- 147. 1957. HepsBerG, O. Pollen morphology in the genus Polygonum L. s. lat. and its taxonomical significance. Sv. Bot. Tidskr. 40: 371-404. 1946. Henpricks, H. V. Torsion studies in twining plants. Bot. Gaz. 68: 425-440. 1919; II. bid. 75: 282-297. 1923. [P. Convolvulus.] Hom, T. Polygonum: sectio Tovara. Bot. Gaz. 84: 1-26. pls. 1, 2. 1927. Justice, O. L. Viability and dormancy in seeds of Polygonum amphibium L., P. coccineum Muhl., and P. hydropiperoides Michx. Am. Jour. Bot. 31: 369-377. 1944. KEARNEY, T. H., Jr. Cleistogamy in Polygonum acre. Bot. Gaz. 16: 314. 1891. [= P. punctatum. Li, H. L. The genus Tovara (Polygonaceae). Rhodora 54: 19-25. 1952. Love, A., & D. Love. Chromosomes and taxonomy of eastern North American Polygonum. Canad. Jour. Bot. 34: 501-521. Martin, A. = Identifying Polygonum seeds. Jour. Wildlife Manag. 18: 514- 520. 195 mae T. ne virginianum. Meehan’s Monthly 7: 181, 182. pl. 10. 1897. Reep, H. S., & I. Smoot. The mechanism of seed-dispersal in Polygonum virginianum. Bull. Torrey Bot. Club 33: 377-386. SALisBuRY, E. J. The Hee nectaries of the genus Polygonum. Ann. Bot. 23: 229-242. pl. 16. 1909. ScHouz, H. feel Saint fiir die Sammelart Polygonum aviculare L. Verh. Bot. Ver. Brandenb. 98-100: 180-182. 1960. ScHotsMAN, H. D. The anatomy of the glands of some pew rege and hybrids. (In Dutch; English summary.) Nederl. Kruidk. 262-276. 1950. [Includes P. Persicaria, P. lapathifolium, P. erent Suarp, A. J. Tovara in Mexico. Rhodora 54: 305, 306. 1952. [P. virginianum. ] SHINNERS, L. H. Polygonum bicorne Raf. instead of P. longistylien Small. Rhodora 59: 265-267. 1957 Stmmonps, N. W. Polygonum L. em. Gaertn. Jn: Biological flora of the British Isles. Jour. Ecol. 33: 117-143. 1945. [Description of sects. occur- ring in British Isles; list of fungi and insects associated with the genus; detailed treatment of life history and ecology of P. Persicaria, P. petecti- cale, and P. lapathifolium SMALL, J. K. A monograph Pr the North American species of the genus Poly- gonum. Mem. Dep. Bot. Columbia Coll. 1: 1-183. pls. 1-84. 1895 SovEces, R. Embryogénie des Polygonacées. Développement de l’embryon chez le Polygonum Persicaria L. Compt. Rend. Acad. Sci. Paris 168: 791— 793. 1919: Le développement de l’embryon chez le Polygonum aviculare L. Ibid. 178: 409-412. 1924. Sperry, J. J. The anatomy of the oe gay viviparum L. (Abstr.) Trans. Tex. Acad. Sci. 26: 55, STANForD, E. E. The inflorescence and Scars in Polygonum, subgenus Persicaria. Rhodora 27: 41-47. 1925. 112 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 . Possibilities of hybridism as a cause of variation in Polygonum. Ibid. 1-89. . The amphibious group oh aati: subgenus Persicaria. Ibid. 109- 112, 125-130, 146-152, 156-1 . Polygonum pti nition ‘and related species. Jbid. 173-184. Polygonum hydropiperoides and P. opelousanum. pot 28: 11-17, 22-29. 1926. . Polygonum Hydropiper in Europe and North America. Jbid. 29: 77-87. 1927. STEWARD, A. N. The Polygoneae of eastern Asia. Contr. Gray Herb. 88: 1-129. pls. 1-4. 1930 ———. The Ne for Polygonum cuspidatum Sieb. & Zucc. Rhodora 32: 223- 225.193 STYLEs, B. sig “The taxonomy of Polygonum aviculare and its allies in Britain. Watsonia 5: 177-214. 1962. Sucano, N., & K. Hayasut. Anthocyanin of the seedlings of a Polygonum. Studies on anthocyanins. XXXII. Bot. Mag. Tokyo 73: 231-233. 1960. [P. Hydropiper Wirts, K. J. The germination of Polygonum species in the field and in the glass- house. Jour. Ecol. 48: 215-217. 1960. [P. Persicaria, P. aviculare, P. Convolvulus. | 4. Polygonella Michaux, Fl. Bor.-Am. 2: 240. 1803. Heather-like, subherbaceous annuals or suffrutescent perennials, mainly of sand hills, scrublands, and waste places. Branches many, erect, adnate to the parent stem for approximately half the length of the internode. Leaves alternate, narrowly linear to spathulate, caducous or persistent, sessile, articulated with and appearing to arise from the free upper edge of the membranaceous, cylindrical ocrea. Inflorescence racemose, terminal on primary and lateral branches, the flowers solitary in the axils of im- bricated ocreolae and slightly to well exserted from them; ocreolae uni- laterally flared to an acute or acuminate apex, the margin entire or bear- ing several short to long, bristle-like hairs. Flowers bisexual or function- ally unisexual (the plants dioecious, gynodioecious, or gynomonoecious) ; pedicel broader and longer than the stipe and subtended by 2 lanceolate, keeled, scarious bractlets, or bractlets wanting. Tepals 5, slightly fused, forming a short floral tube, erect (the outer tepals occasionally tardily reflexed), in 2 whorls of 2 outer and 3 inner ones, or (in subg. Thysanella) included, alike (subg. T/ysanella) or the inner whorl of 3 dilated near the base forming 2 lateral nectariferous teeth (subg. Polygonella); anthers versatile; pollen prolate, 12-40, long, tricolporate, the pores equatorially arranged, the furrows longitudinally elongated, the exine reticulate. Gynoe- cium 3-carpellate; stigmata 3, punctate; styles 3, filiform, free to the base; ovary 3-angled, slightly stalked. Fruit a shiny yellow to brown or gray pyramidal achene. Embryo straight, in one of the angles of the seed; cotyledons oblong, thick and fleshy; radicle equalling or exceeding Oe ne eS yO ae | 1965] GRAHAM & WOOD, GENERA OF POLYGONACEAE 113 cotyledons. (Including Delopyrum Small, Dentoceras Small, and Thy- sanella A. Gray.) Type spectes: P. parvifolia Michx. = P. polygama (Vent.) Engelm. & Gray. (Name a diminutive of Polygonum.) — Jotnt- WEED. Indigenous to the eastern United States, with two subgenera and nine species, all but Polygonella Parksii Cory (endemic to Atacosa and Wilson counties, Texas) occurring in the Southeast. Subgenus THysaNeLta (A. Gray) Horton® (Thysanella A. Gray, Polygonella § Thysanella (Gray) Roberty & Vautier) has inner tepals fimbriate, filaments alike, and pedicels subtended by bractlets within the ocreolae. It consists of Polygonella fimbriata (Ell.) Horton var. fimbriata (T. fimbriata (Ell.) A. Gray), of Florida, Georgia, and Alabama, and var. robusta (Small) Horton (T. robusta ‘Small), 2n = 32, endemic to Florida. Subgenus PoLyGonELLa (§ Delopyrum (Small) Roberty & Vautier, § Dentoceras (Small) Roberty & Vautier), with inner tepals entire, the inner whorl of filaments dilated near the base, and pedicels not subtended by bracteoles within the ocreolae, composes the remainder of the genus. The species are distinguished by such characters as habit, type of margin on the ocreolae, number of ocreolae in a raceme, shape and size of leaves, and color and position of tepals during flowering and fruiting. The procumbent, mat-forming Polygonella myriophylla (Small) Horton (Dentoceras myriophylla Small) is endemic to the southern part of the Florida lake region. Polygonella ciliata Meissn. var. ciliata and var. basi- ramia (Small) Horton (Delopyrum ciliatum (Meissn.) Small and D. basi- ramia Small), both 2m = 22, are also restricted to Florida. Polygonella macrophylla Small, 2n = 28, occurs on the Gulf Coast of Alabama and northern Florida. More widespread are P. gracilis (Nutt.) Meissn. (De- lopyrum gracile (Nutt.) Small, Delopyrum filiforme Small), 2n = 22, 24; P. polygama (Vent.) Engelm. & Gray (P. brachystachya Meissn., P. Croomii Chapm.), 2m = 28; and P. americana (Fisch. & Mey.) Small. The first two are species of the Coastal Plain; the latter is generally pied- montane, occurring sporadically from South Carolina to central New Mexico. Polygonella articulata (L.) Meissn. (Delopyrum articulatum (L.) Small), 22 = 32, is the only species of the genus in the northeastern United States, reaching its southern limit in northeastern North Carolina. It is regarded as the most primitive species of Polygonella by: Horton, who also believes that the modern species of the genus evolved in Florida from an ancestral stock forced southward by Pleistocene glaciations. Internodal branching characteristic of Polygonella is lacking in the rest of the family. The wood anatomy is characterized by a lack of parenchyma in annuals and by axile xylem parenchyma in perennials. Starch is stored in the xylem in an unusual cell type which resembles a *The taxonomic treatment follows the recent revision of Polygonella by Horton. We = also relied on this paper for much of the general information ¢ con- cerning the genus. 114 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 fiber in shape and pitting, but is nucleate and living at maturity. The floral anatomy is, by reduction, the simplest in the family (Vautier). Each tepal has a single trace which connects with the main vascular supply; the other floral parts are supplied by single free traces which end in the floral tube. No anatomical evidence of a double trace to the fused inner and outer tepal can be seen in this genus. Reported chromo- some numbers form an aneuploid series (2m = 22, 24, 28, 32). REFERENCES: Under family references see LAUBENGAYER, VAUTIER, and Woopcock. Horton, J. H. A taxonomic revision of Polygonella (Polygonaceae). Brittonia 15: 117-203. 1963. [Includes distribution maps of each s SMALL, J. K. The relation between the genera Thysanella and “Polygonella as shown by a hitherto unobserved character. Bull. Torrey Bot. Club 24: 47, 48. 1897. [Internodal branching in Thysanella and Polygonella. | Subfam. COCCOLOBOIDEAE Dammer Tribe CoccoLoBEaE [Dumort. | 5. Coccoloba P. Browne ex Linnaeus, Syst. Nat. ed. 10. 997, 1007, 1367. 1759, nom. cons. Dioecious shrubs or trees [vines] of warm sandy coasts and hammocks; nodes often swollen or geniculate. Leaves evergreen [or deciduous}, thick and leathery [or membranaceous], alternate, variable in size and shape, ours reniform to orbicular or oblong to elliptic with apex obtuse and base cordate to acute, the surfaces minutely punctate [or the upper surface pitted and lower with superficial or sunken glands]; petioles short, attached to the stem near the base of [or well above] the partly deciduous tubular ocrea. Inflorescence a many- to few-flowered raceme [spike or panicle], less than to exceeding the leaves, terminal on primary and lateral branches. Flowers regular, greenish or yellowish white, unisexual or functionally so, campanulate, with a short floral tube, the stipe mostly equal to or shorter than the pedicel, the young buds completely covered by the ocreola which ruptures as they mature; pedicels [less than to] exceeding the length of the ocreola. Staminate flowers in 2—5-flowered, bracteate fascicles at the nodes, each flower subtended and surrounded by the persistent ocreola which incloses the adjacent flower and its ocreola, additional ocreolae of aborted flowers occasionally also present, the fila- ments long, exserted, the styles short, included. Carpellate flowers soli- tary at the nodes, subtended by a bract and surrounded by an ocreola, the filaments short, the anthers sterile, the styles long, exserted. Tepals 5, greenish white. Stunmas (7) 8; Slaments broadest at base and united into an annulus surrounding the ovary, without nectariferous teeth; anthers versatile; pollen prolate, ca. 50 X 40y, tricolporate, the furrows extending nearly to the apices of the grain, the pores slightly elongated equatorially, the exine finely reticulate, tectate-perforate. Gynoecium 3-carpellate; styles 3, the dilated apices stigmatic; ovary 1-locular, some- 1965] GRAHAM & WOOD, GENERA OF POLYGONACEAE 115 times incompletely 3-locular at base; ovule on a long funiculus. Fruit an achene with 3 major ridges, surrounded by the fleshy floral tube, the tube persistent, accrescent, rose-purple to nearly black when mature, the lobes covering the apex of the achene [or the lobes alone accrescent, surrounding the achene]; seed filling the upper part of achene, the lower part with spongy tissue. Embryo straight, the radicle short, the cotyledons orbicular; endosperm ruminate. (Coccolobis P. Browne, Civ. Nat. Hist. A are 209. 1756, invalidly published; Guaiabara Mill, Gard. Dict. Abr. . 1754, nom. rejic.) Lectotype species: C. Uvifera (L.) L. (Poly- 7 num Uvifera L .), typ. cons. (Name from Greek, coccos, berry, and lobos, lobed, in reference to the fleshy, lobed floral tube surrounding the fruit.) An indigenous tropical and subtropical American genus of about 170 species and one of the few tropical genera in the family. The greatest number of species (about 85) occurs in South America (about 44 in Brazil); two reach their northern limit in peninsular Florida. The sea grape, Coccoloba Uvifera, 2n = 132, a characteristic shrub or small tree of warm, sandy shores from Florida to northern Argentina, is im- mediately recognizable by its large orbicular to reniform leaves and the dense clusters of sweet-acid, rose-purple fruits (nearly black when fully mature). The pigeon plum, C. diversifolia Jacq. (C. laurifolia Jacq., C. floridana Meissn.), 2n = 22, a tree to about 20 m. tall, with generally smaller, oblong to elliptic leaves and nearly black fruits, inhabits ham- mocks and scrublands in Florida, the Bahamas, the Greater Antilles, and Antigua. Both range along the east coast of Florida north approximately to Merritt Island; on the west coast C. Uvifera occurs at least as far north as Manatee County, but C. diversifolia is restricted to the southern counties. Hybrids of the two are known from Everglades National Park, Florida, and Prov. Pinar del Rio, Cuba. Coccoloba Uvifera also hybridizes with several other species elsewhere in its range. The taxonomy of the genus presents numerous difficulties. Many species have at least two names, one based on a staminate, the other on a carpel- late plant, the unisexual condition having been discovered only recently (cf. Howard, 1949). Morphologically, the flowers vary little among species, while vegetative characters which, for lack of floral differences, must be used in classification, are extremely variable. In collecting Cocco- * Pre ap saan unreported chromosome counts by Prof. R. A. Howard, Arnold Arbo- retum: Coccoloba Uvifera &, 2n = 132 (meiotic count), R. A. & E. S. Howard 10211, Miami B ok. apg? County, Florida, 18 May 1948 (ny); C. rapa Zn = 22, seedling grown in Cambridge, Mass., from fruits of R. A. & E. S. Howard 8976, Arroyo del a Prov. San Juan, Dominican Republic, 15 Sept. 1946 (cx) An earlier count ae 2m = ca. 80 for C occoloba Uvifera (Edman, re Horti Berg. 9: 270. 1929) is open to question, for it is possible that this was obtained from either hybrid or apomictic material of C. Uvifera. Since the name C. diversifolia has been applied incorrectly to several different species, Jaretzky’s count of 2 = ca. 200 (Jahrb. Wiss. Bot. 69: 441. 1928) may have been obtained from another species. Apparently both reports are undocumented by herbarium specimens 116 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 Fic. 1. Coccoloba. a-i, C. Uvifera: a, fruiting branchlet, x ™%4; b, staminate flower, X 6; Cc, ca rpellate flower with staminodia, x 6; d, carpe ate flower in semidiagrammatic een section to show flo ral tu = ‘and basal orthotropous ovule, i seed ah off-center, cotyledons appearing in section, two to right, one to left (ct, “8 }. Xe 3:8 fot in diagrammatic cross section at approximate level . ; “ be spongy tissue, seed coat, and endosperm as in “f,”’ 3; h, seed, X 4; i, embryo, inverted from patie in seed, X 4. In “bd” note acter of stipe with pedicel. 1965 | GRAHAM & WOOD, GENERA OF POLYGONACEAE 117 loba care should be taken to represent the diversity of growth habit and leaf variability within the population. Adventitious shoots, which tend to have longer internodes and larger leaves than the normal stems, should also be included. Since the fruit of many species changes in shape as it matures, comparisons of size and shape can be made only with fully mature fruits A successful ornamental, Coccoloba Uvifera is planted along seasides in warmer regions of the world for its hardiness and attractive fruiting racemes and evergreen foliage. It has been reported in cultivation since 1690 and has become naturalized in several places in southeastern Asia and on some Pacific islands. Two variegated leaf forms are propagated. It is successful also as a pot plant, requiring little care except pruning. Jelly is sometimes made from the fruit. Coccoloba diversifolia, with bronze-colored juvenile leaves, is planted as a street tree in southernmost Florida. Both species are considered good honey plants, and birds feed on the fleshy fruits. Little work has been done on the biology of the genus, although hybridization apparently is frequent, and apomixis is suspected. Fossil leaves of Coccoloba are known from the Paleocene or Eocene of Alaska and also have been reported from Alabama, Mississippi, Texas, Arkansas, Louisiana, Tennessee, Kentucky, and Wyoming. REFERENCES: Under family references see DUKE, LAUBENGAYER, and VAUTIER. Ho.iick, A. The Tertiary floras of Alaska. U. S. Geol. Surv. Prof. Pap. 182: 1-185. pls. 1-122. 1936. [Coccoloba, 112, 113, a £27, 227] Hooker, W. J. Coccoloba Uvifera. Bot. Mag. 59: pl. 1 Howarp, R. A. The genus Coccoloba in Cuba. Jour. hela Arb. 30: 388-424. 1949. [C. Uvifera, 409; C. diversifolia, 423]; Studies in the genus Cocco- loba. II. The identification of Coccoloba Swartzii Meisner and Coccoloba barbadensis Jacquin and their relatives. Jbid. 37: 317-339. 1956; III. h i id. 38 38: 211-242. 1957; V. The genus in Haiti and the Dominican Republic. Ibid. 39: 1-48. 1958; VI. The species from the Lesser Antilles, Trinidad and Tobago. Ibid. 40: 68-93. 1959; VII. A synopsis and key to the species in Mexico and Central America. Jbid. 176-203, 205-220; IX. A critique of the South American species. /bid. 41: 213-229, 231- 258: 1960; . New species and a summary of distribution in South America. /bid. 42: 87-95. 1961; XI. Notes on the species in Asia. [bid. 107-109. a history of the genus Coccoloba in cultivation. Baileya 6: 204-212. 1958 LaMorre, R. S. Catalogue of the Cenozoic plants of North America through 1950. Geol. Soc. Am. Mem. 51. 381 pp. 1952. [Coccoloba, 128.] Lrnpau, G. Monographia generis Coccolobae. Bot. Jahrb. 13: 106-229. pl. 5. 1890 : Ze r Entwicklungsgeschichte einiger Samen. Ber. Deutsch. Bot. Ges. 9: 274-279. pl. 17. 1891. [C. populifolia, 276-279, pl. 17, figs. 7-13.] Nachtrage und Berichtigungen zu meiner Monographia Generis Cocco- lobae. Bot. Jahrb, 14(Beibl. 31): 14-16. 1891. 118 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 SarcENT, C. S. Coccolobis. Silva N. Am. 6: 113-120. pls. 298-300. 1894. YVINGLING, H. C. Some aspects of the anatomy of Coccolobis Uvifera (L.) Jacq. Ohio State Univ. Abstr. Doctors’ Diss. 40: 343-352. 1943. 6. Antigonon Endlicher, Gen. Pl. 310. 1837. Herbaceous to woody, perennial vines; stems lightly to densely pubes- cent, often with geniculate nodes; lateral branches leafy or modified as tendrils; ocreae short, with a nearly entire margin, membranaceous and falling away on old stems or reduced to a fine line; axillary buds 2, one forming a lateral branch, the other an inflorescence, one or both some- times suppressed. Leaves deltoid to cordate or hastate, the blades in ours not decurrent on the petiole. Inflorescence of terminal and axillary ra- cemes bearing lateral tendrils, especially on their distal portions, ~ terminating in a forked fendeil: flowers borne at the nodes, 3—5 in fascicle, the fascicles subtended by an acuminate bract and each ae subtended by a membranaceous, persistent ocreola. Flowers bisexual, pedicellate, the pedicel persistent, the stipe and pedicel nearly equal in length. Tepals 5, forming a short floral tube, bright rose- -red [ purplish crescent, enlarging 3 or 4 times, inclosing the fruit, and often becoming very showy, the outer tepals cordate to ovate, the inner oblong. Stamens (7-9), the filaments bearing scattered glandular-capitate hairs and alternating with small triangular, nectariferous teeth, the filaments and teeth united half their length to form a broad cuplike annulus around the ovary, the base of the annulus fleshy; anthers versatile; pollen sub- prolate, ca. 50 40y, tricolporate, the furrows extending nearly to the apices and with the membranes lightly flecked with scattered granules, the pores elongated equatorially, the exine thick, finely reticulate. Gynoe- cium 3-carpellate; stigmata 3, reniform-capitate; styles 3, filiform, gen- erally included; ovary 3-angled, slightly stalked, surrounded by the fleshy base of the annulus; ovule on a long, thin funiculus. Fruit a 3-angled achene. Seed with mealy, ruminate endosperm; embryo slightly excentric. (Corculum Stuntz, U. S. Dep. Agr. Bur. Pl. Indus. Bull. 282: 86. 1913.) LECTOTYPE sPECIES: A. leptopus Hook. & Arn.; see Hooker & Arnott, Bot. Capt. Beechey’s Voy. 308. pl. 69. 1840, and Britton & Wilson in Sci. Surv. Porto Rico Virgin Is. 5: 265. 1924. (Name from Greek, probably from anti, similar to, and gonu, knee, alluding to the geniculate nodes; cf. Polygonum.) — CORAL-VINE, MOUNTAIN-ROSE. A Central American genus of two or three species: Antigonon guati- malense Meissn. (A. macrocarpum Britt. & Small); A. flavescens S. Wats., perhaps only a greenish-white color form of A. leptopus; and the well- known A. leptopus (Corculum leptopus (Hook. & Arn.) Stuntz), Confeder- ate vine, 2m = 40, ca. 44, 48, a widely cultivated ornamental vine with cordate to hastate leaves and showy racemes of accrescent rose-colored flowers. Antigonon leptopus is planted throughout the warmer parts of 1965] GRAHAM & WOOD, GENERA OF POLYGONACEAE 119 the Southeast and apparently has become naturalized in several scattered localities in Florida. Its natural range is described as Baja California and Chihuahua, south to Oaxaca, Mexico. In Florida the plant blooms con- tinuously from early spring to late fall and may reach a height of 10 m., climbing by means of tendrils. Abundant viable seeds and edible tubers are produced, from which new plants sprout in the spring. The genus is insect pollinated, as the brightly colored tepals and copious nectar indicate. A white-flowered form (= A. flavescens?) is occasionally culti- vated. REFERENCES: Under family references see DUKE, VAUTIER, and WoDEHOUSE. BAILEY, -- Antigonon. Standard Cyclopedia of Horticulture. ed. 2. 1: 304. Hooker, I i“ “Antigonon leptopus. Bot. Mag. 96: pl. 5816. Orpetx, G. S. Coral vine, a perennial honey source. eat Bee Cult. 82: 656, 657. 1954.* Rao, V. S. A contribution to the morphology of Antigonon leptopus Hook. & Arn. Jour. Indian Bot. Soc. 15: 105-114. 1936. [Mega- and microsporo- genesis. | STANDLEY, P. C. Antigonon. In: Trees and shrubs of Mexico. Contr. U. S. Natl. Herb. 23: 247, 248. 1922 7. Brunnichia Banks ex Gaertner, Fruct. Sem. Pl. 1: 213. pl. 45, fig. 2. 1788 Perennial, woody, high-climbing, tendril-bearing vines of riverbanks and low damp ground, mostly growing in trees or thickets. Stems ridged [or nearly terete], glabrous or slightly pubescent; lateral branches short, ending in tendrils; axillary buds 2, one forming a lateral branch, the other suppressed or forming an inflorescence; ocreae short and entire, glabrous or pubescent, continuous with the base of the petiole or reduced to a row of hairs flanking each side of the base of the petiole. Leaves alternate, petiolate, ovate to oblong, the apex acute to acuminate, the base rounded to truncate. Inflorescence of densely flowered terminal and axillary panicles or racemes; flowers borne at the nodes, 3—5 in a fascicle, each fascicle subtended by a bract and each flower by a persistent, hyaline, lanceolate-subulate ocreola. Flowers bisexual, the stipe much longer than the short pedicel, the floral tube and stipe 3-ridged, with one ridge [or two] expanding in fruit to form a flattened, leathery wing. Tepals 5, whitish- or yellowish-green, oblong to elliptic, leathery, slightly united at base, forming a short floral tube, reflexed at anthesis, connivent in fruit, both tepals and pedicel accrescent. Stamens 8 (7-10); filaments ex- serted, persistent; anthers versatile; pollen subprolate, ca. 45 & 35n, tricolporate, the furrows extending nearly to the apices of the grain, the pores slightly elongated equatorially, the exine thick, reticulate, tectate- perforate. Gynoecium 3-carpellate; stigmata 3, obscurely bilobed and irregularly flabelliform; styles 3, exserted, persistent; ovary 3-ridged; ovule basal in the spongy tissue of the locule, the funiculus long and 120 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 runnichia. a-k, B. cirrhosa: a, flowering and fruiting branch, X on Qs eee disarticulated from pedicel, X 8; c, d, stamens, the anther-halves bent forward in “d” to show connective, < 16; e, two views of stigma, < 10; f-h, ary in Sublanratigoat vertical section to show successive changes in orientation of ovule, beginning at anthesis — note incompletely developed outer integument of ovule, X 8; i, mature accrescent calyx with wing, the achene not visible, X 2; j, achene in semidiagrammatic vertical section to show seed sur- 1965] GRAHAM & WOOD, GENERA OF POLYGONACEAE 121 thin, at anthesis anatropously bent so that the micropyle is directed downward, later straightening, the micropyle then directed upward (cf. Ic. 2).8 Fruit a 3—6-angled achene inclosed in the persistent, accrescent floral tube; seed filling upper part of achene, spongy tissue the lower part. Embryo embedded in sagen along one angle rh ing“ cotyle- dons oblong; endosperm mealy, ruminate. TyPE sPEC B. cirrhosa Gaertn. (Named for the Danish aia es iY Aesiaittey 1737-1827.) — EARDROPS, BUCKWHEAT VINE. genus of one American and two African species. The American Brunnichia cirrhosa, 2n = 48, is found generally on rich bottom-land soils and along riverbanks of the pipe eaclka United as from brunnichia Hutch. & Dalz., primarily on the basis of the two- rather than one-winged pedicels. The genus is most closely related to Muehlenbeckia Meissn. and is of no economic importance. REFERENCES: Under family references see LAUBENGAYER, RIDLEY, and VAUTIER. Damme_r, U. Zur Kenntnis der afrikanischen Brunnichia-Arten. Bot. Jahrb. 26: 347-357. 1899. [See also J. HurcuHinson & J. M. Datziet, Bull. Misc. Inf. Kew 1928: 28. 1928, and Fl. W. a. Afr. ed. 2. 1: 139. 1954.] Lewis, W. H., H. L. Srrrptinc, & R. B. Ross. Chromosome numbers for some angiosperms of the southern United Ses and Mexico. Rhodora 64: 147-161. 1962. [B. poreng 150. ] SMALL, J. K. Morphological notes on the genus Brunnichia. Bull. Torrey Bot. Club 21: 131, 132. 1894. [B. cirrhosa.] family characterized by orthotropous ovules, Brunnichia cirrhosa is re- markable in that the long, slender funiculus of the ovule is bent over at the apex, the part of the locule, while a spongy (flotation?) tissue occupies the remainder of the cavity. At maturity the seed has become so large that the funiculus is pushed down- ward, developing a crook or “kink” just beneath the seed. (Cf. Fic. 2, f-h, j. ientati ric thought that the ovule was at first ny aah ai later ing an anatropous position “which it does not seem, however, to keep up to the ioe of its development” (trans- a supplied). In B. cirrhosa the gate i oo bent long before anthesis, n in the very smallest buds which could be disse ee by spongy tissue (stippled) filling fruit, x 4; k, dia Race cross section of fruit and seed cut through c otyledons of embryo (uppe ht), the endosperm stippled, the seed coat hatched, the space between s oa be achene wall black — note three remnants of spongy tissue giving appearance of partial partitions, X 8. 122 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 LETTERS FROM CHARLES SPRAGUE SARGENT TO REGINALD SOMERS COCKS, 1908-1926 * Edited by JosEpH EWAN 82 May 6, 1915. You sent us leaves of Malus from Sardis under three numbers — 1, 2 and 3. These appear to us, however, to be all the same thing and until I went to Selma I thought they might be a form of Malus ioensis, but I col- lected flowers on the road between Selma and Sardis which look very much like the flowers of Malus platycarpa. Now what we want is the fruit to make the determination certain. I suppose you can manage that this sum- mer. I was surprised to find Persea palustris so common about Sardis as I did not know it ever got so far into the interior. I had a letter from Palmer this morning in which he said things were coming out very rapidly and he thought Tilia would be in bloom in a week. I hope therefore you will bear in mind the collections of flowers at Lake Charles and Avery Island. Palmer says there are more different kinds of trees and shrubs about Natchitoches than any place he has yet visited, although Crataegus is not represented there by many groups. How about the name for our Lake Charles Carya? I want to write the description and am depending on you for a very fancy and first-class name. P.S. — There is a man here now preparing a monograph of Smilax, so if you have any doubtful specimens we had best see them. Your Smilax 91 from Wakefield, June 1907, is Smilax lanceolata. 83 May 12, 1915. What is the Tilia situation? Of the three Lake Charles species we have flowers only of #2 and no fruit of 1, 2 and 3. Of the Sardis species which grows on the river bluff, your +6 of 1914, we have flowers and fruit. Of your other species (#41110) we have specimens only in very young bud, but a specimen of yours (no date and no number) is in flower; the label says “the neighborhood of Selma”. This is the tree I found at Selma but did not see at Sardis where I suppose you did not see it. I hope you will be able to manage the fruit of this second species this summer. It cer- tainly ought to be pretty fully grown before you return to New Orleans. Pretty bad times these but there is no use thinking about them when there are trees to think of. * Continued from volume 46, p. 44. 1965] EWAN, LETTERS FROM SARGENT TO COCKS 123 84 May 21, 1915. I congratulate you on having found such a fine tree as Magnolia acu- minata in Louisiana. I do not think it is surprising that it should be there but rather that it should have been overlooked so long. Please tell me the exact place where you found it, town if possible, so that I can enter it properly in my Manual. This does not seem such a remarkable discovery as the finding by Harbi- son of Magnolia in northeastern Mississippi. He is finding a great many interesting things in that state and it would not surprise me if Carya pal- lida reaches eastern Louisiana as he finds it so abundant in western central Mississippi. You ought to make a special hunt for this in West Feliciana. Carya callicoma I think sounds good and I am much obliged to you for the suggestion. 85 June 2, 1915. I have written out the description of C[arya] callicoma. Your #15 from Natchitoches, with very young leaves, looks as if it might be this species. I do not suppose you remember much about it. Is it possible that the Winnfield trees “in dry woods”, that is near the quarry, with young leaves pale tomentose below, can also be callicoma? I think Winnfield needs more investigation as to its Hickories. Your #1 Lake Charles, October 1913, is of course the same as your tree from the man’s yard, and undoubtedly C. arkansana. There is fruit which probably belongs to this species collected “‘near English Bayou” where we did not see C. arkansana. #11, 12 and 16 Natchitoches are C. arkansana, and #3 from Grand Ecore is probably the same. I hope soon to hear from you about Tiélias. 86 June 14, 1915. I am glad to get your letter of the 10th of June which has arrived just as I was going to write to tell you that among some specimens of Tilia sent by Palmer are three numbers from Lake Charles and all apparently our +1. He seems to have missed the other two numbers, at least there is no trace of any pubescence whatever on these specimens. They were gathered too soon but among the lot there are one or two open flowers. I very much hope therefore you have been to Lake Charles and have got the flowers of our three numbers. I should judge that you might have been too early at Avery Island as I have received from a young lady there specimens of what she calls the smooth and rough-barked trees. The flowers, however, will answer our purpose after a fashion. Is it possible that Liriodendron does not reach Louisiana? I have no record of it there and no specimen. 124 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 Can you tell me the range of Quercus velutina in Louisiana? While at Sardis you will look out, of course, for that peculiar Quercus coccinea with the deep cups. It will be interesting to know if this is a common form or only the peculiarity of an individual tree. I do not know what to say about that little Post Oak about Selma in which Harbison is so much interested. The leaves are very tomentose on the lower surface but I do not see that, except in habit, the shrub differs from the large trees in Selma. You have, I hope, the memorandum of the two Plums we found in flower near your place and south of Selma. As I think I told you, I am trying to arrange the Hickories largely on the absence or presence of yellow scales on the leaves and on the involucre of the pistillate flower and fruit. I had supposed that these yellow scales did not occur on the pistillate flowers and young fruit of Carya alba until the other day when I received from Harbison specimens collected in Mis- sissippi of what otherwise looked like Carya alba, but the young fruit was thickly covered with yellow scales. Such scales certainly do not occur on our tree as it grows here, and it is possible that the northern and southern trees ought to be separated. All this is preliminary to a request that you will look at Carya alba in the Selma region and see how the young fruit looks, and possibly dry me a few specimens, especially if you find the yellow scales. Instead of going to North Carolina why don’t you come to Brookline and do some work on the synonymy and bibliography of your Flora? 87 June 23, 1915. I never received a more interesting package of plants from you than the one which arrived this morning — very many thanks. Quercus 800 is surely Small’s Quercus austrina which we have been hunting for so long and is evidently a good species. Would n’t it be pos- sible to get fruiting specimens? 802, Quercus coccinea. Is this the tree on the bank with the peculiar cups I wrote you about? 804, Quercus nigra. This is a distinct form with thicker leaves which are usually three-lobed at the apex. I saw a good deal of it between Selma and Sardis, and I think it ought to have a varietal name. Tilia 788, Sardis, with leaves pale and glabrous below and prominent axillary tufts, seems to equal 2528, West Feliciana, and probably our 2 of Lake Charles. This I think is a common and widely distributed species in the southwest. Tilia 784 and 786, leaves green and glabrous below, axillary tufts very small, looks like 2534 of Lake Charles. 780, 782, 790 and 792, the Tilia with leaves white pubescent below. It would seem as if there were three and not two Tilias in the Selma- Sardis region as I had supposed. I only saw two last spring. Acer 2520 seems to be a form of the Sugar Maple which is common in southern Georgia and in Mississippi. The leaves are paler on the lower 1965] EWAN, LETTERS FROM SARGENT TO COCKS 125 surface but I can see no other difference. We have never had flowers or fruit of this. Quercus 2524, ee Quercus 2526, Q. a Tilia 2530, HEL, nm Charles,” on English Bayou at Miller Place. This has leaves densely pubescent below, but our #1 from Miller Place with checkered bark, collected April 12th, is perfectly glabrous. Tilia 2536 is our #2 of April last. Tilia 2534, bark like a Persim[m]on, is our #1, although you say #2? I do not think we collected your 2530 at all. The pubescence is more rusty than that on the leaves of our #3. This new Lake Charles tree may be one of the Avery Island species. Did you find it on the Miller place? You do not appear to have collected this time our #3 which has leaves white-pubescent below. The trees near the schoolhouse in West Lake Charles and the tree by the river bank in Lake Charles are our #3. This seems to be a common species and no doubt Palmer has the flowers. Let me hear soon what else you are finding of interest at Sardis which seems to be an inexhaustible field. 88 June 28, 1915. The Ptelea which we collected at Lake Charles in March 1911 looks exactly like the Ptelea rhombifolia of Heller and I think should be referred to that species, at least until we know more about the different forms of this troublesome genus. From eastern Louisiana I have one specimen of P. trifoliata collected by you but no other Ptelea at all from Louisiana. Have n’t you met with the genus in other places in that state, and how does it happen that Selma and Sardis, which contain almost every other known plant, cannot boast of a Ptelea, at least there is nothing from that region in our herbarium? 89 July 2, 1915. Very many thanks for the fruit of Carya alba which tells me just what I wanted to know. It is surprising that these yellow scales on the fruit of this species have always been overlooked. Perhaps, however, this is not more remarkable than that every American botanist has described the st of Epigaea repens as a dried capsule although it is really a fleshy erry. There is no one now in this country who knows so little about Tilias as I do, but I should think Palmer could look after the fruit of the Lake Charles species this autumn. As you have read in my last letter, there is what appears to be a fourth species there. Undoubtedly the West Feliciana species are the same as those which grow at Natchez and I do not think you need to worry much more about them. This leaves Avery Island from which we certainly ought to have good fruiting specimens. I wrote to the 126 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 lady who sent the specimens the other day from there, asking her to send me a specimen of one of her numbers showing the young branches but I have had no answer from her. I have no more idea than you do when the fruit of these things ripen. The Scarlet Oak which I want you to look into is a single tree growing on the edge of the bluff, or rather a little way back from the edge. We went from your place to a store, the post-office of Sardis I think, and then took the road to the river. We got out at a negro cabin near the river, walked across a field, and then turned to the left and followed along the cliff for some distance. This particular Oak cannot be more than three hundred yards from where we reached the bluff. The remarkable thing about this tree is the much swollen scales of the cup. If this sort of Scarlet Oak is common in the Sardis region, then it ought to be made a variety of Quercus coccinea. 1 think perhaps I had best send you one of the fruits we picked up under the tree. As you know, Quercus rubra also grows on this bluff, and very far south of any station I have known for it before. How about the two Plum trees of which I wrote you? I hope you will follow them up, and also Quercus austrina. We should be glad to get specimens of any of the woody plants growing about Sardis which you may have time to make. 90 July 22, 1915. You have n’t said anything about having collected at the time of your last visit to Lake Charles fruiting specimens of that remarkable Ash tree with pendulous branches which we saw there. If you did not collect fruit I can ask Palmer to get it when he goes there this autumn. How early do you think the tree is likely to drop its fruit, and can you give any specific directions for finding the trees? We only saw one tree, as I remember, with fruit on it and that was hanging over a cottage close to the road. This of course was as we went down the lake (not towards the bayou) but I should not know what to tell Palmer in order to find this particular tree. Anything new in the tree line at Sardis, and have you given up the idea of coming north this summer? 91 July 23, 1915. How about the Slippery Elm, Ulmus fulva, in Louisiana? It is common from Missouri to Texas and it ought at least to be in western Louisiana, but I have no knowledge of it there. 92 July 29, 1915. A new tree for the Selma region is Carya carolinae-septentrionalis which Harbison collected there this spring. Why can’t you send me now a specimen of the White Oak near your 1965 | EWAN, LETTERS FROM SARGENT TO COCKS 127 place which we thought was Quercus austrina? I should like to see it in its present summer condition. It might be sent without drying I should think. 93 August 7, 1915. I have your letter and the package of specimens, for both of which many thanks. Acer 820 and 828 are A. saccharum. This form with leaves pale and pubescent below appears to be a rather common tree from southern Georgia and northern Florida to Mississippi. Acer 822 is A. floridanum. Carya carolinae-septentrionalis is right. I have already written you that Harbison had found it near Selma. Quercus 814 looks as if it might be a hybrid of Q. imbricaria. Do you know more than one tree? We ought to know more about it and get more material. Quercus 812 Sardis must be Q. Durandii. Is this the small tree on the left-hand side of the road going from your house towards Sardis and at the foot of a rather steep hill where our motor got stuck? This was the tree we took for Quercus austrina. Please answer this. If your 812 is not our little tree, where does it grow? Have you noticed on 812 that the leaves on the lower branches are very different from those on the upper part of the tree? Is your Quercus 800 the same as the 804 mentioned in your letter? 800 looks as if it might be the mysterious Q. Aybrida of Chapman. The Tilia cultivated at Lake Charles is Tilia tomentosa from Hungary. I saw only two Lindens in the Selma-Sardis region but there are evi- dently three species, — Ist, the tree with leaves silvery pubescent below, your 780, 782, 790 and 792. 2nd, the tree with very coarsely serrate leaves, your 784 and 6 of July 1914. I did not find this tree and I do not know if the young leaves are all pubescent, and I am afraid that you do not. 3rd, the tree with leaves not so coarsely serrate as #2, covered below with pale stellate hairs in April and in June entirely glabrous with the exception of the conspicuous axillary tufts. This is your 788. 786 is the same but without the tufts. This was the only Linden I saw on the bluffs of the river at Sardis but I saw plenty of #1 south of Selma. I think now, thanks to you, that we have complete material of these three Lindens. Can’t you induce your friends at Avery Island to collect good fruiting specimens of Lindens there? There are probably two species, one a tree with pubescent branchlets and rusty pubescent leaves which are very scabrate above, your #6 of October 1910. I do not know that we have either flowers or fruit of this. It seems similar to the tree from the coast of Georgia which has passed for T[ilia] pubescens, but I daresay it is not that species. Two, a tree with leaves pale pubescent below and also rough above. This is Miss MclIlhenny’s smooth-barked Linden and your specimen of May 24, 1914. Can this be the same as #1? The dif- ference in the time of year when the specimens were taken may account 128 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 for the different color of the pubescence. Miss McIlhenny’s rough-barked Linden may be the same as #1 but the specimen is so poor that it does not tell much. Palmer is now in St. Louis arranging his collection and I am afraid can- not get off much before the first of September, but I shall write him to get back to Louisiana as soon as possible. The Florida and Alabama Oaks seem to be as puzzling or more puzzling than the Lindens. I wish you were here now for I am living all alone and we might get through a great deal of work together. 94 August 10, 1915. I have received from you a box of Plum fruit i no word as to what it is. I suppose I should hear from you on this subje Your Wakefield, West Feliciana “Tilia americana,” June 1907, is evi- dently the same as Selma #2, and the same as your specimen collected near Alexandria June 1905. I cannot quite follow you on the Lake Charles Tilia. Your #2532 with pubescent branchlets and leaves pubescent below agrees with our #3 of last spring, but your #1 from the Mills’ place (42530) has the same pubescence as your number 3 (2532) although the branchlets are gla- brous. Now unless there is a bad mixup on my part it is impossible that the tree in Mill’s yard which had perfectly glabrous leaves on April 12th should on May 21st have leaves densely pubescent below. I am writing Palmer today to get off as soon as possible and go first to Lake Charles for evidently there is no Linden at Natchitoches which does not also grow at Lake Charles. I hope you are hunting Oaks. 95 August 19, 1915. Many thanks for your two notes and the information they contain. I hope you will continue to watch the Oaks at Sardis. You have said noth- ing about that peculiar form of Quercus coccinea about which I wrote you. If the leaves of Tilias are going to change from glabrous to pubescent as they grow older I am afraid it will be hopeless to find characters by which they can be separated. Too bad about Avery Island for I suppose the weather won’t become cool there or mosquitos disappear until it is too late to collect Tilias. I hope the cotton crop is going to be a success. 96 August 21, 1915. I enclose a leaf of an Oak which Harbison found at Pleasant Hill, near Selma, and which may be an extreme form of Quercus Durandii. He found similar trees at other places near Selma, and I have asked him to write and give you as near as possible the locality of these trees in the hope that 1965 | EWAN, LETTERS FROM SARGENT TO COCKS 129 you will be willing to investigate them before you leave Alabama and obtain mature specimens. According to Harbison Pleasant Hill is a very interesting place, and possibly you may like to visit it. 97 August 23, 1915. Many thanks for yours of the 19th and for the Plums which have not yet arrived but which I suppose will be here tomorrow. It seems to me that this is about the worst of all the groups of our woody plants. We are sending you by express to Tulane a set of woody plants col- lected in Louisiana by Palmer this year, or rather nearly a full set for I understand he has been obliged to leave some of them behind him at St. Louis in his pursuit of Tilia fruit. I will arrange that the herbaceous plants be sent from St. Louis. I take no responsibility for Palmer’s names, and a few of them appear to be incorrect. Notice his Quercus 7473 with leaves only from Natchi- toches, which looks as if it might be Q. Durandii. His Quercus 7448 from Natchitoches seems to be the three-lobed form of Q. nigra common be- tween Selma and Sardis. It needs a varietal name. His Q. stellata var. Grand Ecore 7518 and Chopin 7978 we are much interested in here. The pubescence seems to be like that of the Post Oak in Selma. I do not know if I wrote you about this tree and the importance of collecting fruit of it this autumn, as well as the little dwarf Oak in sandy soil near Selma which is evidently a dwarf form of the larger tree. Palmer makes a point that the bark of his tree is scaly like that of the White Oak and not rough like the Post Oak. Is this true of the Selma tree? Palmer’s Quercus velutina 7443 is stellate-pubescent on the upper sur- face of the leaves. This sort of pubescence so far as I have seen never occurs on the leaves of the straight velutina but is conspicuous on the variety missouriensis of mine from Missouri and Arkansas. On the Mis- souri tree, however, the branches are much more tomentose than Palmer’s 7443, but in Mississippi and Florida there is a velutina without tomentose branches but with stellate pubescence on the upper surface of the leaves at least early in the season. This may be another variety. It has not yet turned up in Alabama and I hope you will be able to look at the velutinas in your region and see if you can find any trees with such stellate pubes- cence. Harbison claims that the trees in Mississippi with the stellate pubescence do not have the orange colored inner bark. This is a fact, however, which needs further investigation. 98 October 20, 1915. Many thanks for your letter of October 15th. No, I have n’t been ill but away on several short journeys this autumn and my correspondence has fallen into a hopeless condition of neglect. I was on the point of writing you when I heard of the storm in Louisiana but did not get around to it. I hope you did not suffer personal inconvenience, but I am afraid 130 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 many of our trees at Lake Charles were destroyed. I hope, however, the Hickories are all right. Have you been out this autumn with Palmer? He wrote me that you thought of doing so. I have n’t heard from him for some time and I fancy that he is now in eastern Texas or Oklahoma as he went early from Louis- iana. I am sorry to say that he found that many of the Lindens from which he collected in the spring were blown down early in the summer. Please let me hear from you soon. a9 November 1, 1915. Many thanks for the collection of Oaks from the Selma district. The more I see of these Selma Oaks the less I know about them and this last collection of yours seems to me the worst lot I have seen. 904 is Crataegus tomentosa. The other specimen of Crataegus in fruit I do not know. 950. I do not know this Oak. Is it some form of the Water Oak? Do you remember the tree? 954. Quercus velutina. 972. Quercus austrina, if there is such a thing. 962 I suppose is the dwarf form of the new Post Oak with scaly bark. As you know, this also grows at Natchitoches. 934. Is this not the common Post Oak? It shows none of the pubes- cence which is so conspicuous on the under surface of the leaves of the new Post Oak. How is the bark of this 934? 952, 968, 960 and 932 seem to be QO. Durandii. 930. Is this Q. austrina again? 938 looks as if it might be a hybrid of Q. imbricaria. It is unknown to me. 956 is the new Post Oak. 958 is nearly glabrous and I suppose is the common Post Oak. I am going to take up the Oak question seriously when Harbison’s col- lection of the past autumn arrives and try and come to some conclusion about some of these troublesome southern species, varieties or hybrids. It is rather curious that this scaly barked Post Oak grows at Selma and then not again until we get to Natchitoches. By the way, Palmer writes that he thinks he has found Gleditsia texana at Natchitoches. This would be a new tree for Louisiana. Apparently anything can be expected in the Natchitoches region. I hope you are all right. 100 November 12, 1915. We have been able to raise two new Chinese Citrus plants which may prove useful. The first of these, Citrus ichangensis,’® has fruit something ° Burkill includes this in C. macroptera, as a race of thick-skinned shaddock-like Citrus. 1965 | EWAN, LETTERS FROM SARGENT TO COCKS 131 like a grapefruit although rather smaller. Some persons think that the fruit has even a better flavor than that of grapefruit but this of course is doubtful. The great point about it is that it is likely to be much hardier than the grapefruit or any other Citrus plant with edible fruit. The fact that our plants are raised from seeds gathered from wild plants growing in the mountains of western China suggests that the plants may be freer from disease than many of the long cultivated forms. The second plant is an Orange and still unnamed. It is said by Wilson to be the best of the tight-skinned Oranges of China. It is a low level plant in the Yangtse Valley but should prove hardier than the ordinary cultivated Orange. I could send you fifty or one hundred plants of each of these if you can use them to advantage among your friends in Louisiana. I think these trees should be introduced. Let me hear about this as soon as you conveniently can. 101 November 22, 1915. I have yours of the 20th and you may be sure we shall send you a set of Palmer’s plants as soon as they arrive. He is now working on them in St. Louis and I expect something from him very soon. It would certainly be remarkable if you have found M[agnolia] cor- data in Louisiana. As you know perhaps, it has recently been rediscov- ered in Georgia. If you have specimens of the leaves I should like to see them as they throw some light on the identity of the plant. We are arranging to send you two hundred and fifty of the Ichang Orange and two hundred and fifty of the Ichang Lemon for distribution among your friends in Louisiana. We can send more if you want them. The Arboretum will take the liberty of paying the express charges on the case to New Orleans as you ought not to be put to any expense in this matter. 102 December 7, 1915. The Citrus fruit started from here last night in two cases by express and I have no doubt will reach you in good condition. They have been in- spected and I do not think there can be anything the matter with them. Now if you can use more of these plants to advantage in Louisiana we can send them to you. I am anxious, of course, to get them as widely dis- tributed as possible. I advise you to try a few of Citrus ichangensis at your place in Alabama. This ought to prove, judging from the locality where the trees grow naturally, the hardiest of all the Citrus species with edible fruit. If you want more you better let me know soon before winter really sets in. Crataegus 964 from Sardis, 1915, is evidently C. tomentosa, although I have n’t seen the flowers, and the same as your 14 Sardis, 1914. These are the only Alabama specimens of this widely distributed tree we have 132 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 and I have no other record of its extending into Alabama, so this is an- other tree the distribution of which you have greatly extended. 103 December 14, 1915. I have just been looking over the Louisiana specimens of Malus. Your Pineville specimen, our +1 west of Opelousas, the Crowley speci- mens and your specimens collected at Natchitoches April 15, 1911 and April 15, 1912, and Palmer’s Natchitoches 7252 we are placing with M{alus| ioensis var. Palmeri. This leaves now only our Malus 42 west of Opelousas, a tree thirty feet high with red-brown bark. This is cer- tainly something different. It may be an extreme form of M. angustifolia but I hardly think so. I had supposed you collected flowers but I do not find them, although you visited the tree the second time to collect fruit. The flowers if possible should be collected in the spring. I have just written Harbison that we have n’t got to the bottom yet of the Selma-Sardis species of Malus and I hope he will visit that locality again next spring. 104 December 16, 1915. In writing you yesterday I made a mistake, of course when I said we had referred your Pineville Malus to ioensis var. Palmeri. It is, of course, the type of ioensis creniserrata. 105 December 18, 1915. I am very glad indeed to hear that the Citrus plants reached you in good condition. We have plenty more of them and I shall be glad to send you another supply if you can use them to advantage. I do not suppose the Orange is going to prove of much value but the so-called Ichang Lemon I believe will turn out to be a plant of much economic value. Yes, I remember about the Crataegus from Louisiana and will get to it before long I hope. Ashe is a man who is apt to make a good deal of trouble and I suggest that you be cautious about sending him material. He has a way of not keeping specimens on which he makes species and then forgetting them himself. How about the Magnolia cordata? 106 December 22, 1915. We have no specimen of Quercus cinerea collected in Louisiana by you, and only one specimen in the state recently collected at Natchitoches by Palmer. Does not this tree occur in eastern Louisiana and have n't you seen it in other places in western Louisiana? 1965] EWAN, LETTERS FROM SARGENT TO COCKS 133 107 December 31, 1915. I have been looking over your Sardis specimens of Quercus. #800 with “bark like the White Oak” is what Harbison feels sure is Small’s Q. austrina, but the old fruit which you picked up from the ground with this specimen is probably Q. texana. The following are also austrina: 804, 958, 956, 930, 950, 934, 902, 932. This seems to be a good species and to be widely distributed. You ought to find it in eastern Louisiana. 944, 960, 972, QO. Duranditi. 898 and 912, ea oe new variety. 942?, 900, O.r 914, ‘Sardis, teh like a hybrid and should be followed up. I cannot guess at its parentage. 932 might be Q. Durandii or austrina. It is sometimes very hard to dis- tinguish these from leaves ong ne. 802 I take to be Q. texa 2524, fea ot. Francisville, Louisiana (no fruit) must be a form of Q. falca Po is Feces for today. 108 January 8, 1916. I have your note about Quercus austrina. I confess that without acorns I cannot tell austrina from Durandii with any degree of satisfaction. Ap- parently what we are calling austrina according to Harbison is sometimes a large tree, and the tree which he considers austrina certainly always has scaly bark. Small, as you know, is often rather vague in his descriptions. Probably there is a good deal more to learn about these trees. P.S.— Do you know anything about an Oak collected by you in St. Tammany, July 1900, No. 6, and called by you Quercus tomentosa or Q. velutina? It is not that species and the leaves are good Red Oak leaves, and the fruit is good coccinea fruit. It is probably a form of the Red Oak with cups shaped like the northern form (Q. ambigua or borealis). I should think this was a tree worth further investigation. 109 January 13, a I enclose a leaf of an Oak specimen gathered by you in New Orleans October 1911. Was this tree planted? This is the same tree as the one which is growing at Natchez, Mississippi, and of which Harbison has found a single planted individual by an abandoned house near Selma. With the leaf is an acorn from one of the Natchez trees. At Natchez most of the leaves are still on the trees and perfectly green, so that if this tree has been planted in the streets of New Orleans you ought to be able to recognize it at this time. The same species grows on the Texas coast and it would be surprising, I think, if it does not turn up in West Feliciana. 134 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 It is certainly an undescribed Oak and one of the largest Oaks of North America. It is well worth hunting for, therefore, in Louisiana. I have only just found out about the persistent leaves which add a good deal to the interest of the tree. 5 I enclose also a copy of Buckley’s description of his Quercus Shumardu which has always been overlooked or disregarded. Judging from his de- scription, it is a tree with foliage a good deal like that of Q. texana but with cups as shallow as those of the Red Oak but covered with very thick scales. I enclose a couple of cups of what I suppose is this tree from Fulton on the Red River in Arkansas. Our Quercus texana collected last April at Lake Charles may be this tree, and Palmer appears to have found it at Natchitoches and Grand Ecore. Your Sardis 1906 and 1915 may be the same thing. ; I doubt if 1900 and 942 of Sardis are Quercus rubra, but your Sardis specimen of August 1913 is certainly this last. The trees which I think are Shumardii have more deeply lobed leaves than Q. rubra and large and very conspicuous axillary tufts of hairs on the lower surface. Typical texana has deep cups with thin scales. Unless some character besides the shape of the cup and the thickness of their scales can be found to separate Q. Shumardii from Q. texana I should think that the former was a variety of the latter, but this group needs a great deal more observation in the field. It is true that Q. rubra evidently grows in Feliciana, but I do not feel at all sure that it is in western Louisiana. It is important to collect more material of the Red Oak group about Sardis if you are there again, and particularly in Louisiana. I am writing to Palmer to pay especial attention to it in Arkansas, Okla- homa, western Louisiana and Texas this year. I should think that the Oaks of West Feliciana especially needed attention. I feel almost sure that you will find the new Natchez species there. If this tree from New Orleans, of which I now return a leaf, was planted where did the trees come from? Probably not very far from the city. This, however, is not a swamp tree so possibly it did not come from the immediate neighbor- hood of the city. Do you remember the peculiar Post Oak with white scaly bark growing in low wet ground west of Opelousas? The leaves are not pubescent on the lower surface like the peculiar Post Oak of the Natchitoches region, and apparently they lose the stellate pubescence from the upper surface very early in the season. We have never had the mature leaves of this Opelousas tree and fruit has never been collected; this is a Louisiana tree which certainly needs further investigation. Either it is a new species or a very distinct form of the Post Oak. I thought at one time the Hickories were bad but I must acknowledge the Oaks are more troublesome, and certainly they are now less well known than the Hickories. Quercus Shumardii.— {This name is followed by the direct quotation of Buckley’s original description, which ends the letter. | 1965] EWAN, LETTERS FROM SARGENT TO COCKS 135 110 January 15, 1916. Referring to the Opelousas Malus, No. 1 is from the clump which the negro carried you to see but which I did not see. This we call ioensis var. Palmeri. No. 2 is the solitary large tree with red-brown bark which we saw later but further east. This we first collected October 10, 1913. We have from you a specimen called No. 2 collected in April 1914, but there is not a trace of a flower on this specimen. On the label of your No. 1 collected April 1914 you say “large tree.” This I take for granted is the negro tree. On the label of No. 2 you say “large tree, not in clumps.” Don’t you think it would be a good plan for you to revisit Opelousas this spring and get good flowering material of this No. 2 which probably has very little to do with variety creniserrata? This will give you an op- portunity, too, to take a look at the peculiar Post Oak which does not grow very far from it and quite near the Quercus ludoviciana. 111 January 21, 1916. I have just finished looking through Palmer’s autumn collection of Oaks. His Natchitoches 8761 and 8762 represent my idea of Quercus Shumardii. The buds show that this tree is in no way connected with Q. rubra, but I expect it will be found to be a variety of Q. texana. This last is well represented by Palmer’s 8835 from Chopin. His Windsor 8935 has, as you see, shallower cups and appears to be approaching Shumardii. His Q. breviloba, Natchitoches 8718, is Q. Durandii and this is the first speci- men of this species which I have ever seen from Louisiana, although it used to be said to grow on the banks of the Red River at Shreveport. If it does grow there I have never seen specimens. This fine tree can now, I think, safely be added to the Louisiana flora. The next time you go to Covington would it be too much trouble to cut for me a piece of a stem of the //licium about six inches long and as large a piece as can be conveniently found? It is wanted here for one of my associates who is studying the anatomy of woods. Have you ever been to Port Hudson and do you know what grows in that region? I should like to go there again myself for it is fifty-three years since I saw it. Have you ever been up the railroad northward from Laurel Hill? This goes on to Woodville, Mississippi, which must be an entirely unknown Oak [region] botanically. 112 January 22, 1916. I have your two notes of January 18th and the two cups. The more shallow one with the much thickened scales is my idea of typical Q[wercus | Shumardii. The specimen with the thinner scales seems to me to be a shallow-cupped texana, that is it has the thin scales of that species. What 136 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 I think is that these things run together and that Shumardii must be considered a variety of texana. It has been planted a good deal in the streets of New Orleans and I have collected there acorns with deep and with shallow cups. You say that your two trees can be distinguished in the field but how? It would be very useful if we could so distinguish Shumardii, otherwise than by the thickened scales of the cup and keep it as a distinct species. How large are the specimens of the Natchez Oak 1! in New Orleans? If these planted trees came from Covington it ought not be too very difficult to find these species growing wild there. Have you noticed whether the New Orleans trees are still retaining their leaves? I enclose an extract from Palmer’s last letter. Do you know these regions? “In Winn Parish, some distance n. e. of Natchitoches, there is a peculiar region with limestone bluffs and saline springs, so different from any of the surrounding country that it should produce something of botanical interest, also in going north from Natch. I passed through a region of nearly pure sand some miles south of Bienville where Quercus brevifolia seemed to be very common. I only saw it from the train but wished I could have gotten into it. If you are in that part during the summer and it is convenient I should very much like to spend a day or two with you about Natchitoches and Chopin, where there are a great many good things to be seen. The accommodations at Natchitoches are very good and I spent sufficient time about there to know something of the surrounding country.” Have you noticed whether the New Orleans trees are still retaining their leaves? 113 January 26, 1916. Many thanks for the package of specimens. The two Oaks are just my idea of Quercus texana. Malus +2 from Opelousas certainly should be followed up. It seems intermediate between angustifolia and ioensis and nearer the former. The bark as I remember it is different from that of any of our American species. This may be something new. We have a specimen of Magnolia collected by Mohr near Mobile with the cordate leaves of your specimen. The Mobile specimen has always been considered an acuminata, and I think it is probably a variety of that species. The leaves are not at all like what we call cordata. This of course is a tree to follow up. 114 February 8, 1916. I have been trying to group all the Tilia material in this herbarium with the following results as far as your specimens are concerned: Alabama. #1, bluffs of the Alabama River. A species with glabrous branchlets, thin cordate leaves stellate-pubescent below throughout the season. I collected specimens last April near Sardis and you collected it in ™ Q. comptonae Sarg. eS 1965} EWAN, LETTERS FROM SARGENT TO COCKS 137 1914; no other date on the label and no number. This species has not been collected and these are urgently needed. #2, Tilia heterophylla Ventenat, not of Sargent’s Manual, = T. Michauxii of Nuttall but not of Sargent’s Manual. To this species I refer your numbers 780, 782, 790, 792, and many specimens collected by Harbison. #3, a glabrous species with very coarsely serrate leaves, your 6, 784, 786, and 788, “near Selma, 1914,’’ and Selma, June 12, 1914. This might pass for a green-leaved form of 7. floridana if the leaves were not so coarsely serrate. If these specimens had been found in the north they would certainly be referred to T. americana. I feel very doubtful about this tree of which we have good material. Your specimen from West Feliciana, June 1907, called T. americana?, is evidently the same thing. This makes three species for Selma. Louisiana. +1, branchlets glabrous, leaves stellate-pubescent below through the season. Lake Charles, your 2530 from the Mills Place. I had supposed that the Avery Island plant with leaves pale pubescent below belonged here but I find on the upper surface of the leaves the remains of stellate-pubescent hairs. These are very unusual on the upper surface of Tilia leaves and the Avery Island plant may be something entirely distinct. I much want to see the very young leaves of your Lake Charles tree and fruiting specimens of the Avery Island tree. I have placed with the Lake Charles tree Palmer’s following numbers from Natchitoches Parish, — 8747, 7946, 7397, 7474, 7554, 8013, 8021, 7420, and the specimen col- lected by you at Natchitoches April 15, 1911. Palmer’s 7675 from Welsh is probably the same as it has glabrous branchlets. All these specimens are very incomplete and we ought to have more and better Lake Charles material of this number. +2, a perfectly glabrous species with leaves green or more or less glaucous on the lower surface. This is our #2 of last spring from Lake Charles, and Palmer’s 7644, 7673, 8523, 7574, 7342, 7694, 8510, your 2534, your Lake Charles specimen of March 26 and May 12, 1911, and your Shreveport specimen of June 1908. What I believe is the same species but with leaves more or less glaucous on the lower surface are Palmer’s 7923, 8699 and 7569. These specimens with glaucous leaves I do not know how to distinguish from T[ilia] floridana of Small, and I do not see how to distinguish the glaucous and the green-leaved specimens specifically. We have only very incomplete material of this +2 from Lake Charles. #3, a species with glabrous branchlets, young leaves stellate-pubescent but soon becoming glabrous, and usually conspicuous tufts of axillary hairs. Your 2528 from West Feliciana seems to be this species which is our #3 of Lake Charles and Palmer’s 7523, 7956, 7554, 7479, 7474, 7952, 7397, your Natchitoches specimen April 27, 1911, your Lake Charles 2530 and 2536, Palmer’s 8511, 7695, your Lake Charles October 1914, your Winnfield April 6, 1913, our Opelousas April 3, 1913, your Laurel Hill March 1910. This seems a widely distributed species in Oklahoma, eastern 138 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 Texas and Arkansas, and gets into Mississippi. Our Louisiana material is still very incomplete, especially that from Lake Charles. +4, a tree with pubescent branchlets and leaves rusty tomentose below. Your 2532 (3 L. C.) from Lake Charles, and Palmer’s 7674, 8500, and perhaps 8494 but of this the branchlets are glabrous. I suppose that the “Tilia pubescens” of Avery Island belong to this species. There are no fruiting specimens of any of the above numbers and they are urgently needed. I should say then that there are certainly four species in Louisiana. If, as I suspect, the Avery Island specimen with young leaves stellate- pubescent above is a species, there would be five, and if the glabrous tree with very coarsely serrate leaves from West Feliciana can be separated from T. floridana, then there are six species. The important species to work up are the two species from Avery Island, the fruit of +4 from Lake Charles and the young leaves of your Lake Charles #3. It is important to compare young leaves of all the species because the conditions of their early pubescence may be very helpful in distinguishing the species. I had hoped that we had reached the end of the southern Tilias but evidently we are still very far from that I fear. It is astonishing how the peculiarities of the southern trees have been overlooked or neglected. There is confusion I am sure about that Mills Place Lake Charles number 115 February 10, 1916. I am sending you under another cover the cup of a very small Oak fruit and a cup and nut of a larger one. These seem to have been both gathered by me at Laurel Hill when we were there together. We have young leaves which belong to the small cup and these leaves seem to be Q. texana. The shape of the cup is all right for that species but of course is exceptionally small. The large acorn and cup would pass here at the north perfectly for a Red Oak. Unfortunately there are no leaf specimens which belong with this acorn. It appears to be in an exceptionally good condition if it was picked up from the ground in March, which was the time we were at Laurel Hill. If the Red Oak gets at all into Louisiana I suppose it would be in West Feliciana, and I am sending you this material in the hope that it may help you in studying this group. Of course it is desirable to determine, if we can do so, the true range of the Red Oak. Undoubtedly what has been called Red Oak in the south is often Shumardi. Ames and I are starting about the 10th of March for a short trip in Arizona, but my idea is to stop a day in New Orleans to talk matters over with you and then on the way back to stop again, if you think it desirable to do so, and return home by Natchez. All this, however, can be decided when I see you in New Orleans on the way west. 1965 | EWAN, LETTERS FROM SARGENT TO COCKS 139 116 February 19, 1916. I have yours of February 15th. Please take another look at your Tilias 784 and 788 from Selma. The flowering specimens look rather different but I cannot find much difference in the fruiting specimens. How do you distinguish these two trees? The truth is we have got to live with trees in order properly to understand them. Herbarium material is really of very little good except as a record. These two numbers look very much like the Tilia from west Louisiana. In order to meet engagements further west Ames and I would have to be in New Orleans on the morning of March 16th when I hope the crowd will have left. We had best write, I think, for our accommodations at the hotel. I should like to devote a part of the day to going over with you your specimens of Tilia and Quercus. Our plan now is to be back in Louisiana in April, but about this we can confer on March 16th. 117 February 23, 1916. I have just been through all the Hickory material of the 1915 collections. The most interesting thing for us now in Louisiana is your Natchitoches +6 of 1913 and 1914, for which I long ago wrote out a description, but I had so little material that I did not do anything more about it. I believe that Palmer’s Natchitoches 7258 (see this), which he called C[ arya] alba, may be this same tree, and I believe also that the two small trees which we found in dry woods just before we got to the quarry at Winnfield April 6, 1915, may be the same. We thought at the time they were C. alba, but of course they have nothing to do with that species. These trees ought to be followed up, and if you cannot get there in the autumn perhaps Palmer can I suppose that those miserable Monroe Hickories which have bothered us so much, judging by the buds, all belong to C. arkansana, but they did not have good arkansana sn It would pay to get spring material which would show the young lea In spite of the very peauily shaped fruits which used to puzzle us so much I now believe that the following Natchitoches numbers must be referred to C. arkansana: Cocks 10, xi. 12, 16, 20, Palmer 1 and 3, also Chopin 1 of Palmer, Lake Charles 1 of Cocks, and Cocks and Sargent, Lake Charles (yard of J. Mills). Your Grand Ecore 3 of 1913 may also be C. arkansana with small globose nuts. It is important to see your leaves of this number. Can you find the tree or tell Palmer how to do it? Grand Ecore, Palmer 7524, Creston 7016 Palmer seem to be C. arkansana. 1 and 18 of Natchitoches which I had placed with C. megocarpa |sic| are evidently C. arkansana. I feel very doubtful about 2 and 5 of Natchi- toches and I should like to see young leaves of these numbers. 5 may be C. arkansana but I am doubtful about 2. It is possible that the flower- ing and fruiting specimens are of different species. 140 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 I hope you can get together all the specimens about which I have been writing you lately so that we can discuss them on the 16th of March when we shall not have any too much time I fear. 118 April 24, 1916 I got home Saturday night after a successful two days in Natchez and one day in the neighborhood of Jackson where unfortunately Harbison did not turn up. ; Miss Compton has now located in Natchez twenty-nine or thirty indi- viduals of the new Oak. All the old trees are in the neighborhood of build- ings but some of these must be so old that I am inclined to believe that they must be original trees left standing when the forest was cleared. We saw, however, two wild trees, one growing in a field far from any houses and probably thirty or forty years old, and one much smaller in the woods. The relationship of the trees certainly would be Live Oak and I believe should be looked for on high ground and not in swamps. The largest specimen near Natchez is fully one hundred feet high and another tree has a trunk about five feet in diameter. Several of them are much larger than the tree we showed you on the Duncan place.” It is as evergreen as Quercus virginiana and is certainly one of the finest Oaks in the United States. I hope you will be able to locate the wild trees in Louisiana. The Maple tree from which we collected on our way to Natchez and which you thought was Acer leucoderme is evidently the same form 0 Sugar Maple that grows in West Feliciana and is evidently a distinct form. I shall hope soon to hear about your visit to Sardis. I enjoyed our days together immensely and certainly West Feliciana can produce as fine trees as are to be seen in any part of the United States. 119 I think it would pay you when you are in West Feliciana Parish in the autumn to look into the Beech question a little and collect good fruiting specimens from a number of trees. I collected only two specimens this spring and these represent the northern species, F[agus| grandiflora, and the southern variety, var. caroliniana. The curious thing about it is that the under surface of the leaves is very pubescent. The pubescent grandiflora has been described as forma pubescens and apparently has only been collected before in New England. What is more interesting perhaps is that the leaves of the specimen of the variety caroliniana are also pubescent on the lower surface. This has been described as forma mollis. Apparently it has only been collected once before, at Tallahasseé by Nash in 1895, and in Louisiana (New Orleans) by Drummond 1m 1832. * «Duncan Park” of Dr. STEPHEN Duncan of Natchez. : f i E ; 1965] EWAN, LETTERS FROM SARGENT TO COCKS 141 This collection seems to give you two new varieties for Louisiana. It would not be a bad idea to collect Fagus in any part of the state you may visit. I hope you are finding the new Oak. P.S. — The right name of the shrub we found so common and naturalized by the roadsides near Laurel Hill is Pyracantha crenulata Roemer. 120 May 10, 1916. Many thanks for the package of specimens which arrived yesterday. Tilias. 820. This is the species with leaves stellate-pubescent on the lower surface while young, becoming glabrous, and small but conspicuous axillary tufts. This seems to be the commonest Linden of Louisiana and southern Arkansas. 822. The glabrous, early-flowering species I suppose. There is a trace of axillary clusters on this particular specimen, the flower-buds appear although the leaves are not much grown. S54 = 822. 836, T. heterophylla. 832 = 822 and 834, I believe. 4014 is what we have been calling T. floridana. 4016, the species with pale pubescence on the lower surface of the leaves. This, according to the original description, should be T. leptophylla. 4012 = 4016. 4010 = 4014. 2530 = 4016. Alabama 842 = Malus angustifolia. Alabama 844, Amelanchier. Mature leaves and fruit desired if not already sent. Quercus 838, Sardis. This certainly outdoes Shumardii as we have seen it in Louisiana. I notice that you are quite cautious about giving names to your Oaks and I do not know how to tell the young leaves of austrina from those of Durandii. 1 hope you can do so. Notice your 840 which has stellate pubescence on the upper side of the leaves. Is this a Post Oak? What is 830 from Pine Bank Creek with its long-stalked fruit? Can this be the Natchez tree or is it what you call Q. austrina? I hope to hear soon from you again. 121 May 23, 1916. Much obliged for your letter of March [sic] 18th. I am glad to hear that you found Carya pallida in Louisiana. I notice what you say about the lateness of C. alba. 142 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 I have noticed in different places in the South individual trees of what I supposed was C. alba which were very much later in coming out than is usually the case with this tree. I have not been able, however, to follow up any of these trees to find out if they had other peculiarities. I suppose you can recognize in the autumn these late trees and it would be well, of course, to collect fruit and leaf specimens. These may show some characters which would make it possible to distinguish the late trees as a variety. I do not think I ever wrote you about Quercus velutina. My variety missouriensis was based principally on the stellate pubescence on the upper side of the leaves. I saw at Jackson this spring a small tree with these stellate leaves but the inner bark was not yellow like the inner bark of the common form of Quercus velutina. It is possible that this stellate pubescence and the gray inner bark may have some relationship and make it possible to distinguish another form of this tree. Harbison has noticed the same thing in Mississippi and western Florida, and I think it would be well for you to bear this in mind and keep a lookout for Black Oaks. There is certainly an undescribed species of Malus at Sardis, your numbers i., ii. and iii. The leaves on the flowering branches of this tree look exactly like those of M. angustifolia but the mature leaves, especially on the shoots, resemble those of some of the forms of M. ioensis. Possibly the angustifolia which you sent me the other day may belong to this tree. I have n’t got time to look at the specimens this morning however. I will write you later on this subject. Harbison says that the young leaves of Q. austrina are perfectly glabrous. 122 June 1, 1916. Have you ever collected Halesia carolina in Louisiana? If so, we have no specimen in our herbarium. My Manual speaks of it as growing through Louisiana to eastern Texas and Arkansas. West of the Mississipp! we have one specimen collected in northern Arkansas, nothing from Louisiana or Texas. It appears to be common at Selma. It is possible that the description of the range has got to be modified. 123 June 8, 1916. I am very glad indeed to get at last the Avery Island Tilias. They do not look at all as I thought they were going to and seem to be the same as three common species at Lake Charles. So much the better, for the fewer species we can get on with the better. What do you know about the distribution of Asimina triloba? We have Louisiana specimens only from Wakefield and Natchitoches. Does it grow in eastern Louisiana, and how does it happen that it does not grow at Sardis? We have no Alabama specimen whatever in the herbarium. 1965] EWAN, LETTERS FROM SARGENT TO COCKS 143 124 June 19, 1916. I have yours of the 14th from Sardis where I hope you are not suffering too much from the heat. I am just going to begin to arrange this spring’s collecting of Tilia and will write you again on that subject. I have given up trying to understand Q. Durandii and Q. austrina. Harbison, however, thinks he knows these trees as far off as he can see them and I am enclosing your note to him. The leaves of Q. Durandii vary more on the same tree than those of any other Oak I have seen. On the lower branches they are very often deeply lobed and higher on the tree entire, and sometimes green and glabrous below or pale and pubescent, although I believe the pubescent leaves are more common than the glabrous ones. I will ask Harbison to write you just where he found the Natchez Oak near Sardis and I hope you will be able to locate the tree and find others. Palmer has failed to find it where Mohr collected in Texas twenty-six years ago, so we know it only as a tree growing near human habitations. 125 June 30, 1916. Perhaps you will be interested in reading the enclosed from Harbison. Please let me have it back sometime. Notice what he says about the difference in Tilia at the top and bottom of the trees, and also about Quercus. You and he between you ought to be able to get to the bottom of this Quercus austrina business. 126 July 5, 1916. In spite of all our efforts the Louisiana Tilia material here is still very incomplete and unsatisfactory, as you will see by the following. My present idea is that Louisiana specimens can be grouped as follows: A. The perfectly glabrous species without axillary tufts of hairs. Lake Charles. Cocks, May 12, 1911, no number, good flowers. May 21, 1915, 2534, good flowers. Palmer, 7644 in bud, 8523 sterile. Sargent, river bank, April 12, 1915. #1 yard of C. J. Mills, sterile. River bank, April 13, 1915, #1, sterile. St. Francisville. Sargent, April 12, 1916. Laurel Hill. Cocks, March 1910, sterile. Natchitoches Palmer, 7569 in bud, 7523, good flowers. These Palmer specimens have leaves glaucous below but are probably a form of species represented by Group A. 144 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 Please notice that we have n’t a single Louisiana fruiting specimen of this group. The tree is common in Mississippi and reaches Texas and Alabama, and your Sardis 786, of which we have three fruits I think is the same. B. This is the species with large leaves very oblique at the base, early in the season very thin, bluish green, lustrous and slightly pubescent below, especially on the midribs and veins, usually soon becoming glabrous, axillary tufts small, often wanting. My +2 of Lake Charles, April 13, 1915, is my idea of this species. We also collected it March 26, 1911. It is your 4014, Mills Place, April 3, 1916 Lake Charles. Palmer, 7694 in bud, 8510, sterile (bracts but no fruit), 7695 in bud, 8511, sterile. Avery Island. Cocks 4050 and 4052. Natchitoches. Palmer 7574, sterile, 8699, sterile, 7956 and 7963 with flowers. Ish elsh. Palmer 7673, sterile. opin. Palmer 7342, sterile, 7970, poor flowers. Grand Ecore. Palmer 9449, sterile. East Opelousas. Cocks 401, sterile. Shreveport. Cocks, 10, June 1908. D. Coty, no date. West Feliciana. Cocks, 2540, flowers and fruit. I should have liked to have made my #2 of St. Charles the type of this species for the young leaves look so distinct, but if you can get fruit of Avery Island 4052 that might be taken as the type. This tree grows also in Texas and Arkansas but we have no very good material from those states. Sardis. Cocks 822 and 834, April 20 and 22, 1916, and 788, flowers and fruit complete I once Gok 788 belonged to Group C but the leaves are very oblique and bluish below. The tufts of hairs are prominent on this specimen, but I do not know how to distinguish these numbers from the Avery Island tree. Are 822 and 834 from the 788 trees? It would be well to get fruiting specimens of these numbers. Please write me, too, the sort of place in which this tree grows, the height of the tree approxi- mately, diameter of trunk, nature of bark, branching, and any other in- formation you can supply as we may have to take the Sardis plant as the type of the species in case we cannot get more Louisiana material. ieee ee eee 1965] EWAN, LETTERS FROM SARGENT TO COCKS 145 C, Leaves more or less cordate, dull green when young on the lower surface, more pubescent than those in Group B, becoming glabrous, axil- lary tufts large and conspicuous. Lake Charles. Cocks 2536, flowers just opening May 21, 1915. West Lake Charles. Cocks, October 1914, no number, sterile. Near Alexandria. Cocks, June 1905, in flower? Shreveport. Palmer 9479 in bud. Opelousas. In rich woods, Sargent, March 29, 1900, and April 3, 1915, both sterile. Natchitoches. Palmer 7523 and 9416, both sterile. St. Francisville. Cocks 2523 in bud. Sardis. Cocks 820 and 832, April 20, 1916, sterile. We should have more specimens of these num mbers D. Leaves mostly very oblique at the base, rusty pubescent below during the season Lake Charles. Cocks 3, 2530, flowers, fruit and young leaves, and 2532. The branch- lets of 2532 are pubescent and those of 2530 are perfectly glabrous. How about getting fruit of 2532? Avery Island. Cocks 4054 in flower May 24, 1914, is evidently the same but the branches are more pubescent. October 18, 1900, a ree shoot. Welsh Palmer 7674, 8500, 8495, all sterile. E. Leaves smaller than those of the last group, mostly cordate, densely pale stellate-pubescent below during the season but the pubes- cence is not closely attached and can be easily rubbed off. Lake Charles. Sargent 3, April 13, 1915. Several sheets, all sterile. One specimen taken near the schoolhouse in West St. Charles [West Lake Charles in- tended] has pubescent branchlets; those of the other specimens are glabrous. Natchitoches. Palmer 7420 in bud, 7946 in flower, 8747 sterile, 7474 in bud, 7952 in flower, 7479 in bud, 9416 sterile, 8013 sterile, 8021 sterile, 7397 sterile. Cocks, April 27, 1911, in bud, no number; April 15, 1911, sterile, no number. Chopin. Palmer 7554 in bud. Welsh 146 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 Palmer 7675, sterile. East Opelousas. Cocks, 4312 and 4016, sterile. Winnfield. Sargent, April 6, 1913, sterile. We have no fruit of this tree from Louisiana, and you do not appear to have collected any specimen in the state. The specimen of 2530 taken April 3, 1916, at the Mills Place seems to belong to E rather than to D. Are there not two species on that wretched Mills Place? This seems to make five species in the state, although it may be diffi- cult to separate B and C with the material we have At Sardis there is a Linden with oblong finely serrate thin leaves pale pubescent below during the season and obliquely cordate at the base. We have a sterile branch collected by you in 1914, no date and no number. I collected what I think is the same on the bluff near Sardis April 19, 1915, sterile shoots only. I don’t think this is your 788, and I don’t find that you have sent flowers or fruits. Possibly your Sardis 22 and 23 belong here. These two sheets look abnormal. The leaves are very long-pointed, very coarsely serrate and oblique at the base. These trees should be followed up and more material obtained. Your Selma 784 and 6, and also several other unnumbered sheets col- lected in 1914, bother me a good deal for I cannot find any way to separate them from T [ilia| americana. T. americana, however, is a dis- tinctly northern tree as we now understand it and fees not appear to extend south of Pennsylvania and northern Missouri. Please look at these trees and see if the leaves are lustrous below and if there is any pubescence on them except the axillary tufts. I am afraid that your specimen from Wakefield, Louisiana, June 1907, is the same. Have you ever heard anything from our Laurel Hill friends and are they collecting any material for us? Don’t you think it would be a good plan to ask them to get fruiting specimens of all the different Lindens in that neighborhood? Memorandum. Tilia material needed before these trees can be properly described. A. Fruiting specimens from Louisiana and Sardis 786. B. Fruiting specimens from Avery Island 4050 eine 4052, and from Lake — where it appears to be a very common ies. C. Fruiting specimens of Cocks’s 2536, Lake Charles, and fruiting RPE ies at Sardis 820 and 832. Fruiting specimens of Lake Charles 2532, Avery Island 4054. E. Lake Charles fruiting specimens. Sardis. Fruiting specimens of Cocks’s 22 and 23, also of the species collected by Sargent on the bluffs April 1915. Additional material of 784 and 6. 1965 | EWAN, LETTERS FROM SARGENT TO COCKS 147 127 July 12, 1916. Of the Quercus of your last sending two specimens (no numbers) called Durandii and austrina are both Durandii. 958 is my idea of austrina. 960 I suppose is a good Durandii. These species are easily distinguished by the 820 I should call Durandii. 824, probably apne 826, Durandii. 840, Durandii. 830 I do not know but it may be aust 958, 970, 956, 950 and 934 appear to be some iy : the Post Oak. The small tree by the stream at the foot of this hill where we were stuck, and called both austrina and Durandii, I now think is the Post Oak with scaly bark. 4022, West Opelousas, is the Post Oak like that at Natchitoches. I enclose another communication from Harbison. Between you I think you ought to be able to clear up this Oak business this year. I take it that your Sardis Tilia 952 and 954 is the one I collected on the bluff near Sardis about which I wrote you the other day. This tree apparently grows at River Junction, Florida. We have no flowers or You called your Sardis 950 T. floridana. Have you ever seen this before? How does your Hickory which was so late in putting out its leaves look now? If you stay late enough in Alabama it would be a good idea to get some acorns of your peculiar hybrid Oak 938 to plant, as these hybrid Oaks are interesting and it is possible that it might be hardy here. Too bad that you and Harbison did not connect! 128 August 7, 1916. It is a long time since I have heard from you although Harbison wrote me the other day that he had a letter from you telling him of the floods in your region. I hope your crops are not ruined. Farming in the South seems to be a very precarious business. The young leaves of Quercus austrina as collected by Harbison are perfectly glabrous and I think all the specimens which you have sent under that name are either Durandii, which seems to be a common tree with you, or forms of the Post Oak. I trust you will be able to get a good supply of Quercus material so that we can settle this year some of these doubtful points I have pretty much come to a conclusion about Lindens, or most of them, and will write you on the subject in the course of a few days. I wish you would take a look at Magnolia glauca and perhaps make some specimens. Harbison has recently found a shrubby form with perfectly glabrous peduncles, while in the form which grows to a large tree the peduncles are densely covered with snow white silky hairs. At the north the plants are shrubby and the peduncles are glabrous. Most of our 148 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 southern specimens have silky peduncles, especially those from Natchi- toches collected by Palmer. I do not think we have any specimens from Sardis and shall be glad to have them. If it is any hotter in Alabama than it is here today I am sorry for it. 129 August 9, 1916. I have this morning yours of en ae You must have had a horrible time and I fear that your crops are I think we are slowly getting a little light on the Tilia situation, but it is a tedious and difficult job because material is often so fragmentary. There are certainly a number of new species. I notice that your 2530 from Lake Charles has perfectly glabrous branches and that 2532 has pubescent branches. We have flowers and fruit of 2530 but only flowers of 2532. I hope therefore that the new Louisiana collections may include fruit of this number. These numbers must represent the same species in spite of the branches. The lower surface of the leaves is covered with matted stellate pubescence which is easily rubbed off and not like the close persistent tomentum which you find on several of the Sardis species. This Lake Charles tree, which is also common in the coast region of Texas, does not seem to extend east of the Mississippi River. I hope soon to hear that things are in better shape in your part of the world. 130 August 26, 1916. Thanks = yours of the 21st which I find after a short visit in Illinois and St. Loui No, that eae won’t do for C. aprica. It is no doubt an undescribed species. I went to St. Louis to see Palmer who was arranging his spring collec- tion. He has a good deal in the autumn to do in Texas and I have told him that it was not absolutely necessary for him to get back into western Louisiana this autumn. He thought, however, he might do it. He reports very large trees of Quercus Durandii in Natchitoches, so that is another Louisiana tree. Apparently they are doing nothing in St. Louis in the way of distributing his plants and claim they are not going to distribute anything until they can make a critical study of them. I am sorry about this and if you can designate exactly what you need I will make another attempt to get them for you. If you get this note drop me a line as to where you are to be in Virginia as I am going there myself in September and it is possible we might meet. 131 September 28, 1916. I have yours of the 19th. I wrote you on August 9th & 26th & told you of 1965] EWAN, LETTERS FROM SARGENT TO COCKS 149 the safe arrival of the specimens & suggested that we might meet in Virginia whence I have just come. Have you specimens of the late Carya alba about which you have written me or if not can you get them? I will write a real letter when my secretary gets home. 132 November 16, 1916. It is a long time since I have heard from you and I hope that silence does not mean that you are entirely swamped by those Alabama Oaks and Hickories. The Oak business seems to be endless, for now Harbison feels quite sure that he has found a new species near Vicksburg. I have n't, however, seen his material. I am writing especially today to tell you that we are sending you by mail two plants of a Chinese vine, Sargentodoxa cuneata. We think here it is going to be a good thing as in China, at least, it grows vigorously and has fragrant yellow flowers. Will you “place’’ these two plants where they will be likely to be taken care of? I am rather anxious about these because I am afraid this plant will not prove hardy in the north. Hoping to hear from you soon, I am, 133 November 18, 1916. According to Harper, “Mr. Mattoon of the U.S. Forest Service claims to have found Pinus caribaea in eastern Louisiana this year. Dr. Mohr did not seem to have any information about its occurring in Louisiana when he mapped its range, and I did not see any of it west of the Pearl River when I visited New Orleans in 1905.” Have you any views one way or another about this tree in Louisiana? 134 December 30, 1916. I very much fear there are three species of Tilia on Avery Island. First, Tilia floridana, Cocks 4052, May 29, 1916, and 4042, July 28, 1916. Second, Tilia ambigua, n. sp., Cocks 4050, May 29, 1916; Miss Mcllhenny, June 8, 1915, ‘“‘the smooth barked Linden.” This is a common western Louisiana species and our No. 2 of Lake Charles. It is one of the trees which has been placed under lettophila [sic], gcease name I now want to abandon as the identification is not possi Third, Tilia Cocksii, Cocks (4054) May 29, 1916, (4040) July 28, 1916. “The rough barked Linden,” Miss Mcllhenny, June 8, 1915. This looks superficially like some of the Lindens from the coast of South Carolina and Georgia which have been called Tilia pubescens but is distinct from them in its glabrous branches and buds. I have seen nothing like it except these Avery Island specimens and I wish we could find it in other localities. I hope, however, that we are going to Avery Island in 150 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 the spring and then perhaps we shall be able to see more of these trees. This is only the first paragraph of the Linden dose I am going to send you very shortly. With best wishes for the New Year, I am, 135 January 2, 1917. I have determined some of your Alabama Tilias as follows: T. Harbisoni, n. sp., near Selma, 1914, Cock’s 780, 790 and 792. This is a common west Florida species and I think a good one. T. floridana var. oblongifolia, n. var. 788 type, 820, 832, 834, 952, 954, 958, also near Selma 1914. This is a common Louisiana tree about which I shall write you later. T. nuda, n. sp. 95, June 6 and July 28, 1916; 954 (= 956 of 1915), June 28, 1916. Where is 956? I do not find it here. This is a perfectly glabrous species which is common in Mississippi, Louisiana, Arkansas and eastern Texas. Selma appears to be its most eastern station. The trees at Natchez, Mississippi, I have taken as the type. Hutchins’ Run, Cocks’s 970, July 26, 1916, with partly grown fruit. This differs from all other species of the United States’ Tilias which I have seen in the long hairs on the fruit, and if this is a constant character I suppose this is a distinct species. If you can locate the tree so that Harbison can find it in early spring; flowers and mature fruit will have to be collected either by him or by you. 964, 966 and 968, Hutchins’ Creek, Cocks, June 27, 1916. This looks like what we have been calling T. heterophylla but flowers are needed. Can you locate these trees for Harbison? 962 Hutchins’ Creek may be the same but the branchlets are much more slender and the buds smaller. 782, limestone cliffs of the Alabama River at Cocks’s place, June 15, 1915, has slender light yellow branches and seems different from any of the others. 836, young leaves, April 21, 1916. I do not connect this with any of your other specimens. What is it? Your Selma 1110, May 5, 1913, which seems the same as Selma June 30, 1915, I cannot place. Complete material is needed. I am still in doubt about your Sardis species which we have thought might be T. americana, represented by your +6 of July 14 and #784 of June 2, 1915, by specimens collected in June 1914 and by #960, July 25, 1916. Specimens from West Feliciana and from Alexandria seem to be the same. The axillary tufts of hairs which are so conspicuous in Tilia ameri- cana are smaller and often wanting from your specimens. Can you tell me if your #822, April 20, 1916, is from one of these trees? If not, Harbison ought to look them up and get young leaves. I suppose you can tell him where to find them. The fruit appears to be usually de- pressed-globose. I do not know, however, how to really distinguish these dried specimens from T. americana, but central Alabama is a long way from the home of T. americana. 1965 | EWAN, LETTERS FROM SARGENT TO COCKS 151 When we see how difficult it is to get complete material of these south- ern Lindens it is not at all surprising that they have been misunderstood and neglected for so many years. I am afraid there is still a good deal of field work to be done on them. If you can give Harbison the informa- tion how to find the doubtful numbers he can go there this spring and obtain the young leaves, always supposing that you were unable to do so. Of course you could probably find the trees much more easily than he could. In case you can not go will you get into communication with him? 136 January 4, 1917. Will you send me your full name so that it can appear with the description of Tilia Cocksii, also can you give me the Christian name or names of Miss MclIlhenny who sent me specimens from Avery Island? As you know, there are on the Louisiana coast region a number of islands similar to Avery Island. Have they ever been explored botanically, and don’t you think this ought to be done in connection with this Tilia business. It seems fair to presume that the vegetation on these islands would be very similar. We certainly ought to try to find Tilia Cocksii in more than one place. 137 January 6, 1917. I have referred the Louisiana species of Tilia as follows: Tilia ambigua, n. sp. Our Lake Charles #3, Avery Island, Cocks’ 4050; Chopin, Palmer 7554; Creston, Palmer 7420; Welsh, Palmer 7675, 8494: near Winnfield, Cocks and Sargent, April 6, 1913, East Opelousas, Cocks 4012 and 1016; on Little Bayou Teche, Cocks 4018. Please notice that we have neither flowers nor fruits of this species from Lake Charles. We have a very poor flowering specimen from Avery Island but no other flowers from the state, and fruit only from Little Bayou Teche. T. floridana, Small, West Lake Charles, Cocks, October 1914 (no number) ; Lake Charles, Cocks 2536, 2530, 4014; our #2 of Lake Charles; Palmer, Lake Charles, 7695, 8511. Notice that from Lake Charles we have partly expanded flowers on your +2536 but no other flowers and no fruit at all. Avery Island, Cocks, 4052 and 4042. Grand Ecore, Palmer, 9449; Chopin, Palmer 7970, 7342; Natchitoches, Palmer, 7574; Shreveport, Palmer, 9479, 10608; Welsh, Palmer, 7673; East Opelousas, Cocks, 4010, 4020; West Feliciana, Cocks, 2540; St. Francisville, Cocks, 2528; near Laurel Hill, Cocks and Sargent, April 12, 1916; West Plantation, Bayou Sara, Cocks and Sargent, April 14, 1916. There is good material from Avery Island. 4020, East Opelousas, has good fruit. 2540, West Feliciana, has a few open flowers. All the other specimens are very incomplete. 152 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 Tilia floridana, var. oblongifolia, n. var. Laurel Hill, Cocks, March 1910; Natchitoches, Palmer, 9437, 8699, 9416, 7956; Grand Ecore, Palmer, 7523; Chestnut, Palmer, 9462. Natchitoches 7956 has good ae si there are no other flowers and no fruit of any of the Louisiana speci Tilia rhoophila, n. sp. Lake Chars, Cocks 2530 (I am doubtful about this), 2532; Welsh, Palmer, 8500, There are good flowers of 2532 aed no ies and all the Welsh material consists of young shoots. 2530, Lake Charles, has glabrous branches and buds and may be 7. Cocksii. This species, if it is one, needs investigation in Louisiana. Tilia nuda, West Feliciana, Cocks, 4024; Lake Charles, Cocks, May 12, 1911, 2530, 2534, Palmer 7644, 8523, our #1 of Lake Charles. T. nuda, var. glaucescens, n. var. Natchitoches, Palmer, 7569, 7923. The type of this new variety is Cocks Sardis 786. We have no flowers or fruit of the variety from Louisiana, and only two fruits from Sardis and none from Texas With T. Cocksii this makes five Louisiana species without the Sardis species with coarsely serrate leaves like those of T. americana which, judging by the two fragmentary specimens, also grows in Louisiana. On the whole our Louisiana material is not at all satisfactory. When we remember how much time and trouble we have taken to get together the small amount we have it will perhaps be understood why these southern species have been so much neglected and overlooked. Certainly there is no other genus which is more difficult to collect material in a saitsfactory condition. A common Lake Charles tree our #2 with leaves at the end of May very thin, blue-green and nearly glabrous below is interesting. It belongs with the floridana set but perhaps should be distinguished as a variety. Palmer 7694 and 8510 from Lake Charles may be the same but I am not sure, and probably Natchitoches 8697 may be the same form. We have no flowers and no fruit of this tree from Lake Charles and we ought to try and get them. I take it for granted that you know the form I am talking about as you collected it at Lake Charles March 26, 1911 (no number). I wish you would think over the situation and decide if possible if you will be able to obtain some of the material we need this year. The im- portant things to get are flowers and fruits of this Lake Charles Tilia floridana, and material and information of T. rhoophila from Walsh. Palmer has a good deal of Tilia work to do in western Texas this year and it is desirable that it be arranged so that we will not be obliged to go into Louisiana at all this season. I take it for granted that you will be able to get off for a week or ten days about April Ist for a trip into western Louisiana, and also for your proposed expedition with me into the country adjacent to the Illinois Central Railroad. Apparently there are no Lindens in East Louisiana except in West Feliciana; we at least —— ER ate ss tea a 1965] EWAN, LETTERS FROM SARGENT TO COCKS 153 do not seem to have any specimens from that part of the state. How about this? As the nature of the pubescence is the best character by which to distinguish the various species it is essential to collect in the spring the half-grown leaves. It is necessary therefore to visit trees three times, in the spring for young leaves and in the summer and early autumn for flowers and fruit. The color of the branchlets and the character of the winter-buds have been almost entirely overlooked, and it will be laborious and expensive to get these in addition to the spring and summer material. All these southern trees ought to be grown in New Orleans where they could be easily seen and ih ee studied. You need an arboretum there very badly. 138 January 12, 1917. I enclose a provisional conspectus of the United States species of Tilia based on the material already in hand. It will require considerable chang- ing probably when we are successful in getting all the material we need this year. I am sure we have n’t got to the bottom yet of the distribution of these trees. For example, we know nothing of Lindens in Louisiana east of the Mississippi River with the exception of those in West Feliciana; certainly there must be others. And we know nothing of Lindens in any of the coast region of Mississippi or Alabama, Selma and Sardis being the most southern station in Alabama from which I have seen specimens. There must be Lindens on or near the gulf coast of Mississippi. If you do not mind doing so I wish you would ask Tracey [sic] if he has not collected Lindens along the southern coast. Lindens are so common on the Carolina and Georgia coast, and so common in places like Lake Charles, that it would be surprising if they did not occur in the coast region east of the Mississippi River. Many thanks for your letter just received this morning. I will make the corrections you suggest and write you again in a few days. Conspectus of the [Tilia] Species of the United States. Surface of the leaves glabrous at maturity. Surface of the leaves glabrous or almost glabrous as they unfold. Leaves furnished with conspicuous tufts of axillary hairs, their lower sur- face light green and lustrous; pedicels glabrous or on glabrous. . T. glabra. Leaves without tufts of axillary hairs, their lower surface lustrous; pedicels densely hoary tomentose. _T. nuda. Leaves finely serrate, yellow- i oblique or cuneate at the base; tufts of axillary hairs small, rusty bro _ T. littoralis. Leaves minutely dentate, wie green, without tes of males hairs. 4 brevipedunculata. Leaves crenately serrate, ean on the lower surface, the tufts of illary hairs usually wanting or minute. 5. T. crenoserrata. Surface of the leaves Bir exnitescent below early in the season, becoming 154 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 glabrous or nearly naga’ gia summer, glaucescent or green, often without tufts of axillary hai 6. T. floridana. Surface of the leaves pubescent or cern below through the season. Lower surface of the leaves stellate-pubescent. Pubescence short, gray, firmly attached; neg coarsely serrate; tufts of axillary hairs not pine ee someti wanting; ns ds sparsely stellate-pubescent, becoming glabrous. . T. neglecta. Pubescence white, slightly attached. Branchlets and winter-buds glabrous or very rarely pubescent in No. 8. Leaves dark green, finely dentate, oblong-ovate, mostly oblique or cor- date at the base; tufts of axillary hairs minute. 8. T. ambigua. Leaves blue-green, coarsely serrate, orbicular to suborbicular, sym- metrically cordate at the base; tufts of axillary hairs conspicuous. 9. T. phanera. Branchlets and winter-buds pubescent. Leaves finely serrate, oblong- -ovate, obliquely truncate or cordate at the h. base, up to 12 cm. in lengt 10. T. texana. sks very coarsely serrate, ovate to orbicular, mostly cordate at the , hot more than 8 cm. in length. 11. T. grosseserrata. Surface of ie cial tomentose below with close firmly attached tomentum; branchlets glabrou Leaves broadly pa short-pointed, mostly cordate = the base, their tomentum silvery white or grayis . T. heterophylla. Leaves oblong-ovate, long-pointed, mere oblique and ae at the base. Leaves silvery white on the lower surface; flowers at least 10 mm. in length, on sparingly stir pedicels; bumebilets stout; winter-buds compressed, 10-12 in length. 13. T. monticola. sity silvery white or eae on the lower surface; cane 5 or 6 ong, on densely ase erin pedicels; branchlets slender; winter- s terete, 5 or 6 mm. in length. 14. T. Harbison. Surface My i leaves Rien ‘below early in the season, becoming floccose pubescent, the pubescence more or less firmly attached. Young leaves stellate-pubescent on the upper sie Leaves often broader than long, dark green above, silvery white on the lower surface; branchlets —— or xellaepubescen hee the first season ; winter-buds glabrou T. amphiloba. Leaves snot than broad, i oe above, white or eke on the lower surface; branchlets tomentose during their first season, becoming pubes- ae winter-buds covered with rusty brown pubescence. 16. 7. georgiana. Young leaves not stellate-pubescent on the upper surface. Winter-buds glabrous. Branchlets sparingly pubescent early in the season, — glabrous or nearly glabrous; lower surface of the leaves rusty pubescen Leaves coarsely serrate, their pubescence thin, slightly mee some- times entirely deciduous late in the season; — 9 or 10 mm. in length. T. caroliniana. Leaves less coarsely serrate, their pubescence oe attached, persist- ent through the season; flowers 4 or 5 mm. in length. 18. T. Cocksii. Branchlets pale pubescent pinion at least two seasons; lower surface of the leaves white pubescent. 19. T. nivea. 1965 | EWAN, LETTERS FROM SARGENT TO COCKS 155 Winter-buds pubescent. Pubescence on the lower surface of the leaves and seni gray or rusty, that of the leaves only slightly attached. 0. T. rhoophila. Pubescence on the lower surface of the leaves silvery an the pubes- cence slightly attached. Branchlets covered with close gesonierage becoming glabrous in their ‘ second year; flowers 7 or 8 mm. in length, the corymbs densely stellate-pubescent. 21. T. cheiophila. Branchlets covered with long rigid hairs mostly deciduous at the end of the first season; flowers 5 or 6 mm. in length, the corymbs covered with straight hairs. 22. T. trichoclada. 139 February 9, 1917. Hickories. 6 of Natchitoches, We used to think this was a new species but I am convinced that the flowering specimen collected in April 1914 is C[arya| alba, and that the fruiting specimen collected in October 1913 and the fruit is C. arkansana. #1 of Natchitoches, October 1913, seems to be C. ovata. The Carya in dry woods before reaching the quarry at Winnfield, a tree with slender glabrous branches, young leaves nearly glabrous above and hoary-tomentose below. This is not arkansana, alba, ovata, ovalis, megacarpa, callicoma or luchieieilc What is it? It has always bothered P me. Can’t you settle this problem this year? Carya #5, Natchitoches, Cocks, October 1913. This has stout glabrous branches, buds the size of those of C. ovata, and stellate pubescent rachis and midribs of the leaves. The fruit is pear-shaped with a thick husk splitting tardily and looking like some of the forms of C. arkansana. Unless there is some mixup here I do not know it, but if the fruit and leaves do not go together I do not know the leaves unless they may be some extreme form of ovata. 2 and 17 of Natchitoches seem the same. I once thought that 17 was a callicoma but these Natchitoches numbers have small obtuse buds covered with yellow scales and are entirely different from the buds of callicoma. The young leaves of 2, however, show no trace of the yellow scales, so that it cannot be an arkansana. Another species? I hate to think so. Some of the other Natchitoches and Grand Ecore numbers which I have placed with arkansana may be the same thing and it is possible, of course, that this may represent another species for western Louisiana seems capable of anything in the tree line. n Have you found out anything more about that Alabama alba which i leaves out so late? I think I saw the same thing once near Natchez. It may be some variety which we have not got hold of yet. I wish you would go over your Hickory material and see if you can throw any light on any of the numbers I have written to you about today. I had hoped that we had finished the Louisiana Hickories but there seems to be still a little to do at Natchitoches and especially at Winnfield. 156 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 140 February 14, 1917. Here are some of the results of Rehder’s visit to New York: Vitis rufotomentosa. The Louisiana specimen referred by Small to this species of his was collected by Hale near Alexandria and was called by him V. aestivalis. The best character by which V. rufotomentosa can be distinguished from V. aestivalis is in the pubescent branchlets and pedicels. According to this Palmer’s 7974 from Chopin, Louisiana, would be V. rufotomentosa. Harbison has found it at River Junction, Florida, so there is no reason why it might not grow in Louisiana. I shall ask Harbi- son to look for it in Alabama and Mississippi. If you want therefore to keep V. rufotomentosa distinct from V. aestivalis you can include it in your Flora on the strength of Palmer’s specimen. I do not believe myself that it is very much of a species. The Vitis ni pubescent stems, how- ever, ought to be looked for in eastern Louisian Of Vitis caribaea, which Small also credits nm Louisiana, there is no specimen in the New York herbarium, and only four from Florida. On one of these some one has written ‘Florida to Louisiana.” I think you can safely drop this species as a Louisiana plant for there is nothing to show that it grows there Bumelia lucida Small. This is based on a specimen collected by Car- penter in 1838. His label reads, “No. 19, Prinos new? Small tree 10-20 feet high; limbs rigid, some short ones resembling thorns. Grows in thick woods, Feliciana, Louisiana. Flowers in June and Ist of July; berries purplish black. Dr. Carpenter, 1838.” Before Small got hold of it some one labeled it “B. lycioides Pursh.”” I should think it probably a form of B. lycioides. It certainly does not look, however, like your specimen of that species collected at Wakefield in 1907. You may be able to find Carpenter’s plant yet. In the shape of the leaves it looks like a sterile specimen collected by Letterman near Opelousas in August 1882 and preserved in the St. Louis herbarium. Bumelia cassinifolia Small is based on a specimen collected by Letterman in August 1883 near Opelousas. We have a photograph of the type speci- men and a single leaf. The leaves are obovate, rounded at the apex, thin, glabrous and not conspicuously reticulate. This specimen is certainly different from that collected by Letterman referred to above and preserved at St. Louis. We have what seems to be the same as Small’s B. cassini- folia from Paradise Key in the Everglades of Florida, from Mosquito Inlet and St. Marks, Florida, and from near Bainbridge, Georgia. This last is a shrub growing in moist soil. These Florida specimens have slender glabrous pedicels and rounded calyx-lobes. I think you ought to be able to find this plant near Opelousas. If not, I should think you could include it in your Flora on the strength of Letterman’s specimen in New Of B. lycioides we have from Louisiana only your Wakefield specimen and we should have more of it and know more about the distribution of this tree in the state. I think it must be common in western Louisiana. 1965 | EWAN, LETTERS FROM SARGENT TO COCKS 157 141 February 15, 1917, Your unnamed Vitis No. 9, Richmond Parish, June 1909, is V. palmata. This seems common in western Louisiana as Palmer got it at Monroe and Natchitoches. I suppose you have Palmer’s specimen No. 8006, V. rupes- tris from Natchitoches, June 1915. I am sorry to hear that there is another den of Tilias in western Louisi- ana for I fear it is filled with trouble. I suppose you will pay a visit there and I shall be glad to hear of the results. 142 February 23, 1917. Do we know enough about C[arya| leiodermis and have we ever had fruit from more than one tree? Except by the color of the young leaves and by the much thicker husk of the fruit can /eiodermis be distinguished from callicoma? Are you sure that the young leaves of leiodermis are not red or sometimes red? We ought to have a great deal more fruit of these two Hickories. Even if the distinctive characters hold good might they not better be considered varieties of one species? Why cannot our un- known Winnfield Hickory be leiodermis, of which we have no very young leaves? 143 April 9, 1917. Biloxi is a wonderful place for Live Oaks and I wish you had been there with us. There are two very distinct forms, one with young leaves nearly grown and pale and slightly pubescent on the lower surface. These trees are like the one which we collected at Springfield. The second form is a tree still covered with last year’s leaves which are dark green and very lustrous on the upper surface, coated below with thick white tomen- tum, strongly revolute and very conspicuously reticulate-venulose. These trees were not in bloom and have lost none of their old leaves. T his second form was what I supposed was the typical Quercus virginiana. We saw nothing like it in Louisiana. It would be strange, however, if it did not extend into the southeastern part of the state and I should expect it at Slidell. Have you ever been there? I understand that your Live Oaks at New Orleans are green and glabrous on the lower surface and I wish you would dry me some specimens show- ing their present condition. I very foolishly omitted to do this and in this herbarium there is only one specimen of Q. virginiana collected by me in western Louisiana many years ago. I do not find in our Texas material specimens with leaves as thick and tomentose as those on the trees of Biloxi but what appear to be a good many intermediate forms. It is evident that Q. virginiana has been shamefully neglected. That curious White Oak which we found in the road from Springfield 158 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 seems to agree with Michaux’s Q.[uercus| alba repanda which is a name for it. It would be interesting to get good leaves and fruit from this tree which seems to be exceedingly rare. We are having today what appears to be the severest snow storm of the winter and I wish I was back in some comfortable place like Opelousas, although I daresay Ames does not share this opinion. I will write again in a few days as soon as I get things a little in order. I think we had a wonderfully good trip this year. I can’t have too many Live Oak specimens from different trees. 144 April 11, 1917. T. W. Adams, Greendate, Canterbury, is the name and address of the New Zealand man who cultivates Conifers and Oaks so successfully and apparently has a wonderful collection of trees. I wonder if your people know him. I have a note that Pinus glabra grows on the Amite River. Did you supply this information and, if so, can you give me the exact locality where you found the tree? This must be the extreme western limit of its distribution. I understand that the only place where you have found Kalmia latifolia in Louisiana is in Washington Parish. If so, this is probably its western limit and the talk in my Manual of its occurring in western Louisiana and southern Arkansas must be incorrect; at any rate we have no specimens here from that region. In the March issue of Torreya which arrived this morning there is an article by Wilbur R. Mattoon of the Forest Service on the extension of the range in Louisiana of Pinus caribaea. According to him there are pure stands of this tree covering mile after mile along the New Orleans Great Northern Railroad, and that logging and turpentine operations are carried on there by several large companies. He observed it between Slidell and Covington. From Covington to Hammond he found it confined to ponds and along streams, and spread over large areas of flat lands further east. About Hammond he noticed it quite common in young stands and scattered groups. This was the most western point this man visited. He says that “its occurrence south of Hammond along the Illinois Central right-of-way suggests the high probability that it occurs to the north and to the west doubtless as far as the natural barriers formed by the extensive outlying swamp and overflow lands of the Mississippi River.” He says, — “The region, it would seem, affords an excellent opportunity for some one to look for a further extension of the western limit of this interesting and very valuable commercial Yellow Pine.” How about this, and if it does grow at Hammond how did we happen to miss it? This seems to be a thing to look into. I will try and get some of Mattoon’s specimens, but as Hammond is of easy access from Louisiana you will no doubt have an opportunity to verify his observations. 1965] EWAN, LETTERS FROM SARGENT TO COCKS 159 145 April 13, 1917, Was it Jlex cassine myrtifolia or Ilex lucida which you said you had not been able to find in Louisiana? We have a specimen of /lex lucida from you, your Pearl River No. 81, no date, with //ex Dahoon on the label. In our herbarium I do not find any /lex cassine myrtifolia west of Alabama. I saw a plant of //ex cassine (Dahoon) growing in the lake at Avery Island but failed to take a specimen. We have no Louisiana specimen of this tree. It does not seem to be anywhere very common, and I should like to know about its distribution in Louisiana and to get Louisiana specimens if possible. Ilex lucida is common at Biloxi. 146 April 16, 1917. Is the Tilia by the river (the tree I did not see) from which you collected this sprig the same as your 2532 (3 Lake Charles) of May 21, 1916? Your 2530, June 1, 1915, from the Mills Place seems the same. If these are all one species T. rhoophila as a Louisiana tree will have to disappear, and we should have to consider that the young branchlets of 7. ambigua are sometimes pubescent and that the pubescence of the leaves of this tree sometimes becomes brown before the end of the summer. Tilia Cocksii. I understand that your 4040 and 4054 are from the tree at Avery Island which you showed me. If this is true it would seem that there is no way of really learning anything about these southern Linden trees except by visiting them every day from March 15th to September Ist. My last visit shows that there is no use trying to do anything with them worth while with only herbarium specimens gathered twice a year to depend on. The Carya by the roadside near Loranger with slender branchlets and stellate pubescent young leaves and evidently small buds may be some form of C. alba but I think you should try and get the fruit. These trees seem to be the only strange-looking Hickories we saw this year. As I understand it C. callicarpa becomes C. leiodermis, C. letodermis becomes a variety, and C. ovalis subglobosa becomes C. leiodermis, which is something gained. I enclose a bit of my specimen of the Winnfield Hickory which does not look like what you showed me. Is it possible that my specimen is C. leiodermis. What seems to be the same tree grows at Natchez. I think it desirable to investigate Winnfield in the autumn. There is that com- paratively smooth Bumelia from Avery Island which needs further inves- tigation. Can you manage that? Don’t you want to lend me the fruiting specimens of the new Oak with rhombic leaves which you showed me the other day? I am afraid that we shan’t be able to get fruit of this tree this autumn owing to spring frosts. [To be continued 160 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 ANATOMY OF THE PALM RHAPIS EXCELSA I. MATURE VEGETATIVE AXIS? Martin H. ZIMMERMANN AND P. B. TOMLINSON DIAGRAMS PURPORTING TO REPRESENT the course of vascular bundles in palm stems are reproduced frequently in textbooks. These seem to be based less on direct observation than on figures copied from earlier publications. Reasons for this are obvious. Palms are essentially tropical organisms and are not commonly available to research workers. More- over, palm stems are too large and too complicated to lend themselves easily to standard botanical microtechniques. Workers of the nineteenth century, despite their limited technical facilities, made much more prog- ress than their modern counterparts. Von Mohl’s classic and unexcelled work (1824-1849) is the prime source of our present understanding of the anatomy of the palm stem, as has been shown in surveys of the history of the subject (Branner, 1884; Monoyer, 1925; Tomlinson, 1961, 1964). The development of this under- apr up to the time of Von Mohl and subsequently has been traced noyer and is best illustrated by the series of diagrams in Fic. 1, cia et largely from his paper, together with Monoyer’s diagrams of the stem of “Cocos botryophora” (Arecastrum romanzofianum (Cham.) Becc., see Moore, 1963), based on his own observations and reproduced here as Fic. 2. In addition to surveying the literature both Branner and Monoyer made significant contributions to the subject. Branner attempted to explain the distinctive curvature of individual leaf traces in palm stems as a mechanical consequence of their appearance in the meristema- tic region of the palm stem (cf. also Schoute, 1903; Arber, 1925, p. 35 a) According to Arber, Von Mohl appreciated the mechanics of the curvature in 1858. Monoyer was concerned much more than previous workers with anas- tomoses between individual bundles and the way in which the vascular network constitutes a pathway for long-distance transport. His own diagrams (Fic. 2) illustrate a regular system of bundle fusions. He *This is a contribution to a major research program in which the as of arborescent monocotyledons is being studied by one of us (P.B.T.). The gram is supported by grants from the National Science Foundation (G.B. 506) aoe the American Philosophical Society (Penrose Fund 2959, Johnson Fund 488). Subsequent articles will deal with leaf, root, and inflorescence insertion, seedling and cae anatomy, and general development. f this paper, including a 12-minute analytical motion picture, was presented in cai guns posium “Long Distance Transport Phenomena” held during the X Inter- a Botanical Congress, Edinburgh, Scotland, August, 19 t should be noted that Arber’s diagram (Fig. xvi) is somnew uit misleading. 1965] ZIMMERMANN & TOMLINSON, RHAPIS EXCELSA 161 shows a presumed bundle fusion (his P/. V/IJ, Fig. 4) without com- menting on anatomical details. Previously Lestiboudois (1840) and Karsten (1847) had recorded presumed fusions, but they were based on macroscopic dissection only. All other illustrations of fusions appear to have been inferred rather than observed (e.g., Fic. 1 E). Our own dye injection experiments suggest that interconnections between different vascular strands are numerous and physiologically very significant. Monoyer’s work was also important in suggesting differences between proximal and distal parts of a single palm stem (Fic. 2). This is a point overlooked by theorists who ascribe a simple “type” of construction to all palm stems. We suspect that the essential features of the distribution of vascular bundles in palm stems has hitherto not been grasped except by Branner (1884). Early concepts of the growth of palm stems, now known to be false, were used by De Candolle (1813) in separating monocotyledons from dicotyledons as a major group. Monocotyledons do indeed differ in their basic construction from dicotyledons, but not in the way envisaged by De Candolle. The so-called “palm-type” of bundle distribution has long been regarded as one of the basic patterns in monocotyledonous construction (e.g., Falkenberg, 1876). It would evidently be useful to establish these concepts more factually. The present article is the result of the reappraisal of the problem at first hand. Von Mohl’s general principles have been established on a quantitative basis for the first time and many new observations made. A much more accurate understanding of the structure of the palm stem and of the pathway for long-distance transport has now been reached. Earlier observers give little indication of methods used in analyzing palm stems. Large scale dissection, particularly of partly decomposed trunks, has evidently been most used. This gives general information about the behavior of individual bundles, but the intimate relation be- tween different bundles is lost, and anatomical details are overlooked. Macroscopic dissection cannot distinguish easily between temporary juxta- position of two bundles, which is frequent where they are congested, and true vascular anastomoses which are physiologically and developmentally important. Serial sections, less used by previous workers, have been mostly of small palms (e.g., Nageli, 1858). Our own study is largely based on serial sections, which means that the relative position and detailed anatomy of vascular bundles is preserved. Anatomical information is stored on ciné film so that large scale analyses which are needed become a relatively simple matter. In addition, this dynamic approach makes it easy to distinguish between important and unimportant vascular contacts. GENERAL MORPHOLOGY OF RHAPIS The problem was simplified by using the small palm Rhapis excelsa (Thunb.) Henry, a native of eastern Asia and a common ornamental in ee ee eee 1965 | ZIMMERMANN & TOMLINSON, RHAPIS EXCELSA 163 south Florida, Aérial stems 2-3 cm. in diameter and up to 5 m. high arise from horizontal rhizome segments up to 1 m. long. Younger, de- veloping parts of each rhizome segment are protected by short, rigid scale leaves which are morphologically equivalent to the sheathing base of the foliage leaves of aérial stems. In older parts of rhizomes the scales decay, leaving annular scars at intervals of 1-3 cm. marking the nodes. Adventitious roots arise at wide but fairly regular intervals along the rhizome, commonly, but not exclusively at the nodes. After a period of hortsontal growth the apical part of the rhizome enlarges appreciably and turns erect. There is a rapid change from rhizome scales to normal foliage leaves, the transitional leaves with increasingly developed blades illus- trating that the scales are essentially foliage leaf sheaths without a blade. In the axils of several of the most distal rhizome scales new vegetative buds grow out as the rhizome apex turns upwards. Most of these lateral branches grow out horizontally as new rhizome segments; occasionally In the distal parts, axillary inflorescences appear, internodes are shortened Pic. 1, casera bundle course of “palm type” according to various classical authors. A—-E, redrawn from Monoyer, 1925; F, from Haberlandt, 1884. - : ; =) — ime ro) | a ra) 3 =) gq » a A= seh ‘a pre (=) rot) ond [a wm i aa 7 oO et » wm =) = mS oo A ns i=) ge < oOo a e€ number of bundles at any one level remains ae per constant. D. enberg (1876). Purporting A illustrate Chamaedorea elatior (and possibly Snives from Nageli, 1858, who studied this species), although, according to his text, Falken- 7 i ° = - o for =F “SD 10 a @ BS; = © wn oO i=) =. un io” ° c 2 "SSS = a io 1] i 2 pl @ 4 ) = Qu. ° (a =] oo a yy — ro) te a! tr & never observed. F. Ane ome (1884). From Falkenberg ( pes via — (1882). A much- re, bie illustration purporting to represent the “palm type. It was actually drawn by Falkenberg to represent Aspidistra elatior C Tihaseack: a statement pollen Ba textbooks. It seems to bear little relation to actua palms. From this date on, the “palm type” of vascular construction has become entirely hypothetical. JOURNAL OF THE ARNOLD ARBORETUM MONOYER 1925 M _ E Hil iE K | 7 ‘N SS HAN \\ 1965 | ZIMMERMANN & TOMLINSON, RHAPIS EXCELSA 165 and the outline of the stem becomes more angular owing to the pressure of developing inflorescences. The anatomy of this part of the stem will be described in a subsequent article. By choosing a palm of this stature the problem was reduced from the analysis of a stem like that of Cocos, which may be up to 0.5 m. in diameter with over 20,000 vascular bundles at any level, to one in which there were about 1,000. The relative ease with which the Rhapis stem could be sectioned freehand was a second reason for its selection. Clearly, principles based on the study of a diminutive palm must be verified by analysis of a much larger and more typical stem. The basic approach used in this work suggests that ultimately this will be possible. However, we already have much evidence from larger palms like Chrysali- docarpus, Sabal, Washingtonia, etc. which suggests that Rhapis is indeed a small-scale model for palm stems generally. CINEMATOGRAPHIC ANALYSIS Cinematographic analysis was accomplished by two different methods, namely, by photographing, frame by frame, (1) sequentially cut surfaces and (2) serial sections through the microscope. Kodachrome II, Type A film has been used throughout. Surface photography. Photography of cut surfaces has been used only to a very limited extent with Rhapis. Pieces of stem were clamped in the specimen holder of a Reichert type OME sliding microtome, and thin shavings were taken with a frequently sharpened knife. The speci- men holder was advanced for each cut along the optical axis of the camera, i.e., manually with the screw which raises the specimen holder. The specimen holder was modified in order to accept long pieces of stem. A scale marked in ink on the stem provided a means of measuring the rate of advance. The camera, a Bolex H 16 REX, was mounted on a sturdy stand above the specimen with the optical axis coinciding with that of the stem. A long focal-length lens (105 mm.) allowed the operator of the microtome ample working space. We developed this method our- selves but learned during the preparation of this manuscript that a similar, but automatic method for the analysis of short tissue pieces had been developed earlier by Postlethwait. In fact, the method seems to have been proposed as long ago as 1907 (Postlethwait, 1962). Photography of serial sections. For the analysis of the course of Base. May be compared with De Bary’s scheme, but with the addition of — bundles (faisceaux de jonction) which unite upper and lower bun sees (solid black) represent dorsal (major) bundles which penetrate most deeply into stem center. Ventral and minor bundles (cross hatched) have same course but penetrate less deeply into stem. : Apex. Essentially ae base but with bundles of both systems all extending more than halfway across stem. A likely explanation of this is given in the text (p. 171). 166 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 vascular bundles over long distances, sections 25—50u thick were cut freehand on the same microtome at intervals of 0.5 mm. The automatic feed was set at 50u and every 10th section kept. For easy processing each of the sections retained was tied with a constant orientation to a sequentially numbered slide by cotton thread. Orientation was facilitated by a shallow longitudinal groove scored along the stem and visible as a slight notch in each section. A small wad of glass wool between cotton thread and section prevented unstained thread marks across the section. Sections were bleached briefly in “Clorox,” washed in water and 70% ethanol, stained overnight in a mixture of safranin (95 parts of a 1% solution in 70% ethanol) and Delafield’s haematoxylin (5 parts), dif- ferentiated briefly in acidulated 50% ethanol, dehydrated, and mounted in neutral ‘“Piccolyte.” Sections at 0.5 mm. intervals proved adequate for general analyses as it was easy to trace a single bundle from section to section. For details of “rapid” changes which occurred at a node a complete series of sections approximately 30, thick were prepared, pass- ing through an entire internode about 2 cm. long which had been embedded in celloidin. A piece of stem much longer than could be held complete in the micro- tome clamp was analyzed in shorter lengths. The upper part of each of these shorter segments was clamped in the normal way and as many sections as possible removed, then the lower part was held firm by freezing, in a 5% gum arabic solution, to the flat stage of a freezing attachment and the sectioning of the segment completed. We have now greatly simplified the procedure by the development of a modified clamp with which very long pieces of stem can be held. The key problem is to match successive sections under the microscope so that a fixed point in the section successively occupies the same position on each frame of the film. So far this has been done by incorporating a camera lucida into the microscope-camera system in such a way that an outline drawing of every 2nd to 6th section is made and used to orient its successors precisely. The drawings are used as an additional per- manent record, with the microscope-stage coordinates marked for about every 10th slide, so that, if necessary, it is possible to return to the slides at any later time and locate the photographed position. A Wild M20 microscope was used, with the parts assembled from below upwards as follows: microscope base with centerable rotating mechanical stage, ob- jective turret, drawing tube, camera tube H, binocular. Viewing and focusing are done in the binocular; photographs are taken with a Bolex H 16 REX, which is mounted on a Wild camera stand. It is often not necessary to have a permanent record of drawings and photographed slide positions. The extra labor of drawing can be eliminated with an optical “shuttle” system. A Wild discussion tube is mounte graphs are taken alternately through each microscope, the double image being used merely to superimpose successive sections precisely. 1965] ZIMMERMANN & TOMLINSON, RHAPIS EXCELSA 167 Analysis of films. For the analysis of films an L-W 224 A Photo- Optical Data Analyzer was used. This is essentially a 16 mm. movie pro- jector modified by L-W Photo, Inc. (15451 Cabrito Road, Van Nuys, Calif.), so that a film can be projected without flicker at any speed from 1 to 24 frames per second, forward or backward. The film can be stopped instantly, and moved frame by frame. Less expensive projectors are available, and action editors are also usable for the purpose. How- ever, the instrument we have used has proved ideal in every respect. The vascular system of Rhapis described below has been resolved by an essentially dynamic process. This is reflected in the subsequent description which refers to the distribution of vascular bundles by a dynamic terminology. This is merely a convention. Changes in direction of mature bundles reflect changes in the angle of differentiation of bundles in meristematic regions. “The “movement” suggested by ciné analysis, particularly of one bundle in relation to another, is quite apparent and implies neither actual movement of bundles nor their growth. GENERAL STEM ANATOMY OF RHAPIS Stem (Fic. 7) divided into a narrow cortex less than 1 mm. wide and a layer except for interruptions below occasional stomata. Cortical parenchyma up to 20 cells deep; cells isodiametric or slightly elongated vertically, becom- ing chlorenchymatous in exposed stems. Innermost layers of cortical paren- vascular bundles forming peripheral mechanical zone. Each sg eel with a massive fibrous sheath; xylem sheathed only by narrow slightly thick- (vertical, bridge, and satellite bundles — see below). iieieps wage igri center of stem gradually less congested and with reduced fibrous sheaths, coming more circular in transverse outline. very rare. Silica bodies in small isodiametric stegmata abundant next to fibers. 168 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 Tannin common in unmodified parenchyma cells throughout the stem; rha- anatomy throughout its length is discussed in detail below): abruptly pen from ground parenchyma by sheathing tissues. Phloem usually directed towards stem periphery, partly sheathed by fibers forming a well-developed fibrous phloem sheath. Fibers becoming sclerotic and conspicuously concentric- -layered tissue, often divided into two equal strands by a narrow sclerotic isthmus, in- cluding sieve tubes mostly 12-15 wide, with compound sieve plates on slightly oblique to oblique end walls, irregularly scattered companion and phloem paren- chyma cells, the former not well differentiated from the latter. Protophloem remains often recognized as an irregular group of sclerotic cells in the peri- pheral part of the sclerotic isthmus. Xylem enclosed by thin-walled, slightly lignified parenchyma less abruptly delimited from ground tissue than phloem sheath; rarely including a few fibers. Cells immediately surrounding vessels with wide pits and distinctly differentiated from remaining xylem parenchyma. Tracheal elements? including one or two metaxylem vessels 40-604 wide, circular in transverse outline, and usually some narrow protoxylem elements. perforation plates with few (4-10) thickening bars on oblique end walls; protoxylem either long imperforate tracheids with annular or helical wall thickenings or somewhat shorter vessels with unspecialized perforation plates transitional to metaxylem elements. Variation throughout a single stem largely involves an increase in lig- nification with age. Stems vary much in vigor of growth and this is re- flected in differences in diameter and length of internodes and total number of vascular bundles. Absolute measurements may be misleading, the best comparative unit is the internode. Differences between axes in juvenile and adult stages of growth have not yet been examined. The following paragraphs deal with overall distribution and types of vascular bundles in the adult mature stem; subsequently anatomical changes along vascular bundles are described. COURSE OF VASCULAR BUNDLES IN THE STEM Cinematographic analysis demonstrates that all vascular bundles behave essentially alike (Fic. 3). Major and minor bundles, with a we agoge transitional series, differ in relative position in the stem and in the number of internodes between their successive branchings. Any one leat is supplied by a larger number of minor and a smaller number of major bundles. From the base to the apex essentially all bundles maintain their indi- viduality and proceed indefinitely up the stem. Each of these vertical *Protoxylem is distinguished from metaxylem entirely by its position and the nature of its elements. Developmental differences are not considered here. 1965 } ZIMMERMANN & TOMLINSON, RHAPIS EXCELSA 169 7. LEAF TRACE ENTERS LEAF E SATELLITE BUNDLES ENTER INF LORESCENCE s e) ss oS 1, VERTICAL BUNDLE NEAR STEM @ 9 Soe vp PERIPHERY ON ITS WAY UP o@ Cz: (le oS s s Bs we Se << y 8 ie) = $ i lia “@e¢ hire: s® Bsc he a aes | © 6. LEAF TRACE AT 5. BUNDLE MOVES TOWARD BASE OF INFLORES- Farha PERIPHERY AND 3 CENCE SURROUNDED UP INTO LEAF TRACE 4, sare B INCREASES BY SATELLITE BUNDLES, 4 Si 6 1s =f | 1s a = Z| @ & I ] le ae j 23 1 Jo ak 10 los. BE 10 1 a ‘ % 1 . =! Ro = re a! ° ea 5 I SF I LEAP 5 9 7 i J I 1 o 1 1 e 3 1 FS 2 - ‘ hibsacatsancaeens 5 Jo 13 mm 20 i 2 3 4 MILLIMETERS hoe 4 (eft). The course of a vascular bundle in the stem of Rhapis, drawn . ni i i un stem periphery (short eta gow, and the measured bundl Open circles indicate presence of protoxylem only; half-open circles indicate ranenee of both protoxylem and metaxylem; closed circles indicate presence of metaxylem only. Fic. 5 (right). Diagram of a major, an intermediate, and a minor leaf trace to show spatial relationship to each other. Bridges and ‘satellite bundles are not shown. Stem axis is foreshortened 3.6 times in relation to the stem radius Measurements were taken as in FIG. 1965 | ZIMMERMANN & TOMLINSON, RHAPIS EXCELSA 171 A feature which cannot be indicated in FicurE 3 is the shallow helix exhibited by all bundles in the central, uncrowded part of the stem. Traced upwards, all central bundles rotate uniformly in a clockwise direction, the steepness of the helix being such that each bundle rotates about 1% of the stem circumference during a distance of about 15 cm. This helical path had been indicated by several earlier workers (e.g., Meneghini, 1836; Niageli, 1858), but irregularly and only for single bundles. Ciné analysis shows it to be a uniform property of all bundles. It is quite obvious that if the distance between two successive leaf contacts of the same (major) vertical bundle is a little over 30 cm. this twisting brings the higher contact to the side of the stem opposite that which bears the lower contact (cf. Mirbel’s diagram, Fic. 1B). If the path is still longer, the bundle may actually return to the original side of the stem. Monoyer’s diagram of the upper part of the stem of Are- castrum (Fic. 2, right) may be simply the expression of either longer distances between leaf-trace intervals and/or a steeper helix. DETAILS OF VASCULAR BUNDLE ANATOMY FicurE 5 shows three bundles drawn to scale, a major, a minor, and an intermediate one. It can be seen that the three traces enter the leaf at different angles. Therefore there is no distinct nodal plexus. During the oblique outward curve, a bundle forks repeatedly (Fics. 3, left; 6, 10, 11). The protoxylem-containing branch goes into the leaf as the leaf trace, the metaxylem-containing branches become respectively bundles. These four types of branches are discussed in the following paragraphs. Leaf trace. This originates by gradual transformation of the vertical bundle in the following way. Following any vertical bundle upward in the stem, at a point about 10 cm. below the next leaf contact, the first evidence of protoxylem is reached (position 3 in Fic. 3). From this point upward the amount of protoxylem increases continuously, The bundle then reaches its maximum size (position 4, Fic. 3), turns sharply toward the stem periphery and breaks up into several branches. The whole of the metaxylem, but normally none of the protoxylem, passes into these branches. The protoxylem-containing bundle is the direct continuation of the vertical bundle and goes into the leaf as the leaf trace. This means that there is no continuity of metaxylem between stem and leaf. The leaf is irrigated solely by protoxylem.* Vertical bundle. Ordinarily, the vertical bundle is the first branch to break off the leaf trace, shortly above the sharp outward turn. It then follows the leaf trace and near the stem periphery abruptly turns upward and continues its course up the stem, repeating the cycle (Fics. 3-6) In most cases a single vertical bundle arises from a trace plexus. Oc- “See footnote 3, page 168. We MAJOR LEAF TRACE JOURNAL OF THE ARNOLD ARBORETUM AF BASE WITH INFL RUDI ORESCENCE MENT VERTICAL BUNDL BRIDGES — Sk Sa BUNDLE ~~ ‘ iy VERTICAL BUNDLE NM! SPLITS OFF MINOR BUNDLE In leaf trace to its neighboring bund elation to s e > as x MAJOR BUNDLE ae hs iagrammatic representation (not to scale) of the relation of a major i em axis foreshortened about XY. f u i our times 1 ini oxylem- i i . The i ion and continuity of vascular tissue, they do not imply continuous vessels (see text p. 173). This figure represents stages 4~7 of FIG. 3. casionally, however, none of the branches develops into a vertical bundle, in yet other cases two or even three vertical bundles may arise. This emphasizes the dynamic flexibility of the palm stem: the total number of bundles does not have to be exactly the same along the stem; in fact, Future investigation should answer this question. it is certainly affected by environmental conditions. Whether the initial to this mechanism of “multiplying vertical branches’? we do not know. leaf contact is less than 10 cm. drastic increase in stem diameter from seedling to mature axis is due Occasionally one can find a vertical bundle carrying not only metaxylem but also protoxylem. This indicates that the distance to its next higher The very small fibrous bundles in the cortical area (Fic. 7) are vertical bundles from minor leaf traces. They become smaller, ‘“move” towar s the stem periphery (rather than towards the center) and end blindly further up the stem. Bridge bundles. From each departing leaf trace, two to six bundles branch off and connect to vertical bundles in the neighborhood (Fic. 6). [voL. 46 1965] ZIMMERMANN & TOMLINSON, RHAPIS EXCELSA 173 Bridges are short, the longest one measured 4.5 mm. Occasionally they are so short that leaf trace and vertical bundle appear to fuse directly (Fic. 10). Bridges are always oriented obliquely upwards in the same way, so that the end of the bridge attached to the leaf trace is below the end attached to the vertical bundle. Bridge bundles normally come directly off the leaf trace, but they also may break off a satellite bundle or even the vertical bundle. Oc- casionally a single bridge bundle breaks up into two branches which then connect to two vertical bundles (Fic. 6 Physiologically, bridge bundles not only provide vertical vascular con- tinuity via other vertical bundles, but they also provide extensive cross connections. Injection experiments in which dyes are introduced via petiole, root, or bore-hole in the stem, illustrate these lateral interconnec- tinuity of individual vessels. Vessels are of limited length and overlap within any one bundle, just as they do within the xylem of a diffuse- porous dicotyledonous tree. Satellite bundles. Several satellites branch from larger leaf traces in the region of bridge production (Fics. 3, left; 6, 11). Major traces produce up to 10 or more satellites; minor traces produce fewer or none, according to their size. Satellites accompany the trace to the periphery of the central cylinder. Some may “rotate” around the parent leaf trace and, in addition, may be inverted temporarily. They retain their in- dividuality but may become temporarily enclosed within the fibrous sheath of the parent leaf trace or a nearby vertical bundle without making vascular connection. Ultimately they form a “halo” around the leaf trace (Fic. 3, position 6; Fic. 9). Immediately below the leaf base they turn abruptly and tangentially, and enter the inflorescence rudiment. It is obvious that satellites are best developed if a fully developed inflores- cence is (or was) present. However, even inflorescences which had aborted very early in their development and are macroscopically not visible can be detected by their satellites, even in the lower “vegetative” part of the aérial stem (Fic. 8). With a good deal of experience, one can often easily distinguish, in a single transverse section, vertical bundle, satellites, and bridges. How- ever, this is not always possible, and one must follow the bundles up- ward in order to differentiate with certainty between the three types of branches that come off the leaf trace. One cannot help getting the im- pression that the three types, vertical bundles, satellites, and bridges, are developmentally equivalent. This problem will undoubtedly be met again, and hopefully be solved, when the apex is analyzed. COMPARISON OF RHAPIS WITH OTHER MONOCOTYLEDONS Our study of Rhapis has been carried out conjointly with the examina- tion of a number of larger palms. The reason for this is quite obvious. 174 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 Our findings differ so fundamentally from the “classical” diagrams (cf. Fics. 1 and 3) that the question arose as to whether we were dealing with an unusual object. Comparative studies have shown, however, that this is not the case. Whatever we have studied in other species has corresponded to Rhapis, although a great deal of work still remains to be done. We are now in a position to do it with the cinematographic method. Literature search has revealed that the actual observations of former botanists do not differ from our findings. The way in which these early workers put together diagrams by extrapolation and inference proved misleading. Interestingly enough, good observers like Von Mohl, Karsten, etc. describe the limits of their work precisely. Distortions arose by the inaccurate citation of older authors, by frequent re-drawing of diagrams, and by the failure to distinguish what was observed from what was in- ferred. Von Mohl’s original diagrams (1824) were correct but incomplete. Those showing the blind ending of bundles in a basipetal direction (Fig. VIII.2 in Tomlinson, 1961, p. 362) can be re-interpreted by assuming that Von Mohl’s upper series were of a major bundle (Tomlinson’s Figs. 2a-f) and the lower series were of a minor bundle (Tomlinson’s Figs. 2g- i). This is very possible because Von Mohl merely drew a series of bundles from a single section and referred them to a hypothetical scheme. It was only subsequently (Von Mohl, 1849) that he considered the pos- sibility of bundle fusions. This hypothetical suggestion was seized upon by later workers and incorporated in their diagrams (e.g., De Bary, 1877; our Fic. 1E). Branner (1884) is the only nineteenth century worker who fully understood the continuity of vascular bundles, simply because he was able to examine a great deal of material at first hand. He dis- covered that the continuity of vertical bundles was only evident when the apical regions were examined. Macrodissection suggested the blind ending of bundles because connections are tenuous and easily broken in mature parts of the stem. It is a remarkable feature of botanical history that Branner’s very accurate work based on the careful examination of many palms in the field has been entirely overlooked in favor of wholly hypothetical schemes proposed by authors who had little opportunity of examining the problem at first hand. While one of us (M.H.Z.) is interested in the vascular anatomy of palms from the point of view of long-distance translocation, the other (P.B.T.) approaches the subject as a systematic anatomist. The un- ravelling of the palm stem via Rhapis has given a much sounder theoretical background for an appreciation of certain features tentatively proposed as diagnostically significant (Tomlinson, 1961; cf. also Stenzel, 1904, where the important problem of identifying fossil palm stems is discussed in detail). Palms have been subdivided into three main categories de- pendent on the presence, in any small sample of the stem, of a majority of vascular bundles with either one, or two, or many vessels. A multi- 1965] ZIMMERMANN & TOMLINSON, RHAPIS EXCELSA 175 dimensional understanding of the palm stem explains this variability. We have seen in Rhapis that these three conditions represent successive stages repeated in the same sequence in all bundles over and over again. In Rhapis all bundles, traced from below upward and beginning a little above a leaf contact, show initially (Fie. 3, stage 1) one metaxylem vessel, except for areas in which two vessels within one bundle overlap, for example in bridge contacts. Continuing upward, however, the bundle then shows two vessels separated by a parenchymatous isthmus (some- what below stage 4, Fic. 3). Finally (stages 4 and 5), prior and during the breaking up of the leaf-trace plexus, many metaxylem vessels are present. This increase in the number of vessels upward along a single bundle goes parallel with the increase in the amount of protoxylem. Diagnostic significance can only be attached to a comparison of bundles of the same stage (preferably stage 1) of two species. Thus Rhapis, as well as Ancistrophyllum, Areca, Borassus, etc. (Tomlinson, 1961, p. 336) can be referred to as ‘‘one-vessel palms.” ‘‘Two-vessel palms” can be recognized by protoxylem-free bundles which contain two vessels separated by a parenchymatous isthmus which indicates that the vessels are not merely overlapping ends. There may even be ‘“many-vessel palms,” but with the present state of our knowledge comment is not justified. One great danger of using this difference in number of vessels per vascular bundle (of the stage 1) as a diagnostic feature lies in being satisfied that a small region of the stem is an adequate sample of the whole. In palms, there are considerable quantitative changes from base to apex of a single axis. It is quite conceivable that the same stem could be a ‘“‘one-vessel’”’ type basally, yet ‘“‘two-vessel”’ distally. The genus Rhapis cannot yet be compared easily with monocotyledons like Zea (e.g., Kumazawa, 1961) or Tradescantia (Scott and Priestley, 1925) which have a plexus of vascular tissue at each node. Non-nodal monocotyledons, however, may be similar, although few of these have been described in sufficient detail to merit comparison. An exception is Alstroemeria aurantiaca (Amaryllidaceae) analyzed by Priestley, Scott and Gillett (1935). This shows some of the basic features of the Rhapis type. Median leaf traces in Alstroemeria are derived from continuous vertical bundles, although this is not obvious because forking occurs many internodes below the node of exsertion. Forking of bundles many nodes below the level of their exsertion as leaf traces is illustrated in Falken- berg’s diagrams of several monocotyledons intended as representative of the “palm type.” But we know that the one palm (Chamaedorea) which Falkenberg claimed to have studied was analyzed incorrectly. We feel justified in being a little suspicious of Falkenberg’s diagrams. Rhapis still has to be analyzed in terms of developmental physiology in the manner of Priestley and his associates. No doubt in the future many other monocotyledonous shoots will be better understood with the Rhapis stem as am 176 JOURNAL OF THE ARNOLD ARBORETUM [VoL. 46 SUMMARY The vascular anatomy of the mature, vegetative aérial axis of Rhapis excelsa (Thunb.) Henry is analyzed quantitatively by a cinematographic method. This involves photographing, frame by frame, either sequentially cut surfaces or serial sections stained and mounted permanently. The resulting film is analyzed in a data analyzer. All vascular bundles have essentially the same course and construction. In the uncrowded inner part of the stem they describe a uniform shallow helix continuously through the stem as vertical bundles, tending always toward the center of the stem. At intervals each turns sharply toward a leaf insertion and at the same time breaks up into several branches. The leaf trace proceeds into the leaf, satellite bundles go into the inflorescence, bridge bundles are minor, intermediate, and major bundles, the last reach farthest into the stem center and have the longest distances between successive leaf contacts. ACKNOWLEDGMENTS Routine serial sectioning on which this analysis is based was carried out by Lesley Jackson, Vroni Oswald, and Gabriel A. Vargo. We are indebted to them for their patience and perseverance with this tedious task. The generosity of Mrs. A. R. Jennings, who permitted unlimited access to the Coconut Grove Palmetum for supplementary experiments, is grate- fully appreciated. LITERATURE CITED Arper, A. 1925. Monocotyledons, a morphological study. 258 pp. Cambridge University Press. Bary, A. DE. 1877. nagar ag Anatomie der Vegetationsorgane der Phanero- gamen und Farne. xvi + 663 pp. Wilhelm Engelmann, Leipzig (English nana gr ieee Press, Oxford, 1884 BRANNER, J. C. 1884. The course and growth of the fibrovascular bundles in palms. Proc. Am. Phil. Soc, 21: 459-483. CANDOLLE, A. P. bE. 1813, Théorie élémentaire de la botanique. viii + 527 pp- Déterville, Paris. DesFonTAINES, R. L. 1798. Mémoire sur l’organisation des monocotylédons, ou plantes 4 une feuille séminale. Acad. Sci. Paris Mém. 1: 478-502 pls. 2-6. FALKENBERG, P.- 1876. Vergleichende Untersuchungen iiber den Bau der Vegeta- tionsorgane der Monocotyledonen. 202 pp. 3 pls. Ferdinand Enke, Stutt- gart. HaBERLANDT, G. 1884. Physiologsiche Pflanzenanatomie. Wilhelm Laer Leipzig (English translation: Macmillan, London, 1914). "Way 1965] ZIMMERMANN & TOMLINSON, RHAPIS EXCELSA 177 Karsten, H. 1847. Die Vegetationsorgane der Palmen. Abh. Akad. Berlin Physik 1847: 73-236 Kumazawa, M. 1961. Studies on the vascular course in maize plant. Phyto- morphology 11: 128~1 Lestipoupots, T. 1840. + sur needa et la physiologie des végétaux. Lille. [Original not seen. Cited by Mono MENEGHINI, G. 1836. Ricerche sulla aces a caule nelle piante monocotile- doni. 110 pp. 10 pls. Minerva, Padua. MirseL, C. F. B. pe. 1843-44. Recherches anatomiques et physiologiques sur quelques végétaux monocotylés. Compt. Rend. Paris 16; 1213-1235; 19: 689-695. Mont, H. von. 1824. De palmarum structura. /n: K. F. P. von Martius, Historia Naturalis Palmarum 1: pp. I-LII. 16 pls. . 1849. On the structure of the palm stem. Rep. Ray Soc. 1849: 1-92. Monoyer, A. 1925. eiicer - Cocos botryophora Mart. Mém. Acad. Roy. Belg. Sci. Coll. 8°. 8(7): 1 Moore, ef “4 1963. An ne check list of cultivated palms. Principes 7: 119-1 NAGELI, e = 1858. Ueber das Wachsthum des Stammes und der Wurzel bei den Gefasspflanzen. Beitrage zur Wissenschaftlichen Botanik. Heft 1. PosTLETHWAIT, S. N. 1962. Cinematography with serial sections. Turtox News 40: 98-100. PriestLey, J. H., L. I. Scott, & E. C. Gmitet. 1935. The development of the ra t in Alstrocmeria and the unit of shoot growth in monocotyledons. n. Bot. II. 44: 161-179. ng 7 1882. Vorlesungen iiber Pflanzenphysiologie. 991 pp. W ory Engel- mann, fae (English translation: Clarendon Press, Oxford, 1887). Scuoute, J. C. 1903. Die Stammesbildung der Monokotylen. ‘ 92: 32-48. 1. Vv. Scott, L. I., & J. H. Prrestiey. 1925. Leaf oe oe anatomy of Tradescantia duminencis Vell. Jour. Linn. Soc. Bot. 47: 1 STENZEL, K. G. 1904. Fossile Palmenholzer. so Palaont. Geol. Oesterreich- Vora Orient. 16: 107-288. Tomurnson, P. B. 1961. Palmae. Jn: C. R. Metcatre, ed. Anatomy of the Monocotyledons 2. xv + 453 pp. Clarendon Press, Oxford. 1964. Stem structure in arborescent monocotyledons. In: M. ZIMMERMANN, ed. The formation of wood in forest trees. pp. 65-86. Academic Press, New York. HARVARD UNIVERSITY ABOT FOUNDATION, PETERSHAM MASSACHUSETTS 01366 AND FAIRCHILD TROPICAL GARDEN Miami, Fioripa 33156 178 JOURNAL OF THE ARNOLD ARBORETUM [ VoL. 46 EXPLANATION OF PLATES PLATE I re 7, A small rie of the Rhapis stem showing crowding of bundles near the stem periphery. Near the center of the photograph a leaf-trace plexus con- sisting of leaf trace ( LT), vertical bude (VB), and satellite bundles (S). This corresponds to position 5, FIG. 3. PLATE II Fics. 8-11. Transverse sections of stem of Rhapis; all magnifications iden- tical. Fic. 8 (top left). A section at the level of the leaf base, corresponding to position 7, FIG. 3. The aborted inflorescence between stem and leaf. Fic. 9 (top right). A section just below the leaf base, corresponding to position 6, Fic. 3. The leaf trace (LT) is surrounded by a halo of satellite bundles. Fic. 10 (bottom left). A short bridge between a leaf trace (LT) and a neighboring — bundle. Frc. 11 (bottom right). A leaf-trace plexus consisting of leaf e (LT), vertical bundle (VB) and ne bundles (S). This photograph panos to position 5, FIG. 3. PLaTE I Jour. ARNOLD Ars. VOL. 46 ve oe Ok +5 | yy See Phos ’ v6 » ~) Stee er oe - SereSee:. np es Lat } whe Sree ase! ees pease Rar ao: ih : rr tee eerie’ x . Ssieege ee LY tae a Pat hy 8," @® ans A SG bah 4 gern Fes nse se O25, ot, yews ig, HO a oe é. ee ZIMMERMANN & TOMLINSON, THE PALM RHAPIS Pirate II Jour. ARNOLD Ars. VOL. 46 1 MILLIMETER , THE PALM RHAPIS NN & TOMLINSON ZIMMERMA ~~ 1965 | DUDLEY, STUDIES IN ALYSSUM 181 STUDIES IN ALYSSUM: NEAR EASTERN REPRESENTATIVES AND THEIR ALLIES, II. SECTION MENIOCUS AND SECTION PSILONEMA T. R. DuDLEY A SYSTEMATIC SURVEY OF the genus Alyssum has long been needed, not only because it is the largest cruciferous genus in Turkey, accounting for approximately one fifth of the taxa of the Cruciferae in that area, but also because it has been the subject of some revisionary work (at least in the perennial species of sections ALyssum and OpONTARRHENA). The very extensive, and mostly unidentified Near Eastern collections of P. H. Davis, A. Huber-Morath, and K. H. Rechinger have added impetus to the need for a reassessment of morphological characters and variation, and a more stable definition of the specific and infraspecific taxa. Previously, the only practical procedure for identifying Alyssum was to compare and contrast individual specimens with “correctly determined’ herbarium material. The lack of conformity in treatments because of hierarchies and extreme “splitting” of Fenzl in Tchihatcheff (1860), J. Baumgartner (1907-1911), and E. J. Nydrady (1926-1949). Taxa such as Alyssum linifolium, A. alyssoides, A. minus, A. minutum, A. montanum, A. sibiricum, A. tortuosum, A. murale, etc., having very wide geographical distributions are subject to considerable morphological variation within and between populations, as well as between individuals. With the exception of section TETRADENIA, whose component species are not known to occur in Turkey, and section PstLtoneMA which is repre- sented in Turkey only by the widespread Alyssum dasycarpum (also possibly A. alyssoides, cf. page 199), more species (eighty-eight) and endemic species (forty-nine) are found in Turkey than in any other land area of comparable size. Of the seven species assigned to section MENI- ocus, six occur in Turkey, and of these three are endemics. These endemics (A. blepharocarpum, A. stylare and A. huetii) are characterized by a unique fruit indumentum of setae, found only in section MENtocus. The other species of this section having the same type of fruit indumentum is A. heterotrichum from Iran, Afghanistan, and a part of Kazakh S. S. R. The remaining three species of section MENIocUS are glabrous-fruited. Of these, A. aureum and A. meniocoides have continuous distributions from southern Turkey to Syria, Lebanon, Palestine, Iraq, and Tran, while A. linifolium is one of the most widespread in the genus, extending from England (where it was probably introduced) throughout Europe and Asia, and east to India. It is probable that as Turkey has the largest 182 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. 46 representation of species and endemics of section MrENiocus, it has been the center of speciation for that section. Certainly, Turkey is the current center of diversity. On the other hand, this is not true for section PsILONEMA, Of which only one species is known definitely to occur in Turkey. Whereas Alyssum granatense is found only in Spain and North Africa, A. damascenum and A. homalocarpum, although both Oriental species, are to be considered as Saharo-Sindian elements, and are not known from Turkey. The widespread taxa Alyssum linifolium of section MeENtocus, and A. alyssoides and A. dasycarpum of section PstLoNeMa are characterized by a wide range of environmental tolerance. Such species as A. aureum and A. meniocoides which have more or less continuous distributions from Turkey into adjacent regions and are steppe-inhabiting or Irano-Turanian elements, are relatively limited by particular habitat preferences. This is also true for the Turkish endemics of section MENtocus, which although found in spatially open habitats, are characterized by a narrow range of environmental tolerance. All species of sections MENtIocus and PsILONEMA are essentially calcicolous. This feature is particularly pronounced in Turkey where the species of section MrENnr1ocus are usually found on calcareous substrates, and is correlated with their preference for the steppe, the soils of which are primarily of limestone structure. The very wide- spread A. linifolium and A. dasycarpum attain their maximum develop- ment in the calcareous steppe and serpentine mountain screes, but tolerate and are relatively abundant in the saline steppe. Similarly, A. blepharo- carpum is known to occur in the salt impregnated portions of the steppe which surrounds many lakes of the Turkish interior. In this and succeeding papers, the strictly European or Asiatic species, which are most closely allied to those occurring in Turkey are included in the keys, and are treated systematically. However, these studies of Alyssum must be regarded primarily as a systematic revision of the Turkish species. To assist in interpretation of the keys and descriptions, in this paper and in those to come, three plates showing the morphological diversity of petals, filaments, fruits, and trichomes are included. These plates illus- trate only the major types of diversity of these structures throughout the genus. Prates II and III show some of the diversity of the distinguishing floral and fruit characters. PLATE IV illustrates basic hair types which comprise the indumentum. These types are arranged more or less in a progressional sequence, from the simplest to the most complex. This does not imply, however, that one type necessarily evolved directly from any other. Many of the fruits pictured in Prater III are normally cov- ered with a characteristic indumentum (Fics. a, c—g, j-o). This indumen- tum has been omitted in order that the fruit shape and configuration not be obscured. I am grateful to Dr. P. H. Davis, University Department of Botany, Edinburgh, Scotland, for permitting the use of these three plates, which were drawn for the forthcoming Flora of Turkey by Mrs. A. Dyer. 1965 | DUDLEY, STUDIES IN ALYSSUM 183 Included in this paper are keys to all currently recognized species as- signed to sections MENrocus and PstroneMa. In the main, all specimens seen have been cited, with the exception of the very widespread and common taxa, such as Alyssum linifolium or A. alyssoides. The specimen citations of these two species, as examples, from many areas in Europe could easily number several hundreds. In such cases I have cited only a few representative specimens; in most large herbaria, however, there will be found abundant additional material. Although I have omitted specimen citations for common European species, the citations of speci- mens of these taxa from the countries of the Near East are as complete as possible since floristic studies of the Near East are still in the explora- tive stages. This policy will be followed also in further papers on Alyssum. The abbreviations for herbaria given by Lanjouw and Stafleu (eds.), Index Herbariorum, ed. 5 (Regnum Vegetabile 31. 1964) are used in this paper. Several not listed in an earlier paper (Jour. Arnold Arb. 45: 58, 60. 1964) are given here. Herbarium of the Botanical Institute of the Academy of Sciences of the Kazakh S. ma-Ata, U.S. S.R. (Aa). Laboratoire de eens de la Faculté des Sciences, Alger, Algérie (AL). Botanisches Museum, Berlin-Dahlem, Germany Instituto Botanico de Barcelona, Barcelona, Spain (Bc). Istituto Botanico dell’ Universita, Bologna, Italy (BoLo). a Museum of Natural History, Department of Botany, Budapest, Hungary (BP). Botanical Institute and Herbarium of J. E. Purkyné University, Brno, Czecho- slovakia (BRNU). Brigham Young University, Provo, Utah, U. S. A. (BRY). Botanical Institute of the University of Coimbra, Coimbra, Portugal (cor). Botanisches Institut der Karl-Marx Universitat, Leipzig, German ny (LZ). : University Herbarium, University of Michigan, Ann Arbor, Michigan, U.S.A. MICH). ; The New England Botanical Club, Inc., Cambridge, Massachusetts, U. S. A. Institutum hotanicute Universitatis Carolinae, Praha, Czechoslovakia (PRC). I would like to acknowledge with gratitude the valuable assistance of Dr. G. K. Brizicky for his translation of pertinent Russian literature and labels, and the editorial advice of Dr. B. G. Schubert. I. Section Meniocus (Desvaux) Hooker KEy TO ALL SPECIES OF SECTION MENIOCcCUS A. Fruits anne smooth, without setae or papillae. B wingless; styles 0.1-0.5 mm. long; petals 1-2(-2. ‘bs x 0.5 m pale ‘vello ae aa ye Ns ide ed eas SR ed Ae ar linifolume ; Seeds winged (wings 0.1-0.4 mm. wide); styles 0.5-1.8 mm. long; petals (1.5—)2-6.5 & 1.5-2 mm., go C. Styles 0.5-1 mm. long; petals oo (1.5-)2-3.5 mm. long; leaves ca. 1.5 mm. wide, + conduplicate. .........- 2. A. meniocoides. to 184 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 C. Styles 1-1.8 mm. long; petals retuse or bilobed, 4—6.5 mm. long; leaves generally 2-4 mm. wide, always fla . were 3. A. aureum. A. Fruits setose, at least on margins, and often papillos D. mae more than 1.5 mm. long; petals ceply bilobed (sinuses 1.5-3.5 m. deep), or retuse; sepals 2-3.5 m E Fruits elliptic, or obovate, (4~)5~7 mm. long, obtuse, setae (0.5-1 mm. long) and papillae always present and dense; seeds wingless; Peta CRO WOO. on ob 8 a Shia os 5 oi as tees 5. A. stylare. Fruits orbicular or ovate, 3.5-5.5 mm. long, emarginate (rarely obtuse) or truncate, setae (0.2-0.5 mm. long) sparse on face of valves, or if only on margins then papillae always present; seeds winged, (wings 0.2-0.3 mm. wide); petals r — bocca eee Ree eer ee Tre ee reer ee A, blepharocarpum. . Styles less than 1.5 mm. long; “pe shallowly eae (sinuses 0.3-0.5 mm. deep) or emarginate; sepals 1-2 . lon F. Petals shallowly bilobed, 2-4 mm. tani seeds winged (wings 0.1- .2 mm. wide); sepals deciduous; leaves increasing in size upwards ; fruits broadly elliptic to orbicular, obtuse or truncate; angen spreading, more than 5 cm. long. A, huetii. F. Petals emarginate, (0.7—)1-1.5 mm. long; seeds aren nee sepals + persistent; leaves decreasing in size upwards; fruits narrowly ellip- tic or obovate, acute or attenuate; inflorescence compact and dense, oR AO ent bee wees eee ke A. heterotrichum. o 1. Alyssum linifolium Steph. ex Willd. Linn. Sp. Pl. ed. 4. 3(1) : 467. 1800(!).— Benth. Fl. Austral. 1: 71. 1863. — Boiss. Fl. Or. 1: 286. 1867.— Ruprecht, Fl. Caucasi. 105. 1869. — Willk. & Lange, Prodr. Fl. Hisp. 3: 834. 1880.— Brandza, Fl. Rom. 136. 1883.— Cosson, Comp. Fl. Atl. 2: 239. 1887. — Fedtschenko, Fl. Turkest. 1: 47. 1906. — Bornm. Verh. Zool.-Bot. Ges. Wien 60: 74. 1910. — Hayek, Prodr. Fl. Penin. Balc. 1: 437. 1925. — Javorka & Csapody, Ic. Fi. Hung. pl. 211, fig. 1589. 1930. — Post & Dinsmore, FI. Syr., Palest. & Sinai, ed. 2. 1: 89. 1932. — Krause, Ankaranin Floru 73. 1934. — Thiébaut, Fl. Lib.-Syr. 1: 72. 1936. Heywood, Repert. Sp. Nov. 64(1): 54. 1961. — Quezel & Santa, Nouv. Fl. Algér. 1: 410. 1962. — Rech. Ark. Bot. 5(1): 169. 1963. — Dudley in Rech. Fl. Lowland Iraq 307. 1964. — Ball & Dudley in Flora Europaea 1: 299. 1964. Syntypes Crimea and Caucasus, Tauria et Armenia, Stephan s.n. (B, non vidi). Lectotype, Tauria, Stephan s.n. (B, non vidi); isolectotype (G-pc). Annual, with many erect, ascending or rarely prostrate stems. Leaves linear and lanceolate (5— )8-10(- 25) & (0.5-)1-2.5 mm.; the lower gradually attenuate, the upper short petiolate. Sepals 12-2 mm. long. Petals only slightly longer than sepals, 1.2-2.4 mm. long, emarginate, and pale yellow to whitish. Long filaments unidentate, 1-1.5 mm. long. short filaments 0.8-1.5 mm. long; appendages 0.3-0.4 mm. long. Fruits obovate or broadly elliptic, obtuse, 3.5-7 « 2-4.5 mm., glabrous; locules 4—6(-8)-ovulate. Styles 0.1-0.5 mm. long. 2n = 14-16. Fl, Feb—July. Two varieties are recognized. —< ase 1965] DUDLEY, STUDIES IN ALYSSUM 185 A. Plant erect; stems stout, greenish; indumentum of stellate ng with few and long rays; fruiting racemes elongate, multi-branched, 3-7 cm. long; leaves 10-25 1-2.5 mm. 2... 0405 ccse cea scans a. a. Var. linifolium. Meniocus serpyllifolium Desv. Jour. Bot. 3: 173. 1814, nomen nudum — non Alyssum linearifolium LaGasca, Gen. & Sp. Pl. 19. 1816. (!). Holotype, jai locis ardis prope Moxente oppidum Regni Valentini, LaGasca 146 on vidi); isotype (G-pc sre linifolius (Steph ex Willd.) DC. Syst. Nat. . ge 1821 (!); DC. Prodr. 1: 165. 1824 raga Se Select. Pl. 2: tab. 42. 1823.— Pri in Ledebour, FI. Alt. 3: 1831; in ‘a chong Fl. Ross. 1: 134. 42.— Mora, Fl. Fan. Esp. . Portugal 6: 572. 1873.—Colmeiro, Pl. aid Hisp,-Lusit. 157. 1885.— Busch in Kuznet., Busch & Fomin, FI. Cauc. Crit. 3(4): 607. 1909; FI. Sibir. & Orient. Extr. 6: 558. 1931; in Fl. U.R.S.S. 8: 359. 1939. — Cadevall & Sallent, Fl. Catal. 1: 147. 1915.— Popov, Man. Fl. Tashkent, fasc. 1 & 2. fig. 222. 1923-1924. — a peieg Fl. Ross. Austro-Orient. 5: 441. pl. 389. 1931.— Grossh. Fl. Kavk. ed. 2 4: 220. tab. 25, i 7. 1950.—I. V. Pavlov (Ed.), Fl. Kazakhstan 4: 282. tab. 35, fig. 12 Mesocus paleo eee Turcz. Bull. Soc. Nat. Moscou 27(2): 297. 1854 (!). olotype, West Australia, Nova Hollandia, collection no. 4, Drummond 127 (LE, non vidi); isotypes (BM, G, GH, K, OXF, W DIsTRIBUTION AND HABITAT: a common and widespread weed of ruderal and cultivated lands, roadsides, vineyards, sandy and conglomerate hillsides, gravelly plains, steppe, Macchie, calcareous and gypsum out- crops throughout most of central, western, and southern Europe, North Africa, and the Middle East, Caucasia, and extending east to Afghanistan and Pakistan, and north to Siberia; alt. 50-2700 m. Naturalized and relatively common in New South Wales of Australia. Spain. Valley of Segura, Bourgeau 577 (E, G, GH, K, W); Cerros del seca nr. Aranjuez, 600 m., Font Quer & Gros 26 (BM, E, G, GH, K, W); Prov. Teruel, Sierra de Valanche, Reverchon s.n. (E); Prov. Almeria, Muria, 1200 m., Rever- chon 851 (£, w); nr. Cazorla, 1200 m., Reverchon s.n. Oe North Africa. Oran, Plateau le Koide, 1000 m., 17 Apr. 1911, Faure (£); Oran, Warion s.n. (E). Russia. Crimea, 1816, Bieberstein (c—pc); Tauria, 1820, Steven (c—Dc) ; Crimea, Busch sn. (w); Odessa, 1846, Nordmann (x); ibid., Rehmann s.n. (E) Podolia, 1820, Andrezejowski (G-pc); Podolia, Besser s.n. (x, w); Illyria, Willdenow s.n. (G-DC); Russian Armenia, Talin, Karmrashen (Karaburun)-Ashnak, 9 June = (cH); Caucasus, Callier 539 (BM, K, W); ibid., Callier 4211, 258 (G); Daghestan, 1874, Becker (K, w); Azerbaidjan, Pichler 150 (G); Transcaucasus, Holmberg 533 (w); ibid., 800 m., 1888, Conrath (c); Tanaim, Goldberg s.n (c-vc); Nachitschevan, Dzhulfa-Darosnan, 3 May 1934, Karjagin (a); ibid., nr. Aliablast, 13 May 1934, Grossheim & Gurvitsh (x); Georgian Caucasus, 186 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 1831, Hohenacker (BM, K, w); ibid., 1838, Hohenacker (c, K, w); Achabzich, Radde 412 (x); Prov. Elisabethpol, nr. Elisabethpol, Apr. 1900, Fedossejew (A, G, W); Prov. Fergana, dist. Kokand, S. of Kanibadam, 1913, Minkwitz (GH); Kurtuk, Kurtu river, 12 May 1930, Serowa & Ryschowa (a); Kirghiz S. S. R., Prov. Semirechensk, dist. Pishpek, Atbashansk, Chu river, Tsintserling (A, GH); Siberia, Suddagh, Pallas sn. (pM); Uskut, Pallas sm. (pm); Dauria, nr. Astracan, 1819, Fischer (G-pc); ibid., 1828, Prestcott (kK); Songarei, 1831, Schrenk (G, GH, w); Siberia, 1819, Sprengel (c-pc); Catherinoslavo, Borysthenes nr. Alexandrovsk, 1865, Griimer (pm); Altai, Ledebour 329 & Meyer 46 (w). Romania. Basarabia, dist. Ismail, nr. Satu, 2-10 m., Borza 654 (£). Turkey. A3: Prov. Ankara, NW. of Beypazari-Nallihan, Kiihne 395 (stu); Kirzbepe, Kiihne 730 (stu). A4: Prov. Kastamonu, Seker-Képrii (Kure-Kastamonu), 1892, Sintenis 3773 (Gc, K, w). A5: Prov. Amasya, Amasya, 400-600 m., 1889, Bornmiiller 1340 (£, K, w). A6: Prov. Tokat, Tokat, 600-700 m., 1893, Born- miller 3244 (G, Kk, w). A7: Prov. Giimiisane, Bayburt-Giimiisane, 1700 m., Stainton 8222 (E). A8: Prov. Giimiisane, Giimiisane, Bourse 39 (w). AQ: Prov. Kars, Kagizman-Tuzluca, Sauer 269 (£, istF); Tusuz, 1800 m., Davis 29576 (A, BM, E, K). A/B8: Prov. Erzurum/Giimiisane, Bayburt-Erzurum, 1853, Huet (e, c). B1: Prov. Manisa, Sipyli (Manisadag), 700-900 m., Born- miiller 9070 (Gc, w). B2: Prov. Usak, Ouchak (Usak), 910-940 m., Balansa 1251 (BM, E, G, GH, K, Ww). B3; Prov. Eskisehir, Eskisehir, Turkish Sugar Co. 475, 477 (BM, E); Prov. Afyonkarahissar, dist. Emirdag-Bolvadin, 10 km. S. of Emirdag, 1100 m., Huber-Morath 13726 (£, HM). B4: Prov. Ankara, Ankara nr. Judyie, 385 m., 1929, Bornmiiller 13853 (BM, Gc); Aokaen, Freres E. C. 244 (G); 13 Apr. 1958, Kahae (stu); Merton 3279, 3286 (£, K); Gérz 15 (BM, G); Prov. Konya, Yavsan Memlehasi nr. Tuz golii, Davis 18691, 31807 (BM, E, K); Prov. Nigde/Konya, 4 km. from Halkenli kéy, W. side of Tuz gélii, 1000 m., Dudley, D. 35928 (&). B5: Prov. Kayseri, plaine de Césarée (Kayseri), 1107 m., Balansa 990 (£, G, GH, K, w). B6: Prov. Maras, Elbistan, 1500 m., Davis 27642 (A, BM, E, K). B7 : Prov. Elazig, Egin (Kemaliye), June 1853, Huet (6, w). B8: Prov. Bonurein: Erzurum, June 1853, Huet (G); Zohrab 376 (kK). C2: Prov. Antalya, Elmali-Korkuteli, 5 miles from Elmali, 1120 m., oe D. 35211 (£). C3: Prov. Burdur, Burdur, May 1845, Heldreich (BM, E, K, G, W); Prov. Konya, Konya-Beysehir, 4 km. from Beypazari, Dudley, D. pe (E). C4: Prov. Konya, Konya-Sultanhani, 18 miles from Konya, 1050 m., Dudley, D. 35918a (a, E); Cumra distr., Kiiciik bug Helbaek 2557 (£); Nigde, Ala dag, nr. Cukur dag, 1900-2060 m., Parry 64 (kg). ‘gndacont a too specimens: Huber-Morath 10990, 12817, 13721, 13724, 14803 (HM); y, D. 35937, 35840 (£); Caria, Kirk s.n. (g): Lycia, Sorkoon (?), a p « K); Acmenia, Szowits sn. (G, GH, K, W); 1867, Calvert & Zohrab (gr, Gc, K); idem 45, p-P- (OxF). Syria. Quaryatein, Davis 5726B, 5602 (£, K); Quaryatein -Rat Tush, Davis 5634 (£, K); Russell s. n. (pm, cG-pc). Palestine. Wadi Musa-Moan (Transjordan), 1219 m., Davis 8691 (gz, K); Plaine de Amalites, Apr-May 1846, Boissier (G); Ain-Musa, 16 Apr. 1929, Eig et al. (G, w); 20 km. S. of ee: 18 gf 1923, idem on w); Wadi Hasa-Ain Musa, 17 Apr. 1929, idem 306 (kK); Jazira, 20 km. NW. of Falija, Guest & Rawi 13649 (xk); Mosul- Kirkuh, Guest 625 (x); Karbala liwa, ca. 15 km. W. of Karbala, Gillett & Rawi 6375 (K); 18 km. S. of Rutba, 640 m., Rawi 14634 (x); Thukhaib, 280 m., Rawi 14795 (kK); 50 km. E. of Samarra. Asila, 10 km. N. of Sha haikh SS ee ee 1965 | DUDLEY, STUDIES IN ALYSSUM 187 Mohamma, 50 m., Rawi 20480 (x); ibid., Rechinger 13452 (w); 4 km. E. of Samarra, 65 m., Rawi 20329 (x); ibid., Rechinger 13500 (w); Truleal-e-Has, Rechinger 4330 (w); Mesopotamian desert, 15 Sept. 1919, Watson (x); Oguhah, Graham 25 (pM); Euphrates, Meskare-Der-es-Sor, Sabcha-Tibne, 250- 350 m., Handel-Mazzetti 542 (w); Kaijum-Abukenal, 120-180 m., Handel- Mazzetti 652 (w); Mejadin-Salbije, 180 m., Handel-Mazzetti 632 (w); Assyria, Kerkcik, 1893, Bornmiiller 894 (Gc); Jabal Hamrun, Muqdadija (Sharaban)- Jalaula, Rechinger 14212 (w). Iran. Prov. Khorasan, Turbat-e-Haidari, 1300 m., Rechinger 4346 (Gc, w); ibid., ca. 30 km. from Meshhed, Rechinger 1502 (w); Mt. Kopet Dagh, nr. Alamli, 2000 m., Rechinger 4670, p.p. (w); Djenaran- Kucan, Rechinger 7519, p.p. (w); Schiras-Kamareyi, 200-600 m., Apr. 1868, Haussknecht (BM, G); Sultanabad nr. Kaswin, 1524 m., Lindsay 29, 32 (pm); Transcaspian, 1900, Sintenis 165 (G, K, W); 10 km. E. of Zorab, 1667 m., Cowan & Darlington 1615 (x); Prov. Hamadan, Kharaghan (Hamadan), Sabeti 94 (w); Faghire nr. Hamadan, Sabeti 225 (w); Aq Bulaq, ca. 100 km. N. of Hamadan, Rioux & Golvan 210, 213 (w); Mt. Elburs, Demawend, 2640 m., 22 July 1936, Gilli (w); Keredj, Rechinger 526 (w); ibid., 27 Apr. 1934, Gauba (w); Mazanderan, Bashm Kuh (Shahmirzad), N. of Baslm, 2700 m., Wendelbo 1341 (BG, £); Haraz valley, W. of Siah Bisheh, 700 m., Wendelbo 428b (xe, E); Bakhtiari, Oregon, Damane-Kuh, 2300 m., Wendelbo 1727 (Bc, £); Kerman, Kerman-Saidabad (Sirdjan), Mashiz-Khan-e-Sorck, 2000-2580 m., Rechinger 3051 (£, G, K, w); Mt. Djamal Bariz, Bam-Djiroft, Deh Bakri, 2100 m Rechinger 3805 (G, K, W); Rescht-Teheran, Ruiobar-Mendschel, 300-400 m., 1902, Bornmiiller 6243 (G). Kuweit. Shaiba, 23 Aug. 1919, Watson & Sharples (kK). Afghanistan. Hari-rud valley, Aitchison 160, 430, (GH, kK); Istalif, 1800- 1840 m., Gilli 1087, p.p.; 1088 (w); Kamardtal, W. of Duab, 1630 m., Gilli 1078 (w); Prov. Kabul, Chord Kabul, E. of Kabul, 2280 m., Gilli 1090 (w); Koh-i-Asmir, 1900 m., Wendelbo 2733 (BG); ca. 15 km. W. of Sarobi, 1100 m., Wendelbo 2791 (Bc); Sarobi, Volk 2454 (w); 20 km. E. of Lataband, 1800 m., Wendelbo 3022 (pc); Qual-e-Eslan, 25 km. S. of Kabul, 1830 m., Wendelbo 3204 (Bc). Pakistan. North-West Frontier, Parachinar, Kurram Valley, 2134 m., Stewart 28117 (micH); Stewart 14738 (kK); Peshin, 1550 m., Lace 3308 (E, K); Murqudochen, Stokes 942 (Kk). Specimens intermediate between the two varieties are known to occur in Afghanistan: Otipore, Chokey and Korobat, Griffith 1519, ex hb. Lehmann (K); Griffith 1415 (GH). Alyssum linifolium is one of the most widespread species in the genus, and is quite variable in the size of its leaves and fruits, and in its stature. However, with the exception of the variety which follows (var. teherani- cum), the minor variations do not warrant nomenclatural recognition. One interesting variation occurs on plants (e.g. Davis 18691 and Dudley, D. 35928) growing in extreme saline habitats. The leaves of these plants are very narrow and conduplicate, resembling those of A. meniocoides. This character, however, is not constant, for the plants assume a normal appearance immediately outside the salt areas. nor The closest allies of Alyssum linifolium are A. meniocoides and A. aureum, both of section Mrentocus. From these, A. linifolium may be easily distinguished by its wingless seeds, much shorter styles, and smaller floral parts. 188 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 Professor Zohary (Palest. Jour. Bot. Jer. ser. 2(2/3): 162. 1941) maintains that Alyssum minimum L. (Sp. Pl. 2: 651. 1753) is the correct binomial for this species, and that A. linifolium is a synonym. Examina- tion of Linnaeus’s specimen of A. minimum (LINN. 828:8) proves with- out any doubt that this is not the case; A. minimum L. can only be treated as a synonym of Lobularia maritima (L.) Desv. (Basionym: Clypeola maritima L. Sp. Pl. 2: 652. 1753). A. minimum sensu Willd. (Willd. Linn. Sp. Pl. ed. 4. 3(1): 464. 1800) clearly is not the same taxon as A. mini- mum L., but may be identified rather as A. desertorum Stapf. b. Var. teheranicum Bornm. Bull. Herb. Boiss. II. 4: 1269. 1904 (!). —Parsa, Fl. Iran 1: 746. 1952. Holotype, Iran, in vallis oppidi Teheran, 1150 m., 20 Feb. 1892, Bornmiiller 2155 (B, non vidi); isotypes (BM, E, G, K, OXF, W). Alyssum (Meniocus) cupreum Freyn & Sint. Bull. Herb. Boiss. II. 3: 695. 1903 (!).— Fedtschenko, Fl. Turkest. 1: 47. 1906. Holotype, Russia, Regio Transcaspica, Krasnowodsk in arenosis montium, 17 Mar. 1900, Simtenis 18 (BRNU, non vidi); isotypes (BM, E, G, K, W Meniocus Aegon f. microcarpus Busch in ‘Kuaet,, Busch & Fomin, FI. Cauc. Crit. 3(4): 610. 1909. Holotype, Russian Anneni, in tractu Bort- et in Somchetia, 1837, Koch 143 (LE, non vidi). Alyssum linifolium var. cupreum (Freyn & Sint.) Dudley in Hedge, Arbok Univ. Bergen, Mat.-Naturv. No. 13: 6. 1963 (!). DisTRIBUTION AND HABITAT: scattered in extreme steppic conditions, loose gravel, dry limestone hillsides, desert, and serpentine substrates of Turkey, Syria, Iraq, Iran, Caucasia, Azerbaidjan, Afghanistan and Paki- stan; alt. 200-2000 m. Turkey. A2 (E): Prov. Istanbul, Rumel Hissar, 16 May 1915, Post (G). A4: Prov. Ankara, Cubuk, 1000 m., Markgraf (z). B4: Prov. Ankara, 7 Apr. 1958, Kiihne (stu). C4: Prov. Konya, Agios Philippos (Hagios Phili- bos), Post 15 (G). C5: Prov. Nigde/Adana, Ulukisla-Pozanti, 900 m., Davis 300 (BM, E, K). Cappadocia, 1834, Montbret (x). Asia Minor, "Aucher 280 (G, K); ibid., Aucher 4100 (BM, G, K, W). Syria. Palmyra, 200 m., Dinsmore 22497 (K). Iraq. Dist. Kiruk (Kurdistan), ad confines Persiae, Khanaquin, Rechinger 14128 (w); Mosul, 200 m., Bugloss 8 (kK). Iran. Prov. Teheran, Kishlak (Garmsar), Seman-Teheran, 900 m., Rechinger 2773b (w); nr. Teheran, 1220 m., 1892, Bornmiiller 2154 (£, Gc, K, w); ibid., Schmid 5102 (c); Kom, 1892, Bornmiiller 2153 (G); Chononsar, 1900 m., 1892, Bornmiiller 2151 (G, K, w); Bornmiiller 2156 (BM, E, G, K, OXF, W); Teheran-Davudieh hills, 1400 m., Wendelbo 97a (Bc, E); Persepolis, Kotschy 1053 (c, w); S. of Tabriz, Gilliat- Smith 1351, 1352, 1336, 1338 (K); 39 km. W. of Kermanshah, 1372 m., Cowan & Darlington 2618 (K); Shershah, Mar. 1859, Bunge (c, K); Kerind, Evans 37 (£); Prov. Kazvin, Keredj, nr. Kalak, 1600 m., Rechinger 2745 (c, w); Mt. Elburs, Keredj-Kalak, Rechinger 143 (w); Prov. Isfahan, Kuh Pah, 1700 m., Rechinger 2714 (w); Abadeh-Daulatabad, 1500-2000 m., Schmid 5313, 5318 (G); Fars, Takht-i-Jamshed, Koelz 14420 (£, w); Prov. Khorasan, Robat Safid, 800-2000 m., Rechinger 7335 (w); Mazanderan, Haraz valley above Panjab. 1300 m., Wendelbo 303 (BG, E); Prov. Baluchistan, Khash (Vasht)-Iranshahr 1965 | DUDLEY, STUDIES IN ALYSSUM 189 (Bampur), Mt. Karvandar, 1500-1600 m., Rechinger 3958 (w). Russia. Azer- baidjan, Zelizabethpol (Elizavetpol), 1882, Pichler (c, cH, w); Krasnowodsk, 1900, Sintenis 17 (£, G, K); 1900, Sintenis 19 (Gc). Afghanistan. Kabul-Pagh- man, 1880-1910 m., Gilli 1089a, 1089b (w); Kabul, Scher Darwasah, 1790- 1840 m., Gilli 1081, 1082, 1083, 1085 (w); ibid., Neubauer 537 (w); ibid., Koh-e-Tschelsotun, 1810 m., Gilli 1084 (w); Kabul, Gilli 1091 (w): E. of Kabul, Budchak, 1770 m., Gilli 1080 (w); Dschmal Baba, S. of Kandahar, 1000 m., Gilli 1077 (w); Kodananebene-Istalif, 1750 m., Gilli 1079 (w); Istalif, 00 m., Gilli 1087, p.p. (Ww); Sarobi, Volk 2455, p.p. (w); Pule Surkh nr. Tsharikar, Neubauer 542 (w); nr. Kabul, NW. of Aliabader Mt., 1800 m., Gilli 1086 (w); ibid., Collett 12 (Kk). West Pakistan. Quelta, Kitta Aboulla, Duthie 8577, 8578 (G, K). The differential characters of Alyssum linifolium var. teheranicum are consistent in small and scattered populations throughout the southeastern range of the species. The type specimens of var. teheranicum are morpho- logically identical to those of A. cupreum, the latter name is, therefore, placed in synonymy. Bornmiiller used yet another varietal name for this taxon in exsiccatae, but without a Latin description. This epithet re- ferred to the dense, metallic-colored indumentum. bo . Alyssum meniocoides Boiss. Ann. Sci. Nat. Paris II]. 17: 158. 1842 (!).—Boiss. Fl. Or. 1: 286. 1867. — Handel-Mazzetti, Ann. Naturh. Mus. Wien 27: 52. 1913.— Boul. Fl. Liban. & Syr. 32. pl. 38, fig. 11. 1930.— Post & Dinsmore, FI. Syr., Palest. & Sinai, ed. 2. 1: 89. 1932. — Thiébaut, Fl. Lib.-Syr. 1: 72. 1936. — Parsa, Fl. Iran 1: 747. fig. 621. 1952.— Rech. Ark. Bot. 5(1): 170. 1963. — Dudley in Rech. Fl. Lowland Iraq 307. 1964. Holotype, Meso- potamia, Aucher 281 (c); isotypes (BM, K, OXF). Meniocus filifolius Jaub. & Spach, Ill. Pl. Or. 1: 107. tab. 53B. 1843 (!), non Alyssum filifolium rating 1826. (Cf. Dudley, Jour. Arnold Arb. 45(3): 372. 1964). lotype, —— 281 (Pp, non vidi); isotypes (BM, G [holotype of A. psa Alyssum tetraspermum Bertol. aera Bot. 2: 12. 1843 (!). Holotype Turkey, C6: a ex campis ad Portum ego ve of Birecik), Mar. 1836, Chesney 25 (Boo, non vidi); isotypes (BM, Alyssum kermanshahensis Cowan ex Parsa, Fl. Iran 1: 733. ra eE 1952 (!). Holotype, Persia, 39 miles E. of Kermanshah, 1372 m., 29 Mar. 1929, Cowan & Darlington 354 (x). Annual, stems slender, 3-10 cm. tall. Leaves (5.5—)10-18 mm. long, + conduplicate. Petals obovate, entire, (1.5-)2-3.5 mm. long. Long filaments 1.5—2 mm. long, terminated with bifid apices 0.3—-0.4(—0.5) mm. long. Short filaments 1-1.5 mm. long, with always bifid appendages 0.5-1 mm. long. Fruits glabrous. Styles 0.5—-1 mm. long. Seed wings 0.1-0.2 mm. wide. Fi, Feb.—Apr. DIsTRIBUTION AND HABITAT: fallow fields, steppe, limestone slopes, and sometimes associated with Quercus aegilops forests in Mesopotamia of 190 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 southeastern Turkey, the Syrian desert, Lebanon, Palestine, Iraq, Iran and Afghanistan; alt. 100-2000 m. Turkey. C6: Prov. Gaziantep, Aintab (Gaziantep), Apr. 1886, Shepard (cH, K); 10 Apr. 1884, Post (Bm); Yonas, Euphrates, 25 km. E. of Gaziantep, 914 m., Haradjian 1770a (£, G, K, w); Bal Zus (Balkis) nr. Birecik, — m., Har- adjian 1043 (Gc, w); Merza nr. Birecik, 1888, Sintenis 131 (BM, E, G, K, W); Prov. Hatay, Amurk nr. Hassa, Amanus dag-Kurt dag, Haradjian 280 fe w). C7: Prov. Gaziantep, Rum Kala (Halfeti), 1888, Simtenis 157, approaching A. aureum (BM, E, G, K, W); Prov. Urfa, Djebel Taktak (Tektek dag) Apr. 1867, Haussknecht (BM, K, W). C8: Prov. Mardin, Mardin, 1894, Post (BM, G). Syria. Isiayah, 1 Apr! May 1900, Post (pm, G); Nebk, Davis 5528, approach- ing A. aureum (£, K); Dayr-’Atiyyah, Post (pM): ibid., 1200 m., Post 13815 (E, G, K); ibid., May 1879, Post (pm); Aleppo, Russell s.n. (pa); ibid., Lesier (Gc, K); Palmyra, 400 m., Dinsmore 20497 (x); ibid., Snoi-Teida, Djebel Abour, tak 2887 (Gc); Pamascis, Apr. 1928, Druce (xr); ibid., 14 Apr. 1894, Péronin (G); ibid., Djebel Kharbi, Gaillardot 1546 (c); ibid., Sasoa-Kisive, 671 m., Feb. 1945, Norris (pM); 10-15 km. from Damascus, 503-914 m., 4 Mar. 1945, Norris (BM); nr. Baalbak, 12 Mar. 1867, Fox (kx); Horms-Hama, 2000 m., Haradjian 4099 (c); Armel Wir’al-Am’El Beidha, 8 Apr. 1890, Post (BM). Palestine. Busrah-Kurayyah, 900 m., Dinsmore 2460, approaching A. aureum (Gc, K); Moab-Qual’at Ziza, Feinbrun _ Zohary 327 (BM, E, G, GH, K, W); Amman-Ziza, 15 Apr. 1929, Eig et al. (G, w); Ziza, 700 m., ‘Dione 11815 (E, G, K); Es-Salt--Amman, Apr. 1895, ae (Gc); Uinel | Amm ud, 15 Apr. 1929, Fig et al. (G, w). Lebanon. Wadi Karn, Yabrud (Zebrad), Post 88 (Gc); Zefer- ya-Beyrout, 21 May 1881, Péronin (Gc); Beyrout plain, 100 m., Maitland 70 K). Iraq. Zawita, Mosul liwa, Polunin et al. 47 (BM, E, G, GH, K); Hieropolis, Mar. 1867, Haussknecht, p.p. with A. aureum (Gc, K, w); Euphrates, Abu Herera. Meskene-Der-es-Sor, 205-350 m., Handel-Mazzetti 424, approaching A. aureum (w); Tuz Khurmatli, Rogers 349 (x); Kirkuh, Rogers 74 (kK); Duleam liwa, 6 km. above Ana, 130 m., Gillett & Rawi 6967 (x); Erbil liwa, Slahaddin, 1000 m., idem 10423 (K); er iReepate liwa, Givija forest, 1100 m., idem 10623 (K); L’Alders, — Hanna (bm). . Teheran, Davudieh hills, 1400 m., Wendelbo 97b (BG, E); distr. oe ee peer aeaiig 22 km. E. of Kermanshah, 1280 m., Bent & Wright 112, p.p., 125, p.p. (w); 39 miles E. of Kermanshah, 1372 m., Cowan & Darlington 256 (x); Prov. Luristan, Durud, Koelz 17103 (£, MICH, w); Teheran, Farahabad, Sabeti 182 (w); nr. Kaswin (Mazraeh), 1200 m., Schmid 5050, 5045 (c); ibid., Schmid 5055 (w); ibid., Stutz 711 (BRY, W)- Afghanistan. Obeh, 1700 m., Keie 3782 (w); Prov. Bamian, Danak Siakr, 10 km. from Doab, 1500 m., W endelbo 3418 (BG, E). In collections containing both Alyssum meniocoides and A. aureum in has any apparent ecological or altitudinal preferences; however, A. aureum replaces A. Eee in central and eastern Anatolia. The petals of some specimens (e.g. Sintenis 131 & 157) approach in size those of A. aureum,; sicseiae the leaves of these specimens are somewhat atypical 1965 | DUDLEY, STUDIES IN ALYSSUM 191 and more or less resemble those of A. aureum. In all other respects, however, these plants possess all the other characters of A. meniocoides. Collectors have frequently confused A. meniocoides with A. linifolium, which occurs commonly throughout the same geographical areas, but the former has winged seeds, longer styles, larger floral parts, and generally narrower and conduplicate leaves. 3. Alyssum aureum (Fenzl) Boiss. Fl. Or. 1: 286. 1867 (!).— Boul. Fl. Liban. & Syr. 32. 1930. — Post & Dinsmore, Fl. Syr., Palest. & Sinai, ed. 2. 1: 89. 1932. — Thiébaut, FI. Lib.- Syr. 1: 72. 1936. — Rech, Ark. Bot. 5(1): 166. 1963. Meniocus aureus Fenzl, Pug. Pl. Nov. Syr. & Taur. Occid. 1: 13. 1842 (!). i Syria, circa Aleppo, 22 Mar. 1841, Kotschy 27 (w); isotypes BM, GH, K, OXF. eee peli RANE Taub, & Spach, Ill. Pl. Or. 1: 105. tab. 53A. 1843 (!). Syntypes, Kotschy 27 & Aucher 4100 (BM, E, G, K, W). as ctotype, Kotschy 27 (Pp, non vidi); isolectotypes (BM, E, G, GH, K, OXF, sa pleiospermum Fenzl, Ill. & Desc. Pl. fut A & Taur. Occid. 54. 3 (!). Holotype, Kotschy 27 (w), isotypes (BM, E, G, GH, K, OXF). ‘ies meniocoides var. aureum (Fenzl) Zohary, Pal. Jour. Bot. Jer. Ser. 2(2/3): 162. 1941 (!). Annual, with stout stems, (3—)5—20 cm. tall. Leaves (8.5—)12-30 mm. long, and always flat. Petals golden, spathulate, retuse or bilobed, 4—6.5 mm. long. Long filaments 2-4 mm. long, wing terminated with lanceolate or bifid apex, 0.5—1 mm. long. Short filaments 2-3 mm. long with lanceo- late or bifid appendages 1-1.5 mm. long. Fruits oot Styles 1-1.8 mm. long. Seed wing 0.2-0.4 mm. wide. F/. Mar.—Jun DIsTRIBUTION AND HABITAT: dry cultivated lands, steppe and marly vineyards in Lycaonia, the Upper Euphrates region and Mesopotamia of Turkey, the Syrian desert of Western Syria and Palestine; alt. 400-2500 m. Turkey. B6: Prov. Malatya, wos (Darende), 18 Apr. 1917, McDaniels (cu). B7: Prov. Erzincan, Siirek, 1890, Sintenis 130b (G); Prov. Elazig, Elazig- Kale, 22 km. E. of Elaziz, 1300 m., Davis 28938 (BM, E, K); Harput, Noé 857 (G); Prov. Malatya, Malatya, 1000-2500 m., Ajtaikovitch (w): Arapkir-Denizli, 1889, Sintenis 153 (£, w). C4: Prov. Konya, Ayos Philibos (Hagios — nr. Konya, Post B16 (c); Cumra dist., Kiiciik kéy, Helbaek 2400 (Ee). C6: Prov. apices Yonas, 25 km. E. of Gaziantep, 914 m., Haradjian 17706 ia C7: Prov. Urfa, Djebel Taktak (Tektek dag), Apr. 1867, Haussknecht, p.p. (sm). Syria. Turmainen nr. pre 396 m., 12 Mar. 1865, Haussknecht, p.p. with A. meniocoides (Gc, w); Aleppo, Haussknecht 98 (c); Yab Sam, 400 m., Dinsmore 20409 (c, K); ok Rogers 599b (kK); Djebel Muhassan, 12 Mar. 1863, Haussknecht (c); El Jebath-E] Beithata, Zz Apr. 1890, Post, p.p. with A, meniocoides (BM); Hauran, Post s.n. (BM). estine. Amman- Ziza, Eig & Zohary 1929 (Huy); Amman-Abu Jaber, 1000 m., a 2838 (Ww q. nr. Mosul of Kirkuk, Guest s.n. (K); Baghdad nr. Kamaracha, 1822, Olivier 192 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 (c); Hieropolis, Mar. 1867, Haussknecht, p.p. with A. meniocoides (Gc, K, W); ibid., 1782, Michaux (c). The species Meniocus pleiospermum and M. grandiflorus are based on the same type material (Kotschy 27) as M. aureus, the basionym of Alyssum aureum, The other syntype of VM. grandiflorus pital 4100) is correctly referred to Alyssum linifolium var. teheranicu Zohary (Pal. Jour. Bot. Jer. Ser. 2(2/3): 162. 1941) ener that the differences between A. aureum and A. meniocoides were very slight, and recombined A. aureum as a variety of the latter species. A. aureum may be distinguished from A. meniocoides by its considerably longer styles, larger and retuse or bilobed petals, larger fruit, and usually flat and wider leaves. Although the distinction in petal size and leaf form occasionally breaks down (cf. note under A. meniocoides), the much longer styles, and the always retuse or bilobed petals remain diagnostic for A. aureum. Furthermore, these species retain their identity in regions of overlap, and they clearly have different flowering times. The inclusion of A. aureum by Parsa (FI. Iran 1: 747. fig. 622. 1952) is probably an error. With the exception of this reference, A. aureum has not been recorded from Iran, and no Iranian specimens have been seen at Kew, upon whose collections Parsa based his work. These ma- terials, and Parsa’s illustration lead me to conclude that A. aureum sensu Parsa is, in fact, A. minutum. > . Alyssum huetii Boiss. Fl. Or. 1: 287. 1867 (!). Syntypes, Turkey, 11 May 1869, Bourgeau; Balansa 1252, Kotschy 206, and June-4 July 1853, Huet. Lectotype, Turkey, B8: Prov. Erzurum, in neglectis circa Erzurum, 1829 m., June—4 July 1853, Huet (c); isolecto- types (BM, K, OXF). Pt. II, rics. i, s. Px. III, Fic. a. Px. IV, Fic. b. TexT-Fie. 1. Meniocus hirsutus Boiss. & Bal. in Boiss, Diagn. 3(5): 32. 1856 (); non Alyssum hirsutum Bieb. Holotype, Turke 2: Prov. Usak O uchak ey, B (Usak), 910-940 m., 21 May 1857, Balansa 1252 (G); isotypes on, GH, K, OXF, W). Annual, with ascending or erect stems up to 40 m. Leaves linear or oblanceolate, (8—)30-40 x (0.5—)2-3 mm., + conduplicate, acute, in- creasing in size upward. Racemes simple, ie rarely branched, (5—) 10-20 cm. long. Pedicels spreading to horizontal, 4-6 mm. long. Sepals early deciduous, 1.7-2 mm. long, acute. Petals obovate-clavate, shallowly bilobed, (2—)3—4 & 1mm. Long filaments 2-2.5 mm. long, with unilateral wings and teeth (teeth 0.5 mm. long), never exceeding anthers. Short filaments 1.2-1.5 mm. long with connate, lanceolate and acute, or bifid appendages, as long or longer than filaments. Fruits elliptic or orbicular, obtuse or truncate, 4.5-6.5 3-3.5 mm., papillose with + sparse, slender 1965 | DUDLEY, STUDIES IN ALYSSUM 193 and simple tuberculate setae (0.3-0.5 mm. long); locules 4~6-ovulate. Styles stout, 0.5-1 mm. long, + dilated basally. Seeds narrowly winged. FI, May-June. DISTRIBUTION AND HABITAT: an Anatolian endemic of disturbed sites, cultivated lands and steppe, scattered mainly in Inner Anatolia, from western, southwestern and central Turkey, and extending east to the Armenian Highlands; alt. 800-2500 mm. Turkey. A3: Prov. Ankara, 10 km. W. of Beypazari, 2 km. W. of Zavije, 1000 m., Kiihne 202 (stv). ‘AS: Prov. Erzurum, Tortum, Calvert 1240 (Gc). A/B4: Pr rov. Ankara, nr. Indize-su, 800-900 m., 1929, Bornmiiller 13853 (BM, GH, K, w). B4: Bias, Ankara, Ankara, Kotte 1021 (kK); Hussein nr. Ankara, Kotte 121 ag ee Haymana, 9 km. W. of Haymana, 1808 m., Huber-Morath 13723 (HM). * Prov: rage Argaei (Erciyas dag) nr. Tpshamaki, 1600—- 2500 m. ee 206 (G, w). : Prov. Erzurum, Erzurum, Calvert & Zohrab 1240 (er, G, K, OXF, w). C2: ai Anatalya, Elmali dag, 11 May 1869, Bour- geau (E, GH, OXF, w); Elmali-Korkuteli, 5 km. from Elmali, 1100-1120 m., Dudley, D. 35212 (£) & Dudley, D. 35230 (a, rE). The closest ally to Alyssum huetii in Turkey is probably A. stylare, which is also an Anatolian endemic. It may be distinguished from A. stylare by its shorter and stouter styles, smaller floral parts, and sparser fruit indumentum of shorter and slender setae, and — It is also Closely related to A. heterotrichum from Iran and Rus The fact that the vast area of east-central eos has been little A. huetii from that area, which intervenes between the presently known areas of distribution of this species. TExt-FIG. 1. Map showing distribution of Alyssum huetii (A), and A. stylare (Q), in Turkey. 194 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 5. Alyssum stylare (Boiss. & Bal.) Boiss. Fl. Or. 1: 287. 1867 (1). Meniocus stylaris Boiss. & Bal. in Boiss. Diagn. 3(6): 16. 1859 (!). Holotype, Turkey, B5: Prov. Kayseri, inter segetes ad basin montis Karamasdagh quinque leucis ad orientem urbis Caesareae siti, c. 1500 m., June & 2 July 1856, Balansa 486 (G); isotypes (A, K, OXF, W). Text-Fic. 1. Annual, similar to, but larger and coarser than Alyssum huetii. Leaves (2-)8-20 x 1.8-3 mm., + conduplicate, obtuse, increasing in size upward. Inflorescence lax, multibranched and somewhat circinate, (8—) 10-20 cm. long. Pedicels spreading to ascending, 5-8 mm. long. Sepals + persistent, 3-3.5 mm. long. Petals deeply bilobed, spathulate and clawed, 4-6 < 2-3 mm. Long filaments 3-4 mm. long, with unilateral teeth (0.5-0.7 mm. long) sometimes exceeding anthers. Short filaments 1.5-2.5 mm. long, with free (rarely connate) lanceolate appendages, as long or longer than filaments. Fruits elliptic to obovate, obtuse, (4—)5-7 x (2.5-)3-4.5 mm., 4-8-ovulate, papillose with very dense, simple or furcate setae (0.5-1 mm. long). Styles slender 2—2.5 mm. long. Seeds wingless. Fl, May-July. DiIsTRIBUTION AND HABITAT: a Turkish endemic of central and eastern Anatolia, the Upper Euphrates, the Cilician Taurus and the Anti-Taurus, in neglected fields, high steppe, and often associated with Quercus- Poterium scrub; alt. 1300-1850 m. Turkey. A6: Prov. hs igre eangr = Be Ae 1600 m., 1858, Tchihatchef (c). : Prov. Giimiisane, Giimiisane, 21 May 1862, Bour- geau (G); Molirva- pe ue: Sorda), fod. laine 5656 ed E, G, GH, K, w). A8: Prov. Giimiisane, dist. Bayburt, Giimiisane-Bayburt, 21 len. from Bayburt, 1620 m., Huber-Morath 14802 (um); Bayburt, 11 July 1862, Bour- geau 171 (Gc). A/B6: Prov. Sivas Tokat, Sivas-Tokat, N. side of Artova pass, 26° 26° 30° 32° E=a 36° s a 2. Map showing distribution of Alyssum blepharocarpum (Q) in urkey. 1965 | DUDLEY, STUDIES IN ALYSSUM 195 Camlibel dag, 1850 m., 14 June 1939, Reese (HM). B6: Prov. Maras, dist. Elbistan, Elbistan-Darende, 28 km. NE. of Elbistan, 1300 m., Huber-Morath 12821 (HM). B7: Prov. Erzincan, Sipikor dag (nr. Kesis daz), Jerbatan, 1889, Sintenis 1543 (Gc, K) & 1890, Sintenis 3123 (Kk). C5: Prov. Nigde, Pursuk nr. Ulukisla), 1300 m., 1898, Siehe 89 (BM, G, K, W). Among the species of section MENtocus, Alyssum stylare has the longest styles, and the densest fruit indumentum which is composed of tubercu- late, simple (occasionally bi- or trifurcate) setae to 1 mm. long. The furcation of the hairs on the fruit is unique in the section. Although Alyssum stylare is sympatric in the eastern part of its range with A. blepharocarpum, it always has larger and obtuse fruits, deeply bifid petals, and wingless seeds. Its larger flowers, denser fruit indumen- tum, larger and broader leaves, stricter habit, and much longer, circinate inflorescences distinguish it from A, huetii - Alyssum heterotrichum Boiss. Diagn. 1(6): 15. 1845 (!).— Boiss. Fl. Or. 1: 287. 1867; in Buser, Suppl. Fl. Or., 50, 1888. — Parsa, FI. Iran 1: 248. fig. 620. 1952 —I. V. Pavlov (Ed.), Fl. Kazakhstan 4: 281. tab. 35, fig. 10. 1961. Holotype, Iran ad muros hortorum prope ruinas Persepolis, 11 Apr. 1842, Kotschy 224 (c); isotypes (BM, E, K, OXF, W). A, bungei Boiss. Fl. Or. 1: 274. 1867 (!). Parsa, Fl. Iran 1: 727. 1952. Holotype, Iran, hab. in Persia, inter Isfahan et Teheran, May 1859, Bunge (G A, SeReaes nah ie Bull. Mosk. Obshch. Isp. Pri. Biol. 52(2): 87. fig. 1 1947. Holotype, Kasahstania, desertum Betpak-dala centralis in loco Kok- aschik, 20 Apr. 1940, Rubtzov (aa, non vidi). Annual, often low and decumbent, rarely more than 10 cm. tall. Leaves oblanceolate-spathulate, flat, 7-25 < 1-4 mm., decreasing in size upward. Inflorescence dense, 5 cm. or less long, frequently strongly branched with lateral branches ca. % as long as terminal. Pedicels ascend- ing, 1.5-3 mm. long. Sepals +. persistent, obtuse, 1-1.5 mm. long, with very sparse indumentum. Petals obovate, emarginate, 0.7—1(-1.5) x 0.5-0.7 mm., greenish yellow. Long filaments 1-1.5 mm. long, with uni- lateral teeth (ca. 0.5 mm. long). Short filaments 0.7-1(-1.5) mm. long, with deeply bifid and basally connate appendages, % as long as filaments. Fruits narrowly elliptic-ovate or obovate, acute or attenuate (rarely otherwise), 3-4.5(-5) 1.5-3 mm.; densely papillose and with simple, slender, and only slightly tuberculate setae (0.2-0.4 mm. long). Styles slender, 0.4-0.7 mm. long, glabrous. Seeds wingless. Fl. Jan—Apr. DIsTRIBUTION AND HABITAT: dry rubble and mountain steppe in Iran, Afghanistan and Kazakhstan in Russia; alt. 1200-2000 m. Iran. Prov. Khorasan, pe geen Rechinger 7519, p.p. (w); Montes Kopet-Dagh, in jugo Alamli, ca. 2000 m., Rechinger 4760, p.p. (w); Montes Hazar Masdjid, Aedak Balen: "1200-1600 m., Rechinger 5047 (w); Prov. 196 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 Hamadan, Aq Bulaq, 35° 36’ N., 48° 27’ E., ca. 100 km. N. of Hamadan, Rioux & Golvan 211 (w); Niriz, Fars, Koelz 14728 (w); Shiras, 1425 m., Pravitz 507 (s); nr. Schiraz, Dilguscha, Mar. 1868, Haussknecht (BM, G, K, W). Af ghanistan. Prov. Pawan, 22 at above Gulhaliar in Panjshi valley, 1700 m., Hedge & Wendelbo 3005 (k£). The type specimens of Alyssum bungei do not deviate in any characters from material of A. hketerotrichum, and accordingly the name, A. bunget, should be treated as a synonym of A. heterotrichum. Boissier considered A. bungei to be the only annual species of section ODONTARRHENA, but he commented that this was an artificial placement based entirely on his observation of its uniovulate fruit locules. I have examined fruits from the type material of A. bungei and have found that the fruit locules con- sistently contain four or five ovules. It is true, however, that only one of these ovules develops into a mature seed, while the others abort. In this case, the fruits are one-seeded and the aborted ovules are visible under magnification. A. heterotrichum is the only species in section MENIOCUS having setae on the fruits, and an entirely extra-Anatolian distribution. 7. Alyssum blepharocarpum Dudley & Huber-Morath, Jour. Arnold Arb. 45(1): 61. pl. J, figs. 1-13. 1964 (!). Pu. I. Text-Fie. 2. Holotype, Turkey, B4: Prov. Ankara, dist. Kadinhan, Sarayoni- Cihanbeyli, Weizenfeld, 13 km. nordéstlich Sarayénii, 870 m., 1 June 1956, Huber-Morath 13722 (HM); isotype (E). Annual, resembling Alyssum huetii, but of more delicate habit. Leaves linear, rarely subspathulate, (3.5—)8-20(-25) XX 0.5-2 mm., usually de- creasing in size upward. Sepals (1.5—)2.5-3 mm. long. Petals 2-4 mm. long, retuse. Long filaments (1.5—)2-3.5 mm. long, bilaterally winged, the wider wings having acute or denticulate teeth (0.5 mm. long), never exceeding the anthers. Short filaments 1.5-2 mm. long with connate, lanceolate or denticulate appendages, 1 or more the length of filaments. Fruits ovate-orbicular, usually truncate, 3.5-4.5(-5) > 2-3(-3.8) mm., papillose, or smooth, and always with sparse, short, slender tuberculate setae, at least on margins. Styles slender, (0.7—) 1-2 mm. long. Seeds winged. F/. Apr.—July. DISTRIBUTION AND HABITAT: an endemic scattered in Inner Anatolia on disturbed lands, cultivated fields, saline steppe, and gypsum outcrops; alt. 500-1620 m The specimens known to date, and a discussion of this species may be found in Jour. Arnold Arb. 45(1) : 61-63. 1964. II. Section Psttonema (Meyer) Hooker KEY TO ALL SPECIES OF SECTION PsILONEMA A. Fruits orbicular, rotund, oblate or ovate, emarginate or obtuse, with indu- mentum and smooth margins; sepals persistent; leaf margins entire. 1965] DUDLEY, STUDIES IN ALYSSUM 197 jee) . Fruit indumentum monomorphic, of stellate hairs only; petals glabrous or rarely with sparse adpressed indumentum; nectaries erect or subulate (up to 2 mm. long), or if + globose, then fruits obtuse. C. Fruit indumentum of non-overlapping, short-rayed stellate hairs; styles glabrous; petals retuse; nectaries subulate and erect; fruits TNE ohn nin Fe acest ng ee Sa ae 8. A. alyssoides. . Fruit indumentum of long-rayed and overlapping stellate hairs; styles with basal indumentum; petals deeply bifid; nectaries short, + globose; fruits obtuse. .................. 9. A. damascenum . Fruit indumentum strigose and dimorphic, of adpressed stellate hairs, and erect, furcate + tuberculate hairs; petals with dense strigose indu- mentum; nectaries always reduced and een less than 0.4 mm. lon D. Seeds winged; styles 0.5—1 mm. long, not basally dilated, glabrous, or with sparse basal indumentum of adpressed stellate hairs; leaves linear-oblanceolate, or elliptic-oblong, increasing in size upwards. 10. A. granatense. D. Seeds wingless; styles (1—)1.5-2 mm. long, strongly dilated at the base with dense dimorphic indumentum; leaves obovate-spathulate, decreasing in size upwards. ................ 11. A. dasycarpum Fruits broadly obovate, truncate, glabrous, with carrie margins; sepals deciduous; leaves minutely denticulate at apices. .... 12. A. homalocarpum. QO oo as ag Alyssum alyssoides (L.)L. Amoen. Acad. 4: 487. 1759 (!); Syst. Nat. ed. 10. 2: 1130. 1759. — Hal. Consp. Fl. Graec. 1: 99. 1909. — Schinz & Thell. Bull. Herb. Boiss. II. 7: 407. 1907.— Beck, Fi. Bosne, Herceg. 2(7) : 302. 1916.— Rydberg, Fl. Rocky Mt. & Adj. Plains, 347. 1923. — Hayek, Prodr. Fl. Penin. Balc. 1: 439. 1925. — Nyar. Magyar Bot. Lap. 24: tab. 1, fig. 24. 1925; Bull. Bot. Grad. Cluj 7: tab. 8, fig. 106. 1927. — Javorka & Csapody, Ic. Fl. Hung. 7: 211. fig. 1590. 1930.— Degen, Fl. Veleb. 2: 187. 1937.— Mansfeld, Repert. Sp. Nov. 46: 114. 1939.— Rech. Fl. Aegaea 225. 1943. — Abrams. Ill. Fl. Pacific States 2: 319. fig. 2149. 1944. — Hylander, Uppsala Univ. Arsskr. 7: 182, 1945.— Fernald, Gray’s Manual of Botany, ed. 8. 699. fig. 1805. 1950. — Gleason, New Britton & Brown Ill. Fl. Ne. U. S. & Adj. Canada 2: 220. 1952.— Lid, Norsk. Fl. 325. 1952. — Parsa, Fl. Iran 1: 744. fig. 617. 1952. — Pawt. Fl. Tatrorum 325. 1956. — Lagerberg, Vilda Vaxter Nord. 2: 762. fig. 394b. 1957.— Markgraf in Hegi, Ill. Fl. Mitt.-Europa, ed. 2. 4(1): 288. tab. 125, figs. 23, 44; tab. 127, figs. 3a, b; fig. 170a. 1960. — Heywood, Repert. Sp. Nov. 64(1): 53. 1961. — Dudley, Jour. Arnold Arb. 45(1): 63-65. 1964; in Rech. Fl. Lowland Iraq 306. 1964. — Ball & Dudley in Flora Europaea 1: 299. 1964 Annual or biennial, with few or many, erect, decumbent or ascending, rarely prostrate stems, up to 50 cm. long. Jndumentum grayish-green (rarely silvery) of adpressed stellate hairs having few and + short rays (density variable). Leaves obovate or linear-oblanceolate, up to 4 cm. long. Racemes elongated, rarely reduced, up to 15 cm. long, and if 198 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 branched, branches never exceeding the terminal axis. Pedicels 2-6 mm. long, spreading or horizontal. Sepals (1.5—)2-3 mm. long, persistent. Petals obovate, usually glabrous, emarginate, 2-3(-4) mm. long, often scarcely exceeding the sepals and persistent with them. Filaments 1-1.5 mm. long, always edentate and unappendaged. Nectaries slender and erect, 0.5-0.8 mm. long. Fruits orbicular, emarginate or truncate, (2—)3- 4(-5) mm. long and wide; valves equally inflated at centers, and with flattened margins, covered with an indumentum of minute (0.2—-0.3 mm. in diameter) adpressed stellate hairs (density variable). Styles 0.3- .6(-1) mm. long, + slender, usually glabrous. Seeds narrowly winged. 2n = 32. Fl. Mar—Aug. Two varieties are recognized: A. Leaves oblanceolate or linear, (3—)10-40 X 1.5-3 mm., with sparse gray- ish-green indumentum; plants always more than 5 cm. tall, usually 15-35 cm.; racemes elongated, never umbellate, (2—-)5-15 cm. long, 10-fruited, ee ee ed tna ea ee ets Se Gwe eee wo a. Var. alyssoides. Leaves obovate, 2-3.5 X 0.5-1 mm., with dense silvery indumentum; plants very reduced, 1-3 cm. tall; racemes very wean a umbellate, OS=) Ch. WO, FFT, no pce ek Ge enw keees b. Var. depressum. ad a. Var. alyssoides. Clypeola alyssoides L. Sp. Pl. 2: 652. 1753 (!). Type, Europe in Austria et Gallia; “2. Clypeola siliculis bilocularibus tetraspermis” (gm, hort. Clifford; Clypeola No. 2, sub “Alysson incanum luteum serpilli folia.’’). Clypeola campestris L. op. cit. 652, 1231. Type protologue refers to Sauvages, Methodus Foliorum Monspeliensis. . . . 71. 1751, reading: “No. 405 Cl annua siliculis bilocularibus dispermis calyce persistente. .. .”, and to eS Bauhin, Pinax, 107. 1623, reading: “Alysson dictum campestre minus. Alyssum calycinum L. Sp. Pl. ed. 2. 2: 908. 1763 (!).— Jacquin, Fl. Austr. 4: tab. 338. 1776. — Reich. Ic. Fl. Germ. & Helv. 2: tab. 18, fig. 4269. 1837- 1838. — Bertoloni, Fl. Ital. 6: 483. 1844. — Boiss. Fl. Or. 1: 285. 1867. — Cusin & Ansb. Herb. Fl. Fr. 2: tab. 312. 1869.— Ruprecht, Fl. Caucasi. 105. 1869. — Ettings. & Pokorny, a Pl. Aust. 9: tab. 871. 1873.— Mora, Fl. Fan. Esp. & Port. 6: 560. 1 — Willk. & Lange, Prodr. Fl. Hisp. 3: 833. 1880. — Brandza, Fl. Rom. 1833. — Schlecht., Lang. & Schenk, Fl. Deutschl. ed. 5. 14: 195. tab. 1387. 1883. — Colmneiro, Pl. Penin. His- pano-Lusit. 155. 1885. — Simonkai, Enum. Fl. Transsil. 91. 1886. — Beck, Fl. Nieder-Osterr. 469. 1892. — Rouy & Fouc. Fl. Fr. 2: 185. 1895. (incl. vars.). — Robinson in Gray, Synopt. Fl. N. Am. 1(1): 115. 1895.— Gre- cescu, Consp. Fl. Roman. 68. 1898 (excl. var. [8] depressum). — Pauletti in Fiori & Pauletti, Fl. Anal. Ital. 1: 457. fig. 1447. 1898-1899. — Busch in Kuznetsov, Busch & Fomin, Fl. Cauc. Crit. 3(4): 601. 1909 (incl. forma) ; in Fl. U. R. S. S. 8: 358. 1939. — Bornm. Bot. Centralb. Beih. 38: 479. 1921. Thiébaut, Fl. Lib.-Syr. 1: 71. 1936.— Palhinha, Fl. Port. ed. 2. 307. 1939. hi A A Fl. Kavk. ed. 2. 4: 220. tab. 25, fig. 6. 1950. — Jessen, Vilde Pl. Nord. 2: 598. fig. 410, fig. 411b. 1950. — Karjagin, Fl. Azerbaid. 4: 277. 1953. Type, Europe, Austria, Gallia, Germania (pm, hort. Clifford; Clypeola No. 2). Alyssum campestre L. op. cit. 909., pro parte (!). —°*° 1965 | DUDLEY, STUDIES IN ALYSSUM 199 Adyseton acy eee tides Fl. Carn. ed. 2. 2: 13. 1772 (!). — Bubani, Fl. Pyr a 3: 209. pac eeii eaten 5 beak Bull. Acad. Sci. St. Pétersb. 7: 132. 1840 (!); Mém. Acad. Sci. St. Pétersb. VI. 6: 15. tab. 2, upper right. 1840; in Ledebour, Fl. Ross. 1: 137. 1842. — Schur, Enum. Pl. Transsil. 630. 1866 (excl. var. depressum).— Fedtschenko, Fl. Ross. Austro-Orient. 5: 440. pl. 388, fig. A. 1931. — Popov, FI. Ukraine 5: 344. 1953. Alyssum jaalerals Jordan, Diagn. Nouv. 198. 1864. Type, Switzerland, hab. in ruderatis et agris es circa Genéve, Jordan (Pp, non vidi). A. vagum Jordan, ibid. T France, hab. in ruderatis et sabulosis agris lugdunensis, Villeurbanne (Rhone), Jordan (P, non vidi). A, sabulosum Jordan, op. cit. 199. Type, France, hab. in sabulosis Beugesi, Thoirette ge Mavier ( non vidi). A. arvaticum Jordan, op. cit. 200. Type, France, hab. in ruderatis et arvis, Delphinatts superioris; eh Grave (Hautes-Alps), Jordan (Pp, non vidi). A, erraticum Jordan, loc. cit. Type, France, hab. in ruderatis et arvis pyren- aeorum Gédre (Hautes- adden ge Jordan (Pp, non vidi). A, sublineare Jordan, op. cit. 201. Type, France, hab. in ruderatis et sabulosis montium Occitanieae, oe Cabardés (Aude), Teka. (P, non vidi DISTRIBUTION AND HABITAT: a widespread weed species of western, central and southern Europe, Russia, North Africa and Afghanistan in ruderal and disturbed habitats and mountain meadows; sea level—2000 (—2800) m. Probably introduced and naturalized in northern Europe, in- cluding the British Isles and Scandinavia. Naturalized in the United States and Canada, and in the Argentine of South America. Recorded from Turkey England. Dirleton Common, 24 June 1835, Macnab (£); Surrey, Wands- worth, July 1835, Hunter (£). France. Montpellier, Herault, 27 Apr. 1894, Galavielle (£, G, W); Ig Arnott (£, K); Paris, Forbes s.n. (£); Zabern, May 1896, Krebs (BM, E, G, K, w); Loches, Apr. 1841, Trevelyan (£, K); Paris, Bois de Boulogne, Cosson i: cH); La Maures, Hyeres, 30 Apr. 1906, Raine (GH). rein Mt. Pietro, Solenol 355 (BM). Spain. Pyrenees, 2 aay 1896, Guillot anes m., Reverchon 1276 (BM, E, 6); Barcelona, Gonzala 5453 (BM); ibid.. 450 m., ‘Gonzala 5434 (BM); ibid., 3 May 1918, Sennen (BM, E); ibid., Can- talejo, Sennen 2964 (pm); Cerdagne, 1380 m., 12 June 1926, Sennen (BM); S Angoustrone Grande Rigole, Sennen 6040 (pm); Sierra de Barza, 1890, all (BM, E, GH, K); Distr. Logrono Sorio, Se Sierra de la eis 1700 m., Dresser 610 (£). Sweden. Skane i Alnarp, June 1841, Palmer (E); Uppsala, Anderssen sn. (e); ibid., July 1866, Ahlberg (cH, on Askersund, 11 June 1887, Wyring (BM, GH, K); Prov. Gotland, Paroecia Alskog, Asplund 877 (BM, GH, w); Stenkyrka, 19 June 1867, Oldberg (cH). Denmark. Isle of Moen, oat a Jéker (GH, w); ibid., May 1846, Tiitein (GH); ibid., Schonid (GH). Jura, nr. Regensburg, 410 m., Rubner 781a (BM, E, K, W); cs | ea an 1829, Rosh (£, W); Thusran, Grierson 30 (£); Baden nr. Wiesloch, 200 m., Zimmerman 8 (BM, E, K, W); Berlin, 14 May 1896, Rehder GH); Kies, 800 m., Zick 781a, b (BM, G, GH, K, Ww); Thurin a, nr tadt, 19 Apr. 1902, Reineck (GH, W). con Zurich, 26 June 1880, Rehder (cH); May 1836, Naegeli (E, w); un, Apr. 1838, Brown (£, K); Vallaris Rhone, May 1858, Balfour (£); Aug. en Dickson (E); Aigle, 4-9 June 1885, 200 JOURNAL OF THE ARNOLD ARBORETUM [ VoL. 46 Hamilton (£); Jura, nr. Neuville, 9 Aug. 1834, Shuttleworth (£); Zermatt- Zmutt, May 1961, Dudley (a); Chur, 700 m., Meisser 340 (BM, G, GH, K, W); _ Naples, 1845, Alexander (£); nr. Bormio, 18 Aug. 1870, Ball (pm, E, K); Flor- ence, Babington (rE); St. Cauzian, Crawford 26 (£); Calabria, Sila nr. San Giovanni in Fiore, 1000 m., 1933, Bornmiiller 80 (A). Sicily, Mt. Nebrodense, Todaro s.n. (G, GH, Ww); 1700-1934 m., 21 July 1874, Stroblysi (BM). Po Kielce, 18 May—30 May 1897, Bodzentyn (£, w); Kiovise nr. Karawajewi, La- zarenko 65 (BM, E, K, W); Prov. Kioviensis, distr. Smila, pr. Jablunilvka, Kleo- pow s.n. (BM, E, K, W). Czechoslovakia. Kaaden, nr. Taltsch, 700 m., Stelzhamer 346 (BM, E, G, GH, K, W). Austria. St. Veit, 6 June 1898, Krebs (£, G, K, W); nr. Graz, 360 m., Kritsch 749 (pM, E, G, K, w); nr. Judenburg, 710 m., Pilhatsch 748 (BM, E, G, K, W); oe Hoare $n. (BM, E, G, GH, K, w); Laibach, Fleischmann s.n. (BM C, Ww). . Mt. Ro kahicty nr. Bekras, 4 June 1922, Degen (w); Sorkut-Toak Bint, 24 June 1922, Degen (w); Chem- oe mgt Ball (BM, £, GH, K). Romania. Transsilvania, distr. Brasov, nr. Har- , 500 m., Borza 651c (pM, E, G, XE, oe distr., Turda Aries, Cheia Turzii, nr. ge | Drin, Struga-Debar, Rechinger 15975 (w); Serbia orientalis, Bela Palanka-Pirot, 600-800 m., Rechinger 15865 (w); ibid. Nischka Banja, nr. Nisch, Rechinger 16072 (w); ibid., Nischa Banja-Bela Palank, 600-800 m., Rechinger 15839 (w); Serbia, Vanjano, May 1898, Adamovic (£, w); Sarlark nr. Pirot, 15 May 1897, Adamovic (£, K, w). Albania. Bertiscus, nr. Pec (Ipek), 500-700 m., Rech- inger & Scheffer 104 (w); Katimi nr. Bukovik, distr. Hati, Baldacci 298 (BM); Kolasia, Baldacci 9 (pm); Kia nr. Skutari, Dérfler 153 (w). Greece. Terkovic- Sliovo, "Rechinger 8 (w); Epirus, Tonschiefer. nr. Arachthos river, Rechinger 23199 (w); Macedonia orientalis, distr. Drama, Boz dagh, nr. Juricik, 300 m., Rechinger 6351b (w); ibid., Lekhani-Kechrokampos, valley of river Mesta (Nestos), 700-900 m., Rechinger 15632 (w); Thrace, Orestias-Visi, Rechinger 22031 (w); ibid., Mt. Rhodope, nr. Jasmos (Jasi-Koi), Rechinger 9579 (w); Mt. Olympus nr. Hagios, Diontsios, 800 m., 21 Mar. 1940, Charworth-Masters 9579 (BM); Xanthi, Lesins 6 (A); Mt. M alevo, Laconiae, Orphanides 2638 (ewe: Parhes, Hagios-Trios, 400 m., Samuelsson 141 (w). Crete. Lassithi, Mt. Lazaro, Gandoger 77 (Bm). ual. Crimea, Yalta, above Nikita, 350 m., Davis 33328 (BM, E, K); Crimea, 1820, Steven (c-pe); Kiev, Zinger 555 (E, w); Konigsburg (East Prussia), Baenitz 5.n. (BM, E, G, GH, K, w); Caucasus, Prov. Terek, —— Balta, Brotherus 86 (sm); ibid., 17 July ne Busch (BM, E, G, GH, w); Prov. Chewsuria, 1 Aug. 1903, Busch (BM, E, G, GH, K, Ww); Daghestan, 6. Becker (BM, K); Azerbaidjan, Ismaily, paae St Ismaily, 24 May 1936, Grossheim (Bm). Bulgaria. Varna, 1846, Noé (Kk); Haskovo- eee (fide Fenzl in Tchihatcheff, Asie Min. Bot. 1(3): 313. 1860). A7: Prov. Trabzon, Boztepe, 21 June 1917, Schischkin (fide Schischk. Ber. Staats-Univ. Tomsk 80: 465. 1929); Prov. Trabzon Giimiisane, nr. Vischera, Kalanema Dere, 00 m., June 1908, Blumencron (fide Handel-Mazzetti, Ann. Naturh. Mus. Wien 23: 156. 1909). B6: Prov. Maras, Elbistan, Asdurian 85 (fide Béguinot & Diartz. Contrib. Fl. Arm. 47. 1912). B10: Prov. Doganbayazit, Dutach-Bur- 1965] DUDLEY, STUDIES IN ALYSSUM 201 nubulak, 26 May 1916, Schischkin (fide Schischk. loc. cit.). C6: Prov. Gaziantep, Killis, Post 335 (fide Post, Bull. Herb. Boiss, II. 3: 163. 1895). Tokat-Erzurum, Aucher sn, (fide Fenzl, loc. cit.). Afghanistan. Prov. Kabul, Kabul-Paghman, 2800 m., Hedge & Wendelbo 3146 (sc, ©). Canada. Quebec. Wolfe County, nr. Lake Aylmer, Wells 37470 ( GH); Missis- quoi County, Philipsburg, 10-11 Aug. 1923, Knowlton (cH); ibid., Marie- Victorin & Rolland-Germain 34152, 43296 (GH). Ontario. Waterloo County, German Mills, Cressman’s woods, Montgomery 198 (cH); Learmington, Macoun 33775 (GH). United States. Maine. Orono, 16 June 1890, Fernald (Nesc); Bar Harbor, 11 June 1899, Rand (NeEBC). Vermont. Chittenden County, Colchester, S. of Porter’s Swamp, Charette & Smith 2430 (NeEBC); ibid., Charette 2439 (NEBC). Massachusetts. Suffolk County, Franklin Park, Boston, 11 June 1891, Manning (GH, NEBC); Jamaica Plain, June 1875, Faxon (GH); Medford-Malden, May 1880, Davenport (cH); ibid., 6 May 1882, Manning (NEBC); Belmont, 6 June 1898, Hoffmann (nEeBC); Medford, 24 June 1880, Davenport (neBc); Somer- ville, 10 May 1878, Perkins (NeBc); Norfolk County, Milton, 6 June 1897, Kennedy (cx); ibid., 17 July 1915, Kidder (NEBc); Barnstable County, Bourne, nr. Monument Beach, 20 June 1911, Knowlton (GH, NEBC); Falmouth, July 1882, Farlow (NEBC); Bristol County, Dartmouth, Sturtevant s.n. (NEBC); ibid., 13 June 1904, Hervey (NEBC); Amherst, 1859, Gray (GH). Rhode Island. Lighthouse Point, East Haven, Eames 216 (GH); ibid., 17 June 1902, Harger (NEBC); Middlebury, Harger 4290 (NEBC); Hartford County, Southington, 31 May 1897, Bissell (NEBC). New York. Washington County, Vaughn, N. of Hudson Falls, 10 June 1915, Burnham (GH); Tuckahoe, 12 May 1894, Pollard (GH); Mt. Beacon, Hudson River, opposite Newburgh, June 1906, Kochler (GH); nr. Plattsburgh, Hunnewell 4667 (GH); Onondaga County, East Green Lake, Jamesville, Wiegand 15516 (GH); Columbia County, Becraft Mt., Green- fort, McVaugh 317 (cH). Pennsylvania. Mt. Airy reservoir, 12 May 1871, Parker (GH); Berks County, Bernville, Stoudt & Hermann 2768 (GH); College Hill, Easton, June 1875, Porter (GH); ibid., 8 June 1892, Porter (GH). West Virginia. Greenbrier County, White Sulphur Springs, Hunnewell 2593 (cH); Berkeley County, nr. Inwood, Hunnewell 19182 (GH). Virginia. Clarke County, nr. Boyce, Allard 106 (cH); ibid., Hunnewell 14991 (cH); Shenandoah County, Pugh’s Run, Artz 830 (cH); ibid., Strasburg, Baldwin 5064 (GH); nr. Cedar- ville, Pease 26574 (GH). Michigan. Near Lansing, 8 June 1887, L. H. Bailey (GH); Berrian Springs, Pease 17777 (cH). Indiana. Tolleston, Umbach 1787 (GH); Fulton County, 1 mile Ne. of Leiter's Ford, Deam 56019 (cH). Wiscon- sin. County highway “D,” S. of Madison, 18 June 1945, Greene (GH). Illinois. Ravenswood, nr. Chicago, 7 June 1883, Arthur (cH); McHenry County, Algon- quin, Benke 5741 (cH); Kankakee County, E. of St. Anne, Jones 11430 (GH). Montana. Many Glaciers, Glacier Park, Pease 22323 (cH); Riverside Park, 975 m., Kirkwood 1129 (cH). Idaho. College campus, Moscow, Henderson 2759 (GH); ca. 2 miles S. of Grangeville on Whitebird Road, Jones 73 (GH); 3 miles E. of Joseph on Joseph-Whitebird Road, Jones 163 (cH); Teton County, Victor, 1829 m., Payson & Payson 2160 (GH). Wyoming. Sheridan County, Red Grade, E. slope of Big Horn Mt., 1981 m. Rollins 57177 (cH). Utah. Newton-Heyde Park, Jones 288 (cH); Salt Lake City, Rollins 3095 (GH); Cache County, Sage E. of U. S. A. C. stadium, Logan, 1433 m., Maguire 34581 (cH). California. Oakland, Brewer 2577 (cH); Siskiyou County, Yreka, Smith 202 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 90, 655 (cu); ibid., Sisson, Heller 8054 (cu); ibid., Parker ranch, Plowman’s valley, 12 June 1948, Parker (GH). Oregon. Selkirk, nr. Nelson, Shaw 663 (GH); Des Chutes river, 5 miles below bend, Peck 1713 (GH). Argentina. Partido de Saavedra, Sierra de la Ermita, Cabrera 5459 (GH). The nomenclatural confusion between Alyssum alyssoides and A. minus (A. campestre sensu multo auct.) has been discussed in an earlier paper (Jour. Arnold Arb. 45: 63-65. 1964). Although these two species are assigned to different sections, to section PsILONEMA and section ALYSSUM respectively, they are frequently confused. The sepals of A. alyssoides are always persistent, its filaments are very slender, edentate and un- appendaged, its nectaries are peg-like and erect, and its styles are usually glabrous. In addition, the symmetrically inflated fruits of A. alyssoides are generally smaller, and the easily displaced indumentum on the fruits is composed of shorter-rayed stellate hairs. Although specimens of A. alyssoides have been recorded from Turkey, I have not seen any Turkish material. Certainly its presence in Anatolia needs confirmation. It seems safe, however, to assume that the records from the Armenian Highlands in eastern Turkey are correct. It is well known from the Caucasus, and its presence in Armenia would be an expected pattern of distribution. Throughout the Levant, A. alyssoides is very rare (cf. Bornmiiller, Bot. Centralbl. Beih. 38: 479. 1921). For the most part, the Caucasian, Turkish and Afghanistan specimens of A. alyssoides are from higher altitudes (e.g. Hedge & Wendelbo 3146, Afghanistan at 2800 m.) than are those normally found in Europe. From the paucity of records and specimens from the Levant it is assumed that this species has been unable to colonize and spread in the ruderal and disturbed types of environments, with which it is normally associated throughout much of Europe. Rather, it is apparently confined to isolate pockets in the mountains. Conversely, it is interesting to note that it has adapted well to the ruderal habitats of North America, and accordingly seems to be commoner than in Turkey. As a common European species A. alyssoides has long been subjected to a very critical examination by numerous workers, many of whom (i.e. Jordan, Sennen, Nyarady, Prodan, etc.) have contributed to the literature approximately thirty taxonomic segregates, mostly of varietal or forma rank, and almost twice as many recombinations. As it would not be in the interests of brevity or clarity to cite all of those minor synonyms, I have referred only to those species of Jordan which subsequently have been recombined many times as subspecies, varieties, formae, and even sub- formae. This species is very polymorphic with respect to plant height, stem length, leaf size, leaf, fruit and sepal indumentum density, and raceme length. The many segregates, excluding var. depressum which follows, have been based chiefly on single character deviations, which are very flexible and unstable according to the varying environmental pres- sures. Many of the characters rarely occur on a population basis, and examination of the type specimens has revealed that, more often than not, ey ee ren ee tea 1965 | DUDLEY, STUDIES IN ALYSSUM 203 the characters are not constant on individual plants. When the morpho- logical variation of A. alyssoides is considered throughout the whole range of distribution, characters such as density of fruit indumentum do not appear to have the stability essential for taxonomic recognition. The one exception is var. depressum which forms small and scattered, but pheno- typically stable populations in Romania, Hungary, Bulgaria, Greece, and Crete. b. Var. depressum (Schur) Dudley, comb. nov. erg Pinstoyes var. [c.] depressum Schur, Enum. Pl. Transsil. 62. 1866 eth otype, Romania, Hermannstadt, Schur s.n. (w). Piles calc um var. pumilum Hal. Denkschr. Akad. Wien Math. Nat 61: 496. 1894 (!). Holotype, Greece, Arcadia, in lapidosis calcareis ieee pl Mt. Chelmos (Aroania vet.) gregarie., 1500 m. (in der Tannen- region by Chelmos oberhalb Sudena, 1200-1500 n,). 15 Apr. 1893, Halacs A, inet var. [8] depressum (Schur) Grecescu, Consp. Fl. Rom. 69. 1898 (!). A. calycinum var. minus Velen. Sitz-ber. Bohm. Ges. Wissen. Prag 27: 3. 1902 (! - Holotype, Bulgaria, in Mt. Tikiski, Balkan, Urumov (prc, non vidi) ; isotype (w). A, pied ha Fil. & Jav tap gine Bot. Lap. 9: 146. 1910 (!); Rep. Nat. Mus. Hung. 107. tab. 1, fig. 2 1910. Holotype, Hungary, in virgultis ca- cuminis montis Nagy- Széndshegy ad pag. Pilissentwany, Comit. Pest, ca. 500 m., 9 June 1909, Filarszky & Kiimmerle (sp, non vidi); isotypes (BM, E, G, K, W). A, cP dtc var. conglobulatum (Fil. & Jav.) Jav. is oe 441. 1924 (!); Jav. & Csapody, Ic. Fl. Hung. 7: 211. fig. 1590a. A. being var. pumilum (Hal.) Hayek, Prodr. 1 ce Balcan, 1: 439. S Crh satin & Rech. Ost. Bot. Zeit. 84: 139. 1935.—Rech. FI. oll #25. A. Sesser a minus (Velen.) Hayek, loc. cit. (!). A. calycinum f{. minus (Velen.) Stoj. & Steff. Fl. ‘Bulg. ed. 2527. 1948 (!). DIsTRIBUTION AND HABITAT: rare on calcareous substrates of moun- tains in Hungary, Romania, Greece and Crete. Hungary. Comit. Pest, Mt. Szénashegy nr. ee (locus classicus), Crete. Pezzuta, 400-500 m., Filarszky & Jdvorka 46 (BM, GH, K, W, Todaro s.n. ‘(w); Mt. Psiloriti, nr. Nidha, Dirfor 774a (Ww). 9. Alyssum damascenum Boiss. & Gaill. in Boiss. Diagn. 3(6): 18 1859 (!).—Boiss. Fl. Or. 1: 285. 1867.— Bornm. Verh. Zool.- Bot. Ges. Wien 48: 553. 1898.— Boul. FI. Lib. & Syr. 32. pl. = 1932. — Thiébaut, Fl. Lib.-Syr. 1: 71. 1936.— Zohary, Pal. Jour. Bot. Jer. ser. ra ie 129 & 161. 1941. Holotype, Syria, in cultis inter Merre amascum (Jardin e Ganchedulu), 18 Mar. 1847, Gaillardot 317 — isotype (A). 204 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 Annual, low growing, resembling Alyssum contemptum in habit; spar- ingly branched from the base, 5-10 cm. in height, Leaves oblanceolate- spathulate, acute, decreasing in size upward, uppermost involucrate; indumentum on the lower surfaces denser, and of smaller stellate hairs than that on upper surfaces. Racemes simple, or sparsely branched from the base, 1-4 cm. long. Pedicels erect or ascending, + basally dilated, 3-5 mm. long. Sepals persistent, ca. 2 mm. long, with narrow scarious wings; indumentum sparse, but with apical tufts of furcate, erect hairs. Petals narrowly spathulate, 2.5-3 0.5—-0.8 mm., deeply bifid. Filaments 2-3 mm. long, edentate and wingless. Fruits ovate or rotund, 3.5—4.5 (-5) < 3-4.5 mm., obtuse; valves + equally inflated, and with + dense indumentum of relatively coarse stellate hairs. Styles 0.6-1 mm., with basal indumentum. Seeds narrowly winged. F/. Mar.—Apr. DISTRIBUTION AND HABITAT: a Saharo-Sindian species of dry_hill- sides and cultivated or fallow fields in Syria and Palestine; alt. 200- 1000(—1900) m. Syria. Aleppo-Aintab (Gaziantep), 610 m., 6 May 1865, Haussknecht (BM, c); Mt. Carmel, Apr. 1928, Druce (oxF); Damascus, Salatie, Péronin 583 p.p. (c); ibid., Kessoue, Péronin 1879 (c); ibid., Davis 5633 (£, K); ibid. Mt. Gebel Khaisoun, Gaillardot 856 (c). Palestine. Jericho, Ain-i-Sultan-Wadi Kilt, 200 m., 1897, Bornmiiller 71 (c, w). Lebanon. Dschebel Sannin, 1700- 1900 m., 10 June 1904, Kneucker (cH). Alyssum damascenum is sometimes confused with A. minus, a weedy species in section ALyssumM common throughout most of Europe and the Near East, because of a resemblance in fruit shape and indumentum. The filaments of A. damascenum, however, are always wingless, edentate, and unappendaged, while those of A. minus are widely winged, appendaged, and usually dentate. A. contemptum from Palestine, another species of section ALyssum is sometimes confused with 4. damascenum due to the annual, low growing habit. A. contemptum has widely winged toothed and appendaged filaments, small globose nectaries, entire or merely emarginate petals, and elliptic fruits whose valves are very asymmetri- cally inflated, similar to those of A. szowitsianum. The fruits of A. damascenum are orbicular with more or less equally inflated valves, its nectaries are erect and peg-like, and its petals are deeply bifid. Zohary and Fahn (Pal. Jour. Bot. Jer. ser. 2(2/3): 130. 1941) and Zohary (op. cit. 161) assign A. damascenum to the group of annual species in section ALyssum (including A. marginatum and A. szowitsianum) characterized by a unique pedicel anatomy associated with the specialized hygrochastic method of seed dispersal. However, in addition to possessing the diagnostic features of section Psttonema, A. damascenum has 4 pedicel anatomy, which (as originally noticed by Zohary & Fahn) is different from that of the other species mentioned. These facts suggest that, although the annual species of section PsttoNEMA and section "~ a i cel 1965 | DUDLEY, STUDIES IN ALYSSUM 205 ALyssuM are closely allied, the phenomenon of hygrochastic dispersal has developed convergently within the two sections. 10. Alyssum granatense Boiss. & Reut. Pug. Pl. Nov. Afr. Bor. & Hisp. Aust. 9, 1852 (!).— Mora, Fl. Fan. Esp. & Portug. 6: 561. 1873. — Willk. & Lange, Prodr. Fl. Hisp. 3: 833. 1880.— Cosson, Ill. FI. Atl. 1: 61. tab. 42. 1884; Comp. Fl. Atl. 2: 236. 1887. — Batt. in Batt. & Trab., Fl. Algérie 1: 47. 1888.— Cadevall & Sallent, FI. Catal. 1: 142. 1915.— Palhinda, Fl. Portug. ed. 2. 307. 1939.— Heywood, Repert. Sp. Nov. 64(1): 53. 1961.— Quezel & Santa, Nouv. FI. Algér. 1: 410. 1962.— Ball & Dudley in Flora Europaea 1: 299. 1964. Syntypes, Spain, in arenosis et cultis regionis alpinae montium Granatensium, Sierra de la Nieve supra Yunquera, Boissier & Reuter (BM, G, W); in Sierra Nevada circa Benalcaza, Boissier (BM, G, w). Lectotype, Sierra de la Nieva supra Yunquera, Boissier & Reuter (G); isolectotypes (BM, W). A. willkommii de Roem. ex Willk. Linnaea 25: 8, 1852 (!). Holotype, Spain, in collibus arenosis siccis prope oppidum Ayamonte, Jan. 1846, de Roemer (uz, destroyed); isotype (BM). A. granatense var. sepalinum Pomel, Nouv. Mat. Fl. Atl. 231. 1874 (!). Holotype, North Africa, a Garrouban, Téniet-el-Haad, Tala-Yezid, Pomel (AL, non vidi) ; isotype (w). A. algeriense Pomel, op. cit., 232 (!). Holotype, apse Africa, de Garrouban a Téniet-el-Haad, Pomel (aL, non vidi); isotype . algeriense var. montanum Pomel, loc. cit. (!). TasGee North Africa, Diehed Endatte, prés de Téniet-el-Haad, Pomel (AL, non vidi); isotype (w). A. hispidum Loscos & Pardo ex Willk. Ill. Fl. Hisp. & Balear. 1: 85. 1882 (!); Suppl. Prodr. Fl. Hisp. 304. 1893.—Syntypes, Spain, en Castelseras particul. en el Cerillo de Calvario, en la huerta de Sg: — Calaceite y nesesar. pasa 4 Catalufa par Caseras, s (WwW); Aranda, Calavia (w). Lectotype, Aragon, airy 28 a 1846, pra Piiex isolectotype (a). A. hispidum var. granatense (Boiss. & Reut.) Willk. Ill. Fl. Hisp. & Balear. 1: 85, 86. 1882 a Suppl. Prodr. Fl. Hisp. 304. 1893 A. marisii Cout. Bol. Soc. Brot. 25: 189. 1910. Holotype, Portugal, hab. in Beira meridional (Castello Branco) Malpica Baixo Alemtejo (Beja), Maris s.n. (cot, non vi A. hieronymii Sennen, Bol. Soc. Arag. 15: 259. 1919 (!). Holotype, Spain, Castillo, Madrid, 15 Apr. 1915, Jerénimo 2411 (Bc, non vidi); isotype BM A. mea var. weilleri Emb. & Maire, Bull. Soc. Hist. Nat. Afr. Nord. 23: 164. 1932. Syntypes, North Africa, hab. in rupestribus ra editis Anti-Atlantis ad Agadir-n-Tigfert, 1700-1800 m., 1931, Weiller (AL, non vidi); in Monte Fidoust, 2000-2200 m., 1931, W eilier (AL, non vidi). Annual, with numerous erect or decumbent stems, up to 20 cm. long. Leaves linear-oblanceolate, lanceolate, or elliptic and oblong, 3—20(—2 5) —3 mm., increasing in size upward, indumentum grayish-green, of adpressed stellate hairs. Fruiting racemes generally simple, elongated, 2-8 cm. long, densely fruited. Pedicels 2-4 mm. long, ascending to erect, 206 JOURNAL OF THE ARNOLD ARBORETUM [VoL. 46 and adpressed to the main axis, indumentum dense and + strigose. Sepals persistent, 2—3.5 mm. long, with wide scarious wings, and often with apical tufts of furcate hairs. Petals clavate, gradually attenuate into claws (3)—4(-6) mm. long, emarginate, with dense strigose indumentum. Filaments very slender, edentate and unappendaged, 2—2.5 mm. long. Fruits orbicular, 3-5 mm, long and wide, emarginate; valves equally inflated at centers and with wide flattened margins, indumentum dimor- phic of tuberculate, simple and furcate hairs, intermixed with adpressed few-rayed stellate hairs. Styles 0.5-1 mm. long, glabrous or with sparse basal indumentum. Seeds conspicuously winged. F/. Feb.—Apr. DISTRIBUTION AND HABITAT: cultivated and fallow lands, and dry mountain screes in eastern and southern Spain, Portugal, and North Africa! alt. 600-2200 m Portugal. Sampaio 2471 (pm). Spain. Aragonia, Boissier s.n. (Gc, w); Murcie, Sierra de Espufia, 1200-1400 m., Jerénimo 7101, 6715 (pm); Almeria, N. ae of Sierra de Maria above Maria, Ellman & Sandwith (BM, GH, K); Pozuelo nr. Madrid, Bucknell (BM, £); ibid., New Castile, 11 Apr. 1907, “White (£); Prov. Valencia, Sierra de Espadan, 1800 m., Reverchon 21 (BM, E, G, K, W); Le Pozo, 1500 m., Reverchon 706 (w); Sierra de Alcaraz, 600-1000 m 1890, Porta & Rigo (Bu, C, a. North Africa. Maroc, Djebel Lalla Aziza, Ibrahim, 1883, Cosson (BM, E K, W); Djebel Kerher, 850 m., Sennen & Mauricio 9237 (pm); Batna, 4 hee 1867, Dukerley (BM, GH, K): Nogen Atlas, Ain Kahta, 1850 m., Jahandiez 322 (a, E, K, w); Algiers Sidi- bel. Abbés (nr. Oran) Warion 114 (BM, £, w); ibid., 21 ‘Apr. 1874, Warion (£, GH, K); Oran, Balansa 535 (BM, E, K, w): Médéa, 900 m., Gay 2395 (BM, G, GH, K). The only other annual species of Alyssum occurring in the Iberian Peninsula and North Africa, and having dimorphic fruit indumentum and persistent sepals is A. strigosum (sect. ALyssum), which is frequently confused with A. granatense because of a similarity in habit and facies. The major differences between them are detailed in the following table: A. GRANATENSE A. STRIGOSUM LEAVES Linear-oblanceolate, lanceo- Oblanceolate, or broadly late or elliptic-oblong, gray- obovate-spathulate, green- ish-green with dense indu- ish with + sparse indumen- mentum of adpressed and tum, often _ strigose, of strongly branched elliptic sparingly branched, or un- hairs. branched, stellate _ hairs. PEDICELS Ascending or and Widely divergent and pat- often on ge main ent, or horizontal. axis. SEPALS Always persistent and erect, When persistent, widely 2-3.5 mm. long, lanceolate, spreading, 1.5-2 mm. long, with + adpressed indu- ovate-elliptic, with dense mentum and apical tufts of overall strigose indumen- strigose hairs. tum. ae ee Te ree 1965 | DUDLEY, STUDIES IN ALYSSUM 207 PETALS Emarginate with dense stri- Primarily bilobed, glabrous gose indumentum, 4-6 X or with sparse adpressed 1.5-2 mm. geome 2-3.5 X 0.4- FILAMENTS Very slender and wingless, Always with wide wings, edentate, and unappend- teeth, and connate append- aged. ages. Fruits Orbicular, up to5 X 5mm., Orbicular or oblate, up to values always equally in- 6 X 7 mm., valves more or flated. less ook inflated. From Alyssum alyssoides, the closest ally to A. granatense in section PsILoneMaA, the latter is distinguished by its always larger floral parts, larger fruits, globose and reduced nectaries, erect and adpressed position of the pedicels, and the dimorphic fruit indumentum. il, — Alyssum dasycarpum Steph. ex Willd. Linn. Sp. Pl. ed. 4. 3(1): 469. 1800 (!).— Fenzl in Tchihatcheff, Asie Mineure Bot. 1(3): 314. 1866.— Boiss. Fl. Or. 1: 285. 1867.— Fedtschenko, FI. Turkestan 47. 1906.— Busch in Kuznetsov, Busch & Fomin, FI. Cauc. Crit. 3(4): 600. 1909; in Fl. U.R.S.S. 8: 358. 1939. — Popov, Man. Fl. Tashk., fasc. 1-2. fig. 221. 1923-—1924.— Post & Dins- more, Fl. Syr., Palest. & Sinai, ed. 2. 1: 88. 1932.— Bornmiller, Repert. Sp. Nov. Beih. 89(1): 58. 1936. — Thiébaut, Fl. Lib.-Syr. 1: 71. 1936.— Grossheim, Fl. Kavk. ed. 2. 4: 218. tab. 25, fig. 5. 1950. — Parsa, Fl. Iran 1: 742. fig. 616. 1952.— Karjagin in FI. Azerbaid. 4: 274. 1953.—I. V. Pavlov (Ed.), Fl. Kazakh. 4: 282. tab. 35, fig. 9. 1961. — Dudley, Notes Bot. Gard. Edinb. 24(2): 157. fig. 1B. 1962. — Rech. Ark. Bot. 5(1): 168. 1963. — Ball & Dudley in Flora Europaea 1: 299. 1964. Holotype, Russia, in Siberia ad Kamam et Volgam fluvium, Stephan s.n. (LE, non vidi); isotypes (BM, G—DC, K). Annual, with many erect stems, up to 25 cm. long, rarely prostrate. Indumentum + dense, of coarse stellate hairs with long and few (but branched) rays, often appearing strigose. Leaves decreasing in size up- ward; the upper attenuate, obovate to oblong-oblanceolate, 7-11(—25) x (2.5-)4-9 mm. the lower wide spathulate and short petiolate, 20-35 < 12-15 mm. Inflorescence racemose or paniculate, up to 15 cm. long, often branching widely. Pedicels (1.2—)1.5-2 mm. long, divergent to as- cending, often subappressed to the rachis, with dense strigose and dimorphic indumentum. Sepals 2-3 mm. long, + persistent, with dense dimorphic and strigose indumentum. Petals obovate-spathulate, bifid or retuse, 2.5-3(-3.8) mm. long. Filaments 2-2.5 mm. long. Fruits elliptic-obovate or orbicular, truncate, 2.5-3(-4) > 2.5-3 mm.; valves equally inflated with thick flattened margins (ca. 1.5-2 mm. wide), and with dense dimorphic indumentum. Styles (1—)1.5—2 mm. long, stout, strongly dilated 208 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 at bases, and with dense dimorphic indumentum on the lower half. Seeds wingless. 27 = 16. Fl. Mar—June. Two varieties are recognized: A. Inflorescence elongate and many-flowered; plants ss leaves oblanceo- pee es eho es oa sy one eee eee . Var. dasycarpum. A. Sr condensed, few-flowered; plants age leaves broadly WOME SoG rss, coe o dw ok Kke Ch ReneS b. Var. minus. a. Var. dasycarpum. Px. a Fics. b, h. Pt. III, 6, Fu. LV, Fm. ser gui dasycarpum (Steph. ex Willd.) Meyer in Ledebour, Fl. Alt. 3: 150. 1831 (!); in Ledebour, Fl. Ross. 1: 127. 1842; Ic. Pl. Fl. Russ. 3: tab. ae 1831.— Fedtschenko, Fl. Ross. Austro- Orient. 5: 44. pl. 388, fig. B. 1931 Alyssum calycinoides Hausskn. in Bornmiiller, Repert. Sp. Nov. Beih. 89(1): 58. 1936, pro syn. (!). DisTRIBUTION AND HABITAT: widespread in disturbed and _ ruderal habitats, roadsides, fallow cultivated fields, vineyards, limestone ridges and screes, and steppe throughout southwestern Asia, including Caucasia, Turkey, Syria, Palestine, Iraq, Iran, Transcaspia, Afghanistan and Paki- stan; alt. 100-2600 m. Turkey. A2(E): Prov. Istanbul, oe 7 June 1916, Aznavour (G); Serai-koi, Frizaldszky sm. (wu). A4: Prov. Cankiri, valley of Cakmakli bah 800-900 m., 1929, Bornmiiller 18860 jai G, K), 13859 (BM, GH, K). A6: Tokat, Tokat, Aucher 4096 (BM, G, K, OxR). A7: Prov. Giimiisane, ae Guans nr. Sobran (Kovans), 1894, —— 6143 (BM, G, E, K, w); ibid., 9 June 1862, Bourgeau (w). A8: Prov. Erzurum, Horasan nr. Hopik, 1600 m., Davis 29378 (A, BM, E, K); ibid., Beibout (Bayburt), May 1843, Huet (Gc). A/B8: Prov. Giimiisane Erzurum, Bayburt- Erzurum, valley of Kassuklu, 1524— 1829 m., May 1853, Huet (BM, x). B3/4: Prov. ‘Ankara, Ankara-Polatli, 40 km. from Sakarya, 13 km. SW. of Polatli, 720 m., Huber-Morath 13732 (HM). B4: Prov. Ankara, Angoradur Monasteri, Ankara, 10 May 1907, Fréres E. C. (G); ibid., Kotte 1019 (x); Prov. Konya, Yavsan Memlehasi, nr. Tuz Gdli, Davis 18706 (z, K); Prov. Nigde Konya, Sultanhani-Cihanbeyli, 4 km. from Halkanli W. side of Tuz Gélii, 1000 m., Dudley, D. 35927 (a, £). BBS: Prov. Kayseri, Incesu Develi, 3 km. S. of Incesu, 1050 m., Huber-Morath 10984 (uM); ibid., Calasse (Talas) nr. Kayseri, Balansa 489 (c, w); ibid., Erdschias dagh (Excivas dag), Lerca dag nr. Kononia, 1600 m., May 1902, Zedebour (w); Prov. Kayseri Yozgat, Kayseri-Yozgat, Kopriilii, 1200 m., 1890, Pema 1936 (BM). B6: Prov. Sivas, 4 km. W. of Sivas, 13 June 1939, Reese (HM). B7: Prov. Erzincan, Beahican nr, Albuschikchan, 1890, Sintenis 2176, sub A. calycinoides (w); Prov. Erzurum, Erzurum, Zohrab 375 (x), C2: Prov. Denizli, Tavas-Denizli, 800-900 m. , Dudley, D. 35560a (A, E). C4: Prov. Konya, Konya, 4 June 1937, Reese (uM). C5: Prov. Nigde, Nigde- Ulukisla, S. side of pass, 47 km. from Ulukisla, 1450 m., Huber-Morath 12818 (um). C4: Prov. Konya, Konya, Post 14 (c); ibid., Cumra, Kiicik koy, Helbaek 2406 (£); ibid., 4 km. from Konya, nr. Sille, 1040 m., Huber-Morath HM). Armenia, Erzurum- Tokat, Aucher 4098B (Gc); ibid., Aucher 4098A (Gc, w); Calvert & Zohrab 45—p.p. (oxF). Anatolia, Noé 947 (c). Syria. Hafar- 1965] DUDLEY, STUDIES IN ALYSSUM 209 Syrian desert, 100 m., Dinsmore 20309 (g, K); Nebk, Davis 5527 (A, BM, E, kK); Damascus, Leamid Atiyeh, Post 1109 (pm); Jebel Abur Rejmein, 2 May 1900, Post (eM, G, K); S. of Jarud, 180 m., Dinsmore 22478 (K). Transjordan. Above Wadi Musa, — Moan, 1219 m., Davis 8 8677 (A, BM, E, K); Ein Musa, 1372 m., Davis 8884 (BM, E, k). Iraq. 18 km. W. of Suleimani, 825 m., Eig & Zohary (avy, non vidi); Moan, Moan-Ain Musa, 18 Apr. 1929, Eig et al. (, w). Kurdistan, Oguhah, Graham sm, (K). Iran. Bakhtiari, Oregon, Damane- caw range, 2300 m., Wendelbo 1726 (Bc, E); Prov. Khorasan, Turbat-e-Haidari, m., Rechinger 4329 (Gc, w); ibid., Rechinger 4362 (x, w); Robat Safid, iene m., Rechinger 7336 (w); nr. Kaswin, 3 May 1882, Polak (k, c, Ww); ibid., 1200 m., Schmid 5044 (G); ibid., 1200-1300 m., 1902, Bornmiiller 6233 (BM, G, W); Prov. Fars, Shiraz-Kazerun, Gauba & Sebati 197 (w); Chiraz (Shiraz), Aucher 4091 (c, K, OXF); Shiraz-Persepolis, 800-1000 m., Schmid 3518 (G); Prov. Kerman /Fars, Saidabad-Cafut, 1900 m., Rechinger 3194 (c, w); Prov. Shahrud/Bustam, Khosh-Jaila, ca. 73 km. from Shahrud, 2000-2200 m., Rechinger 5442 (Gc, w); Prov. Hamadan, Aq Bulag, ca. 100 km. N. of Ham- adan, Rioux & Golvan 212 (w); Abedeh-Daulatabad, 1500-2000 m., Schmid 3 (w); 5 miles N. of Daulatabad, 1372 m., Cowan & Darlington 1096 (K); i Elburs, Keredj, Rechinger 244 (x, w); Scharabad, May 1858, Bunge (c, GH, K); Tabriz, Gilliat-Smith 1773-1782 (x); Keredj, 16 May 1934, Gauba (w); Emirabad, nr. Ibrahirabad, 1829 m., Cowan & Darlington 588 (kK); 10 miles E. of Zorab, 1219 m., idem 1775 (kK); 20 miles W. of Zorab, idem 1734 (K); Sultanabad, 1524 m., Lindsay 31 (Bm); Isphahan, Aucher 4091A (G, K, w); Isphahan-Teheran, May 1859, Bunge (Gc); Transcaspian, Aschabad, Annaju- Gjaurs, 1900, Sintenis 51 (BM, G, GH, E, K, W); Persia, ; ibid., Kotschy 183 (Gc). Afghanistan. Herat, 1100 m., Kéie 4 1600 m., Koie 4164 (w); Kabul, Kabul-Tangi Gharu, 1700 m., ae 1076 (w); E. of Kabul, Budchak, 1770 m., Gilli 1075 (w); Prov. Bamian, Band-i-Amir, Lake Band-i-Panir, 2800 m., Rechinger 18394 (w); 2800-2900 m., Rechinger 18226 (w); Wendelbo 4783 (BG, E); 2900 m., Wendelbo 4761 (Bc, E); 3200 m., Volk 2767 es Afghanistan/Iran. Hari Reid Valley-Khorasan, Aitchison 104 (BM, G, GH, E, K); Jouvnal-Sinab, Griffith 299 (xk); Griffith 1366 (Kk, w). peak ee 1525 m., Lace 3574 (gE, K); 1600 m., 17 fi 1888, Lace (e); W. Baluchistan, Ziaret, 2338, Stewart 28087 (mMIcH); Nichara, Stocks 910 Caucasus, Bunge 75 (K, w); ibid., Taltisch nr. Swant, Meyer 1601 (G, K); ibid., 9 May 1947, Grubner (BM); ibid., nr. Codshadoi, Swant, June 1838, Hohenacker (BM, G, GH, K); nr. Khabadian, 610-914 m., 1883, Regel, p.p. (B M); Transcaucasus, Nachitschevan, dist. Dzulfa, Darry-Dagh- Dzhulfa, Apr. 1934, Karjagin (a); Erivan, Buhse 118 (G, w); Azerbaidjan, dist. Salma, Dehr- man, 2 Apr. 1828, Szovits (Gc); Turkistan, Chuma, Karolif & Krause s.n. (G); Transcaspian, Kisil Arsah, 1885, Becker (k, w); in deserto Caspio, Pallas s.n. rd Caspium nr. Astrachan, 1819, Fischer (c-pc, K); Astrachan, 1820, Steven (G-Dc); Soviet Armenia, Kotairk region, Vokhchabad, Darabulal, 26 May 1956, Mulkidzhanyan (pM, £); Vedi region, Arazdian-Kiarki, 29 June 1960, Takhtajan et al. (gM, w); Kazakh S.S.R., Prov. Syr-Dariia, Perovsk, Akkum Tashk. desert, 11 May 1916, Tsintserling (a); Prov. Semiretschensk, Karatal river, 4 May 1902, Saposhnikov (pM, GH); Turkmenian S.S.R., dist. Krasbovodsk, steppe nr. Kizil Arvat, Androsov 2565 (GH); Siberia, Songarei, Schrenk s.n. (G, GH); Soongoro- Kirghisici, nr. Ajagus, Karelin & Kiriloff 73 (8M, E, GH, K); Siberia, Pallas s.n. (pm); Altai, Ledebour 302 (c, w); ibid., Politons s.n. , w); Russia/ China, Songaria Chin. Lake Saisang-Nor, Meyer s.n. (E, GH, K, W 210 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. 46 From all the other oriental species in section PstLoNEMA, Alyssum dasycarpum may be distinguished by the dimorphic and strigose fruit indumentum, longer and basally dilated styles, and wingless seeds. The always orbicular fruits of A. granatense from the Iberian Peninsula and North Africa are at least twice the size of those of A. dasycarpum. Also, the tuberculate hairs on the fruits of the western European species are sparser and longer, resembling more those of A. hirsutum (sect. ALYSSUM). Alyssum strigosum (sect. ALyssuM) is sometimes confused with A. dasy- carpum because of the similar dimorphic fruit indumentum. However, the fruits of A. strigosum are larger and are covered with longer furcate hairs. In addition, the winged, dentate and appendaged filaments, larger floral parts, winged seeds, and widely divergent pedicels are diagnostic for A. strigosum. No specimens which have winged seeds have been found throughout the range of A. dasycarpum. A plausible explanation for the Ukranian A. dasycarpum var. pterospermum (Bordz. Bull. Jard. Bot. Kiev 7-8: 17. 1928) is that unless the seeds are examined under magnification, the lighter colored and often translucent radicle may be misinterpreted as a wing formation. b. Var. minus Bornm. ex Dudley, Notes Bot. Gard. Edinb. 23(2): 157. fig. 1A. 1962 (!). Holotype, Iran, inter Ispahan et Hamadan, ad pagum Mohammedi, 1800 m., 17 Mar. 1892, Bornmiiller 2174 (£); isotypes (BM, G, K, OXF, W). DISTRIBUTION AND HABITAT: a scattered and limited distribution in desert and fallow lands, and stony hillsides of Syria and Iran. Syria. Nebk-Quaryatein, 5 Apr. 1890, Post (Bm). Iran. ue -Yesd nr. Bambis, 1900 m., 1892, Bornmiiller 2173 (Gc); nr. Kom m., 1892, Born- miller 2175 (BM, E, G, K, OXF, W); nr. Dalechi, Kotschy 181 (BM, G-p.p., K, W); Uenak ne. Teheran, Kotschy 64 (Gc, w); Karawanseri, Kaswin, 30 Apr. 1892, Pichler (c, K, W); nr. Kaswin (Mazraeh), 1200 m., Schmid 5048 (w); ibid., Schmid 5007 (Gc); S. of Tabriz, Gilliat-Smith 1356, 1381, 1783 (K). Intermediate between the varieties of A. dasycarpum: Iran. Prov. Kerman, Kerman-Sultanabad, Sirdjian (Saidabad), Mashiz-Khan-e-Sorck, 2000-2580 m., Rechinger 3006 (G, w). 12. Alyssum homalocarpum (Fischer & Meyer) Boiss, Fl. Or. 1: 285. 1867 (!).— Muschler, A Manual Flora of Egypt 1: 422. 1912.— Post & Dinsmore, Fl. Syr. Pal. & Sinai, ed. 2. 1: 87. 1932,— Burtt & Lewis, Kew Bull. 3: 283. 1949. — Parsa, Fl. Iran. 1: 744. 1952. — Montasir & Hassib. Ill. Man. Fl. Egypt 157. 1956. — Jafri, Notes Bot. Gard. Edinb. 22: 95. 1956.— Rech. Bot. Not. 115: 37. 1962; Ark, Bot. 5(1): 169. fig. 14. 1963. — Dudley in Rech. Fl. Lowland Iraq 306. 1964. Psilonema homalocarpum Fischer & Meyer, Ind. Sem. Hort. Petrop. 6: 63. 1840 Holotype, semina in Arabia Petraea, May 1837, Schimper (LE, non vidi— cultivated in Hortus Petropolitanus from seed collected by i seme 1965 | DUDLEY, STUDIES IN ALYSSUM 211 Schimper). Authentic specimens (ex hb. Gay) cultivated (1837) in Jardin Luxembourg 7 Sra same seed collection have been examined in the herbaria of G, GH, : Alyssum herelscien ci Ann. Sci. Nat. Paris, II. 17: 156. Sta (!). Holo- type, Sinai, Mt. Horeb, Aucher 257 (P, non vidi); isotype Alyssum ‘neesili Velen. Sitz-ber. Bohm. Ges. Wissen. Prat 11: 45 1911 (!).— Velen. Repert. Sp. Nov. 13: 25. 1913. ery Arabia, in distr. Harara et Wudijan, Drejheme et Zerko, 1909, Musil (pr Alyssum nomismocarpum Rech., Aellan & Esfand. Phyton 2: 36. 1951 (7), Holotype, Persia, Prov. Lars, "Hadjiabad p prope Tarum, in declivibus siccis saxosis, ca. 900 m., 29 Apr. 1948, Rechinger et al. 3272 (Ww); isotypes (E, G, K). Annual, with numerous, erect but brittle stems, 5-20 cm. long. Leaves oblong-spathulate, 1-4(—5) & (0.5—)1-3 cm., acute, minutely denticulate above the middle, indumentum dense and ashy, of stellate hairs with + divergent and long rays. Jnflorescence corymbose, 4-10 cm. long with numerous patent or ascending fragile branches. Pedicels 1.3-2 mm. long, horizontal, with sparse indumentum. Sepals deciduous, 0.8-1 mm. long, with dense ashy indumentum of + long-rayed stellate hairs. Petals linear-cuneate, subemarginate or entire, 0.5—-1 mm. glabrous. Filaments subulate, edentate, ca. 1 mm. long. Fruits broadly obovate, truncate, glabrous, 3-7 mm. long and wide, margins prominently papillose, and coloring reddish-purple when dry; valves equally inflated, with conspicuous venation. Styles 0.5—0.7 (-1) mm. long, dilated basally. Seeds narrowly winged. Fl. Feb—Apr. DISTRIBUTION AND HABITAT: a Saharo-Sindian species of dry silty river beds, limestone and sandy slopes, basaltic screes, sandstone, desert and calcareous cultivated lands in Egypt, Saudi Arabia, Kuweit, Palestine, Iraq, Iran, and West Pakistan; alt. (120—)200—1000(—1600) m. Egypt. Mergheb, May 1904, Muschler (kK); Ouadi Aschar (Wadi Isleh), 1903, Muschler (c); nr. Belbeyi, 1837, Schubert (K). Saudi Arabia. Arabian desert, Schweinfurth 127 (Kx); Galalah, Schweinfurth 8 (Gc); Talpine camp- -Hafarol B ? } Batin-Mahazul, 100 miles W. of Kuweit, 120 m., Dickson 511 (x); Wady Batin, 1219 m., Fitzgerald 15610/3 (BM). Syria. Jabal-Tenf, Gombault 1642 (p, non vidi). Jordan. Naqb Ishtar, Hunting Aero Survey 16 (£). Palestine. 40 km. S. of Maan, 1929, Eig & Zohary (vj, non vidi). Iraq. ca. 400 km. W. of Baghdad, 1933, idem (Huy, non vidi); Rutba-Ramadi, 15 km. from Rutba, 750 m., Rechinger 9894 (w); 10 km. SW. of Rutba, 500 m. Rawi 21047 (K); 18 km. S. of Rutba, 640 m., Rawi 14637 & 14899 (x); dist. Diwaniya, 40 km. WNW. of Shabicha, 380 m., Rechinger 13642 (w); Shabicha, 200 m., Gillett & Rawi 6276 (Kk); Sharaban (Shabicha), Beluchi, Aug. 1886, Jennings (xk); 40 km. NW. of Shabicha, 390 m., Guest, Rawi & Rechinger 19306A (kK); As- Salman, nr. Ansab, 145 km. SE. of As-Salam (Southern desert, ad confines Saudi Arabia) 340 m., Rechinger 13810 (w); 62 km. WNW. of Ansab, 135 km. SSW. of As-Salman, 360 m., Rechinger 13781 (w); Darb Al’ Haj, Saddi border, 360 m. Guest, Rawi & Rechinger 19083 (K); 12 km. ESE. of As-Salman, 240 m., 212 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 idem 18739 (K); nr. Ansab, 145 km. S. of As-Salman, idem 1893 (K). Shershah, Mar. 1859, Bunge (Gc, K); Prov. Baluchistan, nr. Zahedan, aon m., Gauba & Sabeti 196 (w); Montes Karvandar-Khash (Vasht or Kwash), irae shahr (Bampur), 1500-1600 m., Rechinger et al. 3958 (w). West Pakistan. North Baluchistan, pee Landon (p M). This is the only glabrous-fruited representative of section PSILONEMA having deciduous sepals, and it is the only Alyssum with minute teeth towards the apices of the leaves. Rechinger (loc. cit.) notes and illustrates the papillose fruit margins of A. homalocarpum. This character, although rare in the genus, is not unique. A Greek endemic, A. euboeum (sect. OpONTARRHENA) also has fruits, the margins of which are papillose, especially towards the apices. The fruit size of A. homalocarpum varies considerably; the smallest measure circa 3 < 3 mm., and the largest 7 X 7 mm., but without any apparent geographical pattern. None of the char- acters said to distinguish A. nomismocarpum from A. homalocarpum, i.., longer racemes, fragile stems, larger fruits, and shorter pedicels, indicates a satisfactory specific separation when the entire range of morphological variation is taken into consideration. Similarly, A. musili and A. horebicum do not possess sufficient discontinuities to separate them from A. homalo- carpum. EXPLANATION OF PLATES PLATE I Alyssum blepharocarpum: Holotype, Huber-Morath 13722 (HM). PLATE II Fics. hs Petal types, X 20: a, Alyssum desertorum var. — (Davis 27622); ce gh var, dasycar pum oo - 29378); c, A. pinifolium (Kirk) ; . minutum (Davis 18040); apfii (Davis 28694); f, A. lepidotim (dais 183914): g, A. ge moadicome (Davis 38893). -r. Long filament types, X 20: h, Alyssum ag ap igen jon ees pum yes genet i, A. hueti CDudley” D. 35230); j, A. minutum (Da 1894 0)5 bes A. xanthocarpum (Davis 19411); 1, A. lepidotum (Dai 183010): m, A. a vides (Davis 20328); n, A. strictum (Sintenis 5614); 0, A. repens var. pie Frauen al (Davis 30295) ; Pp, - filiforme (Davis 29072); i A. giosnanum — ne 1381); 1, A. pinifolium (Kirk S. S-y. Short filament types, < 20: S, sce huetii (Dudley, D. 35230) 2 he stribrnyi (Dudley, D. 34558); u, A. pseudo-mouradicum (Davis 38893) ; v, A. corningii (Siehe 241); w, A. vw utum (Davis 18940); x, A. sei or co (Davis 27622); y, A. strigosum subsp. strigosum (Dudley, 638). *Parsa, in his Fl. de Iran 1: 745. 1952, records Alyssum homalocarpum from the Shah Sam he Lut desert of Iran, and indicates that the original publication this record is to be found on page 326 of the Bornmiiller paper titled “Aus Pflanzenwelt Pi inner-Iranischen Wiistengiirtels” (Repert. Sp. Nov. 40: 1936 ). Bornmiiller, however, does not mention A. homalocarpum, or in fact, any species of Alyssum in that particular paper. Bare i iii tial 1965 | DUDLEY, STUDIES IN ALYSSUM 213 PLATE III Fruit types, X 8: a, Alyssum huetii (Dudley, D. 35230); b, A. desertorum var. desertorum (Davi 27622); ¢, A. dasycarpum var. dasycarpum (Davis 9 ; 7 A. hi rescott pseudo-mouradicum (Dws 38893 ) : A. corsicum pied 7 h, A, fili- forme (Davis 31609); k, A. haussknechti Date’ é oie t, wag (Dudley, D. 35860); m, A. murale var. murale (Dudley, D cassium (Kiihne 1446); 0, A, asiele ‘(Pinard): p, A. Ponca, iadiey, BD. 36151). PLATE IV Hair types: a, Alyssum chondrogynum (Davis 3083), from the fruit, x 200; b, A. huetii (Dudley, D. ois from the fruit, X 200; c, A. stapfii (Davis 28694), from the fruit, x 5 d, A. xanthocarpum (Davis 19411), from the ins leaf surface, X 50 pf pseudo- mouradicum (Davis 38893), from the upper surface of sterile at leaf, X 50; f, A. eriophyllum (Haussknecht), from leaf of sterile shoot, & 50; g, A. s. strigosum ‘subsp. strigosum (Dudley, D. 34638), from the stem, X 50; h, A. strigosum subsp. strigosum (Dudley, D. 34638), . a m, A. mouradicum (Balls 186), from lower surface of basal cauline leaf, < 50: n, "A. szowitsianum (Dudley, D. ’35210a), from fruit, X 50. Jour. ARNOLD Arps. VoL. 46 Pate I = t | Aa i ch Fh ih ol G. Wet, - DR. A. HUBER-MORATH { Flora Helvetica é 4 oa a leg. Or, A. Huber-Morath DupLey, Stupres in ALyssum Jour. ARNOLD Ars. VOL. 46 Prate II ATE DupbLey, STUDIES IN ALYSSUM Jour. ARNOLD Ars. VoL. 46 Pate III Dupbtey, Stupres in ALyssum Jour. ARNOLD Arp. VOL. 46 PLaTE IV DupLEY, STUDIES IN ALYSSUM 218 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 THE GENERA OF VALERIANACEAE AND DIPSACACEAE IN THE SOUTHEASTERN UNITED STATES * I. K. FERGUSON VALERIANACEAE Batsch, Tab. Affin. Reg. Veg. 227. 1802, nom. cons. VALERIAN FAMILY) Annual or perennial herbs, sometimes woody at base. Leaves in basal rosettes or opposite, pinnately divided or entire, exstipulate, the bases often arene ih Inflorescence a monochasium, thyrse, or many-flowered compound dichasial cyme, sometimes condensed and capitate, bracteate and shee goreene rer Flowers irregular or almost regular, bi- sexual or unisexual. Calyx obsolete or developing late and becoming con- spicuous only in fruit, annular [or toothed], adnate to ovary. Corolla 4 [bilabiate]. Stamens epipetalous and alternate with the corolla lobes, vary- ing in number, usually 3 in our genera [1, 2, 3, or 4]; anthers versatile, 2- or 4-lobed, 4-locular, introrse, dehiscing longitudinally: pollen tricolpate, echinate. Gynoecium syncarpous, ovary inferior, 3-locular, with two locules usually suppressed and one fertile, with a solitary, pendulous, an- atropous ovule; style 1, stigma simple or lobed. Fruit dry, indehiscent, the calyx often developing into a winged, awned, or plumose pappus. Seed 1; endosperm absent; embryo large, straight, the cotyledons oblong, the radicle superior. TypE GENUs: Valeriana L. A family of about ten genera and 370-400 species, widely distributed but occurring mainly in the North Temperate regions and absent from Australasia; three ogi native and one introduced in North America; two genera in our a Valerianaceae are a natural family closely related to Dipsacaceae but ‘Prepared for a generic flora of the southeastern be aie States, a joint project 0 Louisiana. The descriptions are base marily o e plants of this area, with any a material in brackets. Rested which the author has not seen are y an as The Leger is ecu to Dr. Wood for his aid and valuable criticisms; to George K. Brizicky, for his guidance and — and to Mrs. Gordon W. Dillon, for her help i in the preparation of the typescript + 1965] FERGUSON, VALERIANACEAE AND DIPSACACEAE 219 distinguished by the 3-locular ovary (one locule fertile), cymose inflores- cences, and seeds without endosperm. The family also has affinities with the Caprifoliaceae and Rubiaceae. Members of the family often have a very characteristic unpleasant odor. Volatile oils occurring mainly in the root and rhizome have been investi- gated extensively, especially in Valeriana. The cytogenetics of the family have not been very fully investigated, although chromosome numbers for some species of Valeriana (21 species of 300) and Valerianella (11 species of 60) are available. However, only a few chromosome counts are reported for four additional genera: Centran- thus, 2n = 14, 32 (three species); Fedia, 2n = 32 (one species) ; Patrina, 2n = 22 (one species); and Astrephia, 2n = 32 (one species). The family is of little economic importance. A few members, among them Centranthus ruber (red valerian) and Valeriana officinalis (garden heliotrope or common valerian), are grown as ornamentals. The latter is used as a source of the drug “valerian.” Corn salad (Valerianella Locusta), as its name implies, is used in salads. Spikenard (Nardostachys Jatamanst) yields an oil which has been used in perfumery in the East. REFERENCES: ARNAL, C. Différenciation basipéte des faisceaux libéroligneux stylaires de Cen- tranthus angustifoliuns DC. Compt. Rend. Acad. Sci. Paris 222: 674-676. 1946. ASPLUND, E. Studien iiber die Entwicklungsgeschichte der Bliten einiger Vuletioraneer, Vet.-akad. Stockholm Handl. IV. 61(3): 1-66. 1920. BarLton, H. Valérianacées. Hist. Pl. 7: 504-518. 1879. BenTHram, G., & J. D. Hooker. Valerianeae. Gen. Pl. 2: 151-156. 1874. BorstnI, O. E. Valerianaceae. Inst. Paran. Catal. Gén. 12: 1, 2. 1956. . Revision de las Valerianaceas de Brasil. Lilloa 31: 149-170. 1962. BucHENAU, F. G. P. Ueber die Bliithenentwickelung einiger Dipsaceen, Vale- rianeen und Compositen. Abh. Senckenberg. Naturf. Ges. 55: 106-132. pls. 5, 6. 1854. CANDOLLE, A. P. pe. Valerianeae. Prodr. 4: 623-642. 1830. émoire sur la famille des Valérianées. 24 pp. pls. 1-5. Paris. 1832. Cuatin, J. Etudes hater eo chimiques et médicales sur les Valérianées. 148 pp. pls. 1-14. Paris. 1872 Dempster, L. T. Dim ene ism in the fruits of Plectritis, and its taxonomic implications. Brittonia 10: 14-28 58. Duresne, P. Histoire naturelle et médicale de la famille de Valérianées. 61 pp. pls. 1-3. Montpellier. 1811. a Dyat, S. C. Studies in the family Valerianaceae. Abstr. Theses Cornell Univ. 1941: 333-336. 1942.* Ewan, J. Centranthus: a new immigrant genus to California. Leafl. West. Bot. 2: 195. 1939. : GRAEBNER, P. Die Gattungen der natiirlichen gros der Valerianaceae. Bot. Jahrb. 37: 464-480. 1906. [See also ibid. 436-4 Hock, F. Beitrage zur jane eae Gruppirung nd Geographischen Verbrei- tung der Valerianaceen. Bot. Jahrb. 3: 1-73. 1 . Valerianaceae. Nat. ncaa tV.. 45 172- cy 1891. 220 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 . Verwandtschaftsbeziehungen der Valerianaceen und Dipsacaceen. Bot. Jahrb. 31: 405-411. 1902. Kitup, E. P. Valerianaceae. In: J. F. Macsrive, Flora of Peru. Publ. Field Mus. Bot. 13(6): 287-321. 1937. ‘ale O. B. N. Valerianaceae. Jn: E. WARMING, Symbolae ad floram Brasiliae . Vid. Medd. Naturh. For. Kjgbenh. 34: 139, 140. 1883. rs K. Experimental and cytological studies in Centranthus. Bot. Not. 111: 301-305. 1958. Nietsen, S. D. Systematic studies in the Valerianaceae. Am. Midl. Nat. 42: 480-499. 1949. [ Plectritis. | PizzoLonco, P. Ricerche sulla cariologia del genere Centranthus e loro impor- tanza tassonomica. Delpinoa II. 1: 149-164. pls. 1, 2. 1959. Poucaurs, M. L. Recherches caryologiques sur les Rubiales. Revue Gén. Bot. 56: 97-138. 1949. [Le noyau des Valérianacées, 101-109; Fedia, Vale- rianella, Valeriana, Centranthus. | RENDLE, A. B. The classification of flowering plants. vol. Dicotyledons. 636 pp. Cambridge, England. 1925. Meeseeapee ge 567-5 SCHERMERHORN, J. M. W. By, eds. Lynn ee Monograph V. 67 pp. Boston. 1962. Neier 62-67 ral Sources, R. Embryogénie des Valérianacées. Développement de 1l’embryon chez les Centranthus. Compt. Rend. Acad. Sci. Paris 212: 718-720. 1941. VEsQUE, J. Développement du sac embryonnaire des phanérogames a spermes. Ann. Sci. Nat. Bot. VI. 6: 237-285. pls. 1-8. 1878. [Valerianella pumila, Centranthus ruber, 255, 256 Vat, L. Contribution 4 l’anatomie des Valérianacées. Ann. Univ. Grenoble 15: 561-605. 1903.* Wacenitz, G. Pollenmorphologie der mitteleuropadischen Valerianaceen. Flora 143: 473-485. pl. 5. 1956. [Valeriana, Valerianella, Centranthus.| . Die systematische Stellung der Rubiaceae. Bot. Jahrb. 79: 17-35. 1959. [Gentianales = Contortae + Rubiaceae; Dipsacales = Caprifolia- ceae, a ceae, Valerianaceae, Heuiak | alerianaceae. Jn: H. Metcuior, Engler’s Syllabus der Pflanzen- familien. a 12. 2: 475-477. 1964. WEeBERLING, F. Die Infloreszenzen der Valerianaceen und ihre systematische Bedeutung. Akad. Wiss. Lit. Abh. Math.-Naturw. Mainz 1961: 151-281. 1961. KEY TO THE GENERA OF VALERIANACEAE Calyx inrolled in flower, developing in fruit to form a conspicuous pappus with a umose awns; fruit 1-locular; plant perennial. ........ 1. Valeriana. Calyx obsolete; fruit 3- locular: plant annual or biennial. .... 2. Valerianella. 1. Valeriana Linnaeus, Sp. Pl. 1: 31. 1753; Gen. Pl. ed. 5. 19. 1754. Perennial herbs or vines with thickened, strong-scented, fleshy [woody or tuberous] roots, sometimes stoloniferous; glabrous or sparsely hairy with short, simple hairs. Leaves petiolate ‘to nearly sessile, basal and cauline, undivided, pinnate, pinnatifid, or ternate. Inflorescence usually a dichasial cyme or thyrse, terminal or axillary, sometimes lax or sub- capitate, bracteate and sii cteckaie Flowers irregular, bisexual or uni- 1965 | FERGUSON, VALERIANACEAE AND DIPSACACEAE 221 sexual. Calyx inrolled in flower, enlarging and developing in fruit to form a pappus with short, sessile, patelliform, membranaceous limb and 5-15 plumose awns [or the limb cupuliform and irregularly dentate]. Corolla funnelform or campanulate [salverform or rotate], the tube slightly sac- cate at the base [or straight], usually more or less hairy in the throat, 5[3 or 4]-lobed. Stamens 3 [rarely 4], inserted toward the top of the corolla tube, usually exserted (but included in V. scandens); anthers 2- or 4-lobed. Stigma simple or 3-lobed; ovary 3-locular, with two locules suppressed and one fertile with a solitary ovule. Fruit a unilocular, com- pressed achene with 3 nerves on the outer and 1 on the inner surface. LeEcToTYPE SPEcIEs: Valeriana pyrenaica L.; see Britton & Brown, Illus. Fl. No. U. S. ed. 2. 3: 284. 1913. (Name Medieval Latin, from valere, to be strong; said also to be named for the Roman emperor Valerianus). A genus of about 300 species of Europe, Asia, Africa, and America, centered mainly in the temperate regions of the Northern Hemisphere and the mountains of South America; eleven species in the United States, two in our area. Of the seven sections recognized by Hock, six (Pseudiastrephia Hock, Hybocarpus Hock, Valerianopsis Wedd., Phyllactis Pers., Porteria Hook., and Aretiastrum DC.) are entirely South American and have been various- ly treated by different authors as distinct genera or as sections within Valeriana. The North American species belong to sect. VALERIANA (§ Euvaleriana Hock). Different authors have recognized a varying number of series within this section (cf. Héck, Meyer). The genus is well defined by the combination of perennial habit, three stamens, and inrolled calyx usually enlarging and forming a pappus on the one-locular fruit. It is probably most closely related to Valerianella and to Centranthus, which is distinguished by one stamen and spurred corolla. Valeriana scandens L., readily recognized by its climbing habit, occurs in Mexico, Central and South America, and the West Indies, northward in peninsular Florida to a shell-mound in Duval County. It seems likely that this species is a calciphile, but it does not appear to occur on the Florida Keys. It is placed by Meyer in the Sorbifoliae, a series distin- guished chiefly by included stamens and two-lobed anthers. Meyer recog- nizes two varieties (which some authors have treated as distinct species) : var. scandens, with tripartite leaves, and var. Candolleana (Gard.) Muell., with entire leaves. The widespread var. scandens occurs in Florida, but var. Candolleana occurs northward only to Hispaniola and Cuba. Inter- mediate specimens with both entire and tripartite leaves have been found, however, both in our area and elsewhere. Valeriana pauciflora Michx., distinguished by its long, funnelform corolla tube and stoloniferous habit, occurs from Pennsylvania to Illinois, Tennessee, and Virginia. It is placed in the series Officinales Hick, separated by the rhizomatous or stolonifer- ous roots and four-lobed anthers. 222 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 Sexual polymorphism occurs in the genus in several ways, the plants being completely hermaphrodite, polygamous, polygamodioecious, or di- oecious, with dioecism least common. The presence of unisexual as well as bisexual flowers is easily detected by the differential lengths of the corolla. The mode of pollination appears to be unknown but is probably ento- mophilous, Nectaries are reported to occur in the saccate base of the corolla tube. The occurrence of proterandry assures cross-pollination in bisexual flowers. e cytogenetics of Valeriana officinalis, 2n = 14, 28, 56, has been extensively investigated, but the cytology of the genus as a whole is not well known. Chromosome numbers of 2n = 14, 16, 18, 28, 32, 56, and 64 have been reported. REFERENCES: Under family references see PoucguEs, WaAGENITz (1956), and WEBERLING. ARTIUSHENKO, Z. T., & I. N. Konovatov. Morphology of fruits belonging to the nut and nutlet types. (In Russian.) Acta Inst. Bot. Acad. Sci. URSS. 7. Morphol. Anat. Pl. 2: 170-192. 1951. [Valeriana fruit a dry syncarpous drupe, 185, 186, 188. ] it O. E. Addenda a las Valeriandaceas argentinas I. Lilloa 12: 23-27. 946. [Races of V. scanden kes del estado de Santa Catarina (Brasil). Sellowia 15: 123- 136. 1963. Exzenca, G. Investigation of viability and germination rate of Atropa bella- donna, Digitalis purpurea, Digitalis lanata, Valeriana officinalis and Viola inicalor in connection vith different seed treatments. (In Dutch.) Herba 11: 73-78. 1952.* FaucoNNET, L. Variations saisonniéres chez Valeriana eee L. (German summary.) Bull. Soc. Bot. Suisse 57: 122-131. 1947 GSTIRNER, F. Evaluation of valerian root. Planta Med. ” 81-85. 1958.* Havuscuitp, F. Die Problematik der sedativen Baldrianwirkung. Pharmazie 13: 420-422. 1958.* Hetwic, B. Uber die Frage der Heterorhizie bei Radix Valerianae officinalis. Ber. Deutsch. Bot. Ges. 46: 595-609. 1928 IrmiscH, T. Beitrag zur Naturgeschichte der einheimischen Valeriana-Arten insbesondere der Valeriana officinalis und dioica. Abh. Naturf. Ges. Halle i: 4. pls. 1-4. 1854. eit io F. Aunteniical data on the localization of essential oil in some Valeriana species. (In Ukrainian; Russian summary.) Bot. Zhur. Kiev 10: 81-86. 1953. KrePINsKY, J., V. Herout, & F. Sorm. Plant substances. VI. The isolation of neutral products from the root of valerian (Valeriana officinalis L.). Collect. Czech. Chem. Commun. 24: 1884-1896. 1959.* Kuunuottz-Lorpat, G. Contribution 4 la biologie de Valeriana tripteris L. Feuille Nat. II. 3: 101. 1948.* Lattanzi, V. Valeriana officinalis L., ene and adulterations. (In Italian.) Fitoterapia II. 26: 590-605. 19 LAWALREE, A. Le groupe du Valerian officinalis L. en Belgique. Bull. Jard. Bot. Bruxelles 22: 193-200. 1952. 1965] FERGUSON, VALERIANACEAE AND DIPSACACEAE 223 MEIJeErs, T. Een onderzoek van het a Valeriana officinalis L. in Nederland. 122 pp. 23 pls.’s Gravenhage. 1957. . Methods to determine different cytotypes of Valeriana officinalis L. s. lato. Ba Dutch; English summary.) Pharm. Weekbl. 95(9): 269-278. 1960 MEVER, : G. Valeriana in North America and the West Indies (Valerianaceae). Ann. Missouri Bot. Gard. 38: 377-503. 1951. [An extensive taxonomic revision ’ The genus Valeriana in East Tropical and South Africa. Jour. Linn. Soc. Bot. 55: 761-771. 1958. [Distribution and taxonomy. ] Puitipson, W. R. Studies in the development of the inflorescence. III. The thyrse of Valeriana officinalis L. Ann. Bot. II. 11: 409-416. pl. 3. 1947. Pirtier, H., & E. P. Kite. Venezuelan species of Valeriana, section Porteria. Jour. Wash. Acad. Sci. 16: 422-428. 1926. ProtassenyA, G. D. Zur Frage des karyologischen Unterschieds einiger von Valeriana officinalis L. abgesonderter Baldrianarten. (In Russian.) Arbeit. Bot. Kab. Centr. Moorversuchsstat. Minsk 1: 83-92. 1930.* [See German summary in Bot. Centralbl. 162: 52. 1932.] Runguist, E. Zytologische und morphologische Valeriana-Untersuchungen. Hereditas 23: 279-286. 1937. aah N. H. Quantitative studies in angiosperm taxonomy. X. Valeriana. Geranium. XII. Mimulus. Castanea 29: 138-150. 1965. eesti M. Polyploidy in oe officinalis Linn. in relation to its ecology and distribution. Jour. Linn. Soc. Bot. 53: 159-186. 1947. Studies in cyto-eco ses, geographic distribution and evolution of Valeriana L. Bull. Acad. Polon. Sci. Lett. Sci. Math. Nat. B. I. 1950: 149- 175. pls. 3-5. 1951. Meiosis in a polyhaploid twin plant and a hexaploid hybrid of Valeri- ana sambucifolia Mikan. Acta Soc. Bot. Polon. 23: 359-374. pl. 1 5 Cytological studies in a chromosomal aberrant of oe and its derivatives. Polsk. Akad. Nauk. Folia Biol. 10: 155-168. pls. 1, 1962. SpraGuE, T. A. The British forms of Valeriana officinalis. Watsonia 2: vee 147. 1952 STROH, G. Valeriana L. Provisorische Liste der alterweltlichen Arten. Repert. Sp. Nov. 40: 225-233. 1936. Topp, B. H. The cytology of the Valerianaceae with special reference to the status of the British forms of Valeriana officinalis. Thesis, Univ. Durham. L* Troirsky, N. Résultats des recherches morphologiques et biologiques sur le Velurtana alliariaefolia Vahl du Caucase. (In Russian; French summary.) Moniteur Jard. Bot. Tiflis 46/47: 77-100, (101). 1919. VoroskiLov, V. N. Medicinal Valeriana. 160 pp. Moscow. 1959 Wattuer, E, Zur Morphologie und Systematik des Arzneibaldrians in Mittel- europa. Mitt. Thiir. Bot. Ges. Beih. 1: 7-108. 1949. [V. officinalis. ] 2. Valerianella Miller, Gard. Dict. Abr. ed. 4. 1754. Small annual or biennial, glabrous or pubescent herbs with dichasial branching, appearing dichotomous, but the terminal bud aborting in the lower branches. Basal leaves forming a rosette, petiolate to nearly sessile, undivided or dentate; cauline leaves connate, undivided, dentate [or rarely 224 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 incised-pinnatifid]. Inflorescence capitate, terminal, usually a dichasial cyme subtended by lanceolate to oblong, connate bracts. Flowers almost regular, bisexual [or unisexual] in the axils of lanceolate to oblong, con- nate bracteoles. Calyx obsolete [or forming a narrow, toothed rim]. Corol- la funnelform, narrowly campanulate or tubular, the tube slightly saccate at the base, 5-lobed. Stamens 3, inserted toward the top of the corolla tube, exserted, anthers 4-lobed. Stigma simple or shortly 3-lobed; ovary 3-locular, one locule fertile with a solitary ovule. Fruit glabrous or pubes- cent, 3-locular, two of the locules empty, the other 1-seeded. LectoTyPE species: Valeriana Locusta L. = Valerianella Locusta (L.) Betcke; see Britton & Brown, Illus. Fl. No. U.S. ed. 2. 3: 286. 1913. (Name a diminu- tive of Valeriana.) A genus of temperate parts of the Northern Hemisphere, including about 60 species in two sections; 19 in North America, ten in our area The shape, relative size of the sterile locules, and indumentum of the fruits are the most important characters used to distinguish the species. The indumentum of the plant and the shape of the bracts (rounded, pointed, or acuminate) are also important characters. Specimens with both fruit and flowers are necessary for the determination of species. Four species of sect. SPHONELLA Krok, marked by salverform corollas and strongly glandular fimbriate-serrulate bracts, occur in the southeastern United States. Valerianella Bushii Dyal and V. ozarkana Dyal are strictly calciphile and known only from Arkansas and Missouri, while V. Nuttallii (Torr. & Gray) Walp. and V. longiflora (Torr. & Gray) Walp. are known only from western Arkansas and eastern Oklahoma. The other six species of our area belong to sect. VALERIANELLA, with funnelform corolla and glabrous bracts with entire or ciliate margins. Valerianella Locusta (L.) Betcke (V. olitoria (L.) Poll., Valeriana Locusta L. a olitoria L.), 2n = 14, an introduced European species widely scattered throughout our area, is readily recognized by the corky mass on the fertile locule of the fruit. Valerianella umbilicata (Sulliv.) Wood and V. patellaria (Sulliv.) Wood occur from North Carolina and Tennessee, northward to New York, Ohio, and Illinois. They are distinguished from the other species in the section by the breadth of the sterile locules in comparison with the fertile locule of the fruit. Valerianella intermedia Dyal, which ranges from Massachu- setts to Kentucky and Illinois, occurs in scattered localities in North and South Carolina, apparently as an introduction. It is closely related to V. radiata (L.) Dufr., a widespread, polymorphic species occurring throughout our area and extending to Pennsylvania, Kansas, and Texas. Valerianella Palmeri Dyal, occurring only in Arkansas, appears to be closely related to V. intermedia but is separated by its longer fruit with abortive sterile locules. Somie authors have thrown doubt on the validity of some of the species recognized in North America. Dempster has found dimorphism in the fruits of species of the closely related Plectritis and has revised the taxono- 1965] FERGUSON, VALERIANACEAE AND DIPSACACEAE 225 my of that genus. In view of this, the species pairs Valerianella Bushii— V. ozarkana and V. umbilicata—V. patellaria need further investigation, for in each pair the two sometimes grow together, show a similar, limited geographical distribution, and are distinguished by their fruits. There is a lack of cytological information relating to North American species of Valerianella. Chromosome numbers of 2n = 14, 16, 18, 32 have been reported for European and Asiatic species. The method of pollination appears to be unknown; insect visitors have been observed and self-pollina- tion reported to occur. Valerianella has affinities with Valeriana. It is very closely related to Fedia and Plectritis, and early authors treated these variously as one or more genera. More recent authors have treated Fedia as a monotypic Mediterranean genus distinguished by its irregular corolla, two stamens, and dimorphic fruits. Valerianella is separated from the predominantly western North American Plectritis by the typically 3-locular ovary, 3- lobed stigma, and dichotomously branched inflorescence vs. the 1-locular ovary, 2-lobed stigma, and capitate or interruptedly spicate inflorescence. REFERENCES: Under family references see DemMpsTER, Poucgues, VESQUE, WAGENITZ (1956), and WEBERLING BARTHOLEMEW, E Three spermatophytes new to Virginia. Castanea 5: 111. 1940. 7 olitoria (= V. Locusta), V. umbilicata. Dyat, S.C. Valerianella in North America. Rhodora 40: 185-212. 1938. [See also tbid. 465-467. ] Etvers, I. Chromosomenzahlen in der Gattung Valerianella nebst einigen sys- tematischen Bemerkungen. Acta Horti Berg. 11: 81-87. 1932. FREEMAN, O. M. Notes on the flora of Polk County, North Fei Castanea 20: 37-57. 1955. [V. intermedia as a field and roadside wee GarJEANNE, A. J. M. Valerianella. (In Dutch.) Levende Nat. 53: 101-107. 1950.* Goopman, G. J. Notes on Oklahoma plants. Proc. Okla. Acad. Sci. 32: 39, 40. 1952. [V. ozarkana recorded in Oklahoma. Graves, C. B. Valerianella in New England. Rhodora 4: 195, 196. 1902. Gray, A. Contributions to North American botany. Proc. Am. Acad. Arts Sci. 19: 1-96. 1883. [Valerianaceae-V alerianella, 81-83.] Krox, T. O. B. N. Anteckningar till en Monografi ofver Vaxlfamiljen pant aneae. 1. Valerianella. Vet.-akad. Stockholm Handl. IV. 5: 1-105. pls. 1- 1864. 3 Porter, T. C. The Fedias of the northern United States. Am. Nat. 6: 385-388. 1872. [Valerianella.] ; ee Paes genaged oe Shinners, sp. nov. Field Lab. 21: 93 953. [Described from T Xas. bai ae, R. Embryogéni “— Valérianacées. Développement de l’embryon chez le Valerianella ones Poll. Compt. Rend. Acad. Sci. Paris 176: 1081— 1083. 1923. [V. Locusta. ] Sover-WILLEMET, H. F. Essai monographiques sur les Valerianella de France. Précis Trav. Soc. Sci. Lett. Arts Nancy 1829-1832: 67-73, 288. 1833. 226 JOURNAL OF THE ARNOLD ARBORETUM [ VoL. 46 DIPSACACEAE A. L. de Jussieu, Gen. Pl. 194. 1789, “Dipsaceae,” nom. cons. (TEASEL FaMILy) Mostly perennial herbs, the flowers bisexual, irregular, with a calyx- like involucel of fused bracteoles, and borne in the axils of imbricate recep- tacular bracts, the calyx small [variable|, the corolla gamopetalous, 4[5|- lobed, the stamens 4 [5], inserted toward the top of the corolla tube, and the ovary inferior. TyPE GENUS: Dipsacus L. An Old World family of about eight to twelve genera and 250-300 species, centered in the Mediterranean and Near East, extending to North Europe, East Asia, and South Africa. About ten species in the genera Dipsacus, Cephalaria, Scabiosa, Knautia, and Succisa are naturalized in North America; one species of Dipsacus occurs in our area. Dipsacaceae are placed in the Rubiales by most authors and appear to be a very natural family, with the exception of the isolated genus Morina, which some authors have placed in a separate family. Dipsacaceae have close affinities with Valerianaceae. The family is distinguished from Valerianaceae by the unilocular ovary and by the capitate inflorescence surrounded by an involucre of bracts, characters which suggest affinities with Compositae. Chromosome numbers of 2” = 10, 14, 16, 18, 20, 34, 36, 38, 42, 43, 44, 46, 50, 54, and 64 have been reported. Ehrendorfer has suggested that the basic chromosome number of the family is x = 9 and that euploidy, aneuploidy on the diploid level, polyploidy followed by aneuploidy, gene-, chromosome-, and genome-mutations have been important in the evolution of Dipsacaceae. The variation in the receptacular bracts and the diverse development of the involucel and calyx are connected with fruit dispersal and give good characters for taxonomic division of the family into genera. e family is of little economic importance. A few genera are of horti- cultural interest, and Scabiosa is widely cultivated for its ornamental value. REFERENCES: ALVARA ADO, S. Constitucién morfolégica y filagenia del caliculo de las “dipsaca- ceas.” Trab. Mus. Ci. Nat. Bot. Madrid 21: 1-29. 1925. [Scabiosa, M wee = ——.. morphologische Aufbau des Hiillkelches der Dipsacaceen. Bot. =a “61(Beibl 138): 10-21. se BaILton, H. Dipsacacées. Hist. Pl. 7: 519-534. 1880. Baxsay, L. Monographie der Gattung a Ann, Hist.-Nat. Mus. Hungar. II. 2: 237-260. 1952. . Anatomische und systematische Untersuchungen iiber die Gattung Succisella. Ibid. 6: 167-176. 1955, BENTHAM, G., & J. D. Hooker. Dipsacaceae. Gen. Pl. 2: 157-161. 1873. CANDOLLE, A. Pp. DE. Dipsacaceae. Prodr. 4: 643-664. 1830. nie oN R. ‘ Scabiosa columbaria in central New York. Rhodora 45: 220, i ae 1965] FERGUSON, VALERIANACEAE AND DIPSACACEAE 227 Coutter, T. Mémoire sur les Dipsacées. Mém. Soc. Phys. Hist. Nat. Genéve 2(2): 13-60. pls. 1, 2. 1826. Dott, W. Beitrage zur Kénntiis der Dipsaceen und Dipsaceenahnlicher Pflan- zen. Bot. Arch. 17: 107-146. 1927. EHRENDORFER, F. Beitrige zur Phylogenie der Gattung Knautia ( Dipsacaceae ) | a Cytologische Grundlagen und allgemeine Hinweise. Osterr. Bot. Zeitschr. 109: 276-343. 1962. [An extensive study of the genus in Europe. | , ee Taxonomie und Evolution bei Samenpflanzen. Jn: W. B. TURR ed., Vistas in botany 4: 99-186. New York. 1964. [Darstellung enon: nein Probleme am Beispiel der Dipsacaceae, 121-124. Rela- tionships and evolution of the genera of Dipsacace FIscHeER, J. Beitrage zur Systematik der Dipsaceen. Lotos Sitz-ber. Naturw.- Med. Ver. Bohmen IT. 26: 78-102. 1906. Fopor, F. Beitrage zur Kenntnis der Histologie der Gattung Cephalaria. (In Hungarian: German summary.) Bot. Kézlem. 9: 171-199, (45)-(46). 1910. [For leaf anatomy see F. Varca, Bot. Kézlem. 22: 28-34, (17). 5.] 192 Francois, L. Semences et jeunes poy chez les Dipsacées et les Composées. Ann. Agron. III. 17: 674-718. 1947.* Gutntuart, A. Beitrige zur Bliitenbiologie der Dipsaceen. Flora 93: 199-250. 1904. [Scabiosa, Knautia, Succisa, Cephalaria, Dipsacus. Herr, C. Contribution a |’étude du pollen des Dipsacées. Thése Doct. Univ. (Pharmacie) Strasbourg. 1934.* Hock, F. Dipsacaceae. Nat. Pflanzenfam. IV. 4: 182-189. 1891. . Verwandtschaftsbeziehungen der Valerianaceen und Dipsacaceen. Bot. Jahrb. 31: 405-411. 1902. JAEGER, P. Nouveaux cas de gynodimorphisme chez les Dipsacées. Compt. Rend. Acad. Sci. Paris 199: 1239-1241. 1934. [Knautia, Scabiosa, Succisa, Cepha- laria, Dipsacus; for Succisa see P. Macnus, Bot. Jahrb. 3: 187. 1882.] Etude de la sexualité chez les Dipsacacées. Imprimerie Arts graphiques modernes. Nancy. 1937.* Kacuipze, N. Karyologische Studien iiber die Familie der Dipsacaceae. Planta 7: 482-502. 1929. : LaviaLLe, P. Sur le sac embryonnaire des Dipsacées. Compt. Rend. Acad. Sci. Paris 180: 1127-1129. 1925. [See also ibid. 1606-1608. | . L’étamine chez Knautia arvensis Coult. Polymorphisme des fleurs et des capitules. /bid. 192: 176-178. 1931. . La génération haploide female chez les Dipsacées. Bull. Soc. Bot. France 83: 274-277. 193 . L’androcée et la weubeation haploide male chez les Dipsacées. /bid. 562-570 & P. JaEceER. L’origine du fruit dans ses rapports avec la pollinisation chez Knautia arvensis Coult. Compt. Rend. Acad. Sci. Paris 192: 147 1476. 1 & ———. La fertilité et la stérilité de l’androcée: leurs rapports avec le polymorphisme staminal chez Knautia arvensis Coult. Ibid, 198: 114- 116. 1934. [See also cytology of pollen, ibid. 199: 485-487. 1934. | . Ktetn. Pistils bi-ovulés et soudure des deux ovules chez Knautia arvensis Coult. Bull. Soc. Bot. France 77: 593-597. 1930. Lunt, J. R. Succisa pratensis in Massachusetts. Rhodora 14: 174. 1912. Puitipson, W. R. Studies in the development of the inflorescence. II. The 228 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 capitula of Succisa pratensis oe and Dipsacus fullonum L. Ann. Bot. II. 11: 285-297. 1947. [D. sativ Prouvier, V. Sur la recherche du méthylglucoside B chez quelques Dipsacacées et Rhamnacées. Compt. Rend. Acad. Sci. Paris 256: 1397-1399. 1963. PoppUBNAJA-ARNOLDI, V. eon cin in der Familie der Dipsacaceae. Planta 21: 381-386. 1933. [Scabiosa.] Razr, B. A., & K. SuskAMANYAM. Embryology of the Dipsacaceae. Proc. Indian Acad. Sci. B. 36: 249-257. 1952. [Cephalaria. | RISsE, " Beitrige zur Zytologie der Dipsacaceen. Bot. Arch. 23: 266-288. 192 eee J. W., & M. W. Quimpy, eds. Lynn Index. Monograph V. 67 pp. Boston. 1962. [Dipsacaceae, 20-22. ScunarF, K. Vergleichende Embryologie der Angiospermen. 350 pp. Berlin. 1931. [Dipsacaceae, 209, 210.]| Sovéces, R. Embryogénie des Dipsacacées. Développement de |’embryon chez ns a columbaria L. Compt. Rend. Acad. Sci. Paris 245: 465-468. Embryogénie des Dipsacacées. Développement de l’embryon chez le Cephalaria tatarica Schrad. Ibid, 256: 45-48. 1963. [Knautia, ibid. 1190- 1194. SzaBo, Z. Nouvelles observations concernant Vhistologie et le développement des organes sur les espéces du genre Knautia. (In Hungarian; French sum- mary.) Bot. Kézlem. 9: 133-148, (25)-(41). 1910. . A Knautia génusz Monografidja. Mat. Természett. Kézlem. 31. 1911.* The development of the flower of the Dipsacaceae. Ann. Bot. 37: 325- 334. 1923. Organografiai és genetikai vizsgdlatok dipsacacedkon (Organographische und genetische Untersuchungen an “nea Magyar Tudom. Akad Ertek. Természett. Koréb. 53: 571-636. 193 ae, a a der Gattung Cephalaria. om Természett. Kézlem. 38. 1940 TAMMES, T. Notiz iiber das Vorkommen von Dipsacan bei den Dipsaceae. Rec. Trav. Bot. Néerl. 8: 369, 370. 1911. TiecHEM, P. van. Remarques sur les Dipsacacées. Ann. Sci. Nat. Bot. IX. 10: 148-200. 1909. Varca, F. Vergleichende anatomische Untersuchung der Gattungen Succisella und Succisa mit Rucksicht auf die verwandten Gattungen. (In Hungarian; German summary.) Bot. Kézlem. 21: 32-47, (4)—(8). 1923 VietH, J. Beitrag zur Kenntnis Dipsacaceenbliite. Ann. Univ. Sarav. Sci. 7: 215-274. 1959. WaceEni1z, G. Die systematische Stellung der Rubiaceae. Bot. Jahrb. 79: 17-35. 1959. [Gentianales = Contortae + cme Dipsacales = Caprifolia- ceae, Adoxaceae, Valerianaceae, Dipsacaceae. | Dipsacaceae. Jn: H. MELCHIOR, Engler’s Syllabus der Pflanzenfamilien. ed. 12. 2: 477, 478. 1964. 1. Dipsacus Linnaeus, Sp. Pl. 1: 97. 1753; Gen. Pl. ed. 5. 43. 1754. Stout, erect, biennial herbs with spiny or prickly stems. Leaves simple, entire [to pinnately divided], opposite, exstipulate; basal leaves petiolate, stem leaves connate [petiolate]. Inflorescence a head with 1 or 2 rows of > 1965] FERGUSON, VALERIANACEAE AND DIPSACACEAE 229 linear to lanceolate [narrowly triangular], subulate, erect or speading, spine-tipped involucral bracts; each flower subtended by a chaffy recep- tacular bract and surrounded by a 4-angled tubular involucel of fused bracteoles, which is grooved with a median rib and truncate at the apex, with an almost obsolete denticulate margin. Calyx cup-shaped, 4-angled, with a ciliate margin. Corolla imbricate, lobes often unequal. Stamens 4, alternate with the corolla lobes, exserted; filaments free; anthers 4-locular, introrse, longitudinally dehiscent; pollen tricolpate, echinate. Stigma lat- eral, entire; style filiform; ovary unilocular, ovule solitary, anatropous, pendulous from the apex of the locule. Fruit indehiscent, dry, inclosed within the calyx-like involucel and crowned with the persistent calyx. Seed 1; endosperm fleshy; embryo straight, the cotyledons oblong or ovate, the radicle superior. Embryo sac development of the normal (Polygonum) type. Lectotype species: D. fullonum L.; see Britton & Brown, Illus. Fl. No. U. S. ed. 2. 3: 289. 1913. (Name from Greek, dipsa, thirst, be- cause the united cup-shaped bases of the leaves in some species hold water.) — TEASEL. About 12 species of Europe, western Asia, and Africa, three naturalized in North America. Dipsacus fullonum L. (D. sylvestris Huds.) ,? 2n = 16, 18, a native of Europe, is probably widespread and scattered throughout the United States, occurring sporadically in our area. Dipsacus fullonum has close affinities with D. sativus (L.) Schkuhr (D. fullonum auct. non L.), 2n = 16, 18, also a native of Europe, locally naturalized in the north- eastern United States. The involucral bracts curve upward in D. fullonum and the receptacular bracts end in a long straight spine, while in D. sativus the involucral bracts spread more or less horizontally and the receptacular bracts end in a stiff, recurved spine. Some authors have treated D. sativus as a subspecies of D. fullonum. The very different laciniate-leaved D. laciniatus L. is established in the northeastern United The genus is mainly a well-defined one recognized by the presence of stem prickles and spine-tipped bracts. It is most closely related to Cepha- laria, which is also distinguished from the rest of the family by chaffy receptacular bracts and a four-angled involucel. Grenier and Godron placed D. pilosus L. in Cephalaria, but few subsequent authors have adopted this treatment. The mode of flowering in the genus is characteristic, the flowers opening first about halfway up the head and developing in sequence both upward and downward. The species are separated on the basis of the leaves, which may be con- nate or stalked, entire or divided; the inflorescence, which may be oblong, ovate, or globose; and the involucral bracts, which may be linear-lanceo- late, subulate, keeled with spines on the margins and midrib, or narrowly triangular, shorter or longer than the inflorescence. * The application of the name Dipsacus fullonum L. will be discussed in the next issue of this Journal. 230 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 The ripened inflorescences of D. sativus, the fuller’s teasel, are used in fulling cloth (raising the nap), and the plants are grown to a limited ex- tent for this purpose. The flora and fauna of the rain water caught by the connate leaf bases have been studied more or less extensively, but most biological aspects of the genus have been neglected. Chromosome numbers of 2” = 16, 18, 36 have been reported. REFERENCES: Under family references see GUNTHART, JAEGER (1934), and PHILIPSON. BEAL, W. J., & C. E. St. Jonn. A study of Silphium perfoliatum and Dipsacus laciniatus i in regard ie insects. Bot. Gaz. 12: 268-270. ms Curisty, M. The common teasel as a carnivorous plant. Jour. Bot. 61: 33-45. 1923. eeriariny « on whether the water held in ee connate leaf bases of D. fullonum operates as an insectivorous mechani GREENE, E. L. Nomenclature of the fuller’s teasel. PiHlonts 3: 1-9. 1896. Grenter, M., & M. Gopron. Dipsacaceae. Flore de France 2: 67-81. 1850. [D. iilosas transferred to Cephalaria, 69. | Grossin, F. Virescence spontanée chez Dipsacus silvestris Mill. et virescence induite expérimentalement chez Zinnia elegans L. Compt. Rend. Acad. Sci Paris 242: 1349-1352. 1956. [Experiment with 2,4-D on D. jullonum.J Howe, J. T. Dipsacus sylvestris Huds. Leafl. West. Bot. 2: 16. 1937. [Oc- currence of D. fullonum in western | Jurica, H. S$. Development of head and flower of Dipsacus sylvestris. Bot. Gaz. 71: 138-145, 1921. [D. fullonum.| Macurre, B. Aquatic biotas of teasel (Dipsacus sylvestris) waters. Ecology 40: 506. 1959. [Algae, protozoa, nematodes, and other organisms found in the water of the connate leaf bases of D. fullonum. | Mu tins, D. Teasel growing, an ancient ie World Crops 3: 146, 147. 1951+ [D. sativus as D. fullonum. | PLanTEFOL, L. La soudure des feuilles dans la torsion de contrainte de la cardére (Dipsacus pri Huds.). Compt. Rend. Acad. Sci. Paris 256: 3917-3921. 1963. [D. fullonum. . La torsion de cane. des tiges de cardére (Dipsacus silvestris — et la théorie des helices foliaires multiples. /bid. 3398-3403. fullonum. | ROBINSON, B. B. Minor fiber industries. Econ. Bot. 1: 47-56. 1947. [Includes Dipsacus.] erste I. Adatok a Dipsacus genus anatomiai es fejlodestani ismeretehez [Beitrige zur Anatomie und Entwicklungsgeschichte der Gattung Dipsacus |. Kiilén. Egyet. Természett. Szévet. Evkényv. 1910: 1-32. 1910.* [See German summary in Bot. Centralbl. 116: 482. 1911. Snow, R. Phyllotaxis of flowering teasels (Dipsacus laciniatus). New Phytol. 53: 99-107. 1954. Sovéces, R. Embryogénie des Dipsacacées. Développement de l’embryon chez le Dipsacus sylvestris Mill. Compt. Rend. Acad. Sci. Paris 256: 2268- 2273. 1963. [D. fullonum.| SPOEL-WaLvius, M. R. vAN ver, & R. J. pE Vries. Description of Dipsacus fullonum L. pollen. Acta Bot. Neerl. 13: 422-431. 1964. STEYERMARK, J. A. An albino form of Dipsacus sylvestris. Rhodora 60: 174, 175. 1958. [D. fullonum.| 1965] FERGUSON, VALERIANACEAE AND DIPSACACEAE 231 Toncrorc1, E. Un caso di virescenza del — silvestris Mill. Nuovo Gior. Bot. Ital. II. 45: 221-223. 1938. [D. fullon Varca, L. Recherches limnologiques sur la ihe des réservoirs de la entities (Dipsacus silvester Huds.). (In Hungarian; French summary.) Erdész. Kisér. 30: 353-369, 410. 1928. [D. fullonum. | VieTH, J. Persistance de morphoses induites chez Dipsacus fullonum. Bull. Sci. Bourgogne 20: 70-81. 1960. [D. sativus. | . Sur la feuille terminale présentée par Dipsacus silvestris. Ibid. 21: 39-44. 1962. [D. fullonum Voropyser, B. A. The inhabitation of larvae of Diptera in water bodies in axils of the leaves of Dipsacus. (In Russian.) Entomol. Obozr. 39: 799-801. 1960. Wartiez, N. Contribution 4 l’étude biochimique des dipsacées. Présence dans Dipsacus arvensis de méthylglucoside et de scabioside. Bull. Soc. Chim. Biol. 8: 501-507. 1926.* ARNOLD ARBORETUM AND Gray HERBARIUM 232 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 A PRELIMINARY REPORT ON FUNIFERA Lorin I. NEVLING, JR. THE EARLIEST RECORDED INFORMATION concerning the plants now recog- nized as members of the thymelaeaceous genus Funifera C. A. Mey. was published in 1820 (Atti Soc. Ital. Sci. Modena 18: 391), by the Italian botanist Raddi, In that publication, by means of a rather sophisticated and complete description, Raddi defined a new Brazilian species of Daphne, D. brasiliensis (to be transferred below to Funifera). In nearly a century and a half which has since elapsed, our knowledge of this and related species has not increased much; the chief contribution being descriptions of a few additional species. At the suggestion of Mr. Guido Pabst, Director of Herbarium Bradean- um, I started what was intended to be a comprehensive revision of the genus Funifera. However, the poor quality of most of the existing collec- tions (i.e., flowers and fruits lacking from most of the sheets) has frus- trated such a project. I hoped the difficulties would be at least partially overcome when I had a large suite of specimens assembled. Since, how- ever, the additional material was not very satisfactory I am presenting as a report the information now available, hoping it may serve as an out- line for field work and eventual new collections which will make possible a precise and detailed study. The species of the genus Funifera are not conspicuous components of the vegetation in the areas of Brazil where they grow; nor are they of great economic importance even at the local level. They are intriguing because so little is known of their geographic and ecologic distribution, their anato- my, and their morphology. In addition, the medicinal potential of these plants found in references from folk-medicine, has both practical and theoretical application in modern phytochemistry. Various properties attributed to plants of the genus Funifera are similar to those ascribed to other genera of the Thymelaeaceae. The tough cortical fibers of Funifera are employed as cordage (Penna, 1946) a common prac- tice in related genera of the family throughout the geographic distribution. The fibers of the American genera are never used in paper making, this use being restricted entirely to the Old World members of the family. A decoction of the macerated bark of Funifera is used externally as 4 vesicant (Dragendorff, p. 204. 1898) in the same manner as that of Daphne mezereum L. of Europe. If accidentally ingested the decoction is especially caustic to the mucous membrane of the mouth. Other American Thymelaeaceae known to have a similar vesicating effect are Daphnopsis (one species has the common name “Burn Nose”), Dirca, and Ovidia. A decoction of the leaves of Funifera may be used internally as an 1965] NEVLING, REPORT ON FUNIFERA 233 emetic or cathartic according to Hoehne (pp. 204, 205. 1939). Similar results may be obtained also from the New World genera Daphnopsis, Dirca, Lagetta, Lasiadenia, and Ovidia. Based on a very limited number of reports, in Daphnopsis, Dirca, and Ovidia, such a decoction, or the com- parable use of fruits, should be avoided as it presents more than a little danger to the patient (both violent illness and death have been reported). Hoehne (p. 206. 1920) says that Funifera is ‘““extremamente toxicas.” He further notes that cattle are adversely affected by grazing both Daphnopsis and Funifera and are stricken particularly from May to September (during the vegetative growth phase). An active principle has not been isolated but several compounds, mez- erein, daphnin, and umbelliferone, have been reported in several genera of the Thymelaeaceae. Hoehne (p. 206. 1920; p. 204. 1939) has reported both mezerein and daphnin present in Funifera. If these compounds, or chemically related ones, are more consistently present in the family than has been realized, they may provide the chemical basis for the similarity of application found in folk medicine throughout the world. The chemical composition of mezerein is unknown, most authorities referring it to the category of an “acrid resin.” Both umbelliferone and daphnin belong to a large group of related compounds having coumarin (for a most informa- tive paper on this compound see Brown, 1963) as a nucleus. Since the coumarin derivatives are found in many groups of plants their use as supplemental taxonomic characters is less important than might have been expected. Umbelliferone is known to absorb ultraviolet light and has been used as sunburn protection, and as a whitening agent in soap powders. Specific uses for daphnin have not been reported. Coumarin has been used commercially as an adulterant or substitute for vanilla, as an adulterant of tobacco, and in the perfume trade. It has a fragrant odor and burning taste. A coumarin derivative could account, in part, for the peculiar and distinctive odor of many thymelaeaceous plants. In general, coumarin is considered nontoxic to mammals although Pammel (p. 552) points out that “Dr. Millspaugh states that in large does, [it] causes nausea, vomit- ing, vertigo, great depression of the heart’s action and cold extremities.” Dean (p. 241) further states that . . . “larger animals (e.g., dogs, horses) can be killed by coumarin, but moderate quantities have no very marked effect on man except that it has been reported recently that coumarin has a true curare-like activity.” It is impossible to say whether a single compound or a group of compounds is responsible for the similarity of medicinal uses to which thymelaeaceous plants are put but the latter al- ternative seems more likely. It is clear that more precise diagnostic and physiological data are needed for all the compounds mention MATERIALS This study is based on specimens from the following herbaria, the ab- breviations for which are taken from Lanjouw & Stafleu’s Index Herbar- iorum, Part I. Ed. 5 (Regnum Vegetabile, 31. 1964). 234 JOURNAL OF THE ARNOLD ARBORETUM [voL. 46 A Arnold Arboretum of Harvard University, Cambridge, Massachusetts 1 pms British Museum (Natural History), London E Royal Botanic Garden, Edinburgh F Chicago Natural History Museum, Chicago FI Herbarium Universitatis Florentinae, Firenze c-pc Herbier DeCandolle, Conservatoire et Jardin botaniques, Genéve cH Gray Herbarium of Harvard University, Cambridge, Massachusetts 1 HB Herbarium Bradeanum, Rio de Janeiro K Royal Botanic Gardens, Kew a Rijksherbarium, Leiden LE Herbarium of the Komarov Botanical Institute of the Academy of Sci- ences of the U.S.S.R., Leningrad M Botanische Staatssammlung, Miinchen Mo. Missouri Botanical Garden, St. mes NY New York Botanical Garden, N P Muséum National d’Histoire Naturale Paris PI Istituto Botanico della Universita, Pis R Divisao de Botanica do aes Setacat Rio de Janeiro RB Jardim Botanico, Rio de Jan s Naturhistoriska oats “Stos kholm us U.S. National Museum (Department of Botany), Smithsonian Institution, Washington I wish to thank the directors and curators of the institutions listed for allowing me to examine the specimens in their care. Special thanks are given to Mr. Guido Pabst whose encouragement was responsible for my carrying the study to this stage. TAXONOMY Funifera C. A. Mey. Ann. Sci. Nat. Paris II. 20: 46, 47, 49. February, 1843 (Type species: Lagetta funifera Mart. & Zucc.); Bull. Aca St. Pétersb. Classe Physico-Math. 1: 355, 357. June, 1843. Neesia Mart. ex Meissn. Jn: Mart. Fl. Bras. 5: 67. 1855, non Blume, pro syn. Dioecious shrubs with flexible leathery branches, the stems often di- chotomously branched, the cortex containing many fibers. Leaves irreg- ularly pseudo-whorled, opposite, subopposite, or rarely alternate, simple, entire, pinnately veined, petiolate, estipulate. Inflorescences usually borne from the younger leafy or bracteate shoots, umbelliform or racemiform. Flowers unisexual, tetramerous, perigynous; calyx tube cylindrical or urceolate, the interior villous below and glabrous above; calyx lobes much shorter than the tube, in unequal pairs, generally erect at anthesis; petals 0; disc of 8 free lobes or the lobes connate and coronate. Staminate flowers: stamens 8, in two whorls, the upper 4 antisepalous, included to exserted, the lower 4 alternisepalous, included, the anthers longitudinally dehiscent, introrse, basifixed; pistillode present, densely villous. Pistillate *The abbreviation a-cu, used for citation of certain microfiches, refers to the collection in the combined library of the Arnold Arboretum and the Gray Herbarium, Harvard University. 1965 | NEVLING, REPORT ON FUNIFERA 235 flowers: staminodia 8 or 0, the upper whorl sometimes bearing aborted anthers; pistil 1, superior, the ovary uniloculate with 1 anatropous ovule, the style chiankcie. the stigma obscurely bilobed or capitate, included or exserted. Fruit a berry enclosed by the persistent and accrescent calyx ube KEY TO THE SPECIES a. Mature leaves 4 to 7 times longer than broad, narrowly ong to oblanceo- late, 4-14 cm. long, 1-2.5 cm. broad, densely sericeous ben b. Leaves irregularly pseudo- whorled, opposite, pie woneinig or rarely alter- nate, the apex acute and often minutely apiculate; staminate inflores- cen ces 5—6-flowered, the primary peduncle ca. 3 mm. long, stout; disc the staminate flowers with 8 free lobes; pistillate eon with 8 stam co! ee PCO meres) Pete were Pure LN er : ae prone ao . Leaves subopposite or alternate, the apex long acuminate; staminate inflorescences 10-35-flowered, the primary peduncle 10-35 mm.