D. d ES NOVEMBER 1. | Price Gs. FS THE JOURNAL Oa OF THE LINNEAN SOCIETY. j Vor. XXXVII. BOTANY. No. 257. "uu CONTENTS. Page I. List of the Carices of Malaya. By C. B. CLARKE, (007 cgo 7 Pete TUNE qe 1 II. Some Bicarpellary Beans. By Erro DrRABBLE, D.Sc., Peas CW GIES) E 2. nee see rne tr rh nn een 17 III. On the Species of Impatiens in the Wallichian Herbarium of the Linnean Society. By Sir J. D. Hooxer, G.C.S.I., F.R.S., F.L.S. ........................ 22 IV. On the Blaze-Currents of Vegetable Tissues: a Week’s Holiday with a Galvanometer and some Plants. By Dr. A. D. Warrer, F.R.S. (Communi- cated by Prof. J. B. Farmer, F.R.S., F.L.S.) (With 8 fiom.) "LRL apis acs inva ein rat nont hn hon n 32 See Notice on last page of Wrapper. LONDON: SOLD AT THE SOCIETY'S APARTMENTS, BURLINGTON HOUSE, PICCADILLY, W.; AND BY LONGMANS, GREEN, AND CO., AND WILLIAMS AND NORGATE. 1904. LINNEAN SOCIETY OF LONDON. LIST OF THE OFFICERS AND COUNCIL. Elected 24th May, 1904. PRESIDENT. Prof. W. A. Herdman, D.Sc., F.R.S. VICE-PRESIDENTS. Frank Crisp, LL.B., B.A. Albert O. Seward, F.R.S. Dr. A. Günther, F.R.S. Prof. S. H. Vines, F.R.S. TREASURER. Frank Crisp, LL.B., B.A. 1 SECRETARIES. E D. H. Scott, M.A., Ph.D., F.R S. | Rev. T. R. R. Stebbing, M.A., F.R.S. COUNCIL. V. H. Blackman, M.A. Dr. W. G. Ridewood, F.Z.S. Charles Baron Clarke, M.A., F.R.S. G. S. Saunders, F. E.S. . Frank Crisp, LL.B., B.A. Dukinfield H. Scott, M.A., Ph.D., Prof. J. B. Farmer, M.A., D.Sc., E.R.S. F.R.S. Dr. A. Günther, F.R.S. A. C. Seward, M.A., F.R.S. Prof. W. A. Herdman, F.R.S. Dr: Otto Stapf. E. Daydon Jackson, Esq. Rev. T. R. R. Stebbing, M.A., F.R.S. F. G. Parsons, F.Z.5. Prof. Sydney H. Vines, D.Sc., F.R.8. GENERAL SECRETARY. B. Daydon Jackson, Esq. . LIBRARIAN. CLERK. A. W. Kappel. A. R. Hammond. LIBRARY COMMITTEE, This consists of nine Fellows (three of whom retire annually) and of the officers ex officio; the former are elected annually by the Council in June, and serve till the succeeding Anniversary. The Committee meets at 4 p.m., as required during the Session. The Members for 1904-1905, in addition to the officers, are :— Herbert Druce, F.Z.8. A. B. Rendle, M.A., D.Sc. Antony Gepp, M.A. A. G. Tansley, M.A. W. Botting Hemsley, F.R.S. Prof. H. Marshall Ward, F.R.S. Dr. G. Henderson. Dr. A. Smith Woodward, F.R.S. P. Chalmers Mitchell, M.A., F.Z.S. THE JOURNAL OF THE LINNEAN SOCIETY. 30 TA N Y. VOL. XXXVII. LONDON: SOLD AT THE SOCIETY'S APARTMENTS, BURLINGTON HOUSE, PICCADILLY, W., AND BY LONGMANS, GREEN, AND CO., AND . WILLIAMS AND NORGATE. 1904-1906, Dates of Publieation of the several Numbers included in this Volume. No. 251, pp. l- 50, published November 1, 1904. » 258, » 51-115, » July 1, 1905. », 259, » 116-274, » September 30, 1905. » 200, » 275-882, " July 23, 1906. » 261, » 923-400, » October 18, 1906. » 203, „ 407-544, Index, Title-page, and Contents, published November 1, 1906. PRINTED BY TAYLOR AND FRANCIS, RED LION COURT, FLEET STREET, LIST OF PAPERS. Page Baker, EpMunp GILBERT, F.L.S., Spencer Le MARCHANT Moonz, F.L.S., and ALFRED Barton ReNDLE, D.Sc., F.L.S. The Botany of the Anglo-German Uganda Boundary Commission. (Plates1-4)......... see een n nn etna 116 Baker, Joun GILBERT, F.R.S., F.L.S. A Revised Classification of Roses, 1905 ...... 0.0 eee eee eee 70 Brown, A. F., Director of Woods and Forests in the Soudan. Some Notes on the “ Sudd "-Formation of the Upper Nile. (Communicated by C. H. Wright, A.L.S) .... esee 51 Brown, RosgnT N. RupMoss, B.Sc. The Botany of Gough Island. (Communicated by W. Botting Hemsley, F.R.S., F.L.S.) I, Phanerogams and Ferns. (Plates 7-9 and fig.). . .... 238 II. Cryptogams (excluding Ferns and Unicellular Algz). With C. H. Wrigut and O. V. DARBISHIRE ...... 263 CLARKE, CHARLES Baron, F.R.S., F.L.S. List of the Carices of Malaya ...... een n 6n 1 —— See STAPF, O. COTTON, ARTHUR DissnowE, F.L.S., Assistant, Herbarium, Royal Botanic Gardens, Kew. On some Endophytic Algæ. (Plate 12)....... sss 288 Page Darpisnire, Orro Vernon. See Brown, R.N. R. Dawe, Monrzy Tromas, Notes on the Vegetation of Buddu and the Western and Nile Provinces of the Uganda Protectorate. (Communicated by Dr. Otto Stapf, FLAS.) oe ccc eee eee ees 533 ——-— See STAPF, O. DRABBLR, Eric, D.Sc., F.L.S. Some Bicarpellary Beans (with 6 figures) ...........-..0005. 17 FARMER, Jour DngrLAND, F.R.S., F.L.S. See WALLER, A. D. Ginss, Miss Lintan Suzette, F.L.S. A Contribution to the Botany of Southern Rhodesia. (Plates 17-20) ........ DEMNM MM 495 GipsoN, Rosert J, Harvey, M.A., F.L.S. The Axillary Scales of Aquatie Monocotyledons. (Plates 5 & 6). 998 Groves, Henry, F.L.S., and James Groves, F.L.S. On Characeæ from the Cape Peninsula collected by Major A. H. Wolley-Dod, R.A. (Plate 1) ................s ee eee 985 Groves, JAMES, F.L.S. See Groves, HENRY. llanes, Henry HaskErroor, F.L,S, On two new Species of Populus from Darjeeling. (With figures) ee eee ec eee hai 407 TTavata, Bunzo. On Taiwania,a new Genus of Conifers from the Island of Formosa, (Plate 16.) (Communicated by Maxwell T. Masters, M.D., F.R.S, F.L.S.) oonan anana EEEE 330 HEMSLEY, WILLIAM Borrina, F.R.S., F.L.S. See Brown, R.N. R. HooxER, Sir Josepa Darton, G.C.S.L, F.R.S., F.L.S. On the Species of Impatiens in the Wallichian Herbarium of the Linnean Society ........ TP MS 22 Ik RANZLIN, Fivz, Ph.D. Cyrtaudracesie Malayaneæ insularis nove. (Communicated by Dr. Otto Stapf, ELS) i. sisi 275 Masters, MaxwELL Typex, M.D., F.R.S., F.L.3. Notes on the Genus Widdringtonia ........... esse. ees 267 A Correction of Widdringtonia equisetiformis to Callitris robusta a.nns ce cee cc ee eee eee eee a ue th een 332 On the Conifers of China ....... llle 410 —-— See Hayara, B. MoonxE, Spencer Le Marcuant, F.L.5. A Second Contribution to the Flora of Africa.—Rubiacew and Composite. II. (Plates 13-15) ............00. 02200000. 208 —— See BAKER, E. G., and others. PRAIN, Davin, Lieut.-Col., I.M.S., F.R.S., F.L.S. Mansonie&, a new Tribe of the Natural Order Sterculiacex. (Plate 10). . ., .. aeaaeae HMM 250 RENDLE, ALFRED Barton, D.Sc., F.L.S. See BAKER, E. G., and others. RoLFE, ROBERT ALLEN, A.L.S. See STAPF, O. , ) ) SPRAGUE, THOMAS ARCHIBALD, B.Sc., F.L.S. See STAPF, O. Srapr, Orro, Ph.D., F.L.S. Contributicns to the Flora of Liberia ................ sese 79 —— Plante Nove Daweane in Uganda lecte. In collaboration with T. A. SPnAGUvE, B.Sc, F.L.S.,, R. A. ROLFE, A.L.S., C. B. CLARKE, M.A., F.R.S., F.L.S., M. T. Dawn, and C. H. Wmnianr, A.L.S.; and with an Appendix by M. T. Dawe. (Plates 21-22, and Map) .......... eese 495 Turcuer, WILLIAM JAMES, P.L.S. Descriptions of some New Species, and Notes on other Chinese Plants oo eee n Hasen 58 WALLER, AUGUSTUS DESIRÉ, M.D., F.R.S. On the Blaze-Currents of Vegetable Tissues: a Week's holiday with a Galvanometer and some Plants. (Communicated by Prof. J. B. Farmer, F.R.S., F.L.S.) (8 figures)... 32 vi Page WorLEv-Dop, Major ANTHONY Ilurt. See Groves, HENRY. WoopneapD, THomas WiiiraM, F.L.S. Ecology of Woodland Plants in the Neighbourhood of Hudders- field. (With 70 figures) 0.2... .. 6 cece eee n nn 333 WRIGHT, CHARLES Henry, A.LS. See Broun, A. F. —— See Brown, R. N. R. LIST OF THE ‘PLATES, 1. TRIMERIA MACROPIIYLLA, Baker f. STYASASIA AFRICANA, S. Moore. EnvrTuüRococca Paxu, Rendle. Hemantnus RApcLIFFEI, Rendle. oor WD ‘. | AXILLARY SCALES of AQUATIC MONOCOTYLEDONS. . Gover IsLAND. VEGETATION on Govan ISLAND. . QOorvLA coucusnsis, R. N. R. Dr. 10. Mansonia Gacr, J. R. Drumm. 11. Cnuana ranYGLocuis, H. & J. Groves. 12. Enpopuytic ALGE. 13. POLYSPHÆRIA ZOMBENSIS, S. Moore. ls Bempycopium ATüANASLE, Kunze. B. SPHENOGYNE BRACHYLOBA, Kunze. 15. DICOMA RADIATA, Less. 16. TarwaNiA CRYPTOMERIOIDES, Hayata. 17. VEGETATION at VicroRiA FALLS. oon 19. | Vegetation of Maroro HrLrs. 21. WanRBURGIA UGANDENSIS, Sprague. 22. BarsauxociTRUS Dawzi, Stapf. 23. Mar to illustrate Mn. Dawz's Journey. ERRATA. Page 20, line 50, for Hausemannia read Hansemannia. » 90, „ 5 from bottom, for Juspia read Jurpia. » 60, „ 28, for alternata read attenuata, » 72, , 6 from bottom, for WATSONINIANA read W ATSONIANA. » "4, „ 8 from bottom, for piscicarpa read pisocarpa. , 81, , 21, for Tetraceras read Tetracera. 191, ,, 15, for SorunEA read Sorun1A. » 192, , 9 (the same correction). „ 243, lines 8 & 9, for coronifolia read coronopifolia, 275, line 14, for parvifolia read parviflora, » 292, ,, 31, for Jardinianwm read Jadinianum. 295, ,, 8, for lacertum read /aceratum. 296, lines 18 & 14, for parasiticus, Sauv., read parasiticum, De Toni. „ 826, line 21, for AuTUNESII read ANTUNESH. Plate 9, headline, for Bruce read Brown, - THE JOURNAL OF THE LINNEAN SOCIETY. | / List of the CanrcEs of Malaya. By C. B. Cr&fxz, F.L.S. (Read 3rd March, 1904.] Tuis List comprises all the species of Carex known to me from Malaya, with short diagnoses of those considered new. By “ Malaya ” is understood the Archipelago from Sumatra to the Philippines and New Guinea, together with the Malay and Tonkin Peninsulas. The material used is that in the Kew Herbarium. Dr. A. Zahlbruckner, of the Vienna Hofmuseum, to whom I was introduced by Dr. O. Stapf, kindly sent for my examination several types of Zollinger, of great value for estab- lishing the names of Zollinger and Miquel. I regret that I have not been able to make any use of the fine collection in the British Museum. To avail myself of these, I must have trans- ported the whole of my MSS. thither ; and it would have taken me more time than I can give to work out the specific names of the critical species from my own technical descriptions. Further, Ido not think it would have strengthened this paper to have added such determinations; I should not have felt sure enough of the exact agreement of the plants. It would of course have been different if I could have laid the plants side by side, when the identification would have been more certain as well as more speedy. The Malay Carex consists (here) of 54 species; of which 36, LINN. JOURN.—BOTANY, VOL. XXXVII. v 2 MR, C. B. CLARKE ON including all the new ones (11), belong to one subgenus, Cari- candra. This subgenus includes, in the World, 125 species ; of these, 19 recede from the usual habit of the subgenus by the stem having only 1-5 spikes; these 19 show affinities to species in various other subgenera of Carex. The main, typical, Cari- candre, 106 in number, with numerous spikes, sometimes called the Section “ Indice,” form a natural group of species, very homogeneous, and therefore very difficult to separate into species. They are an essentially Tropical group, of which 77 occur in South-east Asia, 16 in Tropical Africa, 13 in Tropical America ; there is no Carex in Palearctica or Nearctica that can be com- pared with them ; the culm bears numerous (10-200) spikes. In the Flora of British India, the group Indice contains 52 species out of a total of 142 Carew in India. That the Malay Flora should possess 36 species of Indice, out of a total of 54 Carex, is therefore what might have been expected. The group Indies, however, becomes much scarcer in China, and only a few reach Japan, where other sections of Carex abound. The Central American and Tropical African species of the group Indies are so closely allied structurally to the South-east ‘Asiatic, that I have not found it possible to give absolute diag- noses to distinguish them ; thus Boott made a Madagascar plant conspecific with Carew bengalensis, Roxb. At the same time, I do not find that the closely-allied Carea of this group from different continents match; I have therefore treated these species geographically. Several of the species described below as new were catalogued in Dr. Stapf's Kinabalu plants (Trans. Linn. Soc. ser. 2, Bot. vol. iv. [1893] p. 246) by me under old names : the corrections in the present paper are attempts to define geo graphic subspecies as species; that is to say, the Kinabalu Carex (group Indies) do not exactly match the closely-allied Indian. The group Hemiscapose, as to the plants typieally belonging to it, appears very plainly defined; there are at the very base and on the sterile tufts long, well-developed leaves; the stems have the peduncle-holding bracts short, and the stem-leaves, if any, are similar to and hardly longer than the bracts. Many species are thus definitely hemi-seapose. But the character in many other species is less marked, and we have finally a large number of species which may be put in some other group alto- gether or may be attached to the Hemiscapose. Dr. Theodore Holm believes that he is always able to locate the species in its THE CARICES OF MALAYA. 3 true affinity when he is able to cultivate it for a few seasons; but this does not help the arrangement of Carex in a large herbarium. I have run out this list short, and have eited (mainly) new synonyms only, not those which have been cited before. One ofthe most curious is Carex glaucescens, Elliott, a plant of the South-east United States, which belongs to a section that always has a trifid style. Boott, however, says (Carex, pp. 89°& 90) that it has occasionally a bifid style. The American authors also say it has a trifid or bifid style. I found the example with bifid style in herb. Boott, collected at the Isle of Pines, with a note by Boott that the Isle of Pines is close to Cuba. But this sheet was colleeted by Milne in the Isle of Pines near New Caledonia, and is totally unlike the American C. glaucescens, Elliott. In C. glaucescens the style is always trifid, in C. phacodes, Spreng., it is always bifid. CoNsPECTUS of the MALAY CAREX. Subgenus I. VIGNEANDRA. Style bifid. Spikes (many of them) female at base, male at top. Sect. A. BnEvIsPICX. Spikes less than 4 in. long. l. C. Thomsoni. 2. C. nubigena. Sect. B. LowarsPrcz. Spikes linear-lanceolate, an inch long or more. 3. C. brunnea. 4. C. Graeffeana. Subgenus II. VIGNEGYNE. Style bifid. Spikes (many of them) male at base, female at top. Sect. 1. Remora. Spikes sessile, less than 2 in. long. 5. C. alta. 6. C. remota. Sect. 2. Cernuz. Spikes peduncled, 1-14 in. long. 7. C. cernua. Subgenus III. EUVIGNEA. Style bifid. Terminal spike wholly male. a. Female glumes muticous. 8. C. Gaudichaudiana. b. Female glumes aristellate. 9. C. phacodes. 10. C. pruinosa. Subgenus 1V. CARICIMIA. Style trifid. Stem with one spike only. ll. C. rara. 12. C. capillacea. n2 4 MR. C. B. CLARKE ON Subgenus V. CARICANDRA. Style trifid. Terminal spike (and often many others) male at top, female at base (no spikes male at base, female at top). Sect. A. HEMIscAPOsE. Flowering stems scapiform, with short (or no) leaves; the basal leaves, and those on sterile tufts, often very long. 13. C. scaposa. 14. C. eryptostachys. Sect. B. Ponystacuex. Stems not scapiform, with 10-200 spikes. * Brevispicz. Spikes short, rarely up to 10-15 mm. long. + Stramentitie. Spikes straw-coloured. 15. C. indica. 16. C. Dietrichie. 17. C. stra- mentitia. 18. C. repanda. 19. C. perakensis. 20. C. malaccensis. Tt Cruciate. Spikes not straw-coloured. 21. C. cruciata. 22. C. gembolensis. 23. C. Raffle- siana. 24. C. scaberrima. 25. C. fuirenoides. 26. C. Horsfieldit. 27. C. Balansai. 28. C. Cumingii. ttt Filicine. Spikes not straw-coloured, rather slenderer than in Tt Cruciate. 29. O. filicina. 30. C. continua. 31. C. spatiosa. 32. C. neo-guinensis. 33. C. saturata. 94. C. rhizomatosa. ** Longispice. Spikes linear-cylindrie, more than 2 em. long. a. Peduncles (some) 2 or more from one bract-sheath. 35. C.arridens. 36. C. hypsophila. 37. C. Havi- landi. 38. C. turrita. 39. C. sumatrensis. b. Peduncle never more than one from each bract- sheath. 40. C. baccans. 41. C. curvirostris. | 42. C. ton- kinensis. 43. C. composita. Sect. C. OnrcosrAcuy;. Stems not scapiform, with 1-5 (rarely 8) spikes. a. Psilosacce. Utricles glabrous. 44. C. borneensis. 45. C. speciosa. f. Trichosacee. Utricles hairy. 46. C. Loheri. 47. C. madoerensis. 48. C. longibracteata. THE CARICES OF MALAYA. 5 Subgenus VI. PROPRIA. Style trifid. Terminal spike wholly male. a. Utricles glabrous. 49. C. olivacea. 50. C. Jackiana. 51. C. tumida. 52. C. Pseudocyperus. 5. Utricles hairy. 53. breviculmis. 54. C. Wallichiana. Subgenus I. VrGNEANDRA. 1. Carex Tuomsont, Boott, Carex, i. (1858) p. 1. t. 1; Franch. én Nouv. Arch. Mus. Par. sér. 3, viii. (1896) pp. 212, 239 (Thompsoni). Tonkin; Balansa n. 2820. Distrib. North-east India. Ichang. 2. CanEX NUBIGENA, D. Don, in Trans. Linn. Soc. xiv. (1825) p. 326. Distrib. From Cabul to Ceylon, Yunnan and Japan. Var. (3. fallax ; stricta, culmo 4-6 dm. longo; foliis in sicco arcte convolutis; inflorescentiá 12 mm. longá, 8-10 mm. lata, pyramidali, densi; bracteá inferiore 3-4 cm. longá, suberecta ; utriculo in rostrum subito angustato, in parte superiore interdum a glandulis consperso ; rostro lineari, basi lineari-conico, in mar- ginibus scabro subserrulato. C. fallax (sp.), Steud. in Zoll. Verz. Ind. Archip. ii. p. 60, Cyp. p. 189; Maiq. Fl. Nederl. Ind. ii. p. 847; Boeck. in Linnea, xxxix. (1875) p. 57. C. pterolepta, Franch. in Nouv. Arch, Mus. Par. sér. 3, viii. (1896) pp. 215, 239. Java; alt. 1750 metr. Zollinger n. 2554 fide Miquel; sine habitat, Kurz (in hb. Kew). Distrib. Yunnan. 3. CAREX BRUNNEA, Thunb. Fl. Jap. (1784) p. 38; Drake, Fl. Ins. Pacif. p. 336; Hook. f. in Trimen, Fl. Ceylon, v. p. 108; Franch. in Nouv. Arch. Mus. sér. 3, viii. (1896) pp. 241, 259. Celebes ; alt. 2500-3000 meir., Everett n. 38. Distrib. Masearenia, India, China, Japan, Australia, Poly- nesia to the Sandwich Islands ; a common plant. 4. CAREX GRAEFFEANA, Boeck.! in Flora, lviii. (1875) p. 123. Ins. Philippine; Benguet, Loher n. 699. Distrib. Samoa. 6 MR. C. B. CLARKE ON Subgenus IJ. Vrenrcyne. 5. Carex ALTA, Boott, in Proc. Linn. Soc. i. (1845) p. 254; Carex, p. 59, t. 153. C. Brizopyrum, Kwnze, Suppl. Schkuhr, p. 169, t. 43. Java; Horsfield n. 10, Forbes nn. 1110, 1120. Distrib. North-east India; and a var. (Rochebruni, Franchet) in Japan. 6. Carex REMOTA, Linn. Aman. Acad. iv. (1759) p. 298; Sp. Pl. ed. 2, p. 1383. Common in the cool North Temperate Zone of the Old World, extending to Sitka and Japan. Var. Rochebruni, C. B. Clarke, in Hook. f. Fl. Brit. Ind. vi. p. 706 (syn. Franch. et Savat. excl.). Java; Zollinger n. 3192. Distrib. Abundant in the Himalaya; also in China. 7. CAREX CERNUA, Boott, Carex, iv. (1867) p. 171, t. 578, non Franchet. C. lobolepis, F. Muell.! Fragm. Phyt. Austral. viii. p. 258. C. phacota, Franch, in Nouv. Arch. Mus. Par. sér. 3, ix. (1897) pp. 117, 158, non. Spreng. Tonquin; Balansa nn. 221, 2826. Distrib. North-east India, China, Japan. New South Wales. Subgenus III. EuviGNEA. 8. Carex GavunricHaUDIANA, Kunth, Enum. Pl. i. (1837) p. 417 ; Hook. f. Fl. Tasm. ii. p. 99, t. 151 a. C. vulgaris, var. Gaudichaudiana, Boott, Carex, p. 169 partim. C. esespitosa, R. Brown, Prodr. p. 242. C. contracta, F. Muell. Fragm. Phyt. Austral. viii. p. 258. Neo-Guinea ; Guilianetti et English. Distrib. Australia. New Zealand. 9. CAREX PHACODES, Spreng. Syst. ii. (1826) p. 826, errore typogr. phacota; Boott, Carex, i. (1858) p. 683, t. 168; Franch. in Nouv. Arch. Mus. Par. sér. 3, ix. (1897) p. 117 pro parva parte; Hook. f. in Trimen, Flora Ceylon, v. p. 104. C. lenticularis, D. Don, in Trans. Linn. Soc, xiv. (1825) p. 331, non Michaux. C. platycarpa, Steud. Cyp, p. 214. THE CARICES OF MALAYA. 7 C. rubro-brunea, Franch.! in Nouv. Arch, Mus. Par. sér. 3, ix. (1897) pp. 123, 159, non C. B. Clarke. C. glaucescens, Boott! Carex, p. 89, quoad exempla stylo 2-fido. (The name “ phacodes” is a translation merely by Sprengel of * Jenticularis.") Java; Horsfield, H. O. Forbes n. 426. Distrib. Cape of Good Hope. India. China. Japan. Poly- nesia. 10. Carex PRUINOSA, Boott, in Proc. Linn. Soc. i. (1845) p. 255; Carez, p. 65, t. 174. C. pruinosa, var. B. picta, Boott, Carex, p. 198; Franch. in Nouv. Arch. Mus. Par, sér. 3, ix. (1897) pp. 155, 162. C. picta, Boott, in Mem. Amer. Acad. n. s. vi. (1859) p. 418. C. dimorpholepis, Steud, Cyp. p. 214. C. Maximowiezii, Mig. Ann. Mus. Lugd.- Bat. ii. p. 150. Java; Horsfield. Distrib. Khasia. China. Japan. Subgenus IV. CARICIMIA. 11. CAREX RaRA, Boott, in Proc. Linn. Soc. i. (1845) p. 284, Carex, p. 44, t. 109; Hook. f. in Trimen, Fl. Ceylon, v. p. 105; Stapf, in Trans. Linn. Soe. ser. 2, Bot. iv. (1894) p. 246. Borneo; Kinabalu, alt. 3500 metr., Haviland n. 1393. Distrib. Ceylon. North-east India. 12. CAREX CAPILLACEA, Boott, Carex, i. (1858) p. 44 t. 110; Benth. Fl. Austral. vii. pp. 486, 437 ; Franch. in Nouv. Arch. Mus. Par. sér. 8, viii. (1896) pp. 197, 209 (synn. C. nana, C. onta- kensi excl.); Meinsh. in Act. Hort. Petrop. xvii. (1901) pp. 286, 316 partim. C. capitellata, Boiss. FI. Orient. v. p. 399. C. simplicissima, F. Muell. Fragm. Phyt. Austral. ix. p. 191. Ins. Philippine; Benguet, Loher n. 705. Distrib. South-east Asia, extending to Lazistan, Sachalin, and New South Wales. Subgenus V. CARICANDRA. 13. Carex scaposa, Hook. f. Bot. Mag. cxiii. (1887) t. 6940 Franch. in Nouv. Arch. Mus. Par. sér. 3, viii. (1896) pp. 255, 260. Distrib. China. 8 MR. C. B. CLARKE ON Var. B. baviensis, Franch. ! in Nouv. Arch. Mus. Par. sér. 3, viii. (1896) p. 255. Tonkin ; Mons Bavi, alt. 800 metr., Balansa n. 2815. 14. CAREX CRYPTOSTACHYS, Brongn. in Duperrey, Voy. Coquille, (1829) p. 152, t. 25; C. cyrtostachys, C. B. Clarke, in Hook. f. Fl. Brit. Ind. vi. p. 714, errore typogr. Penang; Wallich n. 3383. Perak; alt. 1000 metr., hb. King n. 8517. Singapore; Ridley n. 1720. Tonkin; Balansa nn. 2822, 2823. Java; Zollinger, fide Miquel. Distrib. Also in Hongkong and Waigiou. 15. Carex INDICA, Linn. Mant. alt. (1771) p. 574; nee Kunth nec Nees. C. Moritzii, Steud.! in Zoll. Verz. Ind. Archip. Heft 2, p. 60, Cyp. p.207; Mig. Fl. Nederl. Ind. iii. p. 350; J. Schmidt, in Bot, Tidsskr. xxiv. (1901) p. 38. Trang; Kunstler n. 1883. Penang; alt. 250 metr., hb. King n. 1496. Kedah; alt. 750 metr., Ridley n. 5147. Java; Zollinger n. 318. Cochinchina; Pierre n. 1881. Var. B. lete-brunnea, C. B. Clarke, in Hook. f. Fl. Brit. Ind. vi. p. 715, in Journ. Linn. Soc., Bot. xxxiv. (1898) p. 115. C. indica, Boott, Carex, ii. (1860) p. 87, t. 251; Boeck. in Linnea, xl. (1876) p. 347 partim ; Hook. f. in Trimen, Fl. Ceylon, vi. pp. 102, 108. C. bengalensis, Thwaites! Enum. Pl. Zeyl. p. 355 partim. C. longi-aristata, Kurz, in Journ. Asiat. Soc. Beng. xlv. pars 2 (1876) p. 160. C. Thwaitesii, Boott ! MS. C. fissilis, Boott ! MS. partim. Pahang; Ridley n. 2145. Johore; Ridley n. 4099. Distrib. Ceylon. Tenasserim. 16. Carex Dretricurx, Boeck.! in Flora, lviii. (1875) p. 122. C. indica, Linn., var. B., Boott, Carex, ii. (1860) p. 87, t. 254. C. indica, Linn., var Milnei, C. B. Clarke, in Hook. f. Fl. Brit. Ind. vi. p. 715. C. fuirenoides, Boott, Carex, ii. (1860) p. 156, non Gaudich. C. Milnei, Boott! MS. C. dispar, Boott! MS. Pahang; Ridley n.21434. Borneo; Beccari n. 2741. North Borneo; Creagh. Distrib. Queensland. Polynesia. THE CARICES OF MALAYA. 9 17. CAREX STRAMENTITIA, Boeck. in Linnea, xl. (1876) p. 351. Tonkin ; Balansa n. 2819. Distrib. Assam. Chota Nagpore. Himalaya, west to Nepaul. 18. Carex nEPANDA, C. B. Clarke, in Hook. f. Fl. Brit. Ind. vi. (1894) p. 720. Khasia. Var. (. émplumis ; paniculis partialibus pluribus evolutis, apertis, compositis, polystachyis. There are here no ripe examples of this plant, which may be a new species. Perak; alt. 75 metr., Wray n. 1982. Distrib. Also in Yunnan. 19. Carex PERAKENSIS, C. B. Clarke, in Hook. f. Fl. Brit. Ind. vi. (1894) p. 720. Perak; Wray. Selangor; Ridley. 20. Carex MALACCENSIS, C. B. Clarke, in Hook. f. Fl. Brit. Ind. vi. (1894) p. 722. Langkawi; Ridley n. 1669. 21. Carex CRUCIATA, Wahlenb. in Vet.- Akad. Stockh. Handi. xxiv. (1803) p. 149, non Nees; Franch. in Nouv. Arch. Mus. Par. sér. 3, viii. (1896) pp. 253, 260. C. bengalensis, Roxb. Fl. Ind. iii. p. 572; Benth. Fl. Hongk. p. 401. C. valida, Nees, in Wight Contrib. p. 123. C. indica, Munro, in Seem. Voy. * Herald,’ p. 423, non Linn. C. indica, “ forma altera," Hook. f.! in Trimen, Fl. Ceylon, vi. p. 109 in Obs. C. vacua, Boeck. in Linnea, xl. (1876) p. 343 partim ; Boott! MS. C. Bruceana, Boott! MS. ; cf. Boott, Carez, ii. (1860) p. 85. C. stramentitia, Franch. in Bull. Soc. Philom. sér. 8, vii. (1895) p. 33, fide Franchet, non Boott. C. condensata, Franch. in Nouv. Arch. Mus. Par. sér. 3, vii. (1896) p. 252; Boott! MS. C. canaliculata, Boott MS. Perak ; alt. 150 metr., hb. King n. 1906. Distrib. India (common). Madagascar. Formosa. Var. f). argocarpus, C. B. Clarke, in Hook. f. Fl. Brit. Ind. vi. p. 716. C. benghalensis, Boott,Carex,ii. (1860) p. 85, pro magná parte tt. 240-242, 10 MR. C. B. CLARKE ON Carex vacua, Boott MS. partim. C. condensata, Boott MS. partim. Tonkin; Balansa nn. 2816, 2817. Distrib. Assam to Nepaul, common. 22. CAREX GEMBOLENSIS, sp. nova; utriculis 2-23 mm. longis, copiose (stepe usque ad basin) hispido-pilosis, enervatis aut fere enervatis ; ezeteroquin ut C. filicina, Nees. C. Rafllesiana ?, Joott ! MS. Panicule partiales 5-6, pyramidales ; pedunculi usque ad 10- 14 em. longi, interdum geminati. Spiee 8-10 mm. longs, basi 2-5-nucigere, plures apice masculo. Utriculi ellipsoidei trigoni, paullo eurvati, in rostrum lineari-conieum angustati. Java; prov. Modjokerto, Mt. Gembolo 4300', Zollinger, ser. 2, n. 7. Var. B. timorensis; inflorescentiá debiliore, laxá ; paniculis partialibus 5 em. longis, 3 em. latis, admodum laxis. Timor; alt. 1000 metr., F. Newton. 23. Carex RAFFLESIANA, Boott! in Trans. Linn. Soc. xx. (1846) p. 132, Carex, i. (1858) p. 12, t. 33, var. B incl. Java; Horsfield; Mt. Tidjeng 2000', Zollinger, Forbes n. 785. Celebes ; Minahassa, Koorders nn. 16674 6, 16669 £. Var. B. tenuior; magis attenuata, foliis angustioribus, paniculis partialibus lineari-lanceolatis ; utriculis fere glabratis aut in rostro parcius pilosis. C. virgata, Mig. Fl. Nederl. Ind. iii. (1856-9) p. 351, neque Hook. f. (1853). C. bengalensis, var. 8. virgata, Boeck.! in Linnea, xl. (1876) p. 347. C. vesiculosa ? Boott! MS. Java; Mt. Tidjen, alt. 1750 metr. Zollinger. Sumatra; alt. 2750 metr., Forbes n. 2389. Ins. Molucca; Ternate, alt. 1000 metr., Moseley. 24. CAREX SCABERRIMA, sp. nova; inflorescentià 45 cm. longa, 4 cm. latá ; panieulis partialibus usque ad 8 cm. longis, 15 mm. latis, compositis, ramis ramulisque erectis hispide pilosis ; spi- carum fasciculis pernumerosis, congestis; spicis innumerosis, parvis. C. bengalensis, var. y. scaberrima, Boeck, ! in Linnea, xl. (1876) p. 347. Peduneuli usque ad 10-15 em. longi, interdum geminati. Luzon; Albany, Cuming n. 936. THE CARICES OF MALAYA. 11 25. CAREX FUIRENOIDES, Gaudich.! in Freycinet, Voy. (1826) p. 412; Kunth, Enum. Pl. ii. p. 508 ; non Boott, Carez, iv. (1867) t. 507. C. fibrata, Vidal! Phanerog. Cuming Philipp. p. 67, Pl. Vasc, Filip. p. 286; Boott! MS. C. indica, var. bengalensis, F. Muell.! Fragm. Phytogr. Austral. viii. p 2949. | C. fissilis, Benth.! Fl, Austral. vii. p. 441, non Boott. Ins. Philippine ; Cebu, Cuming n. 1764; Merrill n. 109. Distrib. Queensland. Ins. Marianne. 96. Carex HonsrrELDII, Boott! in Proc. Linn. Soc. i. (1845) . 257, Carex, i. (1858) p. 11, t. 32; Mig. Fl. Nederl. Ind. iii. p. 949. Java; Horsfield. ge 27. Carex BALANSAIL Franch.! in Nouv. Arch. Mus. Par. sér. 3, viii. (1896) pp. 258, 260. Tonkin; Mt. Raui, Balansa n. 2828. 28. Carex CUMINGII, Vidal! Phanerog. Cuming (1885) p. 156, Pl. Vase. Filip. p. 286, non Boott. Robusta. Folia usque ad 16 mm. lata. Inflorescentia 35 em. longa, stricta; panicule partiales 4, remote, 5-6 cm. longe, oblongo-lineares. Spice 7-10 mm. longs, ovoidem. Utriculi (rostro incluso) 3-32 mm. longi, ovales, trigoni, multinervati, pilosi, rostrum eum 4 parte utriculi vix equilongum. Luzon; Albany, Cuming n. 1408. Luzon Central; Loher nn. 704, 712. 29. CAREX Fitictna, Nees! in Wight, Contrib. (1834) p. 123; Boott, Carex, iii. (1862) p. 105 (varr. a, y), tt. 311, 312; Franch. in Nouv. Arch. Mus. Par. sér. 3, viii. (1896) pp. 254, 260; Hook. J. in Trimen, Fl. Ceylon, v. pp. 102, 110. C. cruciata, Thwaites, Enum. Pl. Zeyl. p. 355 partim. Java; Horsfield n. 29; Gedeh, Usteri n. 214. Luzon Central; Loher n. 707 B. Distrib. India. China. A common species. 30. Carex continua, C. B. Clarke, in Hook. f. Fl. Brit. Ind. vi. (1894) p. 717. C. Bruceana, Boott! MS. partim. Luzon Central; Loher nn. 707, 708, 709, 710. Distrib. Yunnan, Sikkim. Nepaul. 12 MR. C. B. CLARKE ON 91. CAREX sPATIOSA, Boott! Carex, i. (1858) p. 86; ii. (1860) p- 246; Boeck. in Linnea, xl. (1876) p. 349. Cochin- China ; Gaudichaud. Var. B. bogorensis ; utriculis ovoideis; rostro cum 4-3 parte utrieuli &quilongo, scabro. C. indica, Mig. Fl. Nederl. Ind. iii. p. 350, non Linn. Xerotes latifolia, Hasskarl, MS. Java; fide Miquel. (I have seen the example ** Hort. Bogor.” of S. Kurz.) 32. CAREX NEO-GUINENSIS, sp. nova; paniculis partialibus in spicas compositas rigidas, oblongas congestis; utriculis (rostro incluso) 3 mm. longis, quam ei C. Lindleyane gracilioribus ; ceteroquin fere ut C. Lindleyana, Nees. Culmi 5-8 dm. longi. Folia 5 mm. lata. Panicule partiales e. 5, 3 cm. longe, 8 mm. late, rubre. Utriculi oblongo-ellipsoidei, glabri, rubentes ; rostrum cum 3-7 parte utrieuli equilongum, in marginibus hispidum, dentibus 2 longis lineari-lanccolatis. Neo-Guinea ; Guilianetti et English. 33. CAREX SATURATA, sp. nova; paniculis partialibus 3-4, late oblongis, 6 cm. longis, 24 mm. latis, perdensis, saturate sanguineo-purpureis ; spicis 10-14 mm. longis, basi fremineis; utriculis (rostro incluso) 3-32 mm. longis, ellipsoideis, glabris ; rostro eum utrieulo fere equilongo, lineari, glabro. C. filicina, O. Stapf, in Trans. Linn. Soc. ser. 2, Bot. iv. (1893) p. 246. Folia 1 cm. lata. Glume fomines saturate rubro-purpuree. Utriculi purpureo-maculati. Borneo; Kinabalu, alt. 3000 metr., Haviland n. 1402. 34. CAREX RHIZOMATOSA, Steud. in Zoll. Verz. Ind. Archip. Heft 2 (1854) p. 60, Cyp. p. 206 ; Mig. Fl. Nederl. Ind. iii. p. 348; Franch. in Nouv. Arch. Mus. Par. sér. 3, viii. (1896) p. 256. C. Cumingiana, Steud. Cyp. p. 206; Boott, Carex, iii. (1862) p. 107, tt. 324, 325; Boeck, in Linnea, xl. (1876) p. 367 ; Vidal, Pl. Vase. Filip. p. 286. C. capitulata, Boott! MS. Tonkin; Balansa n. 3824. Sumbawa; Zollinger n. 3447 fide Steudel. Ins. Philippine; Negros Ins., Cuming n. 1795. Distrib. Assam. Burma. THE CARICES OF MALAYA. 13 35. CAREX ARRIDENS, Hook. f. Fl. Brit. Ind. vi. (1894) p. 726. Larut-Perak ; alt. 900 metr., Kunstler (hb. King n. 2801). Distrib. Pegu. 36. CAREX HYPSOPHILA, Mig.! Fl. Nederl. Ind. iii. (1859) p. 394. C. tartarea, Ridley, in Journ. Bot. xxiii. (London, 1885) pp. 3 * C. decore affinis nisi eadem," Boott, MS. in Miquelii typo. Pedunculi ex unica bracteá sspe plures, Folia breviuscula, 6 mm. lata. Sumatra; alt. 2750 metr., Forbes n. 9444. Java; hb. Miquel; Gedeh, Usteri n. 101 ; Buitenzorg, Usteri n. 102. , 5. 97. Carex HaAVILANDI, sp. nova; foliis elongatis, 5 mm. latis ; inflorescentia 18-stachyá, pedunculis 3-4-nim fasciculatis ; spicis 3-32 cm. longis, linearibus, castaneo-rubris, pluribus basi foemi- neis apice masculis; utriculis (rostro incluso) 4 mm. longis, gracile lanceolatis, fere glabris; rostro cum i parte utriculi :quilongo, recto conico-lineari. C. hypsophila, O. Stapf, in Trans. Linn. Soc. ser. 2, Bot. iv. (1893) p. 246. Borneo; Kinabalu, alt. 3250 metr., Haviland n. 1403. 38. CAREX TURRITA, sp. nova; inflorescentia 5 dm. longá, 40- stachyà; pedunculo imo (interdum geminato) usque ad 14 em. exserto, gracili, apice 3-5 spieas approximatas gerente; spicis 5-7 em. longis, linearibus, pluribus basi feemineis apice masculis, bruneis, nutantibus ; utriculis (rostro incluso) 5 mm. longis, ellipsoideo-lanceolatis, glabris ; rostro quam utriculus vix breviore, conico-lineari, scabro. Ins. Philippine; Benguet, Loher n. 700. 39. CAREX SUMATRENSIS, sp. nova; planta 7 dm. alta, 25- stachya; folis 5 mm. latis; pedunculis inferioribus exsertis, interdum fasciculatis ; spicis 5 cm. longis, linearibus, pluribus basi foemineis apice masculis; utriculis rostro incluso 5 mm. longis, anguste ellipsoideis, glabris; rostro cum utrieulo æqui- longo, conico-lineari, scabro. Spice in apice pedunculi plures approximate, Utriculi basi attenuati, conspicue stipitati. Sumatra; Mons Dempo, alt. 2570 metr., Forbes n. 2388. 14 MR. C. B. CLARKE ON 40. CAREX BAcOANS, Nees! in Wight, Contrib. (1884) p. 122; Franch. in Nouv. Arch. Mus. Par. viii. (1896) p. 251; Hook. f. Bot. Mag. t. 7288, in Trimen, Fl. Ceylon, v. pp. 102, 107. Sumatra; Beccari n.21. Java; Horsfield, Goering. Cochin- china; Gaudiehaud. Tonkin; Balansa n. 2813. Luzon ; Loher nn. 706, 1948. Distrib. North India. Formosa. Var. B. siccifructus, C. B. Clarke, in Hook. f. Fl. Brit. India, - vi. p. 723. Java; alt. 2000-2500 metr., Koorders. Distrib. Khasia. 41. CAREX CURVIROSTRIS, Kunze, Suppl. Schkuhr (1840-50) p. 79, t. 20. C, recurvirostris, Steud. in Zoll. Verz. Ind. Archip. Heft 2, p. 60; Cyp. p. 207. Java; Zollinger. 42, CAREX TONKINENSIS, Franch. in Nouv. Arch. Mus. Par. sér. 3, viii. (1896) p. 251. Tonkin; Mons Bavi, 900 metr. alt., Balansa n. 2814. 43. CAREX COMPOSITA, Boott, Carex 1. (1858), p. 3, t. 8. Java; Buitenzorg, Usteri n. 100. Distrib. Assam. 44. CAREX BORNEENSIS, sp. nova; culmo 4 dm. longo, 6-stachyo ; spicis 2 em. longis, lineari-lanceolatis, apice masculis basi remote 5-6-utriculigeris, utriculis (rostro incluso) 6 mm. longis, lineari- lanceolatis, glabris; rostro eum 3 parte utriculi squilongo, lineari, apice inflato scarioso in altero latere fisso. C. fusiformis, O. Stapf, in Trans. Linn. Soc. ser. 2, Dot. iv. (1894) p. 246. Folia 5 mm. lata. Spies virides ; pars mascula ferruginea. Borneo; Kinabalu, alt. 3500 metr., Haviland n. 1404. 45. CAREX SPECIOSA, Kunth, Enum. Pl. ii. (1837) p. 504. C. peduncularis, Wall. MS., List n. 3391. C. grammocarpa, Sprengel MS. fide Miquel. Borneo; Banjarmassing, Motley n. 1222. Distrib. From South Madras to Nepaul and Assam, frequent. 46. Carex LoHERI, sp. nova; culmis 1-3 dm. longis, per- tenuibus, 3-l-stachyis ; spicis 7-12 mm, longis, ellipsoideis, THE CARICES OF MALAYA. 15 apice masculis, basi paucinucigeris; utriculis ellipsoideis, tri- quetris, undique pilosulis, basi stipitatis, apice in r. ..um longiusculum oblongum attenuatis. Folia 4 mm. lata, in facie inferiore glandulosa. Spice virides; apex masculus viridis. Glume fæmineæ utrieulis breviores, ims autem apice elongate interdum bracteiformes. Utriculi 5-6 mm. longi. Ins. Philippine ; Benguet, Loher nn. 701, 702, 703; Luzon Central, Morong n. 708 bis. 47. CAREX MADOERENSIS, sp. nova; eulmis 20-25 cm. longis, 3-7-stachyis; spicis omnibus sessilibus, 8-10 mm. longis, apice masculis basi paucinucigeris; utriculis ovalibus, apice angus- tatis, undique densius pilosis; rostro cum 3 parte utriculi sequilongo. Folia 5-6 mm. lata. Utriculi 4 mm. longi, fusce virides, basi obpyramidales. Rostri brevis dentes 2 lineari-lanceolati. Java; ins. Madoera, Zollinger n. 1778 (in herb. Wien). 48. CAREX LONGIBRACTEATA, Steud. in Zoil. Verz. Ind. Archip. Heft 2 (1854) p. 60, Cyp. p. 205; Mig. Wl. Nederl. Ind. iii. p. 348, non Boeck. Java; Mt. Waliran, alt. 10,000 feet, Zollinger n. 524 Z. I have seen no specimen. Subgenus VI. PROPRIE. 49. CAREX OLIVACEA, Boott, in Proc. Linn. Soc. i. (1846) p. 286; Carex, p. 56, t. 149. C. confertiflora, Boott, in Mem. Amer. Acad. n. s. vi. (1859) p. 418, Carex, p. 184; Franch, in Nouv. Arch. Mus. Par. sér. 3, x. (1898) pp. 75, 98. Java; Preanger, alt. 1250 metr., Forbes n. 1091. Distrib. North-east India. Japan. 50. CAREX JackrANA, Boott, in Proc. Linn. Soc. i. (1846) p. 260, Carex, p. 9, t. 25; Franch. in Nouv. Arch. Mus. Par. sér. 3, x. (1898) pp. 43, 96. C. oxyphylla, Franch.! in Nouv. Arch, Mus. Par. sér. 3, x. (1898) pp. 57, 97. C. instabilis, Boot? MS. Java; Horsfield. Distrib. Khasia, Yunnan.—Also a var. (B. minor) in Ceylon and the Nilgiri Hills. 16 MR. C. CLARKE ON THE CARICES OF MALAYA. 51. CAREX rvMIDA, Boott, Carex, i. (1858) p. 66, t. 181. C. baviensis, Franch.! in Now. Arch, Mus. Par. sér. 3, x. (1898) pp. 77, 98. Tonkin ; Mt. Bavi, Balansa n. 2825. Distrib. Sikkim. Khasia. Yunnan. 59. Carex PSEUDOCYPERUS, Linn. Sp. Pl. ed. 1, p. 987, ed. 2, p. 1387 ; Boott, Carex, p. 140, tt. 451, 452 ; Benth. Fl. Austral. vii. pp. 437, 448 ; Kuek. in Engler, Bot. Jahrb. xxvii. (1899) pp. 549, 563; Franch. in Nouv. Arch. Mus. Par. sér. 3, x. (1898) p. 59. C. fascicularis, Boott! Carex, i, (1858) p. 53, pro magna parte i. e. t. 139 et t. 140 fig. sinistr.; Hook. f. Fl. Tasm. ii. p. 101. Java; Horsfield (a Boott “ C. fascicularis" inscripta). Distrib. Palearctica (vulgaris) Nearctica. India, Japan (rarior) Australia. Neo-Zeeland (frequens), Mexico.— Var. f. comosa is common in North America; varr. y. platygluma et ò. Haenkeana are common in South America. 53. CAREX BREVICULMIS, R. Brown, Prodr. (1810) p. 242; Franch. in Nouv. Arch. Mus. Par. sér. 3, ix. (1897) p. 192, x. (1898) p. 95. C. Royleana, Nees! in Wight, Contrib. p. 127 ; Boott, Carex, i. (1858) p. 6, t. 19. C. leucochlora, Bunge! in Mém. Sav. Etrang. St. Pétersb. ii. (1835) p. 142. C. Langsdorfii, Boott! in Trans. Linn. Soc, xx. (1847) p. 144. C. puberula, Boott! in Perry, Exped. ii. (1857) p. 324. C. discoidea, Boott! in Mem. Amer. Acad. vi. (1859) p. 419. Neo-Guinea ; Mt. Seratchley, alt. 3750 metr., MacGregor. Distrib. Ussuri. Japan. India. China. Australia. New Zealand. 54. Carex WALLICHIANA, Nees, in Wight, Contrib. (1834) p. 129; Boott, Carex, i. (1858) p. 16, t. 18; Franch, in Nouv. Arch. Mus. Par. sér. 8, x. (1898) pp. 87, 99; Prescott MS. in Wall. List, n. 3380 (nomen). C. hirta, Boiss, Fl. Orient, v. p. 491; Boeck.! in Journ, Linn, Soc., Bot. xviii. (1881) p. 105, non Lenn. Tonkin. Distrib. Cabul. North India. Yunnan. [I have seen this plant in a Tonkin collection, but there is no further information in my note. | ON SOME BICARPELLARY BEANS, 17 Some Bicarpellary Beans, By Eric Drank, D.Se., F. L.S. [Read 21st January, 1904.] Dvnrwa the autumn of 1903 an interesting series of fruits of the French Bean, Phaseolus vulgaris, Savi, was obtained from a garden on the clay of the Middle Coal-measures of North Derbyshire. These beans were bicarpellary in nature, but the degree of development attained by the second carpel varied considerably. In the simplest case (fig. 1) it was present merely Fig. 1. A È i A as a small pod on the posterior aspect of the normal carpel, and adhered to the latter in its lower portion in such a way that a bilocular ovary was formed. The posterior carpel was without seeds. In other cases (figs. 2 & 3) the posterior carpel was more extensively developed, and the basal fusion with its anterior fellow resulted, not, as in the last case, in a bilocular ovary, but in a unilocular one. Distally the two carpels were free from one another. - In the bean shown in fig. 4 the same sort of structure was attained, but the posterior carpel was as large as the anterior one and bore one or more well-formed seeds. The distal extremities of the carpels were free and divaricated widely, recalling somewhat the kind of ovary met with in the Saxifrages. A particularly striking fruit was that shown in fig. 5. Here the midrib of the posterior carpel was normally developed in the distal region, but more proximally it was greatly reduced, and passed gradually towards one of the lateral sutures of the fruit. LINN. JOURN.—- BOTANY, VOL. XXXVIT. C 18 DR. ERIC DRABBLE ON In some cases the midrib became extinct before reaching the lateral suture; in others it fused with the marginal vein of the upper carpel. At the same time the distal quadrangular seetion of the fruit gave way proximally to a triangular form. Fig. 2. In these specimens the fusion of the two carpels was nearly complete, only the distal extremities being free. Finally, a single specimen (fig. 6) was met with exhibiting the triangular cross-section throughout. That we are not here dealing with a tricarpellary fruit, as might at first sight be SOME BICARPELLARY BEANS. 19 supposed, is shown not only by the mode of development indicated in the last case, but also by the arrangement of the veins, as figured in fig. 5. Anomalous though it may seem, it is diffieult to eseape the conviction that we are confronted with a fruit composed of a carpel and a half. The first notice discovered of bicarpellary fruits of Phaseolus was that by Moquin-Tandon (1) in 1841, mention being made of the occurrence of two or even three carpels. In 1844, Kirschleger (2) described a form much like that shown in fig. 4, ascribing it to * Phaseolus vulgaris digynus”; Wydler (4), Fig. 5. e 5. n — | — D n DL—— i \ a a b ce ) VW Vc Fig. 6. a 1 \ i _ Cj * [L] [] L] e 1860, and Sehlotthauber (8) described double-fruited beans; while Du Hamel (5), in his * Physiologie des Arbres,’ very roughly figures a case similar to that in fig.4. Schlechtendal (6) describes a case with two carpels fused below and free above, and only one-seeded. Dr. Maxwell Masters (7) in his ‘Teratology’ gives a list of leguminous plants in which two or more carpels have been found. This list includes representatives of the Mimosoides, the Cesal- pinioidez, and the Papilionoidez. Fermond (9) described a form which would appear to closely c2 s- 20 DR. ERIC DRABBLE ON resemble fig. 6. He says: “Enfin il arrive frequemment que les 2 cótés d'un carpelle (celui qui eroit proportionnellement moins que l'autre) sont considérablement écartés à ce point que bientót ils se trouvent sur un méme plan forment avec les 2 cótés de l'autre carpelle un fruit à 3 angles, et à une seule loge." .... “ Dans ce cas, les trophospermes sont placés aux 2 angles contigue aux 2 faces du carpelle qui se sont placés dans un méme plan; le 3* angle ne porte point de placenta.” He also mentions a triearpellary condition with a triloeular ovary and axile placentation. Penzig (11), 1890, and Taubert (12), 1892, refer to bicarpel- lary fruits of Phaseolus. An important point was raised by Moquin-Tandon. He says: “Quand ces fruits ont acquis une certaine grosseur, et qu’ autour d’eux il ne reste plus de traces des parties florales il devient alors presque impossible de reconnaitre s'ils ont été produits par deux ou plusieurs fleurs et par symétrisation.” In the beans figured and described above, there can be no doubt that they arose from a single flower, as the calyx was, in most cases, still present and comprised the usual 5 sepals. In the same way, so far as the calyx is concerned, no indication of any doubling of the flower was apparent. With reference to the possible meaning of the facts recorded above, but little ean be said. "There is not much doubt that the Leguminose are derived from a polycarpellary stock, and in some degree at least the bicarpellary nature may be regarded as a reversion to an ancestral state. This, however, cannot be said for the synearpous nature of the ovary. In view of the fact that all the normally polycarpellary members of the order (Prosopis sp., Hausemannia sp., and others) and most recorded cases of anomalously polycarpellary forms, exclusive of Phaseolus, are apocarpous, it is at least open to question whether we have in the forms above described a truly primitive condition. If so, then the series read from 4, through 3, 2, 1, and culminating in the ordinary bean-fruit, may possibly represent the manner of reduetion to the present monocarpellary condition. The form shown in fig. 5 would in this ease be quite anomalous, as indeed it . would appear to be under any conditions. Although in our present state of knowledge no special importance ean be attached to the soil upon which the plants were cultivated, it seemed advisable to place it upon record in SOME BICARPELLARY BEANS. 21 the hope that at some future time a coordination of soil and special development may become possible, and it may be sug- gested that it would in all cases be wise to indicate not only the place but also the geological formation and nature of the soil upon which any teratological specimens may occur. References. (1) Moquis-TawpoN, C. H. B. A. Élém. Térat. Vég. p. 277 (1841). (2) Krrscuiecer, F. Flora, xxvii. 1844, p. 130. (3) Gopron, D. A. Mém. Soc. Sci. Nat. Cherbourg, vol. xvi. p. 35. (4) Wynter, H. Flora, xliii. 1860, p. 20. (5) Du Hamer pu MowckaAv, H. L. Physiol. des Arbres, pl. 13. figs. 318, 319. (6) ScunkcuTENDAL, D. F. L. vos. Botanische Zeitung, xiii. 1855, p. 823. (7) Masters, M. T. Teratology, 1879, p. 264. (8) SCHLOTTHAUBER, A. F. Bonplandia, viii. 1860, p. 45. (9) Fermonp, V. Essai de Phytomorphie, Paris, vol. i. 1884, p. 168. (10) Auxqvuisr, S. Bot. Centralb. xxix. 1887, p. 93. (11) Penzte, O. Pflanzenteratologie, Bd. i. p. 402 (1890). (12) Tausert, P. Leguminoss in Engler & Prantl, * Die natür- lichen Pflanzentamilien,’ iii. 3, pp. 94-96 (1891). 22 SIR J. D. HOOKER ON THE SPECIES OF On the Species of Zmpatiens in the Wallichign Herbarium of the Linnean Society. By Sir J. D. Hooker, G.C.S.I., F.R.S., F.L.S. [ Read 2nd June, 1904.] Tur species of Impatiens in the Wallichian Herbarium, con- sisting almost exclusively of those collected by Wallich or his employés in Nepal and Silhet, together with those of the missionaries in Malabar, made during the latter half of the 18th and beginning of the 19th centuries, represent nearly all that was known of the Indian Balsams at the date of the distribution of that Herbarium. They amount to 48 ticketed species (Nos. 4729-4775 & 7274, 7275), of which 18 are from Malabar, as many from Nepal, 10 from Silhet, 6 from Burma (including Rangoon and Tavoy), and 2 from Sirmore. Though comparatively few in number, for the genus is now known to contain upwards of 200 British Indian species, they foreshadow the remarkable fact of a segregation of these in the several phytogeographical regions of India in which they occur *, which has no parallel in any other large genus of plants known to me. * Of the 200 British Indian species (many of them as yet unpublished) there are, in the Kew Herbarium, approximately 23 from the Himalaya west of Nepal, 63 from the Eastern Himalaya (inclusive of the Valley of Katmandu in Central Nepal), 52 from Burma (inclusive of Assam and Silhet), 58 from Malabar, and 21 from Ceylon. Of the 23 West Himalayan species only 11 have been found in the Eastern Himalaya, of which 8 alone enter Sikkim. In other terms, of nearly 80 Himalayan species only about one-tenth are known to be com- mon to both sections of the range, a proportion which will be greatly reduced when the unexplored ranges East of Sikkim are botanized. Of the 52 Burman species only 10 have been found in the Eastern Himalaya. The 7 Malayan Peninsular species are absent elsewhere in India. Of the 53 Malabar species only J. Balsamina, L., oppositifolia, L., and chinensis, L., are found elsewhere in British India; only 8 of the 58 inhabit Ceylon. This segregation of species extends in a marked degree to that of the two great divisions of the genus, namely those species with capsules turgid in the middle, and those with capsules linear or clavate. Every Malabar and Ceylon species belongs to the first of these divisions; every Western Himalayan (except I. Balsamina) to the second. Of the Eastern Himalayan the great majority belong to the second ; of the Burman the great majority belong to the first, as do all the Malay Peninsular. I have little doubt that when the Himalayan ranges Hast of Sikkim and the mountain regions of Burma come to be explored botanically, the number of British Indian species will approach 300. IMPATIENS IN THE WALLICHIAN HERBARIUM. 23 Thus, of all the species contained in the Walliehian Herbarium, only one, J. Balsamina, L., is common to the five of the regions indicated in the note on the preceding page, namely, Eastern Himalayan, Western Himalayan, Burmese, Malabarian, Ceylonese and Malayan Peninsular. Of the 18 Malabarian 3 only were found in other regions: namely, J. Balsamina, L., in four others ; I. chinensis, L., in two; and J. oppositifolia, L., in one, Burma. Of the 18 Nepal species, two alone were colleeted in the adjoining region to the westward. Of the 10 Silhet species, two alone are Malabarian. The consultation of the Wallichian collection of Balsams presents great difficulties which it is the object of this com- munication to alleviate. In many cases two or more species are fastened down under one number and name; and in not a few cases one species occurs under several numbers. This is partly due to the great difficulty in distinguishing badly preserved mounted specimens of a puzzling genus without moistening and removing flowers of the most delicate consistency and making a microscopical examination of their organs; but a far more serious source of confusion is one the origin of which can only be conjectured: namely, that a part of the collection which, after being sorted for mounting (by Bentham, who appears to have ticketed the species), and transferred to the Society’s custody, had either fallen from the hands of a custodian or been swept off a table, and the scattered specimens and tickets had been dealt with in a haphazard way by an ignorant mounter *. It seems to me impossible otherwise to explain the occurrence upon one sheet, under one number (4730), of four species so totally dissimilar in habit, foliage, inflorescence, and flowers as J. racemosa, DC., I. bicornuta, Wall, I. tingens, Edgew., and I. odorata, D. Don; or of such three under No. 4729 as I. bicornuta, Wall., T. bicolor, Royle, and J. sulcata, Wall.; or the mounting side by side under No. 4770 excellent individuals of J. leptoceras, Wall., and J. Balsamina, L., which have no resemblance. l Further sources of difficulty are the misplacement of MSS. tickets, other than the lithographed numbered ones, that are attached by pins to the sheets of the species to which they belonged or were supposed to belong, the minuteness of the * There is no other instance known to me in the Walliehian Herbarium of a similar confusion of specimens and tickets. 24. SIR J. D. HOOKER ON THE SPECIES OF flower of some species, and the absence of fruit of others. Taking all sources of difficulty into aecount, the following summary of their effects is, T hope, approximately correct :— Cases of two or more species occurring under one number: 5 species in 1 case; 4 species in 1 case; 3 species in 2 cases ; 2 species in 10 cases. Cases of single species occurring under two or more numbers : 1 species under 7 numbers; 2 under 3 numbers; 10 under 2 numbers. ENUMERATION OF SPECIES, WITH OBSERVATIONS, No. 4729. “ IMPATIENS LONGICORNU, Wall. Napalia 1821.” One sheet with three species. On the right T. bicornuta, Wall. See also 4730, 4765. On the left T. bicolor, Royle * (I. umbrosa, Edgew.). In the middle 7. sulcata, Wall. (I. gigantea, Hdgew.). See also 4761. No. 4730. “ IuPATIENS racemosa, DC. Napalia 1821." "Three sheets with five species. Sheet 1. On the left I. tingens, Edgew. t On the right I. bicornuta, Wall. See also 4729, 4765. Sheet 2. On the upper right hand Z. bicornuta, Wall. On the left 7. racemosa, DC. On the lower right hand J. racemosa, DC. On the lower left I. odorata, D. Don. Sheet 3. * No. 4730 B. Sillet, W. G.” T. Balsamina, L., with double flowers. See also 4731, 4770. No. 4731. *IwPaArrENS BarsAMINA, Linn. H.R.C." With double flowers. See also 4730 B, 4770. * The name J. bicolor, Royle, placed in the ‘Flora of British India as a synonym under Z. umbrosa, Edgew., must I think be retained, though the description and plate in the ‘ Botany of the Himalaya’ do not suffice to distinguish it from several others. It is the commonest species in the Western Himalaya. T This being the dominant specimen of those ticketed Z. racemosa, DC., by Wallich, I took it for that plant and reduced Z. tingens to a synonym in the ‘Flora of British India. From information kindly supplied by M. Casimir de Candolle, I find that I erred in doing so. No. No. No. No. No. No. . 4741. “ Impatiens INCONSPICUA, Benth. Hb. Heyne.’ IMPATIENS IN THE WALLICHIAN HERBARIUM. 25 4732. “IMpaTIENS COCOINEA, Sims.” = I. Balsamina, L.— Both A. and B are of var. rosea, Hook. f., = I. rosea, Lindl. See also 4730 B, 4731, 4733, 4734, 4735, 4770. A. " Napalia" with MS. ticket “Impatiens glandulifera, W all., an I. Balsamina, L., spontanea? Thaukote, Aug. 1821.” B. *Sirmore, G. Govan, M.D." 4733. “Impatiens CORNUTA, Linn." = I. Balsamina, L. See above. A. “I. cornuta, Herb. Wight.” B. “I. cornuta, Heyne.” C. * I. Tiramira et I. Lacca, H. Ham. Nathpur.” D. * Ripa Irawaddi, 1826." 4734. “ IMP. LONGIFOLIA, Benth. Hb. Wight.” = I. Balsa- mina, L. See above. 4735. * IMP. ARCUATA, Benth. Imp. sinensis, Hb. Madr. Mysore.” = I. Balsamina, L., var. 4736. “ IMP. scaBRIUSCULA, H. Heyne. Two species. Upper right and left and lower right hand specimens, J. scabriuscula, with MS. ticket “ Impatiens ramosa, 23rd Sept. [18]16," and in another hand * scabriuscula.” Lower left hand, I. pendula, H. Heyne. See also 4744. 4737. “Imp. LATIFOLIA, Linn.” One sheet with two species. A. “Imp. latifolia, Hb. Madr. e Travancore.” B. “Imp. latifolia, Hb. Wight.” = I. lucida, H. Heyne. See also 4738. . 4738. “Imp. tucipa, Hb. Heyne?” With MS. ticket ‘‘ Impatiens lucida, Courtallum, ab amicissimo Heyne." See also 4737. . 4739. “ IMPATIENS LESCHENAULTII, De Cand. Hb. Wight.” Two species. A. I. Leschenaultii, DC. B. Nilghiry dom. Noton, with MS. ticket ‘ Impatiens, Noton. Neilghery, a small branch." = J. latifolia, L. See 4737. . 4740. “Impatiens HISPIDULA, Benth. Napalia 1821." = I. puberula, DC. See also 4767. ' Bee also 4743. . 4742. “ IMPATIENS MULTIFLORA, Benth. Pundua Mont. De Silva." = I. tripetala, Roxb. See also 4752. 26 SIR J. D. HOOKER ON THE SPECIES OF No. 4743. “ Impatiens MYSORENSIS, Koth.” Four species. A. “I. oppositifolia, Hb. Madr.” The upper specimen, on the left, with MS. ticket “I. oppositifolia, e Kut- talam, ab amicissimo Mitchell,” is true I. oppositifolia, Wall., with a scrap of T. Kleinii. See also 4746. B. “I. oppositifolia, Hb. Wight, etiam I. rosmarinifolia Dec. e Courtall.” The three specimens on the right of the sheet, with the MS. ticket ** Balsamina Rosmarini- folia, DC., Courtallum, 2 Feb. 1827, in moist places amongst long grasses," = I. inconspicua, Benth. C. “Impatiens mysurensis, Hb. Wight,” consists of frag- ments in the lower left hand corner of the sheet to which a MS. ticket applies (though attached to a specimen of I. oppositifolia), bearing the inscription in Rottler’s handwriting: ‘Impatiens Maysorensis, nob. Pedunc. l-foris, axillarib. solitariis binisq. Fol. altern. ovato- lanceolat. serratis." This is the true Z. mysorensis, Rottl. No. 4744. * IMPATIENS PENDULA, Hb. Heyne," with MS. ticket of Heyne. “Impatiens pendula. Pedune. uniflor. axill. s. lateral. post anthesin reflexis. Fol. altern. ad basin attenuat. margine mucronatis. 22d Sept. (18]16." See also 4736. No. 4745. “ IMPATIENS PUSILLA, Hb. Heyn." = I. inconspicua, Benth. See also 4741 & 4743. No. 4746. * IMPATIENS TENELLA, Herb. Heyn." Two species ; two upper specimens and right hand lower with MS, ticket of Heyne. “ Impatiens sp. longiflora. Pedune. solitar. axill. longissim. Folia opp. oblonga, serraturis cuspidat. 22d Sept. [18]16." B. “Impatiens fasciculata? Hb. Madr. Travancore.” Lower left hand specimen = J. Kleinii, Wight & Arn. See also 4743. No. 4747. “I. RUFESCENS, Benth. Herb. Wight.” = I. tomen- tosa, Heyne, non Wall. Cat. No. 4748. I. HETEROPHYLLA, Wall. Two sheets, two species. A. “Sillet, DeSilva & H. D." — 7. chinensis, L. See also 4749. B. * Taong Dong, 1826." Upper right and left hand speci- mens to which the MS. ticket '* Balsamina amena, Wall.," belongs. The three other specimens— Z. diversifolia, Herb. Heyne. See 4749. IMPATIENS IN THE WALLICHIAN HERBARIUM. 27 No. 4749. Impatiens DIVERSIFOLIA, Hb. Heyne. Two sheets, No. No. No. No. No. No. No. two species. B. “Impatiens malabarica, Hb. Madr.” with one specimen (the middle lower) of J. chinensis, L., and MS. tickets “ Impatiens fasciculata prox. Cuttalam, Sept. 1814,” and “ Impatiens sp. malabarica, mihi. Caul. debilis basi “repens. Fol. opposita cordato-lanceolata acuta serrata "serraturis mucronatis inferiore glauca. Pedune. uni- “floris aggregatis nectariis florum parum longioribus. “ Quilon, Oct. 1814.” The mihi I suppose to be Klein. C. “ Herb. Wight." . 4750. *IwPaTIENS RETICULATA, Wall. Rangoon 1826." = I. oppositifolia, L. See also 4743, 4751. 4751. * IMPATIENS TOMENTOSA, Herb. Heyne," = I. oppositi- Jolia, L. See above. This is an obvious misplacement ofa ticket. The plant, of which there are two specimens ou the sheet, is perfectly glabrous in all its parts. Two MS. tiekets are attached to the sheet, one inscribed “Impatiens tomentosa capsulis totis tomentosis"; the other “Impatiens fasciculata, aurant. ad pedem mont Bauvaputa." 4752. IMpATIENS TRIPETALA, Roxb. Three sheets. A. *Sillet" F. D. A B. “I. ternifolia, H. Ham., with Hamilton’s ticket “ Gualpara, 8th May, 1808." C. * H. B.C.” 4753. IMPATIENS LVIGATA, Wall. Sillet, F. D. Two sheets. B. “ H. B. C. e Sillet.” 4754. “Impatiens ACUMINATA, Benth.” Sillet, H. B. 4755. “Impatiens NATANS, Willd. Hb. Wight." = Hydro- cera triflora, W. & A. See also 4756. 4756. “Impartens TRIFLORA, Willd.” Three sheets. = Hy- drocera triflora, W. & A., 4755. A. “I. triflora, Hb. Madr.” B. Sillet, De Silva. C. Ripa Attran, 1827. . 4757. “ IMPATIENS UMBELLATA, H. Heyn.,” with MS. ticket “ Impatiens formosissima, 1st Aug. [18]18." 4758. * IMPATIENS SCAPIFLORA, Herb. Heyne." B. “ Impatiens cordifolia, Hb. Madras,” with MS. ticket * Impatiens sp. n. cordifolia, mihi. Pedunculis multiflor. e Kuttalam apportat. ab. amiciss. D. Mitchil.” 28 No. No. No. No. No. No. . 4763. “Impariens RADICANS, Benth. Sillet, F. D. SIR J. D. HOOKER ON THE SPECIES OF . 4759. ** IMPATIENS GRANDIS, H. Heyne." B. Impatiens grandiflora, Herb. Madr., with MS. ticket “ In montosis prope Cuttalain, Sept. 1818." 4760. “ I. BRACcTEATA, Coleb. Mont Silhet, F. D. & H. B.,” with MS. name 7. trichocephala, Wall. . 4760 in pencil; error for 4766 g.v. No. 4761. “Impatiens Jurpra, H. Ham. e Morung.” Two sheets. B. Sillet, F. D. . 4762. * Iur. FRUTICOSA, Lesch. Nilghery, Leschenault de Latour." " Two sheets. Two species. B? * Mt. Sillet, F. D. & W. Q.” = trilobata, Coleb. . 4764. “IMPATIENS SULCATA, Wall.” Two sheets. Two species. Specimens very bad and identification doubtful. A. Napalia 1821. Right hand specimen J. Royleana, Walp. ; left hand Z. sulcata, Wall. See also 4729. B. *Sirmore, Dr. Govan.” Upper specimen Z. Roylei, Walp. ; two specimens J. sulcata, Wall. ? See also 4729. 4705. * IMPATIENS BrconNUTA, Wall. Napalia 1821." Two sheets. See also 4729, 4730. . 4766. “Imr. instants, DC. Napaulia 1821," on blank sheet. See 4760 in pencil. 4767. “Impar. DISCOLOR, De Cand. Napaulia 1821.” Three sheets. Three species. Sheet 1. T. discolor, DC., four specimens. J. Wallichit, Hook. f., two specimens. Sheet 2. T. discolor, DC., one specimen. J. Wallichii, Hook. f., two specimens. Sheet 3. 7. puberula, DC., with MS. name Z. mollis, Wall. See also 4740. 4768. “Impat. vugrICIFOLIA, Wall. Gossain Than." 4769. * IuPaT. scABRIDA, DC." Two sheets. A. Napaulia 1821, with two MS. tickets ** Impatiens punc- tata, Wall.” and “I. calycina, Wall., both inscribed * Chandagiri, Aug. 1821.” B. * Sirmore, Dr. Govan.” 4770. *IwPATIENS LEPTOCERAS, DeC. Napaulia 1821." Three sheets. Two species. Sheet 1. J. leptoceras, DC. IMPATIENS IN THE WALLICHIAN HERBARIUM. 29 Sheet 2. Right hand specimen T. leptoceras, DC.; left hand J. Balsamina, L. Sheet 3. I. Balsamina, L. No. 4771. “ IMP. SERRATA, Benth. Napalia 1821.” No. 4772. “Imp. CIRCÆOIDES, Wall.” Two species. A. Rangoon 1826. B. Prome 1826. There are four specimens of two species on the sheet; the two upper with a simple stem, marked a in pencil, no doubt belong to an adjoining MS. ticket inscribed “I. simplex, Wall., Prome, Nov. 1826." The two lower are of a very different and much branched species, in an undeterminable condition, with a ticket inscribed * I. ciresoides, Wall., Legi ad fossas coet. prope Rangoon, Aug. 1826”; it has no resemblance to a Circea. From the above, it would appear that the A of the lithographed ticket is the Prome plaut, and B, the Rangoon, an unnamed one. There are specimens of the Prome plant in the Herb. Kew from Moulmain. No. 4773. “Imp. TAVOYANA, Benth. Tavoy, W. Gomez.” No. 7274. * IuPATIENS RACEMULOSA, Wall. Silhet, Gomez." No. 7275. “Impatrens PORRECTA, Wall. Silhet, Gomez." An additional Nepal species of Impatiens collected by Wallich in 1821 was found by me when examining some bundles of specimens which remained in the rooms of the Linnean Society after the completion of the great distribution in 1831? It isan undescribed species, for which I propose the name of I. preter- missa*. In the same bundle I found a form of T. bicolor, Royle (I. umbrosa, Edgew.), or of an undescribed species, with the MS. * [ praetermissa; herbacea, ramulis foliisque iminaturis puberulis, foliis 1-14 poll. longis alternis ovatis ovato-lanceolatisve acuminatis serratis basi rotundatis v. in petiolum brevem angustatis, glandulis stipularibus pulvinatis, peduneulis axillaribus gracilibus folis multo brevioribus fureatis 2-floris pubescentibus, basi et ad furcam ebracteatis, pedicellis pedunculo longioribus medio bracteatis, bracteis lanceolatis, floribus mediocribus, sepalis 2 amplis oblique oblatis membranaceis aristatis, vexillo erecto orbiculari costa medio dorso longe cornuta, alarum lobo basali antice rotundato basi cuneato, lobo distali paulo longiore anguste dolabriformi, labio eum caleare incluso 1-1} poll. longo late infundibulari in ealear gracile attenuato, capsula pollicaris linearis acuminata, seminibus subglobosis ,'; poll. diam. subrugulosis. Belongs to a section of the genus with Z. scabrida, DC., and others, ineluding the European and North American ones, in which the ebracteate peduncle is forked and the bracts usually placed on the pedicels above their middle, not at their base. 30 SIR J. D. HOOKER ON THE SPECIES OF ticket of “ I. longicornu, Wall., no. 4729," which differs from the ordinary form of 7. bicolor in the leaves being serrulate and the flowers much smaller. In the type collection, No. 4729 Impatiens longicornu, Wall., is represented by three species, Z. bicolor, Royle, I. bicornuta, Wall., and T. sulcata, Wall., all of which are remark- able for the shortness of their spurs. The true J. longicornu, Wall., in Roxb. Fl. Ind. Ed. Carey is 7. leptoceras, DC., and longicornu hence disappears. The two above-mentioned plants I have placed in the cover with Nos. 7274-5 of the Walliehian Herbarium. INDEX. I. acuminata, Benth. No. 4754. Silhet. I. amena, Wall. 4748 B — I. diversifolia, Wight. I. amphorata, Edgew. 4729 in part—I. bicolor, Royle. I. arcuata, Benth. 4735=I, Balsamina, L, var. I. Balsamina, L. 4730 B, 4731, 4732, 4733, 4734, Nepal, Sirmore, 4735, 4770 in part. Silhet, Malabar, Burma. I. bicolor, Royle, 4729 in part. Nepal. I. bicornuta, Wail. 4729 in part, 4730 in part, 4765. Nepal. I. bracteata, Cole), 4760. Silhet. I. calycina, Wall. 4769 A =I. scabrida, DC, I. chinensis, L. 4748, 4749 in part. Silhet, Malabar. I. circæoides, Wall, 4772 in part. Prome. I. coccinea, Sims, 4732=I, Balsamina, L. LI. cordifolia, Herb. Madr. 4758=[1. scapiflora, Herb. Heyne. I. cornuta, L. 4733=[I. Balsamina, L. I. discolor, DC. 4767 in part. Nepal. I. diversifolia, Herb. Heyne, 4748, 4749. Malabar. I. fasciculata, prox. 4749—1. diversifolia, Herb. Heyne. I. fasciculata? Herb. Madr. 4746 B—I. Kleinii, W. § A, I. fruticosa, Lesch. 4762. Malabar. I. gigantea, Edgew. 4729, 4764 —I. sulcata, Wall. I. glandulifera, Wall., MSS. 4732 A. I. grandiflora, Herb. Madr. 4759 B=I. grandis, Herb. Heyne. I. grandis, Herb. Heyne, 4759. Malabar. I. heterophylla, Wall. 4748=I. chinensis, L., and I. diversi- folia, Herb. Heyne. I. hispidula, Benth. 4740=1. puberula, DC. I. inconspicua, Benth. 4741, 4743. Malabar. I. insignis, DC. 4760, 4766. Nepal. I. Juspia, Herb. Ham. 4761. Nepal. I. Kleinii, W. 4 A. 4743 in part, 4746 in part. Malabar. I. Lacca, Herb. Ham. 4133 C— I. Balsamina, Z. I. levigata, Wall. 4753. Silhet, I. latifolia, L. 4737 in part, 4739 in part. Malabar. IMPATIENS IN THE WALLICHIAN HERBARIUM. I. leptoceras, DC. I. Leschenaultii, DC. I. longicornu, Wall. Cat. I. longifolia, Benth. I. lueida, Herb, Heyne, I, malabarica, Herb. Madr. I, mollis, Wall. I, multiflora, Benth. I. mysorensis, Roth, L natans, Willd. I. odorata, D. Don, T. oppositifolia, Z. I. pendula, Herb. Heyne, I. porrecta, Wail, I. puberula, DC. I, punctata, Wall. IL. pusilla, Herb. Heyne, I. racemosa, DC. I. racemulosa, Wall. I. radicans, Benth. I. reticulata, Wall. I, rosea, Lindl, I. rosmarinifolia, DO. I. Roylei, Walp. I. rufescens, Benth. I. scabrida, DC. I. scabriuscula, Herb. Heyne, I. scapiflora, Herb. Heyne, I. serrata, Benth. I. simplex, Wall., MSS. I. sinensis, Herb. Madr. I. sulcata, Wall. I. tavoyana, Benth. I. tenella, Herb. Heyne, I, ternifolia, Herb. Ham. I. tingens, Edgew. 2. Tiramira, Herb. Ham. L, tomentosa, Wall. I. tomentosa, Heyne, I. trichocephala, Wall., MSS. I. triflora, Willd. I. trilobata, Coleb. . tripetala, Roxb. . umbellata, Herb, Heyne, . urticifolia, Wall. . Wallichii, Hook, f. = m m Sp. dub. I. prietermissa, Hook. f. 4755, 4756=Hydrocera triflora, Willd. { 4770 in part. 4139 in part. 4729, 734=I. Balsamina, L. 4737 in part, 4738. 4749 B=I. diversifolia, Herb. Wight. . 4767 in part —I. puberula, DC. 4742—1. tripetala, Roxb, 4743 in part. 4730 in part, 4743 in part, 4750, 4751. 4736 in part, 4744. 7275. 4740, 4767 in part. 4769— I. scabrida. 4145—1I. inconspicua, Benth. 4730 in part. 7274. 4763. 4750=I. oppositifolia, L. 4732—1. Balsamina, Z. 4743 B —I. oppositifolia, L. 4764 in part. 4747 =I, tomentosa, Heyne, 4769. 4736 in part. 4758. 4771. 4772 A=I. cirezoides, Wall. 4735=I. Balsamina, L. 4729 in part, 4764 in part. 4773. 4746 in part. 4152 B—I. tripetala, Rorò. 4730 in part. 4733 0. 47151 =I. oppositifolia, L. 2747. 4760=I. fimbriata, Coled, 4756 — Hydrocera triflora, W. § A. 4763 B ? 4742, 4752. 4757. 4708. 4767 in part. 4772 B. Unticketed, 31 Nepal. Malabar. Malabar. Malabar. Silhet, Burma, Malabar. Nepal. Nepal. Malabar. Silhet. Nepal. Malabar. Nepal. Silhet, Bilhet. Rangoon. Sirmore. Malabar, Nepal, Sirmore. Malabar. Malabar. Nepal. Rangoon. Nepal. Tavoy. Malabar. Nepal. Malabar, Silhet. Silhet. Malabar. Nepal. Nepal. Rangoon, Nepal. 32 DR. A, D. WALLER ON THE Number of ticketed sheets in the Wallichian Herbarium ... 48 » species from Malabar ................. + Leeeeeeaeaens 18 » » » Nepal ..... RH 18 » » » Bilhet ............eeeeee e 10 » " », Burma (Rangoon, Prome, Tavoy).. 6 » » , Birmore ....... etre 2 102 Species common to Nepal, Sirmore, Malabar, Silhet, and Burma, 1 (I. Balsamina). Silhet and Malabar, 4 (I. Balsamina, chinensis, natans, i. e. Hydrocera triflora). Nepal and Sirmore, 2 (I. Balsamina, scabrida). Morung and Silhet, 1 (I. Jurpia, Ham). On the Blaze-Currents of Vegetable Tissues: a Week's Holiday with a Galvanometer and some Plants. By A.D. Wanter, M.D., F.R.S. (Communicated by Prof. J. B. Farmer, F.R.S., F.L.S.) [Read 16th June, 1904.] $1. Frrst Day.—I set up the galvanometer and keyboard in a convenient corner of the room, and verify the circuit. The two unpolarizable electrodes À and B and the galvano- meter are couneeted with the central keyboard in such a way that current from B to A through an interposed object shall give deflection of the galvanometer-spot to my right, which I call the ** positive " direction. The compensator and the induction-coil are likewise connected with the central keyboard, with a reverser in each of these two circuits, so disposed that its position towards the right and left indicates “ positive" and “ negative" directions of current. For the induction-coil, this convention is made to apply to the break current. Its reverser is placed in the primary circuit, and the latter also includes a double-contact key ; so that we may take at will either single currents or alternating (tetanising) currents. The direction of the latter is indicated by the position of the reverser ; if it is, e.g., tilted to the right, we know—or can at once verify—that the break current passes through the object under examination in the “ positive” direction from B to A. Finally, I verify the conventional “ positive " and “ negative " direetions between B and A by touching B with a bit of zine, BLAZE-CURRENTS OF VEGETABLE TISSUES. 33 A with a finger of the other hand; when the galvanometer-spot flies off to my right. $2. The galvanometer, being in a strange place, must now be calibrated by means of the compensator. Here are the calibration-numbers taken at the outset of the present obser- vations :— August 4th.—Galvanometer of 7000 w R. Compensator, Deflection. — 0:001 +0:001 — Sem, + 3cm. —0:002 +0002 —6, +6, — 0:003 4- 0:003 —9 , + 9 ,, — 0004 4-0:004 —]12 , +12 , —0:005 +0-005 —15 , 15 , Swing 8 Steady Deflection 2 One double oscillation = 7 secs. Quite at rest in 30 secs., after 4 double oscillations. The instrument is now sufficiently adjusted for our purposes ; the fact that it is not very sensitive (1 em. deflection indicating rather less than 1077 ampère) is an advantage under our present circumstances, A photograph of the Laboratory corner as it stands at the end of the first day shows the apparatus unpacked and set up ready for use (fig. 1, p. 34). A diagram of the circuit is seen to the right; the apparatus itself—galvanometer, seale, shunt, electrode-holder A B, cells, compensator, coil, reversers, keys, and, in the foreground, the keyboard—is arranged on the packing-case, and is not much more bulky or troublesome to pack and unpack than a dressing- bag. As a travelling nécessaire, I regard it as quite indis- pensable ; and it serves to demonstrate that electro-physiology can be studied elsewhere than in beautifully-arranged Institutes of Physiology. I have often described the use of the various parts of the apparatus; and we shall see in the next few days what can be done with it. I shall not now redeseribe it; all I shall do will be to say that the photograph exhibits two additional pieces on the travelling-box, to the right and left of the electrode-holder (which are not represented in the diagram), and to make a remark intended to guard against any misunderstanding as to the * voltage " of physiological currents. “LINN. JOURN.—BOTANY, VOL, XXXVIT. D 34 Paty agua vi sn POTU ZPN A eS / [prier peni pan prn nang] = [o2 E [n S s o! LI oz ‘7, ‘BY urezis e[qiZe[ Jo poonpoaded st 3y “MOTEG e[Q*j eq uo pesduv1e enqvaedde og} jo uv[d-punoaS v seats Jous09 puvq-jq511 eq) ur weiseip ep, 'sjuv[d-pue[ tra sjuv[d-ves Jo uosrreduioo em) 10] eouvjsur quesead eq; ur : ABpI[OY eprsvos v Jo sesodand 04} 10j dn jos pue xoq s}! woaz poyoudun snjeaudde oy} Jo qdwe2o030qq —[ 3 BLAZE-CURRENTS OF YEGETABLE TISSUES. 35 Compensator. -units e$ téns M o == ? © Le 1/94 EC TM | Il iul i ij UU > Keyboard. The plant is placed between the Electrodes, A, B. Excitation is delivered to it from the Sec. Coil via the Keyboard; the direction of excitation is deter- mined by the Reverser R, in the primary circuit, The Key S'T in that circuit serves to deliver single or tetanising currents. The Compensator serves to counterbalance any incidental plant-current and to standardise the deflection ; the direction of its current is determined by the Reverser R,. The Galvanometer, by the movements of a spot of light on the transparent scale, indicates the direction of exciting current or currents, and the direction and magnitude of the responsive or blaze-currents. D 2 36 DR. A. D. WALLER ON THE $3. Of the additional pieces of apparatus just referred to, the first (marked R) is for the purpose of letting a blaze-eurrent into the galvanometer-circuit at a short and regular interval after excitation by a break induction-shock. The other (marked B A C) is for the purpose of taking the separate post-anodie or post-kathodie effects at A or at B, after excitation ihrough A and B. By a movement of the switch, contact at one or other of these excited points is replaced by a contact C with an indifferent point. $4. The purpose of a standardising deflection is two-fold :— A deflection taken before and after experiment serves to show whether the resistance has altered, and how much. The currents observed in different experiments are, to some extent, brought to a common denominator when they have been referred to and expressed in terms of a standard voltage. I say “to some extent," for it is evident that a reduc- tion to zero by a compensating current, or a comparison of a blaze deflection with a standardising deflection, can never give us the true electromotive value of our physiological currents. We do not know how much of our mass is actually active, nor in what proportion the mass offers a shunt to the galvano- meter. Nevertheless, with this distinct understanding, that a de- nomination by voltage does not signify that we have measured a physiological voltage, it is evidently better to specify effects in terms of voltage than in terms of eurrent. $5. Seconp Day.—The first object I have in view is to compare land-plants with sea-plants as to their blaze-currents ; so I visit the garden to select some convenient land-vegetable from whieh to obtain a constant supply of fresh material: c'est lembarras du choix; finally I select peas as my principal vegetable, although I foresee that it will be quite impossible to remain attached to peas alone. $ 6. I visit the sea-shore at low tide, to select some convenient sea-vegetable from which to obtain a constant supply of fresh material: the choice is not so various. Finally I select a sea- weed, the name of which I do not know, but which exists in luxuriant abundance attached to the rocks at half-tide, and is charaeterised by quantities of olive-shaped ovaries, somewhat like pea-pods. They will therefore afford a quite satisfactory BLAZE-CURRENTS OF VEGETABLE TISSUES. 37 term of comparison with my land-peas. I call them sea-grapes for the present *. $7. First Pair of Experiments—The half-ripe pod of a "land-pea? (Pisum sativum) is placed symmetrically between the electrodes A B, to be submitted to the blazo test. Its accidental current is compensated. A break induction-shock in the positive direction, from B to A, is followed by a blaze-current off-scale in that’ same positive direction. A break induction-shock in the negative direction, from A to B, is followed by a blaze-current off-scale in that same negative direction. In both cases the blaze-current has been homodrome with the exciting current; it has been of quite considerable magnitude, ‘‘ off-scale " signifying in this case “ greater than 0:02 volt”; it was completely abolished after immersion of the pod in hot water. Deflection. 0:001 volt through galvanometer .................. sees 40 0:001 ,, » » -Felectrodes........................ o0 0:001 ,, » » » +pea-pod ......... 0:9 0:001 The same after the double blaze .......................- 1:5 The same after tetanisation .............. esse 1:7 The same 4 hour later ............ eese 2 $8. A “ sea-grape ” is placed symmetrically between the same pair of electrodes A B, to be submitted to the blaze test. Its accidental current is compensated. Break induction-shocks in the positive and in the negative directions do not arouse any blaze-current. The conductivity of the pod is not appreciably altered by the induction-shocks nor by strong tetanisation. § 9. Commentary.—This first pair of experiments, which were several times repeated with precisely similar results, show in the clearest possible manner a remarkable contrast between a land- plant and a sea-plant. And obviously the next step will be to test other land- and sea-plants for this contrast, and learn whether it is justifiable to say that land-plants give blaze- currents, but sea-plants do not. * Prof. Farmer has informed me since that they were probably Asco- phyllum, 98 DR. A. D. WALLER ON THE $10. From a second visit to the garden and to the sea-shore respectively, I bring back two bundles—one of land-plants, the other of sea-plants. Everything from the garden gives me blaze-currents; nothing from the sea-shore does so. I am entitled to say that land-plants blaze, and that sea-plants do not blaze. I do not at this stage commit myself to any quantifying prefix to the subjects of these two propositions ; more extensive trials will be required before we shall venture to say whether only “some,” or ** most," or “all”? land- and sea-plants do or do not blaze. $11. So we set ourselves the task of finding, on the one hand, a land-plant that does not blaze, and, on the other, a sea-plant that does blaze. We hardly expect to find a hard- and-fast “yes” or “no” in every case we try, but rather a differenee of degree, perhaps; so we shall take note of the voltage uf the blaze-currents, by comparing their galvanometric effects with the effect through the same circuit of a known standard voltage. Of course this will not mean that we have arrived at any true electromotive value of the blaze-current, but it will be better than nothing. § 12. I think of water-plants, and ferns, and fungi as possibly non-blazing land-plants. So I fetch some watercress, and leaves of hart's-tongue fern, and a fungus from a tree-stump in the garden. $ 13. I begin with the fungus :— 0'001 volt gives a deflection............ = 25mm. Exc, by Br+10000 gives homodrome blaze ...... = +250 » Br—10000 ,, nothing at all. » 2Br4-10000 ,, a smaller blaze » Br—10000 ,, nothing at all. The fungus evidently gives a blaze-current; but beyond this bare fact there are already several points of detail that must arouse reflection and provoke further study. The blaze-current is only in one direction, from B to A, “ positive," and aroused only by current in that same positive direction. On looking to see how the fungus is orientated between the eleetrodes, I find that the dorsal surface is on B and the ventral surface is against A; so that the blaze just witnessed has been from dorsal to ventral surface. It has been aroused by a break induction-shock in the same direction, having therefore the anode at the dorsal surface; it has not LE ennt =+ 60 I BLAZE-CURRENTS OF VEGETABLE TISSUES. 39 been aroused by a shock in the opposite direction, with the kathode at the dorsal surface. From these points I provisionally infer that the dorsal surface of my fungus is more excitable (leistungsfähig) than the ventral surface, and that the exciting current has been more effective at the anode than at the kathode. But we had better not go off into a discussion of this point now. $14. I take a leaf of hart's-tongue, lay it on its face, and lead off from its dorsal surface by the electrodes A and B applied to the midrib 4 or 5 centimetres apart. The trial comes out as follows :— mm 1/100 volt........... essen 20 Exe. by break 100004 ................ esse 1004- » » — eee eene 80-- Strong tetanisation. Exe. by break 100004 ................. se. trace-+ » » LEM nil 1/100 volt.................. eese 30 Three points are seen in this experiment: (1) that this leaf manifests strong homodrome blaze-currents; (2) that these eurrents are abolished by strong tetanisation; and (3) that the resistance has been considerably reduced. $15. And now, again, questions press upon us. Is the effect of tetanisation temporary or permanent? Will a leaf give a succession of blaze-currents of diminishing or undiminishing magnitude? What relation is there between strength of excitation and strength of blaze? What are the time-relations ? What is the cause of the diminished resistance ? All these questions clamour for their answers ; but they must wait. $ 16. The next object I undertook to test was some duckweed. The results of its trial are as follows :— 1/100 volt through electrodes. » ” n and a leaflet. Blaze after break-shock 10000-4 ............ is + LE "n LE LCLLLLLLLLLTTLITT After strong tetanisation :— Blaze after break-shock 10000+ .......... 4 oni 1 2» 305 0000000 cs99538859835 959 It is not a convenient plant to work with, because the leaf is so small. I try whether a mass of overlapping leaves give blaze, and find that they do not. 40 DR. A. D. WALLER ON THE $17. Clearly land-plants are more interesting than sea-plants. But the contrast between the two kinds, whatever its meaning may be, is a fact of primary importance. I want to confirm it, or the reverse, by every means in my power. So I offer at the next meeting of my family to give two prizes—one for any blazing sea-plant, the other for any non-blazing land-plant. I make the proviso as to the latter that it must be a plant capable of conducting an electrical current, not a non-conducting dry leaf or bit of wood. I also stipulate, as conditions of the prizes, that the winning land-plant shall give less than 0°001 volt, the winning sea-plant more than 0:010 volt. Many things are brought to the galvanometer in a short space of time, from the garden and from the sea-shore. Nothing from the garden fails to blaze ; nothing from the sea-shore gives any- thing like a blaze. "True, a holly-leaf from the garden gave no clear blaze; but its electrical resistance was so great that current sufficiently strong to excite could not traverse it. And, on the other hand, a leaf from among the seaweed which did give a blaze was pronounced to be an intruder from a neigh- bouring stream. The competitors abandoned their quest as hopeless, being even more convinced than I am that land-plants blaze, while sea-plants do not blaze. 1 am indeed satisfied that sea-plants do not blaze anything like land-plants ; but I am not sure—indeed, I think it quite unlikely—that no sea-plants give any blaze at all *. $18. An Ivy Petiole versus a Geranium Petiole—All land- plants are not equally vigorous as regards their electromotive effects ; some are more vigorous than others. I found, a year or two ago, that seeds lose power in this respect by mere lapse of time. An ivy-leaf petiole as compared with a geranium-leaf petiole, under physical conditions as nearly as possible identical, affords a convenient illustration of the general principle that amount of blaze bears some general relation to degree of vitality or vivacity. * Ultimately a long narrow seaweed, called “ boot-laces" by the fishermen, Chorda Filum by its museum name, was brought in, and gave typical homodrome effects in both directions, amounting to as much as 0:02 volt. The peculiarity is possibly due to the transverse septa subdividing the length of weed. Another jointed seaweed coralline was subsequently found, giving about OOl volt. BLAZE-CURRENTS OF VEGETABLE TISSUES. 41 The ivy petiole is comparatively sluggish; the geranium petiole is comparatively lively. Under similar conditions, the response or blaze of the former is below 41, volt, while that of the latter is upwards of 45, volt. $19. Turg» Day.—Some of the questions asked in $ 15 require to be answered, and I shall make use of peas for the purpose. Peas in their pod afford us an unlimited supply of individual test-objects, similar, and protected from accidental injury. The relation between strength of excitation and strength of blaze-current, and the time-relations of the latter, are easily examined. The observations have to be made in a systematic and leisurely fashion; so I have reserved them for a wet day, offering no external temptations. Fig. 3.—Pxa (Pisum sativum). Excitation by a single break induction-shock from A to B. Blaze-current from A to B, i. e. homodrome. 1 1 1 | 1 fi L L | 1 L 1 ji 0 ! 2 3 4 5 10 minutes. The ordinates are in scale-divisions of 1 cm.; with the resistance in circuit, each 1 cm. indicated an approximate voltage=1/1000 volt. § 20. The observation of each individual blaze-current requires 5, 10, or 15 minutes, according to its strength. A blaze aroused too soon after a previous blaze is smaller than normal. One should allow the zero to be recovered of itself, and not impa- tiently return to a false zero by manipulating the compensator. 42 DR. A. D. WALLER ON THE This rule has been observed throughout the series of to-day's readings. In the first series I have taken readings every half-minute and every minute after excitation of five different strengths, omitting in each case the very first deflection, and beginning at the first half-minute after excitation. The series of observations has lasted 13 hour, without appre- ciable “ fatigue ” or alteration of resistance. The dotted curve gives a second blaze by 1000 at end of series. Fig. 4.— Relation between strength of Stimulation and magnitude of Blaze. . Volt. V 9.02 4 peer TT Ort Bow err cn cmos cence te ee een e D E S F Soo, ^ a ‘ 1 L i L L ] 1000 2000 3000 4000 5000 10000 Stimulation. Fig. 5.- -Ditto, ditto. Volb. ce æ- . < t=- -- --2--- - -- -- 7579 a 0-0247 r wq) r c | a f good | œ ] L L L L L J 1000 2000 3000 4000 5000 10,000 Stimulation. In the above two series (figs. 4 & 5) the readings taken are those of the galvanometer-swing produced when the blaze is let through by means of the key R, immediately after passage of the exciting current. The effects are homodrome, 7. e. in the same direction as the exciting current; and the range from minimal to maximal is not a very extensive one—from, e. g., 600 to 2000 units of a Berne inductorium fed by two Leclanché cells. BLAZE-CURRENTS OF VEGETABLE TISSUES. 48 $21. In dealing with animal tissues it is very difficult, if not impossible, to get two isoelectric contacts when the electrodes are first applied; there is always more or less accidental current, by reason of unavoidable injury or from other causes. Precisely the same difficulty presents itself in the case of vegetable tissues. It is advisable to leave them undisturbed in contact with the electrodes for some time before an experi- ment is begun; at the end of an hour they may generally be considered to have settled down to a stable state. $22. The “current of injury," whether accidental or provoked, is very considerable ; it gradually subsides, as in the case of animal tissues, and can then be renewed by renewed injury. The following curve gives the decline of E.M.F. of a pea, led off by a freshly-cut and an intact surface. Fig. 6.—Pza (Pisum sativum). Decline of Current of Injury after a preliminary rise. Volt.) — 0-074 0:06- 0-054 0-054 0:024 0:014 L L 1 1 L i fi 1 1 1 L 1 1 L GO 5 10 i5 20 25 30 Time, o 1! hour. l B 25. At first—z.e. for the first 5 minutes—the current of injury increases. This preliminary increase, amounting to 0:005 to 0010 volt, calls to mind the similar increase recently pointed 44 DR. A. D. WALLER ON THE out by MacDonald and Sowton in the case of the current of injury of mammalian nerve. Clearly, however, the explanation invoked by these authors for their case—viz., augmentation of concentration-current by cooling—will not apply to the present case. I have no very dogmatic explanation to offer; it looks to me very like an imbibition current at a eut surface. § 24. A blaze-current provoked in a vegetable organ during its manifestation of a current of injury is, in general, opposite to it in direction, whatever may have been the direction of the exciting current. Du Bois Reymond would probably have called it a negative variation of a previous current. Blaze-currents caused by excitations of moderate strength at sufficiently long intervals in the course of a declining injury- current are of diminishing magnitude. The smaller the injury- current, the smaller is its negative variation. Fig. 7.—PA (Pisum sativum), Negative variations of Current of Injury. Volt 0-05 Q:O4— 0-034 one O:01- 6-01 = /\ L | 1 L | L L ] f ihour D TI § 25. Fourra Day.—We intend to find out to-day what sort of magnitudes the currents of different peas present: whether they differ at different stages of maturity, whether individual peas of the same pod vary more or less than peas of different pods, and so forth. $26. I begin with a comparison between two very different pods: one quite ripe and fully distended, of which the cooked BLAZE-CURRENTS OF VEGETABLE TISSUES. 45 peas would be rather hard and passés ; the other immature, with tender young peas at, I should say, a quarter to half their normal bulk. The contrast, galvanometrieally, is very striking indeed. The adult peas give large current of injury and large blaze-current. The infant peas give small and irregular current of injury and no blaze-current to speak of. 7 Peas. A mature pea-pod, taken from S acne cie nil nil -0100 nil $27 before backwards. Current of Blaze by 1000— «~ injury. — (1/100 volt — 13 em.). volt. cm, .. +0:0210 —5 .. +0°0430 —]11 . 4-0:0450 > — 20 .. +0:0500 —]14 .«. +0:0520 > —?20 ... +0:0570 > —28 ... +0°0430 —13 nil +3 nil 6 Peas. An immature pea- pod. Ourrent of Blaze by 1000 — «— injury. — (1/100 volt — 13 em.). volt. em. — 0:0150 nil — 0:0060 nil +0:0100 —2 +0-0040 -1 — 0:0080 nil 4-40:0100 nil nil +3 (2) +0010 -5 (7) +0050 -5 (4) . I think the result is of considerable importance. If 46 DR. A. D. WALLER ON THE there has been no mistake—and it will be an easy matter to make sure of this by further trial—we have in our hands a very clear and convenient field within which to test a relation between physiologieal states and electrical responses. $28. Asa rule, the middle peas of a pod give the largest blaze- currents; the peas at both ends, being as a rule less developed than the others, give smaller effects, if any. We may pick out pods more or less completely ripe, and find only the middle peas giving eurrents. And it is interesting in this connection to note that the rule applies to injury-currents as well as to blaze-currents. $99. Firra Day.—Young peas have a truly enormous elec- trical resistance ; a tenth of a volt through pea and galvanometer in series gives a hardly appreciable deflection. This is not merely a skin-resistance, for it persists after a slice has been taken off each end of the pea; it is a resistance attributable to deficiency of electrolytes. An adult pea is a much better conductor; an adolescent pea is of intermediate conducting value. The resistance is greatly reduced in consequence of a single electrical shock; and I find reason for believing that the alteration is partly due to the multiplication of electrolytes by dissociation. Alterations of conductivity of old and of young seeds in consequence of the passage of induction-currents. Full-grown Peas. Young Peas. n. After one After nu After one After Initial break- tetanisa- | Initial break- tetanis: ductivity. res etanisa-| conductivity. Ae etanisa- con J- shock. tion. | Y: shock. tion. 50 70 90 | 2:95 22-5 55 40 70 70 | 0:5 25 50 60 70 75 25 22:5 65 70 75 05 30 55 45 4hr. later. 45 1 hr. later. 65 after tetanis. 50 4 hr, later. 60 after tetanis. 55 4 hr. later. 65 after tetanis. Next day 85 100 after tetanis. 95 1 hr. later. BLAZE-CURRENTS OF VEGETABLE TISSUES. 47 $30. A sufficiently obvious fallacy may arise from this source. Suppose we have a quite young pea on the electrodes, with a fresh-cut surface at B, so that the current of injury is positive: the resistance is very high, the deflection is small, and we have to use a considerable voltage from the compensator in a negative direetion in order to bring the spot back upon zero. A strong induetion-shoek is now sent through the pea in the negative direction, the after-effect of which is a large deflection in that same direction. We find that the resistance is greatly diminished, and the large deflection to the left may evidently be due to our large compensating voltage. While the resistance was high, we might easily have over-compensated; so that the over- compensation would come into evidence when the resistance was reduced. Of course, to do this, the alterations of resistance would need to be very great. And they are so,—it is not uncommon to see the resistance reduced to yy of its original value by a single strong induction-shock. $ 31. I have seen the fallacy in another form. A young pea, with its base cut, has a current of injury in the positive direction which, by reason of the enormous resistance, gives a barely appre- ciable deflection. A single induction-shock is sent through the pea in the positive or in the negative direction; an apparent blaze-current is manifested in the positive direction, which, how- ever, I do not consider to be a true blaze-current, but only a deflection due to the positive current of injury through a suddenly diminished resistance. $32. The different behaviour of young peas and old arouse many reflections. The old are conductors ; the young are non- conductors. Conduction is by electrolytes. Salts—or their ions—are electrolytes par excellence. Are there more salts in old peas thanin young? And what is the meaning of Loeb’s recent investigations into the influeuce of salts upon the development of ova? I have no means of reply; so the questions must wait. $33. The Holly-leaf expertment.—Here is a curious ex- periment on resistance, with a holly-leaf. I say curious, because it has aroused my euriosity, and I cannot rest satisfied with the explanation that offers itself. The leaf is placed between the electrodes as usual, in series with the galvanometer and secondary coil. I send an induction- 48 DR. A. D. WALLER ON THE shock through the cireuit—first a make, then a break ; the make deflection is comparatively small, the break deflection compara- tively large. And now, if alternating make and break shocks are passed, the galvanometer-spot flies off in the direction of the break. The familiar physiological effect known as von Fleisehl's deflection is exactly simulated. But it is not a physiological effect in this case; for it comes off on a boiled holly-leaf and on a thick piece of glazed note- paper. The irresistible conclusion is that we have to do with a case of variable resistance, smaller to the break than to the make shock, as is said to occur in the passage of alternating currents through a vacuum-tube. Whether this be a valid explanation or not, the fact itself is well worth our notice, lest we should confuse a simple physical phenomenon independent of the living state with a similar physical phenomeuon characteristic of that state. § 34. Sıxru Day.—We have reached our last day, and have made little or no use of the accessory piece of apparatus (B A C, § 3) set up on the first day. The apparatus has not “ earned its living"; and though we may not hope to make it do so in a single day, we may at least do something to justify its inclusion in the travelling-box. We shall use it to-day for two or three localisation trials, preliminary to a more systematie inquiry next year. $35. Experiment to deterinine the direction and magnitudes of the separate electromotive changes at the two poles of an exciting current. B C A — L L 1 =f. Excitation by a single break induction-shock 10,000—, from A to B, 7. e. so that A is anode and B is kathode. Compensation established between A and B, B and C, and A aud C, before and after excitation. -B I — A+ C BLAZE-CURRENTS OF VEGETABLE TISSUES. 49 The compensation-values are :— Changes Before. After. Difference. in the Pea. Between A and B...... —0:0090 +0:0030 4-0:0060 — 0:0060 " B and C...... — 0:0240 — 0:0450 — 00210 +0:0210 » A and C...... +0:0140 +0:0490 +0°0850 — 0:0350 Exc? z210 B A 350 140 The events in the pea itself have been as follows :— Excitation from A (anode) to B (kathode). Partial blaze at A ..... MEME = —('0350 Partial blaze at B ..................... = -FU0210 Their algebraic sum ......... ees = —0'0140 Total observed | .................. e. = —(0:0120 i. e., both partial blazes have been ingoing ; the total blaze has been homodrome with the exciting current; the post-anodic homodrome blaze has been greater than the post-kathodie antidrome blaze. Excitation B < C A Total blaze « 0:0120 Post-kathodie partial blaze —20210_,, „2:0350 Post-anodic partial blaze §36. Experiment to determine the alterations of resistance separately at the two poles of an exciting current.—Excitation by a single break induction-shock, 10000 — , from A to B. Deflection observed by jl;th volt between A and B, B and C, and A and C, before and after excitation. The conductivities are as follows :— Before. After. ^ Oonductivity. Between A and B ......... 2 20 x10 » B and € ......... ll 13 x 12 » A and C ......... 15 10 x 67 Between À and B ......... 9 38 x 42 » B and C ......... 8 9 x l'l » A and C ......... 5 10 x 20 LINN. JOURN.— BOTANY, VOL. XXXVII. E Mo. i ot. arac ia 1908 50 ON THE BLAZE-CURRENTS OF VEGETABLE TISSUES. These two observations were taken on a young pea; the next is on an old pea :— Before. After. Between A and B ......... 95 100 » BandC......... 75 80 » A and C ......... 75 75 The conductivity is increased at the anode and at the kathode, more so at the former than at the latter. | i UJ. db E I I E $37. Experiment on a geranium-stalk to see whether the alteration of resistance affects the substance as well as the skin, whether it is interpolar as well as polar.— Deflection by -hth volt through II before and after tetanisation through E E: Before Exe. After Exc. II .........e- 2:5 65 EE ................- 15 50 i. e., the interpolar region I I is affected by the excitation through E E; the alteration is not an exclusively polar effect. The effect of excitation II upon the resistance EE is, of course, much less pronounced : Before Exc. After Exc. II sesers 1:75 12 EE ............... 25 2 INTERNATIONAL CATALOGUE OF SCIENTIFIC LITERATURE. Tne Royan Society has been engaged for some years past in arranging for the publication of an International Catalogue of Scientific Literature, beginning from the lst January, 1901. Each science will be represented in an annual volume containing lists arranged under authors and subjects, of all books and papers published during the year; these will be contributed through official channels of information—abroad, by direct control of the respective governments—at home, by means of the various Societies which devote themselves to particular sciences ; those Societies whose domains over- lap having arranged for mutual cooperation. The collection of title-slips for the United Kingdom of Great Britain and Ireland as regards BOTANY has been undertaken by the Council of the Linnean Society, and they appeal to all botanic workers for support in their endeavour to compile a complete record, by sending notices promptly of all botanic issues to the undersigned. The second volume, for 1902, is now in the press. B. DAYDON JACKSON, General Secretary, Linn. Soc. NOTICE. Vor. XXXV. has been completed by the issue of Part 248. Vor. XXXVI., of which Parts 249-254 are already published, is reserved for the completion of the Chinese Flora. Attention to this announcement is specially requested, to prevent application to the Librarian for unpublished Parts. The CarALoavE of the LrsRARY may be had on application. Price to Fellows, 5s. ; to the Public, 10s. All communieations relating to the general business of the Society should, as heretofore, be addressed to the “ SECRETARIES,” but letters on library business only should be addressed to the * LIBRARIAN.” ed: JULY 1. Price 6s. THE JOURNAL OF THE LINNEAN SOCIETY. Vor. XXXVII. BOTANY. No. 258. . CONTENTS. Page I. Some Notes on the “ Sudd "-Formation of the Upper Nile. By A. F. Broun, Director of Woods and Forests in the Soudan. (Communicated by C. H. HR OU eeeriete hh COR E e 51 II. Descriptions of some New Species, and Notes on other Chinese Plants. By W. J. Turcuer, F.L.S. ......... 58 III. A Revised Classification of Roses, 1905. By Joun GILBERT BAKER, F.R.S., F.L.S. —...................... Ei IV. Contributions to the Flora of Liberia. By Dr. Orr uE AE o ue. de i 79 See Notice on last page of Wrapper. LONDON: SOLD AT THE SOCIETY'S APARTMENTS, BURLINGTON HOUSE, PICCADILLY, W.; AND BY LONGMANS, GREEN, AND CO, AND WILLIAMS AND NORGATE. 1905. LINNEAN SOCIETY OF LONDON. LIST OF THE OFFICERS AND COUNCIL. Elected 24th May, 1905. PRESIDENT. Prof. W. A. Herdman, D.Sc., F.R.S. VICE-PRESIDENTS. Charles Baron Clarke, M.A., F.R.S. Horace W. Monckton, F.G.S, Frank Crisp, LL.B., B.A. Dr. A. Smith Woodward, F.R.S. TREASURER. Horace W. Monckton, F.G.S. SECRETARIES. D. H. Scott, M.A., Ph.D., F.R S. | Rev. T. R. R. Stebbing, M.A., F.R.S. CENERAL SECRETARY. B. Daydon Jackson, Esq. COUNCIL, R. Assheton, M.A. Prof. F. W. Oliver, D.Sc., F.R.S. V. H. Blackman, M.A. Clement Reid, F.R.S. Gilbert C. Bourne, D.Sc. Dr. A. B. Rendle, M.A. Dr. Horace T. Brown, F.R.S. Dr. W. G. Ridewood, F.Z.8. Charles Baron Clarke, M.A., F.R.S. Dukinfield H. Scott, M.A., Ph.D., Frank Crisp, LL.B., B.A. F.R.S. Prof. J. B. Farmer, D.8c., E.R.S. David Sharp, F.R.S. Rev. Canon Fowler, M.A. Dr. Otto Stapf. Prof. W. A. Herdman, F.R.S. Rev. T. R. R. Stebbing, M.A., F.R.S. B. Daydon Jackson, Esq. A. Smith Woodward, F.R.S. Horace W. Monckton, F.G.S. LIBRARIAN. CLERK, A. W. Kappel. A. R. Hammond. LIBRARY COMMITTEE. This vonsists of nine Fellows (three of whom retire annually) and of the officers ex officio; the former are elected annually by the Council in J une, and serve till the succeeding Anniversary. The Committee meets as required during the Session. The Members for 1904-1905, in addition to the officers, are :— Horbert Druce, F.Z.S. A. B. Rendle, M.A., D.Sc. Antony Gepp, M.A. A. G. Tansley, M.A. W. Botting Hemsley, F.R.S. Prof. H. Marshall Ward, F.R.S. Dr. G. Henderson. Dr. A. Smith Woodward, F.R.8, P. Chalmers Mitchell, M.A., F.Z.S. ON THE “SUDD "-FORMATION OF THE UPPER NILE. 51 Some Notes on the “ Sudd "-Formation of the Upper Nile. By A. F. Brow, Director of Woods and Forests in the Soudan. (Communieated by C. H. Wniaur, A.L.S.) [Read 3rd November, 1904.7 Mvcnm has already been written about the vegetation of the marshes of the Upper White Nile. Sir William Garstin's admirable report on the Bahr el Jebel, published in 1901, not only describes the chief constituents of the “ Sudd,” but also the circumstances in which vast masses of floating vegetation are moved hither and thither and block the waterways, by forming dams (Arab “ sudd”) across them. A trip in a sailing-boat undertaken in 1903, which kept me for about four weeks in that country, and another shorter trip recently made, have enabled me to add the following notes, which may be of interest as tending to show the process of Sudd-formation. The mouth of the Bahr el Jebel near Lake No, and at 627 miles by river above Khartoum, may be taken as the northern gate of the Sudd-region—a vast country of swamps, which stretches westward across the mouths of all the tributaries of the Bahr el Ghazal, which, with a stretch of the White Nile joined at Lake No, roughly form the northern boundary. Eastward the swamps reach in their northern portion as far as the some- what higher land between the Bahr el Jebel and the arm of the Nile known as the Bahr el Zeraf. To the south of the spot where this arm leaves the Bahr el Jebel, at about 240 miles by river to Lake No, the swamps stretch to the east perhaps as faras the Upper Sobat. Southward the swamps extend to beyond Gondokoro, but they are of lesser width at Bor, a Dinka village about 380 miles by river from Lake No. In the Sudan, when speaking of the “Sudd,” it is generally understood that the portion of the river between Lake No and Shamba is meant. Shamba is a military post situated on the left bank 266 miles from Lake No. It was in this section that, after the overthrow of the Dervish rule, the channel was found to be blocked at several places by masses of floating vege- tation, the clearing of which has cost the Egyptian Government considerable sums of money, and the Sudan Government the health and lives of many of its employés. In this portion the LINN. JOURN.—BOTANY, VOL. XXXVII. F 52 MR. A. F. BROUN ON THE river twists and turns through swamps which often stretch as far as the eye ean reach from the top of the mast of a sailing-boat or from the upper deck of a steamer. Near many of the bends sheets of water, sometimes several square miles in extent, are to be seen, and glimpses of other subsidiary channels are also obtained. Here and there, especially in the southern portion of the Sudd, lines of trees are visible in the far distance, showing the limit of the swamps. In places, spits of low land dotted with termite-hills, and with some trees and bushes, reach into the swamps close to the main channel. There are also occasional islands with similar vegetation; one of these is at Hillet en Nuer, about halfway through the Sudd, and it served as base camp for the recent Sudd operations. A short way to the south of this, the clearing of one block has still to be completed, and boats have now to make their way through a chain of large lakes. Near Shamba the left bank, covered with an open forest of large trees, comes to within a mile or so of the main channel. The main channel is usually marked by a constant string of floating vegetation, whieh may consist of solitary Pistia Stratiotes, Linn., of masses of Cyperus Schimperianus, Schrad., and other sedges, or of large or small clumps, usually varying according to the state of the weather, of Papyrus, Panicum pyramidale, Lam., or Typha australis, Schum. & Thonn., the most common being twisted strands of rhizomes of Panicum pyramidale, enclosing smaller floating weeds such as Azolla nilotica, Decne., and Pistia. Starting from Lake No, the channel is, for many miles, lined on both sides by a continuous border of Papyrus, dotted here and there by occasional clumps of Ambatch (Herminiera Ela- phroxylon, Guill. & Perr.). This lining of tightly packed shoots, some 10 to 15 feet high, is, in places where water is of the right depth, of considerable thickness, but in shallower water large stretches of Typha are visible behind the Papyrus hedge. These stretches are dotted here and there with clumps of Papyrus and Ambatch filling the deeper pools. In this portion of the Sudd region Panicum pyramidale is not very common. On inspection of the Papyrus fringe near the channel it will be found that the mass is tightly bound together by numerous twiners and climbers, the most common being Vitis ibuensis, Baker, Vigna nilotica, Hook. f., and three species of Ipomea, viz. I. Lindleyi, Choisy, I. palmata, Forsk., and J. reptans, Poir. The last named has a hollow-jointed rhizome, which floats on the “ SUDD "-FORMATION OF THE UPPER NILE. 53 surface and by means of its shoots binds together those given out by the rhizomes of other plants. Among the Cucurbitacez, Luffa egyptiaca, Mill., and Adenopus breviflorus, Benth., are also common. Amidst the tangle of rhizomes of the Papyrus other plants also spring up, and their roots and the soil which they gather fill up the interstices of the mass. The most showy of these are Hibiscus diversifolius, Jacq., and Melanthera Brownei, Rohr. The little bays and inlets and placid pools are covered with Pistia, Azolla, Trapa bispinosa, Roxb., Hydrocotyle natans, Cyr., Utricu- laria Oliveri, Kam., Jussiea diffusa, Forsk., J. pilosa, H. B. & K., Ceratophyllum sp., Nymphea Lotus, Linn., sedges, and others enumerated in the list at the end of these notes. Towards Hillet en Nuer the banks of the channel are often better defined, and Papyrus is frequently replaced by the tall reed Phragmites communis, Trin., while Panicum pyramidale, Lam. (called “ Om-Suf,” or “ Mother-of-wool,” by the Arabs on account of the irritant hairs at the base of the leaves), which has the faculty of growing both in shallower and in deeper water than Papyrus, not only occupies the land which is uncovered or nearly so at low Nile, but forms a fringe in front of the Papyrus in the bed of the channel. South of Hillet en Nuer Phragmites is still more common, and is festooned by the same twiners mentioned above, the most abundant being Ipomea palmata. The islands and spits of land, which are only covered at high Nile, are usually grown over with thick grass and sometimes with open woods of small Acacia Suma, Kurz, or with scattered trees, frequently growing on the top of termite-hills, chiefly Cratæva religiosa, Forst., with some Borassus Jlabellifer, Linn., Hyphæne thebaica, Mart., Euphorbia Candelabrum, Trém., and occasional Kigelia æthiopica, Decne., and Trichilia emetica, Vahl. On the somewhat higher ground afforded by the termite-mounds are also found a pretty white-flowered shrub, Clerodendron near neriifolium, Vahl, also Capparis tomentosa, Lam., Sanseviera guineensis, Willd., Withania somnifera, Dun., and bulbs of the showy Hemanthus multiflorus, Martyn. The Ambatch is much more abundant in the southern portion of the Sudd, and forms woods of some extent, usually lining channels of lakes. Above Shamba the course of the river is, for some distance, even more tortuous. At first the bends keep nearer to the left F2 54 MR. A. F. BROUN ON THE bank, but above Kenissa, the site of the old Austrian Mission of Ste. Croix mentioned by Sir S. Baker, the course gradually approaches the right bank, which only becomes visible far in the distance some way above Shamba. At Bor it actually flows under the high right bank. Above this point the main channel goes from one bank to the other and back again, and, although the marshes between the high banks on either side are still of con- siderable width (at least as far as Mongalla, our southernmost post, 474 miles above Lake No), the boundaries of the Nile Valley above Bor are well defined. As regards the vegetation of this part, it is striking that some way above Shamba the banks begin to be better defined and that Papyrus gives way gradually to Om-Suf, the former being, in the southern swamps, only represented by clumps or belts denoting depressions in the surface of the ground. The grass-covered surface is dotted here and there with trees, either solitary or in clumps, chiefly Crateva religiosa, Forst., Kigelia æthiopica, Deene., and Ficus Sycomorus, Linn., the low banks being often covered with fairly dense fringes of Sesbania egyptiaca, Pers. It appears to me most likely that, at some time more or less distant, the Nile debouched near Bor into a large lake, the northern limit of which was marked by the rise of the ground towards the Nuba hills in Kordofan, still marked by the Bahr el Ghazal and the White Nile flowing from west to east as far as the Sobat mouth. It is possible, and even probable, that this lake did not, at first, occupy so wide an area as is now covered by the swamps, but that it gradually became filled with silt, the cone of dejection spreading northwards from Bor. This would account for the present disappearance of Papyrus im the southern portion of the swamps. It is also easy to conceive how, as the waterin the lake became shallower, its surface was covered with vegetation. Among the plants growing on the shores of the lake were some furnished with more or less tubular rhizomes which they sent out over the surface of the water, and, being numerous, crossed and recrossed each other and interlaced, thus forming rafts, sometimes of con- siderable thickness and buoyancy. The most important of these plants are now the chief Sudd plants: they are Cyperus Papyrus, Panicum pyramidale, Phragmites communis, and Typha australis. Of these the Panicum and the Phragmites have not only the lightest and most tubular, but also the longest rhizomes. I have * SUDD "-FORMATION OF THE UPPER NILE. 55 seen the whole channel of the Bahr el Arab choked with rhizomes of the Panicum, while the rhizomes of the Phragmites, floating in the Nile channel, are often 15 to 20 yards long. The rafts formed by these plants were further strengthened by twiners, which bound them and the shoots they gave out together. Among these Ipomea reptans, which, as I have mentioned before, has also tubular rhizomes and is thus able to reach the most distant shoots, is the most important; but the other twiners, as well as other plants, were soon able to germinate in the soil deposited on the surface of the raft. Their roots and the soil which they gathered, as well as floating waterweeds and vegetable rubbish brought by the wind or currents, filled the interstices of the raft and further eonsolidated it. These masses of vegetation dropped their roots into the soft ooze at the bottom of the lake; but a rise in the lake or a storm broke away portions of the rafts, whieh drifted away until they anchored themselves in other shallow places, where they again spread. These movements in the masses of vegetation still take place, and have been fully described by Sir William Garstin. Although Ambatch is, at certain places in the swamps, e. g. the southern part of the Sudd and in the Bahr el Ghazal swamps, very abundant, Sir William Garstin is of opinion that it has no claim to rank in importance, as a Sudd-forming plant, with the species mentioned above. I quite agree with him and believe that it only appears, with other plants, after the rafts of floating vegetation have formed themselves, and that. at the best, it only acts as anchors wbich help in fixing the rafts in one place and thus form firm land. That the soil is slowly forming and filling the bed of these swamps, there is little doubt. The Nile in flood at Gondokoro is well nigh as turbid as the Blue Nile when the latter is in flood ; but a large proportion of this silt gets deposited on the roots of the plants in the swamps, and in still waters the mud sinks to the bottom. Successive layers of decomposing vegetation should add materially in forming soil, and seeds of large riverside treos, washed from above, may germinate and help in the work of consolidation ; but, unfortunately, fierce fires sweep over the whole country and the ashes of the plants are carried awav by the wind and a large portion deposited in the channels and carried away, while what seedlings of trees do germinate are: with few exceptions, killed outright by the fires. 56 MR. A. F. BROUN ON THE The work done by Nature, described above, if sure, is too slow for our progressive times ; the immense volume of water which spreads itself and is wasted in these vast swamps is wanted lower down, to change desert and howling wilderness into smiling fields. Mueh has been done by the great Irrigation Department of Egypt, and among gigantie tasks which still await it is that of draining the swamps of the Upper Nile. Khartoum, 26th April, 1904. List oF PLANTS SEEN IN THE SWAMPS OF THE UPPER NILE. (a) Chief * Sudd”-formers. Cyperus Papyrus, Linn. Panicum pyramidale, Zam (“ Om-Suf.”) Phragmites communis, Trin. Typha australis, Schum. & Thonn, (b) Lwiners and Climbers (all common). Vitis ibuensis, Baker. Vigna nilotica, Hook. f. Luffa egyptiaca, Mill. Adenopus breviflorus, Benth. ? Melothria cordifolia, Hook. f. Ipomcea reptans, Poir. I. palmata, Forsk. I. Lindleyi, Choisy. (c) Accessory “ Sudd "-formers (including water-plants and plants which grow on the floating vegetation). Nymphæa Lotus, Linn. Hibiscus diversifolius, Jacq. Herminiera Elaphroxylon, Guill. $ Perr. Jussiza diffusa, Forsk. J. pilosa, H. B. & K. Trapa bispinosa, Roxb, Hydrocotyle natans, Cyr. Melanthera Brownei, Rohr. Sonchus Bipontini, Aschers. Utricularia Oliveri, am. Polygonum lanigerum, R. Br. P. tomentosum, R. Br. Ceratophyllum sp. Vallisneria spiralis, Zinn. Boottia scabra, Benth. * SUDD "-FORMATION OF THE UPPER NILE. 57 Ottelia ulveefolia, Walp. Commelina nudiflora, Linn., cum var. Pistia Stratiotes, Linn. Lemna sp. Potamogeton aff. pectinato, Linn. Cyperus Schimperianus, Schrad. C. dichrostachys, Hochst. C. latifolius, Poir. C. nudicaulis, Poir. C. articulatus, Linn. Marsilea diffusa, Lepr. Azolla nilotica, Decne, (d) Plants growing on low, periodically inundated land. [Norr.—The trees the names of which are printed in small capitals are often found growing in the midst of the Sudd, and, although now probably fixed, may originally have started growing on the masses of floating vegetation.) Capparis micrantha, Rich. Shrub (scrambling), C. Rothii, Oliver. Shrub. CRATJEVA RELIGIOSA, Forst. Tree. Malachra radiata, Linn. Herb. (Mouth of Bahr el Zeraf.) Harrisonia abyssinica, Oliver. Shrub. Trichilia emetica, Vahl. "Tree. ZEschynomene aspera, Linn. Shrub. Indigofera orthocarpa, Baker. Undershrub. Sesbania egyptiaca, Pers. Shrub. S. aculeata, Pers. Herb. (Bahr el Ghazal.) Psophocarpus longepedunculatus, Hassk. Herb; climber. Canavalia ensiformis, DC. Herb; climber. ACACIA SEYAL, Del. Tree. A. Suma, Kurz (syn. A. campylacantha, Hochst.). Tree. A. verugera, Schweinf. Tree. - Combretum sp. (near aculeatum, Vent.), Scrambling shrub. Mollugo Spergula, Zinn. Herb. Mitragyne africana, Walp. Tree. (On termite-hills, Bahr el Ghazal,) Oldenlandia sp. (near decumbens, Hiern). Herb. Enhydra fluctuans, Zour, Creeping herb. Ethulia conyzoides, Zinn. Herb. (Mouth of Bahr el Zeraf.) Marsdenia rubicunda, N. E. Br. Woody climber. Cordia abyssinica, R. Br., var. Shrub. C. subopposita, DC. Shrub or small tree. Withania somnifera, Dun. Herb. Kigelia æthiopica, Deene. Tree. Asystasia coromandeliana, Nees. Herb. 58 MR. W. J. TUTCHER ON NEW SPECIES Nelsonia campestris, R. Br. Herb. Clerodendron sp. (near neriifolium, Vahl). Shrub. (On termite-hills.) Ocimum gracile, Benth. Herb. Euphorbia Candelabrum, 7rém. Tree. ? Fluggea microcarpa, Blume. Shrub. (On termite-hills.) Ficus Sycomorus, Zinn. Tree. F. capreefolia, Del. Shrub. Celtis integrifolia, Zam. Tree. Sanseviera guineensis, Willd. Herb. (On termite-hills.) Hemanthus multiflorus, Martyn. Herb. (At foot of termite-hills. ) Hyphzne thebaica, Mart. Tree. BORASSUS FLABELLIFER, Linn., var. HTHIOPICA, Warb. Tree. Juncellus alopecuroides, C. B. Clarke. Herb. Panicum (cf. $ Brachiaria). Deseriptions of some New Species, and, Notes on other Chinese Plants. By W. J. Tureen, F.L.S. [Read 2nd February, 1905.] THE species mentioned in the following paper were found on the island of Hongkong, with two exceptions—one from Kowloon, and the other from Wei-hai-wei. NOTES ON THE FLORA. The island of Hongkong lies just within the tropics, about 22° North latitude and 114° East longitude. It consists of irregular granite mountain-ridges, the principal of which lies east and west, and is broken up into several peaks—Mt. Victoria in the west being the highest, 1800 feet, whilst Mt. Parker in the east is not much less. The area of the island is about 29 square miles. There are two well-marked seasons—the rainy from May to October, and the dry from October to May. The average rainfall is about 85 inches a year, and most of this falls during the south-west monsoon; the dry season corresponds to the north-east monsoon. The hills are intersected with numerous ravines, and it is in these ravines that the vegetation is richest, more especially in the ravines on the north side of the island. There are one or two exceptions to this, as the woods of the Happy Valley on the north-east and those of Little Hongkong on the south side prove. Approaching Hongkong from the south, one is struck with the apparent bareness of the place, and it is not AND OTHER CHINESE PLANTS. 59 until one gets near that this idea is dispelled. Most of the vegetation, however, which strikes the eye is due to the work of the Afforestation Department, as millions of Pinus Massoniana have been planted during the last 30 years. In the ‘ Flora Hongkong- ensis, published in 1861, Bentham enumerated 1053 species as having been found on the island, 159 of which had not, at that time, been obtained elsewhere. Bentham, however, remarked that probably many of these 159 would be found on the main- land of China when the Flora of that country was investigated. This has proved to be tbe case, for I suppose that at the present time not inore than 50 of these 159 remain peculiar to Hongkong. Since the * Flora Hongkongensis' was published many additions have been made to the Flora of the island by various collectors, so that at the present time it amounts to about 1400 species. Of these 350 or 400 additional species, about 50 have not been observed out of the island, making a total of about 100 endemic species. No doubt many of these will be found on the adjacent mainland as we get to know more and more of its flora. It is interesting to note that some of the species found in Hongkong have not been procured nearer than Hupeh, a distance of about 600 miles north. I might mention examples in Machilus macro- phylla and Limacia sagittata. In all probability these species will be found in all the intervening country and perhaps still farther north and south. Rhodoleia Championi, first discovered by Champion in Hongkong, had not been obtained elsewhere until Dr. Henry collected it in Yunnan, about 1000 miles west of Hongkong. The Indian Quercus polystachya, found in Yunnan by Dr. Henry, has been recently collected in Hongkong. Neither of these plants has been collected at any place between these two points, so far as I know. Lasia heterophylla, a Cingalese and Indian plant, comes as far east as Tai-mo-shan in the New Territory, and it has been found on the island of Cheung I, just outside of Hongkong harbour, but it has not been found in Hongkong itself. Additional species are constantly being found, and not a year passes but some plant is added to the Flora. Botanizing in Hongkong is a very difficult business, and it will be many years before our knowledge of its Flora is complete. It is quite impossible to get up many of the ravines on account of the huge boulders that block the way. To proceed up the sides one has often to eut a way through masses of Bamboo and strong-growing creepers and shrubs. Many of the sides of the 60 MR. W. T. TUTCHER ON NEW SPECIES hills are almost perpendicular, and after cutting a path through to a likely spot, one is often confronted with a perpendicular rock which bars all progress. The only thing to do is to make the assault from another direction ; but this will give some idea of how it is species are being constantly added to the Flora and the difficulty of the collector. Ferns are represented by over 100 species and grasses by nearly as many. There are between 90 and 100 species of Leguminosz, and between 70 and 80 of Cyperaces. The Com- posites are between 60 and 70 and Orchids about 60. Although many additions are found in almost inaccessible places, others, on the other hand, are obtained in the most unlikely places, such as alongside publie roads which are traversed daily. Sloanea hongkongensis, for instance, grows quite close to and overhangs a road leading from Mt. Vietoria to Aberdeen, a village on the south side of the island. Lagerstremia Fordii, which was first discovered on the island of Lantao, grows close to the road leading from Pokfulum to Aberdeen. A curious instance of overlooking à species which had been found by Champion 50 years ago may here be noticed. Champion stated that Quercus Eyre? was abundant in the Wong-nei-cheong wood, east of the Happy Valley; but some years later Hance said he could uot understand Champion making such a statement, as he could not find the plant there at all, and came to the conclusion that an oak which he found in the Happy Valley wood was the species Champion referred to. Hance’s error is pointed out by Skan in the ‘ Index Flora Sinensis, where Hance's plant is described as Quercus alternata. All other collectors as well as Hance were unable to find Champion's plant, and the only specimens known were those in the Kew Herbarium. In the spring of last year I was looking up the Hongkong Oaks, and wanted to get living specimens of as many as possible, as some of them were not in the Hongkong Herbarium, so I thought I could not do better than look for Quercus Eyrei in the wood mentioned by Champion, where to my surprise I found it without difficulty. It was quite common in the wood, many of the trees being 30 feet high, and some of them overhanging the public road. Mucuna macrobotrya, stated by Hance to grow near the Buddhist Temple, Causeway Bay, I have never been able to find, although I have often looked for it. The general aspect of the Hongkong Flora is that of a more northern latitude, although most of the species are tropical. AND OTHER CHINESE PLANTS. 61 This is in a great measure due to the abundance of the Pine-tree, which is by far the commonest tree in the island. Generally speaking, the shrubs and trees have a stunted appearance, and that luxuriance which one associates with the tropics is entirely wanting. This is accounted for by the poorness of the soil, which is nearly all disintegrated granite, and makes it all the more remarkable that so large a number of species should be found on such a small island. Showy flowering shrubs and creepers are exceedingly numerous and the flowers are produced in the greatest abundance. In every month of the year the hillsides are bright with something or other. Perhaps there are more shrubs in flower in April and May than at any other time, but even in December and January there are such things as Lurya japonica, Eurya Macartneyi, aud Litsea citrata in flower. Eurya japonica is a most evil-smelling plant when in flower, and it is not at all uncommon to hear people complain of the bad drains when they pass a specimen of it, as it has exactly that kind of smell. That it will give some people **drain-throat" I can testify, having suffered from it myself in this way. The New Territory, in which is included the island of Lantao, leased to Great Britain by China in 1898 has an area of about 300 square miles, or ten times the area of Hongkong. The general features are those of Hongkong on a larger scale, as there are several mountain-ridges split up into numerous peaks, the highest of which is Tai-mo-shan, 3000 feet. Between these ridges are several valleys which are highly cultivated. Looking at the southern boundary of the New Territory from Hongkong, it appears almost as barren as Aden, but on closer inspection it is found that there are numerous ravines in which there is a considerable amount of vegetation. Lantao is an island some- thing like Hongkong, the highest peak of which rises to 3050 ft. Many of the ravines are well-wooded and will, no doubt, yield many interesting plants when thoroughly explored, and the same may be said of the ravines on the mainland. Rubus hexagynus, an Indian species, has been found on Tai-mo- shan, but nowhere elsein China. Rhododendron Championi was considered to be confined to Hongkong until it was discovered on one of the hills in the New Territory last year. Uncaria hispida was found in Lantao a few years ago and in Hongkong last year, but it is not known from elsewhere in China. Gironniera 62 MR. W. J. TUTCHER ON NEW SPECIES nitida, first discovered in Hongkong, was found in Lantao a few years ago, but it is not known from elsewhere. Rhododendron Fordii and Rhododendron Westlandii have not been found out of Lantao. This gives some idea of what an interesting ground the New Territory is likely to prove. Economic Botany. Tree-planting has been carried out on a considerable scale in Hongkong, but owing to the poorness of the soil it has been confined principally to one species, Pinus Massoniana. Many other kinds of trees have been tried but without success, unless in particularly favoured situations. A beginning has also been made in the New Territory, and this will probably be considerably extended. There is very little land suitable for agriculture in Hong- kong, but in the New Territory there are some fine level tracts, and there sugar and rice are the principal crops. Within the last few years improved varieties of sugar-cane have been introduced by the Government and are proving a success. Since the British occupation pineapple cultivation has greatly extended on the southern slopes of Tai-mo-shan and on the island of Cheung I. Experimental cotton-growing has been takeu up within the last year, but I am not in a position to state with what result. Irnrurorvw DUNNIANUM, Tutcher, n. sp. Frutex 3-4-pedalis. Folia verticillata, lanceolata, acuminata, 4-5 poll. longa, 1-1 poll. lata, supra glabra, subtus pallida ; petioli crassi, 4 poll. longi, purpurei. Flores solitarii in axillis foliorum dispositi, pedicellis $ poll. longis (post anthesin longi- oribus); braetes purpures, maxim: circiter 1 lineam longs. Perianthii segmenta 17, in:qualia, minute ciliolata, minima 14 lineas longa, maxima 44 lineas longa et 3 lineas lata, obtusa, ineurva, extus purpurea vel rubella, intus luteo-rubella. Stamina 24, biseriata, antheris filamentis fere æquilongis. Carpella 12, rostrata. Fructus non visus. (Herb. Hongkong Bot. Gard. n. 966.) This species differs from Jllictwm micranthum, Dunn, in having twice the number of stamens and in the greater number of carpels. The above plant was found growing gregariously on the AND OTHER CHINESE PLANTS. 63 banks of a stream at Sam-tam-lo in the British New Territory, Kowloon, in March 1908, and I have named it after Mr. S. T. Dunn, F.L.S., Superiutendent of the Botanical and Afforestation Dept., Hongkong, from whom I have received every assistance and encouragement in any botanieal work I have undertaken. CAMELLIA BANKSIANA, Lindl. ? Ihave been unable to find a complete description of this plant, but Champion in the ‘Kew Journal of Botany,’ vol. iii. p. 310, says :—“ Camellia Banksiana, Lindl., is, I believe, a Hongkong species, but unknown to me. Some specimens found on a hill near Mount Parker, and sent home by Mr. J. Bowring previous to my arrival, were considered as belonging to the C. Banksiana. Mr. Bowring mentioned its having sweetly-perfumed flowers, so that it could scarcely be the same as the C. assimilis, of which the flowers are scentless, notwith- standing a general affinity in habit and in the white pendulous flowers." Several specimens of a Camellia which were found on Mt. Parker in December, 1903, were referred by me to C. assimilis, but Mr. Dunn pointed out that they differed from that species, and on further examination I concurred with him. The specimens differ from C. assimilis in having the styles free for a greater distance, the capsule silky-hairy instead of glabrous, the flowers sweet-scented, and in the hairiness of the stamens, From Champion’s description quoted above, I am inclined to believe that the specimens belong to C. Banksiana. (Herb. Hongkong Bot. Gard. n. 968.) U CAMELLIA CRAPNELLIANA, Tutcher, n. sp. Arbor parva, 15-20-pedalis, ramis cortice lateritio-purpuraceis, ramulis glabris. Folia 4-5 poll. longa, 13-2 poll. lata, obovata, emarginate acuminata, serrulata, revoluta, glabra, subcoriacea, venis subtus conspicuis; petioli crassi, circiter 3 poll. longi. Flores in apicibus ramulorum, solitarii, sessiles, albi, 3—4 poll. diametro. Bractez orbiculatz, extus sericew. — Sepala orbiculata, extus sericea. Petala 6-8, obovata, extus leviter sericea. Styli 3, distincti, glabri. Ovarium lanatum. Fructus ignotus. (Herb. Hongkong Bot. Gard. n. 967.) This species differs from Camellia reticulata, Lindl. in having free, glabrous styles. 64 MR. W. J. TUTCHER ON NEW SPECIES One tree only found on the south side of Mt. Parker in April 1903, but not in flower then. When visited again in the early part of December in the same year, the flowers were just beginning to fade. The brick-coloured bark of this tree makes it a very conspicuous object at a distance, and it was by this means that it was discovered. Named after Mr. A. E. Crapnell, Hongkong, who has often aceompanied me on my botanieal excursions. ZANTHOXYLUM OVALIFOLIUM, Hight * Speeimens of, apparently, this species were found on the southern slope of Mt. Parker in April, 1903. They differ from the description of Z. ovalifolium in the rusty-tomentose inflorescence and much thinner (papery) leaves. Further specimens are required before the species can be determined with certainty. (Herb. Hongkong Bot. Gard. n. 965.) | CHISOCHETON HONGKONGENSIS, T'u£cher, n. sp. Arbor parva, circiter 12-pedalis. Folia 1 ped. longa; petioli 6 poll. longi. Foliola 12-16, 4-6 poll. longa et 2-21 poll. lata, oblonga, papyracea, opposita vel subopposita, obtusa vel emargi- nata, supra glabra, subtus opaca, basi oblique acuta vel rotundata, margine anguste revoluta; nervi 8-12-dupli; petioluli 1 poll. longi. Ramuli juniores fulvo-pubescentes. Flores paniculati, 2 lineas longi; panicule ad apices ramulorum, axillares, erecte, 5-6 poll. longz. Calyx 5-dentatus, pubescens. Petala 5, valvata, alba, extus pubescentia. Staminorum tubus truncatus extus intusque sericeus. Antherz 8, sessiles, in tubo incluse. Stylus glaber; ovarium pubescens. Fructus maturus glaber, pyriformis, 12. poll. diametro. Semina oblonga, eastanea, aril- lata, fere 1 poll. longa. (Herb. Hongkong Bot. Gard. n. 969.) The fruits of this plant are somewhat like those of Chisocheton patens, Blume, but the mode of inflorescence and the flowers, which are 5-merous instead of 4-merous, are quite different. It appears to be a very distinct species. Discovered on the south side of Mt. Parker in April, 1903, in fruit, but flowering specimens were not obtained until December of the same year. The genus has not previously been recorded from Chiua. AND OTHER CHINESE PLANTS. 65 Mocuna Cuampiont, Benth. The bracts of this species have not been noticed before, so far as I am aware. They are four in number at each node of the raceme, dark purple, broadly ovate and hairy on both surfaces. The outermost bract is the largest, 1} in. long, and covers the three others, which are somewhat smaller. The bracteoles are two to each flower, the same colour as the bracts, and hairy on both sides, narrow ovate or lanceolate in shape and about à inch long. Both the bracts and bracteoles are very deciduous. The racemes are produced on the previous year’s growths, and not on the very old wood as is the case with some other species of the genus. The leaflets are stipellate. Mvcvxa Brrpwoonprana, Tutcher, n. sp. Frutex sempervirens, alte scandens, ramulis ultimis ferrugineo- pubescentibus. Folia trifoliata, subcoriacea; petioli 3-4 poll. longi. Foliola petiolulata, exstipellata, ovato-oblonga (lateralia obliqua), breviter acuminata, 4-6 poll. longa et 2-23 poll. lata, supra glabra, subtus leviter pubescentia. Bractex 1 lin. longi, ovate, citissimo decidus; bracteole 3 lineas longs, ovate, cito deciduz, ferrugineo-pubescentes. Flores eburnei, 3-34 poll. longi, carnosi, racemosi, racemis 1-1} ped. longis, 20-30-floris, in ramis vetustioribus fasciculatis interdum in ramis junioribus productis, pedicellis 4 poll. longis, crassis, ferrugineo-pubescenti- bus. Calycis tubus fere 3 poll. longus, extus et intus ferrugineo- pubescens, limbo bilabiato, labio superiore } poll. longo integro vel interdum bidentato, labio inferiore trilobato, lobis tri- angularibus infimo 4 poll. longo, lateralibus infimo brevioribus. Vexillum 17 poll. longum ; al: 2? poll. long; carina 3-31 poll. longa, unguiculata ; margines vexilli et alarum ciliolatæ. Legumen lignosum, pubescens, 9-10 poll. longum, 1j poll. latum, inter semina leviter constrictum suturis ambobus bi- canaliculatis ; semina 5-6. (Herb. Hongkong Bot. Gard. n. 980.) This species differs from Mucuna macrocarpa, Wall., in the colour of the flowers and in the shorter, hairy pod. It has been confused with Mucuna macrobotrys, Hance, from which it differs in the colour of the flowers and in the entirely different pod. The plant was found many years ago by Mr. Ford on Mt. Parker, and last year (1903) I found it at the same place, 66 MR. W. J. TUTCHER ON NEW SPECIES as well as on Mt. Gough, flowering in April, and previously on Tai-mo-shan in the New Territory, in fruit. It is named out of eompliment to Colonel Birdwood, 110th Mahratta Light Infantry, an enthusiastie botanist, who has contributed many specimens to the Hongkong Botanic Garden. ARISTOLOCHIA Tawarresii, Hook. f. This plant is figured in the * Botanical Magazine’ under t. 4918, and is stated to bea native of Ceylon. Under t. 5295 in the same work this habitat is said to be wrong, and under t. 5908 Old Calabar is given as the native place of the species. Many years ago Mr. Ford collected the plant in Hongkong, but no specimens can be found at Kew from that place, although there are specimens in the Herbarium of the Hong- kong Botanic Gardens collected by Mr. Ford. Last year (1903) I found several specimens growing under Bamboos in a wood on the south side of Wong-nei-cheong village, and later on many other plants were found growing in crevices of the almost perpendicular cliffs of the Black Mountain, a place whieh is still likely to yield additions to the flora of the island, as several have been made from that district during the last year or so. (Herb. Hongkong Bot. Gard. n. 979.) BnipELIA BALANS#, Tutcher, n. sp. Arbor 15-20-pedalis; caudex spinosus, spinis 1-143 poll. longis, crassis. Folia 3—4 poll. longa, oblonga, acuminata, acuta, leviter revoluta, basi attenuata, supra glabra, subtus glauca, venis primariis lateralibus $-10; petioli 4 poll. longi. Flores 9 parvi, pedicellati, dense glomerati, in axillis foliorum dis- positi, pedicellis crassis, rugulosis, 1 lin. longis, calycis lobis deltoideis extus intusque pubescentibus, l lin. longis. Flores d non visi. Fructus ovoideus, purpureo-niger, fere 3 poll. longus. (Herb. Hongkong Bot. Gard. n. 974.) This plant has been confused in the Hongkong Herbarium with Bridelia tomentosa, Blume, from which it is quite different in foliage and fruit. Balansa found the plant in Tonkin and his specimens in the Kew Herbarium bear the numbers 4122, 4123, and 4124. In Hongkong I know of only two trees, one in the Happy AND OTHER CHINESE PLANTS. 67 Valley Woods above the Bowen Road, and the other in a wood on the southern slope of Mt. Parker. CLEIDION JAVANICUM, Blume. One speeimen only of this species was found in a ravine south of Sheko Gap in February of this year (1904). Not previously recorded from China. (Herb. Hongkong Bot. Gard. n. 963.) FICUS ALTISSIMA, Blume. It is difficult to say whether this is a native of Hongkong or not, but there are several fine specimens in the island and on the Kowloon peninsula. Hance found it in the province of Kwangtung near temples, and Dr. A. Henry collected it in Hainan. (Herb. Hongkong Bot. Gard. n. 970.) FICUS INFECTORJA, Roxb, This species is frequently met with in Hongkong and is, apparently, indigenous, although frequently planted for shade purposes. Dr. A. Henry states that it is commonly planted near shrines in China, but that he did not meet with it east of Patung in Hupeh. QUERCUS ITEAPHYLLA, Hance. Fruits of this species were collected in March of this year (1904) from trees growing in the woods on the east side of the racecourse, Wong-nei-cheong, and others from trees found on the hills above the Buddhist Temple, Causeway Bay. It belongs, as Hance suspected, to the section Pasania. The cup is scarcely 6 lines in diameter and 3 lines deep, silky on the inside. The connate bracts are in 6 to S series. The acorn is ovoid, glabrous, about 72 lines long including the umbo, which is 3 line. (Herb. Hongkong Bot. Gard. n. 982.) QUERCUS POLYSTACHYA, Wall. Three trees of this species, about 20 feet high, were found on the Black Mountain in February of this year (1904). Pre- viously collected by Dr. A. Henry at Mengtze in Yunnan, but not recorded from any other place in China. The acorns in the Hongkong specimens are about j in. in diameter, somewhat larger than those collected in other places. (Herb. Hongkong Bot. Gard. n. 981.) LINN. JOURN.—BOTANY, VOL. XXXVII. G 68 MR. W. J. TUTCHER ON NEW SPECIES CASTANOPSIS Evnzr, W. J. Tutcher. (Quercus Eyrei, Champ. ex Benth. in Hook. Kew Journ. Bot. vi. (1854) p. 114.) This plant was described in the ‘Kew Journal of Botany’ fifty years ago, but when its description was made out Bentham had not seen the female flowers or fruits, and although Champion states that the tree was abundant in the Wong-nei- cheong wood on the east side of the Happy Valley, it apparently has never been again collected until last March (1904). This appears the more remarkable when it is stated that the tree is still common in the wood referred to by Champion and that speeimens about 30 feet high overhang the publie road, which must have been used by various collectors from time to time. At the time of my visit the trees were past fruit, but I was successful in finding a part of an involucre which was sufficient to prove that the plant belongs to Castanopsis as constituted at present. ‘his fragment showed the iuvolucre to be about ? in. long with about 5 rows of compound spines. Better material is needed before a complete description of the fruit can be given. ALSOPHILA PODOPHYLLA, Hook., var. PROCUMBENS, Tutcher, n. var. Caudex procumbens, 10 poll. longus et 2 poll. diametro, frondibus pinnatis, 2-24 ped. longis, stipite 1-14 ped. longo. (Herb. Hongkong Bot. Gard. n. 977.) Specimens of this fern were found in a ravine on the south side of Sheko Gap, in February 1904. It was found previously by Balansa, no. 33 (in Kew Herbarium), in Tonkin. It struck me as being a new species on account of the procumbent caudex and the simply-pinnate fronds; but Mr. C. H. Wright, of the Herbarium, Royal Gardens, Kew, considers it to be a variety of Alsophila podophylla only. I may mention that there were no specimens of typical A, podophylla seen in the locality where these plants were collected. Young plants of true A. podophylla assume an upright position and produce bipinnate fronds at a very early stage, but it is several years before they bear fertile tronds. Portypopium (8 Puymaropes) Marurwit, Tutcher, n. sp. Rhizoma repens, gracile, squamis castaneis lanceolatis. Stipes gracilis, fere niger, glaber, 1-2 poll. longus. Frons ovata, AND OTHER CHINESE PLANTS. 69 integra, obtusa, j-3 poll. longa, 3-3 poll. lata, subcoriacea, supra et subtus nuda, margine revoluta, costa et venis primariis lateralibus conspicuis inira margines anastomosantibus, areolis et venulis tenuissimis, soris 2-3, majusculis, uniseriatis inter costam et marginem propius ad costam quam marginem. (Herb. Hongkong Bot. Gard. n. 854.) This small fern was collected in Wei-hai-wei by Staff-Surgeon C. G. Mathew, R.N., of H.M.S. ‘Eclipse? who has presented useful eollections of plants from Corea, Wei-hai-wei, and Shanghai to the Hongkong Herbarium. He has also made several additions to the Flora of Hongkong. Mr. C. H. Wright has kindly informed me that the species 1s near P. oodes, Kunze. It is much smaller than P. oodes and the sori are in series instead of being scattered. GYMNOGRAMME ELLIPTICA, Baker, var. FURCANS, Tutcher, n. var. ; pinnis superioribus multo fureatis, inferioribus multo incisis. Several plants were found in a ravine on the south side of Mt. Parker in December, 1908. POSTSCRIPT. Dunnta, gen. nov. Rubiacearum, Tutcher. Frutex 5-pedalis, ramis crassis, ramulis junioribus pubes- centibus. Folia opposita, oblanceolata, breviter acuminata, 8 poll. longa, 2 poll. lata, brevissime petiolata vel sessilia, supra glabra, subtus opaca, costa venisque leviter pubescentibus, venis primariis lateralibus utrinque 16-20, conspicuis, intra margines anastomosantibus, venulis reticulatis. Stipule late, laciniate, persistentes, pubescentes. Flores cymosi; cymæ corymbosz, terminales. Calycis lobi 5, minuti, acuti, persistentes; circiter unum lobum bracteiforme in omni eymula; lobus bracteiformis albus, ovatus. Corolla? Fructus capsularis; capsula crus- tacea, globosa, } poll. diam., septicide dehiscens, 2-valvis, valvis 2-partitis. Semina numerosa, orbicularia, peltata, fimbriata, rugulosa, 1 lin. diam. This genus has afiinity with Emmenopterys and Mussen- dopsis. It differs from both in having persistent calyx-lobes G2 70 MR. J. G. BAKER: A REVISED and in the fruit and seeds. It also differs from Hmmenopterys in its persistent stipules. DUNNIA SINENSIS, Tutcher, species unica. (Herb. Hongkong Bot. Gard. n. 910.) This plant was found by Mr. Dunn’s native collector at Sanning, Kwangtung, last year (1904). The specimens were in fruit only. 17 February, 1905. W. J.T. A Revised Classification/of Roses, 1905. By Joun GILBERT Baker, F.R.S., F.L.S. [Read 16th February, 1905.] THREE valuable Rose-monographs which have recently been published to a large extent fill the gap which was left by the long-expeeted monograph of the late Professor Crépin never having been completed. These are Dr. Keller's account of the Roses, in Ascherson & Graebner’s ‘Synopsis of the Flora of Central Europe’; the very careful and elaborate account of the French Roses, in Rouy & Foucaud's new * Flora of France’; and Dr. Focke’s description of the Roses, in the third edition of Koch’s * Synopsis,’ now in course of publication under the editorship of Dr. Hallier. Reference should also be made to Dr. Christ's book on the Swiss Roses, and his account of the Oriental Roses in the supplementary volume of Boissier's ‘Flora Orientalis, and to Burnat & Gremli's * Roses des Alpes Maritimes, and its Supplement, and to Burnat’s * Flore des Alpes Maritimes.’ The consensus of opinion in all recent investigators of the genus points to the conclusion that a great deal of the difficulty that arises is due to the facility with which Roses hybridise. I have therefore drawn up the following Catalogue of the species, varieties, and principal hybrids, to replace that which I contributed to the ‘ Gardeners’ Chronicle’ twenty years ago, and was printed in that journal, August 15, 1885, p. 199, and reprinted in the * Journal of Botany’ for the same year, pp. 281-286. 71 CLASSIFICATION OF ROSES. "umjuoo Soy} solvads Auvuuad Jo asquinu əy} oiu sourvu əy} Suimo[rog sjoxouq ot Ur sdaquinu at id [} 59! . : : qovuq eq ur l f [$] 'essourdiquar ‘TX ae QQ3vouoq Ivpupuryls AIOA SOAVUO'T [e] ‘SS0TILA "X MEME C DOIMVVVVVMVYMVMVVVV VV VV Katwq Ka0A SOAVO'T [S] 'eurue) "XT (ntt tte nn tt nt t ter t nn nn Kanu Angsrs ao snodqv|8 soAvorT "wogun porojuos SOP W ILLATYOONOH [c] "e ueor[Te5 "ITI A a om mon ast ros à o nom rog on o 5C oso, . om on o, ee jn04s ‘paxooy solyorud Xo [S] 'ewrssrsoutdg ‘ITA (ttt ttt tt titt quorers *ropuo[s *2do[ sopporad. aoSawm "qenbouu £494 *pozogyvos soppuuq P HENVOVAALAH [16] ‘ROULOWLVUUID ‘TA RPM wor oos o e n om not n] os oto os on oa st o. s UC uty} e YUM ‘pod dim [1] 'eq£qdoxoryr “A Peres sees sees esse reese ses UDS MOLD e ygn *ugo18 dig 'snouqu[5 pny [c] ^&jeejovig ^AI 11177 pastour A[daep *popaoao sjovag — "Kareq. Ap[juogsrsaod. pug sared avpndys ur uojjo sopqouq CP ILLNFOFId 'e[orjod əq4 03 o3vupv so[ndigg; [e] 'euersXxueg ‘ITI TE TE e ma ee eee ++ gnonpioop *oo1g so[ndrg 'pepuajoad qonur you *oo1j so[ 438 [Or] 'erp43s£g "II Uto ctttt177 xsrp aq} puo£oq popnagoad st qorqa uuin[oo v ur payun so[43g 'eqv[ndrjs *punoduioo saavary (11 'aT[ojyroipdurrg "To m t s n t on s mmm tira o hse e oh im os t o e oh onn PPP oje[udisxo ‘ordus S9AUO'T m 'sdno4p) ay} o} fay yoouhpup 72 MR. J. G. BAKER: A REVISED SUBSPECIES AND VARIETIES. PRIMARY SPECIES, PRINCIPAL HYBRIDS. Group I. Simplicifoliz. 1. R. SIMPLICIFOLIA, Salisb,— Orient. (R. berberifolia, Pall.) R. Hardii, Cels. (simplicifolia x involucrata). Group II. Systylæ. 2. R. ARvENSIS, Huds.—Europe. (R. repens, Scop.) R. capreolata, Neill. R, perviridis, Gren, (arvensis X sempervirens). R. rusticana, Déségl. (arvensis X stylosa). R. Melsini, Towndr. (arvensis X canina). co R. SEMPERVIRENS, Linn.--South Europe. R. scandens, Miller. R. prostrata, DC. R. longicuspis, Bert. —India. 4. R. LEsCHENAULTIANA, Wight & Arn.—Neilgherries. e R. MOSCHATA, Herm.—S. Europe, N. Africa, Asia. R. Brunonis, Lindl. R. Pissardi, Carriére—Persia. R. abyssinica, R. Br.— Abyssinia. R. nivea, Duport. (moschata x gallica). 6. R. PHÆŒNICEA, Boiss.—Orient. R. byzantina, Dieck. ( phanicea x gallica). 7. R. MULTIFLORA, Thunb.—Japan. R. Lucie, Franch. & Savat. R. anemoneflora, Fortune. (multiflora x Banksie). R. polyantha, Hort. (multiflora X indica). 8. R. WICHURÆANA, Crépin—Japan. Wichureana x rugosa and many others. 9. R. WATSONINANA, Crépin—Japan. 10. R. seriGEnA, Michx.—N. America. setigera x gallica. ll. R. srvrosa, Desv.— Europe. (R. systyla, Bast.) R. virginea, Ripart. CLASSIFICATION OF ROSES. 73 ) SUBSPECIES PRIMARY SPECIES, > PRINCIPAL HYBRIDS. AND VARIETIES. Group III. Banksianæ. 12. R. BAxxsrz, R. Br.— China. R. lutea, Hort. R. Collettit, Crépin—Birma. R. Fortuneana, Lindl. ( Banksic x levigata). 13. R. Microcarpa, Lindl.— China, (R. indica, L., ex parte.) 14. R. LÆVIGATA, Michx.— China. (R. sinica, Lindl., non Ait.) Anemone (lævigata X indica). Marie Leonidas (levigata X bractescens). Group IV. Bracteatæ. 15. R. invotucrata, Roxb.—Ind. Or. (R. Lyellii, Lindl.) (E. clinophylla, Thory.) 16. R. BRAcTESCENS, Wendl.—China. CR. lucida, Lawr. non Ehrh.) Group V. Microphyllæ. 17. R. MICROPHYLLA, Lindl., non Retz.—China, Group VI. Cinnamomeæ. Old World. 18. R. CINNAMOMREA, Linn.— Europe, N. Asia. (R. majalis, Herm.) (R. davurica, Pallas.) R. francofurtana, Munch. (R. turbinata, Ait.) (cinnamomea X gallica). 19. R. MACROPHYLLA, Lindl.—-Himalayas, China. (R. Davidi, Crépin.) 20. R. PnATTI, Hemsl.—Tibet. 21. R. SERICEA, Lindl. — Himalayas, &c. (R. tetrapetala, Royle.) 22. R. WEBBIANA, Wall.—Himalayas. 23. R. ACICULARIS, Lindl.—N. Europe, N. Asia, N. America. (R. carelica, Fries.) (R. Gmelini, Bunga.) (R. Engelmanni, S. Wats.) R. nipponensis, Crépin. 74 MR. J. G. BAKER: A REVISED SUBSPECIES AND VARIETIES. PRIMARY SPECIES. PRINCIPAL HYBRIDS. Cinnamomeæ (cont.). Old World. 24. R. LAXA, Retz. non Lindl.—Siberia. 25. R. FEDSCHENKOANA, Regel—Turkestan. 26, R. BEGGERIANA, Schrenk—Central Asia. R. anserinefolia, Boiss. 27. R. ELYMArTICA, Boiss.—Persia. bo d E t. RUGOSA, Thunb.—Japan Ke. R. kamschatica, Vent. R, Twara, Sieb. & Zucc. (rugosa x multiflora). R. calocarpa, André. (rugosa x indica), and many others. few World. 29. R. CAROLINA, Linn.—Eastern States. (R. hudsoniana, Thory.) 30. R. viraintana, Miller, non K. Koch—Eastern States, (R. lucida, Ehrh.) 9l. R. numis, Marsh.—Eastern States. (R. parviflora, Ehrh.) R. lancifolia, Small. 32. R. NITIDA, Willd.—North-eastern States. 33. R. FOLIOLOSA, Nutt.—Texas and Arkansas. R. mexicana, S. Wats.—Mexico. 94. R. GYMNOCARPA, Nutt.— Western States. 35. R. CALIFORNICA, Cham. & Schlecht.—California. R. piscicarpa, A. Gray. 36. R. Woopsr, Lindl.—Rocky Mountains. 37. R. Fenpiert, Crépin—Rocky Mountains. 38. R. NUTKANA, Presl—N.W. America. Group VII. Spinosissime. 39. R. EGLANTERIA, Linn.—Orient. (R. lutea, Miller.) R. punicea, Miller. CLASSIFICATION OF ROSES. 75 SUBSPECIFS PRIMARY SPECIES. - PRINCIPAL HYBRIDS. AND VARIETIES. 40. R. HEMISPHÆRICA, Herm.— Orient. (R. sulphurea, Ait.) R. Rapini, Boiss, 41. R. PLATYACANTHA, Schrenk—Central Asia. (R. xanthina, Lindl. ?, Crépin.) R. Ece, Aitch. 42. R. SPINOSISSIMA, Linn.—Europe, N. Asia. (R. pimpinellifolia, Linn.) R. altaica, Willd.—Siberia. (2. grandiflora, Lindl.) R. Alberti, Regel—Turkestan. R. myricantha, DC.—S. Europe. R. hispida, Sims—Siberia, (R. lutescens, Pursh.) R. rubra, Hort, R. rubella, Sm. (R. gentilis, Sternb.) (R. reversa, Waldst. & Kit.) (spinosissima X alpina). e. involuta, Sm. (R. Sabini, Woods.) (R. coronata, Crépin.) CR. Robertsoni, Baker.) (R. Wilsoni, Borr.) (spinosissima X tomentosa). R. dichroa, Leresch. (spinosissima x mollis). R. sabauda, Rapin. (spinosissima X glauca). R. hibernica, Sm. (R. Schultzii, Ripart.) (spinosissima X canina). R. biturigensis, Boreau. (spinosissima X rubiginosa). 43. R. ALPINA, Linn.—Central Europe. R. pyrenaica, Gouan. R. Malyi, Kerner. R. oplisthes, Boiss.—Caucasus. 44. R. MINUTIFOLIA, Engelm.—S. California. 45. R. STELLATA, Wootton—New Mexico. 76 MH. J. G. BAKER: A REVISED SUBSPECIES PRIMARY SPECIES. PRINCIPAL HYBRIDS. AND VARIETIES. Spinosissimæ (con/.) 46. R. BLANDA, Ait.—N. America. (R. virginiana, K. Koch, non Miller.) R. lava, Lindl., non Retz. R. fraxinifolia, Borkh, R. Willmottiana, Baker. R. Sayi, Schwein. R. arkansana, Porter. blanda X rugosa. Group VIII. Gallicane. 47. R. GALLICA, Linn.— Central and South Europe. R. provincialis, Miller. R. centifolia, Linn. F. muscosa, Miller. R. pomponia, DC. R. parvifolia, Ehrh. R. damascena, Miller. R, bifera, Poir. (gallica x moschata). ( R. arvina, Krock, | R. geminata, Rau. 4 R. hybrida, Schleich. | R. gallicoides, Baker. LR. Polliniana, Spreng. (gallica X arvensis). 48. R. sancta, A. Hich.—Abyssinia. Group IX. Canine. 49. R. ixpicA, Linn. ex parte— China. R. semperflorens, Curt. . Lawrenceana, Sweet. . odoratissima, Sweet. . pseudo-indica, Lind], . longifolia, Willd. Naas R. Bourboniana, Thory. (indica x gallica). | R. Noisettiana, Thory. | R. floribunda, Andr. (indica x moschata). R, reclinata, Thory. (indica x alpina). R. Ruga, Lind). (indica X arvensis). CLASSIFICATION OF ROSES. 77 SUBSPECIES PRIMARY SPECIES, PRINCIPAL HYBRIDS. AND VARIETIRS. 90. R. GIGANTEA, Collett—Upper Birma, &c. 51. R. CANINA, Linn.— Europe, West Asia. . caucasica, Pall, , dumalis, Bechst. . urbica, Leman. . dumetorum, Thuill. . Montezume, H. b, K. . obtusifolia, Desv. and many others. eU md y RS EDoRd ( R. alba, Linn., | R. incarnata, Miller, 4 R. collina, Jacq., | and many others. | R. macrantha, Desf. (canina x gallica). 52. R. evauca, Vill.—Europe. (R. Reuteri, Rapin.) R. subcristata, Baker. R. coriifolia, Fries. R. Watsoni, Baker. and many others. R. salevensis, Rapin. (glauca x alpina). 53. R. TOMENTELLA, Leman—Europe. 54. R. FERRUGINEA, Vill.—Centrai Europe. (R. rubrifolia, Vill.) 55. R. MONTANA, Chaix—Central Europe. R. rhetica, Keller. R. abietina, Gren. R. Chavini, Rapin. 56. R. PovziNI, Tratt.—Southern Europe. R. Diomedis, Gren. R. Beatricis, Burn. & Gremli, R. gallinaria, Burn. & Gremli. Group X. Villosæ. 57. R. POMIFERA, Herm.—Central Europe. (R. villosa, Linn., ex parte.) R. recondita, Puget. R. Dicksoni, Lindl. ( pomifera X cinnamomea). 78 A REVISED CLASSIFICATION OF ROSES. SUBSPECIES PRIMARY SPECIES, PRINCIPAL HYBRIDS. AND VARIETIES. Villosæ (con.). 58. R. MOLLIS, Sm.—Europe. (R. mollissima, Fries, non Willd.) 59, R. ORIENTALIS, Dupont—Orient, 60. R. TOMENTOSA, Sm.—Europe. R. subglobosa, Sm. R. scabriuscula, Winch. R. cinerascens, Dumort. R. omissa, Déségl. j spinulifolia, Demar. R, hawrana, Knet. (tomentosa x alpina). R. Mareyana, Boul. (tomentosa X gallica). 61. R. HEckELIANa, Tratt.—Sicily to Greece. Group XI. Rubiginosz. 62. R. JuNDziLLI, Bess.— Central Europe. (R. trachyphylla, Rau.) 63. R. nuniarNosa, Linn.—Europe. R. jemensis, M. Schultz. Lady Penzance (rubiginosa X lutea) and many others. 64. R. MICRANTHA, Sm.— Europe. (E. nemorosa, Libert.) R. micranthoides, Keller. R. hungarica, Kerner. R. Driggsii, Baker. 65. R. aGrestis, Savi—Central and Southern Europe. (R. sepium, Thuill.) R. elliptica, Tausch. R. graveolens, Gren, R. nodosa, Vries (R. Klukii, Bess.). R. caryophyllacea, Bess. R. arabica, Crépin. 66. R. GLUTINOSA, Sibth. & Sm.--Orient. R. sicula, Tratt.—-Sicily. 67. R. SERAPHINI, Viv.— Corsica. 68. R. FEROX, Bieb.—Crimea. 69. R. ASPERRIMA, Godet— Persia. ON THE FLORA OF LIBERIA. "79 Geographical Distribution.—Five Roses are found south of the Tropic of Cancer, viz.: A. moschata and R. sancta in Abys- sinia, R. Leschenaultiana in the Neilgherries, and R. mexicana and R. Montezume, H. B. K., in Mexico. The latter is only a form of R. canina, so probably it has been introduced. In the North Temperate Zone there are six well-marked Rose-regions, a considerable proportion of the species inhabiting each being endemie. These are:—(1) Europe, including North Africa; (2) the Orient, excluding the European eountries included in Boissier's * Flora Orientalis"; (3) Northern and Central Asia, including China and Japan; (4) India; (5) Western United States and Rocky Mountains; and (6) the Eastern United States. Of these districts the richest in Roses is Europe with North Africa, which has 29 species. North and Central Asia, including China and Japan, closely follows it, with 26 species. The other districts have much smaller numbers—the Asiatic portion of the Orient having 18, India 9, the Western United States 10, and the Eastern United States 6 species. Contributions to the Elora of Liberia. By Dr. Orro Starr, F.L.S. [Read 16th March, 1905.] Tu subject of this paper includes a number of new species, some of them belonging to new genera, which were collected by Mr. Alexander Whyte in the Republic of Liberia during several expeditions undertaken by him last year on behalf of Sir H. H. Johnston acting for the Monrovian Rubber Company. As I wish to confine myself in this place to technical descriptions of the novelties in Mr. Whyte's collections, I abstain from an attempt to give a general account of the flora of Liberia, reserving that for another occasion. It may suffice to say that the collections were made in the following localities :— (1) At Monrovia and in its vicinity, within a radius of 6 miles, in February during the dry season, the ground being described as covered with second forest-growth on abandoned farms, and as flat with many lagoons and backwaters. 80 DR. OTTO STAPF ON (2) In the hinterland of Monrovia, within a radius of 20 miles of a place called Karkatown. (3) In the basin of the Sinoe River, from Sinoe as far as Soyos’ Town in the Kuru (or Kulu) country, about 80 miles inland, in March and April, during a period of extremely hot weather. “ Physical features of the country flat and covered with dense virgin-forest, except where native gardens have been made and where the forest-growth rushes up with amazing rapidity. Rainfall very abundant; climate moist and humid all the year round. Country under water in very wet weather and travelling almost impossible, as no roads exist, or even traees of forest footpaths.” As there are no special labels with the specimens, I am obliged to give the localities in a summary way. The collections comprise over 260 species. To these Mr. D. Sim, an employee in the service of the Monrovian Rubber Company, also con- tributed; but as I have already described the novelties dis- covered by him in * Flora of Tropical Africa,’ vol. iv. pp. 595- 610, his name does not appear in this paper*. Out of these 260 aud odd species I have to record 4 new genera and 58 new species, or, including those described previously, 67 new species, certainly a very considerable percentage. This will, however, appear less surprising if we bear in mind that Liberia is, perhaps with the exception of the French Ivory Coast, the least explored part of the West-African littoral, and that up to 1904 not more than about 200 species were known from the whole territory of the Republic, the area of which is estimated at 36,800 square miles, or a little more than two-thirds the area of England. Moreover, almost the whole of the earlier collections were made at Grand Bassa and Cape Palmas, that is outside the districts explored by Whyte and Sim. That Liberia is likely to yield a great many new species is also suggested by the results of Mr. Dinklage’s excursion to Grand Bassa and Cape Palmas in 1898, numerous new species having been described by the Berlin botanists from his collections. At the same time, it must be remarked that all the new species described here belong to types of a higher order, characteristic of the West-African flora, or extending over larger parts of the tropics. As might be expected, Liberia, to judge by the little we know, presents itself * See note on p. 115. THE FLORA OF LIBERIA. S1 as a section, with rather artificial boundaries, of the great natural region which extends from the Senegal in the north to the Kunene in the south. Its flora has, no doubt, a strong local colouring, but it is rather of the specific than the generic order. Its nearest affinity is, of course, with Sierra Leone, and not a few of the species described here are representative forms of that country. Even three of the four new genera proposed by me are not endemic in Liberia. One, Urobotrya, extends to Sierra Leone, whilst the other two, A£rovxima and Afrodaphne, range over a stil greater portion of West Africa aud comprise species hitherto referred to old genera. Finally, I have to remark that the whole of Whyte’s and Sim’s collections were presented to Kew by Sir H. H. Johnston, and to thank Mr. C. B. Clarke and my colleagues Messrs. N. E. Brown, T. A. Sprague, and C. H. Wright for their help in working out certain orders, and Dr. James Clark for his assist- ance in dissecting species which had to be taken into account for purposes of comparison. The sequence of genera is that of Bentham and Hookers * Genera Plantarum. Tetraceras leiocarpa, Stapf (sp. nov.); affinis T. potatorie, Afz., a qua differt imprimis carpellis glaberrimis vel pilis paucis minutis aspersis, folliculis levissimis, arilo magis lacerato, semine fere duplo majore. Folia et panicule ut in T. potatoria nisi folia glabra vel subtus parcissime pilosa. Sepala 5-9, late rotundato-elliptiea, majora 3 lin. longa, extus glabra vel sparse minute pubescentia margine ciliata, intus sericeo-tomentosa. Petala 2-3, elliptica, caducis- sima. Carpella circiter 4, glaberrima vel pilis paucis minutis aspersa. Folliculi obovoidei, abrupte in cuspidem (stylum persistentem) contracti, 3 lin. longi, lævissimi, nitidi, olivacei, Semina subglobosa, ad 23 lin. dimetientia, testa rugosa, arillo pallido ad vel ultra medium lacerato, laciniis tenuibus. Karkatown, Whyte. The follicles of T. potatoriu are very conspicuously sulcato- striate, and the largest seed I have seen was only 14 lin. long. Popowia Whytei, Stapf (sp. nov.); affinis P. Vogelii, Baill., differt foliis firmioribus subtus vix glaucis, floribus sericeo- tumentosis. 82 DR. OTTO STAPF ON Ramuli floriferi 14 lin. crassi, cortice atro-cinereo. Folia subelliptiea, obtusa, 33-6 poll. longa, 2-21 poll. lata, subcoriacea, glabrescentia, supra pallide viridia, infra subglauca, nervis lateralibus 8-9 obliquis venarum reticulatione tenui; petiolus 2 lim. longus, erassiusculus. Flores 2-5, fasciculati; pedicelli bracteaque minute, fulvo-tomentoss. Sepala late ovata, tenuiter tomentosa, 3 lin. paulo longiora. Petala exteriora vix 2 lin. longa, extus tenuissime tomentella, interiora dimidio minora, glabra, omnia crassa. Stamina 9 ; anthere subsessiles, connectivo quadrato, obseure glanduloso, loculis lateralibus ; staminodia nulla. Carpella numerosa, tomentosa, 2-ovulata. Sinoe Basin, Whyte. The leaves are, as in Popowia Vogelii, more or less pubescent when young, the fine adpressed and scattered hairs disappearing later on. Kolobopetalum ovatum, Stapf (sp. nov.); affine K. auriculato, Engl., sed foliis ovatis basi rotundatis integris, paniculis robusti- oribus, filamentis 6 ad fere apicem in columnam brevissimam connatis distinctum. Planta glaberrima. Folia ovata, basi rotundata, abrupte in acumen obtusiusculum contracta, ad 4 poll. longa, ad 22 poll. lata, tenuia, nervis lateralibus utrinque 4-5, infimis e basi ortis uti venis vix prominulis. Panicule ad 8 poll. longe, divaricato- ramoss, ramis infimis ad 2 poll. longis. Flores d minuti, 1-2 in axillis braetearum lanceolato-caudatarum 1 lin. longarum; pedicelli pertenues, breves. Sepala 6, interiora longiora, late elliptica, ad 1 lin. longa. Petala 6, obovata, apice rotundata, exunguiculata, à lin. longa. Filamenta fere tota longitudine connata; anther loculis confluentibus transverse dehiscentes, valva interiore minore. Flores 9 fructusque ignoti. Sinoe Basin, Whyte. Nymphza Lotus, Z., var. sinoéensis, Stapf (var. nov.). Folia tenuia, integra. Flores vix 3 poll. diametro. Sepala extus ex toto vel preter marginem album pulchre violacea, ad 14 poll. longa, ad 5 lin. lata. Sinoe Basin, Whyte. Alsodeia prasina, Stapf (sp. nov.); affinis 4. Afzelii, Engl., differt floribus minoribus, calyce pro ratione majore, antherarum appendiculo cochleariformi. THE FLORA OF LIBERIA. 83 Planta glaberrima. Ramuli novelli pallide virides. Folia oblonga vel oblongo-lanceolata, basi obtusa vel subaeuta, apice acuminata, margine plerumque magis minusve serrata vel crenato- serrata, ad 8 poll. longa, ad 3 poll. lata, subcoriacea, pallide viridia, nervis lateralibus utrinque 7-10, venarum reticulatione subtus laxa, supra plerumque aretiore ; petiolus ad 1 poll. longus. Panicule ad 33 poll. longs, fere a basi ramos; rami infimis exceptis 4 poll. vix excedentes; bracteæ parve, ovate, sub- scariose ; pedicelli brevissimi. Sepala rotundato-ovata, interiora ad 13 lin. longa, ciliolata, prominenter nervosa. Petala oblonga, 2 lin. longa. Staminum tubus brevis, truncatus, utrinque glaber ; authere in tubi margine sessiles, eum connectivi appendice exteriore ovato 1 lin. longæ, appendiculo interiore rhomboideo cochleariformi. Within 6 miles of Monrovia, Whyte. Engler (in Bot. Jahrb. xxxiii. p. 140) enumerates Zenkers no. 1778 from Bipinde under Alsodeia kamerunensis, Engl., which is described as having a 5-lobed staminal tube with broad, deltoid, truncate lobes. The specimen at Kew from Zenker’s collection, no. 1778, has a staminal tube very like that of A. prasina, and it also resembles it in other respects; but it differs in the nerves of the sepals not being raised and the hairiness of the inner side of the staminal tube and the edges of the bases of the thecæ. Alsodeia Whytei, Stapf (sp. nov.) ; affinis A. brachypetale, Turez., sed foliis minoribus, petiolis brevioribus tenujoribus, inflorescentiæ pubescentia tenuissima, antherarum thecis obscure mueronatis, connectivi appendice latiore acuto differt. Ramuli pallide virides, minute puberuli vel elabri. Folia lanceolata vel elliptico-lanceolata, acuminata, basi subaeuta, margine magis minusve serrulata, 2-4 poll. longa, 14-2 poll. lata, tenuia, utrinque glabra vel inferne in dorso sparse puberula, pallide viridia, nervis lateralibus utrinque 6-9 uti venis tenuibus ; petiolus gracilis, 4-9 lin. longus. Paniewle 12 poll. longe, pubescentes, breviter ramos» ; bracteæ ovate vel lanceolate, ad 1 lin. longe; pedicelli ad 14 lin. longi. Sepala late ovata, interiora ad 1j lin. longa, tenuiter pubescentia, margine albido- ciliolato. Petala anguste ovato-oblonga, 2 lin. longa. Staminum tubus brevis, truncatus; anthere in tubi margine sessiles, cum LINN. JOURN.—BOTANY, VOL. XXXVII. H 84 DR. OTTO STAPF ON connectivi appendice ovato-lanceolato acuto fere 1j lin. longe, thecis minutissime mucronatis, mucrone pilosulo. Within 6 miles of Monrovia, Whyte. Alsodeia Johnstonei, Stapf (sp. nov.); affinis A. kamerunensi, Engl. (e descriptione), sed foliis tenuioribus lanceolatis, amoene viridibus, nervis lateralibus intra marginem eleganter arcuato- conneetis, paniculis ovatis brevioribus, andracii tubi lobis ovatis distineta. Ramuli juniores pubescentes, mox glabrati, fusci. Folia lanceolata vel oblongo-lanceolata, basi obtusa vel subacuta, acute acuminata, magis minusve serrulata vel subcrenata, ad 7 poll. longa, ad 2- -91 poll. lata, ameene viridia, tenuia, glabra, nervis lateralibus utrinque 8-12 tenuibus eleganter arcuato- connectis, nervo eolleetivo 2-3 lin. a margine distante, venarum reticu- latione tenui; petioli 21-5 lin. longi, pubescentes. — Panicule terminales, cum pedunculo brevi 1-3 poll. longæ, densiuscule, elabre vel subglabre ; bracteæ ovate, 1 lin. longe ; pedicelli ad 13 lin. longi. Sepala ovata, obtusa, 11-21 lin. longa, tenuiter albo- marginata, ceterum exsiccata fusco-nigrescentia, glabra. Petala oblonga, 2-8 lin. longa. Andræcii tubus brevis, 5-lobatus, lobis brevibus ovatis; anthere intra lobos tubi sessiles, cum con- nectivi appendice lanceolato-acuto 2 lin. longs, thecis ad apices contiguis. Capsula 9 lin. longa, glabra, apiculato-acuminata. Semina ambitu triangulari-obovata, compressa, ad 3 lin. longa. Sinoe Basin, Whyte. Oncoba brevipes, Stapf (sp. nov.); affinis O. glauce, Hook. f. (quoad plantam Beauvaisianam), sed foliis ad vel supra medium latissimis abrupte latiuscule obtuse acuminatis, breviter petiolatis, floribus haud earmineis diametro 4-5 poll. distincta. Planta glaberrima, novellis resinosis. Folia elliptico-oblonga vel obovato-oblonga, subito in acumen brevem obtusum contracta, basi acuta, ad 7 poll. longa, ad 3} poll. lata, pallide viridia, haud glauca, nervis lateralibus obliquis utrinque 8-9, venis laxe reticulatis obscuris; petiolus 3-1 poll. longus. Racemi 2-3- flori, brevissime pedunculati; pedicelli subrobusti, ad 1j poll. longi. Petala circiter 10-12, obovato-lanceolata, cum ungue ad 2 poll. longo, ut videtur alba. Anthere acute, 24 lin. longe. Near Monrovia, Whyte. THE FLORA OF LIBERIA. 85 ATnOXIMA *, Stapf (gen. nov. Polygalacearum). Aflinis. Carpolobie, G. Don, differt petalis subæqualibus, infimo haud naviculari, fruetu duro, pericarpio crustaceo, semini- bus endospermate destitutis. Sepala 5 structura simillima, magnitudine inequalia, interiora majora. Petala 5, subæqualia, lanceolata, haud unguiculata, infimum quam estera vix vel paulo latius et ut ea leviter con- "avum, ima basi tubo staminali adnatum, duo summa multo altius et oblique adnata intusque infra medium sericeo-villosa. Stamina 5, monadelphia, tubo staminali in latere superiore tota longitudine fisso; anther sessiles vel filamentis distinctis insideutes. Ovarium sessile, 3-loeulare, loculis l-ovulatis ; stylus filiformis; stigma punctiforme. — Fructus subglobosus, durus; pericarpium crustaceum (more Xanthophylli). | Semina subglobosa; testa tenuiter molliter pilosula ; endosperma nullum, Embryo cotyledonibus crassis sectione transversa subsemiorbicularibus. Arbuseula vel frutices, glabre. Folia alterna, coriacea. Flores in racemis plerumque gracilibus longiuseulis axillaribus, 'arius paniculati. Species note 4 in Africa tropica occidentali. A. liberica, Stapf (sp. nov.) ; affinis A. macrostachye, Stapt (Carpolobia macrostachya, Chod.), sed foliis majoribus magis conspicue reticulatis, breviter petiolatis, inflorescentiis fori- busque preter sepala ciliolata glabris distincta. Ramuli graciles, cortice pallido. Folia oblonga, aeuta vel subobtusa, apice subcaudato-acuminata acumine obliquo, 4—6 poll. longa, 14-27 poll. lata, tenuiter coriacea, nervis lateralibus utrin- que cireiter 6 valde obliquis areubus a margine 2-3 lin, distan- tibus connectis, reticulatione venarum arcta utrinque conspicua prominula ; petioli erassiusculi, nigrescentes, 2 lin, longi, Racemi solitarii vel bini in axillis foliorum, circiter 2 poll. longi; bractez late ovate, minute ; pedicelli 1 lin. longi. Sepala lite ovata, obtusa, minute ciliolata, interiora 9 lin, longa. Petala 3 lin. longa. Anthere sessiles. Sinos Basin, Whyte. * arpotimos, not eatable; in allusion to the fruit, H2 86 DR. OTTO STAPF ON The other species which I refer to A£rozima are :— 1. A. Afzeliana, Stapf (= Carpolobia Afzeliana, Oliver, in Fl. Trop. Afr. i. p. 186); 2. A. macrostachya, Stapf (= Carpolobie macrostachya, Chod. in Bull. Herb. Boiss. v. p. 117); 9. 4. Zen- keri, Stapf (=Carpolobia Zenkeri, Gürke, MS.). As the latter is still undescribed, I add here a short diagnosis :— Atroxima Zenkeri, Stapf (sp. nov.). Kamuli graciles. Folia elliptico-oblonga, basi breviter contracta acuta, apice abrupte acuminata, ad 54 poll. longa, ad 2 poll. lata, tenuiter coriacea, nervis lateralibus utrinque circiter 7 uti venarum reticulatione tenuibus indistincte arcuato-connectis ; petioli subgraciles, 2— 3 lin. longi. — Racemi 1-4 in axillis foliorum vel in ramulis nonnullis ob folia (rudimentaria ?) premature decidua quasi in paniculas collecti, griseo-pubescentes. Flores non nisi ex alabastris noti. Fructus 4 poll. diametro; pericarpium extus opacum, fusco-fulvum, intus nitens, 4-2 lin. crassum. Semina 3 lin. diametro. Cameroons: in primeval forest near Bipinde, Zenker, 1240. The fruit of Carpolobia, as represented by the species C. alba, D. Don, and C. lutea, D. Don, is a true berry. The globose or subglobose seeds are covered with a rieh silky tomentum, and possess an ample, fleshy albumen. The embryo consists of a very short subclavate radicle and two very thin foliaceous cotyledons almost as long and as wide as the seed. Garcinia epunctata, Stapf (sp. nov.); affinis G. punctate, Oliver, differt foliorum glandulis eanaliformibus nervos secun- darios venasque transgredientibus longis tenuibus subundulatis (haud punetiformibus vel striiformibus), pedicellis robustioribus distinetis, floribus paulo majoribus, sepalis interioribus magni- tudine minus diversis, phalangibus circiter 8-nas anthera gerentibus. Folia oblonga vel elliptico-oblonga, basi acuta, apice abrupte caudato-acuminata (acumine ad 3 poll. longo), ad. 5 poll. longa, ad velultra2 poll. lata, tenuiter coriacea, nervissecundariis numerosis, circiter 1 lin. distantibus obliquis ut venis tenuibus, glandulis canaliformibus nervos venasque transgredientipus longis tenuibus subundulatis, petiolus 2-3 lin. longus. Flores solitarii vel geminati; bracteæ parvule, late; pedicelli 1 lin. longi. Sepala 4 rotundata, pallide viridia, venosa, interiora majora, 9 lin. diametro. Petala 4 rotundato-elliptica, saturate lutea, ad 4 lin. THE FLORA OF LIBERIA. 87 longa. Stamina (in d) in phalanges 4 antheras circiter 8-nas arcte approximatas distinctas gerentes collecta. | Ovarium (in 9 ) turbinato-globosum, 11 lin. longum ; stigma discoideum, margine minutissime crenulatum. Sinoe Basin, Whyte. The differenees in the glandular system of the leaves of Garcinia punctata and G. epunctata are very striking, and can, without further preparation, be seen in young leaves by holding them up against strong light. Hibiscus Whytei, Stapf (sp. nov.); affinis H. lunariifolio, Willd., differt bracteis epicalycis semper 10 angustis a basi sub- linearibus, corolla minore 14 poll. longitudine vix excedente. Caulis superne pilis stellatis conspersus, demum glabrescens. Folia intermedia deltoideo-ovata vel ovata, basi rotundata vel subtruncata, acuta vel in acumen lineare producta, crenato-den- tata dentibus alternis multo majoribus binis ad diametrum folii maximum positis magis minusve productis, 13-33 poll. longa lataque, utrinque pilis aut stellatis aut simplicibus parce con- spersa; petioli graciles, 4 ad ultra 1 poll. longi, pubescentes ; stipuli laneeolati, 3 lin. longi. Folia floralia ovato-lanceolata vel lanceolata, basi acuta vel subeuneata, quam intermedia minora, breviter petiolata, ezeterum illis similia. Pedicelli 2-5 lin. longi, primo dense stellato-pilosi, demum subglabrati. Bractee epi- calycis 10, a basi sublineares, superne sensim attenuate, 5-7 lin. longe, 1-14 lin. late, sparse stellato-pilose. ^ Calyv 7-8 lin. longus, profunde 5-fidus, segmentis late oblongis acuminatis 2-3 lin.latis extus sparse stellato-pilosis intus villoso-tomentosis. Petala obovata, 14 poll. longa, e siecato lutea, basi purpureo- maculata, extus molliter pubescentia. Capsula ovoideo-globosa, fulvo-strigoso-tomentosa, submatura 7 lin. longa. Within 20 miles of Karkatown, Whyte. Gomphia amplectens, Stapf(sp. nov.); affinis G. Mannii, Oliver, differt foliorum auriculis magnis ultra basin petioli productis post eum imbrieatis et simul ramulum amplectentibus, floribus paniculatis, sepalis sub maturitate majoribus. Folia lanceolata, basi auriculato-cordata auriculis magnis ultra basin petioli productis post eum imbricatis et simul ramulum amplectentibus, acuminata, margine supra medium minute ser- rulata, ad fere 1 ped. longa, ad 3 poll. lata, tenuiter coriacea, costa supra prominente utrinque linea impressa addita, nervis 88 DR. OTTO STAPF ON lateralibus utrinque circiter 16 prorsus curvatis obliquis supra prominulis, venis inconspicuis ; petiolus crassus 3 lin. longus. Inflorescentia paniculata, semipedalis, ramis ad 22 poll. longis subpatulis. Flores sub anthesi ignoti, plerumque 3-ni, fasciculati, fasciculis superioribus 3-2 lin. distantibus; pedicelli 6-7 lin. longi, 1-2 lin. supra basin disartieulati. Sepala oblonga, sub- obtusa, sub maturitate 6-7 lin. longa, ad 3 lin. lata, patentia. Fructus ignotus. Within 20 miles from Karkatown. Gomphia subcordata, Stapf(sp. nov.) ; affinis G. congesta, Oliver (quoad plantam e Sierra Leone), sed foliis basi subcordatis, petiolo erassissimo, ramulis et pedunculis (ima basi excepta) stipulis aphyllis destitutis. Folia late lanceolata vel oblongo-laneeolata, acuta, basi sub- cordata, margine fere a basi minute serrata, ad 6 poll. longa, ad 23 poll. lata, coriacea, costa supra distincte infra vix prominente, nervis lateralibus utrinque circiter 10 subpatulis prorsus arcuatis supra prominulis, venis ineonspieuis; petiolus crassissimus, ad 3 lin. longus. Racemi suberecti, 4 ped. longi pedunculo 1 poll. longo incluso, rhachi subgracili. Flores sub anthesi ignoti, inferne circiter 6-ni fasciculati, fasciculis inferioribus 3-6 lin. distantibus; pedicelli 5-6 lin. longi, 1-1 lin. supra basin disarti- eulati. Sepala sub maturitate ovato-oblonga, obtusa, 4-5 lin. longa, 23-8 lin. lata, patula. Fructus (submaturus) fere globo- sus, 2 lin. diametro. Within 20 miles of Karkatown, Whyte. Olax major, Stapf (sp. nov.) ; affinis O. Mannii, Oliver, differt foliis minus acuminatis, floribus duplo majoribus. Planta glaberrima. Ramali virides. Folia elliptica vel late lanceolata, symmetrica vel asymmetrica, basi obtusa vel sub- acuta, magis minusve acuminata, 34-5 poll. longa, 13-2 poll. lata (acumine 2-6 lin. longo), subcoriacea, utrinque viridis, nervis secundariis utrinque circiter 6 obliquis tenuibus areuato- connectis arcubus circiter 2 lin. a margine distantibus, venis tenuissimis inconspicuis ; petioli 1-2 lin. longi. Racemi ad 4 lin. longi, interdum brevissimi, fasciculiformes, ad 7-flori, basi bracteis paucis vel numerosis squamiformibus vacuis muniti; bractez florales ovate vel lanceolate, 4 lin. longs: pedicelli 4—3 lin. longi. Calyx brevissime cupularis, truncatus. Petala linearia, subacuta, 4 lin. longa. Stamina perfecta 8, 2 alternipetala, THE FLORA OF LIBERIA. 89 loppositipetalum : stamina antheris effotis 5, 4 oppositipetala, lalternipetalum ; filamenta petalis fere tota longitudine adnata, bullata (imprimis staminum imperfectorum). Within 6 miles of Monrovia, Whyte. The petals of Olax Mannii are under 2 lin. long. Uroporrya, Stapf (gen. nov. Olaeacearum). Affinis Opilie, Roxb., sed toto habitu, racemis longissimis, disco annulari indiviso, filamentis quam petalis duplo longioribus distincta. Calyx ad marginem obscurum redactus. Petala 4-5, valvata tandem libera et reflexa. Stamina petalis opposita, hvpogyna, filamentis eapillaribus longissimis ; anthers ovoides, basi affixe. Discus annularis, crassus, undulatus vel truncatus. Ovarium l-loeulare; stigma stylo brevi imposito vel sessile, parvum ; ovulum 1, e placenta ascendente ex apice loculi pendulum. Fructus mihi ignotus; a Dryandro bacca descripta. Frutices glabri, nisi interdum racemorum rhachides pubes- centes. Folia subcoriacea vel membranacea. Racemi longi, pen- duli; braetez nulle vel hyaline. Flores fasciculati, pedicellati. U. angustifolia, Stapf (sp. nov.). Frutex glaberrimus. Folia lineari-lanceolata vel oblonga, acuminata, basi acuta, 6 poll. longa, 1-2 poll. lata, subcoriacea, nervis lateralibus utrinque circiter 7-8 tenuibus uti venis ineonspieuis vel exsiccando prominulis; petioli crassiusculi, ad 21 lin. longi. Racemi ad 10 poll. longi, ebraeteati; pedicelli graciles, 3 poll. longi. Petala ovato-oblonga, subaeuta, 12 lin. longa. Filamenta 3 lin. longa. Stigma sessile. Within 6 miles of Monrovia, Whyte; near the St. Paul's River, 60 miles inland, Reynolds. Also in Sierra Leone, Smeath- man. It is the fruit of the latter which was described by Dryander as baccate in a note in the herbarium of the British Museum. U. latifolia, Stapf (sp. nov.). Frutex racemis exceptis glaber. Folia late elliptica, basi ro- tundata, apice breviter acuminata, 5 poll. longa, 3 poll. lata, fere membranacea, nervis lateralibus utrinque 5, curvatis, tenui- bus distinctis, venis inconspicuis. Racemi ad 8 poll. longi, braeteati; rhachis minute pubescens; braetes ovate, acute, i-3 lin. longs, hyaline, viridi-flavescentes. pedicelli gracillimi, * 90 DR. OTTO STAPF ON 2 lin. longi. Petala 1 lin. longa. ovato-oblonga, obtusa. Stigma sessile. Within 6 miles of Monrovia, Whyte. Professor Engler mentions, in * Natürliche Pflanzenfamilien,' Nachtr. p. 143, under Opilia a species with long pendulous racemes which he calls O. Afzelii. It is evidently a species of Urobotrya; but, in the absence of a description, I am not able to identify it. Another species of the same genus was col- leeted by Kalbreyer near Victoria, Cameroons. I add here its description :— Urobotrya minutiflora, Stapf (sp. nov.). Frutex glaberrimus, 4-6-pedalis. Rami virides. Folia ob- longa vel elliptica vel ovato-oblonga, acuminata, basi rotundata, 4-6 poll. longa, 13-2 poll. lata, firme membranacea, pallide viridia, nervis lateralibus utrinque 5-7 obliquis sub margine in- distincte arcuato-connectis tenuibus, venis inconspicuis ; petiolus erassiuseulus, 1-12 lin. longus. Racemi gracillimi, ad 8 poll. longi, ebracteati; pedicelli gracillimi, 2 lin. longi, 4-6-ni. Petala late oblonga, obtusiuscula, } lin. longa, viridia. Filamenta ad 3 lin. longa. Ovarium in stylum brevem productus. In shady bush, near Victoria, Cameroons, Kalbreyer. Iodes reticulata, Stapf (sp. nov.); affinis Z. africane, Welw., differt foliis breviter petiolatis, eximie reticulatis, inflorescentia fulvo-tomentosa. Rami scandentes, fulvo-pubescentes; fibrilli extra-axillares. Folia ovato-elliptica vel obovato-oblonga, acute acuminata, basi rotundata vel subcordata, sepe asymmetrica, 23-5 poll. longa, 11-21 poll. lata, membranacea, subtus paulo pallidiora, nervis lateralibus utrinque circiter 6 curvatis obliquis uti venarum reti- culatione arcta utrinque prominulis ; petioli 3 lin. longi, fulvo- pilosi. Panicule sspe dichotome subcorymbose, pedunculo 1-2 poll. longo suffultz, terminales, interdum nonnullis minori- bus e foliorum summorum axillis additis, fulvo-tomentose, bracteæ minimze; pedicelli 4 lin. longi. Calyx 4-fidus, segmentis minutis. Corolla 4-fida, extus pubescens, ad 23 lin. diametro, segmentis ovatis subacutis. Sinoe Basin, Whyte. Ampelocissus gracilipes, Stapf (sp. nov.); affinis 4. salmonea, Planch., sed indumento tenui, pedicellis gracilibus glabrescen- tibus, corollis longioribus distinctus. THE FLORA OF LIBERIA. 91 Planta scandens. | Ramuli novelli einnamomeo-floccoso- vel araneoso-tomentell, mox glabrescentes; cirri dichotomi, longi. Folia late cordato-ovata vel subpentagono-cordata, acute acumi- nata, sinu lato, margine serrato-crenata, ad 6 poll. longa, ad 5 poll. lata, membranacea, in gemma dorso cinnamomeo-tomen- tosa, citissime glabrescentia indumento parco araneoso persistente, hine inde purpurascentia, e basi 5-nervia, nervis lateralibus e costa ortis utrinque 4-5 omnibus uti venis tenuibus; petioli paulo ultra 2 poll. longi, graciles. Panicule oppositifolie, pedunculo ad 44 poll. longo suffulte, densiflora, ad 14 poll. longs, 2 poll. late, ramo infimo in cirrum dichotomum commutato, ramis ramulisque primo cinnamomeo-araneosis cito glabratis; bracteæ ovate, tenuiter membranacez, ad 1 lin. long ; pedicelli graciles, ad 1 lin. longi. Calyx patelliformis, } lin. diametro, glaber. Petala 5, oblonga, cucullata, ad 1 lin. longa, glabra. Stamina 5. Ovarium 10-suleatum ; stylus brevissimus, late conicus. Sinoe Basin, Whyte (flores & ). Eriocelum pendulum, Stapf (sp. nov.); affinis Æ. racemoso, Baker, differt floribus multo minoribus, racemis flexuoso-pendulis longissimis. Rami juniores fusco-hirtelli, mox glabrescentes. Folia 2-3- juga; rhachis 13-2 poll. longa, fulvo-hirtella, glabrescens ; foliola infima ad rhachis basin quam reliqua multo minora ceterum iis simillima, intermedia et summa oblonga vel lanceo- lato-oblonga, interdum subobliqua, acuminata, basi subacuta vel obtusa, 21-4 poll. longa, 11-14 poll. lata, chartacea, preter costam sparse adpresse hirtellam glabra, nervis lateralibus utrinque 8-9 prorsus curvatis uti venarum retieulatione tenui subtus prominula; petioluli 1-2 lin. longi. Racemi pedales, graciles, flexuoso-penduli, densiflori, ex toto fulvo-hirtelli ; rhachis fere filiformis, a basi florifera; bracteæ lanceolate, parve; pedi- celli i-i lin. longi, fasciculati. Sepala ovata, subacuta, fulvo- hirtella, 4 lin. longa. Petala anguste lanceolata, 1 lin. longa. Filamenta 13 lin. longa. Near Monrovia, Whyte. Deinbollia polypus, Stapf (sp. nov.); affinis D. insigni, Hook. f., sed foliolis angustioribus, basi magis acutatis tenu- issime acuminatis, panicula exsiecando nigrescente, floribus paulo minoribus pedicellis medio articulatis, staminibus paucioribus differt. Foliola lanceolata, tenuissime acuminata, basi acutata, ad 92 DR. OTTO STAPF ON 8 poll. longa, ad 23 poll. lata (acumine angusto ad 4 poll. longo), chartacea, glaberrima, exsiccata supra saturate viridia, subtus fuscescentia, nervis lateralibus utrinque 7-9 obliquis tenuibus uti venarum retieulatione utrinque (imprimis subtus) eleganter prominulis; petioluli 1 lin. longi. Paniculæ ad 7 poll. long: : rhachis stricta, glabra vel parcissime papilloso-puberula papillis minutissimis; rami divaricati, ad 4 poll. longi; flores numerosi fasciculati in ramulis 1-2 lin. longis minutissime papilloso- puberulis; bractez lanceolate, minute ; pedicelli ? lin. longi, medio artieulati. Calyx 3 lin. longus, sepalis rotundato- ellipticis dorso parce papilloso-puberulis cæterum glabris. Petala elliptico-oblonga, basi in unguem attenuata marginibus ad constrictionem inflexa, 2 lin. longa. Stamina 10; filamenta pilosa, ad 14 lin. longa; antheræ 4 lin. longs. Sinoe Basin, Whyte. The name Deinbollia polypus was chosen in allusion to the numerous persistent pedicel-bases at the ends of the short branchlets of the panicle. The flowers of D. insignis are on the whole less numerous, although they appear just as crowded as in D. polypus, on account of their being larger and supported by stouter pedicels. The latter are, moreover, articulated close to the base. Bersama leiostegia, Stapf (sp. nov.); affinis B. paullinioides, Baker (quoad plantam e Sierra Leone), differt foliolis firmioribus basi obtusis, stipulis multo longioribus lanceolato-acuminatis marginibus ciliolatis exceptis glaberrimis, pedicellis calycibusque parce pubescentibus. Jamuli glabri, robusti. Folia 6-8-juga; rhachis magis minusve interrupte alata alis interdum angustissimis, 1 ped. longa vel ultra; foliola infima 13-24 poll. supra rhachis basin sita, oblonga, anguste acuminata acumine lineari 6 lin. longo, basi obtusa vel rotundata, ad 5 poll. longa, ad 2 poll. lata, firme membranacea, glabra, nervis lateralibus utrinque cirea 11 tenui- bus, venarum reticulatione utrinque prominula; petioluli vix l lin. longi; stipule lanceolat:e, acuminate, marginibus exceptis glaberrims, ad 14 lin. longe. Racemi rigidi, fere 6 in. longi ; rhachis ad 13 poll. a basi nuda, superne parce pilosula ; bracteæ subulatz, laxe pilose; pedicelli tenuiter parce pubes- centes, ad 3 lin. longi. Calyx 2 lin. longus, parce pubescens ; lobi ovati, subobtusi, infimi duo fere ad apicem connati. Petala oblonga cum ungue 4-5 lin. longa, lamina fulvo-cinerea tomen- THE FLORA OF LIBERIA. 93 tella reflexa. Stamina 4; filamenta per paria basi connata, lanceolata, pubescentia. Ovarium 4-loculare, sericeo-tomen- tosum, in stylum superne glabrum contractum. Sinoe Basin, Whyte. Spiropetalum triplinerve, Stapf (sp. nov.) ; affinis S. hetero- phyllo, Gilg, differt foliis 3-4-jugis, foliolis supra subglaucis subtus einnamomeis triplinervibus (7. e. nervis secundariis infimis 2 fere a basi ortis, ceteris perpaucis remotis), venis transversis subhorizontalibus utrinque eleganter prominulis. Ramuli florifert robusti, primo minutissime ferrugineo - tomentelli, deinde glabrescentes, cortice fuscescente. Folia 3-4-jugi ; rhachis 2 ad fere 4 poll. longa, tenuissime pubescentia, deinde glabrescens ; foliola oblonga vel elliptico-oblonga, breviter acuminata, basi subacuta vel obtusa, ad 2 poll. longa, 1 poll. lata, coriacea, supra glabra, subtus in costa tenuiter pubescentia exterum glabra, nervis lateralibus utrinque 3 admodum obliquis tertio a secundo longe remoto, venis utrinque prominulis ; petioluli 1-1j lin. longi, tomentelli. Racemi densi, sepe compacti, solitarii vel fasciculati, }-1} poll. longi, a basi flori- feri ubique fulvo-velutini ; braetes ovate, parve; pedicelli brevissimi. Calyw 13-1? lin. longus, ad medium 5-partitus, segmentis oblongo-ovatis obtusis. Petala lorato-linearia e basi latiore, ad 5 lin. longa. Stamina longiora calyce breviora. Carpella velutino-tomentosa cum stylis calycem sequantia. Near Monrovia, Whyte. Connarus libericus, Stapf (sp. nov.); affinis C. floribundo, Schum. & Thonn., differt imprimis inflorescentia, nempe pani- culis brevibus gracilibus plerumque racemiformibus versus ramorum apices fasciculatis. fami floriferi robusti, cortice cinerascente lenticellato. Folia circiter 4-juga, glaberrima; rhachis 3-4 poll. longa, teres; foliola elliptica vel obovato-elliptica, subabrupte acuminata, basi obtusa vel subacuta, 24-8 poll. longa, 14-14 poll. lata, tenuiter coriacea, nervis lateralibus utrinque circiter 5 valde obliquis, venis transversis tenuissimis; petioluli 2 lin. longi. Panicule 13-2 poll. longze, sepe racemiformes vel saltem ramis longioribus paucis, a j—l poll. supra basin divisæ, tenuiter ferrugineo- pubescentes, graciles, fasciculate, numeross versus ramorum apices; bractez minimz; pedicelli ad 1 lin. longi, graciles, medio artieulati. Calya vix 1 lin. longus, profunde 5-partitus, segmentis ovatis subobtusis tenuiter pubescentibus. Petala 94 DR. OTTO STAPF ON lanceolato-ligulata, 3 lin. longa, ? lin, lata. Stamina majora calycem :wquantia. Carpella sericeo-villosula, cum stylis gra- cillimis 14-1} lin. longa. Within 6 miles of Monrovia, Whyte. Connarus Reynoldsii, Stapf (sp. nov.); affinis C. floribundo, Schum. et Thonn., differt foliis conspicue nervosis, inflorescentiis axillaribus et terminalibus quam foliis brevioribus densissimis, petalis longioribus, staminibus stylisque multo brevioribus. Frutex ramis cortice brunneo tectis multi-lentieellatis. Folia 3-4-juga ; rhachis gracilis tenuiter rufo- vel fulvo-pubescens vel tomentella; ima juga circiter 2 poll. a basi distantia; foliola oblonga vel elliptieo-oblonga, breviter acuminata aeumine cuspi- dato, basi rotundata vel subacuta, 2-32 poll. longa, 3-14 poll. lata, coriacea, supra glaberrima, nitidula, infra in costa nervisque magis minusve tenuiter fulvo-pubescentia esterum fere glabra, nervis lateralibus utrinque 2-3 valde obliquis et prorsus curvatis, venarum retieulatione laxa prominula; petioluli tomentelli, lj lin. longi. Panicule numerose in foliorum axillis et ad ramorum apices dense congeste, cirea 2 poll. longe, densiflore, fulvo-velutine ; braetez ovate vel lanceolate, acute, pedicellos 2-12 lin. longos paullo superantes. Calyw fulvo-velutinus, 12 lin. longus, ad medium 5-lobus, lobis ovatis subacutis vel obtusiuseulis. Petala anguste lorato-linearia, superne longe attenuata et erispo-undulata, circiter 7-8 lin. longa, ad 14 lin. lata. Stamina perfecta 10; longiorum filamenta paulo ultra 13 lin. longa, breviorum duplo breviora, Carpella 5, fulvo- tomentella, eum stylis parum brevioribus vix ? lin. longa, stigmata subdisciformia. Ovula 2, collateralia. Liberia, near the St. Paul’s River, about 70 miles from the coast, H. Reynolds. Dalbergia Ecastaphyllum, Taub., forma trifoliolata, Stapf, a typo differt foliolis 3, terminali quam lateralibus multo majore. Foliorum rhachis 1-14 poll. longa; foliola oblonga vel ovato- oblonga, basi rotundata, apice sensim breviter acuminata, ter- minali ultra 6 poll. longa, ad 2? poll. lata, lateralia minora, coriacea, supra opaca, subtus subglaucescentia, pilis minutissimis sparsis adpressis, nervis lateralibus utrinque circiter 9, ceeterum ut in typo. Flores in specie majusculi, ceterum ut in typo. Sinoe Basin, Whyte. This is a rather striking form, which might be taken at the THE FLORA OF LIBERIA. 95 first glance for a distinct species; but the structure and general aspect of the leaflets (apart from the size) and of the flowers are absolutely as in Dalbergia Ecastaphyllum. The only other pinnate-leaved specimens of D. Ecastaphyllum examined by me are from Liberia (Grand Bassa, Vogel ; Dinklage, 1986) and the Cameroons (Zenker, 2153). In these the leaflets are exactly (also in size) like those of the typical unifoliolate leaves of D. Ecastaphyllum. They were, probably on account of the number of leaflets, distributed as D. Monetaria, which has glabrous glossy long acuminate leaflets with a more marked venation. I have seen no examples of D. Monetaria from Africa, Heudelot’s specimens quoted under this name in ‘ Flora of Tropical Africa,’ ii. p. 236, belonging to a distinct species, for which I propose the name D. Heudelotit, Stapf. I add a description of it :— Dalbergia Heudelotii, Stapf (sp. nov.); affinis D. Ecastaphyllo, Taub., differt foliis normaliter 3-7-foliolatis, fructibus asym- metrice obovatis majoribus crassioribus sublignosis, bullatis, ferrugineo-velutinis. Frutex ramis cinerascentibus vel fuscescentibus. Folia 3- vel sepius 5- rarius 7-foliolata ; rhachis ad 34 poll. longa, tenuiter ferrugineo-pubescens, demum glabrata; foliola ovata vel elliptica, breviter obtuseque acuminata vel subobtusa, basi rotundata, terminale maximum ad 6 poll. longum, ad 34 poll. latum, supra glabra arcte tenuiterque reticulata, subtus pallide fuscescentia, pilis minutissimis adpressis aspersa, nervis lateralibus (in ma- joribus) utrinque 8-10 prorsus arcuatis tenuibus uti venis prominulis; petioluli 2-3 lin. longi. Racemi vel paniculi pauci- ramosi ad 14 poll. longi, ubique rufo-villosuli; bracteæ bracteo- leque minute, oblonge ; pedicelli ad 13 lin. longi. Calyæ 13 lin. longus, breviter perlate dentatus, rufo-tomentosus. Corolla 5 lin. longa, petalorum unguibus demum e calyce exsertis; vexilli lamina late orbieulari-ovata, 2-loba. Legumen immaturum dense rufo- vel fulvo-tomentosum marginibus latis undulatis, maturum asymmetrice obovatum, ad lj poll. longum, ad 10 Jin. latum, uno latere convexum, bullatum, ferrugineo-velutinum, sutura incrassata, valvis ad 2 lin. crassis lignosis; stipes tenuis fere 2 lin. longus. Semen unicum. Senegambia, Rio Nunez, Heudelot, 623. Sierra Leone, Kah- reni, Scott Elliot, 5626*, Without precise locality, Afzelius. D. Monetaria, Linn. f., differs in having glabrous (in the 96 DR. OTTO STAPF ON typical form), distinctly and often long acuminate leaves, and glabrous, neither undulate nor bullate fruits. Ostryocarpus major, Stapf (sp. nov.); affinis O. ripario, Hook. f., differt foliolis 7-foliolatis, foliolorum venis utrinque plane ineonspieuis, panieule ramis longioribus, floribus paulo majoribus. Folia T-foliolata; rhachis gracilis, basi incrassata, circiter 4 poll. longa, glabra; foliola infima 3 poll. supra basin, terminale elliptieum, breviter aeuminatum, basi subobtusum, 34 poll. longum, 2 poll. latum, coriaceum, glaberrimum, nervis later- alibus utrinque 5-6 obliquis pertenuibus, venis plane incon- spicuis ; petiolulus 3j lin. longus. Panicule pedales, breviter pedunculate, ubique dense tenuissime ferrugineo-pubescentes ; rhachis subrobusta, curva; rami racemos densos exhibentes ad 2 poll. longi, patuli; bractew ad ramorum bases mox deeilusm, ad pedicellos vix ullæ ; bracteole ad calycum bases 2, minute; pedicelli inferiores sspe geminati, graciles, 1-14 lin. longi. Calyx obovoideus, 2-23 lin. longus, eodem indumento induto ae rhachis et ramis panieule, dentibus latissime brevissime tri- angularibus. Corolla 5-53 lin. longa, exsiecando nigrescens ; vexillum orbiculare, emarginatum, brevissime unguiculatum ; ale et carina vexillo æquilongæ, unguibus 2 lin. longis. Sta- minum tubus 33 lin. longus. Ovarium dense minute pubescens ; ovula 8. Within 6 miles of Monrovia, Whyte. Macrolobium obliquum, Stapf (sp. nov.); affinis JM. diphyllo, Harms, differt folis 7-jugis, floribus majoribus, longius pedi- cellatis. Folia 7-juga; rhachis ad 10 poll. longa, basi incrassata, atro- ferrugineo-hirtella, demum glabrescens ; foliola ima 3 poll. supra rhachidis basin, vix 23 poll. longa, intermedia et summa ad 6 poll. longa, ad 13 poll. lata, omnia oblique lanceolata vel lanceolato- oblonga, oblique tenuiter acuminata, basi obtusa, dimidio extus spectante quam altero fere duplo latiore, glabra, nervis lateralibus utrinque 13-15, 1-12 lin. sub margine nervo collectivo eleganter connectis subtus prominentibus, venarum retieulatione utrinque tenui prominula ; petioluli llin. longi. Panicula ultra à-pedalis ramis paucis divaricatis ad 3 poll. longis ubique tenuiter ferru- gineo-hirtella; bracteæ obovate, naviculares, 5 lin. longe, mox THE FLORA OF LIBERIA. 97 decidum ; bracteole 2 flores involuerantes obovate, fere 6 lin. longs, obtusiuscule ; pedicelli ad 6 lin. longi. Calyx fere ad basin 5-fidus, segmentis lineari-lanceolatis ciliolatis 21-21 lin. longis. — Vexillum longiuseule unguiculatum, lamina rotundatum profunde 2-lobum, petala cetera calycis segmentis simillima eisque :equilonga. Stamina fertilia 3; filamenta pilosa ; anthere 2j lin. longs; staminodia pauca, minutissima, denti- formia. Ovarium rufo-villosum ; ovula circiter 5. Sinoe Basin, Whyte. Acioa Whytei, Stapf (sp. nov.); affinis A. padlescenti, Baill., differt folis minoribus, inflorescentiis glabris rhachi parce pilosula excepta. famuli glabri, cortice brunneo. Folia oblongo-elliptiea vel ovato-lanceolata, subacuminata, basi rotundata, 21—4 poll. longa, 1-13 poll. lata, coriacea, pallide viridia, nervis lateralibus utrinque 6-7 nti venarum reticulatione laxa utrinque prominulis ; petiolus crassus, l lin, longus. Racemi rhachi parce pilosula excepta glabri, vix pollicares, a basi densiflori ; bracteæ ovato-lanceolate, ad 2 lin, long, bracteoleque pallide, tenuiter membranacez, glabre ; pedicelli graciles, glabri, 3-4 lin. longi. Jieceptaculum tenue, glaberrimum, 8-9 lin. longum. Calyw 4-5 lin. longus; sepala elliptiea, obtusissima, extus preter margines in alabastro imbrieatos glaberrima, intus albo-velutina. Petala elliptico- oblonga, breviter unguiculata, 2-loba, 4i lin. longa. Filamenta in laminam liguliformem 12 poll. longam 4 lin. latam juneta, parte summa libera 3—4 lin. longa excepta. Within 6 miles of Monrovia, Whyte. A specimen collected by G. F. Scott Elliot (no. 5521) near Kafogo, Sierra Leone, after flowering but with some remnants of inflorescences, seems to be identical with A. Whytei. Cassipourea czsia, Stapf (sp. nov.); aftinis C. parvifolie, Stapf (=Dactylopetalum parvifolium, Scott Elliot), differt foliis supra cesiis, calyce cæsio, petalis dense lanatis, disco intrastaminali multo minus alto lobulato lobulis cum filamentis aiternantibus. Frutex ramulis gracilibus primo adpresse pubescentibus cito glabratis, cortice brunneo. Folia anguste elliptica vel obovata, iatiuseule acuminata, basi cuneata vel acuminato-cuneata, 2- kz poll. longa, 13-2 poll. lata, subchartacea, glabra, supra eximie cæsia subtus viridia, nervis lateralibus utrinque circiter 6, 1- 98 DR. OTTO STAPF ON 2 lin. sub margine arcuato-connectis tenuibus uti venarum reticu- latione tenuissima utrinque vix prominulis; petioli 2-23 lin. longi, subadpresse hirtelli. Stipulæe lanceolate, fulvo strigilloso- hirtelle, ad 2 lin. longs. Flores 1—4 in axillis foliorum congesti ; bracteze minute, ovate, hirtelle ; pedicelli 4 lin. longi, glabri. Calyx semiglobosus, 2 lin. longus, extus glaber, cæsius, intus pallide virescens et sericeo-velutinus, ad medium 5-fidus, seg- mentis ovatis subacutis. Petala anguste lanceolata, ad 23 lin. longa, fimbriata fimbriis ad 2 lin. longis dense albo-lanatis. Stamina 15, epipetala 5 ad 23 lin. longa, ima basi disco intra- staminali brevissimo intus piloso lobulato, lobulis eum filamentis alternantibus adnata. Ovarium tomentosum, 3-loculare. Within 20 miles of Karkatown, Whyte. Eugenia Whytei, Sprague (sp. nov.) ; affinis E. calophylloidi, DC., differt ramulis glabris, foliis minus coriaceis, subtus glan- dulis prominentibus punctatis, receptaculo glabro styloque minore. Ramuli glabri, cortiee cinereo vel subfusco. Folia elliptico- ovata vel oblongo-ovata, apice obtuse acuminata, basi in petiolum attenuata, 24-44 poll. longa, 1-13? poll. lata, tenuiter coriacea, glabra, supra olivacea vix nitidula, subtus pallidiora opaca, nervis lateralibus primariis utrinque 7-9 subtus prominulis, secundariis iis parallelis hine inde interjectis, tertiariis supra ineonspicuis subtus subtiliter reticulatis; glandule utrinque precipue subtus prominentes; petiolus 2-3 lin. longus, supra sulcatus. Flores plures in axillis fasciculati; pedicelli 1-3 lin. longi, glabri, nigrescentes ; bracteole deltoidez ciliate. Recep- taculum glabrum. Sepala orbiculari-ovata, obtusissima, ciliolata, exteriora 3 lin., interiora i lin. longa. Petala breviter obovata, 2 lin. longa, ciliolata. Stamina iis E. callophylloidis similia. Ovarium biloculare ; stylus 2 lin. longus, stigmate peltato. Sinoe Basin, g, Whyte ; within a radius of 6 miles from Monrovia, $, Whyte. Osbeckia liberica, Stapf (sp. nov.); aflinis O. senegambiensis, Guill. & Perr., differt bracteis haud scariosis, citissime deciduis, sepalis subulatis, connectivo basi distincte elongato. Herbe annua (?). Caulis gracilis, subquadrangularis, minute adpresse strigilosus, pilis plerumque nigrescentibus. Folia lanceolata, rarius ovato-lanceolata, apice longe attenuata, sub- THE FLORA OF LIBERIA. 99 acuminata, basi rotundata vel subacuta, 2-21 poll. longa, $-3 poll. lata, membranacea, utrinque adpresse strigilosa (infra tenuius), 3-5pli-nervia; petioli graciles, 3 lin. longi. Panicule pauciflore, ramis demum ad 14 poll. longis gracilibus, eodem indumento indutis ae cauli nisi densiore, internodiis intermediis sub maturitate 2-3 lin. longis; braetes tenuiter membranacex, lanceolate, 23 lin. longs, mox decidum; pedieelli 1 lin. longi. Receptaculum ovoideo-globosum, sub anthesi 24 lin. longum, deinde paulo elongatum, ad medium strigillosam supra medium squamulis apice stellato-setosis additis. Sepala subulata, 2 lin. longa, patule setosa, mox decidua. Petala rosea, late oblique obovata, 4 lin. longa. Stamina subæqualia ; filamenta 2 lin. longa, antheræ connectivo basi producto antice appendiculo parvo 2-lobo instructo, connectivi continuatione et appendice in 5 alternis longioribus. Ovarium apice pilosum. Within 6 miles of Monrovia, Whyte; Sinoe Basin, Whyte. Dissotis paucistellata, Stapf (sp. nov.); affinis D. petiolate, Hook. f., differt indumento rudiore copiosiore, florum capitulis majoribus, receptaculo stellato-setoso. Caulis glanduloso-hirsutus pilis patulis apice glanduligeris. Folia ovata, subacuminata, basi rotundata vel subcordata, 11-2 poll. longa, 1-14 poll. lata, 5-nervia, utrinque strigilloso-hirsuta ; petioli ad 3 poll. longi, dense hispidi. Flores in capitulos cir- citer 10-floros diametro 1 poll. foliis suffultos bracteatos collecti. Bractee tenuiter membranacex, exteriores late ovate ad 4 lin. long, interiores angustiores, ciliate et in dorso sparse pilose. Receptaculum e basiangustata oblongum, anthesi peracta ovoideo- oblongum, supra medium constrictum, ad 7 lin. longum ima basi setis tenuibus simplieibus glanduligeris supra usque ad medium setis stellatis sparsis instructum. — Sepala lanceolata, acuta, 3 lin. longa, parce glanduloso-setosa, persistentia. Petala rosea, late cuneato-obovata, ad 10 lin. longa. Stamina 5 filamentis 4 lin. longis et connectivo 3 lin. ultra antherarum bases producto antiee ima basi 2-lobato postice ecalcarato, alia 5 filamentis 4 lin. longis connectivo brevissime producto antice 2-lobato postice minute ealearato. | Ovarium apice setosum. Within 6 miles of Monrovia and in the basin of Sinoe, Whyte. Memecylon Simii, Stapf (sp. nov.); affinis M. polyanthemo, Hook. f., differt floribus multo majoribus in cymis densioribus. LINN. JOURN.— BOTANY, VOL. XXXVII. I 100 DR. OTTO STAPF ON Rami cortice fuscescente. Folia oblongo-elliptica, abrupte in acumen obtusum vel subaeutum ad 6 lin. longum latiusculum contracta, basi breviter aeutata, 21-5] poll. longa, 14-24 poll. lata, coriacea, opaea, subtus pallidiora, nervis venisque plane obseuris, costa subtus prominente obtusa, supra tenuiter cana- lieulata; petioli vix 1 lin. longi. Cyme@ axillares, ad ł poll. longi, a basi fere divise, dens: ; bracteæ ovate, ad 1 lin. longe, citissime decidue; pedicelli 4-14 lin. longi. Receptaculum obovoideo-globosum, 1} lin. longum. Sepala latissime ovata, 1-2 lin. longa, ad 14 lin. lata. Petala elliptico-oblonga, crassa, vix l1 lin. longa, obtusa. Sinoe Basin, Whyte. Homalium molle, Stapf (sp. nov.) ; affinis H. stipulaceo, Welw., differt folis subtus molliter pubescentibus, petalis in fructu maturo multo brevioribus. Rami juniores molliter pubescentes, demum magis minusve glabrescentes, cortice brunneo vel fusco. Folia lanceolata vel oblonga vel elliptica, acuminata, basi rotundata, subaeuta, margine repando-serrata, 4-63 poll. longa, 13-33 poll. lata, subcoriacea, supra preter costam primo minutissime puberulam glabra, subtus molliter pubescentia pube persistente, nervis lateralibus utrinque cireiter 9 magis minusve prorsus curvatis venis transversis laxiuseulis uti retieulatione subtus subprominula ; petioli 2-3 lin. longi, tomentelli; stipule admodum varie mox decidue vel subpersistentes, lanceolate usque orbiculari-auriculate sepissime oblique, ad 6 lin. longs, majores foliacez.. Panicule floribunde terminales et e foliorum superiorum axillis, ad 10 poll. longa, ubique tenuiter molliter cinereo-pubescentes, ramis gracilibus sspe subpendulis ad 6 poll. longis; florum fasciculi approximati vel inferiores magis distantes; bractew minime ; pedicelli brevis- simi. Receptaculum turbinatum, molliter pubescens, 3 lin. longum. Sepala triangulari-ovata vel ovato-lanceolata, acuta, receptaculo paulo breviora. Petala sub anthesi oblonga, 3 lin. paulo excedentia, in fructu maturo ,indurata obovato-oblonga, 1 lin. longa, albo-pubescentia. Stylus apice brevissime 3-4-fidus. — H. stipulaceum, Mast. in Oliver, Fl. Trop. Afr. ii. p. 498 (in part). Sinoe Basin, Whyte. Also in Sierra Leone, by the Bagroo River, Mann, 881. Mann's specimen was referred by Masters, l. c. to H. stipu- laceum, Welw. (no. 2495). This has, however, glabrous leaves, THE FLORA OF LIBERIA. 101 petals which in the fruit are over 3 lin. long and thin, and much more deeply divided styles. ANDROSIPHONIA, Stapf (gen. nov. Passifloracearum). Affinis Paropsie, Noronha, differt inflorescentia terminali paniculata foliata, filamentis pubescentibus inferne dilatatis et in tubum ovarium cingentem connatis. Flores hermaphroditi. Calycis tubus (receptaculum) turbi- natus, brevis; sepala 5, oblonga, acuta, tenuissime velutina. Petala sepalis æquilonga et simillima nisi angustiora et paulo tenuiora. Corona fere ad basin multipartita, segmentis late linearibus copiose tenuiterque fimbriato-laceratis. Stamina 5; filamenta inferne dilatata et in tubum ovarium cingentem e gyno- phoro summo ortum connata, pubescentia. Ovarium breviter stipitatum, ellipsoideum ; styli 3, liberi, antheras attingentes ; stigmata globosa; ovula 2 in unaquaque placenta, collateralia. Capsula (immatura) subglobosa, tenuissime velutina. Semina ignota.— Frutex sempervirens. Folia alterna, obscure serrulata, basi biglandulosa. Stipul@ nulle. Flores griseo-virescentes, in cymulas paucifloras sessiles vel subsessiles arcte contractas racemose dispositis, racemis ad ramorum apices in paniculas foliaceas collectis; folia floralia flores squantes vel (inferiora) longiora, basi glandulis 2 magnis nigrescentibus notata. A. adenostegia, Stapf (spec. unica). Rami glabri, cortice brunneo tecti. Folia oblonga, sensim in longiusculum acumen attenuata, basi breviter constricta vel sub- acuminata, margine obseure remote serrulata, ad 6 poll. longa, 2 (vel ultra) pollices lata, chartacea, glaberrima, ima basi utrinque glandula nigreseente notata, nervis lateralibus utrinque circiter 8 obliquis sub margine areuato connectis uti venarum reticu- latione eleganter prominulis. Panicula rigida, circiter 4 poll. longa ; rami (racemi) stricti, 1-23 pull. longi, tenuissime fulvo- velutini, rhachi gracili ; folia floralia subcoriacea, ovata vel elliptica, acute acuminata vel mucronata, cum petiolo distincto 3-1 poll. longa glandulis 1 lin. diametro. Cymule sessiles vel subsessiles arcte contracts, pauci- vel uniflore ; bracte minute ; pedicelli basi disartieulati, 13-23 lin. longi, demum elongati et nutantes. Calyx 4-6 lin. longus, post anthesin magis minusve auctus ; receptaculum 14 lin. altum; sepala sub anthesi basi vix 1 lin. lata. Petala. linearia, 4 lin. lata, grisea. Corona 12 102 DR. OTTO STAPF ON paulo ultra 1 lin. alta. Staminum tubus ovoideus, vix 13 lin. altus; filamentorum pars libera 2} lin. longa uti tubus extus pubescens; anthera oblonge, 2-5 lin. longe, medio dorso affixe. Ovarium 3 lin. longum; styli pertenues 3 lin. longi. Fructus (immaturus) in pedicello recurvo pendulus, subglobosus, 3 lin. longus. Within 6 miles of Monrovia, Whyte. Soyauxia grandifolia, Gilg et Stapf (sp. nov.); affinis S. gabonensi, Oliver, ct S. glabrescenti, Engl., differt foliis lineari- oblongis amplis et spicis densis, a S. glabrescente etiam sepalis fulvo-sericeis. Ramuli juniores tenuissime fulvo-tomentelli. Folia lineari- oblonga, acuminata acumine acuto longiusculo, basi obtusa vel subaeuta, fere pedalia, 13-3 poll. longa, coriacea, utrinque glabra, viridia, eosta validiuseula, nervis lateralibus utrinque 13-15 obliquis sub margine ipso arcuato-connectis Subtus prominen- tibus; venis transversis et reticulatione elegante utrinque pro- minulis ; petioli crassi, 2-24 lin. longi. Spice densissime, a basi floriferz, 24-5 poll. longs, cylindric, undique fulvo-velutine ; bracteæ ovato-lanceolate acuminate, ad 1 lin. longe. Sepala ovata, 14 lin. longa, extus velutina, intus glabra. Petala late elliptica vel obovato-rotundata, ad 13 lin. longa. Filamenta ad 8 lin. longa. Styli 3 ad lin. 3 longi. Ovula 2 in unaquaque placenta. Capsula pedicello valde inerassato semigloboso insi- dens, basi calyce persistente cincta, valvis 3 late obovatis duris 1 poll. longis fere 1 poll. latis dehiscens. Sinoe Basin, Whyte; Grand Bassa, Dinklage, 2051. Modecca tenuispira, Stapf (sp. nov.); affinis M. Manni, Mast., differt foliis minoribus subtus magis conspicue recticulatis, ey mis breviter peduneulatis, floribus triplo majoribus, calyce minus profunde diviso. Planta scandens, glaberrima. Folia oblonga vel elliptico- oblonga, breviter subacuminata, basi subaeuta, 21-3 poll. longa, 1-13 poll. lata, subehartacea, nervis lateralibus utrinque 4 uti venarum retieulatione subtus prominula; petioli 4-6 lin. longi, apice 2-glandulosi. Cyme axillares, 2-4-flore ; pedunculi 14- 3 lin. longi; pedicelli 1-2 lin. longi. Flores g: Receptaculum depresso-globosum, vix 13 lin. altum. Calya tubulosus, 5-lobus, tubo (receptaculo) excluso ad 6 lin. longus, lobis oblongis obtusis THE FLORA OF LIBERIA. 108 23 lin. longis, marginibus hyalinis integris undulatis. Petala lineari-lanceolata, basi longe attenuata, fimbriata, 5 lin. longa. Corona tenuiter membranacea, annuliformis, tenuissime laciniata laciniis pilosulis. Glandule 5, breves, clavato-filiformes. Fila- menta 3 lin. longa; anthere 3 lin. longs. Flores Q ignoti. Semina compressa, oblique ovata, 5 lin. longa, 3 lin. lata, secun- dum margines obscure tuberculata. Sinoe Basin, Whyte. Begonia Whytei, Stapf (sp. nov.); habitu B. Scutulo, Hook. f., simillima, differt floribus minoribus, petalis flavis, filamentis basi in eolumnam connatis, capsulis multo magis basin versus attenu- atis 4-alatis. Caulis brevis, repens, radicans, hirsuto-villosus, cataphyllis ovatis tenuibus fimbriatis 2 lin. longis. Folia peltata, oblique ovata, acuminata, ad 5 poll. longa, ad 3 poll. alta, margine obscure repando-serrata, membranacea, supra pareissime infra et in marginibus copiosius setulosa (imprimis in nervis), circiter 7-nervia, nervis seeundariis venisque tenuibus ; petioli ad 5 poll. longi, magis minusve hirsuti. Zaflorescentie terminales, cymis 3-4-floris longe pedunculatis racemose dispositis, magis minusve hirsute ; rhachis communis ad 1 poll. longa; bractew ad pedun- culorum bases eataphyllis simillimz, bracteæ cy mas subtendentes ovate vel oblonge fimbriatw, quam inferiores multo minores; pedunculi graciles ad 3 poll. longi; pedicelli florum g 4-5 lin. longi, lorum 9 vix ulli. Flores d ante 9 aperti, mox decidui, plerumque 3 in unaquaque cyma: sepala 2, rotundata, flava, 4 lin. diametro. Filamenta basi in columnam distinctam brevem connata, parte libera anthere equilonga, } lin. longa. Flos 9 unicus in unaquaque cyma: sepala ut in d. Styli 4 simplices, basi connati. Stigmata 4, semilunaria, haud torta. Fructus obpyramidalis, basi longe attenuatus, ad 9 lin. longus, 4-alatus, alis apice obtusis ad 24 lin. latis. Placente integre. Semina ellipsoideo-globosa, laxe reticulata. Sinoe Basin, Whyte. B. Scutulum, Hook.f., is described as having bipartite placentas. There is no female flower with the specimen which we have at Kew, and only one fruit, which I do not wish to sacrifice. I am therefore not able to decide whether the affinity of B. Whytei and B. Seutulum is really as great as their general similarity would suggest. 104 DR. OTTO STAPF ON Begonia Simii, Stapf (sp. nov.) ; affinis B. auriculate, Hook. f., differt caulo erassiore succulento purpureo, foliorum auriculo tota longitudine libero (haud petiolo adnato) nervis basalibus e sinu intimo ortis, stipulis ellipticis deciduis. Caulis carnosus, purpureus uti tota planta glaberrimus. Folia oblique ovata, acuminata, basi inequaliter auriculato-cordata, auriculo exteriore 3-5 lin. longo rotundato a basi libero, margine obseure dentato, dentibus in setulam minutam abientibus, ad 4 poll.longa, ad 2 poll. lata, carnosula, nervis preter costam circiter 5 ex ima basi ortis, nervis secundariis e costa utrinque 2-8 ; petioli 4-7 lin. longi, carnosuli ; stipule decidue, ovato- vel oblongo-elliptiez, parce ciliato-dentate, tenues, 23 lin. longs. Cyme axillares, pauciflore, 1 poll. breviores, pedunculo gracili suffulte, bracteate, bracteis similibus nisi latiores. Flores d pedicello gracili 2-24 lin. longo suffulti: sepala 2, orbicularia basi obscure cordata, 3 lin. diametro, purpurea. Filamenta libera, 4-8 lin. longa; antheræ oblonge, circiter } lin. longs. Flos 9 brevissime pedicellatus: sepala ignota. Styli tres, basi con- nati ; stigmata hippocrepiformia, longiuscule papillosa. Capsula in pedunculo nutans, trigona, trialata, 9 lin. longa, alis demptis 3 lin. diametro, alis rotundatis maxima 3 lin. lata. Sinoe Basin, Whyte. Mussaenda conopharyngiifolia Stapf (sp.nov.); affinis M. tenui- flore, Beuth., differt indumento inflorescentiarum patule hirsuto, foliis supra preter costam setulosam glaberrimis etiamque infra costa nervisque exceptis setulosis glaberrimis, sepalis linearibus acuminatis, corollis anguste infundibuliformibus tubo latiore limbo majore. Ramuli hirsuti mox glabrescentes. Folia late elliptica, breviter acuminata, basi obtusa, ultra 6 poll. longa, ad 34 poll. lata, preter costam (utrinque) et nervos laterales subtus parce setu- losos glaberrima, nervis lateralibus 7-8, venis transversis laxis ; petioli 2 Jin. longi, setulosi; stipule profunde bifida, ad 4 poll. longs, segmentis lineari-subulatis setulosis. Cym@ corymbose disposite, ubique setuloso-hirsutz ; peduneulus brevis; bracteæ lineares, superne attenuatz, fere l poll. longi, setulose; pedi- celli brevissimi. Receptaculum hispidulum setulis nonnullis longioribus additis. Sepala linearia, superne attenuata, acuta, 4-6 lin. longa, 3 lin. lata, setoso-ciliata, pauca uniuscuiusque inflorescentie foliacea lutea, late elliptica, acuminata, petiolata, THE FLORA OF LIBERIA. 105 ultra 3 poll. longa, 2 poll. lata, hispidula. Corolle aurantiace tubus 10-11 lin. longus, infundibuliformi-tubulosus, superne dilatatus, ad 1} lin. latus, patule flavo-setulosus; lobi latissime ovati, apieulati, 3 lin. longi, ore aureo-tomentosi. Sinoe Basin, Whyte. Mussaenda macrosepala, Stapf (sp. nov.); affinis M. tristig- matice, Cummins, differt sepalis et corollis multo majoribus. Ramuli subgraciles patule hirsuti. Folia elliptica vel elliptico- oblonga vel sublauceolata, breviter acute acuminata, basi obtusa vel aeuta, ad 4 poll. longa, 2 poll. lata, supra adpresse setulosa, subtus in costa setulosa in nervis minute adpresse pilosa, margine hirsutiuscula, nervis lateralibus utrinque circiter 17, prorsus curvatis, nervis transversis tenuibus; petiolus dense hirsutus, 21 lin. longus; stipulew bipartite segmentis subulatis, appresse hirsute, 4-6 lin longe. Cyme densiflorz, ad ramulorum apices 2-3, dense undique flavo-hirsute ; pedunculi 3 ad ultra 1 poll. longi; bractes lineari-lanceolate, acuminate, ad 4 lin. longe ; pedicelli brevissimi. Receptaculum densissime flavo-hispidum, vix 3 lin. longum. Sepala lanceolata, acute acuminata, 10-13 lin. longa, 3 lin. lata, utrinque hirsuto-tomentosa. Corolle aurantiace tubus subcylindricus, supra medium ampliatus, ultra 1 poll. longus, dense aureo-tomentosus; limbus lobis ovatis apiculatis extus sericeo-tomentosis, 1 poll. diametro, ore aureo- velutino. Sinoe Basin, WAyte. Sabicea discolor, Stapf (sp. nov.) ; affinis S. venose, Benth., differt foliis subtus albidis dense araneoso-tomentosis, calycis segmentis quam receptaculo, preecipue in fructu, brevioribus. Ramuli fusco-cinerei adpresse hirsuti. Folia ovata vel elliptica, acuminata, basi rotundata, ad 4 poll. longa, ad 1? poll. lata, chartacea, supra sparse hirsuta, subtus albida, araneoso-tomentosa et in nervis adpresse hirsuta, nervis lateralibus utrinque 12-15, sub margine prorsus curvatis venis trausversis obscuris; petioli ad 7 lin. longi, adpresse hirsuti; stipule late ovate, acuminate, 3 lin. long, utrinque minus magisve adpresse hirsute. Cyme multiflora, axillares, laxe, 1-1} poll. diametro, pedunculo 4-6 lin. longo suffulte, undique albo-hirsute ; bracteæ oblonge, ad 11 lin. long: ; pedicelli brevissimi vel ad 3 lin. longi. Recepta- culum dense albo-strigillosum, 2 liu. longum. | Calycis segmenta 106 DR. OTTO STAPF ON j liu. longa, in fruetu haud aucta, ovata, obtusiuscula. Corolla angustissime infundibuliformis, 5 lin. longa, extus sparse adpresse pilosa; lobi ovati, acuti, 1 lin. longi. Fructus globosus, 14 lin. diametro. Near Monrovia and in the Sinoe Basin, Whyte. Also in the Gold Coast near Akwapim, at 1400 feet, Murphy (herb. Johnson, 679). This might be treated as a variety, like the following species. Sabicea lasiocalyx, Stapf (sp. nov.); affinis S. ferruginec, Benth., differt ramulis pedunculisque patule hirsutis, receptaculo et calycis tubo æque adpresse albo-tomentosis et preterea uti calycis segmentis longe hirsutis, capitulo compacto. Ramuli adpresse hirsuta pilis longis patulis additis, ferruginei. Folia late oblonga vel elliptiea, subaeuminata, basi rotundata, 6-8 poll. longa, 23-4 poll. lata, supra costa hirsutiuscula excepta glabrata exsiccando nigrescentia, subtus cinnamomea adpresse lanato-tomentosa pilis tenuibus longioribus nonnullis additis, nervis lateralibus utrinque cireiter 20, marginem versus prorsus curvatis venis transversis obscuris; petioli 3-1 poll. longi, dense hirsuti; stipule foliacee late ovate, acuminate, ad i poll. longe, dorso hirsute. Capitula densa, pedunculo 11-4 poll. longo eodem indumento ae ramulis induto; bractex in- volucrantes rotundato-ovatæ, longe tenuiter acuminate, 3-1 poll. longs, dorso tomentose et simul patule hirsute. Flores arcte congesti, sessiles. Receptaculum uti calycis tubus æque adpresse albo-tomentosum preterea pilis longioribus hirsutum. Calycis tubus brevis, demum paulo elongatus; segmenta subulata ad à poll. longa, albo-tomentosa insuper pilis tenuibus patulis ad 13 lin. longis additis. Corolla anguste infundibuliformi-tubulosa, circiter 2 poll. longa, basi glabra, medio albo-pubescens, superne et in loborum dorsis dense albo-tomentosa ; lobi vix 1 lin. longi. Within 6 miles of Monrovia ; Sinoe Basin, Whyte. This might perhaps be treated as a variety of S. Jerruginea with an exceptionally copious indumentum and sessile flowers. Webera gracilis, Stapf (sp. nov.) ; affinis W. congensis, Stapf (Tarenna congensis, Hiern), sed foliis tenuioribus, longius acu- minatis, nervis lateralibus 7-8 in utroque latere, inflorescentiis puberulis strictioribus, ramis longioribus, floribus multo minori- bus, alabastris adultis 6—7 lin. longis. THE FLORA OF LIBERIA. 107 Ramuli graciles, glabri, cortice pallide fusci. Folia elliptico- oblonga vel elliptico-lanceolata, acuminata (acumine ad 4 poll. longo), basi obtusa vel breviter acuminata, 23-4 poll. longa, 14-1}? poll. lata, membranacea, glaberrima, nervis lateralibus utrinque circiter 8 valde obliquis sub margine prorsus curvatis uti venis tenuibus; petioli 1-2 lin. longi; stipulw latissime, rotundate, apiculate, 1 lin. alte. Inflorescentie corymbiformes plures ad ramulorum apices; pedunculi graciles, ad 1 poll. longi, uti rami pedicellique tenuissime pubescentes ; rami primarii circiter 4 poll. longi; bractew ovate vel lanceolate, minute, pubescentes; pedicelli 2-4 lin. longi, minute braeteolati. Ke- ceptaculum obovoideo-globosum, } liu. longum, uti calyx tenuis- sime pubescens. Calycis tubus brevis lobis late triangulari- ovatis acutis. Corolle glabre tubus subcylindricus, supra medium ampliatus, ad 5 lin. longus; lobi oblongi, obtusi, 3 lin. longi. Anthere 3 lin. longe. Stylus glaber. Fructus globosus, 1j lin. diametro. Sinoe Basin, aud without precise locality, Whyte. Oxyanthus tenuis, Stapf (sp. nov.); affinis O. pallido, Hiern, differt ramis gracilioribus, foliis tenuioribus, petiolis gracilioribus, stipulis 2-3 lin. longis, coroll: tubo tenuiore, lobis brevioribus. Planta glaberrima. Rami graciles, teretes. Folia oblongo- lanceolata, acuminata, basi acuta, 4-6 poll. longa, 14-2 poll. lata, membranacea, tenuia, nervis lateralibus utrinque circiter 6, uti venis tenuibus; petiolus 3 lin. longus, gracilis; stipulae e basi triangulari subulate, 2-3 lin. longw, membranaceg. Cyme 3-4-flore, subsessiles vel pedunculo crassiusculo ad 2 lin. longo suffulta; ; bractew lanceolate, tenuiter acuminate, ad 2 lin. long. Pedicelli brevissimi vel ad 4 lin. longi. Receptaculum ovoideo- oblongum, 1 lin. longum. Calycis tubus brevis; segmenta subulata, 1 lin. longa. Corolle tubus tenuis, cylindricus, ultra 5 poll. longus; lobi lineares, angusti, 9 lin. longi. Anthere vix 2 lin. longa. Sinoe Basin, Whyte. Also in Sierra Leone, Limba Country, Madina, Scott Elliot, 5571. Ixora congesta, Stapf (sp. nov.); allinis Z. daviflore, Smith, differt foliis majoribus, floribus sessilibus ad apices ramulorum brevium corymbi trichotomi aggregatis. Planta glaberrima preter iuflorescentiam interdum sparse papilloso-puberulam. KRamuli teretes, fuscescentes. Folia lan- 108 DR. OTTO STAPF ON ceolato-oblonga vel elliptiea, in eodem ramulo valde varia, acute acuminata acumine ad 1 poll. longo, basi longe cuneato-attenuata vel summa rotundata vel subcordata, ad 1 ped. longa, 12-4 poll. lata, coriacea, nervis lateralibus utrinque 15-18 subpatulis sub margine prorsus arcuatis uti venarum retieulatione laxa utrinque prominulis ; petioli ad 1 poll. longi vel summi brevissimi, crassi ; stipule basi in tubum brevem connate, late ovate, subulato- acuminate, ad 4 poll. longs. Corymbi multiflori, densissimi, pedunculo 3-5 poll. longo suffulti; rami primarii 5-9 lin. longi, interdum uti ramuli obscure papilloso-puberuli ; bractew bracteo- læque lanceolato-subulate, infime ad 3 lin. longe. Flores 2 vel plures ad apices ramulorum brevium braeteatorum sessiles. Jieceptaculum globosum, i lin. altum. Calycis segmenta late ovata, acuta, 3 lin. longa. Corolle tubus tenuis, cylindricus, 6-7 lin. longus; lobi lineares, acuti, 4-5 lin. longi, basi papillosi. Anthere ad 3 liu. longe. Stylus 4 lin. e corolla tubo exsertus, tandem ad 14 lin. divisus. Siuoe Basin, Whyte. Ixora atrata, Stapf (sp. uov.); affiuis Z. Soyauxii, Hiern, differt ramulis strictis divaricatis, foliis glaberrimis, nervis venisque infra pulchre prominulis, alabastris acuminatis, coroll segmentis longioribus. Planta preter inflorescentias petiolosque interdum parce pilosulos glaberrima, exsiecando nigricans. Ramuli graciles, teretes. Folia elliptico-lanceolata vel oblongo-lanceolata, acu- minata, basi acuta, 34-44 poll. longa, 14-2 poll. lata, mem- branacea, nervis lateralibus utrinque 4-5 valde obliquis sub margine arcuato conneetis uti venis transversis subtus pro- minentibus; petioli 11-3 lin. longi, graciles; stipule e basi latissima brevi abrupte tenuissime subulatæ ad 2) liu. longs. Cyme laxi, teichotomie, pauciflore ad ramulorum apices; pedun- culi graciles, 4-9 lin. longi, uti rami et pedicelli glabri vel parce rufo-pilosuli; rami 2-4 lin. longi; bractew lanceolate vel subulate, parvw; pedicelli ad 24 lin. longi. Receptaculum vix + lin. longum. Calyx vix 4 lin. longus, segmentis ovatis obtusis vel acutis. Corolle tubus 23-3 lin. longus, cylindricus, superne paulo dilatatus ; lobi oblongi, 4 lin. longi. Within 6 miles of Monrovia, Whyte. Coffea nudiflora, Stapf (sp. nov.); affinis C. melanocarpe, Hiern, sed floribus folia precedentibus, corolla ore glabra, fructibus rubris distincta; a C. divaricata, K. Schum., et THE FLORA OF LIBERIA. 109 C. rupestri, Hiern, quibus habitu simillima, imprimis calyculo minuto differt. Planta glaberrima. Rami graciles, cortice cinerascente vel albieante. Folia decidua oblonga vel ovata vel obovato-oblonga, acuminata, basi acuta, 31-5 poll. longa, 14-12 poll lata, tenuiter membranaeea, nervis lateralibus utrinque 5 obliquis arcuato- connectis ut venis tenuissimis; petioli 13-2 lin. longi; stipule e basi lata subito subulato-contracte vel apiculate, ad 1 lin. longs, deinde indurate albieantes. Flores in ramis annotinis ante folia hornotina evoluti, solitarii vel bini in ramulis brevissimis sessiles; bractew infimee (2 paria) membranacez in annulum brevem subbilobum connate, summe (2 paria) calyculum formantes magis minusve herbacez, ovate, acute, plerumque parve, rarius in folia parvula ecrescentes. Calyx truncatus annuliformis. Corolle glabre tubus cylindricus, 1 poll. longus ; lobi elliptico- oblongi, 24 lin. longi; lobi 6, oblique truncato-elliptici, 33-4 lin. longi. Anthere inclusw, lj lin. longe. Stigma inclusum. Fructus ruber, globosus, 2-24 lin. diametro. Within 6 miles of Monrovia and in the Sinoe Basin, Whyte. Coffea ligustrifolia, Stapf (sp. nov.); aflinis C. scandenti, K. Schum., sed ramis rigidis angulo recto divaricatis, foliis angustioribus, valyculi bracteis summis foliaceis, calycis margine crenato corollae tubo multo longiore. Rami uti tota planta glaberrimi, cortice castaneo tecti, angulo recto divaricati. Folia sempervirentia lanceolata, obtusiuscule acuminata, basi acuta, 24-8 poll. longa, 3 ad fere 1 poll. lata, subcoriacea, nervis lateralibus utrinque 4-5 sub margine eleganter arcuato-connectis uti venarum retieulatione utrinque prominulis ; petioli 1 lin. longi; stipule brevissimz, lata, apiculate. Flores solitarii vel bini, axillares sessiles; braetez irfimæ (2 paria) in annulum obscure bilobum connate, summe (2 paria) calyeulum formantes foliacez, elliptiez vel ovato-elliptieze, acute, ad 3 lin. longe. Calyx brevissimus crenulatus. Corolle glabre tubus tenuiter infundibuliformis, ad 7 lin. longus, ore fere 2 lin. diametro ; lobi 5 lineari-oblongi, 54 lin. longi. Anthere exsertæ 21 ad fere 3 lin. longe. Stigma coroll; os 3—4 lin. excedens, Sinoe Basin, Whyte. Tylophora liberica, W. E. Brown (sp. nov.); T. conspicue, N. E. Br, simillima, sed glabra et lobis basalibus foliorum brevioribus. 110 DR. OTTO STAPF ON Caulis scandens, glaber. Folia glabra; petiolus 3-1 poll. longus; lamina 33-44 poll. longa, 2-24 poll.lata, oblonga vel elliptico-oblonga, acuta vel acuminata, basi cordata vel emarginata, lobis vel auriculis ad 1 poll. longis. Pedunculi ad flexuras racemum brevem sessilem gerentes, glabri. Pedicelli 4-5 lin. longi, glabri. Sepala 1 lin. longa, ovata, glabra. Corolle lobi 2 lin. longi, 1 lin. lati, oblique elliptico-oblongi, apice rotundati. Coronc tuberculi subglobosi, fusci. Columna staminum basi non dilatata. Sinoe Basin, Whyte. Lankesteria brevior, C. B. Clarke (sp. nov.); L. eleganti, T. Anders., affinis, sed corolle tubo multo breviore distincta.— L. elegans, C. B. Clarke, in Dyer, Fl. Trop. Afr. v. p. 70, pro parte. Rami imo in parte superiore lignescentes. Folia anguste obovata (in L. eleganti elliptica, apice magis acuminata). Strobili bractee iis L. elegantis similes, sed plus minus pubescentes, in margine ciliato-villose. Calyæ į poll. longus, aut parum longior. Corolla alba, in centro lutescens (teste W. H. Johnson); tubus 3 poll. longus vel brevior (in Z. eleganti 1 poll. longus vel ultra), limbi lobi quam Z. elegantis minores. Anthere e coroll tubo brevissime exserte. Capsula 2 lin. longa stipite cylindrico, apice rotundata, complanata, copiose hygroscopice hirsuta. Near Monrovia, Whyte. Also in the following localities :— Gold Coast, Akim, W. H. Johnston, 257, 261, and without precise locality, Burton and Cameron. Ashanti Country, near Kumassi, Cummins, 50, 199. Cameroons, in primeval forest, Staudt, 538 (issued as L. Barteri, Hook. f.). AFRODAPHNE, Stapf (gen. nov. Lauracearum). Affinis inter | Lauraceas gerontogeas Beilschmiedia, Nees, differt receptaculo cupulari vel turbinato distineto, filamentis pro ratione brevioribus vel subnullis, ovario in receptaculo sub- immerso, panieulis laxioribus, sepe amplis; inter neogeas accedit ad Hufelandiam, Nees, et Aioueam, Aubl., sed a priore receptaculo, ab altera perianthio magis herbaceo, receptaculo haud carnoso-incrassato nee persistente recedit. Flores hermaphroditi. | Perianthium herbaceum, post authesin totum deciduum ; receptaculum eupulare vel turbinatum ; segmenta 6, wqualia, parva. Stamina ordinis primi et secundi THE FLORA OF LIBERIA. 111 qualia, fertilia, basi eglandulosa, magis minusve papilloso- pilosula ; filamenta lata, brevia vel subnulla; anthere late ovate vel subquadrate, introrsum 2-locellate; stamina tertii ordinis vel fertilia filamento distincto subeylindrieo pilosulo, antheris extrorsum 2-locellatis, vel at filamenta conica brevia dense papillosa circa ovarium in conum conniventia reducta, semper basi utrinque glandula carnosa crassa instructa; stamina quarti ordinis semper sterilia, vel ad fila pilosula vel ad glandulam cordatam carnosam reducta. Ovarium sessile in receptaeuli basin subimmersum, anguste ovoideum, sensim in stylum longi- useulum gracilem attenuata. Drupa pedicello haud inerassato insidens, basi nuda, oblonga. Semen cotyledonibus magnis plano-convexis coreulum parvum ineludentibus. Arbores vel frutices, gemmis nudis sericeo-velutinis. Folia alterna, rarius nonnulla subopposita, coriacea glabra, pennivenia. Flores parvi vel minimi in panieulas multifloras laxiuseulas axillares parce vel vix bracteatas dispositi, pedicellis brevibus vel longiuseulis. Species circiter 15, omnes Africæ occidentalis. Although very homogeneous in general appearance, the genus consists of two clearly distinct sets which may be treated as sections, the difference being chiefly in the presence of 9 or of 6 fertile stamens. Where the third series of stamens is reduced to staminodes, the latter connive more or less into a cone surrounding the ovary. I propose for these sections the names Ennearrhena and Hexarrhena respectively, the names explaining themselves. To the section Ennearrhena I refer the following species :— l. A. ErATA, Stapf (= Beilschmiedia elata, Scott Elliot). 2. A. FRUTICOSA, Stapf (= Beilschmiedia fruticosa, Eagl.). 3. A. GRANDIFOLIA, Stapf (= Cryptocarya ? grandifolia, Zngl.). 4. A. Manni, Stapf (= Beilschmiedia Mannii, Hook. f.= Oreo- daphne Mannii, Meissn.). 5. A. MINUTIFLORA, Stapf ( = Beilschmiedia minutiflora, Hook. f. = Oreodaphne minutiflora, Meissn.). 6. A. NITIDA, Stapf (= Beilschmiedia nitida, Engl.). 7. A. Preussin, Stapf (= Beilschmiedia Preussii, Hng/.). 8. A. SESSILTFOLIA, Stapf (= Beilschmiedia sessilifolia, Engl.). 9. A. Srauprir, Stapf (= Beilschmiedia Staudtii, Engl.) 10. A. ZENKERI, Stapf (= Beilschmiedia Zenkeri, /ng/.), 112 DR. OTTO STAPF ON To the section Hexarrhena belong :— 11. AFRODAPHNE CALABARTCA, Stapf (Mann, 2255, in part). 12. A. CAUDATA, Stapf (sce below). 13. A. EURYNEURA, Stapf (see below). 14. A. aABOONENSIS, Stapf(=Beilschmiediagaboonensis, Hook.f. =Oreodaphne gaboonensis, Meissn.). 15. A. OBSCURA, Stapf ( — Beilsehmiedia obscura, Engl.). Afrodaphne caudata, S/apf (sp. nov.); inter species sectionis Hexarrhene foribus majuseulis, staminibus quarti ordinis ad filamenta pilosa eglandulosa distincta. Folia oblonga, caudato-acuminata acumine acuto ultrapollicari, basi obtusa, 9 poll. (vel ultra) longa, ad 34 poll. lata, coriacea, glaberrima, lucidula, nervis lateralibus utrinque circiter 8 areuato-connectis, nervo collectivo submarginali altero addito, utrinque pulehre laxe retieulata; petioli erassi, 3 lin. longi. Panicule laxe, ad 8 poll. longs, superne fulvo-puberule ; pedunculi ad 3 poll. longi; braetes pauce, decidusm, ovate, concave, ad 2 lin. longæ, dorso fulvo-tomentelle ; pedicelli 1-2 lin. longi. Perianthium turbinatum, 13 lin. longum, extus fulvo-tomentellum ; receptaculum j lin. altum, basi haud con- strictum : lobi ovati, subacuti. Stamina primi et secundi ordinis filamentis brevissimis ad loborum bases insertis, antheris apicu- latis apice papillosis; tertii ordinis ad staminodia anguste conica dense papillosa reducta, glandulis majusculis; quarti ordinis ad filamenta pilosula basi eglandulosa reducta. Pistillum e flore haud exsertum. Sinoe Basin, Whyte. A. euryneura, Stapf (sp. nov.), affinis A. gaboonensi, Stapf, differt foliia majoribus basi rotundatis nervis lateralibus dis- tantibus, panieulis majoribus. Folia oblonga, breviter tenuiterque acuminata vel fere aristu- lata, basi rotundata, ad 8 poll. vel ultra longa, ad 34 poll. lata, coriacea, glabra, nervis lateralibus utrinque circiter 8 arcuato- connectis subtus valde prominentibus, nervo collectivo sub- marginali altero addito (haud semper conspicuo), venarum re- tieulatione laxa utrinque prominula; petioli ad 6 lin. longi, crassi. JPanieule 4-8 poll. longs, laxæ, superne fulvo-cinereo- puberule; pedunculi 1-8 poll. longi; bractew paucæ, ovate, acute, concave, ad 2 lin. longe, dorso fulvo- vel cinereo-tomen- THE FLORA OF LIBERIA. 118 telle ; pedicelli graciles, 1 lin. longi. Perianthium breviter turbinato-globosum, vix 1 lin. longum, extus fulvo-cinereo- tomentellum ; receptaculum 4-i perianthii altitudine ; lobi late ovati, subacuti. Stamina primi et secundi ordinis filamentis brevissimis ad loborum bases inserti antheris exapieulatis apice papillosis; tertii ordinis ad staminodia eoniea dense papillosa reductis glandulis crassissimis ; quarti ordinis ad glandulam cordatam carnosam apiee et seeundum lineam mediam utrinque papillosam reducta. Pistillum stylo apice jam ante flores apertos exserto. Sinoe Basin, Whyte. Cleistanthus liberica, JV. E. Brown (sp. nov.) ; affinis C. ango- lensi, Muell. Arg., cortice cinereo, cuspide foliorum longiore angustioreque distincta. Cortex cinereus, nec brunneus. Folia alterna, breviter petio- lata, 21-32 poll. longa, 11-1} poll. lata, oblonga vel elliptico- oblonga, cuspidato-acuminata, basi obtusa vel late cuneata, glabra cuspide 7-9 lin. longa basi 12-2 lin. lata lineari obtusa, Racemi }-1 poll. longi, ferrugiueo-puberuli. Flores d fascieu- lati. Pedicelli 11-2 lin. longi. Sepala 1} lin. longa, lineari- oblonga, acuta. Petala minuta, vix 3 lin. longa, linearia, apice dentata, glabra. Discus crassus, integer, apice dense pubescens. Flores 9 non vidi. Sinoe Basin, Whyte. Phyllanthus profusus, W. E. Brown (sp. nov.); P. floribundo, Muell. Arg., habitu simillimus, differt foliis ovatis acuminatis, floribus numerosioribus albidis vel pallide viridescentibus. Rami foliiferi 34-5 poll. longi, glabri. Folia breviter petio- lata, 2-3 poll. longa, #-13 poll. lata, ovata vel elliptico-ovata, acute acuminata, glabra, subtus pallida. Flores d numero- sissimi fasciculati, albi vel pallide viridescentes, glabri ; fasciculi in racemos interruptos fasciculatos 23-5 poll. longos dispositi. Pedicelli 1-2 lin. longi. Sepala 4, 3-3 lin. longa, elliptico- oblonga, obtusa. Glandule 4, minutissime. Stamina 4, libera. Flores 9 non vidi. Sinoe Basin, Whyte. Croton dispar, N. E. Brown (sp. nov.): C. macrostachyo, A. Rich., habitu similis sed gracilior et glabrior. Rami floriferi 1 lin. crassi, minute et sparse ferrugineo- 114 DR. OTTO STAPF ON lepidoti, demum glabri. Folia elliptico-ovata vel oblonga, obtuse subcuspidato-acuminata, basi leviter emarginata vel latissime rotundata, triplinervia, utrinque fere glabra, squamulis minutis stellatis sparsissime conspersa; petiolus 3-1 poll. longus; lamina 23-8 poll. longa, 11-2 poll. lata. Racemus terminalis, solitarius, 5-6 poll. longus. Flores masculi 5-meri ; femineos non vidi. Pedicelli 2 lin. longi, minutissime stellato- puberuli. Sepala et petala 5, revoluto-reflexa, subequalia, fere 1 lin. longa; sepala subacuta, glabra, apice ciliata; petala obtusa, dorso glabra, intra lanata. | Glandule 5, subulate. Stamina 10-11; filamenta 13 lin. longa, glabra. Receptaculum pilosum. | Ovarium rudimentum nullum. Near Monrovia, Whyte. Crotonogyne caterviflora, JV. E. Brown (sp. nov.); proxima C. Manniane, Muell. Arg., sed foliis basi rotundatis nee longe cuneato-aeutis, et racemis multo brevioribus distincta. Folia oblongo-oblanceolata, breviter acuminata, basi obtuse rotundata, utrinque sparsissime lepidota vel supra fere glabra ; petiolus 5-8 lin. longus; lamina 4-7 poll. longa, 14-23 poll. lata. Racemi masculi 2-6 poll. longi, dissite glomeruliflori, lepidoti ; femineos non vidi. Pedicelli 3-1 lin. longi, dense lepidoti. Sepala 3, subinzqualia, 1-1} lin. longa, elliptica, densissime lepidota. Petala 5, elliptica, obtusa, 11 lin. longa, glabra. Stamina 15; filamenta 4-1 lin. longa, basi dilatata, connata. Sinoe Basin, Whyte. Erythrococca aculeata, Benth., var. acutissima, N. E. Brown (var. nov.) ; ab typo differt foliis acutissime (nee obtuse) acu- minatis. Around Monrovia, Whyte. Hemanthus longitubus, C. H. Wright (sp. nov.); affinis H. multifloro, Martyn, perianthii tubo multo longiore differt. Folia elliptica, breviter abrupteque acuminata, tenuiter mem- branacea, 6 poll. longa, 23 poll. lata; petioli 1l poll. longi; vagine 3 poll. longz, maculate. Pedunculus lateralis, 8 poll. longus, multiflorus ; spathz e basi ovata longe acuminate, rubre ; pedicelli tenues, 6 lin. longi. Perianthium rubrum; tubus cylindrieus, tenuis, 14 lin. longus; segmenta linearia, tubo THE FLORA OF LIBERIA. 115 equilonga, 2— lin. lata, 1-nervia. Stamina perianthii segmentis equilonga. Ovarium trilobum ; ovula solitaria. Sinoe Basin, Whyte. This species, at first sight, much resembles the widely-spread H. multiflorus, Martyn, which is distinguished by its perianth not exceeding half an inch in length. Dracena prolata, C. H. Wright (sp. nov.); a D. bicolore, Hook., foliis basi non longe attenuatis, vaginis parvis, petiolisque brevioribus crassioribus differt. Caules leves, siccando straminei, nitidi. Folia elliptica, basi apiceque breviter acuminata, obtusa, coriacea, circa 7 poll. longa, 3 poll.lata; petioli superiores ad 1 poll. longi. Panieula brevis, congesta; bracteæ late triangulares, acute; pedicelli 2 lin. longi. — Perianthii tubus cylindricus, 3 lin. longus, 3 lin. diam. ; segmenta oblonga, apice cucullata, mucronata, 4 lin. longa, 1 lin. lata. Anthere oblongz, 1l lin. longg. Ovarium ovoideum ; stylus antheras paullo superans. Sinoe Basin, Whyte; within 6 miles of Monrovia, Whyte. Culcasia liberica, V. E. Brown (sp. nov.) ; affinis C. scandenti, Beauv., differt caule crassiore, foliis majoribus, floribus numerosi- oribus. Caulis X poll. crassus. Folia lanceolata vel elliptico-lanceo- lata, acuta vel subeuspidata, basi cuneata; petiolus 44-5 poll. longus ; lamina 8-9 poll. longa, 3-32 poll. lata. ^ Pedunculi 7-8 in fasciculos terminales dispositi, 13-22 poll. longi, 1-14 lin. crassi. Spatha li poll. longa; tubus ovoideus ; lamina eymbi- formis, subaeuta. Spadix e spatha exserta, 2 poll. longa, parte feminea 2-23 lin. longa, subglobosa, parte mascula a feminea longe distante, cylindrico-clavata. Ovarium subglobosum ; stigma sessile. Sinoe Basin, Whyte. P.S.—The specimens mentioned on pp. 89 and 94 as collected by H. Reynolds form part of a small collection received at Kew since this paper was read. 12th May, 1905. O. STAPF. LINN. JOURN.— BOTANY, VOL. XXXVII. K INTERNATIONAL CATALOGUE OF SCIENTIFIC LITERATURE. Tue Hovar Society has been engaged for some years past in arranging for the publication of an International Catalogue of Scientific Literature, beginning from the lst January, 1901. Each science will be represented in an annual volume containing lists arranged under authors and subjects, of all books and papers published during the year; these will be contributed through official channels of information—abroad, by direct control of the respective governments—at home, by means of the various Societies which devote themselves to particular sciences; those Societies whose domains over- lap having arranged for mutual cooperation. The collection of title-slips for the United Kingdom of Great Britain and Ireland as regards BOTANY has been undertaken by the Council of the Linnean Society, and they appeal to all botanic workers for support in their endeavour to compile a complete record, by sending notices promptly of all botanic issues to the undersigned. The third volume, for 1908, is now in the press. B. DAYDON JACKSON, General Secretary, Linn. Soc. NOTICE. Vor. XXXVI., of which Paris 249-254 are already published, is reserved for the completion of the Chinese Flora. Attention to this announcement is specially requested, to prevent application to the Librarian for unpublished Parts. The CarALoGuE of the Liprary may be had on application. Price to Fellows, 5s. ; to the Public, 10s. All communications relating to the general business of the Society should, as heretofore, be addressed to the SECRETARIES, but letters on library business only may be addressed to the * LIBRARIAN.” CIS Enty dinis os eee SEPTEMBER 30. ; Price 16s. THE JOURNAL OF THE LINNEAN SOCIETY, Vor. XXXVII. BOTANY. No. 259. CONTENTS. Page I. The Botany of the Anglo-German Uganda Boundary Commission. By E. G. Baxer, F.L.S., S. Moore, F.L.S., and A. B. Manus D.Se., F.L.S. (Plates 1-4) 116 II. The Axillary Seales of Aquatic Monocotyledons. By R. J. Harvey Gissow, M.A., F.L.S., Professor of Botany in the University of Liverpool. (Plates 5 & 6) 228 III. The Botany of Gough Island.—l. Phanerogams and Ferns. By R. N. Rupmose Brown, B.Sc., Botanist of the Scottish National Antarctic Expedition. (Com- municated by W. Borrine Hemsuey, F.R.S., F.L.S8.) (Plates 7-9)... eR emn 238 IV. Mansoniez, a new Tribe of the Natural Order Stereu- liaceæ. By Lt.-Col. D. Pratn, I. M.5., F.R.S., F.L.S. CPlate 10)... raai thor ht He EERASRAR AMOUR HN 250 V. The Botany of Gough Island.—II. Cryptogams (ex- cluding Ferns and Unicellular Alge) By R. N. RupwosE Browy, B.Sc., C. H. Wriaut, A. L.S., and O. V. DARBISHIRE. (Communicated by W. Borrina CETOXLR Dig qe 00-00 49 99 — Y CERO DU canes 263 VI. Notes on the Genus Widdringtonia. By MaxwerL T. Masters, M.D., F.R.S., F.L.S., Correspondent of the Institute of France ....................... serere. 267 See Notice on last page of Wrapper. LONDON: SOLD AT THE SOCIETY'S APARTMENTS, BURLINGTON HOUSE, PICCADILLY, W.; AND BY LONGMANS, GREEN, AND CO., AND WILLIAMS AND NORGATE. 1905. LINNEAN SOCIETY OF LONDON. LIST OF THE OFFICERS AND COUNCIL, Elected 24th May, 1905. PRESIDENT. Prof. W. A. Herdman, D.Sc., F.R.S. VICE-PRESIDENTS. Charles Baron Clarke, M.A., F.R.S. Frank Crisp, LL.B., B.A. Horace W, Monckton, F.G.8. Dr. A. Smith Woodward, F.R.S. TREASURER. Horace W. Monckton, F.G.S. SECRETARIES. D. H. Scott, M.A., Ph.D., F.R 8. | Rev. T. R. R. Stebbing, M.A., F.R.S. CENERAL SECRETARY. B. Daydon Jackson, Esq. COUNCIL. R. Assheton, M.A, V. H. Blackman, M.A. Gilbert C. Bourne, D.Sc. Dr. Horace T. Brown, F.R.S. Charles Baron Clarke, M.A., F.R.S. Frank Crisp, LL.B., B.A, Prof. J. B. Farmer, D.Sc., F.R.S. Rev. Canon Fowler, M.A. Prof. W. A. Herdman, F.R.S. B. Daydon Jackson, Esq. Horace W. Monckton, F.G.S. Prof. F. W. Oliver, D.Sc., F.R.S. Clement Reid, F.R.S. Dr. A. B. Rendle, M.A. Dr. W. G. Ridewood, F.Z.8. Dukinfield H. Scott, M.A., Ph.D, F.R.S. David Sharp, F.R.S. Dr. Otto Stapf. Rev. T. R. R. Stebbing, M.A., F.R.S. A. Smith Woodward, F.R.S. LIBRARIAN. A. W. Kappel, CLERK. A. R. Hammond. LIBRARY COMMITTEE. This consists of nine Fellows (three of whom retire annuall y) and of the officers ex officio; the former are elected annually by the Council in June, and serve til the succeeding required during the Session. the officers, are :— Herbert Druce, F.Z.8. Antony Gepp, M.A, W. Botting Hemsley, F.R.S. Dr. G. Henderson. P. Chalmers Mitchell, M.A., F.Z.S. Anniversary. The Members for The Committee meets as 1904-1905, in addition to A. B. Rendle, M.A., D.Sc. A. Q. Tansley, M.A. Prof. H. Marshall Ward, F.R.S. Dr. A. Smith Woodward, F.R.S, THE FLORA OF LIBERIA. 115 æquilonga, 3—} lin. lata, 1-nervia. Stamina perianthii segmentis equilonga. Ovarium trilobum; ovula solitaria. Sinoe Basin, Whyte. This species, at first sight, much resembles the widely-spread H. multiflorus, Martyn, which is distinguished by its perianth not exceeding half an inch in length. Dracena prolata, C. H. Wright (sp. nov.); a D. bicolore, Hook., foliis basi non longe attenuatis, vaginis parvis, petiolisque brevioribus crassioribus differt. Caules leves, siccando straminel, nitidi. Folia elliptica, basi apiceque breviter acuminata, obtusa, coriacea, circa 7 poll. longa, 3 poll. lata; petioli superiores ad 1 poll. longi. Panicula brevis, congesta; bractee late triangulares, acute; pedicelli 2 lin. longi. Perianthii tubus cylindricus, 3 lin. longus, 4 lin. diam. ; segmenta oblonga, apice cucullata, mucronata, 4 lin. longa, 1 lin. lata. Anthere oblonge, 14 lin. longs. Ovarium ovoideum ; stylus antheras paullo superans. Sinoe Basin, Whyte; within 6 miles of Monrovia, Whyte. Culcasia liberica, N. E. Brown (sp. nov.) ; affinis C. scandenti, Beauv., differt caule crassiore, foliis majoribus, floribus numerosi- oribus. Caulis + poll. crassus. Folia lanceolata vel elliptico-lanceo- lata, acuta vel subcuspidata, basi cuneata; petiolus 44-5 poll. longus; lamina 8-9 poll. longa, 8-32 poll. lata. Pedunculi 7-8 in fasciculos terminales dispositi, 12-22 poll. longi, 1-1} lin. crassi. Spatha 1i poll. longa; tubus ovoideus ; lamina cymbi- formis, subacuta. Spadix e spatha exserta, 2 poll. longa, parte feminea 2-24 lin. longa, subglobosa, parte mascula a feminea longe distante, cylindrico-clavata. Ovarium subglobosum; stigma sessile. Sinoe Basin, Whyte. [ P.S.—The specimens mentioned on pp. 89 and 94 as collected by H. Reynolds form part of a small collection received at Kew ince this paper was read. 19th May, 1905. O. SraPr.] LINN. JOURN.—BOTANY, VOL. XXXVII. K 116 THE BOTANY OF THE ANGLO-GERMAN The Botany of the Anglo-German Uganda ndary Commission. By E. Œ. Barek, F.L.S.; S. Moors, F.L.3.; and A. B. RENDLE, M.A., D.Sc., F.L.S. (Prates 1-4.) [Read 16th February, 1905.] THE collections which form the subject of this paper were made by Dr. A. G. Bagshawe, medieal officer to the recent Anglo- German Uganda Boundary Commission under H.M. Com- missioner, Lieut.-Col. Delmé-Radcliffe. Beginning at the mouth of the Kagera River, on the south-western boundary of the Uganda Protectorate, at a point where the river empties itself into the Victoria Nyanza on the west side, a camp was made for three months. The second collecting-centre, about 60 miles from the lake, was Mulema in South Ankole, lat. 1° S., long. 31? E., where there was also a camp for about three months till the middle of 1903. Barumba, where a few plants were collected, is fifteen miles further west. The next colleeting-centre was the district of the high hills of Ruehigga (lat. 1°-1° 10' S., long. 30°- 30° 15' E.) from 5500-7800 ft., which drains into the Congo. With this exception, all the plants in the collection are from lands which drain into the Nile. Two visits were paid to an isolated hill, Irunga, alt. 7160 ft., which lies at the intersection of the English, German, and Congo boundaries. The River Rufüa, whieh drains Lake Karenge, and is in long. 30^ 6' E., and lat. 0° 55' S., and runs S.E. to join the Kagera River, was also visited. The next important collecting-centre was the island of Buvüma, opposite the exit of the Nile from the Victoria Nyanza, where a stay of about three weeks was made. Buvuma is about the size of the county of Rutland. It is hilly, the highest point being 600 feet above the level of the lake ; about half the island is thiekly wooded. The inhabitants differ considerably from those of the mainland—their language is different and allied to that of the Basoga. Sir Harry Johnston, who recently visited the island, suggests that it may have been separated from the mainland for a sufficient period to acquire or retain peculiar forms of vegetation. The earliest collection made in the Uganda Protectorate was UGANDA BOUNDARY COMMISSION. 117 that by Col. Grant, on the Speke and Grant Expedition of 1860 * partly on the north-western shore of the Victoria Nyanza, between what was then called the river Kitangule (now called Kagera) and M'tesas. Other collectors who have visited this region are Mr. G. F. Seott Elliot on his return from his expe- dition to Mt. Ruwenzori, and Dr. Stuhlmann on the Emin Pasha Expedition. Lieut. Stairs, on the Stanley Expedition in 1890, and several others have also collected plants in Uganda. In Sir Harry Johnston’s book on Uganda t, Mr. C. H. Wright, A.L.S., has given a list of the plants known from the Protectorate. Sir Harry Johnston recognises five Botanical Regions in the Uganda Protectorate :— (1) Somarı Reaion.—Includes the arid country in the basin of Lake Rudolf and up the Rift Valley as far as the north end of Lake Baringo. (2) East ArnicAN Reaion.—A land of grass, Borassus, Hy- plene, and wild Date Palms, &c. characteristic of the low- lying parts of German, Portuguese, and British East Africa. (3) CENTRAL AFRICAN Raarow.—This fertile region presupposes an average altitude of 3500 ft. in the equatorial regions of Uganda. (4) West ArnrcaN Forest Recron. — Characteristic of the countries near the shore of the Victoria Nyanza. (5) Puareav or ArPrINE RkGroN.— Everywhere between 6500 and 10,000 ft., with a flora which alternately recalls the trees and plants of temperate South Africa and temperate Abyssinia. The island of Buvüma comes under the West African Forest Region, the higher portion of the Ruchigga district and the hill Irunga reach the Alpine Region; but much of the country covered by the Expedition is in the Central African Region. As a whole, the entire country explored lies in the northern part of the Central Lake Region, Engler's “ Seengebiet,” repre- senting that portion of it which is included in the Nile Land District of Oliver’s ‘ Flora of Tropical Africa.’ The collection comprises 480 species of Seed-plants, of which 67 are new to science, and includes 433 Dicotyledons, 46 Mono- * Botany, by Prof. D. Oliver, in Trans. Linn. Soc. vol. xxix. (1872-t). t Sir Harry Johnston; The Uganda Protectorate, 1902. x2 118 THE BOTANY OF THE ANGLO-GERMAN cotyledons, and one Gymnosperm (Podocarpus milanjiana). Of the Dicotyledons 199 belong to the Polypetale, 183 to the Gamopetale, and 51 to the Apetalz, as recognised in Bentham & Hooker’s ‘Genera Plantarum’ The Monocotyledons include 30 Orchids, and also members of other petaloid orders; no sedges nor grasses were collected. Taken as a whole, the plants show considera! le affinity with the Floras of the West African Coast-region, especially Angola ; about 20 per cent. comprising species hitherto known only from those regions, or new species with a strong West African affinity. Recent investigation has revealed the same state of things as far eastward as Usambara*. Noteworthy examples of this among the Polypetale are a new variety of Capparis Afzelit, a species hitherto known only from Sierra Leone and the Cameroons; a form of Polygala Gomesiana, an Angolan plant; Hugonia platysepala (Upper and Lower Guinea); and Zanthorylon nitens, Clausena anisata var. multijuga, and Illigera pentaphylla, all three Angolan. Among the Gamopetale the following eight species are worthy of mention as having hitherto been known only from Upper Guinea :—Craterispermum brachynematum, Clitandra cymulosa, Alafia grandis, A. Schumannii, A. landolph- ioides, Baissea tenuiloba, Acanthopale decempedalis, Barleria opaca. To the same category belong Peddiea lungiflora (Togo- land), Claoxylon africanum and Habenaria Soyauaxii (Gaboon), a new Polystachya (P. inconspicua) near the Fernando Po species P. alpina, and a new species of Haemanthus (H. Radcliffer) nearly allied to H. Lindeni from the Congo. The Angolan affinity is exemplified by several orchids—Hulophia dichroma, Habenaria foliosa, Disa ochrostachya, and a new species of Mystacidium (M. ugandense). Excluding widely distributed tropical species, the majority of the plants represent an Eastern tropical African element, including a number of Abyssinian types, a few species hitherto known only from Mt. Kilimanjaro on the East (such as Tragia Volkensii), and several from Mt. Ruwenzori on the West, such as the Composites Crassocephalum ruwenzoriense and Senecio ruwenzoriensis, and Liparis ruwenzoriensis, Pteroglossaspis ruwenzoriensis, Poly- stachya nigrescens, and Disa erubescens, four orchids previously collected by Mr. Scott Elliot. A more southern element is * Engler, in Notizbl. k. Bot. Gart. Berlin, iii. 83. UGANDA BOUNDARY COMMISSION. ° 119 represented by a few Nyassa-land types, such as Eulophia missionis, Lissochilus Nyase, and Senecio sabulicolus. There is a slight South-African affinity, the most marked examples of which are a new species of Pappea allied to P. capensis, Pavetta assimilis, and Ohetacanthus Persoonii ; the last two have not previously been found north of the Tropics. ENUMERATION OF SPECIES. DICOTYLEDONES POLYPETALA. (By Epmunp G. BAKER.) RaNUNCULACER, Crematis SrUHLMANNI, Hieron. in Engler, Pflanzenwelt Ost- Afr. C. 180. lrunga. Fl. November, 381. Distrib. Karagwe in the Central African Lake-Region. C. ORIENTALIS, Linn. Sp. Pl. 543. Fl. April, 202. Native name * Munkamba." “ Flowers pinkish.” Var. y. Tuuneerat, O. Kuntze, in Verh. Bot. Brand. xxvi. (1885) 124. Ruchigga. Fl. November, 439. “ Trailing with white flowers." Var. e. WIGHTIANA, O. Kuntze, l. c. Hillside, Simba. Fl. February, 172. Hillside, Burumba. Fl. July, 374. Native name * Lumama." * Climber with white flowers.” Distrib. C. orientalis is widely spread in Temperate Asia, Northern India, &c. THALICTRUM üHyNcHocanPUM, Dillon § A. Rich. in Ann. Se. Nat. sér. II. xiv. (1840) 262. Ruchigga. Fr. November, 456. “ Erect plants 5 ft. high, with inconspicuous flowers.” Distrib. Mountains of Tropical Africa ; also occurs south of the Tropic, in Natal, &e. RANUNCULUS MEMBRANACEUS, Fres. in Mus. Senckenb. ii. (1837) 270; R. pinnatus, Oliver, in Trans. Linn. Soc. xxix. (1872) tab. 2, non Poir. River Rufüa, at edge of pool. "December, 534. Distrib. Nile Land, Lower Guinea, Mozambique District ; also in South Africa. 120 BOTANY OF THE UGANDA BOUNDARY COMMISSION. DILLENIACES. TETRACERA ALNIFOLIA, Willd. Sp. Pl. ii. 1243. Island of Buváma, Victoria Nyanza. March, 602. Distrib. Upper Guinea. The Welwitsch plant from Golungo Alto, no. 1203, formerly referred to T. alnifolia, Willd., is regarded by Dr. Gilg as a distinct species, T. podotricha, Gilg. ANONACES. Uvanra Wetwitsouit, Engler § Diels, Monogr. 18.—Oxymitra ? Welwitschii, Hiern, in Welw. Cat. 10. In gully above Mulema, Fl. April, 244; also Fr. April, 245. Native name “ Mukamira.” “Shrub with greenish-yellow flowers and yellow fruits. Fruit eaten. Wood used in hut-building.” The following is a description of the flowering specimen which, although the type comes from a very different locality, I have referred to this species. It has previously been collected only in fruit :— Frutex ramis demum glabris atropurpureo-corticatis ; foli- orum petiolo brevi inerassato +2 mm. longo; lamina subcoriacea, supra glabra, subtus preter costam glabra, oblonga, apice obtusa, basi late cuneata vel rotundata, 9-12 em. longa, 3°5-4°5 em. lata; costa subtus prominente, nervis lateralibus utrinque 10-14 adscendentibus subtus subprominentibus; floribus solitariis sepissime terminali- bus inter medioeres generis; sepalis late ovatis subacutis 11-13 mm. longis quam petala brevioribus; petalis subzqualibus ovato-suborbicularibus extus fusco-tomen- tellis, longioribus quam latis, + 14-16 mm. longis, 12-14 mm. latis; staminibus brevissimis, + 1 mm. longis, con- nectivo obliquo ; carpellis extus fusco-tomentosis. Distrib, Angola. U. svkonENsrs, Engler, in Pflanzenwelt Ost- Afr. C. 178. Island of Buvima, Fl. March, 639. * Shrub with greenish flowers." Distrib. Central African Lake-Region. UVARIA 8p. Island of Buvüáma. Fl. March, 642. * Shrub with yellow flowers." Further material is necessary of this species. E. d. BAKER: DICOTYLEDONES POLYPETAL.E. 121 Folia oblonga vel ovato-oblonga glabra, 9-11 cm. longa, 8-4 em. lata, subeoriacea. Petiolus brevissimus 2-3 min. longus. Calyx 12-13 mm. longus. Petala subæqualia quam ea U. Kirkit, Oliver, angustiora, 20-22 mm. longa, 15-16 mm. lata. ARTABOTRYS Sp. Island of Buvüma. Fr. March, 662. Climbing shrub in fruit only. Perhaps allied to A. stenopetala, Engler, but flowers required to confirm this. Leaves ovate, glaucous-green, smooth. MENISPERMACE X. TiwosPoRa TENERA, Miers, in Ann. y Mag. Nat. Hist. ser. ILI. xiii. (1864) 322. Island of Buvima. Fl. March, 658. “ Herbaceous climber with greenish flowers. @ only." Distrib. Nile Land, Mozambique District; also in Mada- gascar. FuMARIACER. FUMARIA OFFICINALIS, Linn. Sp. Pl. 700. Ruchigga., Fl. November, 396. Distrib. Europe, North Africa, North Asia. CRUCIFERE. DBnassircA WILLDENOVII, Boiss. in Ann. Sc. Nat. sér. IL. ii. (1842) 88. Near Mulema. Fl. June, 344. Distrib. Asia, temp. and trop. Nile Land, Lower Guinea, Mozambique District. CAPPARIDACEE. CLEOME MONOPHYLLA, Linn. Sp. Pl. 672. Roadside, Mulema. Fl. June, 326. Distrib. Widely spread in Tropical Africa; also at the Cape. C. mirta, Oliver, Fl. Trop. Afr. i. 81. Near River Rufüa. Fl. December, 515. l Distrib. Lower Guinea, South Central, Mozambique District. 122 BOTANY OF THE UGANDA BOUNDARY COMMISSION. MAERUA ANGOLENSIS, DC. Prod. i. 254. Near Mulema. Fl. April, 210. Native name * Mweogola." “Low tree with much con- torted branches ; flowers greenish yellow." Distrib. Upper Guinea, Nile Land, Lower Guinea. M. TRIPHYLIA, A. Rich. Tent. Fl. Abyss. i. 32, tab. 7. Near Mulema. Fl. April, 217. Native name “Moywante.” “Low shrub with white flowers.” Distrib. Nile Land. CADABA FARINOSA, Forsk. FI. ZEgypt.-Arab. 68. Near Kikobe ferry, River Kagera. Fl. March, 176. Native name “ Kakubauga." “ Wood used to make cattle- kraals. Shrub; flowers green.” Distrib. Widely distributed in Tropical Africa. Boscra sarrCiFoLIA, Oliver, Fl. Trop. Afr. i. 93. Hillside near Mulema. Fl. June, 341. * Tree-like shrub with girth of 6 ft. and greenish flowers." A form with generally 6 stamens. Distrib. Nile Land, Lower Guinea, Mozambique District. CAPPARIS spinosa, Linn. Sp. Pl. 503. Near Mulema. Fl. April, 223. Distrib. Spreading through the Mediterranean Region and Eevrt it extends eastwards to Western India. Also Nile Laud. C. TOMENTOSA, Lam. Encyc. Méth. i. 606. Hillside, Mulema. Fl. April & May, 225 & 281. Native names * Mutatsi" and ** Mukoma.” Distrib. Widely spread in Tropical Africa. C. Rora, Oliver, Fl. Trop. Afr. i. 97. Near Mulema. Fl. March & April, 187 & 237. Distrib. Nile Land. C. ERYTHROCARPOS, Zsert, in Ges. Naturf. Fr. Berl. Schrift. ix. (1789) 334, tab, 9. Island of Buvima. Fl. March, 623. Distrib. Upper Guinea, Nile Land, Lower Guinea. C. ArzELII, Pax, var. nov. BUVUMENSIS. Arbor parva glabra, foliis subcoriaceis, oblongis vel ovato-oblongis, breviter petiolatis, stipulis tenuibus, minus acuminatis quam in E. G. BAKER: DICOTYLEDONES POLYPETALA, 123 typo; floribus 1-3 ad apices ramulorum dispositis ; bacca subglobosa, lateraliter apiculata, gracillime pedun- culata. Hab. Island of Buvúma, Victoria Nyanza. Fl. March 11th, 1904. 598. * Low tree not exceeding 20 ft., with white flowers which become purplish.” Distrib. of type. Sierra Leone. BIXINES. Oncosa spinosa, Forsk. Fl. ZEyypt.- Arab. 103. Mouth of Kagera River. Fl. & Fr. February, 150. Distrib. Nile Land, Upper Guinea. PrrTOsPOREX. PrrrosPORUM ABYSSINICUM, Delile, in Ann. Sci. Nat. sér. II. xx. (1843) 49. Ruchigga. Fl. & Fr. November, 416 & 452. Distrib. Nile Land, Lower Guinea. PoOLYGALE. POLYGALA PERSICARLEFOLIA, DC. Prod. i. 326. Coast of Victoria Nyanza, Uganda. FI. February, 570. Distrib. Nile Land, Mozambique District. The following is also closely allied to this species :— Ruchigga. Fl. November, 459. P. Votkensit, Gurke, in Engler, Pflanzenwelt Ost- Afr. C. 234. Near River Rufáa. | Fl. December, 516. “ Herb with flowers that appear white when fully open.” Distrib. Central Atrican Lake-Region. P. Gomesiana, Welw. ex Oliver, Fl. Trop. Afr. i. 126; forma nov. UGANDENSIs. Frutex 4-pedalis, foliis quam in typo aliquid crassioribus, sepalorum margine solummodo colorato, capsulis paullulum quam in typo longioribus, + 55 mm. longis, + 4 mm. latis, alis angustissimis, + 12 mm. longis, + 10 mm. latis; seminibus hirsutis. Ruchigga, alt. 7000 ft. Fl. November, 397. “Shrub with dark red flowers; usually grows amid bracken.” Distrib. of type. Lower Guinea. 124 BOTANY OF THE UGANDA BOUNDARY COMMISSION. Potyaata Fiscuert, Gürke, in Engler, Jahrb. xiv. (1891) 310 ; forma. Kagera Valley near Mulema. Fl. May, 275 & 291. Native name “Ntanakorogondo.” ‘Herb with lilac- eoloured or bluish flowers." Distrib. East Tropical Africa. SECURIDACA LONGEPEDUNCULATA, Fres. in Mus. Senckenb. ii. (1837) 275. On hillside, Mulema. Fl. April, 226. Native name * Mweya.” “ Low tree used for posts and to make charcoal. Flowers pink (sepals and petals).” Distrib. Widely distributed in Tropical Africa. CARPOLOBIA ALBA, G. Don, Gen. Syst. 1. 370. Island of Buvima, Fl. March, 664. Distrib. Upper Guinea, Lower Guinea. CARYOPHYLLACEEX. CERASTIUM AFRICANUM, Oliver, Fl. Trop. Afr. i. 141. Ruchigga, alt. 7000 ft. Fl. December, 470. Distrib. Upper Guinea, Mozambique District. DnyManrA corpata, Willd. ex Roem. § Schult. Syst. v. 406. Ruchigga, alt. 7000 ft. Fl. December, 479. Distrib. Upper and Lower Guinea, Nile Land; also in Comoros, Madagascar, Tropical Asia, and Tropical America. PORTULACEE. TaniNUM CUNEIFOLIUM, Willd. Sp. Pl. ii. 864. Near Mulema in Kagera Valley. Fl. May, 270. Native name “ Mtainagerom.” “Stem chewed for tooth- ache.” Distrib. Widely spread in Tropical Africa. HyPERICINEX. HYPERICUM PEPLIDIFOLIUM, A. Rich. Tent. Fl. Abyss. i. 95. Ruchigga. Fl. November, 415. Distrib. Nile Land, Mozambique District. H. LaraNDrI, Choisy, in DC. Prod. i. 550. Near mouth of Kagera. Fl. February, 551. Distrib. Nile Land, Lower Guinea, Cape. E. G. BAKER: DICOTYLEDONES POLYPETAL.E. 125 PsorosPERMUM FEBRIFUGUM, Spach, in Ann. Sc. Nat. sér. II. v. (1836) 163. Near Musozi. Fl. February, 166. Native name * Kansironsiro." Near Mulema. Fl. April, 257. Native name “ Mkonya." * Low shrub with white flowers and red resin." Distrib. Upper Guinea, Lower Guinea, Mozambique Distriet. Haronea PANICULATA, Lodd. ex Steud. Nom. ed. II. i. 722. Coast of Vietoria Nyanza, Uganda. Fl. March, 572. “Shrub with white flowers and red sap." Tree growing usually in damp soil, Musozi. Fl. & Fr. December, 53. Native name * Mulidida." ** Yields a red gum." Distrib. Widely spread in Tropical Africa; also in Mada- gascar and Mauritius. MALVACEAE. SIDA CARPINIFOLIA, Linn. fil. Suppl. 307. Ruchigga. Fl. December, 438. Distrib. Very widely distributed in Tropical and Sub- tropical Regions. ABUTILON INDICUM, Sweet, Hort. Brit. ed. I. 54. Near Mulema. Fl. April, 195. Native name “ Mwema.” Distrib. Widely distributed in the Tropics. A, ZANZIBARICUM, Bojer, ex Masters, in Oliver, Fl. Trop. Afr. 1. 186. Near Mulema. Fl. May, 268. * Shrubby. Flowers yellow with dark centre.” Distrib. Upper Guinea, Lower Guinea, Mozambique District. PAVONIA MACROPRYLLA, E. Meyer, ev Harvey § Sonder, Fl. Cap. i. 169. Near Mulema. Fl. May, 319. Native name * Akonyeshagum." Distrib. Nile Land, South Central; also at the Cape. 126 BOTANY OF THE UGANDA BOUNDARY COMMISSION. KosTELETZKYA ADOENsIS, Masters, in Oliver, Fl. Trop. Afr. 1. 192. Below Ruchigga, alt. 5000 ft. Fl. December, 492. Distrib. Nile Land, Mozambique District. HIBISCUS micrantuvs, Linn. f. Suppl. 308. Mulema. Fl. April, 200. Native name * Kazinga." “Low shrub 3 ft. Grows on ant-hils. Red flower." Distrib. Widely distributed in the Tropies. H. aossyPrNus, Thunb. Prodr. Pl. Cap. 118. Gully above Mulema. Fl. May, 311. Native name “ Musinga.” “Stem used as a tooth-brush and as handles for spears." Distrib. Widely distributed in Tropical Africa. H. zruroPrCus, Linn. Mant. ii. 258. Hillside, Mulema. Fl. April, 228. Native name “ Niakun." Distrib, Nile Land, Cape. H. piversirouivs, Jacq. Ic. Pl. Rar. t. 551. Kitura Koki. Fl. August, 380. Lake Shore, Musozi. Fl. February, 155. Native name “ Kagei, “Rope made from the bast.” Ruchigga. Fl. November, 441. Distrib. Widely distributed in Tropieal Africa. STERCULIACE E. STERCULIA sp. Island of Buvüina. Fr. March, 656. Tree, flowers not obtained. Follicles woody, shortly stipi- tate, covered externally with a close brown tomeitum, shortly pointed. Dompeya Masrznsrr, Hook. fil. in Bot. Mag. tab. 5639 (1867). Near Kikobe Ferry, R. Kagera. Fl. March, 185. Native name “ Nkarabu." “ Bark used to make rope. A low shrub. Flowers white.” The following appears to be a variety of this species differing from type in the deeply lobed leaves and fewer flowers in the inflorescence :— Near Mulema. FI. May, 304. E. G. BAKER: DICOTYLEDONES POLYPETAL.;. 187 Native name “ Makokwa.” ** Woody stems used in hut- building. Shrubby, 10 ft. high.” Distrib. Nile Land — Central African Lake-Region, Mozambique District. DowBEYA PEDUNCULATA, K. Schum. in Engl. Pflanzenwelt Ost- Afr. C. 269. Ruchigga. Fl. November, 451. Shrub with pink flowers. Distrib. Central African Lake-Region. D. (S Eudombeya) BAGSHAWEI, sp. nov. Frutex ramis validis lignosis brunneo-corticatis, lenticellosis precipue sursum + brunneo-tomentosis; folis modice petiolatis, petiolo manifeste rufo-tomentoso, latissime suborbicularibus, 3- vel sub 5-lobatis, lobis rotundatis, latioribus quam longis utrinque dense tomentosis subtus pallidioribus margine irregulariter denticulatis palmatim 8-9 nerviis ; inflorescentia pedunculata furcata, pedunculo ramulisque rufo-tomentosis ; floribus pro rata longiuseule pedicellatis, bracteolis caducis, ovatis vel oblongo-ovatis, subaeuminatis ; petalis valde obliquis inter mediocres generis; staminibus 15 cum staminodiis 5 liguleformibus alternantibus, stamine medio triadum brevissimo, stylis cinereo- vel flavo-tomentosis, ovario pentamero, extus tomentoso. Hab. Irunga. Fl. November, 1903. 391. * Flowers white with red centres.” Folia 65-7 cm. longa, 11-12:5 em. lata. Pedunculi 5:0- 7:0 em. longi, multiflori. Petala 12-13 mm. longa. Following K. Schumann (Sterculiacee Africans, p. 20) this plant belongs to the first series in the Bect. Eudombeya, in which the peduncle bears more than three flowers. D. ($ Xeropetalum) RETICULATA, Masters, in Oliver, Fl. Trop. Afr. i. 228. Kagera Valley, near Mulema. Fl. April, 213. Native name “ Numkde." “ Wood used for handles for tools. A shrub resembling an almond-tree in blossom ; masses of flower; few leaves; flowers white, with faint pink tinge.” Hillside, Mulema. Fl. April, 261. Native name “ Mukde." Distrib. Nile Land, Mozambique District. 128 BOTANY OF THE UGANDA BOUNDARY COMMISSION. WALTHERIA AMERICANA, Linn. Sp. Pl. 673. Below Ruchigga. Fl. December, 497. Distrib. Generally distributed throughout Tropical Africa. TILIACEA. QREWIA BICOLOR, Juss. in Ann. Mus. Par. iv. (1804) 90. Mulema. Fl. April, 196. Native name *Mukoma." * Wood used for sticks and handles of tools." Island of Buvima. Fl. March, 615. Shrub with yellow flowers. Distrib. Nile Land, Mozambique District. G. SIMILIS, K. Schum. in Engler, Jahrb. xv. (1892) 118. Near Mulema. Fl. April, 221. Native name “ Mukoma." “Shafts for spears. Shrub with blue flower.” Kagera Valley, near Mulema. FI. May, 287. Native name “ Mukomakazi." “ Posts used in hut-building. Shrub with white flowers.” No, 222 is probably a form of this species. Distrib. Nile Land, Mozambique District. G. MOLLIS, Juss. in Ann. Mus. iv. (1804) 91. Island of Buvama. Fl. August, 610. Shrub with pink corolla and white sepals. Distrib. Upper Guinea, Nile Land. TRIUMFETTA PILOSA, Roth, Nov. Sp. 223. Lake shore, Musozi. Fl. & Fr. December, 13. Native name * Binsambwe." “ Fibre used.” Distrib. Nile Land, Lower Guinea. Also in South Africa, India, &e. T. RHOoMBOIDEA, Jacq. Enum. Pl. Carib. 22. Musozi. Fl. & Fr. December, 14. Native name * Binsambwe." Distrib. Widely distributed in the Tropics. T. annua, Linn. Mant. i. 73. Ruchigga. Fl. November, 429. Distrib. Nile Land, Lower Guinea. GLYPHHA GREWIOIDES, Hook. fil. in Hook. Ic. Pl. t. 760. North of mouth of Bakora. Fl. & Fr. January, 136. E. G. BAKER: DICOTYLEDONES POLYPETALA. 129 Native name * Muzira.” “A shrub from which walking- sticks are cut.” Distrib. Upper and Lower Guinea. LINER. HvaoNrA PLATYSEPALA, Welw. apud Oliver, Fl. Trop. Afr. i. 272. Lake shore, near mouth of Kagera. Fl. February, 563. “ Tree with yellow flowers." A form with rather smaller flowers than type; outer sepals curved outwards at the margins, inner shortly apiculate. Distrib. Upper Guinea, Lower Guinea. MALPIGHIACES. FLABELLARIA PANICULATA, Cav. Diss. 436, t. 264. Var. MOLLIS, Engler, Pflanzenwelt Ost- Afr. C. 232. Musozi. Fl. & Fr. January, 83. * A woody twiner. Fruit seen 40 ft. up." Distrib. of variety. Central African Lake- Region. G'ERANIACE X. Monsonta BIFLORA, DC. Prod. i. 638. Hillside, near Mulema. Fl. May, 292. ** Prostrate herb with whitish-blue flowers." Distrib. Nile Land, Lower Guinea; also Cape. GERANIUM SIMENSE, Hochst. ex A. Rich. Tent. Fl. Abyss. i. 116.] Ruchigga, alt. 7000-8000 ft. Fl. December, 469, 472, 528. “ Herb with blue or pink flowers." Distrib. Nile Land, Mozambique District. OXALIS CORNICULATA, Linn.; var. STRICTA, Oliver, FI. Trop. Afr. i. 297. Kagera Valley, near Mulema. Fl. May, 297. Near River Rufüa. Fl. December, 512. Distrib. A weed of cultivation in nearly all warm countries. IMPATIENS BAGSHAWET, sp. nov. Caulis erectus herbaeeus subsimplex vel sparse ramosus, internodiis inferioribus quam foliis 2-3-plo longioribus, superioribus brevioribus, ad nodos s:epe tentaeulis glandu- losis instructis ; foliis breviuscule petiolatis, oppositis, 130 BOTANY OF THE UGANDA BOUNDARY COMMISSION. lanceolatis vel ovato-laneeolatis, sepissime 3-4-plo longi- oribus quam latis, ad apieem attenuatis, basi rotundatis vel leviter subcordatis, discoloribus, costa subtus distincta supra haud conspicua, pennatinerviis, nerviis lateralibus subtus prominulis erecto-patulis arcuatis, margine serratis; pedunculis strietis erectis vel erecto-patulis, tenuibus, glabris, quam folia brevioribus, paucifloris, axillaribus ; floribus parviuseulis, graciliter pedicellatis ; sepalis later- alibus oblongo-ovatis obtusis quam petala brevioribus, labello infundibuliformi in calear incurvatum et applan- atum apiee acutum abrupte exeunte, calearibus quam flores longioribus; fructibus brevibus, inflatis, apice acutis inequiateralibus, lineis paucis longitudinaliter notatis, glabris. Hab. Near mouth of Kagera River. * Herb in swamps with pink flowers." Fl. & Fr. Feb. 1904. 552. Caulis usque ad fere 70 em. alt., internodiis inferioribus , 5:0-9:0 em. longis, superioribus 3:0-4'0 cm. longis. Folia 3:0—5:0 em. longa, 1:0-1:2 em. lata. Petiolus brevis 2:0— 3:0 mm. longus. This plant would fall in Series A of Dr. Warburg's Balsa- minacem Africans (Engler, Bot. Jahrb. xxii. 46) in the section of this series in which the inflorescence is more than 1-flowered. RuTACEX. TODDALIA NOBILIS, Oliver, Fl. Trop. Afr. i. 306. Hillside, Mulema. FI. April, 250. Native name “ Muzo.” ‘Spears and walking-stieks made of this. A spreading shrub with yellow flowers and dotted leaves." Musozi. Fl. January, 110. Native name * Nzo." “Tree- mallets made for hammering bark cloth." Distrib. Nile Land. Mozambique District. The following is probably another species of Toddalia, but fruit is required for determination :— Ruchigga. Fl. November, 428. Island of Buvüma. Fl. March, 621. * Shrub with dotted leaves and yellow greenish flowers.” ZANTHOXYLON NITENS, Hiern, in Welw. Cat. 1. 112. Island of Buvúma. Fr. March, 653. E. G. BAKER: DICOTYLEDONES POLYPETAL:. 131 “Thorny shrub with greenish-white flowers.” Only 9 collected. Distrib. Angola. CLAUSENA ANISATA, Hook. fil. in Niger Fl. 256. Hillside, Mulema. Fl. June, 355. “ Shrub with inconspicuous flowers and dotted leaves.’ Distrib. of type. Nile Land, Mozambique District. Var. MurLTIJUGA, Welw. ex Hiern. in Welw. Cat, i. 116. Irunga. Fl. November, 390. “Shrub with whitish flowers and dotted leaves.” Distrib. Angola. > SIMARUBE EF. HARRISONIA ABYSSINICA, Oliver, Fl. Trop. Afr. i. 311. Kagera Valley, near Mulema. Fl. May, 296. Native name * Ente erungu.” “ Low shrub with greenish flowers.” Distrib. Nile Land, Mozambique district. OcHNACES. OCHNA sp. In forest at edge of swamp near Kanabulem. Fr. February, 170. Native name “ Mowa.” “Used for posts in building. A shrub, flowers not seen." BunsERACEX. COMMIPHORA AFRICANA, Engler, in DC. Mon. Phan. iv. 14. Kagera Valley, near Mulema. Fl. April, 214. Native name ** Mesesi." ‘* Wood used for handles for tools. Low shrub. Flowers red." Distrib. Nile Land. Canarium Scuwermrurtait, Engler, in. DC. Mon. Phan. iv. 145. Musozi. Fl. January, 111. Native name * Mpafu." “Large tree with great spread of branches. Flower green. Fruit a drupe.” Distrib. Nile Land. Upper and Lower Guinea. LINN. JOURN. — BOTANY, VOL. XXXVII. L 132 BOTANY OF THE UGANDA BOUNDARY COMMISSION, MELIACE X. TvRnzA VoaErm, Hook. fil. in Niger Fl. 258. Lake shore, Musozi. Fl. January, 75. Native name * Makansi." ‘ Used for posts and sticks. A low thin shrub; flowers white with white staminal tube; seeds black.” Distrib. Upper and Lower Guinea. EKEBERGIA P COMPLANATA, Sp. nov. Arbor ramulis fusco-pubescentibus; foliis longiuscule petio- latis, apice ramulorum dispositis, 6-7-jugis, foliolis oppo- sitis, brevissime petiolulatis, obovato-oblongis vel oblongis, basi euneatis, apice acute euspidatis, utrinque + pubes- centibus, subtus pallidioribus, nervis secundariis sub- patulis adscendentibus ; paniculis quam folia multoties brevioribus, pedunculatis, puberulis ; ealyee subeampanu- lato, subobtuse 5-dentato, extus puberulo; petalis 5, ovatis, extus pubescentibus; staminibus in tubum petalis paullo breviorem coalitis; antheris 10 in margine tubi sessilibus glabris ellipticis, disco brevi glabro annulari; ovario 4—5-loculari, stylo glabro ovario equilongo, stigmate crasso, obconico 4-5-lobulato. Hab. Island of Buvüma, Victoria Nyanza. Fl. March 1904. 600. “ Tree with white flowers.” Folia 22-32 em. longa. Foliola 5:5-7:5 em. longa, 22-2-7 cm. lata, nervo medio supra impresso. Pedunculus 13- 17 em. longus, inferne complanatus. Petala + 4 mm. longa, 2 mm. lata. Differs from Æ. senegalensis, A. Juss., by the leaflets being in 6-7 pairs aud more or less pubescent above and below. Fruit is required. E. SENEGALENSIS, Á. Juss. var. cortacea, C. DC.; floribus, 5-meris ; calycis dentibus acutis; petalis 5, ovatis, pubescentibus, © 3:5 mm. longis; tubo urceolato tomen- toso, + 2 mm. longo; antheris 10, ovario glabro 5-loculari ; stylo glabro, stigmate capitato. Hab. Kitara. In flower, August 9th, 1903. 378. I have made the above notes, as this variety was described from fruiting specimens. E. G. BAKER: DICOTYLEDONES POLYPETALJ. 133 EkEBERGIA PETITIANA, A. Rich., var. nov. AUSTRALIS; arbor cortice nigrescente, foliis imparipinnatis, foliolis s:epissime 7-9 lanceolatis, basi valde inequilateris, glabris subtus haud glaueis ; racemis folio multoties brevioribus ; floribus 4- vel 5-meris, ealycis dentibus semiovalibus ; petalis oblongis obtusis; ovario 2-loculari. Hab. Ruchigga. In flower December 1903. 466. * Tree with white flowers." Distrib. of the type. Abyssinia. TRICHILIA EMETICA, Vahl, Symb. i. 31. Coast of Victoria Nyanza. Fl. March, 576. “Tree with pale yellow flowers.” Distrib. Widely distributed in Tropical Africa. CHAILLETIACER. DICHAPETALUM BUVUMENSE, sp. nov. Frutex cortice brunneo-lutescente, ramulis novellis puberulis, foliis chartaceis, adultis utrinque glabris, oblongis vel anguste obovato-oblongis, apice acutis, costa superne impressa, nervis lateralibus utrinque 7-8 arcuatim adscendentibus, prope marginem arcuatim connectis, venis reticulatis ; petiolis brevibus, cinereo-pubescentibus ; peduneulis subnullis eymas congestas ferentibus; sepalis oblongo-lanceolatis, pubescentibus, apice obtusis; petalis quam sepala liplo longioribus, longe cuneiformibus usque ad medium bilobis, lobis lineari-oblongis ; stamini- bus quam petala longioribus, ovario ovoideo in stylum superne tenuem filiformem petala superantem contracto, stigmate breviter trifido. Hab. Island of Buvüma, Victoria Nyauza. March 1904. 595. Folia 9:0-12:0 em. longa, 3°8-5*2 lata. Petiolus 2-3 mm. longus. Sepala +3 mm. louga. Petala angusta, 4°5 mm. longa. Stamina + 6 mm. longa. Stylus + 6:5 mm. longus. Species D. acutifolia, Engler, affinis, differt pedunculis brevioribus, foliis longioribus, &c. No. 650, also from the Island of Buvüma, may belong to this genus, but there are no flowers. Dr. Bagshawe notes that the eut surface of the branches give out a smell as of sulphuretted hydrogen. L2 134 BOTANY OF THE UGANDA BOUNDARY COMMISSION. OLACINER. XIMENIA AMERICANA, Linn. Sp. Pl. 1193. Island of Buvima. Fl. & Fr. March, 666. * Shrub with dark red flowers and bright red fruits.” Distrib. Widely distributed in Tropical Africa, also in Tropical America. LEPTAULUS DAPHNOIDES, Benth. in. Benth. & Hook. fil. Gen. Pl. i. 351. Island of Buvüma. FJ. March, 637. * Shrub with white flowers.” Distrib. Upper Guinea. APODYTES BENINENSE, Hook. fil. in Hook. Ic. Pl. t. 778. Mouth of Kagera. Fl. February, 161. Coast of Victoria Nyanza, Uganda, 582. Native name * Kafumbo.” ‘A liane; from the bark are made lines for fishing. Flowers white.” Distrib. Upper Guinea and Lower Guinea. CELASTRACEE. GYMNOSPORIA SENEGALENSIS, Loes. Var. INERMIS, Rich., forma cortacea, Loes. in Engler, Jahrb. xvii. (1893) 541. Near Mulema. Fl. April, 211. “ Shrub with white flowers." Var. INERMIS, forma MACROCARPA, Loes. l. c. 542. Near Mulema. Fl. April, 256. Native name * Munyaburuk." “ Handles made for hatchets. Shrub with white flowers.” Var. spinosa, Engler, Jahrb. xvii. (1893) 542. In gully near Mulema. Fl. May, 293. Native name * Omwah." “Spoons made from the wood. Shrub with white flowers." Distrib. Widely distributed in Tropieal Africa; Mediter- ranean Region and eastward to India. G. FASCICULATA, Loes. in Engler, Jahrb, xix. (1893) 232. Lake shore, Musozi. Fl. February, 162. Native name * Mpoko." * Walking-sticks cut. A large shrub; flowers greenish." E. G. BAKER: DICOTYLEDONES POLYPETALX. 135 Near Mulema. Fl. May, 263. Distrib. Mozambique District. Also in the Comoro Islands. ErxopENDRON ÆTHIOPICUM, Oliver, in Fl. Trop. Afr. i. 365. Irunga. Fl. November, 385. * Trees with greenish flower." Distrib. Lower Guinea; Mozambique District. RHAMNEE. ScUTIA INDICA, Brongn., var. OBLONGIFOLIA, Engler, in Jahrb. xix. (1901) Beibl. no. 47. 37. Near Mulema. May, 303. Native name “ Mugusha.” “Shrub. Fruit eaten." Distrib. Kilimanjaro District. HEriNUS MYsTACINUS, E. Meyer, ex Steud. Nom. ed. IL. 1. 742. Ruehigga. Fl. November, 453. * Climbing shrub with greenish-white flowers." Gorge near Mulema. Fl. April, 262. Native name * Muvimba." “Given to cows to produce milk." Distrib. Nile Land, Mozambique District. AMPELIDEX, Cissus arauta, Hook. fil. in Niger Fl. 261. Musozi. December, 65. Distrib. Upper Guiuea. C. QUADRANGULARIS, Linn. Mant. i. 39. Near Mulema. Fl. April, 193. Native name * Kagenzanda." “Climber; flowers reddish.” Distrib. Widely distributed in Tropical Africa. Also Arabia, Comoro Is., Madagascar, &e. C. ADENOCAULIS, Steud. ex A. Rich. Tent. Fl. Abyss. i. 111. Near Mulema. May, 308. Native name * Kibombo." “A climber: the fleshy root is used as an aperient.” Distrib, Nile Land, Mozambique District. There are also two other species of Cissus insufficiently repre- sented in the collection,—No. 151 gathered on the Lake shore, Musozi, Native name “ Kikasakasa’’; and No. 307 from near Mulema, Native name “ Kibombo.” 136 BOTANY OF THE UGANDA BOUNDARY COMMISSION. Lera avINEENSIS, G. Don, Gen. Syst. i. 712. Island of Buvüma. Fl. March, 612. Distrib. Tropical Africa and Islands. SAPINDACEX. CARDIOSPERMUM GRANDIFLORUM, Swartz, var. ELEGANS, Hiern, in Welw. Cat. Part i. 166. Island of Buvima. Fl. March, 618. “ Liane with pinkish-white flowers." Distrib. Lower Guinea. PAULLINIA PINNATA, Linn. Sp. Pl. 366. Musozi. Fl. January, 115. Native name * Kabugu." ** Climber used to make baskets.” Hillside, Musozi. Fr. January, 118. Distrib. Widely distributed in Tropical Africa; also in Madagascar and Tropical America. ALLOPHYLLUS AFRICANUS, Beauv. Fl. Owar. ii. 54, t. 107. Lake shore, Musozi. Fl. January, 77. Distrib. Upper Guinea. A. SUBCORIACEUS, sp. nov. Frutex foliis parviusculis, breviuscule petiolatis, petiolo demum glabro, foliolis lateralibus quam terminalibus paullo brevioribus, subcoriaceis, oblongis, basin versus angustatis, omnibus subsessilibus, margine integris vel hine inde serratis, utrinque glabris, nervis venisque utrinque subtus prominentibus ; inflorescentia nune simplice pseudo-racemosa, nune ramos utrinque emittente sepissime folium subduplo superante; floribus generis, albis parvulis, in cymulas approximatas dispositis, rhachi pedieellisque pilosulis. Hab. Near Mulema. Fl. May 1903. 264. Native name * Mtete.” “Sticks rubbed together and used to produce fire. A spreading shrub with white flowers." Foliola 3:5—-5:5 em. longa, l:5-2:1 em. lata. Petiolus com- munis, 8-13 mm. longus. Not closely allied to any species of which I have seen speci- mens. Noticeable on account of the rather small, subcoriaceous leaflets and small numerous flowers in few-flowered cymules. E. G. BAKER: DICOTYLEDONES POLYPETAL.. 137 ALLOPHYLLUS LATEFOLIOLATUS, Sp. nov. Frutex ramulis pubescentibus, foliis mediocriter petiolatis, petiolo pubescente, foliolis late ovatis, lateralibus oblique oblongo-ovatis, intermedio basin rotundato vel latissime euneato, apice acuminato, lateralibus basi oblique subeor- datis vel rotundatis, circiter 12 plo longioribus quam latis, ommibus papyraceis, margine grosse serratis, utrinque glabratis, nervis pilosulis, nervis venisque utrinque sub- equaliter prominentibus ; inflorescentia pseudo-racemosa, paullo folium superante ; floribus generis, brevipedicellatis in eymulas inferne remotas dispositis, rhachi pedicellis calycibusque pilosiusculis. Species A. Welwitschii, Gilg, affinis, sed floribus majoribus. Hab. Lake shore, Musozi. February 4th, 1903. 153. Native name “ Kakoto." “Shrub, fish-baskets made. Flowers white.” Foliola 6:0-10*0 em. longa, 4-6°5 em. lata. Petiolus com- munis, 2:5—4*0 cm. longus. A, PSEUDO-PANICULATUS, Sp. nov. Frutex foliis mediocriter petiolatis, petiolo patenti rufo- hirsuto, trifoliolatis, foliolis oblongis vel oblongo-obovatis, intermedio basin versus sensim longe euneato-angustato apiee aeuminato, lateralibus quam terminalibus paullo brevioribus, basi oblique subrotundatis, omnibus papy- raceis, precipue in parte superiore argute serratis, utrinque + pilosulis, nervis venisque subtus prominentibus ; inflorescentia eopiose ramosa pseudo-panieulata folium superante; floribus generis, numerosis in eyinulas dispositis, rhachi rufo-tomentosa, pedicellis calycibusque glabris. Hab. Hill near R. Rufáa. Jannary 1904. 544. * Shrub with white flowers only. & only. “Minute petals covered with white hairs; disc orange-red, 4-lobed." Foliola 6:0-9-0 em. longa, 3:0-3:8 em. lata. Petiolus com- munis, 4/0—5:0 em. longus. Differs from A. stachyanthus, Gilg, especially in the apex ot the leaflets. DEINBOLLIA FULVO-TOMENTELLA, Sp. NOV. Arbor 10-pedalis, foliis magnis eire. 7-jugis, foliolis amplis oppositis vel suboppositis coriaceis, petiolulo erasso brevi 188 BOTANY OF THE UGANDA BOUNDARY COMMISSION. tenuiter puberulo, oblongis, penninerviis, nerviis lateralibus utrinque sepissime 11-13, prope marginem arcuatis et inter se conjunctis, supra impressis, fere glabris, subtus reticulatis et pubescentibus ad apicem attenuatis apice ipso obtusis, basi rotundatis, panicule ramulis precipue inferne + elongatis, tenuiter fulvo-tomentellis ; floribus masculis albis; sepalis concavis imbricatis, dorso rufo- pubescentibus margine subseariosis ; petalis unguiculatis, ovatis + albo-hirtis, ad basin squamula magna tomentosa instructis; staminibus circa 18-20, petalis subsequilongis, filamentis pubescentibus, diseo manifeste conspicuo. Species D. insigni, Hook. fil., valde affinis. Hab. Island of Wema, Victoria Nyanza. Fl. March 1904, d onlv. 592. Folia 50-60 em. longa. — Folicla 16-245 em. longa, 7-9 cm. lata. Sepala 4-5 mm. longa. Petala + 6 mm. longa + 4mm, lata. Filamenta + 5 mm. longa. PAPPEA UGANDENSIS, sp. nov. Frutex ramis vetustioribus nigrescenti-corticatis, foliis m , apicem ramulorum versus confertis, oblongis, apice obtusis, basi rotundatis, nervis lateralibus utrinque 10-17 subtus prominentibus, modice petiolatis, junioribus utrinque brevissime tomentosis, adultioribus supra demum gla- bratis; inflorescentia axillari, spleato-racemosa; calyce cupulari 5-lobato, in flore 3 staminibus 8, exsertis, ovario rudimentario, in flore 9 ovario 2-lobo, 3-loculari hirsuto, disco annulari hy pogyno, stylo brevi crasso erecto superne longitudinaliter tristigmatoso. Species P.capensi, Ecklon & Zeyher, affinis, differt floribus brevissime pedicellatis, racemis foemincis quam racemis masculis longioribus, filamentis longioribus. Hab. Hillside, Mulema. Fl. June 1903. 369. “Shrub with inconspicuous flowers.” Folia 5:0-10 em. longa, 2°3-4°5 em. lata. — Petiolus 6-10 mm. longus. Fl. d filamentis + 1:5 ium. longis sparse pilo- sulis. Fl. 9 stylus + 1:5 mm. longus, DODONÆA viscosa, Jacq. Enum. Pl. Carib. 19. Below Ruchigga, alt. 5,600 ft. FI. December, 521. Distrib. Widely distributed in the Tropics. E. G. BAKER: DICOTYLEDONES POLYPETALX. 139 Bersama Horsti, Giirke, in Engler, Jahrb. xix. (1894) Beibl. n. 47. 36. Hab. Kitara, Koki. August, 379. * Tree with white flowers.” A. plant with unopened flowers, presumably identical with the type which was collected by Holst in Usambara. B. rAULLINIOIDES, Baker, in Oliver, FI. Trop. Afr. 1. 435. Ruchigga. Fl. November, 398. * Free with white flowers." Distrib. Upper and Lower Guinea. ANACARDIACE.E. ANAPHRENIUM PULCHERRIMUM, Schweinf. Fl. Æthiop. Beitr. i. 32. Mulema. Fl. May, 306. Native name * Mukerenge." “Tree with white flowers and white latex." Distrib. Nile Land. Rus vitrosa, Linn., var. TOMENTOSA, Oliver, Fl. Trop. Afr. i. 439. Hillside, Musozi. Fl. January, 123. Native name “ Kakansu." Another form of this species was collected at Musozi. Fl. December, 50. Native name * Msese.” “Small branches used as a tooth stick." Distrib. Widely spread in Tropical Africa. R. arAucEsCENS, A. Rich. Tent. Fl. Abyss. i. 143. Near Mulema, Fl. May, 192. Native name * Msese." * An infusion of the leaves is taken for pain iu the abdomen. Shrub with green flowers.” Distrib. Nile Land, Mozambique District. There is another species of Rhus in the collection from lrunga, 384, of which only 9 fs. were obtained. PSEUDOSPONDIAS microcarpa, Engler, in DC. Mon. Phan. iv. 258. Mouth of Kagera. Fr. December, 71. Lake shore, Musozi. Fl. February, 163. Native name * Nkoba." “Planks eut for canoes. Tree, flowers white. Wood contains resin." Distrib. Upper and Lower Guinea. 140 BOTANY OF THE UGANDA BOUNDARY COMMISSION. LANNEA RUFESCENS, Engler, Jahrb. xxiv. 495. Arbor ramulis floriferis dense rufescenti-tomentosis quam folia brevioribus ; pedicellis brevissimis; calycis segmentis ovatis; petalis oblongo-ovatis, calyce triplo longioribus; staminum filamentis petala brevioribus (in floribus mas- culis); ovario rudimentario stylis coronato. Hab. In gorge near Mulema. June 19083. 267. “Tree with reddish-green flowers, which come out before the leaves." Inflorescentia 5-12°5 em. longa. Petala + 3 mm. longa. Anthere 1:25 mm. longe. A deseription of the flowers is given, as Engler's specimens were sterile. L. STUHLMANNI, Engler, Pflanzenfam. Nachtr. 214. Kagera Valley, near Mulema. Fl. May, 286. Native name “ Mserot." ‘* Wood used for posts in hut- building. Shrub with yellow flowers." Distrib. Mozambique District. L. rutva, Engler, Pflanzenfam. Nachtr. 213. Near Mulema. Fl. June, 347 & 354. * Shrub with greenish or whitish flowers." Distrib. Victoria Nyanza. CONNARACER. AGELHA NITIDA, Solander, ev Planchon, in Linnea, xxii. (1850) 437. Island of Buvüma. Fl. March, 609. Distrib. Upper Guinea. A species of Connarus was obtained near the mouth of the River Kagera, 158, but without flowers. LEGUMLNOS&. CROTALARIA ADENOCARPOIDES, Jaubert, in Engler, Pflanzenwelt Ost- Afr. C. 206? Hab. Ruchigga. November, 404. * Shrubby plant; flowers brown and yellow." Differs from the type in having a short raceme. Closely allied also to C. argyrolobioides, Baker, from the Nyika Plateau. C. IMPERIALIS, Taubert, in Engler, Pflanzenwelt Ost- Afr. C. 206. Hillside, Musozi. F!. & Fr. December, 46. E. G. BAKER: DICOTYLEDONES POLYPETALE. 141 Native name * Ksamba Ndega." “Shrub with flowers greenish-yellow." Distrib. Central African Lake-Region. CROTALARIA CEPHALOTES, Steud. ev A. Rich. Tent. Fl. Abyss.i. 156. Near Mulema. April, 206. Distrib. Upper and Lower Guinea, Nile Land. C. LANCEOLATA, E. Meyer, Comm. Pl. Afr. Austr. 24. Ruchigga, 7000 feet. Fl. November, 421. “ Erect plant, 8 ft. high, with yellow flowers." Distrib. Nile Land, Mozambique District. The following is another species of CROTALARIA, but without pods. Hillside, Mulema. Fl. April, 234. Native name * Munyakishak." ‘“Tubers eaten. Flowers yellow streaked with brown." Hillside, Mulema. Fl. April, 253. Native name * Katarugum." ‘ Root eaten for toothache. Herb with yellow flowers striped with brown." ApENOCARPUS Manntt, Hook. fil. in Journ. Linn. Soc., Bot. vii. (1864) 189. Ruchigga. Fl. November, 424. “Shrub with conspicuous yellow flowers.” Distrib. Upper Guinea. TRIFOLIUM PoLYsTACHYUM, Fresen. in Flora, xxii. (1839) 50. River Rufüa, alt, 4400 ft. Fl. December, 536. Herb in marshy ground with pink flowers. Distrib. Nile Land, Lower Guinea, Mozambique District. T. suBROTUNDUM, Hochst. et Steud. in Flora, xxiv. (1841) i. Intell. 325 A. Rich. Tent. Fl. Abyss. i. 172. River Rufüa. F]. December, 508. * Herb with pink flowers in 6 in. water." The following are also allied to this species:— On a hill near Rufüa. Fl. January, 550. With calyx-tube + 3mm.long, teeth 5 mm. Standard oblanceolate, 7-8 mm. long. Ruchigga, alt. 7000 feet. Fl. November, 413. Herb with red flowers with 10-ribbed calyx and linear setaceous teeth. Distrib. Nile Land, Upper and Lower Guinea. 142 BOTANY OF THE UGANDA BOUNDARY CCMMISSION. INDIGOFERA (§ Acanthonotus) DREPANOCARPA, ZTaubert, in Engl. Pflanzenwelt Ost- Afr. C. 209. Lake shore, Musozi. Fl. & Fr. December, 18. * Shrubby, prostrate; flowers red." Distrib. Central African Lake-Region. InpigoFERx (§ Tinctorim) sp. I. macrophylla, Schum. & Thonn., affinis, differt ambitu foliolorum, vexillo juvenili extus fulvo-tomentello, foliolis sepissime 11-13 etc., et ab T. emarginella, Steud., differt racemis longioribus etc. Hab. Ruchigga. Fl. November 1908. 443. * Shrub with dark orange flowers." InpIGoFERA BaAGSHAWET, sp. nov. Frutex erectus ramosus. Rami virgati erecti vel suberecti sursum + albo-strigosi deorsum glabri, foliis in siccitate nigrescenti-cinereis imparipinnatis szepissime 3-jugis cum impari, rarissime 2-1-jugis cum impari, foliolis oblanceo- latis 3-4plo longioribus quam latis utrinque + albo- strigosis, costa superne impressa subtus conspicua, apice acutis petiolulis brevissimis foliolis lateralibus oppositis ; racemis nune quam foliis brevoribus nune paullo longiori- bus, pedunculis tenuibus albo-strigosis laxiuseule pauci- floris, pedicellis tenuibus brevibus s:pissime quam calyce brevioribus, calyce extus albo-strigoso, calycis lobis anguste lanceolatis acutis tubo subeequilongis, floribus rubris (ex collect.) in speciminibus mihi obviis delapsis, legumine juvenili cinereo recto etoruloso lateraliter subcompresso extus albo-strigoso paucispermo stigmate coronato. I. heterocarpe, Welw., affinis. Hab. Musozi. Shrabby plant. Sand by Lake. Flowers red. Fl. December, 11. Folia 1:0-I:3 em. longa. Foliola 8-11 mm. longa, 2-3 mm. lata. Pedunculi 1:0-2:0 em. longi, pedicelli sepe vix 1 mm. longi. Calycis tubus +15 mm. longi. Legumen juvenile 3-6 mm. longum. The difference between this and J. heterocarpa is very marked in the dried state, the former being silvery-black, the latter remaining green; the former also branches some distance from the base, the latter from the base. E. G. BAKER: DICOTYLEDONES POLYPETALE. 143 INDIGOFERA HIRSUTA, Linn., var. POLYSTACHYA, Welw. ex Baker, in Oliver, Fl. Trop. Afr. ii. 89. Ruchigga, alt. 7200 ft. Fl. December, 523. Shrub with dark red flowers. Distrib, Angola. Teparosta Voorrit, Hook. fil. in Niger Flora, 296. Hillside, Musozi. Fl. & Fr. December, 41. Distrib. Upper and Lower Guinea, Nile Land, and Mozam- bique Distriet. T. Nxasz, Baker fil, in Trans. Linn. Soc. ser. IT. Bot. iv. (1895) 9. Irunga. Fl. November, 389. Ruchigga, alt. 6000-7000 ft. Fl. November, 426. “Shrub with bluish flowers.” Distrib. Nyasa-land. T. PANICULATA, Welw. ex Baker, in Oliver, FI. Trop. Afr. ii. 122. Mouth of Kagera. Fl. December, 26. Ruchigga. Fl. November, 442. “Shrub with orange-coloured flowers.” Distrib. Angola. SESBANIA EGYPTIACA, Poir. Encycl. Méth. vii. 128. Below Ruchigga, alt. 6000 ft. Fl. December, 491. " Trees with yellow flowers and dusky standards." Distrib. Upper Guinea, Nile Land. Extends through Tropieal Asia to North Australia. S. ACULEATA, Poir. Encycl. Méth. vii. 198. Near Mulema. Fl. May, 294. Native name * Munyezenyege." * Yellow flowers." Distrib. Upper Guinea, Mozambique District; also Natal and through Tropical Asia to Australia. DIPHACA tricnocarpa, Taubert, in Engler, Pflanzenwelt Ost- Afr. C. 213. Near River Kagera, Mulema. Fl. April, 198. Native name * Muwendobisa.” ^ * Kraals for cattle are made of this." Distrib. Central African Lake-Region. 144 BOTANY OF THE UGANDA BOUNDARY COMMISSION. ÆSCHYNOMENE SCHIMPERI, Hochst. ex A. Rich. Tent. Fl. Abyss. i. 202. Mouth of Kagera. Fl. & Fr. December, 24. Distrib. Nile Land. Æ. invica, Linn. Sp. Pl. 713. In swamp, River Rufta. Fl. December, 507. Distrib. Widely spread m Tropical Africa. Smirnta Korscuyt, Benth. in Mig. Pl. Jungh. 211. Musozi. Fl. February, 149. Ruchigga. Fl. November, 403. “ Shrub with glandular hairs and yellow flowers." Distrib. Nile Land. S. ÆSCHYNOMENOIDES, Welw. ex Baker, in Oliver, Fl. Trop. Afr. ii. 153. Ruchigga. Fl. November, 395. “Shrub with white flowers with faint blue stripes." Distrib. Angola. ZORNIA DIPHYLLA, Pers. Syn. ii. 318. Hillside, Mulema. Fl. April, 236. Near River Rufüa. Fl. December, 505. Distrib. Widely distributed in the Tropics. Drsmopium Scarrr, DC. Prod. ii. 834. Ruchigga. Fl. December, 478. Distrib. Widely distributed in Tropical Africa. Also Natal, Masearene Islands, East Indies, and Malaya. Pycrospora HEDYSARIOIDES, R. Br. ex Wight & Arn. Prod. 197. Hillside, Mulema. Fl. April, 233. Native name “ Kazuganji." “Used for cough." Distrib. Widely spread in Tropical Asia. Australia. Pseuparturta Hooxkznr, Wight 4 Arn. Prod. Fl. Ind. 209. Musozi. Fl. & Fr. February, 148. Distrib. Nile Land, Lower Guinea, Mozambique District ; also Natal. Vicia sativa, Linn., var. ABYSSINICA, Baker, in Oliver, Fl. Trop. Afr. i. 172. Ruchigga, alt. 7500 ft. Fl. December, 471. “ Herb with bluish-pink flowers.” Distrib. Nile Land. E. G. BAKER: DICOTYLEDONES POLYPETAL.E. 145 ABRUS PRECATORIUS, Linn. Syst. ed. XII. 479. Hillside, Mulema. Fl. June, 370. Distrib. Widely distributed throughout the Tropies. A. PULCHELLUS, Wall. ex Thw. Enum. Pl. Zeyl. 91. Island of Buvüma. Fr. March, 652. Distrib. Nile Land, Upper Guinea, Central African Lake- Region. Also in Tropical Asia. ERYTHRINA TOMENTOSA, R. Br. in Salt, Abyss. App. 62. Ruchigga, alt. 7200 ft. Fl. December, 531. “ Tree 25 ft., with scarlet flowers." Distrib. Nile Land, Mozambique District. E. BAasuaWET, sp. nov. Arbor vel frutex, ramis cinereo- vel albido-corticatis, spinis brevibus nigrescentibus; foliolis 3 ovato-lanceolatis vel suborbicularibus, basi rotundatis, apice acutis, margine integris haud lobatis, nervis lateralibus erecto-patulis, utrinque 8-10, modice petiolulatis, preter nervos glabri- usculis; racemis axillaribus ; floribus coccineis, pedun- culis fusco-tomentellis ; calyce spathaceo, subeoriaceo, glabriuseulo, laciniis brevibus ; vexillo alis longiori, subsessili, ezeteras partes floris amplectente ; alis inzequi- lateraliter oblongis, brevissime unguiculatis, glabris ; ovario stipitato pluriovulato; stylo incurvo; leguminibus basi in stipitem attenuatis, plurispermis, valvis extus coriaceis molliter et breviter tomentosis. Species E. Brucei, Schweinf., affinis, differt racemis floribus paucioribus, pedunculis fusco-tomentellis non lanato- tomentosis. Hab. Sandy soil north of mouth of River Bakora. January 1903. 137. Native name "**Murungu." * Wood used for drums. A shrub or tree, circumference, 5 it. from the ground, 8 ft. ; thorns on trunk ; flowers red, very few leaves." Foliola 7:0-11:5 em. longa, 4°5-6°7 cm. lata. Calyx + 25 em. longus, Vexillum externe glabrum + 27- 2'8 cm. longum, subsessile. Al» + 1 cm. longe. Carina vexillo multo minor suborbicularis 6-7 mm. longa. 146 BOTANY OF THE UGANDA BOUNDARY COMMISSION. PHASEOLUS SCHIMPERI, Tuubert, in Engler, Hochgebirgsfl. 267. Hillside, Mulema. Fl. April, 235. “Trailing plant with purple flowers and large woody rhizome." Distrib. Nile Land. VIGNA FRAGRANS, Sp. Nov. Caulis subvolubilis striatus precipue superne griseo-tomen- tosus, foliis mediocriter petiolatis, trifoliolatis, utrinque tomentosis, terminalibus trilobatis, lobo medio majore, apice obtusis mueronatis, basi late euneatis, petiolulatis, foliolis lateralibus subsessilibus inzequilateraliter trilobatis, nerviis subtus prominentibus, stipulis persistentibus, lanceolatis striatis; pedunculis sparsissime pilosulis, apice 1-2-3.floris ; foribus cæruleis; calyce campanulato, ineiniis triangularibus, subacuminatis, hirtis, quam tubus brevioribus; vexillo extus glabro, suborbiculari, unguicu- lato ealycem pluries excedente; alis oblique oblongo- subobovatis, basi auricula instructis; stylo arcuato, incurvo, infra apicem biseriatim barbato, stigmate laterali, ovali-globoso ; leguminibus plurispermis, extra pilosulis, subtorulosis, calyeem multoties excedentibus. Species V. trilobe, Walp. affinis, differt calyce breviore, ete. Hab. Hillside, Mulema. FI. April, 232. Native name “ Niahenge.” “Seeds eaten. Trailing herb with large fibrous rhizome. Flowers blue and fragrant (keel petals white)." Foliola terminalia 2:0-2:3 cm. longa, lateralia l:5-l's em. longa. Peduneuli 6°0-12°5 cm. longi. Calyx 5-6 mm. longus. Carina 17-18 nim. longa. If the genus Liebrechtsia, De Wildem. (Fl. Katanga, 70), be maintained, it would include the above species. The differeuces in the style and stigma do not. however, seem sufficiently marked to warrant its separation from Vigna. ERIOSEMA MONTANUM, Baker fil., in Journ. Bot. xxxiii. (1895) 142. Ruchigga. Fl. December, 475. * Shrub with yellow flowers, 4 ft. high." Distrib. Nile Land, Mozambique Distriet. E. aroMERATUM, Hook. fil. in Niger Flora, 313. Musozi. Fl. & Fr. January, 73. Distrib. Widely distributed in Tropical Africa. E. G. BAKER: DICOTYLEDONES POLYPETAL.E. 147 RHYNCHOSIA CYANOSPERMA, Benth. ex Baker, in Oliver, Fl. Trop. Afr. ii. 218. Island of Buvima. Fl. March, 635. ‘Twining shrub with wallflower-coloured petals.” Distrib. Mozambique District. Also Mascarene Islands and India. R. rEstnosa, Hochst. ex Baker, Fl. Trop. Afr. ii. 218. Near Mulema. Fl. March, 191. Native name * Kashaka Kalibioya." “ From the leaves a liniment is made. Low shrub with yellow-green flowers and glandular hairs.” Distrib. Nile Land. DALBERGIA UGANDENSIS, sp. nov. Frutex scandens, ramis plumbeo-corticatis lignosis, striatis, glabris, novellis nigrescenti-corticatis, foliolis subeoriaceis, sepissime 8-12, alternis, oblongis, basi rotundatis, apiee rotundatis vel leviter emarginatis, costa superne im- pressa, subtus conspicua, nervis lateralibus tenuibus numerosis, erecto-patulis, utrinque glabris, subtus palli- dioribus, petiolis petiolulisque glabris, stipulis deciduis ; floribus panieulatis, panieula axillari et terminali fulvo- tomentosa; calyce extus fulvo-tomentoso, segmentis ovatis quam tubo pluries brevioribus ; petalis violaceis ; vexillo carinam excedente, ovato-oblongo, unguiculato, carina obtusa naviculariformi, ungue gracili; alis quam carina longioribus : staminibus ut in eongeneribus, an- theribus parvis; ovario pilis albidis tecto; stylo gracili, stigmate parvo terminali ; legumine ignoto. D. saxatili, Hook. fil., affinis, differt panuliculis + elongatis, pedicellis erassiusculis dense fulvo-tomentosis, calycis segmentis brevioribus. Hab. Coast of Victoria Nyanza, Uganda. « Liane with violet flowers.” Fl. March, 587. Petiolus communis 10-13 cm. longus, petiolulus 2-3 mm. longus. Foliola 2:5-44 cm. longa, l1-l'6 em. lata. Calyx fere 5 mm. longus. Vexillum 7-8 mm. longum. Carina 6-6'5 mm. longa. BarurA RADCLIFFEI, sp. nov. Frutex ramis cinereo- vel flavescenti-corticatis, striatulis, glabrescentibus, teretibus ; foliis modice petiolatis, petiolo LINN. JOURN.—BOTANY, VOL. XXXVII. M 148 BOTANY OF THE UGANDA BOUNDARY COMMISSION. sepissime nigrescenti, ambitu B. barombiensis, oblongis vel ovatis, apice subacuminatis, basi rotundatis vel late cuneatis, chartaceis, utrinque glabris, nervis lateralibus, utriusque 8-10 ereeto-patulis, prope marginem inter se conjunctis, subtus prominulis, nervis nervulisque densi- uscule reticulatis; racemis multifloris, demum laxis, pedunculis pedicellisque pubescentibus, bracteolis calyce multo minoribus; calyce membranaceo externe pubes- centi; vexillo oblongo-ovato, breviter unguiculato, quam calyce vix longiore ; ovario glabro, stylo curvato ; legumine ignoto. Species ex affinitate B. polygalacee, Baker, differt primo intuitu folis angustioribus et praecipue petiolis brevioribus. Hab. Lake shore, Musozi. Fl. January, 74. Native name * Matchko.” * Used for mallets to beat bark- cloth. A shrub with white flowers and conspicuous in- florescence." Petiolus 1:0-1:5 em. longus. Folia 80-10 em. longa, 3°5-4:0 em. lata. Vexillum 6-7 mm. longum. PrEROLOBUM LACERANS, R. Dr. in Salt, Abyss. App. 64. Banks of Kagera River near Mulema. Fl. May, 269. Native name * Orge." “Specimens in fruit sent subse- quently.” Distrib. Abyssinia. CASSIA DIDYMOBOTRYA, Fres. in Flora, xxii. (1839) 53. Near Mulema. Fl. June, 366. Distrib. Nile Land, Lower Guinea, Mozambique District. C. FarciNELLA, Oliver, in Fi. Trop. Afr. ii. 281. Near Mulema. Fl. March, 190. Native name “ Kanyeüra." “ Wiry shrub, 12 inches high, with yellow flowers." Distrib. Nile Land. C. Grantit, Oliver, in Fl. Trop. Afr. ii. 279. Mulema. Fl. June, 360. Wiry plant, 18 inehes high, with yellow flowers. Distrib. Mozambique District. C. MIMOSOIDES, Linn. Sp. Pl. 379. Mulema. Fl. June, 365. Distrib, Common in the Tropics. E. G. BAKER: DICOTYLEDONES POLYPETALX. 149 BarkragA Ewriwr, Zaubert, in Engler, Pflanzenwelt Ost-Afr. C. 198. Lake shore, Musozi. .Fl. January, 79. Native name * Musa.” '* Used as a medicine for deafness ; the leaves are heated, and the juice which exudes is squeezed into the ear. "Tree 30—40 feet high." Island of Buvüma. Fl. March, 607. Shrub with white flowers." Distrib. Central African Lake-Region. PaRKIA FILICOIDEA, Welw. ex Baker, in Oliver, Fl. Trop. Afr. ii. 324. Island of Buvima. Fl. March, 643. “Tree with heads of red flowers." Distrib. Upper and Lower Guinea, Mozambique District. ENTADA ABYSSINICA, Steud. in A. Rich. Tent. Fl. Abyss. i. 234. North of mouth of Bakora. Fr. January, 143. Distrib. Nile Land, Lower Guinea. PrPTADENIA AFRICANA, Hook. fil. in Niger Flora, 330. Island of Buvüma. FI. March, 647. * Tree with brownish flowers." Distrib. Upper and Lower Guinea. DricunosrAcHYs NUTANS, Benth. in Hook. Journ. Bot. iv. (1842) 353. Near Mulema. FI. April, 255. . Native name * Muyebe." “Handles for hoes made. Tree with flowers yellow at distal, pink at proximal end." Distrib. Widely distributed iu Tropical Africa. Mimosa ASPERATA, Linn. Syst. ed. X. 1312. Kagera Valley, near Mulema. Fl. May, 277. Native name “ Orge.” “Shrub growing in marshy ground. Inflorescence pink." Distrib. Widely spread through Tropical Africa. Acacta VEREK, Guill. & Perr. Tent. Fl. Seneg. i. 245, t. 56. The following is either a form of the above or a close ally. The legumes are broader and generally fewer-seeded than those of Acacia Verek figured in Reliq. Kotsch. t. 3. Hab. Hill-top, Mulema. May 1903. 266. Native name * Mukonje." *' Handles for hoes made. A tree with white flowers." M2 150 BOTANY OF THE UGANDA BOUNDARY COMMISSION. ACACIA PENNATA, Willd. Sp. Pl. iv. 1090. Below Ruchigga. Fl. December, 498. “ Shrub climbing by thorns, white inflorescences.” Distrib. Widely distributed in Tropical Africa. The following is allied, but has smaller pods :— Bank of River Kagera, near Mulema. Fl. May, 289. Native name * Mutinga.” “ Dug-out canoes made of this. Tree with white flowers and green stem." A. SIEBERIANA, DC. Prodr. ii. 463, forma. Near Kagera, Mulema. Fl. April, 216 & 258. Native names * Mutiaza," * Mwezameno.” ‘“ Dug-out canoes and planks made of this. Tree with white flowers." Distrib. Widely distributed in Tropical Atrica. A. SEYAL, Delile, Fl. Agypt. 216, tab. 52. f. 2 Near Mulema. Fl. March, 188. Native name “ Nigandu.” * Bark used as rope. Shrub. Inflorescence green.” Distrib. Widely distributed in Tropical Africa, A. STENOCARPA, Hochst. ex A. Rich. Tent. Fl. Abyss. 1. 238. Hillside, Mulema. Fl. May, 317. Native name * Musange." “Tree assuming an umbrella- shape; used in building huts. Inflorescence white.” Distrib. Nile Land. The following is allied to A. hebeclada, DC. :— Near Mulema. Fl. March, 189. Native name ** Mutongde.” ‘‘ Wood used for posts and handles of hoes; bark used as rope. A tree; flowers white.” ALBIZZIA VERSICOLOR, Welw., in Oliver, Fl. Trop. Afr. ii. 359. Gully near Mulema. Fr. May, 301. Native name * Mubula.” ‘ Vessels made to hold milk of this. Tree; no flowers." A. BRACHYCALYX, Oliver, Fl. Trop. Afr. ii. 361. Near Mulema. FI. April, 219. * Native name * Mumeya." Used for posts for houses and charcoal. A tree ; flowers white; staminal tube red." Distrib. Nile Land. E. G. BAKER: DICOTYLEDONES POLYPETALEXE. 151 ALBIzZIA Brownel, Walp. Rep. i. 928. Island of Buvüma. Fl. March, 627. * Tree with white petals and red staminal tube. Distrib. Upper and Lower Guinea. ” A. FAsTIGIATA, Oliver, Fl. Trop. Afr. ii. 361. Hillside, Mulema. Fl. April, 207. Grunga. Fl. November, 383. Native name *Munyanohinga." Wood used for handles for hoes and charcoal. A flat-topped tree with whitish flowers.” Distrib. Upper and Lower Guinea. Also occurs in Natal. RosacEm®. PARINARIUM CURATELLEFOLIUM, Planch. in Niger Flora, 333 ; forma. Near Mulema. Fl. April, 208. Native name * Munabiliko." ** Medicine for gonorrhea made from the leaves. A low tree. Petals light blue but fugitive." Distrib. Upper Guinea, Nile Land, Mozambique District. Ruzsvs APETALUS, Poir. Encycl. Méth. vi. 242. Ruchigga. Fl. November, 422. * Climber. Flower with pink stigmas, green perianth." No. 520, from the same locality, is also a species of Rubus. Distrib. Nile Land, Mozambique District. CRASSULACE. KALANCHOE CRENATA, Haw. Syn, 109. Musozi. Fl. January, 116. Distrib. Widely spread in Tropical Africa; also at the Cape. K. STUHLMANNI, Engler, in Pflanzenwelt Ost-Afr. C. 188; var. ? Near Kikobe ferry, River Kagera. Fl. March, 180. Native name * Kyondo.” “ Planted by natives amongst the bananas. White flower." 152 BOTANY OF THE UGANDA BOUNDARY COMMISSION. Succulent herb. This differs in certain respects from Engler’s description, but I have not had an opportunity of comparing it with authentic material. Distrib. Central African Lake-Region. COMBRETACES. COMBRETUM SPLENDENS, Engler, Pflanzenwelt Ost-Afr. C. 289 (1895). Near Mulema. Fl. April, 212. * Shrub with small white flowers." The following is closely allied :— Bank of River Kagera, near Mulema. Fl. May, 290. Native name “ Muramma.” Tree with greenish flowers. Distrib. Central African Lake-Region, Mozambique District. C. BUVUMENSE, sp. nov. Frutex ramis adultioribus glabris, novellis + fusco-pubes- centibus, foliorum petiolo superne canaliculato, lamina membranacea, ovata aeuminata, superne glabra vel glabri- uscula, basi late euneata vel rotundata, nervis lateralibus patulis adscendentibus ; racemis axillaribus, copiose floriferis, abbreviatis; floribus inter mediocres generis breviter pedicellatis, 5-meris, receptaculo inferioretenuiter fusco-pubescente, superiore cylindrico-campaniforme, medio haud constricto ; calycis segmentis zequilateraliter triangularibus, acuminatis; petalis oblongo-ovatis, basi cordatis, margine ciliatis, longioribus quam latis, basi cordatis. Species ad C. paniculatam, Vent., accedens. Island of Buvüma, Victoria Nyanza. Fl. March 1904. 624. A member of Engler & Diels's Section Conniventes. Folia ad 7:5 em. longa an longiora. Petiolus sæpe 7-8 mm. longus. Racemi 1:5-20 cm. longi. Receptaculum superius + 6:5 mm. longum. Petala + 3 mm. longa. Stamina quam petala 3-4plo longiora. ILLIGERA PENTAPHYLLA, Welw. in Trans. Linn. Soc. xxvii. (1869) 206. Island of Buváma. Fl. March. 601. “Liane with pink petals and orange pollen.” 619. * Liane with greenish flowers.” Distrib. Lower Guinea. E. G. BAKER: DICOTYLEDONES POLY PETALA, MYRTACE.E. Evaenia ovixEENsIS, Hiern, Welw. Cat. i. 359. Island of Buvima, Fl. March, 614. “Tree with white flowers; colour due to stamens.” Distrib, Upper and Lower Guinea. E. CORDATUM, JM. A. Laws. in Oliver, Fl. Trop. Afr. ii. 438. Musozi. Fl. & Fr. December, 5. 153 Native name “ Kirungi nsamvu." Burnt to give a pleasant smell to bark-cloth.” Distrib. Lower Guinea, Mozambique District ; also Natal. There is another specimen of Eugenia in the collection from Musozi, 154. Native name “ Ntukuza," evidently allied to E. bukobensis, Engler, but the material is insufficient for identification. MELASTOMACE®. Dissoris MINOR, Gilg, Mon. Afric. Melastom. 12, t. 2 c. Musozi. Fl. January, 86. * Low shrub with a red flower." Distrib. Uganda. D. InvixaraNA, Hook. fil. in Bot. Mag. t. 5149. in gully above Mulema. FI. May, 310. Ruchigga. December, 480. Native name “ Nkurutetabi.” “Erect herb 2-23 Flowers blue or purple.” Distrib. Upper Guinea. MEMECYLON HETEROPITYLLUM, Gilg, Mon. Melastoim. 39. Island of Buvima. Fl. March, 608. ** Shrub with blue flowers." Distrib. Central Atrican Lake-Region. LYTHRACEX. RorALA BREVISTYLA, 8p. nov. Planta natans caulibus crassicusculis simplicibus, inferne spe radicantibus, glabriuseulis, precipue extremitates versus foliosis, foliis internodiis subszquilongis, e basi cordatis ovatis vel oblongo-ovatis, sessilibus, oppositis, glabris, obscure penninerviis, apice obtusis; floribus minimis axillaribus, subsessilibus, foliis floralibus semper 154 BOTANY OF THE UGANDA BOUNDARY COMMISSION. multoties brevioribus, calyce demum campanulato, vix ad medium 4-lobato, lobis brevibus acutis quam tubo brevi- oribus; petalis minutis, anguste spathulatis; staminibus 2 inclusis, stylo brevi; capsulis globosis, 4-valvatis, demum quam calyce longioribus. Hab. Mulema. * Floating plant; flowers minute, red." Fl. & Fr. May, 316. Caules 15-20 em. longi, internodiis 6-10 mm. longis. Folia 7-11 mm. longa, 5-7:5 mm. lata. Calycis lobi + °5 mm. longi. Capsula + 1°5 mm. longa. Allied to R. fontinalis, Hiern; differs from that species in the stem being longer and unbranched—in there being 2 stamens included instead of 4 exserted, &c. AMMANNIA BACCIFERA, Linn. Sp. Pl. 120. River Rufüa. Fl. December, 535. “ Erect herb in 2 in. of water; pink petals.” Distrib. Nile Land, Lower Guinea. Also in India and Madagascar. ONAGRARIEX. EPILOBIUM HIRSUTUM, Linn. Sp. Pl. 347. In swamp near River Rufüa. Fl. December, 537. Distrib. Widely spread in Africa and elsewhere. Jussi#a LINIFOLIA, Vahl, Eclog. Amer. 32. R. Rufta. Fl. December, 538. * Marsh plant 5 ft. high ; yellow flowers." Distrib. Widely distributed in Tropical Africa. Also in Tropical America. J. DIFFUSA, Forsk. Fl. ZEgypt.- Arab. 210. Poo. by Lake Musozi. December, S. Water-plant. Distrib. Upper Guinea, Nile Land, Mozambique District. SAMYDACEEX. TRIMERIA MACROPHYLLA, sp. nov. (Pl. 1.) Frutex inermis, ramis junioribus pubesceutibus, foliis crassi- usculis ovatis vel ovato-ellipticis, apice szpissime acutis, rarissime obtusis, basi cordatis, margine serratis, subtus pilis albidis vestitis, adultioribus superne glabris, sub- palmatinerviis; petiolis cinereo-pubescentibus, stipulis E. G. BAKER: DICOTYLEDONES POLYPETAIL.. 155 magnis deciduis; floribus albis dioicis, spicis masculinis simplieibus, quam folia multoties brevioribus sed quam petiolis longioribus; floribus in glomerulos aggregatis; d brevissime pedunculatis, tetrameris vel peutameris, sepalis minutis lanceolatis, extus pilis obtectis; petalis sepalis sub- similibus paulo majoribus, in fl. tetramero ; staminibus 12 inter glandulas 3-nis insertis, quam petalis longioribus ; floribus 9 capsulis coriaceis, stylis 3 gracilibus, stigmate parvo. Species T. tropice, Burkill, valde affinis, differt foliis majori- bus adultis superne glabris basi magis cordatis, filamentis paullo longioribus. Hab. Near Mulema. Fl. & Fr. June, 346. * Shrub with whitish inflorescence.” Folia 7-0-12°5 em. longa, 5:5-9:50 em. lata. —Petiolus 1'0- 15 em. longus. Spice masculini 4:5-8:5 em. long. Capsula 3-3°5 mm. longa. Tv RNERACEE. WORMSKIOLDIA PILOSA, Schweinf., var. a ANGUSTIFOLIA, Urban, Mon. Turner. 54. Island of Wema. Fl. March, 590. Herb with orange-yellow flowers. Distrib. of type. Upper Guinea. PASSIFLOREX. BARTERIA ACUMINATA, Sp. nov. Arbor humilis vel frutex ramulis striatis tenuiter rufescenti- puberulis, an demum glabrescentibus ; foliis oblongis vel oblongo-ellipticis, coriaceis, fere glabris, apice acuminatis, basi in petiolo attenuatis, petiolo brevissimo, crasso, non stipulato, decurrente, lamina nerviis supra et infra circ. 16-19 ante marginem inter se arcuatim conjunctis, nerva- tione reticulata subtus prominula ; floribus 1-2 axillaribus sessilibus, basi bracteatis, bracteis numerosis arcte imbrieatis cupuliformibus, brunneis, nitidis, margine ciliatis ; sepalis 5 ovato-oblongis, acuminatis, quam petalis longioribus, basi coalitis; petalis albis oblongis, apice mucronatis; staminibus numerosis; stigmate maximo conico-globoso flavo; fructu globoso. 156 BOTANY OF THE UGANDA BOUNDARY COMMISSION. Species Barterie nigritane, Hook. fil., affinis, differt foliis apice sensim acuminatis. Hab. Lake shore, Musozi. Fl. January. 93. “ Shrub or tree. Flowers white, with large yellow stigma. Leaves two-ranked.” Folia 22-94 cm. longa, 6'0-7'0 em. lata. Petiolus X 6'0- 8:0 mm. longus, supra canaliculatus. Sepala 2:5-8:0 em. longa, 10-11 mm. lata. Anthere + 3 mm. longs. ADENIA VENENATA, Forsk. Fl. ZEgypt.-Arab. 77. Mulema, close to a hut. Fl. April, 239. Native name “ Numanyama." ‘A shrub; flowers greenish ; 2-3 ft. high only, but trunk 1 ft. diam. just above ground." CUCURBITACEE. Mxrorunra puncrara, Cogn. in DU. Mon. Phan. iii. 615. Near Mulema. Fl. June, 361. “ Herbaceous twiner: d only. White flowers.” Distrib. Widely spread in Tropical Asia. DEGONIACE E. Begonia EMINI, Warb. in Engler § Prantl, Natiirl. Pflanzen- Jam. il. 6a, 141. Lake shore, Musozi. Fl. December, 2. The following is elosely allied :— In a wood near the mouth of Kagera. Fl. February, 564. “ Trailing plant with white flowers." FrcorpE. Morrvao SPERGULA, Linn. Syst. ed. X. 881. Island of Wema. Fl. March, 591. “ Prostrate herb with white flowers. Entire inflorescence in some cases red.” Distrib. Widely distributed iu Tropical Africa. UMBELLIFERS. HETEROMORPHA ARBORESCENS, Cham. 5 Schlecht. in Linnea, i. 385 Near Mulema. Fl. June, 348. * Shrub 10 ft. high, with inconspicuous flowers." Distrib. Nile Land, Mozambique District. SPENCER MOORE: GAMOPETAL®. 157 PEUCEDANUM FmaxINIFOLIUM, Hiern, in Oliver, Trans. Linn. Soc. xxix. (1873) 79. Near Mulema. June, 334. “Shrub 18 ft. high, with fleshy stem." Distrib. Nile Land, Lower Guinea. Torts GRACILIS, Engler, Pflanzenwelt Ost-Afr. C. 301. Ruchigga. Fl. November, 413. “ Herb with yellowish flowers.” Distrib. Mozambique District. DICOTYLEDONES GAMOPETALÆ. (By Spencer Moore, F.L.S.) RUBIACER. HyMenopictyon Kunnra, Hochst. in Flora, xxvi. (1843) 71. In gulley above Mulema, also Ruchigga. Fl. December, Fr. May, 313 & 485. Native name * Mwamira." Distrib, Nile Land, Mozambique District, Upper and Lower Guinea. NEvROCARP SA LONGIFLORA, S. Moore. (Pentas longiflora, Oliver, in Trans. Linn. Soc. ser. II. Bot. ii. (1887) 335.) Ruchigga. Fl. November, 406. Distrib. Nile Land, Mozambique District. N. Tuomsonu, S. Moore. (Pentas Thomsonii, S. Elliot, in Journ. Linn. Soc., Bot. xxxu. (1896) 435.) Ruchigga, 7500 ft. Fl. December, 473. Distrib. Nile Land. N. PURPUREA, Hiern, Welw. Cat.1. 438. Gorge near Mulema. Fl. April, 247. Distrib. Nile Land, Mozambique District, Lower Guinea. OLDENLANDIA HERBACEA, Roxb. Hort. Beng. 11. Hill-top near Mulema. Fl. & Fr. June, 338. Widely distributed. O. ABYSSINICA, Hiern, in Oliver, Fl. Trop. Afr. ii. 57. Hill-top near Mulema. Fl. June, 336. Distrib. African tropics except Lower Guinea and eastern part of Mozambique District. 158 BOTANY OF THE UGANDA BOUNDARY COMMISSION. Muss£npa ARCUATA, Poir. in Lam. Encye. i. 392. Musozi. Fl. June, 89. Native name “Toke Kuru.” * Used as medicine, leaves pounded with water and drunk for pains in abdomen." Widely distributed: also a Mascarene species. DICTYANDRA ARBORESCENS, Welw. ex Benth. § Hook. f. Gen. Pl. ii. 77. Uganda, shore of Lake Victoria Nyanza at Mutunda. FL Mareh, 578. Distrib. Upper and Lower Guinea. TARENNA AFFINIS, S. Moore. (Chomelia affinis, A. Schum. in Engl. Pflanzenw. Ost-Afr. C. 380.) Musozi. Fl. January, 78. “Shrub with showy white inflorescences.’ * Omwanimwani.” Distrib. Mozambique District (Usambara). > Native name Ranpra Macunata, DC. Prod. iv. 388. Island of Buvüma, Lake Victoria Nyanza. Fl. March, 630. Distrib. Upper and Lower Guinea, North Central. GARDENIA THUNBERGIA, Linn. f. ew Thunb. Diss. Gard. 11. Hillside, Mulema. Fl. April, 227. Native name * Ntalama." —** Used for fences to keep away lions, &c." Widely distributed. G. URCELLIFORMIS, Hiern, in Oliver, Fl. Trop. Afr. iii. 104. Uganda, shore of Lake Victoria Nyanza at Mbazi. Fl. March, 575. Distrib. Nile Land (Niam-Niam). The ripe fruits are globose, and two inches in diameter. G. VISCIDISSUMA, Sp. 110v. Arbuscula resiniflua aspectu G. physophylle, De Wildem. (Randie physophylle, K. Schum.), nisi omnimodo multo minor, folis ovatis obtusissimis basi obtusis vel in petiolum brevem cuneatim augustatis integris tenuiter coriaceis costa centrali subtus puberula exempta et axillis costularum pubescentibus glabris, calyce parvo impariter 8-lobo lobis lanceolatis acutiusculis, floribus solitariis terminalibus, corolla mediocri erassiuscula superne gra- datim amplificata fere adusque j iu lobos 5 ovatos SPENCER MOORE: GAMOPETALE. 159 obtusissimos divisa, antheris angustis sursum curvatis, fructu ambitu oblongo subtereti coriaceo, seminibus in massam unicam confertis. Hab. Edge of copse on hillside, Musozi. Fl. February, 144. Folia 9:0-17:0 cm. long., 6:0-fere 12:0 cm. lat., nitida, subter castanea, supra fusca ibique resina copiosissime obducta. Stipule 0:5 cm. long. Pedunculus 077 cm. long., ut calyx resinosus. Flores lactei, suaveolentes. Calyx 0'8 em., lobi cirea 0:2 em. long. Corolle extus scabriuscule 11:0 em. long., tubus inferne 0:3-0*4 em. diam., faucibus fere 3"0 cm.; lobi 0°25 em. long. Anthere paullo ultra 2-0 em. long. Stylus breviter exsertus, sursum clavato- oblongus. Fructus 2:5 em. diam., vix 4'5 em. long. Nervation of leaf exactly like that of G. physophyila, but the leaves, besides being much smaller, are very obtuse at the top. The calyx and corolla are also very much smaller than those of the species just cited. The 8-lobed calyx and longer anthers may be mentioned among other points of difference. Native name * Mugondo.” ‘Wood used for posts." OXYANTHUS LITOREUS, sp. nov. O, alte fruticosus ramis validis saltem prope nodos applanatis crasse pubescentibus, foliis magnis elliptieis obtusissimis basi valde obliquis neenon cordatis margine undulatis subcoriaceis in fac. sup. prsesertim secus costam centralem scabridis subtus pr:esertim secus costam centralem et costulas hispidulo-pubescentibus, petiolis brevibus crassis hispidulo-pubescentibus, stipulis lanceolato-oblongis acutis extus hispidulo- pubescentibus, cymis plurifloris quam folia brevioribus, bracteis paucis parvis setaceis, floribus mediocribus pedicellatis, calycis limbo adusque ? diviso lobis subulato-setaceis piloso-hispidulis, eorolla sat elon- gata verisimiliter hypocrateriformi laciniis angustis, ovario 2-loculari multiovulato albo-piloso. Hab. Shore of Lake Victoria Nyanza at Musozi. FI. Jauuary, 95. Folia ex schedis cl. detectoris 40:0x 25:0 em. attingunt, ea mihi obvia 22:0 em. long. et 12:0-14-0 em. lat. metiuntur: petioli 1:0 em. long. Stipule 2:0 cm. long. Pedicelli 0'2-0'5 em. long. Bractez 0:1-0:25 em. long. Cyme 5'0 em. deinde 8'0 em. diam., griseo-puberule. Flores 160 BOTANY OF THE UGANDA BOUNDARY COMMISSION. vireseentes. Calycis limbus 0'2 cm., lobi O'G em. long. Corolle nondum expanse 10:0 em., anthere 0°65 em. long. Stigma fusiforme, 0°65 em. long., cruribus O'L em. long. inelusis. Ovarium oblongum, 0:5 em. long. The material of this consists of two leaves with the part of the stem to which they are immediately attached, and two separated inflorescences—one with buds in various stages but no fully expanded flower, the other with very young fruits just after the corollas have fallen away. The length of the inflores- cenees cannot therefore be given, nor that of the flowers at maturity. In spite of this, I have ventured to describe the specimen, as it seems different from any hitherto known. In appearance it is much like O. wnilocularis, Hiern, but the clothing of the leaf, the longer calyx-lobes, the 2-celled ovary, and, to judge from the unopened corollas, the smaller flowers are well-marked differential characters. OXYANTHUS LEPIDUS, Sp. nov. O. fruticosus, glaber, ramulis aliquantulum tortis necnon applanatis olivaceis, foliis brevipetiolatis ovatis cuspidato- aeutiuseulis basi rotundatis chartaceis costis secundariis utrinque circa 10 summis parum aspectabilibus retis sed marginem versus abrupte arcuatis, stipulis a basi lata in appendicem longam lineari-setaceam exeuntibus, corymbis abbreviatis paucifloris, bracteis lanceolatis acuminatis margine ciliolatis pedicellos zquantibus, calycis limbo ad 4 in lobos triangulari-subulatos diviso, corolle tubo mediocriter elongato quam lobi lineares obtusi multo longiore extus glabro intus deorsum piloso sursum fauces versus transversim rugoso, filamentis perbrevibus, antheris anguste linearibus, ovario 2-loculari, stigmate anguste fusiformi. Hab. Wema Island, Lake Victoria Nyanza. Fl. March, 594. Folia 10-12 x 5—6'5 cm., supra subnitentia, in sicco olivacea, subtus pallida; petioli 0'5 em. long., puberuli. Stipulæ extus puberulæ, intus basi pilosæ, pars expansa 0'3- 04 cm., pars attenuata 0'7-1'0 em. long. vel paullulum majus. Inflorescentie axis robustus, summum modo 0'5 em. long. Bractew et pedicelli 0:3 cm. long. Calyx totus 0*4 em. long. Corollæ albe», tubus 8:0-8 5 em. long.; lobi 15 em. long., 0'3 em. lat. Anthere 04 cm., SPENCER MOORE: GAMOPETAL.E. 161 ovarium 0'25 em. long. Stylus glaber, ad 12 em. exsertus. Stigma 04 em. long. Known by the comparatively short leaves rounded below, together with the stipules terminating in a long slender appen- daze, the congested few-flowered inflorescenees, tho narrow lobes of the corolla, and the relative length to them of the tube. Tricatysta BUXIFOLIA, Hiern, in Fl. Trop. Afr. iii. 119. Near Mulema. Fl. & Fr. June, 352. Distrib. Lower Guinea. CANTHIUM SCHIMPERIANUM, A. Rich. Tent. Fl. Abyss. i. 350. Kagera Valley, near Mulema. FI. May, 282. Native name * Mukiraga.”’ Distrib. Nile Land, Mozambique Distriet (Usambara). C. HISPIDUM, Benth. in Hook. Niger Fl. 409. Mouth of River Kagera. Fl. December, 54. Distrib. Nile Land, Mozambique Distriet, Upper and Lower Guinea. C. GoLUNGENSE, Hiern, Welw. Cat. i. 478; var. PARVIFLORA, var. nov.