THE JOURNAL OF THE LINNEAN SOCIETY. BOTANY. VOL. XXX. LONDON: SOLD AT THE SOCIETY'S APARTMENTS, BURLINGTON HOUSE, PICCADILLY, W., AND BY LONGMANS, GREEN, AND CO., AND WILLIAMS AND NORGATE. 1895. [Sag eurin frou Aotea] Dates of Publication of the several Numbers included in this Volume. No. 205, pp. 1-63, published August 28, 1893. » 26, , 63-142, ” February 1, 1894. », 207, ,, 148-217, ” March 19, 1894. » 208, ,, 217-280, ,, July 10, 1894. » 209, ,, 281-330, ,, October 6, 1894. » 210, ,, 331-435, ” February 5, 1895. 4 » 211, ,, 486-561, ,, September 7, 1895. PRINTED BY TAYLOR AND FRANCIS, | RED LION COURT, FLERT STREET. | LIST OF PAPERS. Page BornMUt yer, J. Recent Botanical Exploration in Southern Persia, being the substance of a from J. Bornmiiller to Dr. Otto Stapf. (Communicated by the Secretary, B. Daydon Jackson.) .... 140 BreBNER, GEORGE, late Marshall Scholar in Biological Research, Royal College of Science, London. On the Origin of the Filamentous Thallus of Dumontia jili- formis. (Communicated by George M. Murray, F.L.S.) (Plates XXXV.& XXXVI)... eee eee eee eee 436 On the Mucilage-Canals of the Marattiacew. (Communicated by Prof. D. H. Scott, M.A., Ph.D., F.L.S.) (Plate XX XVII.) 444 CLARKE, C. B., M.A., F.LS. On certain Authentic Cyperaceze of Linneeus................ 299 Drovery, CHar.eEs T,, F.L.S. Notes upon Apospory in a Form of Scvlopendrium vulgare, var. crispum, and a new Aposporous Athyrium; also an additional phase of Aposporous Development in Lastrea pseudo-mas, var. cristata. (Plate XVIT.) oo... co cece eee eee eee 281 Exutot, G. F. Scort, M.A., F.L.S. On the Botavical Results of the Sierra Leone Boundary Com- MISSION .... ee ee ee eee ee eet eee eee e tee eeeeeene 64 Ewart, Miss M. F., B.Sc. Notes on Abnormal Cypripedium Flowers. (Communicated by Prof. F. W. Oliver, D.Sc., F.L.S.) (Plates I. & UH.) ...... 45 Page Guppy, H. B., M.B. On the Habits of Lemna minor, L. gibba, and L. polyrhiza (Communicated by W. Botting Hemsley, F.R.S., A.L.S.) .. 323 Hemstry, W. Borttina, F.R.S., A.LS. On two small Collections of Dried Plants from Tibet. With an Introductory Note by Lieut.-Gen. R. Strachey, C.S.L, F.R.S., FLAS. (Plates IV. & V.) ciee ccc e cece cee eee ete eens 101 The Flora of the Tonga or Friendly Islands, with Descriptions of and Notes on some New or Remarkable Plants, partly from the Solomon Islands. (Plates IX.-XI.) ................ 158 Henstow, Rey. Georas, M.A., F.L.S. The Origin of Plant-Structures by Self-Adaptation to the En- vironment, exemplified by Desert or Xerophbilous Plants. (Plate XIL.)... ccc cece eee rece tee e ee ee tes 218 Lussock, The Right Hon. Sir Joun, Bart., M.P., F.R.S., F.L.S. On Stipules, their Forms and Functions—Part II. (With 7 woodcuts.) ........... eee e eter e eee eet ee ee eeees 463 MARTELLI, Signor U. On the Cause of the Fall of the Corolla in Verbascum. (Com- municated by the President, Prof. Charles Stewart.) (With SB WOOdCUIS.) oe cece e eee erect tere e erect tere seeeens 316 Masters, MaxweE 1 T., M.D., F.RS., F.LS., &e. Notes on the Genera of Tavacee and Conifer@ ....... 0.00005 1 Mt .trr, Dr, J. Thelotremex et Graphidew nove quas presertim ex hb. Reg. Kewensi exponit. (Contributed with the sanction of the Director, Royal Gardens, Kew.) ......... 0... ce cece ee eee 451 PLowricHt, C. B., M.D., and W. Tuomson. On the Life-history of the Acidium on Paris quadrifolia. (Communicated by the President, Prof. Charles Stewart.) .. 43 REnpDiy, A. B., M.A., B.Se., F.LS. Revision of the Genus Mipadites, Bowerb. (Plates VI. & VIL.) 143 A Contribution to the Flora of Eastern Tropical Africa. (Plates XXXI-XXXIV,) oo cec eee eee ceeeeeues 373 is WN Nitin tocbibeaintnttba eins Page . Situ, AnNre Lorrain. On the Anatomy of a Plant from Senegambia. (Communicated by G. F. Scott Elliot, M.A., F.L.S.) (Plate VIIL)........ 155 Spruce, RicHarp. Hepatice Elliottiane, insulis Antillanis Sti. Vincentii et Domi- nica a clar. W. R. Elliott, annis 1891-92, lectee. (Plates XX.- DO. ©: ©) pp 331 Stapr, Dr. Otto. Recent Botanical Exploration in Southern Persia, being the substance of a Letter from Mr. J. Bornmiiller. (Communi- cated by the Secretary, B. Daydon Jackson) .............. 140 Tuomson, W., and C. B. PLowrieurt, M.D. On the Life-history of the Kcidium on Paris quadrtfolia. (Communicated by the President, Prof. Charles Stewart.) .. 43 West, Wo., F.L.S., assisted by G. 8S. Wxsrt. On some Freshwater Alge from the West Indies. (Plates XTII.- DOA 0) rn 264 Wis, J. C., M.A., late Frank Smart Student in Botany of Gon- ville and Caius College, Cambridge. Contributions to the Natural History of the Flower.—Part I. Fertilization of Claytonia, Phacelia, and Monarda. (Commu- nicated by Francis Darwin, F.R.S., F.L.S.) (Plate II.) .. 51 Part II. Fertilization Methods of Various Flowers; Cleistogamy in Salvia Verbenaca. (Communicated by Francis Darwin, F.R.S., F.L.S.) (Plates XVII. & XIX.) ............06.. 284 EXPLANATION OF THE PLATES. Pate : IL | ABNORMAL CypRivepiIuM FLowsrs. II. Fsrriization of Claytonia, Phacelia, and Monarda. IV. { CreEpPIs SOROCEPHALA, Hemsl. Saussurea Tuoroip!, Hemal. VY. Saussurea Aster, Hemsl. VI. Nipapites Burrini, Brongn., from Sussex. VII { NIPADITES LANCEOLATUS, Bowerb. N. Parxrinsonis, Bowerb. VIII. Nuzw Puanr from Senegambia. IX. DysoxyitumM meGALaANtaum, Hemsl. X. Crossostyiis Comins, Hemsl. XI, Sararanaa sinvosa, Hemsl. XII. Onicin or Piant-Srructurss. ° XIII. } XIV. | XV. 4 Fresnwater ALG#& of West Indies. XVI. (XVII. Apvospory in Scolopendrium vulgare, XVIII. Bropi#a rxiomwes; STANHOPEA TIGRINA. XIX. Prwevea, Cotyte ron, Hyprouna, Satvia, Viona. Xx j Prurvanta SPATULIFLORA, Spruce. | LesEUNEA ATROVIRIDIS, Spruce. LEJEUNEA VULCANICA, Spruce. XXI.4 L. euapavurensis, Gottsche. L. BREVINERVIS, Spruce. L&JEUNEA GRAMINICOLOR, Spruce. {r. CORYNANTHA, Spruce. Leseunea Evxiorri, Spruce. {r. PILILOBA, Spruce. { LrsEuNEA pissecTA, Spruce. L, neTEROMORPHA, Spruce. XXv. { Seip SUBSIMFLEX, Steph. C { anaaeenaes’ XXII. XXIII. XIV. Seypryera Exziorti, Spruce. CrrPHaLoziaA WRIGHTII, var. BICORNIS, Spruce. ALOBIELLA DOMINICENSIS, Spruce. MICROPTERYGIUM PORTORICENSE, Steph. Leloscyruus ovatus, Spruce. PiaciocniLa Exviorru, Spruce. ¥ SYzYGIELLA PERFOLIATA, Var. LAVIGATA, Spruce. XXVI. XXVII. XXVIII. i aio Vil PLATE XXIX, { TUsonRaANNTA DOMINICENSIS, Spruce. J. scuistopniia, Spruce. { Scatra ANDINA, Spruce. | SyMpnyoGyNa TRIVITTATA, Spruce. - La > Nurs, Rendle. XXXII. { JAROSIPHON TENUIS, Rendle L. crispus, Rendle. i LAGAROSIPHON HYDRILLOIDES, Rendle. XXXII.) Disperis KILIMANJARICA, Rendle. | ANaracus FIMBRIATUM, Rendle. XXXIII. Anrnericum speciosum, Rendle. - PANDANUS RABAIENSIS ndle. XXXIV. { ; » Re ANEILEMA CLARKEI, Rendle. XXX. ; XXXV. \ Finamentous TuHatuus of Dumontia filiformis. XXXVI. J XXXVII. Mvcinacr-Canaus of MARATTIACES. THE JOURNAL OF THE LINNEAN SOCIETY. Notes on the Genera of Taxacee and Conifere. By Maxwett T. Masters, M.D., F.R.S., F.L.S., Corresponding Member of the Institute of France. [Read 15th December, 1892.] Tux following observations are the outcome of a comparative examination of the morphological characters of all the genera of Taxacee and Conifere so far as I have been able to accomplish it. In most cases living plants have been examined, and in all instances the available museum and herbarium specimens have been studied and the literature relating to them referred to. It is not necessary to state the sources whence I have derived my information, as they are given in my paper on the Comparative Morphology of the Order, published in the Journal of this Society, Botany, vol. xxvii., and again in a List of Conifers and Taxads in cultivation in Great Britain, prepared for the Conifer Congress, and published in vol. xiv. of the Journal of the Royal Horticultural Society in October 1892. Constant reference has also been made to the schemes of arrangement proposed by the older writers, and in more recent times by Eichler, Van Tieghem, and others. In the main, however, I have followed the lines laid down by Bentham, in Bentham and Hooker’s ‘Genera Plantarum,’ iii. p. 420 et seg. (1880). In LINN. JOURN.—BOTANY, VOL. XXX. B 2 DR. M. T. MASTERS ON THE some cases I have felt constrained to deviate from the arrange- ment there made, and to propose others which, according to my judgment, are more in harmony with ascertained facts or with more recently acquired information. Incidentally a few indi- cations of the structural peculiarities of certain parts, which may be available for purposes of distinction, are given, but any attempt to deal fully with comparative histology would be beyond the scope of the present communication. The writings of Ber- trand and of Van Tieghem may be specially pointed out as containing most valuable information on this head. Embryo- logical researches also are of the foremost importance from a classificatory point of view, but they demand for their prose- cution more time and skill than are at the command of the writer. It is, however, to systematic and comparative in- vestigation in these two departments, minute anatomy and embryology, that we must now look for future progress, but that progress will be retarded until the rigidly comparative and systematic methods by means of which the framework of the Natural System has been pieced together shall be carried out also in histology and embryology. And this remark applies as strongly to the methods in which these observations are recorded and published as to the observations themselves. Classifications founded on leaf-structure, on the distribution of the resin- canals and fibro-vascular bundles, or on the structure of the wood even, without reference to all other evidence, are neces- sarily as “artificial” as any other truncated schemes. For details concerning the structure of fossil species the reader will naturally refer to the publications of Williamson, Carruthers, and others. For the indications as to geological distribution I am chiefly indebted to Renault and Starkie Gardner. TAXACEAZ, Lindley, Vegetable Kingdom, ed. 3 (1853), p. 230. Arbores vel frutices. Rami plerumque homomorphi raro dimorphi. Folia persistentia raro decidua. Flores masculi amentiformes. Squame fructifere, ut videtur, simplices, liber, membranaceze seu demum carnosze nunquam lignosee. Ovula erecta v. pendula, e squama emergentes, aril- lata vel raro arillo destituta. Testa sicea seu demum carnosa nunquam alata. GENERA OF TAXACEHX AND CONIFER. 3 Tribus I. SALISBURINEZ. Flores dioici; testa carnosa arillo genuino deficiente vel imperfecte evoluto. Ramuli dimorphi. Folia decidua. Flores masculi umbellati.............. 1. GINKGO, Ramuli homomorphi. Folia persistentia. Flores masculi capitati .............. 2. CEPHALOTAXUS. Flores masculi spicati..............4- 3. TORREYA. Tribus II. TAXINEA, vide p. 6. GINKGO. The earliest mention of this genus is by Kaempfer in 1712, and it was adopted by Linneus, in his ‘ Mantissa,’ ii. 313, in 1771. It is remarkable for its shoots, some of which are elon- gated, others contracted, its fan-shaped, many-ribbed, deciduous leaves, and diccious flowers. The male flowers are stipitate, catkin-like, arranged in umbels on the ends of branches, having short, contracted, or undeveloped internodes and intermixed with leaves. Anther-lobes two, pendulous, divergent, connec- tive prolonged into a knob. Pollen not seen. The female flowers are borne in pairs on the sides of slender stalks. each flower consisting of a single erect. ovule destitute of bract or scale, but arising from a cup-shaped dilatation of the axis with- out any true arillus. The testa of the seed, however, eventualiy becomes succulent and drupe-like, while the inner coat becomes hard and woody, as in Cycas. The seed in the interior is covered with a brown membrane, the lower half of which is adherent to the shell, so that, on removing the latter, the lower part of the perisperm is bared. A similar appearance is presented in the seeds of Cycas. As in the case of Torreya and Cephalotazus, the structure of the ripe seed has a close resemblance to that of Cycas, the nucellus having a well-marked pollen-chamber. According to Van Tieghem, Bull. Soc. Bot. France (1891), there are in the axis of the pith two resin-canals. The same author asserts that fertilization only takes place in the ripe seed after its fall in autumn, and that the embryo is only developed in the following winter. The cotyledons, in germination, remain within the seed, their stalks protruding in an arched manner, the cylindrical tigellum B2 4 . DR. M. T. MASTERS ON THE or hypocotyl descending vertically into the ground, while the plumule ascends, bearing several scattered rudimentary leaves, arranged on the 4 plan before the perfect leaves are formed. Clustered raphides occur in the leaves—a most unusual occur- rence in Conifers. In the anatomical structure of the root the most striking feature is the abundance of callus-thickenings in the angles of the cells of the endoderm. There is a single fibro-vascular bundle. The leaves have no true palisade-cells, the two fibro-vascular bundles of the petiole branch into several which traverse the lamina; each is ovoid in section, with xylem resting on the phloem. Wedge-shaped masses of sclerous tissue occur here and there beneath the epiderm, but no true exoderm. The genus is now represented by a single Chinese and Japanese species, but was once so widely distributed through the colder and temperate parts of the northern hemisphere, that Prof. Heer enumerated no fewer than sixty species in the Secondary and Tertiary strata of various parts of the globe. It seems probable that most of these are mere variations from a single species. CEepHALOTAXUS. A genus first published in 1840 by Siebold and Zuccarini in their ‘Flora Japonica.’ The species are shrubs or trees with spirally arranged leaves, which, on the lateral branches, become twisted into an apparently two-ranked arrangement. The flowers are dicecious, in stalked axillary heads, each flower subtended by a bract. Male flowers numerous, clustered, capitate. Anthers 3-lobed, pendulous from the apex of the filament. Connective prolonged. The female flowers are bracteate and arranged in terminal spikes or heads. The base of each bract becomes more or less fleshy, forming a cup-shaped cavity around a short shoot, which bears one or two erect ovules destitute of aril. The testa of the seed, however, becomes succulent, the inner lining bony, so that the ripe seed resembles a drupe, and is similar to the seed of Cycas or of Ginkgo, as described under the latter genus. In germination the seedling plant shows a thick fusiform tigellum or hypocotyl, two long linear cotyledons, from between which the plumule emerges before the cotyledons are disengaged from the seed-coats, as in Ginkgo, but the primary leaves are opposite and verticillate. GENERA OF TAXACEX AND CONIFERS. 5 The leaf-section (Cephalotaxus drupacea) shows the midrib prominent on the upper surface, a thick epiderm, no exoderm, and a single layer of palisade-cells above the mesophyll. In the centre is the fibro-vascular bundle, in section, of an elliptical shape, with ill-defined endoderm. Immediately beneath the phloem is a single central resin-canal, between which and the lower epiderm two or three rows of cells intervene. The stomata are abundant on the lower surface ; they are of the usual reniform shape, and are arranged in linear series. Van Tieghem notes the presence of a resin-canal in the centre of the pith, a character which separates it from all Conifers except its near ally Ginkgo, which has two in the same situation (Bull. Soc. Bot. France, 1891, April 24). By Bentham and Hooker this genus is placed among the Taxodinee. It is, however, evident that its nearest affinity is with Ginkgo and Torreya, and that it has relatively little in common with Zuzxodinee. The species are natives of China and Japan, extending into Khasia. TORREYA. A genus established by Arnott in 1838, in Ann. Nat. Hist. ser. 1, vol.i. p. 180. The. leaves are spirally disposed, the flowers dicecious. The male flowers are axillary, spike-like, surrounded at the base by imbricate scales ; each stamen has a filament which expands at its upper portion into 2-4 longitudinally dehiscing pollen-sacs ; connective slightly prolonged. The female flowers are solitary, axillary, with imbricating, rather fleshy scales at the base. The erect ovule is gradually invested from below upwards by a fleshy aril, which becomes confluent with the succulent testa, so that in the ripe seed it is no longer distinguishable. Seed drupaceous. Embryo pendulous in floury or ruminate perisperm. The leaf-structure is almost identical with that of Cephalotaxus, but branched sclerous cells occur in the mesophyll, which confer strength in the absence of theexoderm, The fibro- vascular bundle has a semilunar form, and is surrounded py an endoderm with thick-walled cells. The solitary central resin- canal of the leaves resembles that of the species of T'suga. The species are natives of California, South-eastern United States, Japan, and China. In a fossil condition they are stated to occur in beds of the Miocene period. 6 DR. M. T. MASTERS ON THE Tribus IT. TAXINEZ. Flores monoici vel dioici; testa sicca arillata vel exarillata. Ovula semper vel ad postremum erecta. Subtribus 1. Tax 5 #. Ovula exarillata ..........e ee aes seeeee 4, PHEROSPHARA. Ovula arillata. Flores feminei perulati. Ramuli foliaceo-dilatati ............ 5. PHYLLOCLADUS. Ramuli teretes...... ccc. cece cece 6. Taxus. Flores feminei eperulati ............-. 7. DACRYDIUM. Ovula semper vel ad postremum inversa ; testa sicca arillo carnoso circumdata. Subtnibus I]. PopocaRPEA. Pedunculus bracteeque demum carnosx concrescentes ..... os Faewsee ..... 8. Popocarpus. Pedunculus lignosus. Fructus laxe spicati; folia linearia .... 9. Sracnycarpus. Fructus dense aggregati. Folia specie disticha linearia; flores MOMOICL. ... cece eee ee te eee 10. SAXEGOTHAEA. Folia tetrasticha squamiformia; flores C00) cy ll. Microcacurys. PHEROSPH ERA. A genus nearly allied to Dacrydium, established by Archer and Hooker, in Hooker’s ‘Kew Journal of Botany,’ in 1850. The leaves are decussate and spirally imbricate, the flowers dicecious ; male flowers terminal, anthers subsessile, spirally arranged ; connective with a scale-like projection ; pollen sub- globose, somewhat trigonous, or oblong curved, with opaque spaces at the ends or angles. In the female flowers loosely imbricate bracts are alone visible, without any trace of scale externally, but no opportunity of examining the microscopical structure has presented itself. The solitary ovule, at first in- verted, proceeds apparently directly from the base of the bract, which does not become succulent. The seed, which is ultimately erect, is not surrounded by any aril. By Eichler this genus is included under Dacrydium, but the aggregation of the female flowers, the simplicity of their structure, and the absence of an aril, seem to warrant keeping it distinct. The plant is a native of Tasmania and of the Blue Mountains of Victoria (see Hooker's ‘ Icones Plantarum,’ t. 1383). GENERA OF TAXACEX AND CONIFERS. 7 PHYLLOCLADUS. L. C. Richard, in 1826, proposed the establishment of this genus, which has been generally adopted. The species are re- markable for their flattened, entire, or lobed phylloclades. The leaves, from whose axils these organs spring, are small and scale- like. The flowers are monecious or diccious, the males stalked, clustered or solitary, catkin-like, each arising from the axil of a leafy bract. The anthers are imbricate, two-celled, dehiscing lengthwise at the sides; the connective is prolonged into an acute crest. The female flowers are few, or aggregated in close heads along the edges of the phylloclade, each protected by one or more bracts which eventually become fleshy. Ovule solitary, inverted, but subsequently erect, surrounded at the base by a tubular aril which becomes fleshy, but which never entirely con- ceals the erect seed. The two cotyledons are linear, and are followed by primordial leaves of similar form, which soon give place to scale-like leaves. The leaves at the base of the male flowers are broader and strap- like. The leaves have small resin-canals close to the exoderm on the lower surface of the leaf (P. alpinus), and a single central vascular bundle. In the phylloclade of P. alpinus beneath the upper epiderm is a layer of perfect parenchyma, traversed by a central and by numerous lateral fibro-vascular bundles. The species are natives of Tasmania, New Zealand, and the mountains of Borneo; and some species are known to have existed in Nebraska and in Spitzbergen in the Cretaceous epoch (Renault). Taxus. A classical name adopted by Tournefort (1700) and taken up by Linneus (1737) and by all subsequent authors. The leaves are spirally disposed, but so twisted on the side shoots as to appear distichous. On the leader shoots, and in the fastigiate forms, the leaves are not twisted, so that their spiral arrange- ment is always apparent. The flowers are usually diccious, rarely monecious. The male flowers are stipitate, the stipes sur- rounded at the base by imbricating, perular scales, and bearing above a spike of anthers with 4-6 lobes radiating from the under surface of a peltate connective, recalling the appearance of the spore-cases of Hguisetum. The anther-lobes open lengthwise and liberate globose microspores. The female flowers are solitary, & DR. M. T. MASTERS ON THE terminal, and surrounded at the base by numerous imbricate, marcescent bracts. Ovule erect, surrounded by an aril which ultimately forms a fleshy cup in which the ripe seed is nearly enclosed. The two cotyledons are linear, obtuse, leafy, and are succeeded by smaller leaves of the same form. The leaves of 7. baccata have the midrib prominent on the upper surface and the margins are recurved. The upper epi- derm is smooth, the lower papular. There is no hypoderm, but well-marked palisade is present. The vascular bundle is single, elliptic in outline, the pericycle being surrounded by an ill- defined endoderm. There are no resin-canals. The root-fibrils show a loose epiderm with root-hairs, a thick cellular cortex, and a single cylindrical fibro-vascular bundle in which masses of xylem alternate with similar masses of phloem. The pericycle is surrounded by endoderm-cells, the angles of which are much thickened by callus. Bertrand (Ann, Sc. Nat. sér. V. xx. 1874, p. 58) points out the anatomical differences between the several species, especially in the nature of the epiderm and the disposition of the stomates. The species are natives of the northern hemisphere in both continents, while fossil remains in various countries attest the existence of the genus since Miocene times. DacryDIvuM. am name taken up by Lambert (1803) from Solander*. The species have heterotaxic and heteromorphic leaves. The flowers are dicecious, the males at the ends of the branches; anthers sessile, spirally imbricate, two-lobed, dehiscing laterally, and with a prolonged spur-like connective. Pollen-grains oval or oblong. Female flowers solitary, terminal. The uppermost bracts are thickened ; one of them bears an.ovule at first inverted or placed horizontally, but ultimately erect and surrounded at the base by a tubular fleshy aril. The cotyledons are two in number and leaty. The genus is well illustrated in Sir Joseph Hooker’s ‘ Flora of Tasmania,’ and in T. Kirk’s ‘ Forest Flora of New Zealand.’ Lepidothamnus of Philippi is referable to this genus. The species are natives of the Malay Peninsula, Borneo, New Zealand, Tasmania, New Caledonia, and Chile. * Solander in G, Forster, Plant. Esculent. (1786). GENERA OF TAXACEX AND CONIFERS, 9 PopocarPuws. A genus established by L’Héritier in 1788. The leaves vary in attachment and in form. The flowers are monecious or dicecious ; the males solitary on the ends of short axillary shoots, catkin-like, stipitate, stipes surrounded by perular scales. Anthers spirally disposed, sessile, two-lobed, dehiscing longi- tudinally ; connective prolonged into a short point. Pollen- grains with two air-sacs. Female flowers 1 or 2, lateral, stalked, surrounded at the base by afew bracts, which, together with the raphe of the ovule, the peduncle, and the outer coat of the seed, become fleshy. Ovule solitary, anatropous. Cotyledons two, leafy. The leaf-structure has been studied by Bertrand in certain species. In P. Manni and P. chilensis I find a thick exodermal layer (especially so in the first-named), palisade-cells, areolar cells, surrounding an undivided fibro-vascular bundle, beneath which is a solitary resin-canal. The species are natives of tropical and subtropical regions of both hemispheres, and occur also in Tasmania. ‘Traces of their existence occur in the Miocene beds of Central Europe. STACHYCARPUS. Endlicher proposed the name Stachycarpus for a section of the genus Podocarpus, distinguished from the remainder of the genus by the peduncle not becoming fleshy, “ receptaculum carnosum nullum.” In this section he included P. falcata, taxifolia, andina, ferruginea, and spicata. Bentham and Hooker, J. c. p- 435, include only P. spicata and P. andina, Poeppig, in this group. The genus Prumnopitys of Philippi (Linnwa, xxx. 731) is founded on the latter species. Quite recently Van Tieghem has gone a step further ard pro- posed the establishment of a new genus * for species wherein the fruits are on a loose spike, the axis of which does not become fleshy. The distinctions observed in the reproductive organs are correlated with other distinctions in minute structure. Thus, in Stachycarpus there are, according to Van Tieghem, resin-canals in the pericycle of the root outside the phloem of the vascular bundle, as in Araucaria and in Dammara, but not in other Taxinee@, nor in Cupressinea, nor in Abietinee proper. The species are natives of Chile, Australia, and New Zealand. * Van Tieghem, “ Stachycarpus, genre nouveau de la famille des Coniféres,” Bull. Soc. Bot. France, 1891, p. 163. 10 DR. M. T. MASTERS ON THE SaXEGOTHAA. This interesting Chilian genus was founded in 1852 by Lind- ley, who says of it, “ The genus Saxegothea forms a transition of a most remarkable kind, from Conifers to Taxads. Sir William Hooker regarded it as a Podocarp with flowers ina cone. It is in reality a genus with the male flowers of a Podocarp, the female flowers of a Dammara (Agathis), the fruit of a Juniper, the seed of a Dacrydiwm, and the habit of a Yew ” (‘ Vegetable Kingdom,’ p- 229). The foliage and habit are, indeed, those of a Yew, the male flowers are axillary, stalked, catkin-like; the anthers arranged spirally, each opening longitudinally, and destitute of a crest. In the original illustrations, however, such an appendage is distinctly represented. The pollen-cells are globose. The leaves on the cone-bearing branches pass gradually into spirally arranged, but distant scales, which, in their turn, graduate into thick, ovate-lanceolate, loosely and spirally imbricate bracts, simple externally, but each provided near the base with a cavity from the upper end of which hangs a single free ovule. The bracts ultimately become fleshy and consolidated into a more or less pulpy head or spike. In floral structure the genus ap- proaches the Araucarinea, but the habit is more like that of Taxus. Bertrand (in Ann. Se. Nat. sér. V. xx. 187 4, p.69) describes the structure of the leaf as very similar to that of Podocarpus. Microcacueys. Sir Joseph Hooker was the founder of this genus in 1845. It is represented by a Tasmanian shrub with small decussate leaves, which pass gradually at the ends of some of the branches into stamens. The flowers are dicecious, the males in terminal spikes ; anthers stipitulate, two-lobed, dehiscing transversely, and pro- vided with an incurved, scale-like connective. Pollen-grains 3-cornered, or somewhat globose. Female flowers terminal; bracts spirally imbricate, given off at right angles to the axis, apparently simple, bearing an inverted ovule, and ultimately becoming succulent, and as it lengthens carrying the ovule with it, so that the latter becomes placed horizontally. A fleshy, tubular aril grows from below upwards over the greater part of the woody testa of the seed, leaving the apex exposed. Seed 3-sided. It differs from Podocarpus mainly in the form of the pollen- grains, the aggregated fruits, and the woody axis of the spike. From Dacrydium it differs in the aggregated female flowers and in the succulent bract. GENERA OF TAXACEH AND CONIFERE. 11 CONIFER A, Linn. Phil. Bot. p. 28 (plurimum), = PINACE, Lindley, Vegetable Kingdom, p. 226. Arbores seu frutices resiniferze. Flores masculi amentiformes, antheris seu. microsporangiis numerosissimis. Flores feminei strobiliformes. Squame fructiferee duplices e bractea et e squama seu sporophyllo lig- noso rarius carnoso constantes. Bractee nisi ad basin libere vel plus minus cum squama concrescentes. Ovula (seu macrosporangia) erecta vel pendula pauca vel plura. Semina alata vel exalata arillo omnino destituta. Tribus I. CUPRESSINEZ. Folia verticillata seu decussata. Stamina decussata vel ternatim verticillata ; anther loculi globosi. Strobili maturi squamz opposite vel verticillate raro subspiraliter disposite duplices,extrinsecus autem specie simplices. Bractez nisi ad apicem cum squamis concrescentes. Semina erecta. Strobili squamz usque ad apicem concres- centes demum carnose.... Subtribus I. JuNIPERIN«. Strobili squame basi tantum concrescentes, lignescentes. Strobili squame valvate seu verticillatz. Subtribus I]. CALLITRIN &. Strobili squame decussate ...... Subtribus II. Tu urn z. Folia mono- seu sepius dimorpha, opposita vel ternatim verticillata, patentia vel cum ramulis basi concrescentia. Flores mas- culi spicati axillares. Strobili carnosi. Subtribus I. JUNIPERINA. cece cece ceorrcens whee c eee e ee eees 1, JUNIPERUS. Folia mono- vel dimorpha verticillata vel decussata, juventute linearia patentia, adulta cum ramulis compressis seu angu- latis concrescentia. Flores masculi spicati terminales. Strobili lignosi, squame 4-8 verticillata. Ovula pauca vel plura. Subtribus II. CALLITRIN&, Ramuli compressi, strobili solitarii axis haud productus; squame 4 ........ 2, TETRACLINISs. Ramuli angulati, strobili paniculati axis supra squamarum basin productus. Squame 6 ineequales ..........+...+. 3. CALLITRIS. Squamee 8 equales..........++ee004. 4, ACTINOSTROBUS. Squamee 4 wequales ........++.++e0++. 5. WIDDRINGTONIA. 12 DR. M, T. MASTERS ON THE Ramulicompressi vel angulati; folia dimorpha, juniora libera patentia, adulta squami- formia appressa basi plus minus con- crescentia. Strobili squame decussate vel subspiraliter disposite ........ Subtribus II. Tu urn 2. Flores dioici; strobili squamz tenues superiores tantum fertiles .......... 6. Firzroya. Flores monoici; strobili squame incrassate. Strobili squamze basi horizontaliter pa- tentes, apice peltatim expansze, 2- vel plurisperm2® ...... 0. cece cece cece 7. CUPRESSUS. Strobili squamze ascendentes oblonge vel clavato-dilatate. Squame 8-12 plus minus imbricate, semina utrinque eequaliter alata (exalata tamen in § Biota)...... 8. THuya. Squame 6 valvate mediane tantum fertiles; semina apice oblique alata. 9. LiboceDRUus. Tribus II. TAXODINEZ, vide p. 20. Tribus III. ARAUCARINEZ, vide p. 24. Tribus IV. ABIETINES, vide p. 27. JUNIPERUS. A genus known in classical times, mentioned by Pliny and other authors, established botanically by Tournefort in 1700, adopted by Linneus in 1737, and accepted with little or no modification by all subsequent botanists. It is well and sharply defined by its succulent cone-scales, which are confluent one with another at the base. The foliage is polymorphic. The anatomical structure of the free leaves is substantially the same as that of the concrescent leaves. In both cases the stomata are nearly if not quite confined to the upper surface of the leaves. The resin-canals are close beneath the epidermis on the upper face of the leaf. The single fibro-vascular bundle occupies the centre of radiating rows of parenchymatous cells all containing chloro- phyll, and more closely packed in the adnate than in the free leaves. A layer of exoderm cells (hypoderm) surrounds the leaves of both kinds immediately within the epiderm. In Juni- perus drupacea there is one large single resin-canal near the phloem of the fibro-vascular bundle. _. GENERA OF TAXACEEH AND CONIFERS. 13 An examination of the course of development shows that the leaves, as in the case of so-called “ decurrent” leaves, generally are free at the base at first, but subsequently are uplifted with the axis as it lengthens, and are partly concrescent with it instead of becoming detached. The male flowers are catkin-like, axillary or terminal, placed at the end of short shoots, and show no intermediate forms be- tween the foliage-leaves and the sporophylls or stamens. Anther- cells 4-8. The pollen-cells or microspores are globose. The succulent fruits ripen in the second year. Ovules erect, 2 to each scale, wingless. Cotyledons usually 2. Three sections or subgenera have been established, dependent upon the freedom or union of the seeds, the “ decurrence” or isolation of the leaves, and the perulate or eperulate buds. These three sections—Caryocedrus (Endlicher), Oxycedrus (Spach), and Sabina (Spach)—have also met with general accept- ance. Haller’s proposal in 1745 to constitute a distinct genus “ Sabina” has not been adopted; nor has the genus Arceu- thos of Antoine and Kotschy (1854), proposed in order to include J. drupacea (which Endlicher included in his section Caryocedrus), met with any greater degree of recognition. Thuiecarpus of Trautvetter, Plant. Imag. ii. 16, is, judging from the figure, a monstrous form in which the fruit-scales are of a leafy consistence, separate at the tips so as to expose the seeds. Asa malformation by arrest of development it is interesting, and also as showing an intermediate condition between the appearance of the fruit in Juniperus and in Thuya respectively. The distribution of the species extends in the Old World, roughly speaking, from the Arctic regions of Europe and Siberia to Northern Africa, Abyssinia, Arabia, the Himalayas, and China. From West to East the species are found at intervals from West- ern Europe, the Azores and the Canary Isles, to Kamtschatka and Japan. In the New World they occur from the Arctic Circle to Mexico. On the eastern side Juniperus virginiana extends completely from the north to the south of North America. One species is found in the West Indies. Numerically the largest number of species are found in the Mediterranean region and the Levant inclusive of Asia Minor. Next in number are the species found in the Atlantic islands of Madeira, the Azores, and Canaries. Three species are found in 14 MR. M. T. MASTERS ON THE the Himalayas, as many in Japan and China, One inhabits Kamtschatka, and one N.W. America. There are three species in Mexico, four in the Eastern United States, and one in the West Indies. J. communis is distributed throughout the whole of Europe, in Siberia and Kamtschatka, especially in the form called alpina or nana, which occurs also both in Eastern and in Western North America. On the Alps this species extends upwards to 5000 feet. In a fossil condition, according to Renault, Bot. Fossile (1885), p- 185, it occurs in Tertiary strata, in amber &c., but not fre- quently. Hither its remains have not been preserved, or the genus, in spite of the simplicity of its structure, is of relatively late origin. TETRACLINIS. Under the name of Thuya articulata a remarkable Conifer, native of the Barbary States, was mentioned in Shaw’s ‘ Travels in Barbary ’ (1738), and botanically described by Vahl in 1791. It was known to the Romans, who highly valued its timber, which is still used for ornamental purposes. Desfontaines adopted the name given by Vahl, but as the characters of the plant differ in many respects from those of Thuya, it has of late been generally included under the genus Callitris, of which Frenela isa synonym. Richard, Spach, Endlicher, Parlatore, and Ball all attribute the species Callitris quadrivalvis to Vent- enat; but this, as was pointed out by Bentham, is an error, as Ventenat, when founding his genus Callitris upon an Australian evergreen tree, mentioned no species at all byname. Richard is responsible for the name Callitris quadrivalvis, while to Mirbel is to be credited the synonym Frenela Fontanesii. Spach takes it as the type of the genus Callitris in spite of its North-African origin, and Bentham makes it the representative of a section of the genus Callitris under the sectional name Tetraclinis. Tetraclinis articulata, as we prefer to call it, has the articulate stems of Callitris, but differs in its flattened (not triangular), divaricate, Salicornia-like branches, its four-ranked, uniform foliage, and its solitary female cones consisting of four nearly equal decussate or pseudo-verticillate scales, at first somewhat fleshy, and with two equally winged, erect seeds to each scale. Traces of this bract are visible near to the apex of each scale. The GENERA OF TAXACEXH AND CONIFER. 15 axis of the cone is not prolonged beyond the scales as it is in Callitris. Desfontaines, ‘Flora Atlantica,’ t. 252, figures the male flowers as well as the cones. The fossil species from the Tertiary formations named Callitris Brongniarti (Renault, ‘ Cours de Botanique Fossile,’ Coniféres, tab. 15. fig. 9) appears to belong to this genus rather than to Callitris proper. For a summary account of the Gum Sandrac tree of Marocco see Hooker and Ball, ‘Marocco and the Great Atlas,’ Appen- dix D; and Luerssen, ‘ Medicinisch-Pharmaceutische Botanik,’ Band ii. (1882), p. 98, fig. 37. CALLITRIS. A genus established by Ventenat * in 1808 and founded upon a single evergreen Australian tree to which no specific name was attached, but which was probably that now known as C. rhom- boidea. It is synonymous with Frenela of Mirbel, who, in 1826, substituted the name Frenela on account of the similarity of sound between Calilitris and the Myrtaceous genus Calythriz. The genus has been understood in various ways by different authors. Richard, Endlicher, Parlatore, Eichler, Bentham, and others refer the North-African Thuya articulata to this genus, but, for reasons cited under the preceding genus, it seems to me to be preferable to consider the Barbary plant as the representa- tive of a distinct genus (see Tetraclinis). The South-African Widdringtonia (alias Pachylepis of Brongniart) is also often com- prised under this genus, but this also I propose to keep separate. As thus circumscribed Callitris comprises only Australian and New-Caledonian species, but includes the genus Leichartia of Shepherd and the Octoclinis of F. von Mueller. Callitris has its vegetative organs similar to those of its allies, but differs from them in the presence of 6-8 unequal verti- cillate scales to the cones and of numerous seeds to each scale. Unlike what happens in —Flowers yellow. Hill- sides at 16,000 ft. Similar forms, having almost beakless achenes, occur in the Himalayas ; but I certainly think it would be more convenient to separate them specifically, though I would not risk increasing synonymy without a critical examination of all the forms referred to this species. 70. Taraxacum officinale, Wigq., var.—Flowers yellow. Stony wide valleys at 17,200 ft. The Dandelion is recorded in the ‘Flora of British India’ at localities from 1000 to 18,000 ft., and it is now generally dis- tributed in both northern and southern temperate and cold regions. COLLECTIONS OF DRIED PLANTS FROM TIBET. 117 71. Primula tibetica, Watt.—Flowers purple. Close to water at 16,200 ft. Western Tibet north of Kumaon at 15,500 ft., and Eastern Tibet on the Sikkim frontier at 16,000 to 17,000 ft. 72. Androsace Chamajasme, Host, var. coronata, Hook. fil.— Flowers white with yellow centre. Sandy gravelly soil in sheltered spots near water at 17,500 ft. This variety had previously been collected in Western Tibet at altitudes of 16,000 to 17,000 ft. The species ranges all round the northern hemisphere in cold regions. 73. Glaux maritima, Zinn.—At 16,200 ft. North temperate and arctic regions in both hemispheres. 74. Gentiana falcata, Turcz.— Flowers blue. Marsh at 15,000 ft. Siberia, in the Altai and Baical mountains. 75. Gentiana squarrosa, Ledeb.—F lowers blue. Banks of dry rivulet on hill-side at 17,200 ft. Alpine Himalaya, Dahuria, and Siberia to North China. 76. Gentiana decumbens, Zinn.—Flowers greenish white. Sandy soil near water at 15,400 ft. Baltistan eastward to Lahul and through Siberia to Dahuria and Mongolia, 77. Gentiana humilis, Stev.—Flowers whitish blue. Close to water at 16,200 ft. Caucasus to North India, Eastern Siberia, and Dahuria. 78. Pleurogyne diffusa, MJazvim.—Flowers blue. Hill-side close to water at 16,800 ft. ; North Tibet, on the northern declivity of the Burchan Budda chain, in thickets of Myricaria at 10,200 ft. 79. Swertia, sp—*S. Davidii, Franch., affinis.—F lowers blue. Valleys at 15,000 ft. 80. Tretocarya pratensis, Maxrim.—Flowers blue. Earthy stony soil close to streams at 18,000 ft. Western Kansuh at 10,000 ft. 118 MR. W. B. HEMSLEY ON TWO SMALL 81. Microula Benthami, ©. B. Clarke; Hook. Ic. Plant. t. 2257.—Flowers white. Top of pass at 18,000 ft. Western Tibet at 15,C00 to 17,000 ft. 82. Pedicularis alaschanica, M/avim., var. tibetica, Maxim.— Flowers yellow. Broad valley at 16,000 ft. Eastern Himalaya, Mongolia, and Western Kansuh. 83. Pedicularis cheilanthifolia, Schrenk.—Flowers purplish white. Earthy water-logged soil in wide valleys at 17,000 ft. Himalayas, North Tibet, Soongaria, and Western Kansuh. 84. Nepeta longibracteata, Benth.—Flowers blue. Stony soil in old water-course at 17,400 ft. Western Himalaya and Tibet at 14,000 to 17,000 ft. 85. Dracocephalum heterophyllum, Benth.?— Flowers white. Hill-sides at 17,700 ft. This species inhabits Western Tibet and Turkestan at 13,000 to 16,000 ft., and has also been found in Eastern Mongolia. 86. Polygonum sibiricum, Zavm.—Flowers green. Salt- impregnated soil near salt lake at 16,800 ft. Western Himalaya and Tibet to Western China, and north- ward into Arctic Siberia. 87. Stellera Chamejasme, Linn.—F lowers green. Sandy valleys at 15,000 ft. Caucasus to the mountains of North India and Central Asia to Mongolia. 88. Urtica hyperborea, Jacquem.—Flowers greenish white. Rocky hill, amongst stones at 16,200 ft. Eastern and Western Tibet at 12,000 to 17,500 ft. 89. Ephedra Gerardiana, Wall.—F lowers yellow. Salt- impregnated soil close to salt lake, at 16,500 ft. This is united with ZL. vulgaris, Rich., in the ‘ Flora of British India, which thus limited extends from Europe to the moun- tains of North India and Central‘Asia. 90. Iris (§$ Apogon) Thoroldi, Baker. Ad I. humilem, Bieb., magis accedit: differt caulibus dense cespitosis, foliorum vetustorum reliquis copiosis valide fibrosis, COLLECTIONS OF DRIED PLANTS FROM TIBET. 119 foliis firmioribus multo brevioribus, floribus minoribus sulphu- reis, perianthii segmentis exterioribus limbo parvo obovato ungue duplo breviore. Fibri radicales multi, graciles, cylindrici. Folia viridia, anguste linearia, 2-3 poll. longa, 1 lin. lata, marginibus nervis- que incrassatis. Caulis brevissimus, uniflorus. Spathe valve lineares, deorsum membranacee, sursum herbacese. Perianthii tubus 14 lin, longus; limbus vix pollicaris, pallide luteus; segmenta exteriora 3 lin. lata, limbo reflexo ungue ascendente 2-3plo breviore; segmenta interiora oblanceolata, erecta, quam exteriora vix breviora. Styli 6-7 lin. longi, appendicibus parvis. Top of the pass, at 17,800 ft. A well-marked new species, belonging to the group of beard- less Irises with a produced perianth-tube. Its other nearest neighbours are 1. Rossii, Baker, of Corea and Northern China, and I, Ludwigii, Maxim., a plant of the Altai Mountains. 91. Allium senescens, Zinn., var—Flowers pink. Rocky hill among stones at 16,200 ft. Southern European Russia through Siberia to Dahuria and Eastern Mongolia. 92. Juncus membranaceus, Loyle—Flowers white. Close to water, at 16,200 ft. Himalayas from Kashmir to Sikhim at 6000 to 13,000 ft. 93. Triglochin palustre, Zénn.—Close to water at 16,200 ft. Western Europe and North Africa to India and China, and in North America. 94. Carex stenophylla, Wahlenb.—Close to water, at 16,200 ft. South-eastern Europe to Dahuria and North China. 95. Carex Moorcroftii, Falconer.—Sandy gravelly soil at 17,600 ft. Tibet at 14,000 to 17,000 ft. 96. Trisetum subspicatum, Beauy.—At 16,500 ft. Very widely spread in temperate and polar regions both in the northern and southern hemispheres. 97. Poa nemoralis, Zinn.—Sheltered valley at 17,000 ft. All round the northern hemisphere. 120 MR. W. B. HEMSLEY ON TWO SMALL 98. Poa alpina, Linn.—Close to water at 16,200 ft. All round the northern hemisphere. 99. Poa alpina, Zinn., forma nana.—In sandy gravelly soil, at 18,000 ft. (no. 26), at 16,400 ft. (no. 103). 100. Elymus sibiricus, Linn.—Close to water at 16,200 ft. North India, Central and Northern Asia, eastward to the Sitka Islands. 101. Elymus dasystachys, 7rin.—Sandy plain at 16,000 ft., and valleys at 17,000 ft. Siberia, chiefly in the Altai and Baical mountains, and east- ward to Mongolia and North China. 102. Pennisetum flaccidum, Griseb. Afghanistan, North-west India, and Tibet to Mongolia and North China. 103. Stipa purpurea, Griseb. Gesamm. Abhandl., Gram. Hochasiens, p. 800.—Syn. Lasiagrostis tremula, Ruprecht, Sert. Thianschan. p. 35.—16,500 ft. Distributed from the Thianshan Mts. and South-east Kash- mir to East Tibet: Lake Rupshu, at 15,000 feet, Lance!; “ Tibet. oce.,”’ probably near the Lanak Pass, between 15,000 and 17,000 feet, 7. Lhomson!; province of Gnari Khorsum, at Poti, 17,000 feet, Schlagintweit, Cat. 7016!; Bhomtso Mt., up to nearly 18,000 feet, J.D. Hooker!; Thianshan, 8.E. of Chatyr- Kul, between 11,000 and 12,000 feet, Osten-Sacken ! 104. Stipa ($ Lasiagrostis) Hookeri, Stapf. Affinis S. splendenti, Trin. (Lasiagrostis splendens, Kunth), sed robustior, spiculis majoribus, glumis fertilibus longe bifidis, antheris nudis. Gramen perenne. Rhizoma ut turiones cataphyllis ovatis obtusis firmis prominenter striatis imbricatim vestitum. Culmus floriferus ac innovationes basi vaginis vetustis firmis vestitus, gla- berrimus, levis, strictus, cum panicula 1-1} ped. altus. Folia tnnovationum vagina levissima, primo arcte involuta, demum plus minusve aperta, ligula brevissima graciliter denseque fim- briata, lamina convoluta longe setaceo-attenuata, glauca, stricta, 4-10 poll. longa, dorso ecarinata, tenuiter striata, asperula vel COLLECTIONS OF DRIED PLANTS FROM TIBET. 121 levi, facie profunde et multi-sulcata in costulis asperula ; cau- lina longe vaginata, sammum fere ad panicule basin, vagina levi striata, in infimis laxa, lamina abbreviata, summi panicula breviore, ceterum ut in innovationibus. Panicula angusta subcontracta, 4-6 poll. longa, ramulis pedicellisque brevissimis vel ad 1 lin. longisasperulis. Glume vacue subequilonge, lanceolate, acute, tenues, 3-33 lin. longe, glabre ; fertilis lanceolata, vix 3 lin. longa, apice ad 3 bifida, lobis setaceo-acuminatis, inter eos aristata, 7-nervis, tota longe et patule pilosa, arista vix torta, geniculata, asperula, 4~5 lin. longa. Palea lanceolata, gluma fertili paulo brevior, ut ea longe et patule pilosa. Anthere apice nude. Lodicule 3, oblique ovato-lanceolate, libere. Styli subnulli. Stigmata ovario xquilonga. Sheltered nullahs near water, 14,800 feet (no. 124). The same plant was collected by Sir Joseph Hooker in Eastern Tibet, at 16,000 feet. 105. Calamagrostis holciformis, Jawb. et Spach. Tibet and Sikkim at great elevations. 106. Diplachne Thoroldi, Stapf. Affinis Diplachne serotine, Link, foliis caulinis paucioribus, miuus distincte distiche pateutibus, in vaginis superne vel in ore quidem longiuscule pilosis, glumis florentibus exaristatis pilosis. Rhizoma repens ut turiones dense cataphyllis stramineis ovatis brevibus prominenter striatis vel eorum residuis vestitum. Culmus erectus cum panicula 12-16 poll. altus, levis, ad 3 vagi- natus. Folia circa 5-6 cum quoque culmo, 3 basalia vaginis laxe et fere ad eandem altitudinem (1-2 poll.) vaginantibus, sequens magis remotum, summum vagina longissima ad 2-4 poll. vel tota longitudine exserta, vaginis inferioribus patule pilosis, demum glabratis, superioribus in ore et in marginibus superne tantum pilosis; ligula 3 lin. longa, truncata, lacerato-ciliata ; lamina glauca, in foliis infimis et in summo brevi 1-12 poll. longa, rarius longiore, in intermediis 13-4) poll. longa, e basi 1-14 lin. lata longissime attenuata, apice fere setacea, in infimis plana, in ceteris involuta, utrinque tenuiter multistriata, glabra vel in infimis sparse pilosa. Panicula angusta, 2-4 poll. longa, ramis ad 1-1} lin. vel infimo remoto ad 3 lin. nudis primo rachin adpressis, deinde patulis vel subreflexis, spiculis brevissime pedicellatis, pedicellis levibus, 3-floris, flore 4 imperfecto addito, 3-4 lin. longis. Glume vacue tenues, pallide vel yiolacex, 122 MR. W. B. HEMSLEY ON TWO SMALL inferior 13 lin. longa, oblongo-lanceolata, obtusiuscula, uninervis, superior 2 lin. longa obtusa vel erosula, trinervis, sparse pilis longis patulis obsita; florens oblongo-lanceolata, 2 lin. longa, saturate glauco-violacea, apice obtusiuscula minute denticulata et ciliolata, trinervis, nervo medio haud vel brevissime excurrente, patule pilosa. Palea vix brevior acuta, sparse pilosa. Lodicule truncate, obtuse lobulate. Anthere 1 lin. longe. Ovariwm glabrum, stylis paulo brevioribus, stigmatibus aspergilliformibus. Sandy soil in valleys, 15,800 ft. (no. 120). D. serotina has been referred by Mertens and Koch, Grisebach and others to Molinia, and although it does not answer exactly to the type on which the genus Diplachne was founded by Beauvois, i. e. D. fascicularis, Beauv., a North-American species, I prefer to leave it in that genus following the authority of the authors of the ‘Genera Plantarum,’ of Parlatore and of Hackel, and consequently I also refer the new species to the same genus. 107. Atropis distans, Griseb. in Ledeb. Fl. Ross. iv. p. 388, forma nana. Wide valleys, 17,000 ft. (no. 78); close to water, 16,200 ft. (no. 88), 16,500 ft. (no. 111). Var. convoluta, Trautv. in Act. Hort. Petrop. i. p. 282.—Syn. A. convoluta, Griseb. in Ledeb. Fl. Ross. iv. p. 389, forma nana. Close to water, 16,200 ft. The first form differs from the typical Atropis distans, as we usually find it in Europe, only by its dwarfness, the smaller specimens being 2-3 inches high with panicles 3 inch long and broad. The variety convoluta is likewise represented by dwarf specimens about 4 inches high, but they agree otherwise perfectly with the form of Atropis distans which prevails in the North- western Himalaya and which is identical with the plant described by Grisebach under the name of A. convoluta. This variety extends from Tibet to the Altai and westward to South Russia. 108. Festuca ovina, Linn., var. valesiaca, Koch; cf. Hackel in Boissier, Fl. Orient. v. p. 618. At 16,500 ft. (no. 110). A common grass in the higher parts of the Himalaya from the Afghanistan frontier to Sikkim, here from 12,000 ft. ascending to 17,000 ft. (Donkiah Pass, J. D. Hooker). COLLECTIONS OF DRIED PLANTS FROM TIBET. 123 109. Agropyrum Thoroldianum, Oliver in Hook. Ic. Pl. t. 2262. —At 16,500 ft. Not previously represented in the Kew Herbarium. EnuMERATION of the Plants collected on the Kuen-lun Plains at about 17,000 feet by Captain H. P. Picot *. Clematis orientalis, Zinn., var. tangutica, Maxim. Persia to Mandshuria and North China. Berberis ulicina, Hook. f. et Thoms. Previously collected by Dr. Thomson only in Nubra, at 14,000 to 16,000 ft. Christolea crassifolia, Camb. * Western Tibet to Lhassa. Malcolmia africana, Zinn. Mediterranean region to North-west India. Myricaria germanica, Linn. * Europe to Western China. Astragalus, sp. Potentilla Salessowii, Steph. North-west India and Siberia. Potentilla fruticosa, Zinn. * Cold temperate and arctic regions all round the northern hemisphere. Potentilla bifurca, Zinn. Caucasus and Tauras to Siberia and Mongolia. Potentilla sericea, Linn. * All round the northern hemisphere. Sedum crenulatum, Hook. f. et Thoms. Kumaon and Sikkim at 12,000 to 18,000 ft. Chrysanthemum Richteria, Benth. Tibet and Scongaria. * Those followed by a star were also collected by Mr. Thorold. 124 MR. W. B. HEMSLEY ON TWO SMALL Tanacetum tibeticum, Hook. f. et Thoms. * Western Tibet at 15,000 to 17,000 ft. Previously collected only by Dr. T. Thomson and Mr. Thorold. Artemisia, sp. ? Taraxacum officinale, Wigg. * Generally dispersed in temperate and cold regions. Statice, sp.—insufficient for determination. Primula sibirica, Jacq. North India, Siberia, and China. Androsace Chamezjasme, Linn. * Cold regions all round the northern hemisphere. Lindelofia Benthami, Hook. f. Kashmir and Western Tibet, at 11,000 to 15,000 ft. Pedicularis dolichorrhiza, Schrenk. Turkestan and North India to Mongolia. Allium blandum, Wall. Kumaon to Nepal, at 13,000 to 17,000 ft. Allium, sp.— A. Jacquemontii, Reg., affinis. Triglochin maritimum, Zinn. Sea-coasts and salt marshes all round the northern hemisphere. Kobresia Royleana, Boeck. Kumaon to Sikkim. Elymus sibiricus, Linn. * North India, Central and Northern Asia eastward to the Sitka Islands. Elymus dasystachys, 7rin., forma tomentella.* Siberia, Mongolia, and China. COLLECTIONS OF DRIED PLANTS FROM TIBET. 125 Asstract of Tibetan portion of Collection of Plants made by T. E. Winterbottom and R. Strachey in 1847. EXOGEN &. Genera. Species. (Genera. Species. 1. Ranunculacee. 9. Fosacee. 1. Clematis .........cc eee 1. Potentilla....... ee 7 2. Anemone .............eee eee ee 3 2, Chamerhodos ............... 1 3. Thalictrum ..............6665 2 — 8 4. Adonis ............eeceeeeeeeee 1 10. Saxvifragee. 5. Ranunculus .............0.665 6 1. Saxifraga oo... eee 5) 6. Isopyrum oo... cece 1 2. Parnassia... 0.0.0... cece eee ] 7. Delphinium..........0....04. 2 — 6 8. Aconitum oo... eee eee ee 1 Il. Crassulacee, —17 1. Sedum oo... eee 5 2. Fumariacee. 2. Sempervivum ............... 1 1. Corydalis ....... cc ceceeeeeee 3 — 6 — 3 12. Haloragre. 3. Crucifere. 1. Hippuris ..............000008. 1 L. Parrya oo... cece ees 2 —1 2. Cheiranthus ............6.. 1 13. Onagracee. B. Arabis .......ccecceeeeeeee ees 1 1. Epilobium ...........0.0.. 2 4. Alyssum .....ceeccee cee ee eee 1 —2 5. Draba oo... cece ee eeeee enone 4 14. Umbellifere. 6. Cochlearia .............0006 1 1. Trachydium...........0...... 1 7. Sisymbrium .. .........60000. 2 2. Pleurospermum ......... .. 1 8. Erysimum ..............0605 1 —2 9. Christolea ...............66. 1 15. Caprifoliacee. 10. Braya oo... cece eee 1 1. Lonicera oo... 1 11. Lepidium ........... 0.00.6. 1 —1 12. Thlaspi.......... cece eee 1 16. Composite. 13. Iberidella ...........ee. 1 L. Aster ....ccccceeceeeeeeeee eee 2 14. Crambe.. ..........c.eceee eee 1 2. Hrigeron .......... ee 1 15. Chorispora ...... ......eeeee 1 3. Leontopodium............... 1 —20 4. Anaphalis...........0..0.0..... 2 4. Violacee. 5. Allardia ........ cece eee 2 L. Viola ..cccceceeceeseeeeeeees 1 6. Tanacetum ...............4+ 3 —] 7. Artemisia .............eee ee 4 5. Caryophyllee. 8. Senecio oe... eeeeeeeeees 3 Silene cecececcecccecceesceeeee 1 9. Werneria .................2665 1 2. Lychnis ...... cece seen eee 2 i eaussurea weet eeece secs cece: 5 B. Stellaria. we. Da ReRIS peti ! 4. AVeNarla oo... cece ee 3 13. Lactuca Tennessee se se ences 1 5. Lhylacospermum ......+. 1 ~ Lactuca oe. eee eens ‘7 —8 _ —2 6. Tamariscinee. 17. Campanulacee. Lo. 9 1. Campanula ........... 1 1. Myricaria...........-.s-eeeees 2 4 en . —2 18. Primulacee, i. Gerantacee. 1. Primula oo... ee 5 1. Biebersteinia ............68. 1 2. Androsace 2... ccc cececee ee 1 2, Geramium........cccceeeeee eee 1 — 6 —2 19. Gentianacee. 8. Lequminose. 1. Gentiana 20... 6 1. Thermopsis ..........02ee 1 2. Pleurogyne .................. 1 2. Caragana......-. cece ees 2 —7 3. Gueldensteedtia ...........5 1 20. Boraginee. 4, Astragalus .........0.e es 3 1. Lindelofia .....0.....0 20... 1 5. Oxytropis ......ceeeeeee eee 4 2. Eritrichium...........0...... 2 6. Stracheya .........ee eee 1 3. Microula .................0068 1 Y Fan OSes 1 4. Onusma oo. eee cee eee 1 —13 — 5 126 Genera. Species. |Genera. Species. 21. Solanacee. | 26. Polygonacee. 4 > BRITA ......ccececece 1. Polygonum ................. 1. Physochlaina ....... t 1 2. Rhea becteeneeneeeeeeeeeeeees 1 2. Scrophularinee. —_? 1. Veronica i se eeeeeeec eee eas 1 _ 21, Eleagnacea. 1 2. Pedicularis .................. 4 . 1, Hippophaé «0.0.0... _— 1 on Gy), ~~? 28. Euphorbiacee. _ 23. Selaginea. “OL, Euphorbia. vesccscceeeee 2 1. Lagotis ............. eee 1 — 2 > . —1 29. Urticacee. 24. Labiate. 1. Urtica oe... ceeeeeeees 1 1. Elsholtzia.................008. 1 —] 2. Thymus ......... eee 1 30. Cupulifere. 3. Nepeta .....0....c eee 4 1. Betula ........c.cceeeeceeecees 1 4. Dracocephalum ............ 1 —1 5. Marrubium .................. 1 31. Salicinee. 6. Lamium ..................008 1 L. Salix ...cc.ccccccccccccceccecece 1 —9 —1 25. Chenopodiacee. 32. Gnetacee 1. Eurotia .......... eee 1 1. Ephedra ..................... ] 2. Axyris .........c cece eee ee 1 —1 3. Microgynecium ............ 1 33. Conifere. 4. Salsola ............0...ccceeeee 1 1, Juniperus oo... ee. 3 — 4 —3 EnbOGEN 2. 1. Hemodoracee. 7. Cyperacee. 1. Aletris ........0.... ee 1 1. Blysmus ................00085 1 —1 2. Elyna oo... ee eeeees 1 2. Tridee. 3. Trilepis ....... eee 1 DP. Tris... eee eee ] 4. Carex oe. ceeeeeeeee ee 6 —1 —9 3. Liliacee, 8. Graminee. 1. Polygonatum ............... 1 | 1. Piptatherum ............... 2 2. Allium ..........ccce eee eeeeee 4 | 2. Lasiagrostis.. 2.0.0.0... 1 3, Lloydia.............c cece 1 B. Stipa oe 1 . — 6 4. Deyeuxia oe 1 4. Juncacee, 5. Deschampsia ............... 1 1. Juncus «00.0.2... eeeee 3 6. Trisetum ..............0000008 2 — 3s Ce <0) ee 4 5, Alismacee. 8. Schoenodorus ............... 1 1. Triglochin ............... we «2 9. Festuca oo... 2 — 2 | 10. Agropyrum ......... Seeeeeee . 2 6. Naiadacee. Il. Blymus............000.c..cee 2 1. Potamogeton ..........04... 2 12. Hordeum.................064. 2 —2 —21 CrYPToGamaz. 1. Equisetacee. 4. Lichenes. 1. Equisetum .................. 1 1. Lecidea..... eee eee eee ee 4 —1 2. Cladonia ................2.... 1 2. Filices 3. Urceolaria .................. 1 1. Cystopteris .................. 1 4. Parmelia .............. 0... 10 —1 5. Cetraria oe 1 _ 3. Musci. 6. Evernia oo... eee. 1 1, Distichium .......000........ 1 —18 2. Bryum ......... eee 2 3. Hypnum «00... ee 1 —4 R. Srracuey. MR. W. B. HEMSLEY ON TWO SMALL Siberia. COLLECTIONS OF DRIED PLANTS FROM TIBET. TuHoroLp’s TiBET PLANTS. { Tibet Tibet East- only. N. India, ward. West- ward, oo Wide, in- 127 cluding America. 8 LINN. JOURN.—BOTANY, VOL, XXX. Ranunculacee, Clematis alpina Adonis cerulea Ranunculus Cymbalarie ... involucratus ... ” pulchellus ” hyperboreus Delphinium grandiflorum . ” ceruleum ...... Ce rs bP) 8 Papaveracee. Meconopsis horridula......... Hypecoum leptocarpum ...... >) a Fumariacee. Corydalis Boweri ” Hendersonti a ” Hendersonii ...... Crucifere. Parrya lanuginosa Erysimum funiculosum ...... Sisymbrium humile Braya rosea .......cceeeeeeeees » UNI flora... eee eee Draba incompta » alpina Cochlearia scapiflora Sere ve rereeenes Christolea crassifolia Capsella Thomsoni ............ Lepidium capitatum Caryophillee. Stellaria sp. .......... steer eeeees AYenarla SP. ...eeeec cee cece cece | 2 2 Tamariscinee, | Myricaria germanica i 1 | Lequininose. | | Thermopsis inflata | | Astragalus Hendersonii ......) | tribulifolius | eee tel eeeene ” | ” China China Mongolia China OK S * | Mongolia | China Ural Yarkand * | | | { | Caucasus | | | Yarkand | Yarkand Europe 128 MR. W. B. HEMSLEY ON TWO SMALL ‘TroroLp’s TrpeT Pants (continued). — . id . | 7 ribet) NORE | acts | wont Siberia. only. x India, WAT wae | America. _———| —— ~ | —_"| 7 7 8 4 8 Leguminose (cont.) 8 ‘ | Oxytropis microphylla .....) ++ * » densa .........0006+ * % ” physocarpa......... | ” SP. ceeveeseceeeeeees _ » tatarica becee sae bee we ves Yarkand ” SP. cseceeeeeeeeteeeee 14 Rosacee. * Chameerhodos sabulosa ...... bee vee China Potentilla fruticosa ............! . * Ec. ee * 3 Crassulaceeé. CVarkand | rkanc Sedum tibeticum ............... wes ses ve | Afgh. | * » quadrifidum............ wee * Saxifragacee. * Parnassia trinervis ............ tes * Saxifraga parva ..........6..0 * ” SP. seceeeecreeeeeeees Umbellifere. | Pleurospermum stellatum ...) * | Dipsacacee. | Morina Coulteriana............ bes eee |W. imal. L Composite. | Aster tibeticus ..............000 wee * » Thoroldii ............... * », molliusculus ............ we * Hrigeron sp. .....seseeeseeeees Anaphalis xylorrhiza ......... vee * ” SP. cceeeeeseceneeeaes Tanacetum tibeticum ......... * * Artemisia desertorum .. ......)... * a” re Senecio arnicoides ............ —_ * Cremanthodium sp. ......... | Saussurea subulata ............ wee wee ves Yarkand * ” pygimea ............ tee wee Dahuria | ” glanduligera ...... # | 7 tridactyla ......... [| oes * m | ” sorocephala_...... ” Aster .........e..08. * ” Thoroldi............ * _ Crepis sorocephala ............ * | | nn) ee | | Taraxacum officinale .........) ... wes wes | we * 15 21 | 17 16 | CO9 12 7 | | s COLLECTIONS OF DRIED PLANTS FROM TIBET. Troroup’s TrBpet Pants (continued). 129 | | Siberia. 15 18 Tibet a only. IN. India, Tibet nd East- ward. West- | ward. Primulacee. Primula tibetica Androsace Chamajasme ...... Glaux maritima ree eee Se Gentianacee. Gentiana falcata ” squarrosa decumbens ” humilis............... Pleurogyne diffusa Swertia sp. ..... c.ceceeeeeeeeee 6 eee eee eeeeeeee eee eeree eer eeeeee ” Boraginee. Tretocarya pratensis Microula Benthami............ 5) “ weet eeees Serophularinee. Pedicularis cheilanthifolia ... alaschanica ” 2 Labiate. Nepeta longebracteata......... Dracocephalum heterophyl- lum eee ee ee eee ee Polygonee. eeecee Polygonum sibiricum Thymeleacce. peerereee Stellera Chamzjasme 1 Urticacee. see eeereneee Urtica hyperborea 1 Gnetacee. Ephedra Gerardi. na 1 eee eer eee Tridee. Tris Thoroldi 1 eee ee ee Liliac’e. Allium senescens ..........0s66. 17 16 * China Mongolia Dahuria China China China Mongolia China | | 23 16 18 Baltistan Caucasus | Caucasus Europe Mongolia E. Europe 17 K2 Wide, in-) cluding | America. | 130 MR. W. B. HEMSLEY ON PLANTS THoroty’s TrpeT PLants (continued). , — ‘Wide, in- a Tibet | Tibet East- West- | cluding | Siberia. only. India. ward. ward. | America, | \ —_—_— —— | —| — | 2 18 we | 18 | Juncacee, 23 16 | | Juncus membranaceuws......... we | | | 1 | | Alismacee. | | Triglochin palustre ............ * | | 1 | Cyperacee, | Carex stenophylla ............ ves be China Europe » Moorecroftii ............ * 2 Graminee. | Trisetum subspicatum ......... | * | Poa nemoralis ..............4.4. | . | ” alpina weer ewe reee re sseeee eee nee ot ere * Elymus sibiricus ............... wee bee Sitka * »» asystachys............ wes ves China Pennisetum flaccidum......... vee Lee China | Afgh. | | * Stipa purpurea.................. | 5 Hookeri ...........0.0.... * Calamagrostis holciformis ...) ... * | Diplachne Thoroldi......... * , . * Atropis distans ............... vee we ses KE, Europe | Festuca ovina ..............000. vee a * | orepyrum Thoroldianum...| * | | 1 | 2 27 18 | 22 2000 | EXPLANATION OF THE PLATES. Puiare IV, Fig. 1. A plant of Crepis sorocephala, Hemsl., natura) size; 2, 4 capitulum from the cluster of capitula; 3, a flower; 4, a bristle of the pappus- 5, a plant of Saussurea Thoroldii, Hemsl., natural size, also having densely clustered capitula; 6, a capitulum ; 7, a flower; 8, one of the outer shorter bristles of the pappus; 9, stamens. All enlarged. Prats V. Fig. 1. A plant of Saussurea Aster, Hemsl., natural size, having densely clustered capitula; 2, a leaf; 3, a flower; 4, stamens; 5, stigma and upper portion of style. All enlarged. yA, Linn. Soc. JourN.Bor Vou XXX. P: HFitch del et hth. Hermsiey . Cc uU PHALA , Herrsl 3OROC FE: Q ie) 1-4 CREPIS mrge ya , He THOROLDII 5-9 SAUSSUREA Tamm. Soc Journ. Bor. Vou. XXX. Pid. Hemealey. JN.Fitch 2mm . Hitch dele ath. C SAUSSUREA ASTER, Henwley. COLLECTED IN CENTRAL TIBET. 131 Mr. Woopvitte Rockutiry’s Centrat Trper Pants. Collected in 1892. Sryce the foregoing paper on Dr. Thorold’s Tibet plants and Captain Picot’s Kuen-lun plants was read, the Kew Herbarium has been enriched, through the kindness of Professor C. S. Sargent, Director of the Arnold Arboretum, Harvard University, U.S.A., by the collection of dried plants made by Mr. W. W. Rockhill on his last journey in Tibet, in 1892. It is unnecessary to say much concerning this gentleman’s travels in China, Mon- golia, and Tibet. His first journey is described and illustrated in his ‘Land of the Lamas,’ published in 1891; and his last journey, on which the plants enumerated below were collected, is the subject of a work already, I believe, in the printer’s hands. It will be remembered that Mr. Rockhill was last year (1893) awarded the Patron’s cr Victoria Medal of the Royal Geographical Society. I had prepared a brief outline of Mr. Rockhill’s route from a condensed report of his account of his journey read before the Royal Geographical Society in March of the past year; but on the very day of going to press I have received a prefatory note from him, which is much more to the purpose, aud may follow here :— “The object I had in view when making the little collection of plants, which, through Professor Ch. S. Sargent’s kindness, has been examined and classified by Mr. Hemsley, of the Royal Gardens at Kew, was to give some idea of the flora of the country between the Kuen-lun range to the north and the in- habited regions of Tibet adjacent to the Tengri Nor on the south. This region has an average altitude of 15,000 feet above sea- level along the route followed by me in 1892, and had not, prior to my visit, been explored. “The route followed in 1879 by Col. Przewalsky, when tra- velling towards Lhasa, which was nearly parallel to the last that I took, differed considerably as regards the configuration of the country from mine; and consequently I anticipated that notable differences in the flora along the two roads would be discovered. “T traversed this country in the months of May, June, July, and part of August, and heavy snowstorms and nearly daily frosts occurred during this period, though the thermometer rose more than once to 70° F., and even 83° on one occasion in the shade at 2 p.m. The mean temperature from the 17th of May, 132 MR. W. B. HEMSLEY ON PLANTS when we entered the mountainous region to the south of the Ts’aidan, to the 11th of August, when we descended to below the Timber line (13,500 feet above sea-level) on the Ramachu, where I ceased collecting plants, except such as the natives pointed out to me as being used by them either as food or medicinally, is shown in the following table :— 1892. 7 AM. 2 P.M. 7 P.M. May 17 to 81... +37°5 F. +54°°6 F, +37°°3 F. June .......e.eee +35°7 +55°°9 +38°:3 July eee +43°-0 +54°°6 +44°2 Aug. ltoll ... +40°°6 +61°5 +47°°3 “Nearly the whole of the region traversed in this interval was of sandstone formation, the predominating colour of which was bright red. The water was invariably brackish, and in many cases undrinkable; the soil everywhere sandy, or covered with a rather fine gravel, and occasionally a little clay. The grasses grew in bunches, nowhere forming a sod, except around the rare pools of pure water fed by the melting snows we occasionally passed. “T was careful to collect all the flowering plants I saw along my route, and the barrenness of this region may be judged by the very smail number I have brought home with me. ‘The only edible plant we found in this country was a species of onton (Allium senescens)*, which grew in the sand in great quantities at altitudes higher than 15,000 ft. above sea-level, though we looked for it in vain below this level. “‘T may here remark that the rhubarb plant, which I found growing in enormous quantities on the north and north-eastern slopes of mountains on the Ich’u, Lench’u, and other feeders of the Jyama-nu cb’u, thrived at an altitude above sea-level ranging from 12,000 to 13,500 ft. I note this fact as Col. Przewalsky (Mongolia, ii. p. 84) says that this plant rarely flourishes at an elevation of more than 10,000 ft. above the level of the sea. “For further details concerning the habitats of the various plants I brought home with me I beg to refer to the annexed enumeration.” W. Woopvitie Rockit. This is an exceedingly interesting collection, especially when * There was no specimen of this plant in Mr. Rockhill’s collection.— W.B.H. COLLECTED IN CENTRAL TIBET. 133 examined in connection with Mr. Thorold’s; the plants for the greater part being of the same habit and diminutive size. More than half of them, however, are different species; and most of them had previously only been collected by Przewalsky, from whose specimens the lamented Maximowicz described them. Several, it will be seen, too, were previously only known from the extreme western part of Tibet. In all cases where the species are different from Thorold’s their general distribution is given. The localities, altitudes, and geographical positions were supplied by Mr. Rockhill. Mr. Rockhill also brought home leaves of the famous “ white sandalwood tree” of Kumbum, Syringa villosa, Vahl, which he erroneously refers to as Philadelphus coronarius, Linn., in his ‘Land of the Lamas.’ Enumeration of the Plants collected by Mr. W. Woodville Rockhill. 1. Clematis graveolens, Lind/.—Flowers light vellow. Pochu valley ; very abundant at 14,000 ft. Lat. N. 31° 45’, long. E. 94° 45’. Aug. 14, 1892. Western Himalaya at 6000 to 11,000 ft., Tibet and Western China. This form is the same as that named C. orientalis var. tan- gutica by Maximowicz, but if the two are maintained as inde- pendent species it is better referred here. 2. Anemone imbricata, Maxim. Fl. Tangut. i. p. 8, t. 22. ff. 1-6.—Foot-hills of Dangla mountains, N.W. extremity of range at 16,500 ft. Lat. N. 33° 40’, long. E. 90° 35’. June 27, 1892. Previously collected only by Przewalsky on the extreme Upper Yaugtsekiang in Tibet. 3. Ranunculus tricuspis, Maxim. Fl. Tangut.i. p. 12; Enum. Pl. Mongol. i. p. 16, t. 4. ff. 17-27.—Valley of Murus; valley bottom at 15,640 ft. Lat. N. 33° 44’, long. E. 91° 18. June 23, 1892. Mongolia. 4. Delphinium grandiflorum, Zinn.—Kechu valley at 12,700 ft. Lat. N. 31° 25’, long. E. 96° 28’. Aug. 22, 1892. On river-bottom ; fine forest-growth, mostly pines (), on hill- sides ; fine grass. 184 MR. W. B. HEMSLEY ON PLANTS 5. Delphinium Pylzowii, Maxim. in Mél. Biol. ix. p. 709; #7. Tengut. i. p. 21, t.3.—Dangchu valley, river-bottom of gravel ard clay ; good fodder, at 14,500 ft. Lat. N. 32° 12’, long. E. 99° 12’. July 23, 1892. Mongolia. 6. Meconopsis horridula, Hook. f. et Thoms.—Plateau west of Dangla mountains at 16,350 ft. Lat. N. 32° 51', long. E. 89° 44". July 8, 1892. Sandy soil, some clay. 7. Corydalis Hendersonii, Hems/.— Basin of Murus. Extreme head of valley on foot-hills of Dangla mountains at 16,340 ft. Lat. N. 38° 43’, long. E. 90° 50’... June 25, 1892. Sandstone. See description of this species in Thorold’s list at p. 109. 8. Parrya exscapa, Ledeb.—Basin of Murus. Extreme head of valley, on foot-hills of Dangla mountains, sandstone, at 16,340 ft. Lat. N. 33° 43’, long. E. 90° 50'.. June 25, 1892. Altai mountains and Western Tibet. 9. Arabis, sp.? Insufficient for determination.—Valley of Tsacha-tsang-bo-chu at 14,700 ft. Lat. N. 32° 13’, long. E. 90° 14’. July 6, 1892. 10. Erysimum Chamephyton, Maxim. Fl. Tangut. i. p. 63, t. 28. ff. 1-10.—Hill-slope two miles north of Murus river (head- waters of Yangtsekiang); sandy soil, some clay, at 14,750 ft. Lat. N. 33° 53’, long. E. 91° 31’. June 21, 1892. Basin of Murus in lateral valley, sandstone, at 15,700 ft. Lat. N. 338° 45’, long. E. 91° 05’. June 24, 1892. North-eastern Tibet. 11. Eutrema Przewalskii, Maxim. Fl. Tangut. i. p. 68, t. 28. ff. 11-23.—Basin of Murus, in lateral valley, sandstone, at 15,700 ft. Lat. N. 33° 45', long. E. 91° 05’. June 24, 1892. North-eastern Tibet. 12. Myricaria prostrata, Hook. f. et Thoms. in Benth. et Hook. f. Gen. Pl. i. p. 161.—Upper Naichi gol valley near river at 12,130 ft. Lat. N. 35° 52’, long. E. 93° 49’. May 21, 1892. Called “aura kashim” by the Mongols. First plant in flower seen On journey. I have followed Maximowiez in restoring this form to specific COLLECTED IN CENTRAL TIBET. 135 rank. It is restricted to the elevated alpine regions of the Himalayas and Tibet. See Maximowicz (FI. Tangut. p. 95, t.31), where it is fully described and figured. In Hooker’s Fl. Brit. Ind. i. p. 250, it is treated as a variety of ML. germanica. 13. Gueldenstedtia ?, insufficient for determination.—Gela, on Ramongchu at 12,670 ft. Lat. N. 31° 40’, long. E. 94° 36’. Aug. 138, 1892. Fine crops of barley and turnips near by. 14. Astragalus or Oxytropis, sp.? Material insufficient to de- termine the genus.—Toktomai-ulan-muren at 14,340 ft. Lat. N. 34° 09’, long. E. 91° 80’. June 20, 1892. Sandy soil. 15. Astragalus or Oxytropis, sp.? Material insufficient to de- termine the genus with certainty.— Valley of Murus, head-waters Yangtsekiang, at 14,900 ft. Lat. N. 33° 45’, long. E. 91° 20’. June 22, 1892. 16. Potentilla fruticosa, Linn., var. pumila, Hook. f.—Plateau west of Dangla mountains at 16,350 ft. Lat. N. 32° 51’, long. E. §9° 44. July 3, 1892. Sandy soil, some clay. 17. Potentilla anserina, Linn.— Plateau west of Dangla moun- tains ; saidy, some clay, at 16,220 ft. Lat. N. 33° 09’, long. E. 89° 38’. July 2, 1892. This is widely spread in the temperate and cold regions of both the northern and southern hemispheres. 18. Potentilla nivea, Zinn.—Kechu valley; on river-bottom at 12,700 ft. Lat. N. 31° 25’, long. E. 96° 28’. Aug. 22, 1892. Fine forest-growth, mostly pines (?) on hill-sides; fine grass. Alpine and Arctic regions of the northern hemisphere. 19. Sedum algidum, Ledeb., var. tanguticum, J/axim.—Camp north of Tsacha-tsang-bo-chu ; sandy soil at 15,650 ft. Lat. N. 32° 28', long. B. 90° 03". July 5, 1892. The species is a native of the Altai regions of Siberia; the variety was described from specimens from north-western Kansuh. 20. Aster tibeticus, Hook. f.—Valley of Murus, valley-bottom at 15,640 ft. Lat. N. 33° 44’, long. E. 91° 18’. June 23, 1892. Western Tibet and Kashmir at altitudes of 14,000 to 16,000 ft. 136 MR. W. B. HEMSLEY ON PLANTS 21. Inula? Material insufficient for determination.— Foot-hills of Dangla mountains, north-west extremity of range at 16,500 it. Lat. N. 83°40', long. E. 90°35’. June 27, 1892. 22. Leontopodium alpinum, Cass., var.—Bank Chilchang-tso (Lake Glenelg). Hill-side ; limestone and red sandstone ; lake salt, at 16,000 ft. Lat. N. 38° 27', long. E. 90° 10’. June 30, 1892. Alps of Europe, through Central Asia and North India to China, ascending in the Himalayas to nearly 18,000 ft. The variety collected by Mr. Rockhill is a very elegant little plant about three inches high with remarkably spathulate leaves. 23. Leontopodium Stracheyi, C. B. Clarke in Herb. Kew. (L. alpinum, Cass., var. Stracheyi, Hook. f.).—Ruchu valley, in river-bottom, at 12,100 ft. Lat. N. 31° 10’, long. E. 95° 12’. Aug. 16, 1892. Fine crops of barley and turnips now ripe. Also a little wheat. This Western Tibet and Himalayan plant is so easily distin- guished from the other forms that it may well be accorded specific rank. It ranges from Kumaon to Nepal. 24, Anaphalis mucronata, C. B. Clarke.—Basin of Dangchu, right bank affluent. Clay and sand-gravel, at 15,180 {t. Lat. N. 32° 20, long. E. 92° 08’. July 21, 1892. This form is united with A. nubigena, DC., in the ‘Flora ot British India.’ It is only found at great altitudes in the Hima- layas and Tibet. 25, Antennaria nana, Hook. f. et Thoms.—Valley of Murus, head-waters Yangtsekiang, at 14,900 ft. Lat. N. 33° 45’, long. E. 91° 20’. June 22, 1892. Western Tibet in the Nubra and Shayuk valleys at 12,000 to 14,000 ft. 26. Saussurea tangutica, Maxim. in Mél. Biol. xi. p. 247.— Near summit of Gam (or Angti) la at 15,600 ft. Lat. N. 30° 40’, long. E. 98° 13’. Sept. 4, 1892. Tangut and Northern Tibet. The leaves are infused and used by the natives as a tonic. Called in Tibetan Sha-pé gong-t’ag. It is said to grow only on COLLECTED IN CENTRAL TIBET. 137 the west side of this mountain. The Chinese call it “ snow lotus ” (Hsuch lien). 27. Taraxacum palustre, DC.— Valley of Murus, valley-bottom at 15,640 ft. Lat. N. 383° 44’, long. E. 91° 18’. June 23, 1892. This is usually regarded as a variety of the almost ubiquitous T. officinale. 28. Cyananthus incanus, Hook. f. et Thoms., var. leiocalyx, Franch. in Morot’s Journ. de Bot. i. 1887, p. 279.—Kechu valley at 12,700 ft. Lat. N. 31° 25’, long. E. 96° 28’. Aug. 22, 1892. On river-bottom. Fine forest-growth, mostly pines (?) on hill- sides ; fine grass. A Himalayan species, of which this is a naked-calyx variety, also found in Yunnan. ‘The typical form inhabits alpine localities at 12,000 to 16,000 ft. 29. Androsace tapeta, Mavim. in Mél. Biol. xii. p. 754.— Valley of Murus, head-waters Yangtsekiang, at 14,900 ft. Lat. N. 33° 45’, long. E. 91° 20’. June 22, 1892. Kansub and Szechuen, in Western China. 30. Androsace villosa, Zinn., var. latifolia, Ledeb.—Valley of Murus, valley-bottom at 15,640 ft. Lat. N. 33° 44’, long. E. 91° 18’. June 23, 1892. This species is widely dispersed from Asia Minor through Central Asia, North Asia, and the mountains of North Iudia. 31. Gentiana Rockhillii, Hems/., n. sp. Species G. aristate, Maxim., similis sed minor strictior floribus fere cylindricis angustissimis. Annua, erecta, simplex vel pauciramosa, 1—2-pollicaris, glaber- rima. Folia subscariosa, lineari-subulata, vere conduplicata, 3-4 lineas longa, apice breviter aristata, basi semiamplexicaulia, suberecta, cauli fere appressa. Flores cerulei, terminales, soli- tarii, subsessiles, circiter 9 lineas longi; calyx subscariosus, corolla tubum equans, dentibus lineari-subulatis; corolla sursum leviter dilatate, lobi breves, oblongi, vix acuti, erecti, conniventes, intermediis brevioribus albis tenuissimis, fauce nuda; stamina cum pistillo omnino inclusa; styli brevissimi, stigmatibus capi- tatis. 138 MR. W. B. HEMSLEY ON PLANTS Kechu valley at 12,700 ft. Lat. N. 31° 25’, long. E. 96° 28’. Aug. 22, 1892. On river-bottom. Fine forest-growth, mostly pines (?), on hill-sides ; fine grass. 32. Tretocarya sikkimensis, Oliver, in Hook. Ic. Plant. t. 2255. — Basin of Suchu valley, north side, Drayalamo pass, at 14,600 It. Lat. N. 31° 52’, long. E. 93° 17’. Aug. 2, 1892. Limestone; fine grass ; flowers blue, very abundant. Sikkim Himalaya at 11,500 ft., and Western China near Tachienlu. Mr. Reekhill’s specimen is much smaller than the others and nearly glabrous. 33. Pedicularis Oederi, Vah/ (P. versicolor, Wahlenb.).— Valley of Murus, valley-bottom at 15,640 ft. Lat. N. 33° 44’, long. E. 91°18’. June 23, 1892. Alpine and Arctic regions of Europe, Asia, and America. 34. Pedicularis Przewalskii, Maxim. in Mél. Biol. x. p. 84, et xii. p. 787, n. 2. fig. 2.—Large state. Basin of Suchu, valley north side, Drayalamo pass, at 14,000 ft. Lat. N. 31° 52’, long. E. 93° 17'. Aug. 2, 1892. Limestone ; fine grass; flowers blue, very abundant. Eastern Himalaya, Tibet, and China. 35. Lagotis brachystachya, Maxim. in Mél. Biol. xi. p. 300.— Hill-slope 2 miles N. of Murus river, head-waters Yangtsekiang, at 14,750 ft. Lat. N. 33° 53’, long. E. 91° 31’. June 21, 1892. Sandy soil, some clay. Kansuh. 36. Polygonum Bistorta, Zinn.—Pochu valley at 14,000 ft. Lat. N. 31° 45’, long. BE. 94° 45. Aug. 14, 1892. Temperate and cold regions of Europe, Asia, and America. 37. Polygonum viviparum, Zinn.—Pochu valley at 14,000 ft. Lat. N.31° 45’, long. BE. 94° 45’. Aug. 14, 1892. The seeds are parched and ground and eaten mixed with barley-meal (tsamba). Tibetans call it ranpa or ramba. Temperate and Arctic regions of Europe, Asia, and America. 38. Polygonum bistortioides, Boiss —Ramachu valley, hill-side, at 12,800 ft. Lat. N. 31° 48’, long. E. 94° 28’. Aug. 12, 1892. Used by the natives for fuod like P. viviparum. COLLECTED IN CENTRAL TIBET. 139 This species or variety is found in Asia Minor and Persia, Although very distinct from ordinary P. Bistorta, Boissier (Flora Orientalis, iv. p. 1028) subsequently united it with that species. 39. Iris Thoroldi, Baker, ante, p. 118, et Hook. Ic. Plant. ined.—Sharaknyi-gol, hill-slope at 13,800 ft. Lat. N. 35° 50, long. E. 93° 27’. May 29, 1892. Described from specimens collected by Dr. Thorold at an altitude of 17,800 ft. Mr. Rockhill’s specimens furnish better flowers. 40. Tulipa (§ Orithyia) sp. af. T. eduli, Baker.—Sharakuyi- gol, hill-slope at 18,800 feet. Lat. N. 35° 50’, long. E. 93° 27’. May 29, 1892. Tulipa edulis is a native of Japan, and Mr. Rockhill’s one flower is insufficient for satisfactory identification. 41. Carex Moorcroftii, Boott.—Hill-slope two miles north of Murus river, head-waters Yangtsekiang, at 14,750 ft. Lat. N. 35° 53’, long. E. 91° 31’. Sandy soil, some clay. June 21, 1892. Yarkand and Western Himalaya. 42. Kobresia Sargentiana, Hemsl., n. sp. R. schenoidei valde affinis sed bracteis latissimis spicam fere omnino involventibus late scariosis subtruncatis simul emargi- natis. Hill-slope two miles N. of Murus river, head-waters Yang- tsekiang, at 14,750 ft. Lat. N. 33° 53’, long. E.91° 31’. Sandy soil, some clay. June 21, 1892. 43. Miscanthus sinensis, Anderss.— Near top of Feiyuehkuan pass, S.W. of Yachoufu, in West Szechuen, at 3583 ft. Oct. 11, 1892. China, from Japan and Korea to Hongkong and Canton, Luchu and Bonin Islands, Tonquin, Borneo, and Celebes. 44, Stipa, insufficient for determination.—Hill-slope two miles N. of Murus river, head-waters Yangtsekiang, at 14,750 ft. Lat. N. 33° 53’, long. E. 91°31’. Sandy soil, some clay. June 21, 1892. 45. Calamagrostis, sp.— Near top of Feiyuehkuan pass, S.W. of Yachoufu, in West Szechuen, at 3583 ft. Oct. 11, 1892. 140 MR. J. BORNMULLER’S RECENT BOTANICAL 46. Festuca ovina, Linn.?—Hill-slope two miles north of Murus river, head-waters Yangtsekiang, at 14,750 ft. Liat. N. 33° 53', long. E. 91° 81'. Sandy soil, some clay. June 21, 1892. Europe, N. Africa, Siberia, Himalaya, N. & S. America, and mountains of Australasia. Mr. Rockhill’s specimen is a mere fragment. 47. Usnea barbata, Fries.—Valley of Pontramo, east of Ba- tang, at 12,600 ft. Lat. N. 29° 59’, long. E. 99° 42’. Sept. 19, 1892. Sometimes 30 feet long. Hangs only on the oaks called “ green oaks” (ching kang) by the Chinese. This oak is called by the French missionaries “ chéne 4 feuilles de houx.”’ All over the world in temperate and tropical regions. Recent Botanical Exploration in Southern Persia, being the substance of a Letter from Mr. J. BornmtLier to Dr. Orto - Srapr. (Communicated by the Secretary.) {Read 16th February, 1893.] Mr. J. BornMiLyer, a botanist known from his collections in Asia Minor, went, late in 1891, to S. Persia with the intention of exploring the high mountains of Kirman, which were botanically a terra incognita. He reached the province of Kirman early in May of Jast year, and visited first Kuh Sirdsh and Kuh Jupar. As it was still too early in the season for exploring the high alpine region, he repeated his excursions to Kuh Jupar, from which point his narrative begins. Mr. J. Bornmiiller brought his exploring tour in the pro- vince of Kirman to a successful conclusion. The following is a short summary of his excursions after his first visit to Kuh Jupar. He visited Kuh Jupar again in the second week of June, when he reached an elevation of 3850 m. (12,644 English feet), but was unable to get to the very top on account of the excessive steepness of the rocks along the ridge. He encamped for six days at 3000 m. (9842 feet). In the beginning of July he started for Kuh Lalesar and Kuh Hezar, which, like Jupar, had never been visited before by a botanist. EXPLORATION IN SOUTHERN PERSIA. 141 Kuh Lalesar is S.W. of Kuh Jupar, Kuh Hezar S. of it; they rise from level or hilly, generally desert, land to 3900 and 4900 m. (12,793 feet; 16,073 feet). Bornmiiller succeeded in reaching the culminating peaks of both mountains. The general character of the vegetation is the same in Kuh Jupar, Kuh Lalesar, and Kuh Hezar; and it appears to agree perfectly with that of equal elevations in South-west Persia, whilst the more Indian elements, which are found even as far as Western Belu- chistan, seem to be absent. The plains and lower valleys are covered chiefly with a halophytic vegetation, which becomes sometimes very scanty, partly also with masses of Glycyrrhiza, Prosopis, Sophora alopecuroides, and other plants, which prefer a deeper and more clayey soil and occupy usually land which was once in cultivation but is now deserted. Stony hills and gravelly plains abound more especially in wormwoods, Stedlera, Zygophyllum eurypterum, Scorzonera tortuosissima, Anabasis, and other Salsolacez, and occasionally in Rheum Ribes. So far the vegetation of the ‘‘ Biaban” of Kirman agrees perfectly with the character it assumes in the deserts of Central and North Persia. The slopes of the mountains are sprinkled with a scanty “‘jengeli”’ vegetation consisting of scattered shrubs or trees of Ficus carica (to 2900m. ; 9514 feet), various species of Amygdalus, Colutea persica, Daphne acuminata, Lonicera nummulariefolia (to 2800 m.; 9186 feet), Crategus Azarolus, Cotoneaster nummu- laria, Acer canescens (to 3200 m.; 10,498 feet), Pistacia mutica, Ephedra, Rhamnus, &c.; whilst the banks of rivers and torrents are sometimes bordered by thickets of willows and tamarisks, sometimes of Berberis (to 3600 m.), Rosa lacerans (?) (to 3700 m. ; 12,139 feet), and Rubus sanctus. Quercus persica, which forms forest S.W. of Shiraz, and Celtis transcaucasica, which is very characteristic of the higher and moister valleys of South-west Persia, were not met with by Bornmiiller on those mountains. On the other hand, he found Juniperus excelsa in a valley on Kuh Lalesar up to 3400 m. (11,155 feet). The average limit for shrubs hes on Kuh Hezar a little above 3600 m. (11,811 feet). A dwarf Ribes only was gathered as high as 4500 m. Bornmiller found Kuh Hezar very dry, and less rich in vegetation than Kuh Lalesar and Kuh Jupar. But even on Kuh Lalesar, where he met with snowfields 4m. deep (13 feet) in the middle of July, the vegetation con- sisted almost entirely of species which were mere alpine repre- 142 BOTANICAL EXPLORATION IN SOUTHERN PERSIA. sentatives of the xerophilous types of the lower region. The only high-mountain plants of a more northern character were, apart from the Ribes mentioned above, gathered in swampy meadows. Bornmiiller mentions particularly a Primula, evidently P. auri- culata, Pall., which is very common on all three mountains, from 3000-3800 m. (9842-12,467 feet), a small-flowered Parnassia (P. subacaulis, Kar. & Kir. ?),a very fine Pedicularis, and a dwarf annual Gentiana (G. minutissima, Boiss.?). Wery characteristic of the steep rocks were the beautiful cushions of Dionysia. The usual fruit-trees were the apricot, plum, apple, pear, walnut, almond, and a variety of Eleagnus angustifolia with edible fruit. They all bore very well and freely even at Lalesar, at 3010 m. (9873 feet), except the almond-trees, which were mostly sterile. The walnut-tree sometimes attains enormous dimensions. The plane, which is planted frequently, forms very fine trees even at 2600 m. (8531 feet), but it does not reach the 3000 m. (9842 feet) ine. At Chehar Tag, S.W. of Kuh Hezar, at 3000 m., the wheat harvest was in operation on the 7th of August. Bornmiiller left Kirman for Shiraz on the 16th of September. He went by way of Mashiz and Saidabad to Niris. The country was generally desert and very rich in Salsolacew. The hilly parts along the Kuh Paris only were covered with woody vegetation, but of a distinct desert character, whilst Pistacia mutica and Daphne acuminata were observed in a pass at 2350 m. (7724 feet), and in the mountains E. of Niris, where also Acer canescens and, for the first time also, Celtis transcaucasica were found. The low country along the south-east shore of the Niris lake yielded numerous Salsolacez in a fruiting state. But the mountains S. of it abounded in tree- and shrub-vegetation between 1540 m. (5051 feet) and 1800 m. (5905 feet) ; and on the 9th of October Bornmiller passed for several hours through a fine wood of Pistacia mutica with Acer, Amygdalus, Populus euphratica, Ephedra foliata, and a few patches of Nerium and Myrtus. He crossed this range and reached Shiraz by way of Servistan, where, however, no Cupressus—as the name seems to indicate (“ Serv,” Cupressus)—was seen. Bornmiiller intended to leave Shiraz at the end of November for Bushir, whence he wished to go either to Maskat orto Basra. In the spring he will start for Shuster and then try to proceed through the western Zagros to Hamadan. CC REVISION OF THE GENUS NIPADITES. 143 Revision of the Genus WMipadites, Bowerb. By A. B. Renpiez, M.A., B.Se., F.L.S. [Read Ist June, 1893.] (Puates VI. & VIL.) In the ‘ Philosophical Transactions’ for 1757 (p. 8396) Dr. James Parsons gives an account of some fossil fruits and other bodies found in the island of Sheppey by Edward Jacobs, who also sub- sequently published an account of them as an Appendix to his ‘ Plante Favershamienses’ (1777). Of Parsons’s fruits, the only ones bearing any resemblance to Nipadites are figs. 1 and 3 on tab. xv., which, he says, seem to be figs petrified when hard and green. In 1784 Burtin, in his ‘ Oryctographie de Bruxelles,’ describes as coco-nuts some large fossil fruits from the Middle Eocene at Woluwe near Brussels. Numerous specimens have since been brought from the same beds; they are as large as coco-nuts, but broader and more ovate. In 1785 James Douglas published ‘A Dissertation on the Antiquity of the Earth, and on p. 25 refers to the Sheppey fruits, one of which, the property of Sir Joseph Banks, he figures (t. v.) and describes as “ aspecies of almond .... remarkable in the opinion of the most intelligent persons as bearing no analogy in size to any recent specimens of this nut discoverable in any quarter of the globe.” In 1804 Parkinson (‘ Organic Remains,’ p. 458) mentions the same and similar fossils, which, he says, may perhaps be referred to the genus Cocos. In plate vii. he figures two specimens: that represented in figs. 1-3 “was originally in the possession of Sir Joseph Banks, by whom it was presented to the British Museum ;” while the subject of figs. 4 and 5 “ was added to the collection in the British Museum by Mr. Douglas, by whom it is figured and described in his ingenious essay on the antiquity of the earth.” Plate vi. fig. 6 is probably the cast of the interior of a fruit of the same species, and figs. 5 and 7 the surface and section of an unripe specimen. Adolphe Brongniart, in his ‘ Prodrome ’ (1828), was the first to give the fossils specific names; Burtin’s is cited as Cocos Burtini, and Parkinson’s as C. Parkinsonis. He also includes a third species, “C. Faujasii, Faujas, Aun. du Mus. i. p. 445, t. 29.” LINN. JOURN.— BOTANY, VOL. XXX. L 144 MR. A. B, RENDLE’S REVISION It is impossible to say what Faujas’s fruit was, probably not a Nipzdites; Jussieu, Desfontaines, Lamarck, and Thouin, who saw specimens, noticed a great resemblance to fruits of the Areca Palm. Brongniart also includes among Monocotyledons of unknown family a monotypic genus of his own, Pandano- carpum, “very common in the isle of Sheppey.” From the de- scription (p. 135), his P. oblongum might be our Nipadites umbonatus or N. lanceolatus, or perhaps a collective species including all the more elongated fruits. He says it has the greatest analogy with the fruits of Pandanus, and scarcely doubts its belonging to this ora closely allied genus. In the ‘ Lethea Geognostica’ (1837), a mere bibliographical work, Bronn substitutes Cocites for Cocos; while Endlicher (‘Genera Plantarum,’ p. 257) calls it Burtinia, a new geuus of fossil Palms. Brongniart’s Pandanocarpum Endlicher takes up (ib. p. 244), but definitely locates at the end of Pandanex as “ Pandanea fossilis,” immediately after Phytelephas and ipa, which are “ genera Pandaneis affinia.” The genus Wipadites was founded by Bowerbank in 1840, when appeared the first and only published part of his ‘ History of the Fossil Fruits and Seeds of the London Clay.’ He describes thirteen species, one of which, V. Parkinsonis, is Brongniart’s Cocos Parkinsonis ; while he suggests that WV. wmbonatus may be Pandanocarpum oblongum, Brongn. The author shows reason for rejecting the name Pandanocarpum in favour of Nipadites, which emphasizes the relation of the fossils to the recent genus Nipa. So near, in fact, is this relation, that one is tempted to follow Ettingshausen, who, in a list in the ‘ Proceedings of the Royal Society, xxiv. p. 393, reduces all the species to Nipa. The only consideration which leads me to keep up the paleontological suffix is the uncertainty which must always attach to the specific diagnosis of fossils based, as it often necessarily is, on material which would be deemed quite inadequate in the case of recent plants. As Bowerbank points out, the fruits of Mipadites are never ageregated into several seéded portions, as in Pandanus; but closely resemble those of Nipa fruticans, a dwarf palm of some- what doubtful affinity, frequenting the brackish estuaries and littoral marshes of India and Malaya as far as Borneo, New Guinea, and the Philippines. Both are drupes with a thin epicarp, fibrous mesocarp, and somewhat indurated endocarp. OF THE GENUS NIPADITES 145 In the recent fruits, just beneath the epicarp is a dense layer of thin branching longitudiual fibres interspersed with a few thicker, then a looser spongy tissue with thicker fibres running length- wise and also protruding radially from the endocarp. This structure is best seen at the angles where the mesocarp is thickest ; on the broader surfaces the mesocarp is often very thin and the spongy layer absent. I have been able to make out a similar arrangement in many of the fossil fruits. Bowerbank did not recognize the hard endocarp ; but though often the fibrous coat only is perceptible, an innermost more resistant layer can sometimes be seen. In the specimen figured (Pl. VII. fig. C), e indicates the dense endocarp from which the fibres are seen radiating outwards. The single seed is replaced by mineral matter ; and on the surface of the cast may be found remains of the fibrous testa. The names Cocos and Cocites, implying attinity with the modern coco-nut, are undesirable, as there is no indication in the fossil, by pores or otherwise, of the presence of three carpels. Subsequent workers have shown, and Bowerbank himself seems to have realized, that he made too many species. Ad. Brongniart, in the ‘ Tableau,’ while recogaizing that his Pandanocarpum aud Cocos Burtini beloug to Nipadites, at the same time points out that some of Bowerbank’s species are founded on too slight difterences, which may be individual variations or the result of the degree of maturity of the fruit or its position in the in- florescence. In the diagnosis of species it is important to remember that in the modern Nipa fruticans the fruits are densely packed in a spherical head, and hence not only is there considerable variety in the number of sides and the development of the angles, but, as many fruits never reach maturity, all stages of abortion may be found, especially near the base of the head, compressed and dis- torted by the growth of the ripening drupes. There seems no reason to doubt that the fossils were borne in a similar way, as a corresponding variety in shape is evident in one and the same species. Any one who will look at a head of ipa fruits will see at once how several species might be described from it. IL have seen a number of unripe or abortive fruits taken from a head kindly sent by Mr. Ridley from Singapore ; aud a comparison with L2 146 MR. A. B. RENDLE’S REVISION Bowerbank’s figures will show that some of his names, Mipa- dites clavatus for instance, might be accounted for in this way. I have seen no fossils at all like ripe fruits of Wipa fruticans, which has always the broad, often flattened or even incurved apex witha central umbo; and I might point out that the shape of the upper part of the fruit, whether rounded, obtuse, or pointed, is a valuable character for diagnosis, as it is not influenced by compression by surrounding fruits, the top portion being quite free. In his ‘ Synopsis’ (1845) and ‘Genera et Species Plantarum Fossilium’ (1850), Unger follows Bowerbank, and quotes his thirteen species of Nipadites under the Order Pandanex, while under Palmew he cites Endlicher’s Burtinia. Ettingshausen, on the other hand, in the ‘ Sitzungsber. Akad. Wiss. Wien’ (1852), viil., reduces eleven to a single species, which he calls WVipadites Bowerbankii, while the remaining two, VV. semiteres and N. pyra- midalis, he unites as N. semiteres, but in a list published in the Proc. Roy. Soc. 1879, he rejects WV. Bowerbankii, making seven of the eleven species formerly included therein syno- nymous with his Mipa Burtini, restoring two others, and redu- cing two more to 4 single species. He, moreover, calls them all Nipa. In the Ann. Sci. Nat. (Botany), 1862, Saporta describes, from the lower lignite near Beleodéme and Vallée de Vede in the South-east of France, two fruits which he calls Carpolithes and places as “ Monocotyledone ?, incerta sedis ;” but in a subsequent paper in the Mém. Soc. Géol. he makes them congeneric with the Mipadites of the London Clay. I have seen no specimens, but his figures bear no resemblance to fruits of this genus ; more- over, they are not erect, but have evidently been attached at the side of the base. Watelet has described a species from the Paris basin, which he includes, I think rightly, in Mipadites. The latest and most complete revision is that of Schimper in bis ‘Traité Paléont. Végétale.’ Six species are recognized : NV. Burtini, including the large Belgian specimens and six of Bowerbank’s species ; WV. Parkinsoni (sic), including NV. lanceo- latus, Bowerb.; NV. semiteres, Bowerb., with WV. pyramidalis as a synonym, following Ettingshausen; Saporta’s two species and Watelet’s one. Three of Bowerbank’s, one of which, WV. ellip- ticus, iS rather distinct, are omitted. Schimper probably never OF THE GENUS NIPADITES. 147 saw the London-Clay fruits, or he would not have included under his diagnosis of W. Burtini species so obviously distinct as NV. giganteus and N. cordiformis. The genus Paleokeura was founded by Massalongo in 1858, in a short paper entitled “ Plante fossiles nove in formationibus tertiariis Regni Veneti,” from specimens found near Breonio in the Province of Verona. He first referred them to Mipadites ; but in his published account gives reasons for separating them from that genus. Some of these are insufficient, and others are based on the imperfect description of the seed given by Bower- bank. Thus, he says, it cannot be included in Nipadites, as it has not an umbonate apex, the base is not lacerated and hollowed, nor is the endocarp thin and fibrous, but stony; while the seed does not show in section the rows of cells radiating from the centre, nor the central embryo, described by Bowerbank, the embryo being basal. The seeds in the Sheppey fruits have, however, been entirely replaced by mineral in which I can find no indication of radiating cell-rows or a central embryo; on the contrary, the frequently hollowed base indicates the position occupied by a basal embryo; while the stony endocarp described by Massalongo is present in many of the Belgian specimens of V. Burtini and others. Finally, the umbonate apex is not universal, while the contracted entire base resembles that of fresh Nipa fruits. The description is accompanied by admirable figures of the fruit and seed, which correspond well with some from the Brussels Eocene. Schimper does not mention Palgokeura; but it is included by Meschinelli and Squinabol in their recently published ‘ F.ora Tertiaria Italica ’ (1893). In my revision I have carefully compared all the specimens at the British Museum, including Bowerbank’s collection, a good series from the Belgian beds, and anumber which Mr. Shrubsole, of Sheerness, has kindly collected for me; also a few from the London Clay preserved in the Jermyn Street Museum. Unfor- tunately, owing to their strongly pyritized condition, the London- Clay fossils soon “shoot their salts,” as Dr. Parsons says, and crumble on exposure to air. I was induced to look intothe genus byan interesting discovery, by my friend Mr. Clement Reid, in the Bracklesham Beds, at Bracklesham and at West Wittering, Sussex, and more recently at Hengistbury Head, near Christchurch, Hants, of some fine specimens as large as the Belgian ones. Unhappily, it seems 148 MR. A. B. RENDLE’S REVISION impossible to keep them for any length of time, as they consist of a carbonized shell filled with sand, which rapidly crumbles on being taken from the damp sea-beach while the carbon film cracks and peels. One brought by Mr. Reid is preserved at Jermyn Street, and another, which I brought from West Wittering, is at the British Museum. Of the latter, Mr. Worthington Smith made careful natural-size drawings immediately on its arrival, and as he has sketched both the upper and lower surfaces as well as the sides, we have a fair record of the specimen which, in the efforts to preserve it, has become somewhat broken. One result obtained, besides a more satisfactory correlation of species and synonyms than hitherto realized, is the extension of the area of the genus in its largest species, V. Burtini, which, previously recognized only from Sheppey and the Brussels Eocene, is now, by Mr. Reid’s discovery, pushed westward to the west of Hampshire and, in Massalongo’s Paleokeura, eastward to North- east Italy. The conditions under which the fossil fruits are found show that, like their modern representatives, the palms grew in the rich brackish mud of the shore or river-mouth. Associated with the Sheppey fruits are remains of crabs, freshwater turtles, and a great variety of fruits and seeds, with some stems and branches ; such a collection of marine and freshwater organisms as is cha- racteristic of the delta of a large river. The great diversity of forms and the water-worn condition of many of the specimens indicate an assemblage brought together from a large area. On the other hand, the appearance of the fruits which I saw laid bare by the receding tide on the Sussex coast argues strongly for the view that they had grown close by. So uniform was their size and condition, that they might have fallen from one tree, and having germinated almost where they fell, left only an empty shell now filled by sand. Herein, too, they differ from all other specimens I have seen. The mineral cast of the interior of the London Clay and Belgian fruits which replaces the seed was evidently formed by the filtration of dissolved mineral matter through the pericarp, while the Sussex specimens are empty shells filled with sand pushed in from outside through the aperture left in germination. The absence of teredo-borings and water-wear favours this view ; but if this be the case, it is strange that there was no trace of the stem or leaves of the palm, which no one seems to have found in any of the localities. 'Water-worn OF THE GENUS NIPADITES, 149 drift-wood with coniferous and dicotyledonous structure occurred along with the West Wittering specimens. The Belgian and Italian fruits must have floated some time, as they are often much bored, while the associated fossils have a somewhat more marine character than those of the London Clay. In every case there is ample evidence, as Bowerbank, Ettings- hausen, and Schimper have demonstrated, from both the fossil fauna and flora, of a tropical or at least subtropical climate. Nipapites, Bowerbank. Nipadites, Bowerbank, Fossil Fruits and Seeds of the London Clay (1840), p. 1; Unger, Synopsis Plantarum Fossilium (1845), p. 180, Genera et Species Plant. Foss. (1850) p. 327; Ad. Brongniart, Tableau, (1849), p. 88; Ettingshausen, in Sitzungsb. Mat.-Nat. Cl. Akad. Wiss. Wien, viii. (1852) p. 491; Schimper, Traité Paléont. Végét. ii. (1870-72) p- 479; Watelet, Description des Plantes Fossiles du Bassin de Paris (1866), p. 91. Cocos, Burtin, Oryctographie de Bruzelles (1784), p. 118, t. xxx. A; Parkinson, Organic Remains (1804), p. 457 ; Ad. Brongniart, Prodrome (1828), p. 121. Pandanocarpum, Brongn. Prodrome, p. 138; Endlicher, Genera Plan- tarum (1836-40), p. 244. Cocites, Bronn, Lethea Geognostica (1837), i. p. 861. Burtinia, Endlicher, 1. c. p. 257; Unger, Synopsis, p. 187, Gen. et Sp. p- 339; Massalongo, Synopsis Palm. Fossil., in Lotos, ii. (Prague, 1852), p. 205 *. Paleokeura, Massalongo, Plante Foss. Nove (Verona, 1853), p. 13, “Un Nuova Genere di Pandanee Fossile’ (1853), p. 21; Meschinelli and Squinabol, Flora Tert. Ital. (Padua, 1893), p. 160. Nipa, Ettingshausen, in Proc. Roy. Soc. xxix. (1879) p. 393. Drupz lanceolatx, ovate, obovatie, vel cordiformes, 4-5- rarius 3- vel 6-angulares, basi truncate et sepe lacerw, epicarpio tenue levi, mesocarpio fibroso, endocarpio duro, monosperma, embryone basali. Nipapires Burtrni, Ad. Brongn. Tableau, p. 88; Lyell, in Quart. Journ. Geol. Soc. viii. (1852) p. 345, t. xix. figs. 1 & t. xx. figs. 7 & 8; Schimper, partim, /. c. p. 479. (Plate VI.) Cocos, Burtin, 1. c. Cocos Burtini, Brongn. Prodrome, p. 121. Cocites Burtini, Bronn, l. c. * I have ignored Endlicher’s Burtinia as, first, the evidence as to date of publication is not exact, the title-page of the ‘Genera Plantarum’ bearing the date 1836-40, while that of the Preface is 1841, and secondly, his very brief diagnosis will not apply to any species of Nipadites. 150 MR. A. B. RENDLE’S REVISION Burtinia cocoides, Unger, Synops. p. 187, Gen. et Sp. p. 339; Massa- longo, Synops. p. 206 ; Ettingshausen, in Abh. Geol. Reichsanst. (1852) i. Abth. 3, nr. 1, p. 9. N. giganteus, Bowerb. Foss. Fr. Lond. Clay, p. 22, t. vi. f. 1; Unger, Gen. et Sp. p. 329. N. turgidus, Bowerb. partim, J. c. p. 21, t. v. f. 1; Unger, 1. ¢. N. Bowerbankii, Ettingshausen, partim, l. c. p. 8, et in Sitzungsb. Akad. Wiss. Wien, viii. (1852) p. 491. Nipa Burtini, Ettingsh. partim, in Proc. Roy. Soc. xxix. (1879) p. 393. Paleokeura Pellegriniana, Massal. Plant. Foss. Nov. p. 15, et Nuova Gen. (with four plates) ; Mesch. et Squin. Fl. Tert. Ital. p. 161. Drupa magna, semipedalis, obovata, seepius late obovata, vel fere suborbicularis, jam quadrangularis, faciebus duabus oppositis latis subcomplanatis vel convexis, duabus angustis sejunctis, jam faciebus tribus subequalibus, vel una majore; apice obtuso in” terdum umbonato; basi truncata; angulis prominentibus sepe alatis; epicarpio levi; mesocarpio fibroso, hoc in faciebus vix unilinea, angulis autem multo crassiore ; endocarpio, quum rarius apparet, duro, fere unilineo. Size 53-7 inches long, 4-6 inches broad, 14-4 inches thick. There is often a marked depression below the apex, and the prominent angles are often flattened into a thin broad wing. Many of the Belgian and the Italian specimens are much bored by Teredinex. The cast of the interior, which takes the place of the seed, varies with the external ‘shape of the drupe, which it almost equals in breadth and thickness, being more or less irre- gularly oval or oblate with three or more, usually rounded, angles. On the surface are often traces of the irregular fibrous coat like that of recent seeds. Localities. Sheppey (London Clay). Near Selsea and West Wittering, Sussex. Hengistbury Head, Hants (Bracklesham). Woluwe and Schaerbeek, near Brussels (Middle Eocene). Breonio, Verona (Eocene). The Sheppey fruits are strongly pyritized, those from the south coast carbonized, and those from Belgium often enclosed in concretionary nodules. Nreapires Parxrnsonts, Bowerb. l. c. p. 16, t.iv.; Brongniart, Tableau, p. 88 ; Unger, Gen. et Sp. p. 328 ; Lyell, in Quart. Journ. Geol. Soc. viii. (1852) p. 346, t. 19. figs. 5 & 6. (Pl. VII. figs. B, C.) A species of Almond, Douglas, Dissert. on Antiquity of the Earth (London, 1785), p. 25, t. v. OF THE GENUS NIPADITES. 151 Cocos sp., Parkinson, Org. Rem. et t. vi. figs. 5-7, t. vii. figs. 1-5. Cocos Parkinsonis, Brongn. Prodr. p. 121. Cocites Parkinsonis, Bronn, Leth. Geogn. p. 861. N. turgidus, Bowerb. partim, Foss. Fr. Lond. Clay, p. 21, t. v. f. 23 Unger, Gen. et Sp. p. 329. N. Bowerbankii, Hétingsh. partim, in Abh. Geol. Reichsanst. (1852) 1. Abth. 3, no. 1, p. 9, et in Sitzungsb. Ak. Wiss. Wien, 1. c. N. Parkinsoni, Schimper, Traité Pal. Vég. p. 480. Nipa Parkinsonis, Ettingsh. in Proc. Roy. Soc. xxix. (1879) p. 393. Drupa quam precedens multo minor, late obovata, subita acuminata, basi truncata ; haud compressa ; angulis 4-6, seepius 4 “ere equidistantibus ; pericarpio, apice crasso solido, basi multo tenuiore vix unilinea ; epicarpio levi tenue ; mesocarpio cum fibris longitudinalibus sub epicarpio tenuibus et arcte confertis, tum crassioribus et sejunctis, fibrillis transversis intertextis; endo- carpio duro. A distinct species, of comparatively uniform size; the most perfect specimen measured 34 inches in length, 23 inches in breadth, and 53 inches in thickness, and the smallest 23 inches, 2 inches, and 1} inch respectively. The strong longitudinal fibres of the mesocarp are characteristic; they are 3-13 line apart, the intervals being filled with closely-packed transverse fibrils. The cast of the interior is large, from 14-2 inches long, slightly broader, and from 1-1} inch thick, very broadly obovate, with a more or less acuminate apex and 4—5 more or less flattened sides with rounded angles. A fragment of a drupe not quite ripe (fig. C), in which the structure of the interior of the pericarp is preserved, shows the hard endocarp with the fibres passing from its outer surface, as obtains in Mipa fruticans. Localities. Sheppey ; Schaerbeek, Woluwe. The specimen figured by Burtin, Oryct. Brus. fig. C, may be an internal cast of the fruit of this species. NIpaDITES LANCEOLATUS, Bower. 1. c. p. 16, t. iil. fig. 7; Unger, Gen. et Sp. p. 328; Lyell,in Quart. Journ. Geol. Soe. vil. (1852) p. 346, t. 19. figs. 8 & 4. (PI. VIT. fig. A.) N. crassus, Bowerb. l. c. p. 12,t. ii. figs. 4 & 5; Unger, lc. p. 328. N. Bowerbankii, H¢tingsh. partim, 1. c. N. Burtini, Schimper, partim, /. c. p. 479. N. Parkinsoni, Schimp. partim, /. c. p. 480. Nipa lanceolata, Ettingsh. in Proc. Roy. Soc. xxix. p. 393. N. Burtini, Ettingsh. partim, ibid. 152 MR, A. B. RENDLE’S REVISION Drupa lanceolata vel ovato-lanceolata, subacuta vel subobtusa, epicarpio levi, quam in Parkinsoni crassiore; mesocarplo cum fibris longitudinalibus sub epicarpio arcte confertis, tum crassi- oribus sed minus quam in precedente et minus sejunctis, fibrillis oblique transversis intertextis. The specimens vary in size from 23-3} inches long, 13-14 inch broad, and 5-17 lines thick, the thinnest ones being either unripe or very much compressed. As in the last species, the pericarp forms a thick, generally obtuse apical portion, while below it remains thicker than in that species ; the epicarp is also thicker, measuring about } line. The outer layer of mesocarp with the closely-set finer longitudinal fibres extends in the middle of the nut fora depth of about } a line, but is thinner below; the thicker fibres are more rarely exposed, when they are seen to be less coarse than in V. Parkinsonis, and only from 3-3 line apart. The cast of the interior is, in the broader specimens broadly ovate, in the narrower oval. Localities. Sheppey (London Clay). Schaerbeek (Middle Eocene), fide Lyell in Quart. Journ. Geol. Soe. viii. p. 346. NipapiTtEs umBonatus, Bowerb. Foss. Fr. Lond. Clay, p. 9, t. 1; Unger, Gen. et Sp. p. 827. N. clavatus, Bowerb. 1. c. p. 15, t. iii. figs. 4-6 ; Unger, l. c.; ? Lyell, 1. c. p- 346, t. xix. figs. 3 & 4. N. Bowerbankii, Ettingsh. partim, in Abh. Geol. Reichsanst. viii. (1852) p. 491 et in Sitzb. Akad. Wiss. Wien, viii. (1852) p. 491. N. Burtini, Schimper, partim, Traité Pal. Vég. p. 479. Nipa Burtini, Ettingsh. partim, in Proc. Roy. Soc. l.c. Drupa obovata, seepe anguste, et valde compressa, vel clavata, apice obtuso vel rotundato umbonato ; epicarpio levi ; mesocarpio fibroso, fibris longitudinalibus externis densis, internis velut in NV. lanceolato. Locality. Sheppey (London Clay). The specimens are in most cases much compressed, and prob- ably immature. WV. clavatus seems to be a small form. NIPADITES CORDIFORMIS, Bowerb. 1. c. p- 18, t. ii. figs. 6-10; Unger, l. c. p. 328. N. semiteres, Bowerb. 1. c. p. 23, t. vi. figs. 2-4 ; Unger, 1. ¢. p. 329 ; Ettingsh. in Sitz. Akad. Wiss. Wien, viii. (1852) p. 492; Schimper, 1. c. p. 480. N, pyramidalis, Bowerb. 1. c. p. 24, t. vi. figs. 5 & 6; Unger, 1. ¢. p- 329, N. Bowerbankii, Ettingsh. partim, 2. c. OF THE GENUS NIPADITES. 1538 N. Burtini, Schimper, partim, 7. c. p. 479. N. pruniformis, Bower. partim, 1. ¢. p. 14, t. ii. fig. 12. Nipa Burtini, Ettingsh. partim, in Proc. Roy. Soc. xxix. p. 393. Nipa semiteres, Ettingsh. ibid. Drupa minor cordiformis apice subacuto, obtusiusculo vel sub- acuminato, spe vix longior quam lata, interdum angustior; epi- carpio levi; mesocarpio cum fibris densis longitudinalibus. The specimens vary in size from 1-2 inches long, 7-14 inches broad, and 6-10 lines thick. Many are so strongly pyritized and others so worn, that traces of the epicarp are rarely to be seen ; transverse fibrils are nowhere exposed. Bowerbank’s figured specimen of NV. semiteres (Foss. Fr. Lond. Clay, t.iv.f. 2) I have seen in the British Museum ; it cannot, I think, be separated from N. cordiformis. Here, too, I place N. pyramidalis, Bowerb., the figured specimen of which is also extant ; its regular pyramidal shape may be due to its position at the top of a head, or somewhere where pressure is evenly dis- tributed on all sides. NV. pruniformis, Bowerb., probably represents unripe fruits of this or the next species, or both. Locality. Sheppey (London Clay). NipavitEs ELLIPTICUS, Bowerb. 1. c. p. 11, t. ii. figs. 1-8; Unger, Gen, et Sp. p. 827. N. Bowerbankii, Ettingsh. partim, in Sitzb. Akad. Wiss. Wien, viii. (1852) p. 492. ?.N. pruniformis, Bowerb. partim, 1. c. p. 14, t. ii. f. 11. Nipa elliptica, Ettingsh. in Proc. Roy. Soc. xxix. p. 393. Drupa minor, elliptica, compressa, sepe in facie una complanata, apice obtuso; mesocarpio cum fibris longitudinalibus densis, fibrillis transversis subjacentibus in solo specimine apparentibus. The fruits vary in size from 13-1} inches long, 13-15 lines broad, and 5-7 lines thick. Locality. Sheppey (London Clay). NreapitEs Heserti, Watelet, Pl. foss. du Bassin de Paris, p: 92, t. xxix. figs. 2 & 3; Schimper, Traité Pal. Vég. p. 480. Drupa majuscula, late elliptica, quadrangularis, compressa ; epicarpio levi; mesocarpio fibroso. Judging from the figure and description, near NV. elliptica, Bowerb., but more globose. Locality. Issy, near Paris (Calcaire Grossier). 154 REVISION OF THE GENUS NIPADITES. Species non satis nota. Nrrapites acutus, Bowerb. 1. c. p. 14, t. iii. fig. 1 ; Unger, l. ¢. p 328. N. Bowerbankii, Ettingsh. partim, /. c. Nipa Burtini, Ettingsh. partim, /. c. Drupa basi major, superne deminuta, apice acuto; mesocarpio cum fibris densis longitudinalibus, crassioribus interspersis. The specimen figured by Bowerbank is nearly 24 inches long, and 14 inch broad near the base; it is much worn, and shows stronger longitudinal fibres with the intervening spaces filled with finer closely set ones with the same direction, but no trace of transverse fibrils. I can find no second specimen; and the one described is, in my opinion, too misshapen to form the type of a distinct species. Locality. Sheppey (London Clay). Species excludende. NIPADITES PROVINCIALIS, Saporta, in Mém. Soc. Géol. France, 2. vill. (1868) p. 303, fig. 2; Schimper, Traité Pal. Vég. p. 480. Carpolithes provincialis, Saporta, in Ann: Sc. Nat. (Bot.) sér. 4. xvii. (1862) p. 200, t. i. fig. 6. Nipaprres curtus, Saporta, in Mém. Soc. Géol. 1. c. ; Schimper, L. ec. Carpolithes curtus, Saporta, I. c. p. 201, t. i. fig. 7. EXPLANATION OF THE PLATES. Priate V1. Nipadites Burtini, Brongn, From West Wittering, Sussex. Natural size. Puate VII. A. Nipadites lanceolatus, Bowerb. The lower part broken away disclosing the seed-cavity. B. Nipadites Parkinsonis, Bowerb. C. The same in longitudinal section, showing at e fibres passing outwards from the hard endocarp, All natural size, from Sheppey specimens in Bowerbank’s own collection the Geological Department of the British Museum. Lrendle. A.Bro BURTINI, NIPADITES * Susse from 7, AX P| Bot.Vor _ JOURN. n hy H , NIPAD A ANATOMY OF A PLANT FROM SENEGAMBIA. 155 On the Anatomy of a Plant from Senegambia. By ANNIE Lorrain Surra. (Communicated by G. F. Scorr ELuior, M.A., F.L.S.) [Read 15th June, 1893, ] (Puate VIII.) THe unnamed plant described below was brought home from Tropical Africa by Mr. Scott Elliot, who found it on dry sandy laterite, between Sulimania and the Niger, about a quarter of a mile before the Farana ford. He very kindly placed it, along with his notes, at my disposal for investigation. Our object has been to see if anatomical characters would throw any light upon the affinity of the plant, which, unfortun- ately, had neither flower nor fruit to guide us in its determination. The plant is one of striking facies and likely to imprint its appearance on the memory, but no systematist at the British Museum or at Kew has recognized it. The guesses as to its affinity from macroscopic characters alone varied from Lycopo- diacee and Conifere on to Acanthacee, Melastomacee, and Lhyme- leacee. I do not propose to give a very detailed account of the anatomy, since dried specimens do not afford adequate material for such a task. I have, however, made out enough to indicate the interesting character of the subject of examination. On Pl. VIII. fig. 1, I have represented part of the plant reduced by onethird. The root-stock, which is broken off just below the ground, isastout structure with lateral rootleis, and is evidently persistent from year to year. The primary branches are radical and grow from 9 inches to a foot in height ; they have no annual rings, and can be of only one year’s growth. The leaves are de- cussate with a marked stipular line ; the broad bases embrace the stem, and they closely overlap the pair of leaves immediately above (fig. 1). ‘The secondary slender branches have grown about half the height of the primary branches, they are inserted a little to the side of the axil of alternate leaves with a divergence of 4, but only those near the tip of the parent branch attain to full development. These secondary branches are round or only slightly flattened ; the primary branches are flattened to an almost rectangular shape. The older leaves are broad at the base, cordate-acuminate, with parallel veins and a strong midrib ending in ahard point. The leaves of the younger branches are narrower and less developed, aud show three distinct veins, especially after 156 MISS ANNIE L. SMITH ON THE ANATOMY maceration. The veins are interlaced by strands of sclerenchyma (fig. 2), and curve in at the base of the leaf, entering the stem together as one bundle. The whole plant appears to be adapted to desert growth and even to the occurrence of desert fires. The natives are in the habit of burning down the grass every year, and this specimen bore distinct traces of scorching ; the soft tissue of the broad leaves is burned, except the part protected by the leaf below (fig. 2), the sclerenchyma is laid bare, but the leaf as a whole remains ; the secondary branches and leaves, which are unharmed, are of more recent and evidently very rapid growth. The plant was gathered after the first two or three tornados of the rainy season, but before the rainy season proper, i. ¢. just when everything is starting growth. The larger internodes and smaller leaves of the crowded branches indicate very clearly a change in the con- ditions of life. The stipular sheath of the stem is strongly sclerotic ; the same arrangement of cells extends up the back of the leaf over the midrib (Pl. VIII. fig. 4), affording additional protection against external influences. The cortex is crushed, and as a rule the normal phloem is very much reduced. The part of the stem I have chosen for figure 5 showed the phloem more distinctly than atany other point. The outer xylem consists of wood fibres and vessels with bordered pits; the vessels of the inner xylem are spiral. Just inside these, round the periphery of the pith, are groups of phloem (fig. 5) belonging to the leaf-trace bundles. This tissue passes out to form the bi-collateral bundle of the leaf, and remains active in the inner protected portion of the leaf, the normal phloem being much crushed (fig. 4). The smaller bundles, as is often the case, do not seem to be bi-collateral. Scattered irregularly through the xylem is a series of phloem islands developed centrifugally from the cambium (figs. 5 & 6). This is plainly seen from the radial position of the cells of the phloem islands, which consist of sieve-tubes and companion cells with parenchyma. There are also some islands in process of formation which show no trace of internal cambium. The medullary phloem groups show no cambium; the central pith is composed of large pitted cells densely filled with needle crystals of calcium oxalate. I have not drawn any longitudinal sections owing to the state of the material ; callus had in each case formed over the sieve- w Bot. Vou KAA. FITS Linn. Soc. Jou au — eROM SENE GAMBIA Berjeau & Hiqaley are NEW PLAN r OF A PLANT FROM SENEGAMBIAs 157 plates obliterating the perforations, and the medullary phloem was much shrivelled. Dr. Scott, to whom I am much indebted for confirming my anatomical observations, thinks it would be almost impossible to detect perforations in material so long gathered and so dry. I failed to do so with a magnification of 1300 diameters. The lateral roots show interxylary phloem islands ; I have been unable to see very young stages, but I do not think they lie within the primary wood as in Chironia and Strychnos *. According to our present knowledge, this peculiar anatomical structure considerably narrows down the systematic limits within which the plant must come. To decide as nearly as possible what might be its affinity, I have examined plants belonging to the orders Loganiacee, Acanthacee, Peneacee, Gentianacee, and Melastomacee, but without such definite success as would warrant my placing it in any of them. In some respects this essay may be instructive in showing the limits of plant anatomy in affording characters of value to the systematist. My anatomical work has been done in the Laboratory of the Botanical Department of the British Museum, and I have to thank the officials both for advice and material. EXPLANATION OF PLATE VIII. Fig. 1. Part of plant, reduced one third, 2. Broad leaf, slightly enlarged. 3. Narrow leaf, slightly enlarged. 4, Transverse section of leaf mid-rib, x 150. 6. Transverse section of part of primary branch: pf.i., phloem islands; ph.g., phloem groups: x 150. 6. Transverse section of phloem island, x 450. * Scott and Brebner on Internal Phloem in the Root and Stem of Dicotyledons, Annals of Botany, vol. v. p. 281. 158 MR, W. B. HEMSLEY ON THE FLORA Tue Frora oF THE Tones or Frrenpuy Isnanps, with Descrip- tions of and Notes on some New or Remarkable Plants, partly from the Solomon Islands. By W. Borriya Hemstey, A.LS., F.R.S. (Contributed by permission of the Director of the Royal Gardens, Kew.) [Read Ist June, 1895. ] (Puates IX.-XT.) 1. Introductory Note on the Tonga Flora. AurnovueH I have drawn up as complete a list as I could of the vascular plants hitherto collected in the Tonga Islands, the few preliminary observations I have to make on the flora of that group relate chiefly to a collection of dried plants, mostly from the island of Eua, made by Mr. J. J. Lister during the years 1889 and 1890. As Mr. Lister is present and has promised to give you some particulars of the position and composition of the islands, and of the local conditions under which plant-life exists*, ° it will be sufficient for me to point out that the Tonga Islands lie a little south-east of Fiji, that Eua lies to the south-east of the comparatively very low (about sixty feet in its highest part) though much larger island of Tongatabou, and that it has a central ridge of high ground which attains a height of upwards of 1000 feet above the level of the sea. Tongatabou was touched, I believe, by Cook on all three voyages, for there are Tongan plants in the British Museum collected by Banks and Solander, by Forster, and by Nelson. Various other expeditions have touched there, including Wilkes’s United States and the ‘ Chal- lenger’ expeditions ; and as the late Professor Moseley’s collec- tion made there on the voyage of the ‘ Challenger’ has not been published, I propose including it in the Enumeration which follows these observations. The vegetation of Eua differs from that of Tongatabou in having some peculiar species, at least so far as our present data go; and it also includes a considerable number of Polynesian forms not found in the low coral islands. The violaceous Meli- cytus ramiflorus is one of the most interesting plants found in Eua by Mr. Lister. It is a common New Zealand shrub or * Mr. Lister has kindly communicated the substance of his remarks, which will follow my observations on the flora. OF THE TONGA OR FRIENDLY ISLANDS. 159 small tree, which has also been collected in the Kermadec Islands and in Norfolk Island ; and Mr. Lister’s discovery gives the genus a considerable north-eastern extension. Besides the New-Zealand species there are two or three undescribed ones from New Caledonia, but I am not aware that the genus exists in Fiji. Among the Mangroves collected by Mr. Lister is a small specimen which Dr. Stapf has identified with the American Rhizophora Mangle, a species common on the shores of tropical America and also found on the West - African coast. I have compared it myself,and I agree with Dr. Stapf that it is much more like the American than either of the Old World, or eastern species; but it may have been accidentally introduced with ballast. With the permission of Mr. Thiselton Dyer, the Director of Kew, I am able to show you this and a selection of other interesting objects from Polynesia. Several of the plants collected in Eua have not been exactly matched, but my Kew colleagues and I have not ventured to describe many of them because we are unwilling to add unneces- sarily to synonymy, and much research would be involved in thoroughly working them out. Ardisia Listers and Cyrtandra Listeri are inconspicuous species of their genera, but apparently quite distinct from anything described. Graptophyllum Sipho- nostena had previously been found in Fiji and Samoa, and Dr. Stapf has now fully described it. The plant I have doubtingly referred to the Monimiaceous genus Hedycarya is in too young a condition for satisfactory determination. Mr. Lister’s Tongan collection numbers nearly 200 species of vascular plants, of which about 100 were not previously repre- sented in the Kew Herbarium from those islands; nor are they recorded in Seemann’s ‘ Flora Vitiensis’ as being in the British Museum. Moseley collected 76 species of vascular plants in Tongatabou ; and the following enumeration of all the plants collected by Mose- ley, Lister, and recorded by Seemann in his ‘Flora Vitiensis,’ contains 337 species belonging to 234 genera and 82 natural orders. I had not time to work out their distribution in detail previous to reading this paper; but Mr. J. F. Jeffrey has re- written the Enumeration and collected the details of the distri- bution of each species, so far as it could be done at Kew, from manuscript lists and published records, and constructed the table upon which my remarks, at the end of the Enumeration, on the LINN. JOURN.—BOTANY, VOL. XXX. M 160 MR. W. B. HEMSLEY ON THE FLORA composition of the flora are based. A close scrutiny of the Kew and British Museum herbaria would probably yield a few additional species. 2. Notes on the Position, Geology, and Aspects of the Vegetation of the Tonga Islands. By J. J. Lister, Esq., M.A. The Tonga or Friendly group consists of a number of small islands scattered along an axis whose direction is N.N.E. and S.S.W. The length of the group is about 200 miles. Some 400 miles to the W.N.W. are the large islands of Fiji, and the Samoa islands are about 300 miles away in the direction of the axis northward. Tongatabou*, the largest island of the group, is situated at the southern end, and is about 1000 miles from Auckland in New Zealand. The 20th parallel of south latitude passes through the middle of the group. The group is composed in part of a line of volcanoes which traverses it lengthwise. Some of these are still active, while others are only the broken-down remnants of volcanoes which have been long extinct. In 1885. submarine eruption occurred, and a mound of ashes f was built up in the interval between two of the existing islands; but this has since been largely worn away by the sea. The other islands of the group are for the greater part formed of reef-limestone. In some of them an underlying basis is ex- posed, formed of volcanic tuffs laid out beneath the sea, and in a few cases the islands consist entirely of the latter formation. Movements both of elevation and subsidence have occurred, but the more recent tendency has been, in the main at least, one of elevation. Vavau, a limestone island in the northern part, is about 500 feet high; and Ewa, in the south, is over 1000 feet. Some of the voleanoes attain greater heights. The other islands are for the most part little elevated and flat. Generally it may be said that all the high ground of the group is either the summits of volcanoes, active or extinct, or composed of reef-limestones, which were, of course, formed beneath the sea. Soundings show that the islands are situated at the northern end of an extensive plateau which reaches far to the southward * Designated Amsterdam Island in some of the older charts.—W. B. H. t Falcon Island, it has been named.—W. B. H. OF THE TONGA OR FRIENDLY ISLANDS. 161 in the direction of New Zealand. In the interval between its southern limit and those islands is another shoal area, on which the small volcanic Kermadec Islands are situated. It has been pointed out that the volcanoes of the Kermadec and Tonga Islands continue the direction of those of the Taupo zone of volcanoes of the North Island of New Zealand. To the east of the Tonga plateau the sea-bottom descends to 4530 and 4429 fathoms, the greatest measured depths in the southern hemi- sphere. Between Tonga and Fiji there is a depth of 1500 fathoms. At the close of the last century the Tonga Islands appear to have been thickly populated and extensively cultivated. At the present time the population has greatly decreased, and much of the cultivated land has passed into a wild state. The higher part of the island of Hua presents a vegetation of a different character from that met with in the adjoining islands. This island, which is rather more than twelve miles long, is composed of a basis of volcanic tuffs overlain by reef-formation. On the high ridge of the island the limestone appears in isolated masses and pinnacles, which stand out above the red volcanic soil, while in the lower part it forms wide terraces at different levels, the upper ones ending in lines of inland cliffs. The eastern side, presented to the trade wind, rises abruptly from the shore in ranges of limestone cliffs alternating with steep slopes, covered with a dense wind-swept forest-growth. On the western side the surface is scored by watercourses which traverse the volcanic basis. The streams, on meeting the limestone ter- races, dip underground and pursue a subterranean course to the sea. This aspect is in marked contrast with the eastern. In- stead of the dense wind-swept growth, there are wide open spaces alternating with belts of luxuriant bush which cover in the water- channels. The open spaces are often dotted over with Screw- Pines (Pandanus odoratissimus) with their curious prop-like air- roots, and are generally covered with a coarse herbage. In the volcanic part the soil is often bare, and shows bright red, yellow, or brown from the decomposing rock. When it is covered the tracts of voleanic formation can be recognized by the plants which grow on them: Afelastoma denticulatum and some of the Ferns are, I believe, never found elsewhere. The great crowns of the Tree-Ferns are conspicuous features when looking down M2 162 MR. W. B. HEMSLEY ON THE FLORA on a tract of bush from the cliffs. Along the eastern shore the Piu or Biw Palms (Pritchardia pacifica), with their great fan leaves, are very handsome objects. The trees are not generally very high, running to about 60 to 80 feet. After penetrating the outer edge of a belt of bush, festooned with creepers, Hntada scandens and various species of Ipomea, with sulphur-coloured, blue or purple flowers, there was generally no difficulty in making one’s way about. My collections at Eua were made during a visit of a fortnight in December 1889, and in two days which I spent there in June of the same year. In December I think I obtained most of the plants which were in flower at the time in the bigher part of the island, but I have no doubt that there are many plants on the island not represented in the collection. In conclusion I may perhaps be permitted to point out that while the contour of the sea-bottom, with the great southerly extension of the Tonga plateau, suggests that a closer connection may at one time have existed between New Zealand and land to the northward, the geological structure of the present Tonga islands would not lead us to expect that an ancient fauna and flora survives on them; for, as was shown above, all the high ground of the group either consists of the outpourings of vol- canoes, or has been submerged within the period during which the existing reef-limestones have been formed. Mr. E. A. Smith informs me that the land mollusks are quite of a Polynesian character, and show no special affinities with those of New Zealand. With regard to the birds, to which I have devoted some at- tention, the,same statement holds good. Compared with those of Fiji and Samoa, the birds of Tonga are few in number and very little specialized. There are no peculiar genera, and prob- ably only four peculiar species, each of which is represented by allied species, both in Fiji and Samoa. With the very peculiar bird-fauna of New Zealand there are certainly no special affinities *, J.J. Lisrer. * A fuller account of the Geology of the Tonga Islands will be found in the Quarterly Journal of the Geological Society, November 1891, vol. xlvii. p. 590, OF THE TONGA OR FRIENDLY ISLANDS. 163 3. Observations on some Remarkable Plants from the Solomon Islands. The Solomon group consists of a chain of islands, some 600 miles in length, extending in a south-easterly direction, from New Britain and New Ireland (now the New Pomerania and New Mecklenburg of the Germans), nearly to the New Hebrides. The larger islands are from 50 to 100 miles in length, and 15 to 30 miles in breadth, with elevations of 8000 feet in Guadalcanar, in the south, to 10,000 feet in Bougainville, in the north. So far as I know, none of the earlier voyagers landed in these islands ; but they were visited by D’Urville in 1838, and by Denham (H.M.S. ‘ Herald’) in 1853; and William Milne, who was attached to the ‘ Herald’ in the capacity of botanical col- lector for Kew, collected some 200 species of plants in the coast region, nearly all of which are plants of wide range. Within the last decade, however, both English and Germans have reached the mountains of the interior of some of the islands, and discovered a varied and curious indigenous flora. Dr. K. Schu- mann has published some of the novelties collected by Dr. Naumann and other Germans in the ninth volume of Engler’s ‘ Jahrbiicher’; and Dr. H. B. Guppy, in his book entitled ‘The Solomon Islands and their Natives,’ gives a rough list of the plants he collected, from approximate determinations made at Kew. This collection contained a considerable number of novelties, some of which have since been described by Dr. Beccari, Professor D. Oliver, and myself; but many of the specimens were insufficient for description, though evidently un- described. To some considerable extent this collection has been supplemented by the Rev. R. B. Comins, who brought home a small parcel of dried plants in 1890, and who has since sent to Kew two other small parcels. It is not my intention at present to attempt to give an enumeration of all the plants known from the Solomon Islands, because it would be, at best, a very small fragment of a manifestly rich endemic flora, therefore I will confine myself to directing attention to a few of the more re- markable plants contained in the collections to which I have referred. As Dr. Guppy was first in the field, I will begin with some of his novelties. Myrmecophilous plants appear to abound, espe- cially the rubiaceous genera Hydnophytum and Myrmecodia. 164 MR. W. B. HEMSLEY ON THE FLORA They are epiphytes growing chiefly on the mangrove trees and bushes on the coast. The galleried tuberous stems are some- times as much as eighteen inches long. I have here an example of each genus. A new genus of Pandanacex (which Ihave named Sararanga sinuosa, Sararang being the native name of this tree, see p. 216, Pl. XI.), a tree fifty feet high, with long, narrow, almost unarmed leaves, and a panicled spadix of white female flowers, was discovered in Fauro island at elevations of 1600 to 1900 feet, and is, indeed, the most remarkable plant of the col- lection. Dr. O. Beceari collected the same, or a closely allied species, in the island of Jobie, North-western New Guinea, and about twenty degrees west of Fauro. Although the material is incomplete I propose publishing such a description of it as I can. Cominsia Guppyi, Hemsl.*, is a new genus of Scitaminee, allied to Phrynium, and characterized by an elongated inflorescence with the flowers clustered in boat-shaped bracts, and opening one at atime. The elongated tube of the corolla, the dehiscent fruit, and the rugose corky seed with an empty cavity parallel to the curved embryo, are also characteristic. This plant was collected in Fauro Island by Dr. Guppy, and in San Christoval, in the extreme south of the group, by Mr. Comins. Proceeding to some of Mr. Comins’s plants, I will first direct attention to the Sapotaceous genus Chelonespermum, published last year (1892) in the‘ Annals of Botany,’ with illustrations from admirable drawings by Mrs. Thiselton Dyer and Miss M. Smith. The genus was really founded upon the seeds, which are, indeed, among the most curious in the vegetable kingdom. One is con- tained in each fruit, which is pear-shaped, and has a more or less fleshy pericarp, though the ovary is two-celled, with one ovule in each cell. The seed is a dorsi-ventral body, and I may explain that it is erect in the fruit with reference to these two very different surfaces. On the smooth surface or back, the pericarp is quite thin, whereas on the uneven or hilum surface it is pulpy, and of sufficient thickness to form a symmetrical fruit. The embryo fills the whole cavity of the testa, or is enveloped in a very thin film of endosperm, at least in the mature state. It has two very large, fleshy, plano-convex cotyledons applied face to face, and a short thick radicle, which probably emerges in germination through an opening in the testa, the beginning of * ‘Annals of Botany,’ v. p. 501, pl. 27. OF THE TONGA OR FRIENDLY ISLANDS. 165 which may be seen in some of the seeds near the base. I ex- tracted the embryos for examination, and afterwards stuck the testa together again to make museum specimens. I have de- scribed four species of this genus: two from the Solomon Islands, one from Fiji, collected by Mr. J. Horne, the late Director of the Mauritius Botanic Gardens, and one presented to Kew by the Rev. G. Henslow, the native country of which is unknown. According to Mr. Comins’s notes these seeds are known in the Solomon Islands by a name signifying turtle-seed; hence my name Chelonespermum. The next thing I have to notice is Lepinia taitensis, which has one of the most singularly constructed fruits in existence. It is a member of the Apocynacee, and in the flowering-stage the ovary is undivided, but soon the usually four carpels of which it is composed separate from each other except at the tips, and develop long slender stalks carrying up the seed-bearing portions in the form of across. This tree would appear to be exceedingly rare. Decaisne described and figured it as long ago as 1849 from Tahitan specimens, and, so far as I am aware, it has not been found in any locality between Tahiti and San Christoval, in the Solomon Islands, where Mr. Comins collected the present specimen, localities separated by about 3300 miles of water and small islands. I think it must be extremely rare in Tahiti, because we have no specimen in the many collections from that island at Kew, nor does it exist in any of the earlier collections in the British Museum. I will not attempt to explain the object of this curious development. I will conclude these notes with a few remarks on some other new plants from Mr. Comins’s last collection, described at p- 211. Dysoxylum megalanthum, a Meliaceous tree, is remarkable for the large size of its flowers. Crossostyles Cominsit (p. 212) is a member of the Legnotidezx, and the present species is charac- terized by having entire petals and no staminodes. Eugenia salomonensis is a very showy species of its genus, and Ruellia Guppyi must be a very ornamental plant. Finally I may mention Antiaris turbinifera (p. 215), the Humming-top tree of the natives, who make humming-tops of the crustaceous endocarp of the fruit. 166 . MR. W. B. HEMSLEY ON THE FLORA 4, Exumeration of the Vascular Plants of the Tonga Islands, together with their Distribution, Local and General. The following Enumeration is, perhaps, not quite complete, because I have not searched the herbaria systematically through ; that being a task beyond the time at my disposal. Nevertheless it is doubtless very nearly exhaustive so far as Kew and the British Museum Collections are concerned, because I have taken up every species recorded by Seemann (‘ Flora Vitiensis ) as existing in the British Museum from the group. I have included all the lists of Tonga plants in the Kew Herbarium—or rather Mr. Jeffrey, to a great extent, collected this information for me ; and I have specially examined the Kew Herbarium in many instances for plants of whose existence in the Tonga group I had no evidence, although they are known to inhabit either the Fijis or Navigator Islands, or both. Among such as I have not found are Lepidium piscidium, Forst., Kleinhovia hospita, Linn., Afzelia bijuga, A. Gray, Parinarium insularum, A. Gray, P. laurinum, A. Gray, and Flagellaria indica, Linn. But Mr. Lister particularly mentions, p. 162, that he is of opinion that he did not exhaust the botany of the island of Eua. Briefly, the sources are the British Museum and Kew Herbaria; published so far as the former is concerned, and mostly unpublished so far as the latter is concerned. The principal books used are A. Gray’s * Botany ’(1854) of the United States Exploring Expedition, 1838 ~1842 ; Seemann’s ‘ Flora Vitiensis,’ 1865-73 ; Hillebrand’s ‘ Flora of the Hawaiian Islands,’ 1888; and Drake del Castillo’s ‘ Flore de la Polynésie Francaise,’ 1893. I did not think it desirable to consult the earlier works, such as Endlicher’s ‘ Bemerkungen tiber die Flora der Siidseeinseln,’ 1836, and Forster’s ‘ Florule Insularum Australium Prodromus,’ 1786, because without a great expenditure of time it might have led to confusion in the nomenclature. The plan of the Enumeration is to give as near as possible the most generally used name of each plant, with a few of the more important synonyms, such as are employed in the works named, followed by the names of the persons who have collected the plant in the Tonga Islands, and its general distribution *. The * we - . In the distribution, “Samoa” is employed to designate the particular island and “ Navigator” the group. Thi i nc p. This group is reckoned east Tonga or Friendly Islands, Sob the tongs ot OF THE TONGA OR FRIENDLY ISLANDS. 167 collectors who contributed the material here consolidated are about twenty in number. First come the early collections de- posited mainly in the British Museum, though some of them are also represented in other herbaria. These are by Sir Joseph Banks and Dr. D. C. Solander, who accompanied Captain Cook on his first voyage in 1768-1771; Dr. J. Reinhold Forster and his son George, who were naturalists to Cook’s second voyage, 1772-1775 ; and Mr. David Nelson, who was on the third voyage in 1776- 1780. Some of the specimens in the British Museum are simply marked “ Cook,” without any indication of the voyage. Another important collection in the British Museum, of which there is alsoia set at Kew, was made by George Barclay, on the voyage of H.M.S. ‘ Sulphur,’ from 1835-1841. He was primarily a colleetor of seeds and living plants for Kew, and his letters to W. T. Aiton, the director, and J. Smith, the curator, are preserved there. The Tonga plants in the Kew Herbarium include a por- tion of the Forsters’ collection ; a few collected by James Macrae, who travelled for the Horticultural Society of London, 1824-1826; a small number labelled ‘‘ Beechey,”’ who was Captain of H.M.S. ‘ Blossom,’ 1825-1828; a considerable collection made by Alex- ander Mathews, a Chiswick gardener, in 1841; afew by Sir Everard Home in 1851; a collection made by the algologist W. H. Harvey during his sojourn in the southern hemisphere, 1854 to 1856; a collection made by Dr. Greffe, a Swiss gentleman, about 1862; Professor H. N. Moseley’s collection made on the voyage of H.MLS. ‘ Challenger,’ 1873-1876 ; a collection made by Mr. T. B. Cartwright during a cruise in 1888 ; and, finally, Mr. J. J. Lister’s collection, chiefly from the island of Eua, made in 1889 and 1890. In addition to the foregoing, the French collectors or commanders of expeditions, Lesson, Hombron and D’Urville, are sometimes cited; a German named Jensen, who collected some Cyperacee determined by Mr. C. B. Clarke ; and the United States Exploring Expedition ; a set of plants of the last being in the Kew Herbarium. It has not been considered necessary to cite the names of all collectors in each case. The names of a few persons who have contributed one or two plants only are not included in the foregoing account. In answer to my inquiries respecting cultivated plants, Mr. Lister informs me that the cocoa-nut, tobacco, sugar, maize, taro (Colocasia antiquorum), pine-apple, orange, lime, breadfruit, yam, sweet potato, kava (Prper methysticum), paper mulberry (Brous- 168 MR. W. B. HEMSLEY ON THE FLORA sonetia papyrifera), and a number of varieties of banana are cul- tivated; but the only one of these plants that struck him as being also wild was the cocoa-nut. Cotton was formerly culti- vated and has become colonized, but he is unable to say from actual observation whether the cultivation is still continued. With regard to the bamboo, its presence in the island has not been impressed on his memory ; yet there is one, Schizostachyum glaucifolium, Munro, which might be expected to occur, as it inhabits the Fiji, Navigator, Society, and Marquesas Islands. ANONACEX. Cananga odorata, Hook. f. et Thoms.—Greffe; Moseley. Ava, Tenasserim, Java, and Philippines. Cultivated widely in the tropics. Anona squamosa, Linn. — U.S. Expl. Exped. Introduced from America. MENISPERMACES. Stephania hernandiefolia, Walp. (including S. Forsteri, A. Gray).—U.S8. Expl. Exped.; Harvey. Samoa, Tahiti, and Marquesas Islands, Australia northward through tropical Asia to China and Japan; also in tropical Africa. CRUCIFERS. Cardamine sarmentosa, Forst.— Barclay ; Cartwright. New Caledonia, Fiji, Union, Samoa, Society, and Marquesas Islands; also recorded from Peru, about Lima, where it was doubtless introduced. Senebiera didyma, Pers.—Cartwright. A widely-spread weed, the native country of which is un- certain. [Lepidium piscidium, Forst., a characteristic sea-side plant from New Caledonia to the Society and Sandwich Islands, in- cluding many of the small coral islands, is not represented in the Kew Herbarium from the Tonga Islands, nor have I found any record of its occurrence. ] VIOLARIER. Melicytus ramiflorus, Forst.— Lister. Norfolk Island, the Kermadec group, and New Zealand. OF THE TONGA OR FRIENDLY ISLANDS. 169 BIXINEs. Xylosma orbiculatum, Forst.— U.S. Expl. Exped. ; Lister. Fiji eastward to the Marquesas Islands. PITTOSPORES. Pittosporum arborescens, Rich.—U.S. Expl. Exped.; Greffe; Lister. Fiji Islands, Pittosporum Brackenridgei, A. Gray.—Cartwright ; Lister. Fiji Islands. Pittosporum rhytidocarpum, A. Gray.— Lister. Fiji Islands. PoRTULACER. Portulaca oleracea, Linn.— U.S. Expl. Exped. Common, especially in maritime districts, in tropical and sub- tropical regions generally. Portulaca quadrifida, Linn.— Barclay; U.S. Expl. Exped. Warm regions of the Old World. GUTTIFER#. Garcinia sessilis, Seem.—Velson ; Lister. New Caledonia and Fiji Islands. Calophyllum spectabile, Willd.—U.S. Expl. Exped. Fiji and Society Islands, Java, Malayan peninsula, and Ceylon. MAtLvacez. Sida microphylla, Cav.— Forster ; Moseley. New Caledonia, Fiji and Samoa Islands, Pescadores Islands in the Marshall Group, India, and Bourbon. Sida rhombifolia, Linn.—Cartwright ; Lister. Fiji to Marquesas Islands ; and tropical and subtropical regions generally. Sida spinosa, Linn.—U.S. Expl. Exped. Sandwich Islands (“a late arrival from tropical America ’’). Tropics and subtropical regions generally. Urena lobata, Linn.— U.S. Expl. Exped.; Harvey ; Cartwright. Fiji to Marquesas Islands. Tropics generally. 170 MR. W. B. HEMSLEY ON THE FLORA Urena morifolia, DC.—Recorded from the Tonga Islands on the authority of DeCandolle (Prodr. i. p. 442). Fiji Islands. This has been doubtfully referred by Dr. Masters (in Hook. f. F). Brit. Ind. i. p. 830) to U. sinuata. Hibiscus Abelmoschus, Linn.— Barclay ; Lister. Cultivated and wild in most tropical countries, but uncertain where it is really indigenous. Hibiscus Rosa-sinensis, Linn.— Home. Fiji eastward to the Marquesas Islands (cultivated and sub- spontaneous). China, India, Java, and Timor. Hibiscus tiliaceus, Linn.—Harvey; Cartwright; Lister, and others. Fiji to Marquesas and Sandwich Islands; and generally dispersed in the tropics and extending to some subtropical countries. Thespesia populnea, Corr.— Lister. Fiji to the Marquesas and Sandwich Islands, and tropics of the Old World, chiefly on sea-shores, including Western Africa. STERCULIACER. Heritiera littoralis, DC.—Welson ; U.S. Expl. Exped. New Caledonia and Fiji. Littoral in tropical Asia, Australia, and E, Africa. Melochia odorata, Forst.—Forster ; Barclay; Harvey. New Caledonia, New Hebrides, Fiji, and Samoa ; also in Dam- mar and Timor Laut. I. vitiensis, Gray, is probably the same as JL. odorata, Forst. TILIACER. Grewia Mallococca, Linn. fi— Forster; Moseley ; Lister, and others. Fiji, Samoa, and Society Islands. Grewia prunifolia, A. Gray.— Moseley. Fiji Islands. Triumfetta procumbens, Forst.— Mathews; Harvey; Home ; Greffe; Cartwright. Throughout western Polynesia, from New Caledonia to Fiji and Samoa. Also in the Seychelles, Diego Garcia, and the Keeling Islands, through the Malay Archipelago, and in Fitzroy Island off the coast of Queensland. OF THE TONGA OR FRIENDLY ISLANDS. 171 Elzocarpus cassinioides, A. Gray.— U.S. Expl. Exped. Fiji Islands. GERANIACER. Oxalis corniculata, Linn.— Moseley; Cartwright. Dispersed in nearly all warm and temperate regions. RutacEz. Evodia hortensis, Forst.—orster ; Nelson; Home; Greffe; Lister. New Guinea, eastward to Samoa. Micromelum pubescens, Blume, var. glabrescens, Oliver (M. mi- nutum, Seem.).—Forster; Moseley; Cartwright; Lister, and others. New Caledonia and Fiji Islands, and tropical Australia, through the Malay Archipelago, to India and China. Citrus Decumana, Linn.— U.S. Expl. Exped. Introduced and cultivated in the Fiji and Tonga Islands. BURSERACEE. Canarium Harveyi, Seem.— Harvey ; Lister. Restricted to the Tonga Islands. SIMARUBER. Suriana maritima, Linn.— Harvey ; Mathews. Ellice and Phenix groups eastward to the Society Islands ; and widely spread on tropical sea-shores of both hemispheres, though hitherto not found in West Africa. MELIACE2. Vavea amicorum, Benth.—Barclay; Moseley; Cartwright; Lister, and others. Confined to the Tonga Islands. Dysoxylum Richii, C. DC. (D. alliaceum, Seem. non Blume).— Greffe; Forster; Barclay. Fiji and Navigator Islands. Melia Azedarach, Linn.— Lister. Introduced. China, India, and commonly cultivated and natu- ralized in other warm countries. Carapa moluccensis, Lam. (C. obovata, Blume).— Lister. Fiji, N. Australia, through Malaya to India; also in tropical Africa and Madagascar. 172 MR. W. B. HEMSLEY ON THE FLORA OLACINE®. Villaresia samoense, Benth. et Hook. f— Lister. Samoa or Navigator Islands. CELASTRINES. Gymnosporia vitiensis, Seem.— Nelson ; Lister, and others. Fiji eastward to the Society Islands. RHAMNACER. Colubrina asiatica, Brongn.— Lister. Throughout Polynesia from New Caledonia to the Marquesas and Sandwich Islands, and widely spread in the tropics of the Old World. Alphitonia excelsa, Reiss.— Home; Lister. New Caledonia eastward to the Marquesas Islands and Timor Laut, tropical Australia. Cardiospermum Halicacabum, Linn.— Harvey. Throughout Polynesia and throughout the tropical regions of the world. SaPINDACER. Allophyllus Cobbe, Blume (Schmidelia Cobbe, DC.).— Home. Samoa and Society Islands, Northern Australia, and tropical Asia. Cupania glauca, Camb. (Guioa glauca, Radlk.).— Lister. New Caledonia. Ratonia falcata, Seem. (Cupania falcata, A. Gray).— Moseley. Fiji Islands. Ratonia stipitata, Benth. (Cupania lentiscifolia, Pers.).— Forster; U.S. Expl. Exped. Samoa, Society Islands, and East Australia. Pometia pinnata, Forst.— Forster. Fiji, Samoa Islands, and New Guinea. Dodonza viscosa, Forst.— Harvey. New Caledonia and Fiji eastward to the Society and Sandwich Islands. Tropical and subtropical regions of the world almost everywhere ; also extending into some temperate countries, both north and south. OF THE TONGA OR FRIENDLY ISLANDS. 173 ANACARDIACES. Rhus taitensis, Guill. (R. simarubefolia, A. Gray, var. f. tai- tensis, Engl.).— U.S. Expl. Exped. ; Lister. Fiji, Samoa, and Society Islands; also from the Philippines. Spondias dulcis, Forst.— Forster ; U.S. Expl. Exped. ; Lister. Fiji, Samoa, and Society Islands. Cultivated in India and the Mascarene Islands. As limited by Engler the species extends to tropical America and the West Indies. LEGUMINOSS. Indigofera Anil, Linn.— Moseley. Fiji to the Marquesas and Sandwich Islands. Commonly cul- tivated, and native country uncertain. Indigofera tinctoria, Linn.—Cartwright ; Lister. Almost universally cultivated in tropical countries, and appa- rently indigenous both in Asia and Africa. Tephrosia purpurea, Pers. (T. piscatoria, Pers.).— Forster, and others. From Fiji to the Marquesas and Sandwich Islands; and now found in tropical regions generally. Desmodium polycarpum, DC.— Lister. From the Fijis to the Marquesas ; North Australia to Japan, China, and the Himalayas; also in the Mascarene Islands and eastern tropical Africa. Desmodium umbellatum, DC.—U.S. Expl. Exped.; Lister. New Caledonia eastward to Samoa and the Marquesas Islands ; also in North-eastern Australia through the Malay Archipelago to Burma and Ceylon ; and in the Mascarene Islands and eastern tropical Africa. Naturalized in the West Indies. Uraria lagopodioides, DC.— Home; Lister. New Hebrides and Fiji to the Navigator Islands; also in N. Australia and through Malaya to India and China. Abrus precatorius, Linn.— U.S. Expl. Exped. Fiji to the Marquesas Islands. Universally spread in the tropics, and extending into some subtropical regions. It exists 174 MR. W. B, HEMSLEY ON THE FLORA as an escape from gardens in the Sandwich Islands (Hildebrand, Fl. Haw. Isds. p. 98). Clitoria Ternatea, H. B. K.—Lister. Widely spread in tropical countries ; but it is commonly cul- tivated, and uncertain where it is really indigenous. Erythrina indica, Lam.— U.S. Expl. Exped.; Lister. Fiji to the Marquesas Islands ; and Eastern Australia, through Malaya to India. Often planted. Erythrina ovalifolia, Roxb.— Lister. Fiji and Samoa Islands, Java, and India. Mucuna gigantea, DC.— Harvey ; Lister. Fiji eastward to the Marquesas and Sandwich Islands ; also Eastern Australia and through Malaya to India. There is some doubt respecting the specific identity of the specimens from the various localities. Pueraria Thunbergiana, Benth.— Lister. Solomon Islands and Buru, and very common in China and Japan. Cultivated in China for its fibre. Canavalia ensiformis, DC., var. turgida (C. turgida, R. Grah.). —U.S. Expl. Exped. From Fiji to the Marquesas Islands, and common nearly all over tropical Asia, Africa, and America, especially in maritime districts, but apparently wanting in Australia. Canavalia obtusifolia, DC.— Barclay ; Lister. Fiji eastward to the Society Islands, and common on sandy coasts in nearly all tropical countries. Phaseolus adenanthus, G. I’. Mey. (P. truxillensis, H. B. K.). — Moseley. Fiji eastward to the Society and Sandwich Islands, and cosmo- politan in the tropics. Phaseolus Mungo, Linn.— Barclay. Fiji and Samoa Islands, and throughout the tropics of the Old World. Extensively cultivated as a vegetable. Vigna retusa, Walp. (V. lutea, A. Gray) —Cartwright ; Lister. Fiji eastward to the Society and Sandwich Islands. Widely OF THE, TONGA OR FRIENDLY ISLANDS. 175 dispersed on tropical sea-shores in the Old World; and also found in the West Indies. Dolichos Lablab, Linn. (Lablab vulgaris, Savi). — Harvey; Moseley; Cartwright. New Caledonia and Fiji to the Society and Sandwich Islands ; and tropics of the Old World, except Australia. Derris uliginosa, Benth.—U.S. Evpl. Exped. A common sea-side climbing shrub, ranging from Polynesia, including New Caledonia, New Hebrides, and Fiji Islands, to Northern Australia, South-eastern Africa, Madagascar, India and China. Inocarpus edulis, Linn. f—Greffe; Moseley; Lister. From the Fijis to the Marquesas, and in Malaya. Mr. Lister also collected this in Christmas Island, off the west coast of Java, from whose specimens the figure in Hooker's ‘Icones Plantarum,’ t. 1837, was prepared. Sophora tomentosa, Linn.—U.S. Expl. Exped.; Greffe. Fiji to the Marquesas Islands, and generally diffused on tropical sea-coasts, and extending into some subtropical regions. Cesalpinia Bonducella, Fleming.— Mathews. From New Caledonia eastward to the Marquesas and Sandwich Islands ; also widely spread and often very common, especially near the sea, in tropical Asia Africa, and America. Cesalpinia pulcherrima, Sw.— Lister. Cultivated and more or less wild in nearly all tropical countries except Australia; but itis uncertain where it is really indigenous, though probably in some part of the Old World. Cassia occidentalis, Linn.—Cartwright; Lister. Fiji to the Marquesas and Sandwich Islands; colonized. Tropica] America, and also now common in tropical Asia and Africa, though probably of American origin. Cassia Sophera, Linn.— Forster; Home. New Caledonia and Fiji Islands. “Tropical Asia, Africa, and Australia, more rare in tropical America; in all continents most frequently only from cultivation, or escaped from cultivation, LINN, JOURN.— BOTANY, VOL. XXX. N 176 MR. W. B. HEMSLEY ON THE FLORA with some evidence of being really indigenous in Australia and more doubtfully so in Asia and America.” (Bentham, in Trans. Linn. Soe. xxvii. p. 532.) Cynometra grandiflora, A. Gray.— Harvey; Lister. _ Fiji Islands only, where, however, it must be common as it 1s in all collections. Entada scandens, Benth.— Harvey; Lister. Fiji to Samoa, and cosmopolitan in the tropics. Mimosa pudica, Linn.— Barclay; Cartwright ; Lister. Tropical America, and now a common weed in most parts of the Polynesian region. Naturalized also in many parts of tro- pical Africa and Asia. Leucena Forsteri, Benth.—Cook ; Mathews. Polynesia from New Caledonia to Tahiti. Acacia laurifolia, Willd.—Cook; Home; Lister. New Hebrides and New Caledonia to Samoa and Fiji. Serianthes myriadenia, Planch.— Mathews. Fiji and Navigator to the Society Islands. RHIZOPHORES. Rhizophora Mangle, Linn.—Lister. Coasts of tropical America generally, also in west tropical Africa. Perhaps accidentally introduced with ballast in the Tonga Islands, as well as in Stewart Island—see Ann. k.-k. raturh. Hofmus. Wien, iii. p. 251. Rhizophora mucronata, Lam.— Harvey ; Cartwright. New Caledonia, Fiji and Navigator Islands, Northern Aus- tralia, and throughout tropical Asia and Eastern Africa. Bruguiera Rheedii, Blume.— Forster ; Cartwright. New Caledonia, Fiji and Navigator Islands. Common in the tropical mangrove swamps of many parts of the Old World. CoMBRETACER. Terminalia Catappa, Linn.—U.S. Expl. Exped. Fiji to the Marquesas Islands. Tropical Asia, and planted in nearly all tropical countries. OF THE TONGA OR FRIENDLY ISLANDS. 177 Terminalia littoralis, Seem.— Harvey; Lister (var. tomen- tella). New Caledonia, Fiji, Navigator, Marquesas, and Sandwich Islands, and throughout Malaya. Often planted in other tropical countries. Lumnitzera coccinea, Wight et Arn.—U.S. Expl. Exped.; Cartwright ; Lister. New Caledonia, Fiji, and Navigator Islands, Northern Aus- tralia, Malaya, and India. Myrracex. Metrosideros polymorpha, Gaudich.— Barclay ; Beechey. New Caledonia and Fiji eastward to the Marquesas, Sand- wich, end Pitcairn Islands ; also in the Kermadec group. As the specific name indicates, this is an exceedingly variable species, and it is spread all over Polynesia; yet it does not reach New Zealand, the home of the rest of the genus except the out- lying South-African I. angustifolia, Sm. Nelitris Forsteri, Seem.—/athews; Lister. Fiji and Navigator Islands. This should perhaps be included in the following species. Nelitris fruticosa, A. Gray.—U.S. Evpl. Exped.; Greffe. Fiji Islands. Eugenia amicorum, A. Gray.—Oook; Forster; Home. Fiji, Navigator, and Society Islands. Eugenia corynocarpa, A. Gray.— Home. Fiji and Navigator Islands. Eugenia Jambolana, Lam.—Cartwright. Probably cultivated. Widely spread in India and Malaya, both wild and cultivated, and extending to Australia. Eugenia malaccensis, Linn.— U.S. Expl. Exped. This fruit-tree, so generally cultivated in tropical countries, is regarded as indigenous in Polynesia by Seemann, Hillebrand, Drake del Castillo, and other writers on Polynesian botany. Eugenia rariflora, Benth. (forma parviflora).— Home; Lister. Fiji eastward to Gambier Island, in the Low Archipelago, and the Sandwich Islands. n2 178 _ MR. W. B. HEMSLEY ON THE FLORA Eugenia Richii, A. Gray.— Harvey. Fiji Islands. Barringtonia Butonica, Forst. (B. speciosa, Linn. f.).— Forster ; Mathews. New Caledonia eastward to Ducie Island in the Low Archi- pelago; shores of India, Malaya, Australia, Mascarene Islands, and it has also been recorded from Diego Garcia, Chagos Islands. MELASTOMACES. Melastoma denticulatum, Labill.— Lister. New Caledonia and Fiji eastward to the Society Islands. This has often been confused with the Asiatic If. malabathricum, Linn. Memecylon Harveyi, Seem.— Harvey ; Lister. Restricted to the Tonga Islands. LYTHRARIER. Pemphis acidula, Forst.— Harvey; Lister. New Caledonia and Union group eastward to the Society and Sandwich Islands; and an exceedingly common sea-side shrub in many parts of the tropics of the Old World, including the eastern coast of Africa. SAMYDACER. Homalium vitiense, Benth.— Lister. Fiji Islands. PASSIFLORES. Disemma czrulescens, Seem. Fl. Vit. p. 96, in nota.— Cook. This differs very slightly from New-Zealand and Fijian species, and a revision of the group would probably result in the reduc- tion of some of the forms described as species; therefore I forbear giving ita name under Passiflora, of which Disemma is now generally regarded as a section. Carica Papaya, Linn.—U.S. Expl. Exped. Cultivated and naturalized in Polynesia from Fiji to the Society and Sandwich Islands. | A native of tropical America, from Florida to Brazil. OF THE TONGA OR FRIENDLY ISLANDS. 179 CuUCURBITACER. Luffa cylindrica, Roem., 3. insularum, Cogn. (L. insularum, A. Gray).— Barclay; Moseley. Commonly cultivated throughout the tropics. The form L. insularum extends from New Caledonia and Fiji to the Society Islands, and it also occurs in Malaya and Australia. Momordica Charantia, Linn.— Moseley. Fiji and Society Islands. Commonly cultivated and natura- lized in tropical countries and probably indigenous only in tropical Africa. Melothria Grayana, Cogn.— Lister. Fiji, Samoa, and Society Islands. Cucumis Melo, Linn. (Cucumis pubescens, Willd.) U.S. Evpl. Exped. New Caledonia and Fiji eastward to the Marquesas and Sand- wich Islands. Tropical Asia and Africa, and naturalized in other countries, including Polynesia. FIcoIDEs. Tetragonia expansa, Murr.—Lister. Japan, Australia, Tasmania, New Zealand, and South America. Sesuvium Portulacastrum, Linn. (Trianthema polyandra, Blume).—Forster ; Lister. Society and Sandwich Islands. Common, especially on sandy sea-shores, throughout tropical and subtropical regions. UMBELLIFER. Hydrocotyle asiatica, Linn.—Barelay ; Lister. Fiji, Samoa, and Sandwich Islands, and tropical and subtropical regions in both hemispheres. Apium leptophyllum, F. Muell.—Cartwright. East Australia, tropical South and Central America, W. Indies, and southern part of North America. A common weed of culti- vation in warm countries. 180 MR. W. B. HEMSLEY ON THE FLORA ARALIACES. Panax fruticosum, Linn. (Nothopanax fruticosum, J7q.).— Home. . Fiji Islands. India and Malaya, cultivated ; native country uncertain. Panax multijugum, Benth. et Hook. f. (Nothopanax multi- jugum, Seem.).—Lister. Fiji Islands. Meryta macrophylla, Seem.— U.S. Expl. Exped.; Lister. Navigator Islands. RUBIACES. Nauclea Forsteri, Seem.—Lzster. Society Islands, Badusa corymbifera, A. Gray.—Forster ; Lister, and others. Fiji Islands. , Bikkia grandiflora, Reinw. (B. tetrandra, A. Gray, B. australis, DC.).—Mathews; Harvey ; Lister. New Hebrides and Loyalty Islands eastward to the Society and J.ow Islands; also from the Solomon Islands, New Guinea, and Java. It varies exceedingly in the size and shape of the flowers and to a less extent in the foliage. Oldenlandia fetida, Forst.— Forster; Greffe. Savage Island, Navigator group, and Palmerston in the Hervey group. Oldenlandia paniculata, Linn.—Forster; Nelson; Moseley ; Lister. New Hebrides and Fiji to the Navigator Islands; and gener- ally diffused in tropical Asia. Mussenda frondosa, Linn.— Forster ; Harvey; Greffe; Lister. From the Solomon Islands eastward to the Society Islands, and through Malaya to India and China. Randia coffeoides, Benth. et Hook. f. (Stylocoryne coffoides, A. Gray).---Barclay ; Harvey ; Moseley, and others. New Hebrides, Fiji, and Socitty Islands. OF THE TONGA OR FRIENDLY ISLANDS. 181 Tarenna sambucina, T. Durand (Stylocoryne sambucina, A. Gray).—Forster ; Barclay ; Cartwright ; Lister. Fiji, eastward to the Society Islands. Gardenia taitensis, DC.— Forster; Greffe ; Cartwright ; Lister. Fiji to the Society and Marquesas Islands. Guettarda speciosa, Linn.—Cook ; Harvey; Lister, and others. Tropical shores in the Old World, extending to the remotest islands of the Pacific. Plectronia barbata, Benth. et Hook. f. (Canthium barbatum, Seem.).—Forster ; Lister. Fiji, Society and Marquesas, Pitcairn and Elizabeth Islands. Plectronia odorata, Benth. et Hook. f. (Canthium odoratum, Seem.). New Hebrides eastward to the Low Archipelago and Sandwich Islands, including many of the smaller islands. Morinda citrifolia, Linn.— Forster ; Lister, and others. Fiji and Union group eastward to the Marquesas and Sandwich Islands, and widely diffused in tropical Asia, north-east Australia, and tropical Africa, chiefly in maritime districts. Common on many of the small low islands of the Pacific. Morinda Forsteri, Seem.— Home ; Lister. Throughout Polynesia from New Caledonia to Pitcairn Island, though not recorded from the Sandwich Islands. Psychotria amicorum, Benth. et Hook. f., ex Drake, Ill. Fl. Ins. Mar. Pacif. p. 197. (Chasalia amicorum, A. Gray, in Proc. Am. Acad. iv. p. 43; Psychotria asiatica, Forst. f. Fl. Ins. Austr. Prodr. p.16, non Linn. ; Petesia carnea, Forst. f. Fl. Ins. Austr. Prodr, p. 10, et ic. ined. in Mus. Brit., sed non Gaertn. Fruct. iii. p- 66, t. 192, et ergo non Eumachia carnea, DC. Prodr. iv. p. 479.) —Forster; Harvey; Lister. Common in the Fiji Islands, and also found in Samoa. This has been confused with Psychotria Forsteriana, A. Gray, Proc. Am. Acad. iv. p. 44 (Petesta carnea, Gaertn., Eumachia carnea, DC.), a native of the Fijis and Samoa, but hitherto not found in the Tonga group. The two plants are exceedingly alike in foliage, though very different in flower and fruit. 182 MR. W. B. HEMSLEY ON THE FLORA Psychotria insularum, A. Gray.—U.S. Expl. Exped. Fiji, Samoa, and Society Islands. Psychotria macrocalyx, A. Gray.—U.S. Expl. Exped. Fiji Islands. Geophila reniformis, D. Don.— Lister. Fiji to the Society Islands, and through Malaya to India and S. China; also in tropical Africa and America. CoMPosIT&. Vernonia insularum, Benth. et Hook. f. (Monosis insularum, A. Gray).— U.S. Expl. Exped. Fiji Islands. Erigeron albidus, A. Gray.— Harvey; Lister, and others. Fiji to the Marquesas and Sandwich Islands, though originally from tropical America. This is usually included under £. bonariensis, Linn., which is widely dispersed in tropical America. Siegesbeckia orientalis, Linn.— Moseley; Lister. Fiji to the Marquesas and Sandwich Islands, and cosmopolitan in warm countries and extending into some temperate regions. Adenostemma viscosum, Forst.— Moseley. Fiji to the Marquesas and Sandwich Islands, and almost every- where in warm countries, but probably native in the Old World. Dichrocephala latifolia, DC.— Cook. Fiji, Society, and Marquesas Islands ; also tropical and sub- tropical Asia and Africa. Wedelia biflora, DC.— Moseley. New Hebrides, Fiji, and Navigator Islands. A common sea-side plant in the tropical regions of the Old World. Wedelia strigulosa, Benth. et Hook. f.(Wollastonia strigulosa, DC.).— Home ; Moseley ; Lister, and others. From New Caledonia to the Navigator Islands and in the Moluccas. OF THE TONGA OR FRIENDLY ISLANDS. 183 Bidens pilosa, Linn.— Forster ; Moseley ; Lister. Fiji eastward to the Marquesas and Sandwich Islands, and generally spread in warm countries. Sonchus asper, Vill.— Forster; Barclay. Fiji, Society, and Sandwich Islands. All temperate and many tropical countries, wild or introduced. Sonchus oleraceus, Linn.— Moseley. The Marquesas and Sandwich Islands. Distribution of S. asper. GooDENIACER. Scevola floribunda, A. Gray.— U.S. Expl. Exped. Fiji Islands. Scevola Kenigii, Vahl, including S. sericea, Forst.—Cook ; Forster; Moseley ; Lister. Fiji and Union group eastward to the Low Archipelago and Sandwich Islands ; also tropical Asia and Australia. CAMPANULACES. Wahlenbergia gracilis, A. DC.— Lister. New Caledonia, New Zealand and the Kermadecs, Australia, India, Eastern Asia, and S. Africa. MyrsIneEz. Mesa nemoralis, DC.—Cook ; Lister. 7 The Solomon Islands, New Caledonia, New Hebrides, Fiji and Navigator Islands. Mesa vitiensis, Seem.— Lister. Fiji Islands. Ardisia Listeri, Stapf, n. sp.— Lester. See description, p. 214. SaPOTACER. Bassia amicorum, A. Gray.— Forster; Nelson; Lister. Only recorded from the Tonga Islands. Mimusops Kauki, Linn. (M. dissecta, 2. Br.).—Forster. In Polynesia only known from the Tonga Islands ; it occurs also in tropical Australia, Malaya, and Burma. 184 MR. W. B. HEMSLEY ON THE FLORA Isonandra Richii, A. Gray.— U.S. Expl. Exped., Seemann (FI. Vit. p. 150) states that this may prove con- specific with Mimusops Kauki, Linn. He also mentions that there was an undetermined species of this order in the British Museum, collected by Cook. EBENACEE. Maba buxifolia, Pers.—Zister. This is the first record from Polynesia. Occurs also in North Australia, Malaya, India, tropical Africa, and Madagascar. Maba elliptica, Forst.— Forster; Harvey; Lister, and others. New Caledonia, Fiji, and Samoa. OLEACES. Jasminum didymum, Forst. f—Greffe; Lister. New Caledonia and Fiji eastward to the Society Islands, and from Northern Australia to Java. Jasminum simplicifolium, Forst. f. (J. australe, Pers.).— Forster; Moseley ; Lister, and others. New Hebrides and Fiji; also the east coast of Australia and Norfolk Island. APOCYNACER. Melodinus scandens, Forst.—Forster ; Nelson; Lister. New Caledonia and Fiji. Alyxia bracteolosa, Rich.— Nelson ; U.S. Expl. Exped. Fiji and Samoa Islands. Alyxia scandens, Roem. et Schult.—Lister. Fiji, Society, and Low Islands. Alyxia stellata, Roem. et Schult.— Forster ; Nelson ; Moseley ; Lister. New Caledonia, Fiji and Society Islands, Malaya and tropical Asia. Cerbera Odollam, Gaertn.— Lesson ; Moseley. Society, Marquesas, and Low Islands. A sea-side tree of Polynesia, tropical Australia, and Asia. Ochrosia parviflora, Hensl.— Nelson ; Harvey. . New Caledonia and Fiji to the Society Islands; it also occurs in the Keeling Islands, where it was first collected by Darwin. OF THE TONGA OR FRIENDLY ISLANDS. 185 Vinca rosea, Linn.— Lister. Samoa and Sandwich Islands. Supposed to be a native of tropical America; but it is now so firmly established in many parts of the Old World as to have all the appearance of being indigenous. Tabernzmontana orientalis, R. Br.—Nelson ; Lister. New Hebrides, Fiji, Loyalty, and Society Islands ; also eastern Australia and the Philippines. ASCLEPIADACES. Asclepias curassavica, Linn.—foseley ; Lister. Society, Marquesas, and Sandwich Islands; a native of the West Indies, now spread all over the tropics. Hoya australis, R. Br. (H. bicarinata, A. Gray).—U.S. Expl. Exped.; Lister. New Hebrides, Fiji, and Navigator Islands, and Eastern Australia. LoGAaNIACER. Geniostoma crassifolium, Benth.—Zzster. Previously only known at Kew from the Philippines. Geniostoma rupestre, Forst.—Welson ; Barclay; Greffe. New Caledonia eastward to the Society Islands. Fagrea Berteriana, A. Gray.—oseley ; Lister. New Caledonia eastward to the Marquesas Islands. BoRAGINER. Cordia aspera, Forst.— Forster ; Nelson. Fiji and Navigator Islands. Cordia subcordata, Lam.— Barclay; Harvey; Macrae. From the Fiji and the Union group to the Marquesas and Sandwich Islands, coast of tropical Australia, Malaya, Islands of the Indian Ocean, westward to the east coast of Africa. In India it is commonly planted and perhaps nowhere really wild. Tournefortia argentea, Linn. f.—Forster; Lister. Fiji and Union group, eastward to the Society and Low Tslands, and on nearly all coasts, insular and continental, of the Indian and Pacific Oceans, at least in the warmer parts. 186 MR. W. B. HEMSLEY ON THE FLORA CoNVOLVULACES. Ipomeea biloba, Forsk. (I. Pes-capre, Roth).— Moseley ; Lister, and others. Fiji to the Marquesas and Sandwich Islands, and almost every- where on sandy coasts in tropical and subtropical regions. Ipomea Bona-nox, Linn. (Calonyction speciosum, Choisy).— Nelson ; Moseley. Fiji to the Society and Sandwich Islands. Now widely dis- persed in tropical countries, often planted ; but supposed to be a native of tropical America. Ipomeea coccinea, Linn.— Moseley ; Lister. Also introduced into the Sandwich Islands. Cultivated and quasi-wild throughout India. Introduced from tropical America. Ipomea congesta, R. Br. (Pharbitis insularis, Choisy; I. in- sularis, Stewd.).— Nelson ; Mathews; Moseley. New Hebrides; Fiji and Sandwich Islands. Also on the east coast of Australia, Norfolk and the Marianne Islands. Ipomea denticulata, Choisy.— Barclay; Mathews; Moseley. New Hebrides, Fiji and Society Islands, Australia, Malaya, India, and the Seychelles. Ipomea martinicensis, G. F. W. Mey. (I. uniflora, Roem. et Schult.) —Lister. Fiji; and widely dispersed in tropical Asia, Africa, America, and Australia. Ipomeea peltata, Choisy.— Lister. Fiji and Society Islands, East Australia, Malaya, and Mas- carene Islands. Ipomea Turpethum, R. Br.— Forster; Nelson; Mathews. Fiji, Samoa, and Society Islands, and North Australia through Malaya and the islands of the Indian Ocean to the Himalayas and Formosa. SoLaANACER. Solanum amicorum, Benth.— Forster ; Nelson, and others. Confined to the Tonga Islands. Solanum Uporo, Dun. (S. anthropophagorum, Seem. ; S. viride, Soland. non R. Br.).—Barelay ; Lister. New Caledonia, Fiji, Samoa, and Society Islands. OF THE TONGA OR FRIENDLY ISLANDS. 187 Solanum Forsteri, Seem.— Barclay. Easter Island and Tabiti. Solanum oleraceum, Dunal.—Cartwright. Fiji, Union, Samoa, and Sandwich Islands. This is now generally regarded as a variety of the cosmopolitan S. nigrum, Linn. Solanum viride, R. Br.— Nelson. Society and Marquesas Islands. Physalis peruviana, Linn.— Moseley. Society, Marquesas, and Sandwich Islands. Indigenous in tropical America; extensively cultivated in the Old World, and widely colonized. Capsicum frutescens, Linn.—foseley ; Lister. Probably indigenous only in America; now cultivated and naturalized in most other warm countries. GESNERACER. Cyrtandra Listeri, Hemsl., n. sp.— Lister. For description see forward, p. 213. ACANTHACES. Hemigraphis reptans, T. And.— Moseley. New Caledonia, New Hebrides, and Admiralty Islands. Eranthemum insularum, A. Gray.— Harvey. Fiji Islands. . Graptophyllum Siphonostena, F. von Muell.—Lister. Fiji Islands. For description of this species see forward, p. 214. VERBENACES. Premna taitensis, Schau.— Harvey; Lister. Admiralty, Ellice, and Fiji, eastward to the Marquesas Islands. Vitex trifolia, Linn.— Forster ; Moseley. Fiji, Marquesas, and Sandwich Islands, and a common sea-side shrub in the tropics of the Old World. Clerodendron amicorum, Seem. (Faradaya amicorum, Seem.).— Banks § Solander; Barclay; Harvey ; Lister. Navigator Islands only. 188 MR. W. B. HEMSLEY ON THE FLORA Clerodendron inerme, R. Br.—athews;\ Harvey; Moseley ; Lister. New Caledonia, Fiji and Navigator Islands, and a common sea-side shrub in tropical Asia and Australia. LaBIATSs. Ocimum Basilicum, Linn.— Home; Lister. Fiji, Society, Marquesas, and Sandwich Islands, introduced. Widely diffused in tropical Asia and Africa, and commonly cultivated. Ocimum gratissimum, Linn.—Woseley. (Introduced.) Java, India, tropical Africa and America ; commonly cultivated. Leucas flaccida, R. Br. (including ZL. decemdentata, Sm.).— Lister. . Fiji, Society, and Marquesas Islands, and Eastern Australia, Malaya, India. Colonized in the West Indies. Teucrium inflatum, Sw.—forster; Barclay; Greffe, and others. New Caledonia, New Hebrides, and Fiji, and widely dispersed in tropical and subtropical America and the West Indies. Also in the Galapagos. NYCTAGINES. Pisonia inermis, Forst.— Harvey ; Lister. Throughout Polynesia, from Fiji to the Low Archipelago and Sandwich Islands. Also North-east Australia. The Polynesian and Australian species of Pisonia have been so variously limited by different authors, that it is difficult to understand them, except on the assumption that there is one very variable species. Indeed, Drake del Castillo (Flore de la Polynésie Francaise, p. 157) unites them all, except P. aculeata, Linn., under the name of P. wmbellifera, Seem. (Ceodes umbel- lifera, Forst.), and doubtless the only alternative is to make several, or even many, critical species. AMARANTACER, Amarantus caudatus, Linn. (Euxolus caudatus, Mog.).—Cook. Fiji, Society, and Sandwich Islands ; and widely dispersed in the warmer parts of Asia; also occurring in tropical Africa. OF THE TONGA OR FRIENDLY ISLANDS. 189 Amarantus paniculatus, Linn.—Barelay. Fiji, cultivated or an escape in East and West Asia and Africa. A common weed in the warm countries of the Old World. Achyranthes aspera, Linn.— Barclay; Moseley ; Lister. Fiji and Society Islands, and generally diffused in Eastern Australia, tropical and subtropical Asia, Africa, and America. Cyathula prostrata, Blume.— Barclay. New Hebrides and Solomon Islands to Tahiti and the Mar- quesas; and widely spread in tropical Asia and East Africa. Gomphrena globosa, Linn.— Barclay. Probably a native of America, now naturalized in nearly all hot countries. CHENOPODIACES. Salsola Kali, Linn.— Moseley. Very widely diffused in temperate and subtropical regions, both in America and in the Old World, where the soil is im- pregnated with salt. PHYTOLACCACES. Rivina humilis, Linn. (R. levis, Linn.).—Lister. Exceedingly common in tropical and subtropical America, and naturalized in many parts-of the Old World, including Polynesia. PoLYGONACE. Polygonum glabrum, Willd.—Lister. Fiji, Society, and Sandwich Islands; also tropical Asia, Africa, and America. PIPERACE. Piper latifolium, Forst. £— Forster; Moseley. New Hebrides, Fiji and Society Islands ; also from Timor. Piper Macgillivrayi, C. DC.— Nelson; Barclay; Moseley ; Lister, and others. Fiji and Society Islands, common. Piper methysticum, Forst. f— Home; Lister. Cultivated and spontaneous in Fiji, Society, Marquesas, and Sandwich Islands. / 190 MR. W. B. HEMSLEY ON THE FLORA Peperomia leptostachya, Hook. et Arn.—Lister. Society, Marquesas, and Sandwich Islands ; also North-east Australia, MyristTIcacEes. Myristica hypargyrwa, A. Gray.— U.S. Expl. Exped. ; Lister. Navigator Islands. Myristica inutilis, Rich—U.S. Expl. Exped. Navigator Islands. MonIMIACES. Hedycarya ? alternifolia, Hemsl., n. sp.— Lister. Confined to the Tonga Islands. Better specimens are required. For description see forward, p. 215. Hedycarya dorstenioides, A. Gray.—-Lister. Fiji and Navigator Islands; apparently very common. LAvURINES. Cassytha filiformis, Linn.—Zister. Fiji, Union, and Pheenix, eastward to the Society, Low, and Sandwich Islands, and very widely diffused, chiefly in maritime districts, in tropical Asia, Africa, America, and Australia. Hernandia Merenhoutiana, Guill —Moseley ; Lister. Society Islands. Hernandia peltata, Meissn.— Home. Fiji, Union, Society, and Marquesas Islands; also in North Australia, tropical Asia, East Africa, and Madagascar. THYMEL#HACER. Wikstremia rotundifolia, Decne.—Nelson; Harvey; Lister. Confined to the Tonga Islands. Leucosmia acuminata, A. Gray (Drimyspermum acuminatum, Seem.).—Forster ; Nelson; Lister. Fiji and Navigator Islands. Leucosmia Burnettiana, Benth. (Drimyspermum Burnet- tianum, Seem.).— Harvey ; Lister. Fiji and Navigator Islands. OF THE TONGA OR FRIENDLY ISLANDS. 191 LoRaNTHACER. Loranthus insularum, A. Gray.—U-S. Expl. Exped. ; Moseley ; Lister. Fiji, Samoa, and Society Islands. EvPuorBiace&. Euphorbia Atoto, Forst.— Cook; Forster; Lister. New Caledonia and Fiji eastward to the Marquesas and Sandwich Islands. A sea-side shrubby species, also inhabiting the coasts of North Australia, Malaya, and India. Euphorbia obliqua, Endl.— Forster. Norfolk Island. Euphorbia pilulifera, Linn.— Moseley ; Cartwright. Fiji, Society, Marquesas, and Sandwich Islands. Now gen- erally diffused in warm regions of both hemispheres. Euphorbia Sparmannii, Boiss.— Lister. Pitcairn and Elizabeth Islands and New South Wales. This is similar to some of the other species and probably occurs in intermediate localities. Phyllanthus Gaudichaudii, Muell. Arg.—U. S. Expl. Exped. Samoa and Marianne Islands. Phyllanthus ramiflorus, Muell. Arg.—Cook ; Lister. New Hebrides and Fiji eastward to the Society and Marquesas Islands. Phyllanthus simplex, Retz.—Cook; Barclay; Lister. New Caledonia and Fiji, Samoa and Society Islands, and Australia, Malaya, India, China, and Mongolia. Codizum variegatum, Blume (C. moluccanum, Decne.).— Lister New Hebrides and Solomon Islands to Fiji and Navigator Islands, and in the Moluccas ; cultivated in many countries. This is an exceedingly variable plant in its foliage. Seemann (Fl. Vit. p. 231) states that it forms underwood in the larger islands of the Fiji group, where it was, even in his time, commonly cultivated by the natives for ornamental purposes. A great many varieties were cultivated, each having a dis- tinguishing name. Hillebrand (Fl. Haw. p. 393) mentions it as a cultivated plant only in the Sandwich Islands. Drake del Castillo does not mention it, and I find no Tahitian specimens. LINN. JOURBN.—BOTANY, VOL. XXX. O 192 MR. W. B. HEMSLEY ON THE FLORA Bischofia javanica, Blume.— Harvey. New Caledonia eastward to the Society Islands ; also Malaya and India. Bischofia leptopoda, Muell. Arg. (ex descr.).— Moseley. Samoa. Aleurites moluccana, Willd. (including A. triloba, Forst.).— Forster; Barclay; Moseley ; Cartwright; Lister. Generally spread in Polynesia; also North-eastern Australia, Malaya, and India. It is now naturalized in most other tropical countries. Acalypha grandis, Benth.— Home; Hombron. New Hebrides to the Navigator Islands and the Malay Archipelago. Acalypha insulana, Muell. Arg.— Moseley. Fiji and Navigator Islands. Macaranga Harveyana, Muell. Arg.—-Harvey ; Moseley ; Cart- wright ; Lister. Fiji, Navigator, and Society Islands. Macaranga Seemannii, Muell. Arg.— Cook. Fiji Islands. Ricinus communis, Linn.— Moseley. Cultivated and subspontaneous in the Fiji, Society, Marquesas, and Sandwich Islands, and very generally spread in tropical and subtropical regions. Itis uncertain where it is really indigenous ; but Africa is regarded as its native home. Homalanthus pedicellatus, Benth. (H. nutans, Benth. et Hook. f.; Carumbium pedicellatum, Mig.; C. nutans, Muell. Arg.).— Forster; Barclay; Lister, and others. New Caledonia, New Hebrides, Fiji, and Society Islands. Excecaria Agallocha, Linn.—Cook; Home. New Caledonia and Fiji Islands, North-eastern Australia, Malay Archipelago, and the coasts of tropical India. URTICACE®. Celtis pacifica, Planch.—Cook. Only recorded from the Marquesas and Tonga Islands. OF THE TONGA OR FRIENDLY ISLANDS. 193 Trema amboiuensis, Blume.— Grefe. Fiji and Sandwich Islands, and widely spread from North and East Australia, through Malaya, northward to South China and India. Trema discolor, Benth. et Hook. f. (Sponia discolor, Decne.).— Moseley. Society Islands. Broussonetia papyrifera, Vent.— Home. Fiji, Society, Marquesas, and Sandwich Islands, New Zealand, Malaya, India, and Formosa; but often cultivated. Ficus aspera, Forst. f.— Moseley. New Hebrides, Fiji Islands, and East Australia. Ficus prolixa, Forst. f. (F. Forsteriana, Endl.).— Forster. New Caledonia, New Hebrides, Society, and Marquesas Islands. Ficus scabra, Forst. f.— Home. New Hebrides and Fiji Islands. Ficus Storckii, Seem.— Barclay. Fiji Islands. Known only from Kadavu. Ficus tinctoria, Forst. f.— Home; Lister. Fiji, Samoa, and Society Islands. Antiaris Bennettii, Seem.— Home. Fiji, and Tucopia Island, near the New Hebrides. Fleurya interrupta, Gaudich.— Cook ; Moseley. From the Solomon Islands and New Hebrides to the Marquesas and Sandwich Islands; also North-eastern Australia through Malaya to China and India, and in Eastern Africa. Laportea Harveyi, Seem.— Lister. Fiji Islands. Elatostemma sesquifolium, Hassk. (Procris integrifolia, D. Don, non Hook. et Arn.).—Lister. India and Malaya. Pipturus argenteus, Wedd. (P. propinquus, Wedd.).— Forster ; Lister. New Hebrides eastward to the Society and Marquesas Islands, and in North-easteru Australia and the Malay Archipelago. 02 194 MR. W. B. HEMSLEY ON THE FLORA Pipturus velutinus, Wedd. (Sponia velutina, Planch.; P. inca- nus, Wedd.).—Cook ; Greffe. From the New Hebrides and the Ellice group to the Marquesas ; also in the Malay Archipelago and the Nicobar Islands. (There is a specimen of Behmeria platyphylla, D. Don, in the Kew Herbarium labelled “Tongatabou and Tahiti,” but it was probably from the latter island. | CASUARINES. Casuarina equisetifolia, Forst.— Home ; Cartwright. Fiji, Society, and Marquesas Islands, and generally spread in the coast-regions of tropical Asia, Africa, and Australia. It is naturalized in tropical America. HyYDROCHARIDEX. Halophila ovata, Gaudich. (H. ovalis, Hook. f.).—Home ; Moseley. In the sea: Fiji and Navigator Islands, and widely spread in the Indian and Pacific Oceans. ORCHIDER. Dendrobium macranthum, A. Rich.— Lister. Solomon Islands, New Guinea, and Java. Spathoglottis pacifica, Reichb. f—Greffe; Lister. Fiji, Samoa, and the Society Islands. Corymbis veratrifolia, Reichb. f—Moseley. Fiji and Samoa Islands, North Australia, through Malaya to the Philippines and India. ScITAMINES. Canna indica, Linn.— Barclay ; Lister. Fiji, Samoa, and Sandwich Islands, Malaya, India, and Tropical America. Hillebrand (Fl. Haw. p. 485) states that this plant is now spread all over the islands, but he suspected that it was intro- duced soon after the discovery of the islands, yet it has a native name. AMARYLLIDER. Crinum pedunculatum, R. Br. (C. taitense, Red.).— Lister. New Caledonia, Fiji, Eastern Australia, and Malaya. OF THE TONGA OR FRIENDLY ISLANDS. 195 TACCACER. Tacca pinnatifida, Forst.— Moseley. Fiji to the Marquesas and Sandwich Islands, and widely diffused in maritime districts of tropical Asia and Africa, as well as on the coast of North-eastern Australia. DI0scoREACEs. Dioscorea alata, Linn.—Coohk. Fiji, Society, and Marquesas Islands, also Malaya and tropical India, though perhaps not indigenous in the last-named country. Dioscorea sativa, Linn.—Cook. Fiji eastward to the Marquesas and Sandwich Islands, North- eastern Australia, through Malaya to India and Africa. LIviace#. Cordyline terminalis, Kunth.—oseley ; Lister. Fiji, Society, Marquesas, and Sandwich Islands, and generally spread over tropical Asia and North-eastern Australia. Dianella ensifolia, Red.— Lister. New Caledonia, Fiji and Sandwich Islands, Australia, tropical Asia, and the Mascarene Islands. Dianella intermedia, Eudl.— Nelson. Fiji and Society Islands, also New Zealand and Norfolk Islands. The synonymy of these two species is variously quoted by different authors. COMMELINACES. Commelina nudiflora, Linn. (C. pacifica, Vahl).—Forster : Nelson; Barclay. Fiji and Sandwich Islands, and generally diffused in warm countries. PaLMe. Pritchardia pacifica, Seem. et H. Wendl.— U.S. Expl. Exped. ; Lister. Fiji, Samoa, and Marquesas Islands. Cocos nucifera, Linn.— Lister. Throughout tropical Polynesia and almost all littoral regions within the tropics. The cocoanut has long been cultivated in the most distant parts of the world, and it is uncertain how much 196 MR. W. B. HEMSLEY ON THE FLORA of its present geographical range is due to the direct agency of man and where it spread from. PANDANER. Pandanus odoratissimus, Linn. f. (P. verus, Rumph.).—Cook. Fiji, Society, and Sandwich Islands, North Australia, and tropical Asia. AROIDER. Raphidophora pertusa, Schott, var. vitensis, Engl. (Scindapsus Forsteri, Endl.; Epipremnum mirabile, Schott).—Lister. New Hebrides, Fiji, and Society Islands. The typical form is found in Eastern Australia, Malaya, and India. Narapacenz*. Halodule australis, Miq.— Moseley. Marianne Islands, New Caledonia, and the Fijis; also Malaya, India, Mascarene Islands, North Africa, Red Sea, and the Mediterranean. Cyprracer. [By C. B. Clarke, F.R.S.] Mariscus albescens, Gaudich. (Cyperus stuppeus, Forst. £. ; C. pennatus, Lam.).—D’ Urville ; Moseley. Throughout Polynesia abundant; sparsely scattered through the tropics of the Old World, from Guinea to the Sandwich Islands. Mariscus albescens, Gaudich., var. 3. ventricosa, C. B. Clarke (Cyperus ventricosus, &. Br.; C. caricifolius, Hook. et Arn.; C. tongatabuensis, Boeck.).— Home; Greffe. Australia and Sandwich Islands. Mariscus cyperinus, Vahl, var. 6. venusta, C. B. Clarke (M. umbellatus, Horne; Cyperus venustus, Forst. f. non 2. Br.).— Jensen. Throughout Polynesia; common. Mariscus levigatus, Hook. et Arn. non Roem. et Schult (M. flavus, Seem. non Vahl; Cyperus Seemannianus, Boeck. ; C. upo- luensis, Boeck.).— Home. Fiji and Navigator Islands; plentiful. * Until this was in type I had overlooked a record of Cymado cea isoetifolia Aschers. (Engl. Jahrb. vii. p. 446), from Tonga. It inhabits the Indian and Pacific Oceans, OF THE TONGA OR FRIENDLY ISLANDS. 197 Kyllinga monocephala, Rottb.— Moseley ; Jensen Abundant in Polynesia, and a weed throughout the warmer parts of the Old World. Fimbristylis complanata, Link (F. communis, Seem.).— Greffe. Frequent in Polynesia and throughout the warmer parts of the world. Fimbristylis cymosa, R. Br., var. (3. subcapitata, C. B. Clarke (F. pyenocephala, Hillebr.).—Graffe. New Caledonia to Sandwich Islands. Fimbristylis diphylla, Vahl (F. affinis, Pres; F. marquesana, Steud.).— Home; Moseley. Throughout Polynesia and the warmer parts of the world. Scleria depauperata, Boeck.— Harvey. From New Caledonia to Samoa; frequent. Scleria lithosperma, Sw.— Barclay, Common in Polynesia aud throughout the warmer parts of the world. GRAMINER. Paspalum scrobiculatum, Linn.—Cook; Cartwright. New Hebrides, Fiji, Society, and Sandwich Islands, and frequent in tropical and subtropical Australia, Asia, and Africa. Panicum ambiguum, Trin.—U.S. Expl. Exped.; Harvey; Moseley. New Caledonia, Fiji and Navigator Islands; it is also recorded from the Philippine Islands. Panicum pilipes, Nees.— Barclay; Moseley. Fiji Islands. Panicum sanguinale, Linn.—Barclay; Home; Cartwright ; and others. New Hebrides to the Society Islands and Pitcairn Island, and almost everywhere in tropical and subtropical countries. Panicum trigonum, Retz.—U.S. Expl. Exped. New Hebrides aud the Fiji Islands, Malaya, and India. Oplismenus Burmanni, Beauv.— Lesson ; Cartwright. Fiji Islands and tropical regions generally. 198 MR. W. B. HEMSLEY ON THE FLORA Oplismenus compositus, Roem. et Schult.— Moseley. Fiji to the Marquesas, Sandwich, and Gambier Islands (Low Archipelago) ; and generally spread in warm countries. Cenchrus calyculatus, Cav. (C. anomoplexis, Labill.).— Home ; Moseley. New Caledonia eastward to the Society, Sandwich, and Pitcairn Tslands. Coix Lacryma-Jobi, Linn.— Forster ; Moseley. Fiji and Society Islands, and now common in the tropics of both hemispheres. Probably introduced into the New World. Erianthus maximus, Brongn. (Eulalia japonica, Seem. non Trin., teste Hackel).—U.S. Expl. Exped. Fiji and Society Islands. Stenotaphrum americanum, Schrank.—Barclay. Widely spread in Polynesia, though not cited by either Seemann or Drake del Castillo; and widely spread on tropical sea-shores in both hemispheres. Thuarea sarmentosa, Pers. (T. involuta, R. Br.).— Harvey. New Hebrides and Fiji to the Society Islands and Low Archi- pelago; and in the Mascarene Islands, Malay Archipelago, and North Australia. Andropogon aciculatus, Retz — Home ; Cartwright. From New Caledonia eastward to the Marquesas, Low, and Sandwich Isiands, and in Australia, Malaya, China, India, and Mauritius. Eleusine indica, Gaertn.— Barclay: Home; Moseley, and others. Fiji to the Marquesas and Sandwich Islands, and almost every- where in tropical and subtropical regions, and occurring in some temperate countries ; often existing, however, as a colonist only. Centotheca lappacea, Desv.— Barclay. Fiji to the Society and Marquesas Islands, on the eastern coast of North Australia, and generally spread over tropical Asia and Africa. OF THE TONGA OR FRIENDLY ISLANDS. 199 GYMNOSPERME#. CoNIFER®. Podocarpus elata, R. Br.— Harvey ; Lister. East Australia, from Rockingham Bay, southward to the Macquarrie river. CYCADACER. Cycas circinalis, Linn.— Lister. Solomon Islands, New Caledonia, New Britain, Fiji, North Australia, Malaya, India, eastern tropical Africa, and the Masca- rene Islands. With a numerous series of specimens, both wild and cultivated, before us, Mr. W. T. Thiselton Dyer and I find it impossible to sort them satisfactorily into the proposed species: C. circinalis, L., C. Rumphii, Migq., and C. celebica, Miq., to say nothing of the half-a-dozen or more other forms which have been described as such; but it would, perhaps, be premature to unite them without studying them where they grow in abundance. Cycas is rare in Polynesia and unknown in the Society, Marquesas, and Sandwich groups, and most abundant in the Malay Archipelago. FILIcEs. Gleichenia dichotoma, Hook.— Home; Lister. Fiji to the Marquesas and Sandwich Islands, and in tropical and subtropical regions generally. Alsophila lunulata, R. Br.—Lister. New Hebrides, Fiji, and Navigator Islands. Davallia elegans, Sw.— Lister. Society Islands, tropical Australia, Malaya, China, India, Madagascar, and western Tropical Africa. Davallia solida, Sw.— Home; Lister. Fiji and Society Islands, Java, and the Malayan Peninsula. Adiantum Capillus-Veneris, Linn.— Lister. Almost cosmopolitan, except in the colder regions. Adiantum diaphanum, Blume.— Lister. New Caledonia, Fiji and Navigator Islands, New Zealand, Aus- tralia, Malaya, and South-east China. 200 MR. W. B. HEMSLEY ON THE FLORA Adiantum hispidulum, Sw.—Velson. Fiji and Society Islands, and widely spread in tropical regions of the Old World. Hypolepis tenuifolia, Bernh.— Moseley ; Lister. New Hebrides and Society Islands, New Zealand, Australia, and Java. Pteris comans, Forst. f—Nelson ; Lister. New Zealand, Tasmania, New Hebrides, Fiji, and Society Islands ; also Juan Fernandez. Pteris ensiformis, Burm. f. (P. crenata, Sw.) —Cook; Nelson. Fiji and Navigator Islands, and tropical Australia to India and China. Pteris longifolia, Linn.— Nelson; Barclay; Home; Lister. New Hebrides and Fiji, and tropical and warm temperate regions all round the world. Pteris marginata, Bory (P. tripartita, Sw.; Litobrochia tri- partita, Presl).— Lister. Fiji to Society Islands; and in Australia, Malaya, India, west tropical Africa, Mascarene Islands, and Kaffraria. Pteris quadriaurita, Retz.— Nelson ; Home. New Hebrides, Fiji, and Sandwich Islands, and all round the world within the tropics and a little beyond them. Blechnum orientale, Linn.— Home. Fiji and Society Islands, Australia, India, and South China. Asplenium falcatum, Lam.— Nelson ; Home; Barclay; Moseley. New Caledonia, Fiji, and Society Islands, and ranging widely in the tropical regions of the Old World. Asplenium obtusatum, Forst. f—Lister. Society, Marquesas, and Sandwich Islands, New Zealand and Australia, and extending to Peru and Chili. Aspidium aristatum, Sw.— Lister. Fiji eastward to the Society and Sandwich Islands, and in Australia, Japan, Ceylon to the Himalayas, and in Natal. Nephrodium dissectum, Desv.— Home ; Lister. New Caledonia eastward to the Society Islands, and in South- west Australia, tropical Asia, and Madagascar. OF THE TONGA OR FRIENDLY ISLANDS. 201 Nephrodium Henkeanum, Pres].— Nelson. Fiji, Samoa, and Society Islands, Malaya, and Ceylon. Nephrodium Harveyi, Baker.—Lister. Fiji, Samoa, and Society Islands. Nephrodium invisum, Carruth.— Home; Moseley. Common in the Polynesian islands, though not recorded from the Sandwich group. Nephrodium latifolium, Baker.— Lister. Fiji to the Society Islands. Nephrodium molle, Desv. (N. nymphale, Hook. et Arn.).— Nelson; Barclay. New Hebrides, Fiji, and Society Islands, and widely dispersed in tropical and subtropical regions. Nephrolepis acuta, Presl, var. (N. rufescens, Pres!).— Moseley; Lister. New Hebrides, Fiji, and Society Islands, and widely dispersed in tropical and subtropical regions. Nephrolepis exaltata, Schott.—elson ; Home; Lister. New Hebrides to the Marquesas and Sandwich Islands, and widely dispersed in tropical and subtropical regions. Polypodium adnascens, Sw. (Niphobolus adnascens, Kaulf.).— Nelson; Home; Barclay. Solomon and Fiji Islands, and common in the tropical regions of the Old World. Polypodium nigrescens, Blume.— Veitch. Fiji, Samoa, and Society Islands, East Australia, Malaya, and India. Polypodium Phymatodes, Linn.— Home ; Moseley; Lister. Fiji, Society, and Marquesas Islands, and tropical and subtropical regions of the Old World. Acrostichum aureum, Linn.— Moseley. Fiji, Samoa, and Society Islands, and widely dispersed in tropical and subtropical regions of both hemispheres. Schizea dichotoma, Sw.—Cook; Lister. Fiji and Society Islands, and widely dispersed in tropical and subtropical regions. 202 MR. W. B. HEMSLEY ON THE FLORA Ophioglossum pendulum, Linn.— Moseley. Fiji, Society, and Sandwich Islands, and over the tropical regions of the Old World. LycopoDIAcEs. Lycopodium cernuum, Linn.— Lister. Fiji, Society, and Sandwich Islands, and cosmopolitan in the tropics of both hemispheres and extending beyond them to Japan, the Azores, Cape Colony, New Zealand, and St. Paul’s Island in the south Indian Ocean. Psilotum triquetrum, Sw.— Lister. Fiji, Society, and Sandwich Islands, and tropical regions of both hemispheres, extending to Florida, Japan, and New Zealand. 5. Tasutar View of the Distribution of the apparently Indi- genous Plants of the Tonga or Friendly Islands. This Table is restricted to the plants concerning which there is some probability of their being indigenous—that is to say, those which have probably reached the islands without the direct agency of man; but of course this is a point from which uncer- tainty cannot be eliminated. Thus limited, about 50 species included in the Enumeration are not repeated here. The list itself is of plants collected in one or more of the Tonga Islands *. In the first column to the right of the name the star indicates whether the species is found in Polynesia westward of the Tonga Islands; and in the second column an eastward ex- tension. It should be understood that Polynesia is here employed in a broad sense, and includes Micronesia and Melanesia of geographers. The third column, Australasia, includes Australia, New Zealand, and the contiguous islands belonging to these countries. Malaya,in the fourth column, includes New Guinea, the Moluccas, Malay archipelago and peninsula, Siam, and Cochin China. The fifth column indicates a wider range in the Old World, and the sixth an extension to America or the New World. * The Wallis Islands, claimed by the French, are here considered as forming part of the Tonga group; but very few plants have been collected in this group. —_—_—_- OF THE TONGA OR FRIENDLY ISLANDS. 203 Of wider range. Polynesia. Tonga Islands, Austral- Malaya. West, | East *, | °° Old World. Stephania hernandizfolia ...... we * * * * Cardamine sarmentosa....... seeee] O% * vee a see Melicytus ramiflorus ............] 0... vee * Xylosma orbiculatum ............ * * Pittosporum arborescens......... * ” Brackenridgei...... * ” rhytidocarpum * Portulaca oleracea ......... rn * * * * ” quadrifida ...... rn * * * * Garcinia sessilis..............00.0066- * Calophyllum spectabile seeeeeees * * . * * Sida microphylla .......... see eeees * * wee * * » Yhombifolia’.................. * * * * * » Spinosa....... ce eeeens te eeeeees see wee * * * Urena lobata .............ceceeeeeees * * * * * » morifolia ......... se eeeeees * Hibiscus tiliaceus ................6. * * * * * Thespesia populnea ...... see eeeees * * * * * Heritiera littoralis ............ wee * * * * Melochia odorata ...............06- * * * Grewia Mallococea ..............| * * »» prunifolia.................. * Triumfetta procumbens ......... * * * * Eleocarpus cassinioides ......... * Evodia hortensis ...............6.. * * wee * Micromelum pubescens ......... * * * * Canarium Harveyi.—Endemic . Suriana maritima.................. * * * * * Vavea amicorum.—Endemic . Dysoxylum Richii_ ............... * Carapa moluccensis ............ ef see * * * Villaresia samoense .....-......... * Gymnosporia vitiensis............ * * Colubrina asiatica...... se eeeee enone * * * * * Alphitonia excelsa ............44- * * * * Cardiospermum Halicacabum .| * * * * * Allophyllus Cobbe .............+ eee * * * * Cupania glauca............. ees eeees * Ratonia falcata.......cccscceeeceees * » Stipitata ....... eee oe * * Pometia pinnata ............6++66 * * we * Dodonsea Viscosa ......-..seeeeees- * * * * % Rhus taitensis ............ seneeeees * * * Spondias dulcis...........00+++++++ * * Tephrosia purpurea .........+..++- * * * » * Desmodium polycarpum......... * * * * * ” umbellatum......... * * * * * Uraria lagopodioides .....-...... * * * * * 48 40 3l 24 28 22 New World. * 10 * West or east of the Tonga Islands, 204 MR. W. B. HEMSLEY ON THE FLORA Tonga Islands. Polynesia. | Of wider range. Malaya. Ola World. New East. World. 48 Abrus precatorius.............0.0.. Erythrina indica .................. ” ovalifolia ............... Mucuna gigantea .........00....... Pueraria Thunbergiana ......... Canavalia ensiformis ............ ” obtusifolia ............ Vigna retusa ..........00cceececese, Leuceena Forsteri............ seeees Acacia laurifolia Pe eeercccees ” mucronata ......... Bruguiera Rheedii ne Pee eecveeese Ceeeee Pete e reese ceceenns ne Peeters ereweees » Corynocarpa ...... beens » malaccensis » Tariflora, ” Richii Memecylon H arveyi—End. _. oo oe Meryta macroph lla Nauclea F orstori ee he ee 100 oo aN i) Be OK Kk OK Ok Ok ok Ok ok OK ok Ok HOR RK KOK KOK RK HK OK KK! * x * 10 * * eK Ke * we * * HOKE * KOK XH ek kK KR KD * * *: * * * * ROR RR RK kk ND OK OK Ok * * x KOK * * * x * x eK Kk KS * Ok Ok OK Ok * * OK Ok ok Ok * * * * * * x: * * OF THE TONGA OR FRIENDLY ISLANDS. 205 Polynesia. Of wider range. Tonga Islands. w E Austral ytalaya. Old | New eat. | Mast. World. World. 100 84 71 42 55 48 21 Gardenia taitensis ............66 * * Guettarda speciosa ..........0..4- * * * * * Plectronia barbata ............... * * ” odorata ...... seeeenees * * Morinda citrifolia....... de eeeeenees * * * * * ” Forsteri............ eeceee] * Psychotria amicorum ............ * * ” insularum ............ * * ” macrocalyx ......... * Geophila reniformis............... * * “ * * * Vernonia insularum ....... seseeees * Adenostemma viscosum ....... . * * * * * * Dichrocephala latifolia ......... * * vee * * Wedelia biflora....... sececeeeeeeens * * * * * » Sstrigulosa ......... sessed ¥ * * Sceevola floribunda ............... * »» Konigii ........e eee * * * * | # Wahlenbergia gracilis............ * vee * * | * Meesa nemor ilis ............eeeeee * * yo VItIENSIS .o.... eee eee ee ee ee * | Ardisia Listeri.—Endemic ...... Bassia amicorum.—Endemic ... | Mimusops Kauki.................. see * * # Maba buxifolia .................0085 a eee * * * » elliptica 2.0... eee eee * * | | Jasminum didymum... ........... * * * x | ” simplicifolium ...... * . * Melodinus scandens............... * | Alyxia bracteolosa ............... * * | yy SCANENS............. ee reeeeee * * | yy Stellata ....... cece cece * * wee * x | Cerbera Odollam .................. vee * * * * | Ochrosia parviflora ........2-0602. * * eee * Tabernemontana orientalis ...| * * * * | Hoya australis ................60... * * * | Geniostoma crassifolium ......... wee wee tee * ” Tupestre .......0.... * * Fagrea Burteriana ............... * * Cordia aspera ...........c0eeeeee * * » Subcordata ..........60... * * * * * Tournefortia argentea ............ * * * * x Ipomeea biloba ...........:06ee0ee | o* * * * * | * ” congesta ...........eeeeee- * * * * | » Genticulata............+.. * * * * * » fpeltata ....... cesecsernes * a * * x | ” Turpethum ............ * * * * * | » -‘Martinicensis............ * . * * * | # Solanum amicorum.—Endemic. | ” Uporo ...seceeceee seen es * ¥ | ” Forsteri ............00008+ * | ” OleraceuM .........00+06+ * * ” ViTIde 22... ccc eee eee e eee wes * 152 127 109 63 80 67 | D5 206 MR. W. B. HEMSLEY ON THE FLORA Polynesia. Of wider range. Tonga Islands, Austral Malaya! oq | New West. | East, World. | World. 152 127 109 63 80 67 25 Cyrtandra Listeri.—Endemic... Hemigraphis reptans ... ........ * . vee * Eranthemum insularum ......... * Graptophyllum Siphonostena...| * Premna taitensis .............006.. * * vee Vitex trifolia ........... cc ecceeee eee * * * sg * Clerodendron amicorum ......... vee * ”» imerme ............ * * * * * Leucas flaccida ............. saeeeene * * * * * Teucrium inflatum ............... * eee we * Pisonia inermis............seeeee0-. * * * Cyathula prostrata ............... * * eee * * Achyranthes aspera ............4+: * * * * * * Salsola Kali ............... eee cee] wee see * * * * Polygonum glabrum ............ * * . eee * * Piper latifolium ...............06- * * . * » Macgillivrayi............... * * » methysticum ............... * * Peperomia leptostachya ......... * * Myristica hy pargyreea ............ * ” inutilis ...............66. * Hedycarya alternifolia.—End. ” dorstenioides ......... * * Cassytha filiformis ............... * * * * * * Hernandia Meerenhoutiana wee * ” peltata .............. * * * * * Wikstreemia rotundifolia.— End. Leucosmia acuminata ............ * * ” Burnettiana ......... * * Loranthus insularum ............ * * Euphorbia Atoto ....... seveceecees * * * * * ” obliqua ............... wee see * ” pilulifera ............ * * # * * * ” Sparmannii ......... * * Phyllanthus Gaudichaudii ...... vee * % ” ramiflorus ......... * * ” simplex ............ * * * * * Bischofia javanica be seeeeeeceeseeees * * # * ” eptopoda ............... eee * \ Codizeum variegatum ............ * * eee * Aleurites moluccana............... * * * * * Acalypha grandis.................. * * * ” msulana ............... * * Macaranga Harveyana............ * * ” Seemanni ............ * Homalanthus pedicellatus ..... * * Exceecaria Agallocha ............ * see * * * Celtis pacifica oo... ee, eee * Trema amboinensis ............... * * * * * 9 Giscolor .................000- we * Broussonetia papyrifera ......... * * * * * Ficus aspera ...........ecesceceeees- * wee * 204 164 149 82 102 84 31 OF THE TONGA OR FRIENDLY ISLANDS. 207 Polynesia. Of wider range. Austral- ————_———— Tonga Islands. West, | East, | 28% Malaya. Old | New ° ° World. | World. 204 164 149 82 102 84 31 Ficus prolixa...............eeceees -| % * yy SCADA oo. cee eeeeee cece ee * » Storckii ........... seeeaeeees * yy CAMNCHOTIA ...........ecees eee * * Antiaris Bennettii.................. * Fleurya interrupta ............... * * * * * Laportea Harveyi...............+6 * Elatostemma sesquifolium ...... ve wee ve * * Pipturus argenteus ............... * * * * ” velutinus ............... * * wes * * Casuarina equisetifolia............ * * * * * Halophila ovata ..............6608 * * * * * Dendrobium macranthum ...... * we * Spathoglottis pacifica ............ * * Corymbis veratrifolia ............ * * * * * Canna indica .........ceeceeseeeeeeee * * see * * * Crinum pedunculatum............ * wee * * Tacca pinnatifida .................. * * * * * Cordyline terminalis ............ * * * * * Dianella ensifolia .............0.... * * * * * ” intermedia ............... * x * Commelina nudiflora ............ * * * * * * Pritchardia pacifica ............... * * Cocos nucifera ...........s.cceee eee * * * * * * Pandanus odoratissimus ......... * * * * * Raphidophora pertusa, * * var, vitiensis =isf Halodule australis ............... * tee bee * * Mariscus albescens ............... * * * * * ” Cyperinus ...........00. | * * ” levigatus ...........0.4- * * Kyllinga monocephala............ * * * * * Fimbristylis complanata ......... * # * * * * ” CYMOSA .eseeseoeee * * 9 diphylla ............ * * * * * * Scleria depauperata ............... * * » lithosperma ..... be eceeeeee * * * * * * Paspalum scrobiculatum .........) * * * * # Panicum ambiguum............... * * * » Pilipes ........cseeeeeeee * » sanguinale............... * * * * * * ” trigonuM ............+6. * tes * * Oplismenus Burmanni ......... * wee * * * * ” compositus ......... * * * * * * Cenchrus calyculatus ............ * * Stenotaphrum americanum...... * * * * * * Thuarea sarmentosa .........-.-- * * * * * Coix Lacryma Jobi ............... * * * * * Erianthus maximus ............... * * Andropogon aciculatus ......... * * * * * Eleusine indica..............+s0006 * * * * * * Centotheca lappacea............ * * * * * se bS a4 | 139 | uo | 136 | 114 | 42 LINN. JOURN.—BOTANY, VOL. XxX. bP 208 MR. W. B. HEMSLEY ON THE FLORA Tonga Islands. 255 Podocarpus elata .........600eere Cycas circinalis ........-...660- Gleichenia dichotoma ........... Alsophila lunulata ............... Davallia elegans ............0.0++ PE -(0) 06 (; Adiantum Capillus-veneris ...... ” diaphanum ............ » hispidulum ............ Hypolepis tenuifolia ............ Pteris comans ............00e eee ee » ensiformis .........-....606 » longifolia ..........:.60- » Marginata ...........0. » quadriaurita .......6....... Blechnum orientale ............... Asplenium faleatum............ ” obtusatum ............ Aspidium aristatum ............ Nephrodium dissectum ......... » Henkeanum ...... ” Harveyi ..... ” invisuM ............ ., latifolium ......... ” molle ............08 Nephrolepis acuta.................. ” exaltata ............ Polypodium adnascens ......... » nigrescens ......... » Phymatodes ...... Acrostichum aureum ............ Schizea dichotoma ............... Ophioglossum pendulum _...... Lycopodium cernuum ,,.......... Psilotum triquetrum ............ Tora.s ...... 290 Polynesia. Of wider range. Austral- Malaya. asia. Old New West. | Hast. World. | World. 214 189 110 136 114 42 ws wee * % wee * * * * * * * * # * * _ x x | x * * * | * * * % x | * * * * * * | * * * * * | * * * * * * * x La wee * * * * x * * wee % * * * * * % * * * * * * * * * * * vee * * * * * * eee * * wee oes * * * * * * * * * a * * * * * * * % * % * * * % * * * * * * * * * * * * * * * * wee tee * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * % * 246 220 1388 162 141 5d 6. Summary of the preceding Table. This table has been drawn up since the paper was read, and supplements, and in some particulars repeats, the “ Introductory Note.” The most striking features are, first, the large number of species of very wide range, and, secondly, the large number of purely Polynesian species; and the proportion of the latter might have been made to appear even larger by eliminating a number a OF THE TONGA OR FRIENDLY ISLANDS. 209 of other species of wide range which very probably owe their presence to human agency, direct or indirect. Out of a total of 290 species, 105, or more than a third, are confined to Polynesia ; and on examination it will be found that a large number of these species are peculiar to the Fiji, Navigator, and Tonga groups of islands, fewer extending eastward to the Society Islands, or westward to the New Hebrides. Considering the area and elevation of the islands and the extent of cultivation in bygone times, the number of indigenous vascular plants still existing is considerable; and, as in most insular floras, the number of natural orders and genera represented is large in proportion to the number of species. Out of the 202 natural orders recognized in Bentham and Hooker’s ‘ Genera Plantarum,’ 79 are represented by apparently indigenous species in the Tonga Islands; and the 290 species comprise no fewer than 202 genera. From calculations I have published elsewhere *, based upon the numbers of the ‘Genera Plantarum,’ the proportions for the whole world are: 37°50 genera to an order, and 12°65 species to a genus. In the Tonga Islands the proportions are: 2°55 genera to an order, and 1°43 species to a genus. The proportion of species to genera is about the same as in the remote insular flora of St. Helena, calculated from the probably indigenous species. It may be interesting to repeat here the figures for some distant dissimilar and larger areas :— Average number Average number of Genera to of Species to an Order. a Genus. India ...... cece eee cece eee 13°0 6:0 MexiC0... oo... cece eee eee ees 11:0 64 America, north of Mexico... 9°6 6:2 Australia... .........c0ccee eee eee 8:7 6°4 In further illustration of the general distribution of natural orders, it may be added that 174 are represented in British India, 163 in Mexico and Central America, and 144 in Australia. The object of giving the foregoing figures is to demonstrate the composite character of the Polynesian flora; to demon- strate that natural orders have, with very few exceptions, a wide distribution ; to demonstrate that there must have been general migrations ; to demonstrate that the vegetation of the world must be the evolution of the same elements throughout, and not of dif- ferent elements in different parts of the world. In illustration of * ‘Biologia Centrali-Americana, Botany,’ vol. i. Introduction, p. xxiii, 210 MR. W. B. HEMSLEY ON THE FLORA this I will proceed to further details of the composition and cha- racter of the Polynesian flora. I have already stated that the most striking fact brought out by the preceding Table is the large proportion of species—upwards of a third—peculiar to Polynesia. Another fact, which does not come out, either in the Enumeration or in the Table, is the complete absence of peculiar generic types in the Tonga Islands ; and even if we include the Fiji and Navi- gator groups, the number is exceedingly small. Seemann (FI.Vit., Introd. p. xvii) could only adduce 16, and some of these were doubtful. Subsequent discoveries, especially in Malaya, have reduced that number, whilst the more complete exploration of Fiji has not resulted in the discovery, I believe, of a single new genus. And if we include the flora of the Gilbert, Phonix, Ellice and Union groups, the results would be practically the same. In Eastern Polynesia there is a larger proportion of endemic genera, but nothing approaching a natural order. Even in so highly specialized a flora as the Hawaiian, only about 40 genera out of 250 are restricted to those islands, and they are perhaps more remarkable for habit than for floral structure. In the Society Islands there are very few endemic genera, or, going further, peculiarly Polynesian genera, and one of the most singular of these, Lepinia, as pointed out at p. 165, has recently been rediscovered so far westward as the Solomon Islands *. Returning to the peculiarly Polynesian species of the Tongan flora, we find that 55 have both an eastward and a westward extension ; 28 a westward extension only; 12 an eastward extension only; and 10, so far as at present known, are endemic. These are:—Canarium Harveyi, Vavea amicorum, Memecylon Harvey, Disemma c@rulescens, Ardisia Listeri, Bassia amicorwm, Solanum amicorum, Cyrtandra Listeri, Hedycarya alternifolia, and Wikstremia rotundifolia. But some of these may prove either to have a wider range or not to be specifically different when intermediate forms may have been found. On the other hand, it is not probable that unknown species remain to be discovered in these islands. * Drake del Castello, ‘ Flore de la Polynésie Francaise,’ Introduction, p. xix, tabulates the distribution of 588 species, belonging to 79 natural orders. His results are: 161 peculiar to French Polynesia (Society, Marquesas, Pomotou, Gambier, and Wallis Islands) ; 123 extending to Oceania ; 297 to Indo-Malaya ; and 7 to other regions. These figures are to some extent inaccurate, inasmuch as the full distribution is not always given, Me te ae OF THE TONGA OR FRIENDLY ISLANDS. 211 Passing to the distribution of the species which extend beyond Polynesia, we have 188 in the Australasian column, 162 in the Malayan, 141 having a distribution beyond Australasia and Malaya in the Old World, and 55 in the New World. Adding the species having an Australasian or Malayan extension to the purely Polynesian species might, without due consideration, lead us to suppose that the Australasian affinities of the Tonga flora were nearly as strong as the Malayan ; but such a conception is dissipated by the figures in the two last columns relating to species of wider distribution. Furthermore, it is manifest from the very small number of species which extend to some part of Australasia, but are not found elsewhere beyond the limits of Polynesia, that the Australian connections are exceedingly slight. The following are the only species coming under this category :— Melicytus ramiflorus, Ratonia stipitata, Metrosideros polymorpha, Jasminum simplicifolium, Hoya australis, Ipomea congesta, Pi- sonia inermis, Peperomia leptostachya, Euphorbia Sparmannii, Ficus aspera, Podocarpus elata, and Pteris comans. After examining all the facts, and especially the distribution of the genera, and the nearest affinities of the apparently endemic Polynesian species, there is no doubt that the combined Fijian, Samoan, and Tongan flora is eminently Malayan in character. I would only repeat, in conclusion, that the distribution may not be quite complete in all cases; hence the figures may be subject to slight modifications, but the main issues would not be thereby affected. 7. Descriptions of New Plants from the Tonga and Solomon Islands. Dysoxylum megalanthum, Hemsl. (Plate IX.) Species pro magnitudine florum insignis, calyce irregulariter 3-4-lobato. Arbor usque ad 30 pedalis alta (fide Cominsiz). Folia (perfecta non visa) fere glabra vel cito glabrescentia, petiolata, abrupte pinnata, rhachi terete gracili, 7 poll. longa; foliola 8, opposita, breviter petiolulata, papyracea, oblonga, 2-4 poll. longa, inferiora breviora, integra, obtusa vel obtuse acuminata, basi inwqualia, venis primariis lateralibus utrinque circiter 12 sat conspicuis. Flores albi (fide Cominsii) polygami? circiter 10 lineas longi et 12 lineas diametro, racemosi (racemis terminalibus fide Cominsiz), 212 MR. W. B, HEMSLEY ON THE FLORA pedicellis crassiusculis 6-12 lineas longis; calyx cupularis, crasse coriaceus, irregulariter 2-4-lobatus, lobis rotundatis vel minoribus interdum subacutis; petala 4, oblonga, obtusa, extus sericeo-tomentosa, quam calyx dimidio longiora ; tubus stamineus cylindricus, 8-10-dentatus, antheris 8-10 inclusis vel brevissime exsertis petala subequantibus; discus cylindricus, ovarium in- clusus; ovarium dense tomentosum, 4-loculare, loculis uniovulatis, stylo sursum glabro, stigmate capitato. Fructus ignotus. Soromon Istanps: without locality, Rev. R. B. Comins, 241. Crossostylis Cominsii, Hems/l. (Plate X.) Folia fere membranacea. Flores parvi, staminodiis nullis, petalis integris. Arbor usque ad 18 ped. alta (ex Comins), ramulis floriferis glabris. Folia tenuia, glaberrima, longiuscule petiolata, lanceo- lata vel oblanceolata, cum petiolo 6-7 poll. longa et 23-3 poll. lata, obtusa, basi cuneata, venis primariis lateralibus utrinque circiter 8 prope marginem inter se anastomosantibus subtus sat conspicuis. Flores 4-5 lineas diametro, pedicellati, in axillis foliorum fasciculati, pedicellis circiter semipollicaribus; sepala 5-6, crassa, tertia parte connata, intus carinata, hirsutula ; petala totidem, spathulata, integra, quam sepala paullo breviora, cito decidua ; stamina circiter 30, disco cupuliformi inserta ; ovarium sepe 10-loculare, loculis 2-ovulatis, vertice facile calyptratim dehiscens. Fructus ignotus. Santa Cruz, New Hebrides, Rev. R. B. Comins, 279. Eugenia salomonensis, Hemsl. Species ornata, floribus rubris speciosissimis laxissime pani- culatis, paniculis longissime pedunculatis pendulis. Arbor magna (Comins), novellis omnino glabris, ramulis gra- cilibus. Folia breviter petiolata, subcoriacea, minutissime pel- lucido-punctata, oblongo-oblanceolata, 44-7 poll. longa, subite acuminata, vix acuta, basi rotundata, venis primariis lateralibus numerosis 13 lineas ab margine inter se anastomosantibus. Panicule terminales (Comins) cum pedunculo gracillimo sesqui- pedales, ramulis secundariis oppositis divaricatis paribus distant- ibus, ramulis ultimis 2—5-floris. Flores umbellati, subsessiles, cum staminibus circiter sesquipollicares ; calycis campanulati lobi rotundati, circiter 14 lineas longi et lati; petala fere orbi- cularia, circiter 3 lineas diametro, cito decidua; stamina nu- issn ae; een OF THE TONGA OR FRIENDLY ISLANDS. 213 merosissima, filamentis fere sesquipollicaribus capillaribus ut videtur sanguineis. Fioripa Isuanp, Solomon group, Rev. &. B. Comins, 232. A handsome species characterized by the very long, loose, pendulous inflorescence. Cyrtandra Listeri, Hemsl. Species C. Tempesti, Horne, valde affinis, differt imprimis foliis minoribus late ovatis basi rotundatis crenatis. Frutex nanus (fide Lister), ramulis floriferis crassis subcar- nosis primum ferrugineo-puberulis. Folia opposita, squalia, breviter petiolata, crassiuscula, late ovata, interdum fere ro- tundato-ovata, cum petiolo 3-5 poll. longa et 2-3 poll. lata, obtusa, crenulata, utrinque, precipue subtus, puberula, subtus pallidiora, venis primariis lateralibus utrinque circiter 6 crassi- usculis curvatis. Cyme@ axillares, sesquipollicares, pauciflora, pedicellis ferrugineo-puberulis 5-6 lineas longis. Flores me- diocres; sepala 5, sublibera, decidua, breviter inzqualia, ovali- oblonga, circiter 4 lineas longa, subobtusa, corolle tubum fere equantia, extus ferrugineo-pubescentia ; corolla glabra, vix 9 lineas longa, distincte bilabiata, labio inferiore longiore, lobis latis rotundatis ut videtur sinuatis, tubo lato leviter curvato antice ventricoso ; filamenta crassa, geniculata; anthere magne, approximate vel conniventes, primum in corolle ventrum posite ; discus magnus, cupularis; ovarium ovoideum, glabrum, leviter verrucosum, stylo brevi crasso, stigmate maximo bilobato. Ardisia Listeri, Stapf Affinis Ardisie grandi, Seem., sed omnibus partibus minor staminibus in tubum brevem corolle basi tantum adnatum coalitis. Frutex 4-5 ped. altus, glaberrimus, ramulis crassis pallidis apicem versus cicatricibus rotundis crebris notatis. Folia ob- longo- vel obverse-lanceolata, 4-6 poll. longa, 1j-1? poll. lata, in petiolum 3-? poll. longum attenuata, obtusiuscula, nervis lateralibus tenuibus utrinque 12-14. Panicule plures terminales 1 poll. longe, # poll. late, pedunculo gracili 1-12 poll. longo suffulte pedicellis tenuibus. Oalyx quinquelobus, 3-3 lin, longus, lobis rotundatis minute ciliatis. Corolla 1% lin. longa, ad 3 quinquepartita, lobis ovatis obtusis. ilamenta in tubum basi cum corolla connatum in ore antheras cordato-ovatas acuti- usculas gerentem coalita. Ovariwm ovoideum in stylum equi- 214 MR. W. B. HEMSLEY ON THE FLORA longum abiens, stigmate quinqueangulari, ovulis 3 placents immersis.—O. Stapf. Eva, Tonga, in rupibus calcareis summi jugi insule, Mr. J. J. Lister, 1889. Graptophyllum Siphonostena, F. Muwell., Fragm. vi. 87. Frutex glaberrimus ramis mox albicantibus. Folia oblonga vel elliptica vel ovato-oblonga, 1-3 poll. longa, 3-14 poll. lata, basi in petiolum 3-6 lin. longum attenuata, obtusiuscula vel subacuminata, membranacea, nervis lateralibus utrinque 4-7, tenuibus. Cyme axillares et terminales sepe ad flores 2 vel 1 redacte, bracteis minutis, pedicellis gracilibus 2-3 lin. longis. Calyx 1-13 lin. longus, fere ad basin 5-partitus, segmentis tri- angularibus vel triangulari-subulatis. Corolle tubus 1 poll. longus, leviter incurvus, superne sensim ampliatus, lobi sub- equales, ovati, 5 lin. longi, apice pilosuli. Filamenta e fauce breviter exserta. Staminodia minuta. Capsula clavata.—O. Stapf. Eva, Tonga, in rupibus calcareis, Mr. J. J. Lister, 1889; Vavau and Liruxa, W. H. Harvey, 1885; Ovazav, Fiji group, Seemann, 1854. There are only three species known from Australia, and a fourth, which is cultivated all over India and the Malay Archipelago, but has never been found in a wild state. As Sir F. von Mueller gave the name to Seemann’s specimen without a proper de- scription, 1 thought it worth while to describe the plant more accurately. It is nearest allied to G. Earlii, F. Muell., which, however, is a tree with smaller leaves and a shorter and wider corolla-tube and more robust capsules. Ruellia Guppyi, Hemsl. Species ex affinitate R. aruensis, 8. Moore, sed fere glabra foliis longe petiolatis, bracteolis calycis dentibusque brevioribus, corollz tubo subite in limbum patentissimum expanso. Herba vel suffrutex ramosus, sesquipedalis (Guppy), cito gla- brescens. Folia petiolata, fere membranacea, primum minute strigillosa, anguste oblongo-lanceolata, cum petiolo usque ad 7 poll. longa et 13 poll. lata, obtusa vel subacuta, basi cuneata, sepe inequalia, obscurissime late crenata, venis primariis late- ralibus utrinque circiter 8-9 prope marginem inter se anastomo- santibus ; petiolus usque ad 2 poll. longus. Flores flavi, axillares (solitarii?) pedunculo brevi, bracteis bracteolisque parvis sub- OF THE TONGA OR FRIENDLY ISLANDS. 215 ulatis; calycis dentes subulati, semipollicares ; corolla leviter obliqua, fere hypocrateriformis, tubo circiter 3 poll. longo angusto vix curvato, limbo subequaliter 5-lobo, fere 2 poll. diametro, lobis latis. Stamina 4,inclusa. Ovarium multiovulatum. Capsula non visa. -Treasury Istanp, Solomon group, Dr. H. B. Guppy, 186; Unawa, in™the same group, Rev. R. B. Comins, 264. These two localities are almost at the two extremities of this chain of islands. Hedycarya ? alternifolia, Hems/. Species insignis, si hujus generis, foliis alternis. Frutex vel arbor undique glabra. Ramuli ultimi pallidi, ver- ruculosi. Folia alterna, coriacea, in siccis nigrescentia, breviter petiolata, rotundato-ovata vel late elliptica, cum petiolo usque ad 5 poll. longa et 3 poll. lata, obtusa, remote obscureque denti- culata, venis primariis lateralibus utrinque 4-5. Flores ¢ vix sesquilineam diametro, racemosi, racemis parvis axillaribus; perianthium depressum, obscurissime lobatum, margine cilio- latum ; stamina numerosissima antheris subsessilibus. Tonea Isnanps, Mr. J. J. Lister. Antiaris turbinifera, Hemsl. Inter species Polynesiz magnitudine fructus insignis. Arbor 80-100-pedalis (fide Comins). Folia petiolata, sub- coriacea, oblongo-obovata, sine petiolo usque ad 6 poll. longa, integra, obtusissima, basi cuneata, supra nitida, subtus pallidiora, venis primariis lateralibus utrinque 8-9 inter se prope margi- nem anastomosantibus ; petiolus sesquipollicaris. Receptaculum florum ¢ breviter pedunculatum, circiter 9 lineas diametro, ut videtur cupulatum, carnosum, extus verrucosum, multiflorum, involucro brevissime 5-lobato, lobis late rotundatis, margine incrassatis. Perianthium tenuissimum, perfectum non visum ; filamenta brevissima. Fructus ellipsoideus, siccus 23 poll. longus et 2 poll. diametro. Paura, Usawa, etc., Solomon Islands, Rev. R. B. Comins. The Rev. R. B. Comins, to whom we are indebted for this and many other curious plants from Polynesia, sends a specimen of a humming-top made by natives of the Solomon Islands from the fruit of Antiaris turbinifera. After the fleshy part of the fruit has been removed or has decayed, there remains an excessively bard endocarp of ivory-like density. The peduncle disarticulates LINN, JOURN.—BOTANY, VOL, XXX. Q 216 MR. W. B. HEMSLEY ON THE FLORA from the base of the fruit and the continuation into the fruit decays and falls away, leaving a circular aperture, about a third of an inch in diameter, and as clearly defined as though it had been cut. To form the humming-top, the natives bore holes at right angles to this and insert a peg, the exserted part being about as long as the diameter of the fruit. It is spun by a double string wound round the peg in the same direction, the free ends of the string being pulled in opposite directions to impart the impetus to the top. It is probable that this natural hole in the fruit serves as a point of egress for the radicle in germination. PANDANACE®. SARARANGA, Hemsl., gen. nov. Flores, ut videtur, dioici, ¢ ignoti. Florum @ spadix pani- culatus, ramosissimus, amplissimus, ramuli ultimi incrassati, spheroidei, receptacula florifera carnosa sinuosa involucrata for- mantes, involucro brevisinuoso. Flores 2 (perjuveniles tantum visi) omnino immersi, flexuoso-biseriati, et mamillis (stigma- tosis?) superficialibus biseriatis conjuncti; staminodia nulla; ovarium 1-loculare, l-ovulatum. Fructus ignotus. Sararanga sinuosa, Hemsl. (Plate XI.) Arbor 50-60-pedalis, radicibus aereis nullis (ex Guppy). Folra coriacea, lineari-lanceolata, basi 4 poll. lata (Guppy), apicem versus tantum aculeolata. Spatha ignota. Spadix albus, 4-5 ped. longus (Guppy), receptacula (i. e. spadicis ramuli ultimi) circiter 3 poll. diametro, sed forsan fructifera accrescentia. Favro Istanp, Solomon group, at elevations of 1600 to 1900 feet, Dr. H. B. Guppy; Josre Istanp, North-east New Guinea, Dr. O. Beccari. I am indebted to my colleague Dr. Otto Stapf for the drawings of the microscopical dissections of the inflorescence. DESCRIPTION OF THE PLATES. Puate IX. Portion of compound leaf and raceme of flowers of Dysoxylum megalanthum, Hemsl., natural size. Fig. 1. Longitudinal section of a flower, showing the hairy ovary half im- mersed in the cylindrical disk ; fig. 2, cross section of ovary; fig. 3, vertical section of the same. All enlarged. Bor.Vou..XXX Py. 9. JOURN T Linn. Soc tlemsley C.Fitch hth Teresi ANTHUM,é SOXYLUM MEGAT, DY: st Vol OURN BR NN. Soc. 1 lal Ii ue th. Fitch Le Cc. ) Stapf “itch del + ft Anai + ] u SINUOSA , Hens SARARANGA OF THE TONGA OR FRIENDLY ISLANDS. 217 Puate X. Flowering branch of Crossostylis Cominsii, Hemsl., natural size. Fig. 1. A flower; fig. 2, the same, with the sepals and petals and some of the stamens removed, showing the ovary in cross section; fig. 3, the top of the ovary and stylar column. All enlarged. Puats XI. Sararanga sinuosa, Hemsl. Fig. 1. Upper portion of a leaf, which Dr. Guppy describes as 4 inches broad in the lower part : natural size. 2. Tip of the same, enlarged. 3. Small portion of female inflorescence, which is 4 to 5 feet long: natural size. 4. One of the ultimate branchlets of the inflorescence, or receptacles in which the female flowers are embedded, with biseriate mammz and sinuous inyolucre, x 3. 5. The same in longitudinal section, x 8, showing some of the ovaries. 6. A portion of the surface of one of the lobes, x 8. 7. Across section of the same, some distance below the surface, showing the biseriate ovaries corresponding to the superficial mamme, about x 8. 8. A vertical section, x 20, showing ovary and stigmatic tube. 9. Another section, still further enlarged. 10. Ovule, about x 40. LINN. JOURN.—BOTANY, VOL. XXX. R 218 REV. G. HENSLOW ON THE ORIGIN OF PLANT-STRUCTURES The Origin of Plant-Structures by Self-Adaptation to the Envi- ronment, exemplified by Desert or Xerophilous Plants. By the Rev. Georce Henstow, M.A., F.L.S., &c. [Read 16th November, 1893.] (Prats XII.) ConTENTS. I. Introductory Observations on the Characteristic Features of Plants of Deserts and Arid Countries. IT. General Morphological Characters. IIT. Spinescent Characters. IV. Foliage of Desert Plants. VY. Succulent Plants. VI. Protection of Buds. VII. Roots. VIII. Histological Peculiarities of Desert Plants. IX. Assimilative Tissues. X. Ligneous Tissues. XI. Water-storage Tissues. XII. Secretions. XIII. General Summary of Observations on Desert Plants. XIV. Self-fertilization of the Flowers of Desert Plants. I. Introductory Observations on the Characteristic Features of Plants of Deserts and Arid Countries. Tue general facies of the flora of a country with a relatively dry soil and atmosphere is very observable ; and when it is found to be the same in widely separated countries—as in the desert regions of North Africa, in the arid districts of India and Thibet, of Afghanistan, in parts of Australia, in S. Africa, in Brazil*, &c.,— * For an interesting description of the plants growing in a very dry region of Brazil, but probably of less intensity than that of the African deserts, the reader is referred to a paper by M. Ed. Warming entitled, “ Lagoa Santa (Brésil), Etude de Géographie Botanique,” Rev. gén. de Bot. 1893, p. 145. Many peculiarities mentioned by that author correspond more or less exactly with those I have described in this paper. Grisebach also compares the 8, African region (Kalahari) with the Soudan :— “Les autres formes végétales sont les mémes qu’on retrouve dans d’autres steppes et déserts, ou bien elles signalent l’intime affinité du Kalahari et du Soudan. A la premiére catégorie appartiennent les plantes grasses (Euphorbia, Mesembryanthemum), les yégétaux bulbeux qui ouvrent rapidement leurs fleurs aprés les orages d’été (ex. Amaryllis); parmi les arbustes, les formes de Spartium (ex. Lebeckia), de /}Oléandre (par Ventremise d’une Rubiacée, le Vangueria), et du Myrte (par une Ebénacée, VEuclea); enfin, les arbustes & fouilles velues, particulidrement fréquents dans la savane (Zarchonanthus).” (Vég. du Globe, vol. ii. pp. 255, 256.) BY SELF-ADAPTATION TO THE ENVIRONMENT. 219 so that these several countries afford either the same or “ represen- tative plants,” one infers (but of course at first merely on @ priori grounds) that most probably similar causes have produced these similar results. A closer inspection shows that the similarities in the vegetative system of such plants, which often have no affinity between them whatever, can be carried down into the minutest details of histological structure; and that a large proportion of such structures at least are always serviceable to the plants in resisting the deteriorating effects of an insufficient supply of water, as well as of an excess of radiation and other hindrances to such vigorous growth as is maintained in moister climes. We thus begin to suspect, indeed very strongly, that the various peculiarities (such, e. g., as the densely hirsute clothing and the consolidation of the mechanical tissues) are the direct results of the dry climatal conditions surrounding the plants ; and that the unfavourable environment actually brings about the production of just those kinds of structure which are best able to resist the injurious effects of the climate, and so enables the plant to survive under them. Such, at least, is the result of my own observations on the plants of the Egyptian deserts. The distribution of similar forms of plants under similar con- ditions illustrates another fact. We speak of ‘“ chalk-loving,” “‘sand-loving,” and other kinds of plants frequenting special environments ; but these phrases seem to be, to a great extent, misnomers. Plants by no means all or always “love”’ the soils alluded to, in which they are often found. Many flourish quite as well, if not better, in a totally different soil. Having, how- ever, been located in them for many generations, they have become so adapted to the peculiar conditions of the soil and climate, by assuming such structures as are the best under the circumstances, that they now succeed in them; but at the same time many are always tacitly protesting, so to say, against their environment; for they at once show how much more vigorous they can become when they are grown in a different and more congenial soil *. Similarly we might just as well speak of desert-loving plants, corresponding to the term “ xerophile;” but we xnow what an intense struggle for existence they have to maintain. Never- theless they have become so inured to their difficulties, that * See Battandier’s observations, Bull. Soc. Bot. de Fr. 1887, p. 189 R22 220 REV. @ HENSLOW ON THE ORIGIN OF PLANT-STRUCTURES when seeds of desert plants are sown in ordinary garden-soil many fail to grow at all*; just as, while some water-plants grow more vigorously on land, others cannot live if exposed to the air. Or, again, with regard to many maritime plants, they will grow quite as well, if not better, away from the sea, either by altering their structure as Samphire does, by developing flat and thin leaves, or else they may retain their usual features, this being due to heredity. Thus Salsola Kali, the Prickly Glasswort of our sandy sea-shores, has become one of the worst weeds ever introduced into American wheat-fields. One year’s damage in Dakota alone is estimated at 2,000,000 dollars. Itis described as taking complete possession of the soil, while its spiny nature makes it objectionable to horses and other animals (/armer’s Bulletin, No. 10, 1893). In this case the plant has not lost its spines by growing in a richer soil, and illustrates the fact that hereditary influences often, if not always, tend, more or less, to resist the effects of a changed environment. And just as new adaptations can easily become fixed in some cases, but with great difficulty in others, as in cultivating wild plants ; so, conversely, while some features are instantly lost, others are as rigidly retained, though it may be in a modified form. What, however, may be called the general “ plasticity ” of plants is now so well recognized, that it affords us a perfectly adequate means in accounting for the self-adaptation of plants, although it is far from being necessarily applicable in every feature. I will now give some special characteristic features of desert plants, and then compare them with others growing in arid districts to show the more or less great similarity which prevails. II. General Morphological Characters‘. On first entering the desert near Cairo, where plants are to be found,—namely, along the lines of watercourses, which are * According to Dr. E. Sickenberger in Cairo and to my own experiments at home. t The best work Iam acquainted with on the structure of desert plants is one by Dr. G. Volkens, entitled “Die Flora der aegyptisch-arabischen Wiiste auf Grundlage anatomisch-physiologischer Forschungen dargestellt,” and the epitome “ Zur Flora der aegyptisch-arabischen Wiiste, eine vorliufige Skizze.” He gives eighteen plates (4to) illustrative of the anatomy of a well-selected series of types. I have fortunately been able to collect a large number of cne same piants myself n the deserts around Cairo; and for nearly all the BY SELF-ADAPTATION TO THE ENVIRONMENT. 221 dry all the year round excepting in February and March, while no plants occur on the higher ground at all,—the general appearance is of low bushes or isolated tufts of a nearly uniform grey colour. The plants are never crowded or cover the ground like an English roadside. In other words, they do not struggle for an existence with one another, but only with their inhos- pitable inorganic environment*. The grey colour is mainly due to intense hairiness, which subdues the green hue of chlo- rophyll. A few plants only, comparatively speaking, have no hair, and are consequently greener, as the species of Zygophyllum, which are fleshy-leaved plants; but a coating of wax is of frequent occurrence, and this aids in giving a glaucous hue. The hair and the wax, as well as the fleshy character of the leaves, are adaptations to arrest the loss of water by transpi- ration during the summer. IIT. Spinescent Characters. The next obvious feature is the stunted character of the bushes, three feet being about the maximum height (Zilla mya- grotdes), with gnarled stems at the base. This is often coupled with a spinescent character, either in the branches (as Zila), foliage (as Echinops), or stipules (Fagonia) and bracts (Cen- taurea). These features are, we may say undoubtedly, in the rest I am greatly indebted to the kindness of Prof. E. Sickenberger, of the School of Medicine, Cairo; so that I have been able to examine anatomically nearly the whole series described by Dr. Volkens, and have thus been able to supplement his observations in some degree in points he has not recorded. I would also refer the reader to the writings of M. P. Maury, Assoc. Frang. pour I’A vance- ment des Sci., Congrés de Toulouse, 1887 ; also Morot’s Journ. de Bot. ii. 1888, Rev. Bibl. p. 101. * Similarly of the desert regions of Beluchistan, Dr. Aitchison says — ‘‘The barren character of the country and the want of indigenous trees 1s due to the extreme dryness of the soil and aridity of the atmosphere. wee The struggle of plant-life for existence is great. The plants which are seen to exist through it all are either annuals or those possessing great root-stocks, tubers, tuberous roots, rhizomes, bulbs, or other such structural developments as assist them to baffle and survive through the extremes of temperatures.”—‘“A Summary of the Botanical Features of the Country tra- versed by the Afghan Delimitation Commission during 1884-5,” Trans. and Proc. Bot. Soc. Edinb. 1889 (read April 11th, 1889), p. 42. 222 REV. G. HENSLOW ON THE ORIGIN OF PLANT-STRUCTURES main due to a want of water *, which always prevents the for- mation of cellular tissue; while this deficiency of parenchyma is associated with a hardening of the fibro-vascular mechanical elements. The converse conditions are sometimes witnessed, e. g- in a plant of Zilla myagroides, which was the only species raised from seed out of many sown in the Botanic Garden of Cairo by Dr. E. Sickenberger. It not only bore well-developed leaves, but the spines, though formed through the forces of heredity, were very slender and subflaccid instead of being intensely rigid. As an interesting illustration of a highly spinescent plant belonging to the Cucurbitacew, an order in which it would be least expected, is the Narras plant of Caffraria (Acanthosicyos horrida). It grows on the sandy downs on the sea-coast. It has no leaves, but double spines studded all over the branches, forming impenetrable bushes which spread widely and attain the height of a man. It is curious that although the seeds germinate readily, all attempts to cultivate it in Europe have failed; just as Dr. Sickenberger and I failed with the numerous desert plants of Egypt, for, as stated, he was only successful with Zilla mya- groides. Another remarkably spinescent plant is Aciphylla squarrosa of the order Umbellifere, described by Sir J. D. Hooker in his ‘ New Zealand Flora,’ vol. i. p. 87. The hardening of the mechanical tissues generally, which so often results in special spiny processes, is brought about by drought and other conditions of the environment, and is one of the best means for resisting the intense heat of the desert. M. de Candolle ¢ called attention to this tact. He says :— ‘‘Very hard wood resists heat because it encloses but little aqueous juices, so there is but little to evaporate.” Similarly he remarks upon the great value of cork for resisting both extreme * So Mr. Belt, in ‘The Naturalist in Nicaragua,’ says (p. 46) :—* This spiny character of vegetation seems to be characteristic of dry rocky places and tracts of country liable to great drought.” Similarly Grisebach, aescribing the flora of Kalahari, South Africa, and alluding to spinescent species of Acacia, says:—* Tous ils portent des épines qui, chez A. horrida, ont 5 4 8 centimétres de longueur. La division limitée du feuillage et l’exiguité des surfaces sont autant de traits en rapport avec le dé- veloppement des organes piquants et de la sécheresse du sol” (La Végét. du Globe, ii. p. 252). He also alludes to similar features of the Tibetan flora (i. p. 614). + “Essai élémentaire de Géographie Botanique,” Dict. des Sci. Nat. vol. xviii. BY SELF-ADAPTATION TO THE ENVIRONMENT, 223 cold as well as extreme dry heat; as there will be a greater number of cells in the layers of cork filled with air, thus making bad conductors between the external air and the cambium layer and alburnum. Thus old trees resist cold better than young ones. On the other hand, endogenous trees having no bark only grow in warm climates. In some, as the Date, the bases of the leaves, especially if they decompose into a hairy covering, may supply the place of it. The evidence in support of the assertion that spines are the direct outcome of the environment also rests upon the well- known fact that there are many instances of plants losing their spines altogether when grown under other circumstances. This variableness in the spinescent character of plants is no new ob- servation. Thus, G. G. Kiichelbecker, in a ‘ Dissertatio botanico- physica de spinis plantarum’ (a.D. 1756), wrote as follows :— “Sunt autem quaedam plantae quae eundem semper et vbique seruant in extensione superficiei habitum, cum contra ea aliae, pro varia soli et culturae indole, formam hance alias sibi propriam deponant, vel tamen maximam sui partem mutent, ita vt, quae glabrae antea erant, nunc inaequalitatem magis minusue emi- nentem induant, atque suum plane deposuisse videantur habitum superficiei pristinum ” (pp. 9, 10). He then refers to Linneus’s Philos. Bot. p. 215, § 272 :—‘‘Spinosae arbores cultura saepius deponunt spinas in hortis...... Hirsuties loco et aetate facillime deponitur.” Again, 1. ¢. p. 247, § 316 :—‘ Solum mutat plantas, vnde varietates enascuntur, et, mutato eodem, redeunt. Hine Acanthi molles et aculeati: Cinarae aculeatae et non aculeatae.”’ Similarly, at the present day Pears, species of Rose and of Prunus, &c., are well known to lose their spines under cultivation. Ononis spinosa, L., has an excessively spiny variety, horrida, grow- ing in maritime sands. It is much less spiny in waste places by roadsides, &c., and becomes the variety inermis elsewhere. This latter form of the “subspecies” repens can be produced, tem-- porarily at least, at will; for when the ordinary spiny form O. spinosa is grown either in avery rich soil or with an abundance of water and a moist atmosphere, whether the plants be raised from cuttings or seeds, they gradually lose their spines; those first formed under these conditions are much reduced in size and in rigidity. Hereditary influence is too strong to arrest them at first altogether; although none are produced later on vigorous 224. REY. G. HENSLOW ON THE ORIGIN OF PLANT-STRUCTURES shoots growing in a saturated atmosphere. They reappear as soon as the same plants are allowed to grow in the ordinary way. It may be added that when the spines are arrested, the peculiar odour of O. repens is present, and the flowers also are larger and are like those characteristic of that subspecies *. M. Lotheliert carried out very similar experiments with other spinescent plants. He thus found that by growing Berberis vulgaris in a moist atmosphere it bore no spinescent leaves, as the parenchyma was well-formed between the ribs and veins; but in a perfectly arid atmosphere it bore spines only. Intensity of light also favoured their production. He found from a micro- scopical examination that in a section of a spine exposed to moist air the vessels of the xylem are few in number and the pericycle is not lignified. In a dry air the xylem forms a con- tinuous ligneous circle and the pericycle is also lignified. I have myself repeatedly corroborated this observation of the special consolidation of fibrous tissues in several instances in different species of desert plants. This consolidation of the mechanical elements may be perhaps explained as a secondary result of the relative abundance of the assimilative tissues ; the palisade-layers being often two, three, or even four in number, and equally on both sides of a leaf of a desert plant. While there is a great arrest in area of the parenchymatous tissue—due to the feeble water-supply—the organized products are mainly utilized in the lignification of the supportive tissues. Conversely, if plants be etiolated by being grown in the dark, while the parenchymatous tissues are relatively in excess, such as the pith and cortex, M. Rauwenhoff has shown that the mechanical tissues are greatly reduced, the assimilative tissues being quite incapable of any activity t. Another instance is supplied by our common Water-reed, * Such were the results of my own experiments in 1891-92. I raised plants from the seeds obtained from these plants in 1893, to see if the hereditary trait of producing spines becomes less pronounced in successive generations. In August the plants had only grown to about five inches in height in conse- quence of the drought; but up to the date of writing this (Aug. 15th) they had developed no spines. t Revue gén. de Bot. 1890, p. 276; Bull. Soc. Bot. de Fr. 1890, p. 176; Comptes Rendus, 1891, exii. p. 110. } “Sur Jes causes des Formes anormales des Plantes qui croissent dans Yobscurité,” Archives Néerlandaises des Sci. exac. et nat. t. xii. p. 297. BY SELF-ADAPTATION TO THE ENVIRONMENT. 225 Phragmites communis. This grass is very abundant in the Nile Valley, growing in places which are not artificially irrigated. It covers large areas of waste ground outside Cairo, furming a stunted growth, the leaves being very short and sharp- pointed. It has been named var. isarica. Close to the Nile, however, in Rhoda Island, it grows ten or more feet high, with Jong leaves almost exactly like the plants in English rivers. In many places the two forms of leaf are on the same stem, sometimes even alternating with each other, suggesting the idea that the leaves were elongated or abbreviated and spinescent, according as the plant happened to have sufficient water at its disposal or not. ‘ In this plant we have, therefore, a varietal character which is quite inconstant, as it varies repeatedly even on the same stem, and which has not become relatively fixed, though deemed worthy of a name. Such and other facts show how completely relative many varietal and even specific characters may be. They may have every degree of constancy, but they can often and easily change if the environment be altered, till other characters take their place. which may then become relatively fixed in their turn. I have said that all observations tend to prove that the great reduction of the parenchymatous tissues, with a correlated hardening of the mechanical tissues—so that a spinescent cha- racter becomes very characteristic of desert plants—are simply the inevitable results of the action of the environment. That this is so, is corroborated by an experiment of M. Duchartre on the effect of drought upon Dioscorea Batatas *. . Some tubers of this plant produced long shoots, but without being allowed to have any water at all. The effect upon the stem was that it acquired excessive rigidity with no heliotropism nor power of climbing. Although the stem was much more slender than normally, there was a marked predominance of the elements of consolidation. The fibres had very thjck walls and a minute lumen. The periphery of the central cylinder showed a zone of perceptibly greater thickness than usual, consisting of fibres of small diameter, but with walls of greater thickness than usual. This appears to have been the usual pericycular zone so common in herbaceous endogenous stems. Similarly, the collen- * Bull. Soc. Bot. de Fr. 1885, p. 156. 226 REV. G. HENSLOW ON THE ORIGIN OF PLANT-STRUCTURES chymatous cords at the angles were quite as much developed as in normally grown stems. Moreover, all these well-developed mechanical elements of consolidation in the Dioscorea were in a greater relative pro- portion, in consequence of the considerable reduction which the parenchymatous tissues had undergone. The leaves were small and undifferentiated, the parenchyma between the veins being much arrested; the stomata were un- developed and the palisade and lax central parenchyma all alike and unformed in character, except that the two of four larger layers in the centre had no chlorophyll. From the above brief ‘epitome of M. Duchartre’s experiment it will be seen that, so far as the small amount of parenchyma and the great density of the mechanical elements of the fibro- vascular tissues are concerned, we have an exact parallel between desert plants and this Dioscorea grown without water in the dark. Hence this experiment, together with those referred to above, will suffice to completely justify the conclusion that the indurated character of the mechanical system and also the spinescent features of so many desert plants are simply the imme- diate results of the effects of the comparatively waterless cha- racter of the environment. Lastly, it must be borne in mind that the spinescent character is hereditary, and although the rigidity begins to break down under cultivation in a moist climate and a good soil, the spines may still be formed but become gradually modified and finally disappear, reverting to leafy branches, as has occurred with cultivated pears, plums, and Ononis; or back into leaves, as in Berberis. That “xerophilous ” peculiarities of plants are not only cor- related with but are actually caused by the arid conditions of their environment is the conclusion of A. F. W. Schimper *, who has lately studied the flora of Java. He finds that the plants of the shores, alpine plants, those of the solfataras, and epiphytes, although very different from each other in a classificatory sense, all present common characters in being strictly “ xerophilous.” In all, the leaves are small and thick, there is a strong cuticle, * “Ueber Schutzmittel des Laubes gegen Transpiration, besonders in der Flora Java’s” (Mittheilungen aus den Sitzungsberichten der konigl. preuss. Akademie der Wissenschaften zu Ber n, 1890, Heft vii. p. 1045). See also Rey. gén. de Bot. 1892, p. 364. BY SELF-ADAPTATION TO THE ENVIRONMENT. 227 the stomata being at the bottom of “erypts,” the intercellular spaces reduced, hairs more abundant, the water-storage tissues present, &c. The author concludes from his observations that among the causes capable of determining the development of these various adaptations of defence against a temporary or permanent insufficiency of water are the dryness of the atmo- Sphere and soil, a strong insolation and rarefaction of the air, the richness of salts, or a too low temperature of the substratum. The stunted condition of the plants of the solfataras and of the shores would be due to the too great abundance of salts in the substratum, which has the result of reducing the assimilation of carbon. ; As a corroboration of the preceding, it may be added that MM. J. Vesque et Ch. Viet * came to the same conclusion, namely, that the fibrovascular structures were much more developed in a dry than in a moist atmosphere. I will conclude this section by quoting the following cor- roborative remarks of Grisebach f:—“ Au nombre de phéno- ménes généralement répandus figure le développement des épines, phénoméne qui va en croissant avec la sécheresse du climat. Les arbustes épineux plus petits qui habitent les steppes asiatiques et les solitudes du Sahara pénétrent dans les savanes de la contrée basse du Soudan (Zragacantha, Alhagi). L’exemple le plus remarquable de ce fait est fourni par le Sider (Zizyphus Spina-Christi), qui, sous la forme d’un arbuste ou arbre nain, s’étend depuis la Palestine jusq’au Seunaar et au Bornou. Mais dans le Soudan le développement des épines n’est point limité aux arbustes asiatiques de petite taille, ou 4 la forme de Sodada, puisque méme les arbres, notamment les Acacias, aussi bien que les plantes grasses, sont également armés d’organes piquants. Dans la Nubie, la majorité des arbres sont épineux, et il parait que dans certaines parties de VAbyssinie et dans le Bornou il n’est presque point de végétal ligneux sans épines. Une chose semblable est rapportée par M. Livingstone relativement aux contrées confinant avec le Kalahari, tandis qu’au contraire cette organisation s’évanouit sur le partage des eaux dans la direction du Congo.” * “De Vinfluence du Milieu sur la Structure anatomique des Vég¢taux,” Aun. Sci. Nat., Bot. sér. 6, tom. xii. 1881, p. 167. t La Végétation du Globe, ii. p. 197. 228 REV. G. HENSLOW ON THE ORIGIN OF PLANT-STRUCTURES IV. Foliage of Desert Plants. Another result of the deficiency of water is the small size of the leaves of desert plants, thereby lessening the surface of the transpiring organs; or else they are suppressed almost, if not quite, altogether, as in Retama, Anabasis, Ephedra, and Tamaria. Many plants produce moderate-sized leaves iu the early spring as soon as the rainy season commences, but none or very small ones later on, as Zilla, Alhagi, Statice, &c. The inrolled margins of the leaves, which make them assume the form of a more or less closed cylinder, is another common contrivance. : Now while this reduction of surface is beneficial by lessening transpiration, we must remember that it is simply the result of drought. This is proved in several ways: first, one and the same plant will produce much larger leaves in March or April, when a good supply of water is at hand, but minute leaves in June, when the supply is deficient ; secondly, if the same plant be grown in the Nile Valley it ceases to produce the smaller summer foliage and resembles the ordinary herbaceous leaves of temperate climates. Thus Salvia lanigera growing in the Delta has flat leaves, eight inches in length; but when in the Desert they are only about two and a half inches long, with the margins inrolled. A similar variability is a common phenomenon and is well known; for numerous instances might be given of leaves varying in form and structure on the same plant or on different individuals, according as they develop at different times or under different circumstances—or, again, if the plant have been transplanted, say, from a hot toa cold, or from a dry to a wet locality ; or, again, from a low to a high altitude or vice versd. Mr. Groom records a case of an orchid, Renanthera albescens, which naturally scrambles over plants growing on hot open sandy heaths. The specimen had been transferred to the botanic garden, Singapore, where it was growing under the shade of a well-foliaged tree. As might be expected, the form and anatomical details of the new leaves became much altered. Mr. Groom gives a series of comparisons showing how the cuticle decreased in thickness, while the dimension of the leaf increased in length but was diminished in thickness, &c.* Mr. Scott Elliot t has independently arrived at a very similar * «Annals of Botany,’ vii. p. 152. + “ Notes on the Regional Distribution of the Cape Flora,” Trans. Bot. Soe. Edinb, 1891, p. 241. BY SELF-ADAPTATION TO THE ENVIRONMENT. 229 result as to the cause of the small size of the leaves of the “‘ ericaceous’’ type which prevails in parts of the Cape and of Australia. He observes:—“The climate [of the Karoo] is characterized by a long and dry summer and by plenty of wind. Such conditions obviously favour transpiration ... Hence the small and excessively coriaceous leaves of these plants without much spongy parenchyma are thoroughly suited to the climate. We may even, I think, go a step further and say that the physical conditions have produced this form... With regard to leaves certain observations, which are not yet extensive enough for publication, as to the variation in size and texture of the leaf in the same species in different habitats, strongly incline me to believe that the smallness, cuticularization, and want of spongy parenchyma in the leaf all follow directly from such conditions.” Similarly M. E. Warming attributes a similar result to climate, especially the dryness of the atmosphere on the Campos of Lagoa Santa in Brazil. He writes :—“ Lorsque les feuilles ne sont pas tomenteuses sur les deux faces, elles sont ordinairement raides et coriaces ; Jes feuilles de plusieurs espéces, agitées par le vent, font entendre un bruit de cliquetis ou de crécelle trés extra- ordinaire ; tels sont certains Salvertia, Vochysia, le Palicourea rigida ou strepitans et certains Bombax. La majorité des espéces arborescentes ont les feuilles dures et coriaces, 4 quelque famille qu’elles appartiennent” *. Besides the general reduction in size of the leaves, plants in the desert reach the subaphyllous and aphyllous conditions. The degradation in the size and form of the leaf passes through many degrees till the leafless stage is reached, as in Ephedra and Retama. Just short of that, the leaves are minute, scale- like, closely adpressed against the stem, and assume the ap- pearance of the foliage of Thuja, Cupressus, &e. In the desert this form is well seen in Anabasis reticulata and Salsola Pachot. Such leaves may terminate in sharp points, as in Cornulaca macrantha, and as seen in our English Salsola Kali. This last- named genus, as also Zamarix, adopts both forms, just as Retinospora may have them, even on the same plant. That drought is the main cause is inferred from the fact that similar forms occur in plants in rocky and arid mountain regions at high elevations. Thus while the New Zealand nearly * Op. cit. p. 155 [supra, p. 218]. 930 REV. G. HENSLOW ON THE ORIGIN OF PLANT-STRUCTURES aphyllous species of Veronica live at an elevation of 7000 feet, V. thujioides is found at a lower but still high elevation. Similarly Thujas (Biotas) are trees and shrubs of considerable elevation. So many coincidences here offer the same grounds in support of the contention that similar causes have brought about similar results and produced these mimetic forms in genera of widely different orders. Thus Zamarix in Africa may be said to represent Halovylon Ammodendron of Beluchistan and the Oriental steppes (Aitchison and Grisebach), Casuarina of Australia, the Thujas of Japan and California, Veronica thu- jiotdes of New Zealand, &ce. V. Succulent Plants. Although spinescence and hairiness are the prevailing features in the desert plants near Cairo, some few are decidedly succulent, as the species of Zygophyllum. That this feature is one of the direct results of the intense heat (probably influenced by the presence of salts in the soil), inducing the formation of a thick cuticle, which, in turn, involves the retention of water and the development of succulent aquiferous tissues, I think cannot be doubted. The presence of salts has been proved by M. Lesage to be the immediate cause of succulency in maritime plants of temperate climates*; and he succeeded in making plants succulent which are not so ordinarily. On the other hand, the structure of “ rock plants,” such as Sedums, Haworthias, &e., is correlated with their arid and stony surroundings (probably without the aid of salts), and is obviously one of the many adaptations for the storage of water. That the succulency is due to the direct action of the environ- ment, is shown by the results of experiments in which the normal succulency is made to disappear when a new combination of surrounding conditions is supplied to the plant. Thus M. Battandier T cultivated Sedum Clusianum, and the leaves at once began to assume a flatter character, and he remarks as a coincidence that the two species S. stellatum and S. tuberosum, which are not rupicolous in France, but inhabit wet places, have flat leaves. On the other hand, I have found the leaves of S. stellatum growing in cracks between flat slabs of rock in Malta, * The succulency of several members of the Chenopodiacee which frequent saline marshes and deserts may be now attributed to the same cause. t Bull. Soc, Bot. de Fr. 1887, p. 191. BY SELF-ADAPTATION TO THE ENVIRONMENT. 231 in an exposed heated spot, to be more or less cylindrical. M. Battandier also says that two other species which are not rupicolous but grow in dry earth, viz. Sedum rubensand S. Magnoli, have flat leaves in a wet season but cylindrical leaves in a dry one. Similarly the common maritime fleshy Samphire of temperate climates, Crithmum maritimum, when cultivated in a garden became luxuriant and bore flat and smooth leaves. Centaurea crassifolia, a plant peculiar to the Maltese islands, growing in hot rocky valleys, has thick succulent leaves which survive during the hot season; but in March, when it begins to produce its new foliage, before the hot summer has approached, I found that the leaves were nearly as thin as in ordinary plants. As another example, M. Costantin observed * that Salsola Kali, a common inhabitant of maritime salt marshes, grows up sandy rivers, when it passes into S. Zragus by losing its fleshy leaves. The most elaborate series of experiments to test the source of the succulency of the maritime plants has been carried out by M. Lesaget, who shows conclusively that the presence of salt is at least a potent cause in its production. He succeeded in making plants, such as garden cress, succulent by watering with salt water. He also testifies to the hereditary effect, in that seed obtained from plants of cress which were somewhat succulent in the first year’s experiment became still more so in the following. The increased substance of the leaf is accompanied by a greater development of palisade-tissue with diminished inter- cellular passages and a less proportion of chlorophyll. This latter result is correlated with a relative decrease in the amount of starch produced. From all the above facts, natural and experimental, the conclusion is inevitable that while succulency is of benefit to the plants under the conditions in which they grow, especially by enabling them to store water during the hot and dry season, it is in all cases actually brought about by the direct action of the environment itself, coupled with the responsiveness of the pro- toplasm of the plant. * Journ. de Bot., 15 mars 1887. - t Rev. gén. de Bot. vol. ii. pp. 55, 106, 163; see also Comptes Rendus, cxii. 1891, p. 672. 2932 REV. G. HENSLOW ON THE ORIGIN OF PLANT-STRUCTURES VI. Protection of Buds. A feature very characteristic of the African desert grasses may be here mentioned—namely, the retention of the leaf-sheaths, so that the annual buds are carefully protected against drought when they appear in the spring. Similarly the Paronychiacee are provided with scarious stipules completely concealing the buds within them. Similarly Lavandula atriplicifolia has a spike of densely overlapping bracts (resembling the “ wheat-eared ” monstrous form of Dianthus) protecting the flower-buds. M. Warming noticed the same thing in Brazil. He thus writes :— “ Dans le groupe des Glumiflores (Cypéracées et Graminées) les feuilles sont étroites et raides; presque toutes les espéces de ce groupe sont tuniquées au sens ou M. Hackel a employé ce mot pour la premiére fois en 1889, c’est-d-dire que les bourgeons demeurent enveloppés et protégés par la base des feuilles qui persistent pendant longtemps, comme cela a lieu aussi dans le Posidonia oceanica (Andropogon, Rhyncospora, Scirpus sp., etc.).” In the African deserts numerous species of Aristida illustrate this fact. Lastly, it may be added that bulb-scales may become almost “woody ” (Allium Crameri, &c.). VII. Roots. The organs hitherto considered are mostly above ground, but roots also exhibit features of self-adaptation to desert life in the enormous length they sometimes attain. Dr. G. Volkens describes * how a young plant of Monsonia nivea of one year’s growth may be seen between July and January to have a small rosette of three or four leaves, while the roots may be twenty inches in length. Other plants may have roots two or more metres long. The Colocynth, he observes, has an enormous length of root in order to maintain its existence. It stands singly, has large herbaceous leaves without any means of preventing an excess of transpiration, as a cut shoot fades within five minutes ; and yet it flourishes unshadowed through the whole summer. The great length of root in certain desert plants has been also noticed elsewhere. Thus Dr. Aitchison observed in Beluchistan that “several of the Astragali (A. kahiricus, auganus, buchtor- mensis) have long whip-like roots, the bark of which is employed as twine by the people. These roots are extracted in a very neat way, by attaching a loop of twine to the crown, passing a stick through the other end and making it act as a lever ” t. * Op. cit. p. 7 [supra, p. 220]. t Op. cit. p. 431 [supra, p. 221]. BY SELF-ADAPTATION TO THE ENVIRONMENT. 233 M. de Candolle has also called attention to the advantages of long roots in enabling the plants to resist extremes of tempera- ture. He says:—“ L’action de la température est trés-sensible a la surface des sols, et l’est moins & une certaine profondeur ; d’ou il résulte, 1°, que, dans un terrain donné, les plantes a racines profondes résistent mieux aux extrémes de la tem- pérature que celles 4 racines superficielles; 2°, qu’une plante donnée résiste mieux aux extrémes de la température dans un terrain plus compacte, ou moins bon conducteur du calorique, ou moins doué de la faculté rayonnante, que aans un sol ou trop léger ou bon conducteur, ou rayonnant fortement le calorique ; 3°, la nature des plantes et celle du sol étant donneés, les plantes résistent mieux au froid dans une atmosphere séche, et 4 la chaleur dans une atmosphére humide” *. The cause of the long tap-roots of so many desert plants is the well-known responsive power of the apices to moisture, or hydrotropism. Similar phenomena may not infrequently be seen in England. Thus if, for example, the tip of the root of a seedling turnip gets into a field drain-pipe, it may grow to a length of some yards, of course never producing the turnip ft. As water is often to be found at various depths below the surface of the desert, the roots stimulated by ascending moisture continue to grow downwards till they attain very great lengths. Duration.—Some desert plants are usually annuals, the majority being perennials. A feature, bowever, which Dr. Volkens notices is that these characters are particularly liable to change in desert plants according to circumstances. Plants which normally live but one year, as species of Savignya, Poly- carpon, Malva, Trigonella, Ifloga, &c., may survive two or more years ; while perennials, like Capparis, Tamaria, Nitraria, Retama, Acacia, &e., may become annuals. The fact is that it simply depends upon the depth to which the primary tap-root descends so as to secure a more continuous supply of water below the surface, which enables the plant to survive the hot season. Mr. Scott Elliot has observed the same fact in South Africa f. In speaking of the prevalence of the “ericaceous” type, including 350 species of Erica itself, he remarks :—“It is not, I think, * «Essai élémentaire de Géographie Botanique,’ Dict. Sci. Nat. tom. xviii. t+ I have such a specimen, which was brought to the late Prof. J. 8. Henslow by a villager in Suffolk about the year 1890. t Op. cit., Trans, Bot. Soc. Ed. 1891, p. 244 [supra, p. 228). LINN. JOURN.— BOTANY, VOL. XXX. 8 934 REV. G. HENSLOW ON THE ORIGIN OF PLANT-STRUCTURES hard to see why this type should obtain so largely in the S.W. district. There is, to begin with, no winter worthy of the name ; and therefore annuals would, when becoming acclimatised, probably cease dying at the end of the year, because there is no reason why they should. Certain European annuals of cultiva- tion have, in fact, become perennials.” Now the annual, biennial, or perennial character of plants is often regarded as specific; but it is one which can be easily changed, and may then become hereditary: thus the garden form of the carrot is now biennial, but normally, as a wild plant, Daucus Carota, it is an annual. This is simply the result of sowing the seed of the original wild form late in the season. The consequence was that the plants did not blossom till the following year. Then, by selection, this biennial feature has been fixed, and is now hereditary. Poa annua, if grown in plenty of moisture, at once becomes a perennial; as it does also on the Alps, just as several other annuals at lower altitudes, as well as latitudes, become perennial when growing at higher altitudes and latitudes. Mr. Th. Holm * has recorded a number of examples of American plants which are ordinarily annuals, but become perennial under exceptional conditions. They include Hypericum nudicaule, Delphinium Consolida, Cyperus flavescens, Carex cyperoides, and species of grasses and Crucifers which are annuals in Europe, but perennials in the United States, and particularly so near Washington. For example, Arabis dentata, which is typically biennial, and A. /yrata, which is normally annual or biennial, have formed perennial specimens. On the contrary, A. levigata, said to be perennial by Hildebrandt, is not at all rare as a biennial near Washington. This change in duration may be accompanied by a change in the period of flowering ; or the period of flowering may change without a plant altering its duration of life. These alterations may become permanent. Thus, Sir J. D. Hooker? noticed how the stock and mignonette become perennials in Tasmania; on the other hand, the castor-oil becomes an annual in England. As a remarkable instance of a plant having undergone a complete change of season in flowering, Ovalis cernua may be * “On the Vitality of some Annual Plants,” Amer. Journ. Sci. xlii. 1891, p. 804; also Rev. gén. de Bot. 1892, p. 364. t Animals and Plants under Domest. ii. p. 305. BY SELF-ADAPTATION TO THE ENVIRONMENT. 235 mentioned. This is a native of the Cape of Good Hope, and flowers in the winter, 7. e. July; but throughout the whole of the Mediterranean border, where it has become dispersed since 1806, it blossoms from November to April *. Mr. Darwin has so fully discussed, under the head “Acclima- tization’’+, the variability of plants in adapting themselves to climate, and so becoming “ precocious” or “late” in flowering, that I need not enter upon this subject. All I would contend for is that such variations of habit are simply due to the respon- siveness of protoplasm to the environmental conditions, and that, when once acquired, they all tend to and may become hereditary traits. In corroboration of this I will conclude with the following observations by the late Dr. Lindley ¢:—‘ It often happens that, as in peas, the tendency in such plants to advance or retard their season of ripening was originally connected with the soil or climate in which they grew. A plant which for years is cultivated in a warm dry soil, where it ripens in forty days, will acquire habits of great excitability: and when sown in another soil will, for a season or so, retain its habit of rapid maturity; and the reverse will happen to an annual from a cold and wet soil. But as the latter will gradually become excitable and precocious if sown for a succession of seasons in a dry warm soil, so will the former lose those habits, and become late and less excitable.” VII. Histological Peculiarities of Desert Plants. Although the morphological features of desert plants are obviously adaptive, the histological elements illustrate the same fact even in the most minute details §. Eprpermis anp Curicte.—Commencing with the epidermis, a thickened cuticle in various degrees is of well nigh universal occurrence. There is also very frequently a superficial layer of wax. The cuticle is often covered with waved lines or ridges, especially on elevated cells, and the hairs with tubercles. * See my paper on “The Northern Distribution of Ovalis cernua, Thunb.” Proce. Linn. Soe. 1890-92, p. 31. t Animals and Plants under Domest. ii. p. 305. t ‘Theory of Horticulture,’ p. 465. (The italics are mine.) . § The reader might consult M. Vesque’s descriptions and figures of species of Capparis, showing how their anatomical structure conformed to their habits. “ L’Espéce végétale considérée au point de vue de l'anatomie comparée,” Ann. Sci. Nat. sér, 6, t. xiii, 1882, p. 5. 2 ; 8 236 REV. G. HENSLOW ON THE ORIGIN OF PLANT-STRUCTURES M. Dufour* found experimentally that the thickness of the external and lateral walls of epidermal cells is greater under sun- light than is that of the same species when grown in the shade; and the cuticle is also much more developed under sunlight. It need hardly be pointed out that in the desert the sunlight and glare reflected from the sand are very powerful, and therefore, ceteris paribus, the intensification of those elements mentioned is just what would be, on @ priori grounds, expected from M. Dufour’s experiments. But the thickness of the cuticle tends powerfully to prevent the loss of water, which is the end and aim of all desert plants in their adaptations to the climate. Dr. Volkens + observed that a considerable number of xero- phile plants are protected against a too energetic transpiration by the existence on the surface, outside the cuticle, of a thick layer of a sort of varnish, very brilliant in appearance. It is in great part soluble in alcohol, and without doubt of a resinous nature. An interesting fact, from the point of view of geogra- phical botany, is that such plants are peculiar to the southern hemisphere. The presence of wax on the surface of leaves was long ago observed by Mulder to be closely correlated with chlorophyll ; so that when we find the palisadic layers much increased, as they are in desert plants, we should, @ priori, expect a specially increased Jayer of wax. The production of this substance is dependent ultimately upon the increased amount of light [and heat ?|, which appears to deoxidize chlorophyll, and to leave wax as one of the products ; hence the epidermis becomes colourless and the cuticle coated with wax, as the wax increases with the loss of water. Upon this Mr. Herbert Spencer observes :—“ The deposit of waxy substances next to the outer surface of the cuticular layer in leaves is probably initiated by the evaporation [transpiration ?] which it eventually checks ” +. When the external walls of epidermal cells are flat, the cuticle is generally smooth; when, on the contrary, each cell forms a convexity towards the exterior, it is nearly always ornamented, either with parallel and straight or undulating strie, or they may be more or less reticulated, &. The above features * Ann. des Sci. Nat., Bot. sér. 7, tom. v. p. 311, t. viii. t “Ueber Pflanzen mit lackirten Blattern,” Berichte der deutschen botan. Gesellschaft, Bd. viii. Heft 4, p- 120 (1890). } ‘ Principles of Biology,’ ii. p. 245. BY SELF-ADAPTATION TO THE ENVIRONMENT. 237 are common in desert plants. M. Vesque offers the following interpretation * :—“ Tl serait bien difficile de déterminer le role de ces dessins cuticulaires par l’expérience; mais étant donnée cette circonstance singuli¢re que les épidermes plans en sont ordinairement dépourvus, tandis que les parties convexes en présentent presque toujours, il est permis d’émettre une hypo- thése & mes yeux fort plausible. Chaque cellule convexe repré- sente en effet une lentille convergente qui, malgré ses faibles dimensions peut, surtout dans les pays chauds, notablement surélever la température en un point déterminé de la cellule épidermique ; il est donc important, dans ce cas, de remplacer la vitre lisse par une vitre cannelée qui a pour “effet de disperser, d’égaliser la lumiére incidente; de cette mani¢re on comprend pourquoi, dans un grand nombre de cas, les cellules convexes qui avoisinent les stomates ou les poils enfoncés au-dessous du niveau de |’épiderme et celles qui se relévent en petites sallies autour de la base des poils sont striées tandis que les autres ne le sont pas.” As far as sheets of glass with striated and reticulated surfaces can imitate a cuticle, I find that a sheet of sensitive paper is not darkened to the same extent under them as under a clear sheet of glass of the same thickness when fully exposed, and for the same time, to sunlight; though nothing could be deduced from any differences of temperature under the same circumstances. Priostsm.—Since a more or less excessive hairiness is a cha- racteristic feature of the great majority of plants growing in hot and barren deserts, the question arises as to what is the cause. Now any extra outgrowth, even if it be but epidermal tri- chomes, implies the presence of more nutritive materiais at the disposal of the plant at the spot than when they are not formed at all. M. Mer, who studied the question, came to the con- clusion that, ceteris paribus, hairs are due to a localized extra nourishment, and therefore frequently occur upon the ribs and veins, 7. e. immediately over the channels of sap. He thinks this view is supported by such a case, e. g., as Rhus Cotinus, in which the abortive pedicels, which bear no fruit, develop a Jarge amount of hairs, while the pedicels which bear fruit have few or none. The excess of hair is therefore presumably due to a compensatory distribution of sap. * Op. cit. p. 34 [supra, p. 235, note]. 938 REV. G. HENSLOW ON THE ORIGIN OF PLANT-STRUCTURES M. Aug. Pyr. de Candolle came to the same conclusion in 1827 ; for he thus wrote about Rhus Cotinus :—‘ Peut-étre la sve destinée & nourir les fruits ne trouvant plus d’emploi, lorsque ceux-ci ont avorté, produit-elle ce développement extraordinaire de poils. Quelques filets d’étamines (Verbascum, Tradescantia) deviennent aussi poilus quand les anth¢res avortent, et pro- bablement par la méme cause’”’*, Dr. M. T. Masters observes, when speaking of the hair on the barren pedicels of Rhus Cotinus, or the “ Wig-plant,” as it is called :—“A similar production of hair may be noticed in many cases where the development of a branch or of a flower is arrested ; and this occurs with especial frequency where the arrest in growth is due to the puncture of an insect, or to the formation of a gall}. As an illustration of this last-mentioned fact, it may be often noticed how Veronica Chamedrys terminates its shoots with an excessively woolly globular bud. A similar thing happens to a heath, Hrica scoparia, common on the hills around Cannes. In the latter plant the abnormal leaves are broad, ovate, and densely hairy ; while the ordinary leaves are linear and glabrous. These globular structures in both plants are due to the irritation set up by the presence of grubs. The axis and the innermost leaves are arrested at the apex, while in compensation the lower leaves of the bud alter their character, enlarge and become densely elethed with hair f. The fact of hairs being developed over and about the fibro- vascular cords is of common occurrence, and, under the above aspect, becomes very significant in such cases as in desert-grasses, Javender, &c. Pfitzer observes, “Almost all grasses inhabiting very dry localities have leaves with well-marked longitudinal folds” $, the stomata being situated within the grooves, while the ridges correspond to the vascular and fibrous cords. The tooth- like (or branched, as e. g. on the calyx of Lavender) hairs of the grasses project like chevaua-de-frise over the grooves. They have swollen bases capable of imbibing moisture; and so in all * ‘Organographie Végétale,’ tom, i. pp. 111, 112. t ‘Teratology,’ p. 472. + I have elsewhere called attention to this fact in its analogy with the results of the irritating action of the pollen-tube. See ‘Origin of Floral Struc- tures,’ p. 164 seqg. § Quoted by De Bary, Comp. Anat. &e. p. 50. BY SELF-ADAPTATION TO THE ENVIRONMENT. 239 probability absorb dew, as well as protect the surfaces against a loss of water. The marginal inrolling of the blade is, of course, an additional protection. To such an extent does this occur, that many blades of desert-grasses are perfectly cylindrical; the upper surface, which is the especially grooved one, being entirely concealed from view, as may be well seen in Volkens’s figure of Aristida ciliata*. In some cases, as in this grass, certain hairs assume a papillate form immediately over the stomata. As another illustration which seems to support M. Mer’s con- tention that hairs are, ceteris paribus, a result of compensation, M. Lesage +t found in a root of the second order of Phaseolus, which was much longer than the primary root, that the portion outside the water was covered with numerous root-hairs ; zear the water these hairs were elongated, while in the water they were much shorter, and finally disappeared altogether. In a transverse section it was seen that the cortical layers in the air contained smaller elements than those in the water; and in the central cylinder the xylem was proportionally more lignified in the aerial portion. The root of the bean was made the subject of similar obser~ vations. It was found that when numerous secondary roots were suppressed, the primary root was covered with numerous absorbing hairs. The above interpretation will, therefore, satisfacterily explain the existence of the hairiness of plants in the deserts ; for drought, aided by the barrenness of the soil, tends to arrest the develop- ment of parenchymatous tissues; and in proportion as this arrest is excessive, so is the compensating process of the production of hairs. Hence, just as with plenty of water ora good soil, as obtains under cultivation, plants tend to become less hairy than in the wild state, as e.g. the parsnip, 80, conversely, under aridity and a poor soil, hairiness becomes a characteristic and hereditary feature. Here again, therefore, if the above explanation of M. Mer’s be true, the very conditions which bring about the production of an excessive clothing of hair are precisely those against the severity of which this dense clothing is one of the very best of protections. _ The above interpretation receives indirectly an additional * Op. cit. pl. xvii. fig. 4 [supra, p. 220]. t Comptes Rendus, exii. 1891, p. 109. 240 REV. G. HENSLOW ON THE ORIGIN OF PLANT-STRUCTURES countenance from the fact that just those desert plants which do develop much parenchyma and become exceedingly fleshy, as the Aloinee, Euphorbia, Cactacee, Zygophyllum, «c., are generally entirely hairless. Intense hairiness is certainly one of the most conspicuous features of desert plants, and is an invaluable means of lessening the heat by forming a non-conducting surface ; and, on the other hand , is ameans of absorbing dew during the summer when no rain falls. Dr. G. Volkens remarks that a multitude of cases establish the empirical deduction that drought is correlated with the pre- sence of much hair, though what the real causal connection may be is not so clear tohim*. He observed that as transpiration tends to increase, so does the relative quantity of hair, till (it may be added) the quantity becomes so great as to check the very process which may have had something to do with bringing it about. In desert plants the hairs are of different forms. They may be stiff, straight, and adpressed to the surface, all lying in one direction ; or the “needles ’’ are interlacing ; or the hairs may be of a twisted cottony character and cover the surface with a layer of wool; or, again, they may be stellate and flat, the branches interlacing so as to produce a dense coating of felt; or they may be bladdery and filled with water. These latter may finally collapse, dry up, and form a glassy sheet. There are also peculiarities in the structure of the hairs them- selves which are remarkable. In many cases the cavity is quite filled up with the exception of the broad basal part of the cell, while the outer surface may be densely coated with wax, either entirely or with “ gashes” and “ pores ” (Diplotaxis Harra), or else the basal portion is quite devoid of it (Heliotropium luteum), so that water can readily pass from without into the interior. Dr. Volkens observes, with regard to the filling up of the lumen with cellulose matter, that this is connected with the swelling up of the inner membrane, and shows (by treatment with suitable reagents) that it is a substance which, when water is absorbed, can hold it fast with great strength. This, of course, greatly retards transpiration. I have found in some cases, besides the closely applied felt, that there are taller branching hairs standing much above the * He makes no allusion to M. Mer’s observations, which were probably there- fore unknown to him, BY SELF-ADAPTATION TO THE ENVIRONMENT. 241 level (Erucaria aleppica). TheseI take to he specially absorbing hairs, as well as all which have no wax, or at least places where it is wanting. As a special peculiarity, I have found in Cocculus Leeba a basket-like arrangement of hairs round the axillary buds. This would presumably retain a large drop of dew, by means of which the bud would be benefited. Besides being actual absorbents of dew*, it may be borne in mind that not only does the felt collect the dew-drops, which get entangled in it and so get absorbed, but it parts with them by evaporation much more slowly than does a smooth leaf, as I have tested by experiment with many kinds of leaves. Now, as Dr. Volkens observes, though suspected, he cannot say for certain how or why the hairiness is produced. But, besides, the reasonable interpretation of M. Mer which I have given above, that hairiness is a direct result of the environment upon the plant, is established by numerous cases. In the first place we have “ the argument of coincidences,” as I would call it ; but when we find that a change of habitat brings about a greater or less degree of hairiness, the probabilities accumulate till they amount to a moral conviction, which is further established by experimental verification. Thus, for example, the hairiness of wild plants tends to decrease under cultivation, as in the wild parsnip. Linneus observed this fact nearly two hundred years ago for he sayst:—“ Spinas et hirsutiem ... plante sepius exuunt a loco vel cultura.” Ranunculus repens growing in a dry barren gravelly soil is very hairy, but another plant growing in water ten feet from the former, which I have preserved, is scarcely hairy at all. M. Battandier observes + that Bellis atlantica, with leaves covered with a true velvet having a long pile, at the summit of the mountain Blida, when cultivated in Algeria, bore leaves less and less veiv ety till they finally became as glabrousas the Bellis of Algeria. Similarly Allium Chamemoly, when cultivated for eight years, lost its villosity which it had on the summit of Zaccar. So also * For proof that plants can absorb rain and dew by their green parts, [ would refer the reader to my paper “ On the Absorption of Rain and Dew by the Green Parts of Plants,” Journ. Linn. Soc., Bot. xvii. p. 313. t Philos. Bot. § 272. ¢ Bull. Soc, Bot. de Fr. 1887, p. 193. 942 REV. G. HENSLOW ON THE ORIGIN OF PLANT-STRUCTURES Cerastium Boissieri, which is white and tomentose at the summit of Ait-Ouabau, became glabrous and of a beautiful green at Algiers. Hairiness is well known to be a most variable character, and although it is recognized as specific when constant and abundant, as in Verbascum, and therefore less variable, it obviously becomes much less important when it fluctuates. It may be observed here that the hairiness as an hereditary character varies greatly. Thus, M. Battandier found that Bellis atlantica varied when transplanted, yet when raised by seed in Algeria did not show similar modifications ; similarly Allium Chamemoly, though it was less villous, remained more hairy than its congener after cultivation as well as by sowing. On the other hand, seedlings of Pastinaca sativa raised in a prepared border in the botanic gardens of the Cirencester Agri- cultural College became less and less hairy, and finally quite glabrous*. As a spontaneous variety, Malva parviflora may be mentioned ; as grown in the Nile Valley as a culinary vegetable it is not very hairy, the hairs being stellate, having only a few rays, sometimes two only. As a desert plant, where it appears as a small annual for a few weeks only, the stellate hairs increase their number of rays. Similarly Hrodium laciniatum is much less hairy when growing in waste ground on Rhoda Island in the Nile by Cairo than when in the desert. I find that the main differences between the two forms may be summed up as follows :—Hypo- dermic collenchyma is much thicker in the petiole of the desert form. The upper epidermis has its cells similar in size but with more hairs. The lower epidermis has cells which are much smaller than those of the Nile Valley form. The palisade, consisting of two rows of cells, is identical in both cases below the upper epidermis; but while there is a lax mesophyll of rounded cells on the lower side in the Nile Valley form, the corresponding cells are somewhat elongated in shape, thus tending to assume the palisadic type characteristic of the lower side as well as of the upper in the desert furm. Hairs of the latter are more numerous in the proportion of 24 to 10 for the same area. Stomata.—These structures are frequently sunk below the * Prof. Buckman raised the “domesticated ” variety alluded to in 1847, which he called the “Student.” It is still regarded as the “ best variety ” in cultivation, according to Messrs. Sutton & Sons of Reading (1894). BY SELF-ADAPTATION TO THE ENVIRONMENT. 243 level of the outer surface of the leaf and may occur on both sides. The depression in which they lie is either in consequence of the great thickness of the outer wall of the epidermis (Allium Cra- meri, Pityranthus tortuosus, &c.), or because the epidermis itself lines cavities in the parenchyma which are either naked or clothed with hairs covering over the stomata, as in the grass Danthonia Forskali, very like the well-known case of Nerium Oleander, to which M. Vesque adds Capparis Breynia. The Oleander has narrow, rigid, more or less erect leaves which are well suited to live in a dry atmosphere. In a specimen growing at Cannes there was a thick cuticle and two layers of thick- walled hypodermic cells, a palisade-tissue of two layers on the upper side and of one layer on the lower. A lax mesophyll of green cells fills up the central space. The stomata are on the bottom of the epidermal cavities. In a leaf gathered from a tree in Cairo, there were slight differences, in that there happened to be no palisade-cells on the lower side at all, the lax mesophyll reaching to the hypodermic layer. Such differences are probably accidental ; but they show clearly how easily the anatomy of a leaf conforms to slight differences of illumination, &e. In desert plants the guard-cells are often so thick-walled that the lumen is nearly obliterated ; and, contrary to what is gene- rally supposed to be the case in temperate regions, Dr. Volkens shows that they often close during the day and are open at night*. Perhaps the arrested moisture due to the check to trans- piration may cause turgescence by day, which closes the slit, while the cessation at night brings about a relaxation ; or it may be the result of a more complicated action between the guard-cells and the adjacent epidermal cells. It is, however, difficult to say without a very close investigation into the phenomena on the living plants in their natural conditions. IX. Assimilative Tissues. The chlorophyll-tissue of an ordinary dorsi-ventral leaf is typically differentiated into a palisadic layer below the upper epidermis and a spongy layer above the lower epidermis. It is well known that in those plants in which the leaf is normally reversed in position, as Als¢rwmeria, the relative positions of * Die Flora der aegypt.-arab. Wiiste, p. 47. 944 REV. G. HENSLOW ON THE ORIGIN OF PLANT-STRUCTURES these two layers are also reversed. This inversion can be more or less decidedly brought about by compelling certain leaves to develop their surfaces reversed. Moreover, when leaves assume a vertical position, so as to be equally illuminated on both sides, then the epidermis and subjacent palisadic tissue are also more or less alike, as in some grasses, the phyllodes of Australian Acacias, the pendulous falcate leaves of gum-trees, though the first-formed horizontal leaves on the same gum-trees are dorsi- ventral in structure like ordinary horizontal leaves *. The preceding facts all conspire to prove that the differen- tiation of the mesophyll into palisadic tissue is the direct result of illumination; and that, as the upper surface of normally horizontal leaves receives more light than the lower, the differen- tiation of the chlorophyll-cells into an elongated form parallel to the incident light is the direct outcome, ceteris paribus, of that external agent. A significant fact with regard to these alterations is that the change evinces itself by degrees. Thus in young leaves of Alstremeria psittacina, M. Dufour ¢ shows that it is only partially effected. Thus the first leaf stands vertically. It possesses few stomata, and is nearly alike on both sides. The second leaf is still nearly vertical, being only twisted towards the apex. It has stomata on both faces at the base, but at the point shows many upon the upper, but none on the lower side. The subsequent leaves have the petiole twisted and are com- pletely reversed in position ; their petioles have a small number of stomata on the two faces in the limb, exclusively upon the * The chief differences between the two forms of leaves in Eucalyptus I find to be as follow :—In the horizontal leaf the upper epidermis is composed of small cells and there are no stomata. There is a palisade-tissue of one layer of cells, with lax mesophyll below the lower epidermis. ‘This latter has larger cells than the upper and is provided with stomata. The pendulous leaf is a good deal thicker than the horizontal. Both epidermides are provided with a very dense cuticle in which the stomata are deep-seated. There are four rows of palisade-cells on both sides with a chlorophyllous mesophyll between them. The petiole is flattened so that the leaf can swing much in the same way as that of the Poplar. The horizontal leaves in E. Globulus are sessile. M. G. Briosi has written a voluminous work with many plates upon the leaves of Eucalyptus Globulus, Labil., to which the reader is referred for numerous details (‘ Intorno alla Anatomia delle foglie dell’ Eucalyptus Globulus, Labill.,’ 95 p. et 23 pl., Milano, 1891). t Bull. Soc. Bot. de Fr., 23 mai, 1886, p. 269. The author here gives several references to the literature of the subject. BY SELF-ADAPTATION TO THE ENVIRONMENT, 245 superior face, now become the lower in position. Another inter- esting point is that in young leaves the mesophyll is homo- geneous; but the cells under the upper epidermis are rather Jarger at first in the older leaves, showing an hereditary tendency to develop palisade-tissue; but later on the influence of the light soon renders the cells below the—now reversed—inferior side much longer than the others. The needle-like leaves of the Norway Spruce prove, according to the careful investigations of M. Mer *, to be most remarkab.y sensitive to light—becoming more quadrilateral with a uniform palisade-tissue as the leaf grows exposed, but flatter and more dorsi-ventral when it is more shaded on the same tree. He also observes the remarkable result of frost setting up a “habit” in the tree, as follows :—“ Trees planted isolated have some modifications. They become more bushy, from the deve- lopment of lateral buds, which are arrested in forests. When the young shoots are struck by frosts in the spring it happens at times that, without reaching the point of death, they lose their turgescence. They become soft, and their extremity turns towards the ground. Lignification supervenes before they have recovered their turgescence, and they remain thus definitely curved. When the terminal bud is not destroyed it develops its succeeding shoot the following year in this position. If one places it vertically, the terminal bud turns downwards, and the branch in course of development preserves this situation ; or rather its extremity tends to elevate itself by a slight curvature. This depends upon its degree of vigour and the time during growth when the experiment is made.” ‘This passage is very suggestive as a cause of “ weeping” varieties of trees, by a temporary injury producing a permanent effect in the growth, though not to the extent of being heredi- tary. Again, Dr. F. Nollt has shown that external influences determine not only the direction of some organs, but also the position in which they are formed ; as, e. g., the development of the gemma of Varchantia, of aerial roots on climbing plants, &e. In other and more numerous cases the formation of fresh organs appears to be independent of external forces, and to be determined only by the internal, ¢. e. hereditary, forces in the * Bull. Soc. Bot. de Fr. 1883, p. 40. + See Journ. Roy. Micr, Soc. 1891, p. 490. 246 REV.-G. HENSLOW ON THE ORIGIN OF PLANT-STRUCTURES plants; as, for instance, in the dorsi-ventral structure of many parts of plants. In Bryopsis the reversal of the plant brings about a corresponding internal organic transformation. M. Warming notes the same feature in the leaves of plants growing in the arid campos of Lagoa Santa. He says *:—“ La direction des feuilles accuse également la sécheresse du climat ; beaucoup d’entre elles ont habituellement une direction verticale ou sont au moins trés relevées, de manicre a n’étre frappées par les rayons solaires que sous un angle aigu. Certaines espéces ont des feuilles trés réduites et quelques-unes sont aphylles ; elles appartiennent 4 des familles trés différentes.” The reader will perceive that this description would apply equally well to many desert plants of the northern hemisphere. Applying the preceding observations to desert plants :—If a leaf be small, narrow, and moreover assume a more or less vertical position, as is so generally the case, so that it is illumi- nated nearly equally on both sides, we should expect to find on a priori grounds that it would have palisade-tissue on both sides. Such is precisely the case with innumerable desert plants. The only and indeed relatively rare exceptions are in the leaves of such plants as develop their foliage during the rainy season, as annuals, or in “ Nile Valley” plants which happen to secure a place in the borders of the desert. Such have a more or less characteristic spongy parenchyma on the underside ;_ while transitions from this to true palisadic tissue are easy to be found. The typical chlorophyll-tissue in leaves of true desert plants is therefore palisadic on both sides, the cells being arranged in from one to four or even five superposed rows. These sometimes meet in the middle (Zizyphus Spina-Christi), in others there is a central layer of short rounded cells (Cassia obovata), the usual lax merenchyma being entirely wanting. In addition to the typical palisadic cells at right angles to the surface, large cells palisadic in shape may be arranged in cylinders around the fibro-vascular bundles. This is particularly well seen in grasses; but within this cylinder is a second, of chlorophyllous cells, which are quadrate and short in form. This peculiar arrangement of a double cylinder of green cells is not confined to grasses, but occurs in exogens, as Tribulus alatus, Atriplec Halimus, &.; so that in all cases the fibro-vascular cords are densely imbedded in chlorophyllous tissue. * Op. cit. p. 155 [supra, p. 218]. BY SELF-ADAPTATION TO THE ENVIRONMENT. 247 In some cases the inner sheath is colourless and acts as a water-storage tissue (Oligomeris subulata). It is,I think, really homologous with the endoderm. M. Dufour * found that chlorophyllous tissue is much more developed in sunlight than in shade; and it may be added that a similar phenomenon occurs in Alpine plants and plants of high latitudes as compared with the same species growing at low altitudes and latitudes. In these cases the deeper green tint is also due to the uninterrupted sunlight. Hence it is not sur- prising to find the chlorophyll-tissues reach a high development under the intense light in the desert, whether from the sun direct, or reflected upwards from the sand. X. Ligneous Tissues. Messrs. Dr. D. H. Scott and G. Brebner have described f the histology of Acantholimon glumaceum (Plumbaginez), and Prof. D. Oliver had previously examined { that of Acanthophyllum as well (Caryophylleew). These are both desert plants ; and I might now add a great many more anomalous and subanomalous stems of plants growing in the North-African deserts, some of which Dr. G. Volkens has described and figured §. The general con- clusion deducible from a study of their peculiarities is that these are due to the climatal conditions under which they grow. Summing them up, they may be enumerated as follows :—There is a general tendency to lignification; with an absence of me- dullary rays (Zilla myagroides, Bassia muricata); if they be present, they are comparatively few and have thick walls (Farsetia africana, Helianthemum kahiricum, Ochradenus baceata); the fibro-vascular cords may form “islands,” as seen in a trans- verse section, imbedded in dense tissue (Statice pruinosa, Atri- plex leucoclada, Pityranthus tortuosus) ; the “ wedges” of wood may fail to form a regular zone, but be more or less isolated and imbedded in water-storage tissue (Zamarix mannifera, Anabasis articulata). The wood may be deeply indented with cortical invasions, as Prof. Oliver has shown to be the case in Acanthophyllum. Dr. Scott observes that this is due to the fact * Ann. des Sci. Nat., Bot. sér. 7, tom. v. p. 311. t Ann. of Bot. vol. v. p. 259. ¢ Trans. Linn. Soe. vol. xxii. p. 289. § Die Flora der aegyptisch-arabischen Wiiste, &c. 248 REV. G. HENSLOW ON THE ORIGIN OF PLANT-STRUCTURES that the zone does not close up after having early parted with a cord for an appendicular organ. I think we may attribute this failure to a want of activity in the formation of wood, which may be correlated to the insufficiency of foliage during the hot months*. Perhaps the curious cavities described + by Prof. D. Oliver in a species of Acantholimon as occurring in the wood may be attributed to the same cause. As other examples of anomalous stems, Dr. Volkens describes and figures that of @ypsophila Rokejeka: the pericycle (which is often very active in the Caryophyllex) puts on a very dense zone of thick-walled sclerenchyma, in lieu of xylem, which in this plant is comparatively feeble. A similar result occurs in Tele- phium spherospermum, of the allied order Paronychiacee. I have found an analogous result of pericycular activity in Iphiona mucronata. Tuside a cortex there are zones of square, oblong, or irregular shaped patches of sclerogen imbedded in a paren- chymatous tissue t, and only isolated patches of xylem around the medulla, all being imbedded in a “stereome-zone.” ) i if { / ad Hanhart imp ORIGIN OF PLANT STRUCTURES BY SELF-ADAPTATION TO THE ENVIRONMENT. 263 The flowers are small and remain only partially open. Allium Crameri is one fifth of an inch long, A. desertorum one fourth, and Dipcadi erythreum one half. In all of these the anthers crowd round the stigmas and pollinate them in the half-opened flowers (Pl. XII. figs. 30,31). Dipeadi is remarkable for having each stigma bifurcated, as in Euphorbia (Pl. XII. figs. 32-35); while Allium desertorum (P1. XII. fig. 30) has a globular stigma. Fig. DNS oP Oo bo EXPLANATION OF PLATE XIL. - Petal of Erucaria aleppica. . Petal of Malcolmia egyptiaca. - Petal of Savignya parviflora. . Stamens and pistil of Malcolmia egyptiaca. . Stigma of Erucaria aleppica with closely adpressed stigmatic lobes. . Stamens and pistil of Savignya parviflora with globular stigma. . Open carpel and naked ovules of Caylusea canescens. . Portion of flower of Silene villosa showing position of sepal (s), petal. stamens, and pistil on the gynophore. . Pistil and stamens of Polycarpea memphitica. . Pistil and stamens of Reawmuria hirtella. . Pistil and stamens of Cassia obovata, . Hollow stigma of Cassia obovata. . Style and stigma of Astragalus bombycinus. . Stigma of Astragalus bombycinus. . Stigma of Astragalus Sieberi. . Hollow stigma of Pulicaria arabica. . Tufted stigma of Artemisia monosperma. . Anthers and stigmas of Senecio egyptiacus. . Anthers and stigmas of Franceuria crispa. . Floret of Broechia cinerea. . Stigmas of Brocchia cinerea. . Pollen-grain (smooth) of Centaurea egyptiaca. . Cleistogamous flower of Wahlenbergia cervicina. . Corolla of same opened, showing the position of the curved stigmas and anthers, . Corolla of Campanula sulphurea. . Stigmas and style of same, nearly devoid of “ collecting hairs.” . Hollow stigma of Linaria Helava. . Flower of Salvia lanigera. . Relative positions of anthers and stigmas in the flowers of Salvia egyptiaca and of S. danigera. . Three stamens and pistil of Allium desertorum. . Three stamens and pistil of Allium Crameri. 32. Portion of flower of Dipcadi erythreum. 33 & 34. Outer and inner perianth-leaves of same. 35. Bifurcating stigmas of same. 264 MR. W. WEST ON SOME FRESHWATER On some Freshwater Algx from the West Indies. By Wo. West, F.L.S., assisted by G. S. West. [Read 1st February, 1894.] (PuatEs XIII.-XVI.) Turoven the kindness of Mr. George Murray, of the British Museum, I have been enabled to examine some Freshwater Alge which were collected by Mr. W. R. Elliott on the islands of Dominica and St. Vincent—from the former in November and December 1892, and from the latter in May of the same year. The material was preserved in weak spirit, and has proved to be very interesting. The alge were in numbered bottles; these numbers are used throughout the paper, as both material and slides, with corresponding numbers, can be consulted in the Museum. The strictly aquatic species were mostly from warm or hot streams. Three of the gatherings were from mossy trees; these latter must have been very moist, as several species of Desmidiex occurred amongst the various Cyanophycee which formed the greater part of this material. Some of the species were very abundant, e. g. Symploca cuspidata, n. sp.; there were others intermingled with these, either in small patches or solitary, and very sparingly. Summary. Varieties Genera. Species, and Forms. Confervaceze .......... 2 2 Chroolepidee .......... 1 2 Zygnemacew® .......... 1 1 1 Desmidiaceew .......... 3 6 1 Volvocinee............ 1 1 Protococcacee.......... 2 2 Nostocacee............ 1 2 Scytonemace®.......... 3 7 Sirosiphoniacew ........ 2 7 1 Oscillariacee .......... 3 4 1 Chroococcacer ........ 6 12 1 Diatomacee............ 8 14 1 33 60 6 Of the above, 11 species and 4 varieties are new. ALGH FROM THE WEST INDIES. 265 “ Crass. fil.” (as used in this paper) = the diameter of the sheith containing the trichomes. “ Crass. trich.”’= the diameter of the cells without the sheath. Class CONFERVOIDER% ISOGAME. Ord. ConFERVACEE. 1. Conrerva BomBycina, 4g., forma mrtnor, Wille. (Om Hvileceller hos Conferva, p. 21, t. 1. figs. 36-40, t. 2. figs. 55, 56.) Crass. fil. 6°5-7°5 p. Hab. In cold, warm, and hot streams, crater of Grande Soufri¢re, Dominica. Nos. 880, 883, 884, and 908. 2. Rurzoctontum BEraGRentanum, Hauck in Nordst. Freshw. Alg. of New Zeal. §& Australia, p. 17. Var. DOMINICENSE, nov. var. (Pl. XIV. figs. 17-24). Var. filamentis dense intricatis, tenuior, articulis diametro 2-3plo longioribus; membrana cellularum 1 p» crassitudine. Crass. cell. veget. 9-10 p. Hab. In hot stream in crater of Grande Soufriére, Dominica. No. 880. In the more densely intricate portions almost every cell emitted a short lateral branch, usually consisting of one cell. Ord. CHROOLEPIDES. 3. TRENTEPOHLIA OpoRATA, Wittr. (Scand. Victer. t. 4. p. 16.) [Chroolepus odoratum, Ag. Syst. Aly. p. 35.] Crass. cell. veget. 12-20 1; crass. zoogonidang. 33-37°5 pu. T. zoogonidangiis lateralibus, interdum terminalibus, subglo- bosis, sessilibus (hine inde pedicellatis). Hab. On bark, Bow-wood Hills (1580 ft.), St. Vincent. (8-12-91.) No. 23.—On trees, Government House, St. Vincent, 900 ft. (10-12-91.) No. 27. 4. T. vintosi, De Toni. (Syll. Alg. i. p. 239.) [Chroo- lepus villosa, Auetz. Phyc. gener. p.. 234; Species Algarum, p- 428.] Crass. cell. veget. 18-23 p. Hab. On rocks, Soufritre, Dominica. (2-7-92.) No. 20. LINN. JOURN.— BOTANY, VOL. XXX. U 266 MR. W. WEST ON SOME FRESHWATER Class CONJUGATS. Ord. ZYGNEMACE. 5. ZyaNeMwa (ZYGOGONIUM) PACHYDERMUM, nov. sp. (Pl. XIII. figs. 1-16.) Z. cespitibus intricatis; filis flexuosis (et interdum genu- flexuosis) dense intricatis, hine inde ramulis brevibus irregulari- busque cellularam 2-6; cellulis vegetativis diametro 2-23plo (sepe 3plo) longioribus (rarius equalibus); membrana cellu- larum crassa vel crassissima; zygosporis in tubo conjugationis brevi inter cellulas sitis, subglobosis, subellipsoideis, plerumque irregularibus (sepe constrictis); membrana zygosporarum cras- sissima, irregulariter lamellosa ; azygosporis globosis subglobo- sisque, zygosporis similibus sed membrana tenuiore. Crass. cell. veget. 16-23 » (usque ad 26 » et plerumque 20 p); crass. membr. cellularum 1°5-5°5 x; long. zygosp. 25°5-33 p (usque ad 40); lat. zygosp. 19-26 » (usque ad 30,); crass. membr. zygosp. 2°5-6°5 p; diam. azygosp. 23-26 »; crass. membr. azygosp. 1°9-3°8 p. Hab. In mud, warm stream, crater of Grande Soufriére, Dominica. No. 883; also in no. 908, and sparingly from no. 882.—This also occurred, without zygospores, on old wall, Roseau, Dominica. (1-8-92.) No. 473. This species has a thick membrane which at first sight reminds one of a Lehizoclonium or a large species of Conferva, the some- what irregular filaments with short branches resembling the former especially. The many examples of conjugation examined were all scalariform, the zygospores completely filling the short conjugating-tube, their variability being remarkable, Spores were seen (figs. 9-10) which were undoubtedly azygospores, and these were regular in shape. One example was noticed where the zygospore was double. The material was preserved in weak spirit; but the two chromatophores were in most examples distinct and in their normal position. Var. CONFERVOIDES, nov. var. (Pl. XIV. figs. 1-6.) Var. cellulis tenuior, diametro plerumque 1-14plo (rarius 23 plo) longioribus. g Crass. cell. veget. 10-13 p. Hab. In mud, warm stream, crater of Grande Soufri¢re, Do- minica. No. 883. Intermingled with the typical form, but not so abundant. ALGEZE FROM THE WEST INDIES. 267 This variety has the cells of the filaments often somewhat doliform, and closely resembles a Conferva; it does not, how- ever, break up in the peculiar manner of any species of Conferva (vide fig. 6). The binate chromatophores also put it out of this genus. Fig. 4 probably represents an attempt at con- jugation. Two cells are figured (fig. 5) which show an attempt at longitudinal division. Ord. DEsMIDIACESR. 6. Mrsor#nium micrococcum, Airchn. (Alg. Schles. p. 134.) [Palmoglwa micrococea, Auetz. Tabule phycolog. i. p. 20, t. 25. f. 5.] Long. cell. 15°5-17°3 yw; lat. cell. 7°5-9 u. Hab. On trees among mosses, summit of Trois Pitons (4500 ft.), Dominica. Nos. 903 and 904.—On Bow-wood Hills (1580 {t.), St. Vincent. (8-12-91.) No. 23. 7. M. cutamyposporuM, De Bary. (Con). p.75; Cooke, Brit. Desm. p. 47, pl. 18. f. 14; West, Freshw. Alg. of W. Ireland in Journ. Linn. Soc. vol. xxix. p. 131, pl. xxiv. f. 8.) Long. cell. 19-23 p; lat. cell. 12°5-14°5 p. Hab. On trees among mosses, summit of Trois Pitons (4500 ft.), Dominica. Nos. 908 and 904. 8. TeTMeMorts La&vIs, Ralfs. (Brit. Desm. p. 147, t. 24. f. 2.) Long. 57°5-65 p; lat. 19-20 p; lat. isthm. 17°5-18°5 p. Hab. In mud, warm stream, crater of Grande Soufriere, Do- minica. No. 883. 9. CosMARIUM PSEUDOPYRAMIDATUM, Lund. (Desm. Suec. p. 41, t. 2. £ 18.) Long. 46-48 p; lat. 24-25 »; lat. isthm. 9°5-12°5 p. Hab. Ono damp wall of dam, Sharp’s River, St. Vincent. No. 477. *sTENONOTUM, Wordst. in Wittr. et Nordst. Desm. et Edog. in Ital. et Tyrol, p. 32, t. xii.f.8. Forma minor, Racib. (Desmidya w podrozy na oklo ziemi, p. 4, t. i. f. 32). (Pl. XIV. fig. 25.) Long. 36-865 p; lat. 28-24; lat. isthm. 65 p53 crass. 13°5 p. Hab. With the typical form, but much more abundant. 10. C. optiguum, Nordst. (Bidrag till kdnned. om Sydligare t2 268 MR. W. WEST ON SOME FRESHWATER Norges Desm. p. 28, t. 1. f. 8). Forma minor, Nordst. (J. c.). (Pl). XIII. fig. 17.) Long. 14°5 p; lat. 11°53 lat. isthm. 6°5 x; crass. 9°5 pw. Hab. Amongst Symploca cuspidata, n. sp., on trees, summit of Trois Pitons (4500 ft.), Dominica. No. 904. This interesting species seems to have a varied kind of habitat, occurring in small upland tarns and pools, on dripping subalpine rocks, and in the present instance on damp mossy trees. 11. Cosmartum Cucursira, Bréb. (Ralfs, Brit. Desm. p. 108, t. 17.4.7.) Long. 40 p; lat. 21 pw; lat. isthm. 19 p. Hab. On trees among mosses, summit of Trois Pitons (4500 ft.), Dominica. No. 903. Class C@NOBIES. Ord. VoLVocINEz. 12. Euportna staGNnaLe, Wolle. (Freshw. Alg. of U.S. p. 160, pl. clii. figs. 11-21.) Var. cellulis parvis et distantibus. (PI. XVI. fig. 10.) Diam. cel]. 4°8-5°6 x; diam. ccenob. 63 pe Hab. Amongst mosses on trees, with Hapalosiphon intricatus, n. sp., and Symploca cuspidata, n. sp.. summit of Trois Pitous (4500 ft.), Dominica. No. 903. Class PROTOCOCCOIDES. Ord. Protococcaces (incl. Palmellacee). 13. Urococcvs 1nsienis, Kuetz. [Chlorococcus macrococcus, Rabh., et var. aureus, Rabh. Fl. Europ. Alg. ii. p. 33.) Diam. cell. 23-31 p; ¢. teg. 42 p. Hab. On mossy trees, summit of Trois Pitons (4500 ft.), Dominica. Nos. 903 and 904, 14, CERASTERIAS STAURASTROIDES, nov. sp. (PI. XIV. fig. 16.) C. quadriradiata e corpore distincto, radiis elongatis sensim attenuatis ef minute granulatis, apice obtusis. Diam. ¢. proc. 30-35 p; diam. corpor. circ. 9-9°5 p. Hab. With Scytonema javanicum, Bornet, amongst mosses, on lime-trees, Shantord Estate, Dominica. No. 901. The rough arms of this species remind one very forcibly of a small Staurastrum, and sufficiently characterize it. ALGE FROM THE WEST INDIES. 269 Class PHYCOCHROMACES. Subclass Nosrocuinex. Ord. Nosrocacez. 15. Nostoc numiFusumM, Carm, ex Harvey in Hooker's Brit. Flora, ii. p. 399. (Kuetz. Species Algar. p. 8301; Rabh. Fl. Hurop. Alg. ii. p. 183.) Diam. cell. 2-3 » ; diam. heterocyst. 3°5 pu. Hab. On lime-trees, Shanford Estate, Dominica. No. 901.— On trees, summit of Trois Pitons (4500 ft.), Dominica. No. 903. 16. N. spxaricum, Vaucher. (Cooke, Brit. Freshw. Alg. p- 231, t. 91. 8-11; Bornet et Flahault, Révis. des Nostoch. Heétérocyst. p. 208.) Diam. cell. 38; diam. heterocyst. 5-5:7 x; diam. thall. usque ad 10 mm. Hab. On damp wall of dam in Sharp’s River, St. Vincent. No. 477. Ord. ScyTONEMACES. 17. Microcu#te TENUISSIMa, nov. sp. (PI. XIV. figs. 7-11.) Mf. inter algas varias alias reperta; filis tenuissimis, sub- intricatis, contortis; vaginis hyalinis, achrois, amplis; articulis elongatis, diametro 5-9plo longioribus, articulis junioribus bre- Vioribus (circiter diametro duplo longioribus) et crassioribus ; heterocystis intercalaribus, subquadratis vel oblongis. Crass. fil. 4-4-5:1y; crass. trich. 1-1°8; crass. heterocyst. 2-2°4-4; long. heterocyst. 3°5-6°5 p. i Hab. Amongst Symploca cuspidata, n. sp., on trees, summit of Trois Pitons (4500 ft.), Dominica. No. 904. 18. Scyronema savantcum, Bornet (in Bornet et Thuret, Notes Algologiques, p. 148; Bornet et Flah. Révis. des Nostoch. Heétérocyst. p. 95). [Symphyosiphon javanicus, Awe¢z. Species Algarum, p.323 ; Tabule phycolog. ii. p. 13, t. 43. fi.) (PL XTY. figs. 12-15.) Crass. fil. 13-16 x; crass. trich. 8°5-11 p. Hab. On lime-trees, Shantord Estate, Dominica, No. 901.— Anguilla, W. Indies. (23-3-92.) No. 70.—Ou the walls, Rosean, Dominica. (9-7-92.) No. 236. . The above agrees well with this species, the most notable dif- 270 MR. W. WEST ON SOME FRESHWATER ference being in the branches not being aggregate. The hetero- cysts vary from subquadrate to subrotund, and are rather numerous. 19. ScyToNEMA AMPLUM, nov. sp. (Pl. XVI. figs. 14-16.) 8. strato parvo, pannoso, 8-5 mm. lato, fusco ; filis dense in- tricatis ; pseudoramis sparsis plerumque geminatis sed interdum singulis, filo primario tenuioribus; vaginis amplissimis stratis parallelis formatis, in parte exteriore gelatinoso-achrois vel sub- luteolis, in parte interiore abrupte luteo-fuscis ; trichomatibus angustis, luteo-viridibus, ad apicem pseudo-ramulorum crassior- ibus et articulis brevioribus; articulis diametro 33—-Gplo (ple- rumque 4plo) longioribus; heterocystis oblongis, diametro 3-33 plo (rarius 2plo) longioribus. Crass. fil. prim. 19-24 »; crass. ramul. 13°5-16 ; crass. trich. 35-4 p. Hab. On trees, summit of Trois Pitons (4500 ft.), Dominica. Nos. 903 and 904.—Growing about and upon the surface of Symploca cuspidata, n. sp. The nearest species to this with regard to the comparative length and breadth of the cells is S. ambiguum, Kuetz., from which it differs in its much larger size, stouter habit, compa- ratively broader sheath, and in the more unfrequent branches. The branches are usually geminate though not unfrequently single, and they are always thinner than the primary filament. The trichomes at the apices of the branches become almost twice as thick, the cells becoming very much shorter. SS. myo- chrous, Ag., is somewhat similar though larger, and has not such a comparatively broad sheath, and the latter is ocreate. 20. S. amBreutm, Kuetz. (Species Algarum, p. 894; Tabule phycolog. ii. p. 7, t. 26. f. ii.) (PL. XV. figs. 11-15.) Crass. fil. 95-115 w; crass. trich. 2-2°5 p. Hab. On trees, summit of Trois Pitons (4500 ft.), Dominica. No. 903.—On the ground, mostly in old Diablotia holes, Morne Anglais (2300 ft.) (15-7-92.) No. 493. Amongst Symploca. The figure given by Kuetzing (J. ¢.) is very indistinct ; but the specimens agreed well with the description given by Bornet and Flahault (Révis. des Nostoc. Hétérocyst. p. 100), excepting that the filaments were a little thicker (9°5-11°5 p against 6-9 ,). The trichomes were, however, the same. One of its chief characters is its long and narrow cells, which get shorter aud thicker (up to 3'8,) towards the apices of ‘the young ALG FROM THE WEST INDIES. 271 branches; the sheath also becomes hyaline. The cells often appeared like the section of a biconcave lens, owing to the con- traction produced by the dilute spirit in which the material was preserved. ‘ 21. Scyronema FieuraTuM, Agardh. (Syst. Algar. p. 38.) [Secytonema calotrichoides, Kuetz. Species Algarum, p. 307; Tabule phycolog. ii. p. 6, t. 22. fig. ii.; Rabh. Fl. Europ. Alg. ii. p. 253. ] Crass. fil. 15°5-23 p; crass. trich. 18-13°5 p. Hab. On damp wall of dam, Sharp’s River, St. Vincent. No. 477.—On trees, summit of Trois Pitons (4500 feet), Dominica. Nos. 903 and 904. 22. ScyTONEMA, sp. Crass. fil. 28-36 »; crass. trich. 5-5'5 p. Hab. Amongst §. javanicum, Bornet, on lime-trees, Shanford Estate, Dominica. No. 901. Only asmall quantity of this was seen, which was insufficient for accurate determination. The characters were those of S. densum, Bornet, and it probably is a small variety of this species. 23. ToLypoTHRix TENUIS, Kuetz. (Phycolog. gener. p. 228; Tabule phycolog. ii. p. 9, t. 81. f. ii.) [T. pygmea, Awetz. T. flaccida, De Bary. ] Crass. fil. 7-7°7 4; crass. trich. 5°7 «; heterocyst. 7°7 x 5°7 p. Hab. On damp wall of dam, Sharp’s River, St. Vincent. No. 477. Ord. SrROSIPHONIACER. 24. HapaLosIPHON INTRICATUS, nov. sp. (PI. XV. figs. 16-28.) H. cxspitibus parvis, eruginosis ; filis densissime intricatis et variabilibus, adultis vaginis arctis distinctis (interdum paullo indistinctis) e cellulis singulis formatis, sparsim ramosis ; ramis singulis unilateralibus flexuosis, filo primario subsimilibus, vagi- natis vel evaginatis ; cellulis variabilibus, diametro 12-3plo lon- gioribus, sepe equalibus et subrotundis, interdum elongatis : heterocystis intercalaribus, subquadratis vel oblongis (diametro 1~3plo longioribus). Crass. fil. 4-7 » ; crass. heterocyst. 3°8-9'5 p. Hab. In little intricate tufts among the leaves of Leucobryum, on trees, summit of Trois Pitons (4500 ft.), Dominica. No. 903. 272 MR. W. WEST ON SOME FRESHWATER The nearest species to this hitherto described is H. laminosus, Hansg. [‘ Ueber den Polymorph. der Alg.,” Botan. Centralb. 1885, p. 48 (cfr. Bornet et Flah. Révis. des Nostoe. Heétéro- cyst. p. 55)], from which it differs, however, in not being calcified in any way, in being a little larger, in having its single and unfrequent branches of a similar thickness to the primary filaments without any attenuation, and in its peculiar habitat. The heterocysts are of the same breadth as that of the cells (or narrower), whilst those of H. lJaminosus are often broader. Like the latter, the threads are very variable, being sometimes similar to an Anabena, sometimes like a Lyngbya, while at other times the cells are quite irregular and somewhat inflated. Some examples showed the rounded granulose bodies noted by Hansgirg in H. laminosus (Bemerkungen zur Svstematik einiger Stss- wasseralgen, p. 18). These bodies when fully formed appear to be thick-walled (figs. 23-28) and have a diameter of 6-13 p. The cell of the primary filament immediately under a branch always projects more or less into the sheath of the branch. 25. HaPaLostPHON FLEXUOsUS, Borzi. (Alghe d’acq. dole. d. Papua, in Nuovo Notarisia, Apr. 1892, p. 43.) Crass. fil. 7°5-8°5 w; crass. cell. 5-5-6 u; long. cell. 4-4°5 p. Hab. In stream, Grande Soufriére, Dominica. No. 884. 26. H. arporetvs, nov. sp. (Pl. XV. figs. 1-3.) H. inter alias algas repertus ; filis primariis flexuosis, passim ramosis uno latere e cellulis singulis formatis, vellulis diametro 1-}3plo (rarius 2plo) longioribus, vagina arcta, tenui, et acvhroa ; filis secundariis brevibus, crassitudine filis primariis similibus sed subtenuioribus, cellulis diametro subequalibus ; heterocystis quadrato-oblongis, intercalaribus. Crass. fil. 7-103 crass. cell. 7-9°5 pe; heterocyst. 6-9 p x 9-11 p. Hab. On trees, summit of Trois Pitons (4500 ft.), Dominica. No. 903. The nearest species to this is H. fleruosus, Borzi (J. ¢.), from which it differs in not possessing flexuose intricate branches on every side; it is also a rather larger species, with cells often a little longer than broad and never depressed. 27. STIGONEMA HORMOIDES, Bornet et Flah. (Révis. des Nostoc. HHétérocyst. p. 69.) [Seytonema hormoides, Auetz. Sirosiphon brevis, Awelz. Botan. Zeit. 1847, p- 196; Tabule phycolog. ii. ALG FROM THE WEST INDIES. 273 p- 10, t. 34. fii. Sirosiphon hormoides, Awetz. Species Algar. p- 316; Zabule phycolog. ii. p. 10, t. 84. f. iv.] Crass. fil. 9-13 p. Hab. On trees, summit of Trois Pitons (4500 ft.), Dominica. Nos. 903 and 904. Rather scarce-—On rocks, Roseau Valley (1000-2000 ft.), Dominica. (27-6-92.) No. 24. Var. TENUE, nov. var. (Pl. XV. figs. 4-8.) Var. minor, filis tenuioribus, 5'd-7 pe crassis. Hab. With the typical form, but much more abundant. It was intermixed with Symploca cuspidata, n. sp., Stigonema minuta, Hass., Scytonema ambigua, Kuetz., &e. 28. STIGONEMA PANNIFORME, Bornet et Flah. (Révis. des Nostoc. Hétérocyst. p. 71.) {Seytonema panniformis, Agardh ; Sirosiphon panniformis, Awetz.] Crass. fil. 23-26 p. Hab. With the preceding. Another gathering from the same locality had abundant hyphe (an incipient lichen). No. 779. Crass. fil. 15-25 p. 29. S. minurum, Hass. (Hist. of Brit. Freshw. Alg. i. p. 230, t. 67. ff. 3-4.) Crass. fil. 19-25 p. Hab. On damp wall of dam, Sharp’s River, St. Vincent. No. 477.—On trees, summit of Trois Pitons (4500 ft.), Do- minica, no. 903; and on lime-trees, Shanford Estate, Dominica, no. 901. 30. S. rnrormE, Kuetz. (Species Algar. p. 319; Tabule phy- colog. ii. p. 11, t. 38. f. iii.) Crass. fil. 44-52 p. Hab. On trees, summit of Trois Pitons (4500 ft.), Dominica. Nos. 903 and 904. Ord. OsScILLARIACEE. 31. Symproca cusripata, nov. sp. (PI. XVI. figs. 1-7.) S. bryophila, late expansa, eriseo-lutea ; fasciculis erectis, angustis subulatis, aggregatis (passim densis), 8-15 mm. altis, eruginescentibus ; trichomatibus wruginels, flexuoso-intricatis, in strato strictioribus, apicibus versus fasciculorum, 1-3 in lata vagina inclusis, sepe interruptis, distincte articulatis ; articulis diametro 2~4plo longioribus; vaginis amplis, acbrois, pellucidis 274 MR. W. WEST ON SOME FRESHWATER yel interdum stratis parallelis formatis, in ambitu sepe sub- rugosis, apice angustioribus et sepe ramosis. Diam. trich. (s. vag.) 1°9-2°3 w; crass. trich. c. vag. 13°5-25 p. Hab. On trees, summit of Trois Pitons (4500 ft.), Dominica. Nos. 903 and 904. Var. LUTEOFUSCA, DOV. var. Var. ferruginea, strato denso, 1-2 mm. alto, fasciculis brevior- ibus, 4-6 mm. (interdum 10 mm.) altis, articulis crassioribus. Crass. trich. 2°5-3°5 ; crass. vag. 15-40 p. This variety also often has two, three, ur more threads in one sheath. Hab. On rocks, Roseau Valley (1000-2000 ft.), Dominica. (27-6-92.) No. 24.—On bark, windward road to lake (1000- 2000 ft.), Dominica. (25-8-92.) Nos. 513 and 514.—Also on the ground, mostly in old Diablotia holes, Morne Anglais (2300 ft.). (15-7-92.) No. 498. Thicker and coarser sheath. The specimens preserved in liquid strongly reminded one of Sphagnum cuspidatum; they occurred amongst mosses in penicillate tufts which were suberuginose towards the apices, the sheaths, but uot the trichomes, here becoming narrower. Many of the sheaths in the stratum were without trichomes ; and in consequence of this, and the broader sheaths at the base, the stratum was paler than the erect fasciculi. 32. Lynapya PENICILLATA, Kuetz. (Botan. Zeit. 1847, p. 194.) [Leibleinia penicillata, Kuetz. Species Algar.; Tabule phycolog. i. p- 46, t. 81. f. ii.] Crass. fil. 3-3°5 p. Hab. On bed of stream in crater of Grande Soufri¢re, Dominica. No. 881. 33. L. suprie, West. (Aly. of the Eng. Lake Distr. in Journ. Roy. Micr. Soc. Dec. 1892, p. 741, pl. x. f. 58.) Crass, fil. 1°4 pe. Hab. With the preceding. 34. PHoRMIDIUM LYNG@BYACEUM, Kuetz. (Phycol. gener. p- 194; Tabule phycolog. p. 33, t. 46. f. iii; Rabh. Fl. Europ. Alg. ii. p. 124.) Crass. fil. 3°5-4 p. Hab. On damp wall of dam, Sharp’s River, St. Vincent. No. 477.—On old wall, Dominica. (1-8-92.) No. 473. The cells were about as long as broad (sometimes a little longer). A ALGZ FROM THE WEST INDIES. 27 35. PHorMIpIvM, sp. Crass. fil. 1-2-1°5 p. Stratum dense, thin, and dark eruginous. Hab. On damp wall, Roseau, Dominica. (5-9-92.) No. 546. Ord. CuHroococcaces. 36. Curoococcus minor, Naeg. (Gattung. einzell. alg. p. 47, t. la. f. 4. Rabh. Fl. Europ. Alg. ii. p. 30.) [Protococcus minor, Kuetz. Species Algar. p. 198; Tabule phycolog. i. p. 3, t.3.] (Pl. XVI. fig. 17.) Diam. cell. 28-4; diam. fam. 11°5-30 p. Hab. On damp wall of dam, Sharp’s River, St. Vincent. No. 477.—On trees, summit of Trois Pitons (4500 ft.), Dominica. No. S03. Forma minima. (Pl. XVI. fig. 18.) Forma cellulis familiisque minoribus quam forma typica. Diam. cell. s. teg. 1-192; diam. fam. 10-28 p. Hab. Along with the type from the above-named localities, and also on lime-trees, Shanford Estate, Dominica. No. 901. 37. C. conzrens, Naeg. (Rabh. Fl. Europ. Alg. ii. p. 30.) [Pleurococcus coherens, Bréb., 1842. Protococcus coherens, Kuetz.] Diam. cell. 3°8-6°5 4 (usque ad 7°5 x); diam. fam. 30-173 p. Hab. Amongst other alge on trees, summit of Trois Pitons (4500 ft.), Dominica. No. 903. . This usually occurred in small families, but occasionally in very large ones, and now and then in solitary examples. 38. C. rurarpus, Naeg. (Gatt. einzell. Aly. p. 46. Rabh. Fl. Europ. Alg. ii. p. 32.) [Protococcus turgidus, Kuetz. Tabule phycolog.i. p. 5, t. 5. £. i.] Diam. cell. s. teg. 7°5-10°5 p, ¢. teg. 21-23 ie . Hab. On damp wall of dam, Sharp’s River, St. Vincent. No. 477. A rather small form. 39. C. scnizopermaticus, West. (Alg. of the Eng. Lake Distr. in Journ. Roy. Micr. Soc. Dec. 1892, p. 742, pl. x. ft. 61-63.) (Pl. XVI. fig. 19.) Diam. cell. s. teg. 115-13 »; diam. ec. teg. 23-28 p. Hab. With the preceding species, but much more abundant. 276 MR. W. WEST ON SOME FRESHWATER 40. Gimocapsa MuRALIS, Kuetz. (Tabula phycolog. i. t. 21. f.i.; Rabh. Fl. Europ. Al. ii. p. 36.) Long. cell. s. teg. 5°5-7°5 p; lat. cell. s. teg. 25-3; diam. fam. 19-25 p. Hab. On damp wall of dam, Sharp’s River, St. Vincent. No. 477. The cells were longer in relation to their breadth than those figured by Kuetzing (J. ¢.). 41. G. aieas, nov. sp. (Pl. XVI. figs. 11-13.) G. familiis solitariis vel subaggregatis ; cellulis magnis, subglo- bosis vel oblongis, 4-36 in familiis subglobosis consociatis, mem- brana cellularum leva vel subtiliter granulata (vide fig. 11), cytioplasmate exrugineo et granuloso; tegumentis subglobosis, ad exteriorem duris sepe subrugosis, luteo-fuscescentis, lamellis extra cellulas indistinctis, paucis et pallide luteolis. Diam. cell. s. teg. 9-15 (plerumque 11 ),); diam. fam. 44-115 p. Hab. On damp wall of dam, Sharp’s River, St. Vincent. No. 477. This occurred amongst Nostoc sphericum, Vauch., in no definite stratum ; and it is sufficiently distinguished by its large cells and other characters. 42. GLa@ocaPsa, sp. Crass. cell. s. teg. 2-3 p. Nos. 23 and 24. 43. GL@OCAPSA, sp.., Crass. cell. s. teg. 15 u. No. 546. 44, APHANOCAPSA ELACHISTA, nov. sp. (Pl. XV. figs. 9, 10.) A. tegumento minutissimo, subgloboso, firmo, gelatinoso non lamelloso, achroo ; cellulis minutissimis, spheericis, solitariis gemi- natisve, laxe dispositis ; cytioplasmate homogeneo et xruginoso ; tegumentis non aggregatis in thallo distincto. Diam. cell. 1°5-1'8 » (usque ad 22); diam. teg. 26-38 p. Hab. On trees, summit of Trois Pitons (4500 ft.), Dominica. No. 903.—And in stream, Grande Soufriére, Dominica. No. 88+. This species seems characteristically distinct by reason of its minute cells in the very small globose colonies, which were scattered amongst other alge. 45, GL@OTHECE LINEARIS, Naeg. (Gattung. einzell. Aly. p.58, t.1@.f.2; Rabh. Fl. Europ. Algq. ii. p. 60.) ALGZ FROM THE WEST INDIES. 277 Long. cell. sine teg. 5°5-6°5 wp, c. teg. 13°51; lat. cell. sine teg. 1'8 p, ce. teg. 9°5-10°5 pu. Hab. On damp wall of dam, Sharp’s River, St. Vincent. No. 477. Forma cellulis 5-6plo longius quam latius. (Pl. XIV. fig. 26.) Long. cell. s. teg. 11°5-13°5 p, c. teg. 21; lat. cell. s. teg. 2-2°3 u, c. teg. 11°5-12°5 p. Hab. With the typical form. 46. GL@OTHECE LUNATUM, nov. sp. (PI. XVI. fig. 9.) G. cellulis 2-4 in familiis consociatis, subcrescentiformibus, 2plo longius quam latius, apicibus acutis; tegumentis univer- salibus, ovalibus vel ellipticis; cytioplasmate wrugineo et ho- mogeueo. Lat. cell. 2°5-2°'7 »; apicibus cellularum 4°8-5°7 w inter se distantibus ; famil. 325 x19 p. Hab. With Gl. linearis, Naeg., but very scarce. 47. APHANOTHECE sAxicoLa, Naeg. (L.c. p. 59, t. 1 u. f. 2.) Lat. cell. 1:4-1°8 p. Hab. In small masses of 70-120 u in diameter, amongst mosses on trees, summit of Trois Pitons (4500 ft.), Dominica. Nos. 903 and 904. Rather scarce. 48. A. mrcroscoprca, Naeg. (L.c.t. 14. f. 1.) Long. cell. 5°5-7'5 ; lat. cell, 3°8-4'4 4; diam. famil. 108—- 179 p. Hab. On damp wall of dam, Sharp’s River, St. Vineent. No. 477.—Amongst Symploca cuspidata, n. sp., on trees, summit of Trois Pitons (4500 ft.), Dominica. No. 904. 49, TerRAPEDIA TRIGONA, nov. sp. (PI. XVI. fig. 8.) T. cellule triangulares, lateribus concavis, angulis subrotun- datis ; a latere vise elliptice ; cytioplasmate pallide erugineo et homogeneo. Lat. cell. 7°2 w; crass. 3°6 p. 7 Hab. On damp wall of dam, Sharp’s River, St. Vincent. No. 477. Class DIATOMACEF. 50. Eprrnemra WESTERMANN, Auetz. (Pritch. Infus. ed. 1861, p. 760, t. 4. £. 2.) MeN gg Hab. In stream, crater of Grande Soufri¢re, Dominica. Nos. 881 and 908. 278 MR. W. WEST ON SOME FRESHWATER 51. Evnorta Arcus, Ehrenb. (W. Sm. Brit. Diat. ii. t. 33. f, 283.) Hab. Instream, crater of Grande Soufriére, Dominica. No. 908. 52. E. aracizts, Rabh. (Fl. Europ. Alg.i. p. 72.) [Himan- tidium gracile, Ehrnb., W. Sm. Brit. Diat. ii. p. 14, t. 33. f. 285.] Hab. Amongst mosses, on trees, summit of Trois Pitons (4500 ft.), Dominica. No. 903. 58. Oponripium MUTABILE, W. Sm. (Brit. Diat. ii. p. 17, t. 34. f. 290.) Hab. On damp wall of dam, Sharp’s River, St. Vincent. No. 477. 54. DesMocontuM RaBEeNHorsTIANUM, Grun. (Diat. ins. Banka, p. 6, t. 1. f. 1; De Tont, Sylloge Algar. vol. ii. p- 680.) Hab. Abundant in stream (hot and cold), crater of Grande Soufriére, Dominica. No. 908. 55. Nirzscuta parvuLa, W. Sm. (Brit. Diat.i. p. 41, t. 13. f. 106.) Hab. On damp wall of dam, Sharp’s River, St. Vincent. No. 477. 56. N. ryexenxa, Suring. (Alg. Japon. p. 11, t. 1. f. 12.) Long. 33-44 p ; lat. 3:8-4°8 p. Hab. In streams, Grande Soufriére, Dominica. Nos. 881 and 884. 57. N. uivearis, W. Sm. (Brit. Diat. i. p. 39, t. 13. f. 110.) Hab. On damp wall of dam, Sharp’s River, St. Vincent. No. 477.—And in streams, Grande Soufriére, Dominica. Nos. 881 and 908, 58. N. minurisstma, W. Sm. (L. c. p- 41, t. 18. f. 107.) Hab. In stream, Grande Soufri¢re, Dominica. No. 884. 59. NavicuLa cryprocepHaLa, Kuetz. (W. Sm. I. c. p. 53, t. 17, f. 155.) Hab. On damp wall of dam, Sharp’s River, St. Vincent. No. 477. West. pepe TES \ 297 TNT We F ATER ALGAE O DYVEN ALGE FROM THE WEST INDIES. 279 60. Navicuna BoreaLIs, Kuetz. [Pinnularia borealis, Ehrenb., W. Sm. Brit. Diat. ii. p. 94; Rabh. Fl. Europ. Alg. i. p. 216.] Hab. With the preceding species. 61. N. mesorerta, Ehrenb. [Pinnularia mesolepta, W. Sm. Brit. Diat. i. p. 58, t. 19. f. 182.] Hab. On bed of stream in crater of Grande Soufriére, Dominica. No. 881. 62. Frusrutia rtomBorpEs, De Toni. (Sylloge Algar. vol. ii. p. 227.) [Navicula rhomboides, Hhrenb., W. Sm. Brit. Diat. i. p. 46, t. 16. £. 129.) Hab. In stream, crater of Grande Soufri¢re. No. 908. Var. saxontca, De Toni. (L.c.) [Frustulia saxonica, Rabh. Navicula crassinervia, Bréb. in W. Sm. Brit. Diat. i. p. 47, t. 31. f. 271.) Hab. In streams (cold, warm, and hot), crater of Grande Soufriére, Dominica. Nos. 883 and 908.—Also amongst mosses on trees, summit of Trois Pitons (4500 ft.), Dominica. Nos. 903 and 904. 63. GoMPHONEMA TENELLUM, Kuetz. (W. Sm. Brit. Diat. i. p- 80, t. 29. £. 243.) Hab. On damp wall of dam, Sbarp’s River, St. Vincent. No. 477. EXPLANATION OF THE PLATES. Puare XIII. Figs. 1-4. Zygnema (§ Zygogonium) pachydermum, n. sp. Conjugated spe- cimens, 520/1. 5-8. Ditto. Four zygospores, 520/1. 9,10. Ditto. ‘Two specimens, with azygospores, 520/1. 11-15. Ditto, sterile filaments; fig. 11, 520/1; figs. 12-15, filaments, showing short branches, 120/1. Fig. 16. Ditto, showing conjugation between three filaments, 120/1. Fig. 17. Cosmarium obliquum, Nordst., f. minor, Nordst., 520/1. Puate XIV. Figs. 1-6, Zygnema pachydermum, Nn. sp., Var. confervoides, 0. var. Figs. 1 & 2, filaments without cell-contents; fig. 3, filament with contents delineated ; fig. 4, attempt at conjugation ? ; fig. 5, cells showing longitudinal division. Figs. 1-5, 520/1; fig. 6, 830/1. 7-11. Microchete tenuissima, n. sp. 520/l. 12-15. Scytonema javanicum, Bornet. 520/1. 280 Fig. 16. Figs, 17-24. Fig. 25. Fig. 26. FRESHWATER ALGHX FROM THE WEST INDIES. Prate XIV. (continued). Cerasterias stawrastroides, n. sp. 520/1. Rhizoclonium Berggrenianum, Hauck, var. dominicense, n. var. 5207/1. Cosmarium pseudopyramidatum, Lund, * stenonotum, Nordst., f. minor, Racib. 520/1. Gleothece linearis, Naeg., forma. 520/1. PuatEe XV. Figs. 1-3. Hapalosiphon arboreus, nu. sp. 520/1. 4-8. Stigonema hormoides, Bornet et Flah., var. tenve, n. var. 520/1. h, heterocysts. 9-10. Aphanocapsa elachista, n. sp. 520/1. 11-15, Scytonema ambiguum, Kuetz. 520/1. 16-28. Hapalosiphon intricatus, n. sp. 520/1. Figs. 1-7. Fig. 8 9 10. Figs. 11-13. 14-16. Fig. 17. 18. 19. Prats XVI. Symploca cuspidata,n.sp. Fig.1, natural size ; fig. 2, a very small portion of one of the erect fasciculi, 120/1; fig. 3, a specimen showing three trichomes in one sheath, 120/1 ; figs. 4 & 5, exampies with two trichomes in one sheath, 520/1 ; figs. 6 & 7, the apices of two filaments, showing in fig. 6a broad sheath at the apex, and in fig. 7 a narrow one, 520/1. . Tetrapedia trigona, n, sp. 8380/1. . Gleothece lunatum, n. sp. 520/1. Eudorina stagnale, Wolle, var. 520/1. Gleocapsa gigas, n. sp. 520,1. Fig. 11, an example having the cells with finely granulate walls; fig. 12, one where the lamellx round the cells are not visible. Scytonema amplum, n. sp. 520/1. Chroococeus minor, Naeg. 520/1. » » » f. minima, 520/1. ” schizodermaticus, West. 520/1. pot We 74 \\ Ha é Ee \ t i) oO a Lr f nN 366 & 59000009 B08 SS = Vie 4 DG JOUARN pe EE Oe a Deere > so 2? ETE pera as ae eae cee ease: ree | Ress coon) Soom teres fe Pc erst ee SE 5 thes SEL Masgietis’s 3 46 ee AE MR. C. T. DRUERY ON APOSPORY. 281 Notes upon Apospory in a Form of Scolopendrium vulgare, var. crispum, and anew Aposporous Athyrium ; also an additional phase of Aposporous Development in Lastrea pseudo-mas, var. cristata. By Cuartrs T. Drurry, F.L.S. [Read 7th December, 1893.] (Pare XVIL) Tux case of apospory which I have the honour of bringing before the Society this evening is an entirely new one, in so far as it is exhibited by a fresh and apparently most unlikely species, viz. Scolopendrium vulgare, the normal smooth-edged strap-shaped fronds of which seem to offer no outlet for apical apospory, displaying, as they do, not the slightest tendency towards the formation of the slender lateral projections with which the phenomenon is usually associated in its apical form. It was on a special visit to Mr. T. Bolton of Warton, near Carnforth, in August last to inspect a new aposporous Athyrium, to which I will later refer, that I came across the case now under notice in the shape of a variety of Scolopendriwm vulgare, known as Scolopendrium var. crispum Drummonde, found many years ago by Miss Drummond near Falmouth. This is a very singular compound variety in which the fronds are very long and narrow, finely frilled, and bearing, in addition, broad flat digitate crests often 9 inches across. Another peculiar feature is that the midrib is deeply undulated several times perpendicularly to the plane of the frond, and finally the margins of the frills are deeply cut into long fimbriate projections. Mr. Bolton kindly sent mea plant, and on its receipt I at once perceived that in many cases the apices of these projections were bifid and translucent, presenting all the features of inci- pient prothalli. I consequently cut off a number of these with a small portion of attached frond, and inserted them in sterilized soil so that the terminal points or bifurcations were in contact therewith. In a few days evident signs of growth were visible, and in a week or two distinct prothalli of more or less cordate form and of normal size were developed, twin prothalli resulting in some cases, each limb of the bifurcations having developed independently. Root-hairs, however, were very tardy in ap- pearing, due probably to the fact that growth was sustained by absorption through the adherent portions of frond; later on, however, they appeared somewhat abundantly, though hardly in LINN. JOURN.—BOTANY, VOL. XXX. x 282 MR. C. T. DRUERY ON APOSPORY. normal quantity. In some cases thick fleshy prothalli of irre- gular form were developed from quite blunt projections which, when laid down, showed no signs whatever of prothallic growth. In October I sent several specimens of these prothalli to Prof. F. O. Bower at Glasgow ; and upon one of these he found numerous archegonia, both old and new, but no antheridia. The fact of these being true prothaili was, however, thus fully established. Later, in November, I found both antheridia and archegonia in abundance on one of the largest prothalli in my culture, an abnormal feature of which was the production ot these organs in quantity on both upper and under surfaces in conjunction with root-hairs, which, however, predominated on the lower surface, as might have been expected. It remains, of course, an open question whether plants will be produced; but, from general robustness of growth, I have little doubt on this point. I exhibit the plant under notice, upon which the fimbriate projections will be clearly seen, though, unless laid down as de- scribed, they have failed to develop more than quite incipient prothalli even inaclose warm frame. I also exhibit the cultures, two in number, showing developed prothalli, and accompany my notes with sketch of a portion of frond, natural size, and enlarged drawings of material after culture. This constitutes the fourth British species in which apospory has been shown to occur, viz. Athyrium Filix-feemina var. claris- sima, Polystichum angulare var. pulcherrimum (several forms), Lastrea pseudo-mas var. cristata, and Scolopendrum vulgare var. crispum Drummonde. The second exhibit consists of cultures of a new aposporous find of Athyrium Filix-femina by Mr. T. Bolton, who showed some pinne at the meeting of the British Pteridological Society at Lancaster in August last. This had been found in the autumn of 1892 in that district ; and on examining the portions of frond submitted, I was at once struck with its strong resemblance to that form of Athyriwm Filix-feemina upon which apospory was first discovered, and was found in North Devon. The peculiar slenderness and attenuation of all subdivisions and the general make of the two ferns were almost identical, though the second find is easily recognizable by most of the terminals being spiral. When the pinne were handed to me, they were accom- panied by the remark that “the spores never seemed to ripen ;” MR. C. T. DRUERY ON APOSPORY. 283 and upon turning them over, the reason was obvious, since I immediately recognized the peculiarly woolly appearance of the sori indicative of apospory, the masses of incipient prothalli being in this case so large as to give in some places a suprasoriferous appearance to the frond. As Mr. Bolton very kindly provided me with material, I laid down several pinne with the excre- scences next the soil, and under close culture actual development immediately began, so that in a week or two each sorus had pro- duced a crowd of full-sized prothalli. Reporting progress to Dr. F. W. Stansfield, who had also received material, he wrote that in his case apical apospory was also appearing, 2. €. prothalli were being developed from the apices independently of the sori. Examining my material, I found several apparent eases of this, which, however, on closer investigation, I found to be produced from small aborted sori situated so near the terminal point that the resulting prothalli seemed to spring therefrom. I therefore asked Dr. Stansfield to re-examine his material, with the result that in the majority of cases he found the appearance had been deceptive, but in some the apices themselves were certainly dilating. A little later I found an unmistakable instance of an apical prothallus, as can be seen by the culture I exhibit. Otherwise this case is identical with that of Athyrium Filiz- femina, var. clarissima, Jones, except that the growths are much more redundant. I have named the fern therefore Athyrium Filix-femina, var. clarissima, Bolton. It will be observed that in this case, as in all previous ones, the abnormality occurs on wild finds, and has not been induced by culture ; and considering the ease with which the prothalli are produced when the pinn are brought into contact with the soil, and the immense advantage which they must derive from being firmly attached to the frond from the beginning instead of, as in the detached spore, being at the mercy of a thousand and one disturbing factors, it is curious that the specimens found are solitary crowns instead of clumps. This advantage, however, in the struggle for existence is far more than counterbalanced by the absence of scattering spores, which limits the variety strictly to its birthplace. Finally, I exhibit young plants raised from the aposporous pro- thalli of Zastrea pseudo-mas, var. cristata, which I brought to the notice of this Society in November 1892. These are interesting as exemplifying exactly intermediate stages between me oophore 284 MR. J. C. WILLIS’S CONTRIBUTIONS TO and sporophore, the primary fronds in some cases consisting of erect prothalli borne on stalks, while in the case of the parent the primary frond had been eaten off when the plant came under notice, owing to a prothallus forming at tbe tip of the second frond and others subsequently on its edges and surface. This exhibit was therefore necessary to complete the case by supplying the missing step, which it does, I think, very satisfactorily. EXPLANATION OF PLATE XVII. Scolopendrtum vulgare, var. Drwummonde. Fig. 1. Fimbriate projection from edge of frond (much enlarged). p. Pro- thallus developed by culture on moist soil. rh. Root-hairs. c. Thick fleshy cushion. Fig. 2. Portion of frond, natural size, showing fimbriate projections terminating in pp. incipient prothalli. Figs. 3, 4, 5,6. Prothalli (4) developed after severance of fimbriate projections from frond and insertion in soil. Contributions to the Natural History of the Flower.—Part II. Fertilization Methods of Various Flowers; Cleistogamy in Salvia Verbenaca. By J. C. Witzis, M.A., late Frank Smart Student in Botany of Gonville and Caius College, Cambridge. (Communicated by Franots Darwin, F.R.S., F.L.S.) [Read 15th February, 1894.] (Puates XVIII. & XIX.) THs paper comprises the results of observations upon various native and exotic plants. The plants whose methods of fertili- zation were studied belong to the genera Brodiea, Stanhopea, Pimelea, Cotyledon, Nemophila, Hydrolea, Ziziphora. A study of cleistogamy was made upon Salvia. The observations are mostly of similar character to those detailed in the first paper of this series. Since the publication of my former paper *, my attention has been called by Prof. Bessey to a paper by himt in which the movements of the stamens in Claytonia virginica are described : the description tallies with that of the author for C. sibirica, &c. Robertson ¢ has described the fertilization of Ellisia Nyc- telea, L., which seems on the whole to resemble that of the less * Journ. Linn. Soc., Bot. vol. xxx. pp. 51-63. t “Sensitive Stamens in Portulaca,” Amer. Naturalist, vii. 1873, p. 464. ¢ “Flowers and Insects, X.,” Bot. Gazette, xviii. 1893, p. 49. lann Soc Journ Kor Vor XXX. Pe 17 APOSPORY IN SCOLOPENDRIUN WolGaAre THE NATURAL HISTORY OF THE FLOWER. 285 conspicuous species of Phacelia. Hydrolea and Nemophila being considered in this paper, we now know the fertilization of five genera of Hydrophyllacew, though not fully. Lastly, the ferti- lization of Monarda jistulosa was described by Dr. Ida Keller * a short time before the author’s paper was read. Her description differs from the writer’s in that the anthers appear to dehisce in the bud; but this is rather what might be expected in plants examined in November, when the light is feeble and the tem- perature low. No trace of this early dehiscence occurred in the plants examined by the author in summer. FertruizaAtion Mrrnops. 1. Brodiea ixioides, 8S. Wats.—This, the only species of the Californian subgenus Calliprora, was studied in the Botanic Garden, Cambridge, during August 1892 and June-July 1893. It is scarcely visited by insects in England, and requires further study in its native place to be fully understood. The yellow flowers are borne in loose umbels, about 8-16 flowers being open at one time. The six perianth-segments are united below into a funnel-shaped tube about 6 mm. deep and 5 mm. wide at the mouth ; the total width of the limb is about 25 mm. (Pl. XVIII. fig. 1). Down the middle line of each segment on the outer side runs a dark brown line, which shows through the leaf and may act as a path-finder. The tip of the leaf is slightly swollen out vertically. Inserted on the perianth, at the mouth of the tube, are six stamens, three (inner) long, three short. The filament is broad and fleshy, ending in two horns, between which on the inner side is the introrse anther (fig. 2). The stamen 1s prolonged downwards into a fleshy ridge on the inner surface of the perianth-tube. The ovary is borne on a gynophore, and has a style with a capitate trilobed stigma (fig. 4); it bears three rows of hairs opposite the long stamens. The most interesting feature of the flower is the structure of the filaments of the stamens. These have a glossy appearance to the naked eye. Under the microscope it is seen that the horns and filament are covered with long, sometimes twisted, cells (fig. 8), which are very turgid, especially in warm weather, and rupture on being slightly touched with any hard body. A drop of fluid escapes on rupture: it has a pleasant slightly sweetish * “The Phenomena of Fertilization in the Flowers of Monarda fistulosa,” Proc. Acad. Nat. Sc. Philad., Dec. 27th, 1892, p. 452. See also the paper which precedes it, “‘ Notes on Monarda fistulosa,” by Meehan. 286 MR. J. C. WILLIS’S CONTRIBUTIONS TO taste, and is attractive to insects. The flowers swarm with Meligethes and Thrips, licking the stamens. Many Diptera also visit them for the same purpose. The ridge which continues the stamen inside the perianth-tube is also covered with cells like those described above. If astamen be scratched all over with a needle, so much fluid exudes that it trickles down into the base of the flower, and may accumulate there to some depth. These peculiar cells develop early, and are found in quite small buds. Shortly after the flower opens, the three short stamens dehisce and cover the stigma with pollen (fig. 1). Its papilla, however, are at this period very short, and apparently not receptive. Of flowers left to themselves, a large number do not set seed. After- wards the flower increases in size (cf. Pl. X VIII. figs. 1 and 4), but the ovary grows most rapidly till the stigma is level with the long stamens ; its papille are now very long and fully developed. The mouth of the flower is narrow, and insects probing the tube for honey must touch both anthers and stigma. Self-fertilization does not regularly occur; but may happen, especially when the flower finally shrivels up. The only insect visitors observed were those mentioned above, which did not usually effect fertilization ; and it is not altogether evident how the flower is regularly fer- tilized, as it is much easier for insects to get honey by biting the stamens than by probing the flower in the proper way. The above-described fertilization method does not agree very closely with that of any of the hitherto described Liliacew. Allium resembles it in the protandry, and Zloydia and Liliwm also approach it in some measure, especially the latter, whose nectary, inserted on the perianth, gives a possible clue to the peculiar mode of nectar-secretion, if such it can be called, in Brodiea. Stanhopea tigrina, Bateman.—The mode of pollination of 8. oculata was studied by Darwin*; but he failed to fertilize the flowers on account of the narrowness of the stigmatic chamber, which prevented the entrance of the pollinia. 8. tigrina flowered profusely in the Cambridge Botanic Garden during 1893, and Mr. Francis Darwin suggested to the author that it should be carefully studied. This has been done, and the insertion of the pollinia has been easily managed; but it was found impossible to completely understand the flower on account of the want of the insects that in its native place (Mexico) probably pollinate it. The hypothesis as to the mechanism of the flower advanced below * ‘Fertilization of Orchids,’ p. 171. THE NATURAL HISTORY OF THE FLOWER. 287 1s perhaps somewhat strange ; though Coryanthes, a near relative of Stanhopea, is equally extraordinary in its mechanism. It may be hoped tbat some botanist visiting Mexico will study the flower there and complete its elucidation. Stanhopea tigrina is an epiphytic orchid with enormous pen- dulous flowers *. In greenhouse specimens they emerge through the bottom of a hanging basket. If held out horizontally they are liable to break off by their own weight. The five perianth- leaves, other than the labellum, are reflexed from their point of insertion, and are very conspicuous, as seen from the side or from below. They are of a deep chocolate-red colour striped with pale lemon-yellow, and are of a slightly fleshy texture. The labellum and column hang downwards, making a curious cage (Pl. XVIII. fig. 5); they are very thick and fleshy, and are, in consequence, very rigid and elastic. As seen in side view the flower, taken as a whole, is about 16 cm. deep and 16 em. wide across the perianth, but often even larger than this. The “cage” narrows towards the mouth, where the anther and stigmas are (fig. 5). The basal portion of the labellum or hypochil (hyp, fig. 5) forms a curious pocket-like organ, widest at its lower end and with its mouth facing the column. Between it and the latter are wide openings through which an insect can fly into the mouth of the cavity. The mouth is bounded by sharply marked ridges below and at the sides. The inner surface of the cavity is covered with stout hairs of a deep-red colour. Below the bucket comes the middle portion of the labellum, which consists of a stout horizontal mass of tissue (mes, fig. 5), prolonged at the ends into two long bars (f,,) which run down- wards from it, at first outwards then inwards: these form the sides of the cage. The distal part of the labellum is fastened to the middle part by a well-marked joint or thin line ; but if it be bent at the joint, breakage occurs: the joint would appear thus to be merely a survival from jointed ancestors with more flexible labella. This part of the labellum is of a broad oval shape, with a deep groove in it near the apex ; like the rest of the labellum, it is thick and fleshy. The colour of the labellum is a pinkish white, with spots of a chocolate colour. The fourth side of the cage is formed by the column, a very rigid structure (see figs. 5-10). Its basal part 1s nearly cylin- drical ; further out it widens into lateral wings, narrowing towards * Figured in Bateman’s ‘Orchidacer of Mexico and Guatemala,’ pl. 7. Also in Bot. Mag. t. 4197, and Bot. Reg. 1839, t. 1. 288 MR. J. C. WILLIS’S CONTRIBUTIONS TO the apex, where are two horns projecting beyond on either side of the anther (fig. 7). The further end of the column has a sharp keel-like ridge on its anterior (inner) face, and the margins ot the lateral wings are also folded inwards: two gradually nar- rowing grooves are thus formed leading to the stigmatic cavity, whose entrance is a long horizontal slit about 1 mm. wide (fig. 7). The cavity turns downwards inside the column (fig. 10), making a sort of pit. From its base a strand of conducting parenchyma runs downwards through the column. Above the stigma is a sharp ridge (e, fig. 7) bearing the anther. There are two pollinia attached by short caudicles to a pedicel of moderate length (fig. 8), which in its turn is attached to the upper side of the saddle-like portion of therostellum. This is hollow, and contains a quantity of a viscid substance which does not fill the cavity. It is continued into a long narrow projection (fig. 7) which runs downwards parallel to, and just clear of, the edge of the central ridge of the column. Each pollinium has a central hollow opening to the outside at its further end (fig. 9). The labellum and column are, except at one or two points to be mentioned below, of a most extraordinary slipperiness. They feel to the touch as smooth as ice. With this introduction, we may go on to consider the actual mode of fertilization. Ifa sharply pointed pencil be slid side- ways down the column, it follows the crest of the central ridge, and the point goes under the projecting tongue of the rostellum, and then reaches the viscid matter in the saddle. Owing to the hollow shape of this latter, already described, only a smal] portion of the viscid matter touches the pencil ; and as one goes on with the movement, the adhesion is not sufficiently firm, and the rostellum often springs back into place. Occasionally, however, the adhesion is firm; and it can always be made so by pressing the saddle against the pencil. In this case the pollinia are re- moved, leaving the empty anther still in position. The whole, t. e. tongue, saddle, pedicel, and pollinia, stand in much the same straight line. No movement appears to occur after removal. The appearance of the summit of the column after removal of the pollinia and of the anther is seen in fig. 10. There remains a small narrow portion of rostellum (x in fig. 10) projecting down the column. If, now, we again slide the pollinia upon the pencil along the column, they slide with the utmost smoothness along its slippery surface ; and presently, as they reach the central ridge, they go ae THE NATURAL HISTORY OF THE FLOWER. 289 one on each side of it. The smooth grooves lead them up to the stigmatic slit in the most exact manner possible; but we find that they are too broad to enter. This is the chief difficulty that Darwin found in Stanhopea oculata : it has also been noticed by other writers. The author has unfortunately not had an oppor- tunity of examining this species. In it, according to Darwin, the pollinia do not shrink on drying; but in the species under consideration they do shrink considerably. The slit (fig. 9) in the pollinium narrows by the moving together of its sides, and the width of the pollinium is thus reduced sufficiently to enable it to enter the stigma. This process takes from 12 to 24 hours, according to the dryness of the air. The delay thus caused will ensure cross-fertilization. Ifthe experiment with the pencil be tried after this period, it is beautiful to see the manner in which the pollinia slide up to and into the stigma. As one moves the pencil on, the caudicles are stretched at right angles to the pollinia and presently break, leaving the latter buried in the stigmatic slit. Withina day or two the pollinia are sucked right into the cavity out of sight, and a sticky fluid exudes from the mouth. The flower was found by experiment to be self-fertile (in Cambridge). Flowers fertilized from another plant set good seed, even when the pollinia were a week old. The flower itself lasts only three days, even when not fertilized, a somewhat unusual circumstance in these large and showy orchids. We come now, finally, to the mode in which insects are attracted to the flower, and the way in which it is actually fertilized in nature. Unfortunately, as has already been explained, it has not been possible to study the plant in its native place; so that this part of the work must remain doubtful. The flower is very conspicuous from the side or from below, and possesses a very powerful vanilla-like odour, which gives a headache to a person sitting over it for any time, and renders our British bees almost incapable of walking if they are shut in with the flower fora couple of hours. The cage formed by the column and labellum hangs downwards, as already explained; and is extremely slippery inside and out except in a few spots, viz. the stalk portion of the column, the interior of the bucket (the slipperiness begins at the mouth of the bucket), and the back of the same. It was found, by experiments with hive- and humble-bees, that they were unable to climb up the slippery labellum. This fact, taken together with the pendulous position of the 290 MR. J. C. WILLIS’S CONTRIBUTIONS TO flower and the arrangement of the viscid disc, renders it evident that the flower cannot be fertilized by entrance of insects at the mouth of the cage. This suggested that the removal, or insertion into the stigma, of the pollinia was effected by insects sleding down inside the cage. This idea was tested by enclosing the flower in a large glass bell-jar and putting a number of large humble-bees, such as Bombus hortorum and others, in with it. After vainly buzzing round the glass for some time, these began to visit the flower, and their actions were carefully watched. Sometimes they alighted on the petals and wandered about upon them; but others found their way into the bucket by flying straight through the large openings between it and the column, which appear as if specially constructed for this purpose. The bucket contains a great number of juicy hairs, which are pro- bably nibbled by bees, as in other orchids; but our native bees did not seem to understand how to go to work. Presently, there- fore, they tried to return by the way they came. They could not, however, fly straight out, as the opening of the bucket itself was too narrow for them to spread their wings. They, therefore, had to crawl out; and as soon as ever they crawled over the edge, they came upon the highly polished surface upon which they could get no grip. Sometimes they hung by their hind legs for a time; but usually they slipped down at once, slid down the cage and out at the mouth. It was quite ludicrous to see the way in which the bees “ tobogganed”” down inside the cage ; they mostly fell upon the table before they realized what had happened. These experiments seem to favour the “ sliding” theory; but, unfortunately, the pollinia were never removed. The bees usually slid down with their backs to the column; and there is no pro- jecting point on the back of a humble-bee, so that the projecting tongue (c in fig. 7) was never caught. A bee, such as the Luglossa mentioned by Darwin, whose tongue is carried under its body and projects over its back from behind, would very probably remove the pollinia. So also would any bee with any projection on its back. Another possible case would be a bee with a sharp-pointed abdomen falling tail first. Some of the humble-bees experimented on fell tail first, but were too stout to put the tail under the rostellum. The curious hollow shape of the viscid disc also points to the probable occurrence on the visiting insect of some rounded projection (say the tail) which will fit it. THE NATURAL HISTORY OF THE FLOWER. 291 When the flowers were placed in the open air on a warm day in August, they were visited by various Syrphidew and by species of Bombus and Apis. From the general size of the parts and the absence of free honey, it would seem probable that the flower is adapted for large bees. Another point that still requires explanation is that the viscid matter of the rostellum does not set at once, as is usually the case in orchids on exposure to air*, but takes from 12 to 24 hours to do so; hence, unless the pollinia are very firmly pressed on to start with, they are easily knocked off subsequently. The smoothness of the labellum appears to be due to a very thick cuticle on its epidermal cells. This can be stripped off without difficulty in pieces an inch square. The bucket portion of the labellum is lined with stout hairs ; the cells of these and of the outer layers of the subjacent tissue have deep red-coloured cell-contents. They appear also to contain sugar (?). Pimelea decussata, R. Br., var. diosmefolia, Meisn. (in DeCan- dolle, Prod. xiv. p. 503).—The flower was studied upon a fine plant in the Cambridge Botanic Garden; when in bloom it is very conspicuous, every twig ending in a head of about forty pink-coloured flowers somewhat loosely arranged. The reddish- coloured perianth is tubular, 12 mm. long and 1 mm. wide at the mouth, with a spreading limb of four segments whose total width is about 7mm. The anterior and posterior segments are usually bent slightly downwards (Pl. XIX. fig. 11). The tube is hairy outside ; on the lower part the hairs are long and horizontal, on the upper part shorter and directed upwards. The two stamens are inserted on the throat, opposite to the anterior and posterior segments of the perianth ; they project beyond the perianth, and the pollen is thus exposed freely, as in so many dry-climate plants. The ovary consists of one carpel with a lateral style and capitate stigma. The flower has no scent; honey is secreted by four glands, two on the anterior, two on the posterior side of the base of the ovary. It is well protected from rain or short-tongued insects by the long perianth-tube. When the flower opens we see the two stamens sticking out, somewhat diverging from one another, with open orange-coloured * It is of course always, to some extent, exposed to the air before the pollinia are removed at all. 292 MR. J. C. WILLIS’S CONTRIBUTIONS TO anthers (fig. 11); the style is still enclosed in the tube, and its stigmatic papille are very short. The stamens now move slowly outwards and downwards, and the style grows out until the stigma stands at the level formerly occupied by the anthers ; its papille are now three or four times as long as in the earlier stage. The stamens now almost touch the perianth-lobes (fig. 12), and finally come to rest upon them (fig. 13). It is scarcely possible for autogamy to occur, as stigma and anthers are never so placed that pollen can get from one to the other. Pollen may fall from one flower, however, to another upon the same head. Out of 77 flowers examined, only 13 had set seed. Various Diptera (Syrphidx and others) were observed visiting the flowers; most of them confined themselves to feeding on pollen, as their tongues were too short to reach the honey. The flower appears more suited to bees or the long-tongued flies, or Lepidoptera. Cotyledon Umbilicus, Linn.—This is one of the few plants native in Britain whose floral biology has hitherto remained unknown. The plant has a tall raceme of inconspicuous scentless greenish- white flowers, pendent upon short stalks. The calyx is very small, the gamopetalous corolla tubular, about 10 mm. deep and 3 mm. in diameter (Pl. XIX. fig. 14). The stamens, ten in number, are inserted on the corolla near the mouth, those opposite to the petals being a little higher up than the others. This latter point is of some importance, as the flower is obdiplostemonous, and so has the stigmas opposite to the long stamens (fig.15). The five carpels have each a large scale at the base, apparently, as in the rest of the order, secreting honey. This organ is reddish yellow in colour, and is visible, through the corolla, from outside, giving a reddish tinge to the base of the petals (n, fig. 14). The stamens dehisce when the flower opens, standing on a level with the stigmas, which are as yet not fully developed (figs. 15 and 16). There is a very small space separating the anthers from the stigmas. In consequence of the pendulous position of the flower (it hangs at an angle of about 40°-80° with the horizontal), the anthers of the long stamens are a trifle below the stigmas. Later on, the latter develop, growing outwards, as shown in fig. 17. This brings them extremely close to the anthers. Young flowers stand out more horizontally than old ones. An insect probing the flower for honey in the young stage will receive pollen upon its head or proboscis, and will probably THE NATURAL HISTORY OF THE FLOWER. 293 place it upon the stigma in other flowers, the entrance to the honey being very narrow. In the later stage of the flower autogamy will almost certainly occur, whether it be visited by insects or not. The distance between anthers and stigmas is extremely small, and a very slight displacement, such as might be caused by an insect grasping the corolla, will bring them into contact with one another. In the final withering of the flower, the corolla collapses over the stigmas, and autogamy may occur then. Aphides are common in the flowers; one of them was observed in the act of producing autogamy. It became stuck to the stigma in climbing out of the flower, and in its struggles almost emptied the neighbouring anthers of their pollen, which it brought upon the sticky stigmas. The pollen is smooth, but not, apparently, powdery, so that the mechanism cannot be compared with that of Erica or Galan- thus. Although the flower is pendulous, the pollen is often wetted by rain getting in; the water collects in a large drop at the mouth of the corolla, and then gets inside very easily. After a heavy shower, many flowers were found full of water. The flower is thus protandrous, with well-marked self-fertiliza- tion. All the flowers examined had set a full complement of seed. The only insect visitor observed beyond the Aphides above mentioned was Thrips, but only a few racemes could be watched, as there was but little material. Some stalks were found bearing male flowers only at the top of the raceme ; these had small shrivelled carpels. The plant may thus be andro- monecious under certain circumstances. One plant bore, in the middle of the raceme, about twenty flowers with all their organs aborted except the calyx. Hydrolea spinosa, Linv.—This plant belongs to the same family as Phacelia *, and its mechanism is somewhat similar. The flowers are in loose cymes, and azure blue in colour, with no perceptible scent. The stems, leaves, and calyces are covered with glandular hairs. The corolla is rotate-campanulate, 13 mm. across and 10 mm. in depth, the segments long, the tube short and wide. The five stamens are exserted, and bear versatile anthers with a large connective. Each lobe dehisces by a longitudinal slit, but does not at once turn completely inside out as in Phacelia. The pollen is brown. Honey is secreted by a disc below the ovary, and protected, asin Phacelia, by processes of the stamens. These are thick and fleshy, covered with stout papillx, and block up * Cf, antea, p. 53. 294 MR. J. C. WILLIS’S CONTRIBUTIONS TO almost all the space between the corolla and ovary (figs. 19, 20). The latter bears glandular hairs on its upper part, above the staminal processes (figs. 18 and 20), and ends in a bifid style, with terminal stigmas, which are very sticky. There are no conspicuous honey-guides. Although so conspicuous, the flower, at least in England, is excellently adapted to self-fertilization. The stamens are bent inwards, and are usually found touching the stigmas when the flower opens. Soon the styles bend outwards a little, carrying the stigmas just clear of the stamens, and giving a chance of eross-fertilization, and then finally the flower shrivels up, and in so doing almost always pollinates its own stigmas. Every flower sets seed, even in the absence of insects. Nemophila maculata, Benth.—The only specimens available bad been some time in cultivation, and therefore were not suitable for a study of the dichogamy, but the general natural history characters of the flower could easily be made out. The flower is very con- spicuous, owing to the blue spots that terminate the white corolla- lobes ; these are also veined with blue (honey-guides). The calyx is softly hairy. The stamens are included, with versatile anthers, which usually hang so as to dehisce extrorsely ; they shorten and roll backwards on dehiscence. The ovary is hairy, rounded (not flattened as in Phacelia), and glandular ; the style short, bifid, hairy, especially below, and glandular; the stigmas terminal. The corolla bears appendages like those of Phacelia tanacetifolia*, which make a little tube leading to the honey, which is secreted by the disc below the ovary. At the mouth of the tube are several dark spots on the corolla, which may serve as path-finders. The honey is perhaps most easily obtained by probing these tubes, but is accessible, as in Phacelia, between the stamens and pro- cesses of the corolla, at the base of the flower. The flower shows protandry very similar to that of Phacelia divaricata, and is visited chiefly by bees, to which and to long- tongued Syrphide it would appear to be best adapted. The stigma is not quite mature when the anthers dehisce, but soon reaches maturity. The flowers generally set seed, and probably often fertilize themselves (cf. Phacelia divaricata). Nemophila thus resembles the other Hydrophyllacee in its general characters, approaching Hydrophyllum on the one hand and Phacelia on the other. Ziziphora capitata, Linn.—This member of the Labiate bears * Cf. anted, p. 54 and figs. 4, 5. THE NATURAL HISTORY OF THE FLOWER. 295 very small inconspicuous flowers, and appears to fertilize itself regularly. The calyx is tubular, 8 mm. long, constricted at the middle ; the corolla projects about 1-2 mm. beyond it, and opens to a width of 2 mm. across the two lips; the lower lip is 3 mm. wide. The conspicuousness is thus very small; the corolla is of a purplish red. The tube of the corolla is only -4 mm. in dia- meter in its lower half, and about ‘9 mm. above. At its mouth it is crossed by the filaments of the two anterior stamens (the posterior pair are aborted), which run up under the upper lip. The corolla is hairy outside and upon the lower lip, while lines of hairs continue down it, inside, from the insertion of the stamens. The anthers cohere at first like those of Monarda, and the style runs up behind them and curls down over aud between them. The stigma is thus pollinated as soon as the anthers dehisce, and as it soon after shrivels up, there is scarcely any chance of a cross-fertilization taking place. Almost every flower examined had set seed, and most of them their full number. Fourteen flowers gave forty fruits altogether, or over 70 per cent. of the possible number. From the length of the tube and the colour, &c., the flower would seem to have been formerly adapted to bees or Syrphida, but now it apparently receives few or no visits and fertilizes itself regularly. CLEISTOGAMY. Salvia Verbenaca, Linn.—This species exists under two forms, the normal and the variety clandestina. Only the normal form occurs in Cambridgeshire, according to Babington *. The species is, however, a very variable onef. The author has carefully observed many plants growing at Grantchester, Ditton, and other places near Cambridge, and has also cultivated specimens (from wild seed), which behaved similarly to the wild ones. The plants bear a great variety of patterns of flower, the bulk of which, however, during 1892-98 have been, not the normal open flowers, but rather cleistogamous, with reduced corollas. Various types are shown in the figures (Pl. XIX. figs. 21-25). The vast majority of the flowers were of the type shown in fig. 21, the corolla not at all or hardly visible at the mouth of the calyx, and completely closed. At the time of fertilization the corolla was * ‘Flora of Cambridgeshire,’ p. 176. + Timbal-Lagrave, “ Recherches sur les variations que présentent quelques plantes communes dans le département de la Haute-Garonne.” Mém. Acad. Sci. Toulouse (Reprint, not dated). 296 MR. J. C. WILLIS’S CONTRIBUTIONS TO about 2 mm. long, and 14 mm. wide, blue at the further end (fig. 22). The position of the essential organs is shown in fig. 23, the style being curled round the anthers in the lower lip, where it is fertilized. Nearly every flower set a full complement of fertile seed. During the early part of June 1893 all the flowers were of this type. Towards the end of the month appeared flowers with larger corollas, as much as 5 or 6 mm. long; these were slightly open at the mouth, but the style still ran into the lower lip and was fertilized there. Soon after, flowers like fig. 24 were found, with the style projecting at the mouth, but it was found to have been pollinated in the bud. Finally a few flowers were observed like that shown in fig. 25, with the corolla fairly large, its mouth wide open, and the lever mechanism in good working order. These flowers were fully capable of insect- fertilization, and did not pollinate themselves in the bud. No insects were seen to visit them, however, which is not surprising when we consider their inconspicuousness. Later on in the season (August onwards) the plant bore the cleistogamic flowers again, without any intermixture of open ones. On the whole, as the season advanced, the corollas became steadily smaller, till in November they were once more only about 2 mm. long ; the flowers, however, were still fully fertile. One flower was noticed with the stamens completely aborted. It would thus appear that the forms of flower on Salvia Ver- benaca vary considerably with the season of the year at which they are examined, and this fact must therefore be taken into account in considering the forms and varieties of the species. From a natural history point of view, the facts described above are of some interest as throwing light upon the origin of cleisto- gamy. The plants bore flowers with very reduced corollas at the beginning and end of their flowering season, and the larger forms chiefly in June and July. It would thus seem probable that the cause of the variation in size of the corolla is either light or temperature, or both together, but the observations made gave no evidence as to which of the two was the important factor. In many plants with cleistogamic flowers, e. g. Lamiwm amplexicaule, these flowers appear at a season when the light and temperature are lower than at the time of appearance of the normal flowers. In others, e.g. Viola, the cleistogamic flowers appear in the brighter and warmer period, but are found usually to be on the shaded parts of the plant. The first-named facts THE NATURAL HISTORY OF THE FLOWER. 297 leave it undecided whether light or temperature be the main factor; the latter: seem to point to reduced light as the main cause of reduction of the corolla. This has been in a large measure proved by the recent experiments of Vichting *, upon the action of light of various degrees of intensity upon the flowers of plants which had a tendency to cleistogamy. At the same time, this does not seem to explain everything ; for instance, the fact observed many times on Salvia, that normal flowers and cleistogamic flowers occur side by side at the same node, This phenomenon the author is inclined to regard as an effect of varying temperature: the flowers had not usually opened simultaneously, and so had been exposed to different conditions of temperature at corresponding periods of growth. This would cause in them varying degrees of protandry (according to the hypothesis of the origin of dichogamy proposed by Meehan and adopted by the writert). Those that were thus caused to become very protandrous would be fertilized in the bud, and the corolla would probably wither without ever opening. A slightly less degree of protandry would give the type of fig. 24 and a lesser still would give the normal flower. This reasoning of course is not conclusive, as it may be applied to the action of light instead of heat ; but we have a certain amount of evidence in favour of the view that dichogamy is dependent on temperature, and it therefore does not seem a very forced view, especially as the temperature conditions are perhaps more variable than those of light during a short period. Viola tricolor, Linn.—Véchting’s experiments bring out, among other interesting facts, that in a zy gomorphie flower the upper lip becomes reduced, under the action of diminished light, more rapidly than the lower. Some specimens of Viola tricolor, var. arvensis, gathered on Dee. 9, 1893, illustrate this reduction very well. The whole corolla is small in comparison with the calyx, but the upper lip is far more reduced than the lower (fig. 26). Finally, the author wishes to thank Messrs. F’. Darwin, R. I. Lynch, and I. H. Burkill for much valuable advice and assistance kindly rendered. * “Ueber den Einfluss d. Lichts auf die Gestaltung und Anlage der Blithen.”’ Pringsheim’s Jahrb. xxv. 1893, pp. 149-208. 7 +t “On Gynodiecism, with a preliminary note on the origin of this and Similar phenomena.” Proc. Camb. Phil. Soe. viii. 1893, pt. iii, LINN. JOURN.—BOTANY, VOL. XXX. YX 298 21 99 ~—s 23. 24, 25. 26. THE NATURAL HISTORY OF THE FLOWER. EXPLANATION OF THE PLATES. Prats XVIII. Brodiea ixioides. Flower, with half of perianth removed, in early stage; from a photograph. Nat. size. ° Do. Stamen with section of perianth. Nat. size. Do. Tip of horn of stamen. X50. Do. Flower in late stage; from a photograph. Nat. size. . Stanhopea tigrina. Labellum and column in natural pendent position ; the rest of the perianth removed ; from a photograph. 3. . Do. The same organs viewed from below, showing the form of the mouth of the cage ; from a photograph. }. . Do. The apex of the column with the anther, &c. ; from a photograph, x22, . Do. Pollinia, rostellum, &c. from above. Nat. size. . Do. The same from below. x2. . Do. Long. sect. apex of column, with pollinia and anther removed. Nat. size. Puatr XIX. Pimelea decussata, Top of flower in early stage. Nat. size. . Do. The same, in intermediate staze. Nat. size. . Do. Final stage. Nat. size. . Cotyledon Umbilicus. Flower. Nat. size. . Do. Young flower seen from the mouth. x8, . Do. Stigma of young flower, from side. x6. - Do. Stigma of old flower, from side. x6. . Hydrolea spinosa, Ovary and styles. 2. - Do. Part of corolla, with stamens attached, toshow the processes ;_ it has been flattened out somewhat in order to separate these. x14. - Do. Cross section of flower, just above the ovary. X4. Salvia Verbenaca. Cleistogamous flower of the smallest: size, after fer- tilization, The corolla has become detached and moved forwards, so us to be visible beyond the calyx, which was not the case before fertilization. x 4. Do. Corolla of same flower. x6. Do. Corolla with upper lip removed, showing the stamens and style. The dotted lines show their position in the lower lip. x6. Do. Larger flower, open at mouth. x3. Do. Largest flower. x1}. Viola tricolor. Flower with corolla reduced, especially the upper lip. a=anther ; 4=projecting end of column; c=tip of rostellum ; cp=carpel; ¢._par=conducting parenchyma; col=column; d=central ridge of column ; e=ridge, bearing anther; epi=epichil; f=lateral bar of cage; 4=horn of stamen ; kyp=hypochil (bucket) ; 4=tip of ridge after removal of anther ; m=cavity in pollinium ; mes=mesochil ; n=nectary visible through corolla ; pet = pe tal; p.s=perianth lobe; po/=pollinium ; 7=rostellum; s¢=stamen ; stig=stigma ; sty=style; v.b=vascular bundles. Linn Wilus. as ) ice ta ‘ Witlts . il-13, PIMELEA. 14-17, COTYLEL Al-a5 SAI LInn. 22 Soc. Journ Bor Vou. XXX. Pr, 19 YON 18-40, HYDRCIEA VIA 26,VIOLA ON THE CYPERACEE OF LINNEUS. . 299 On certain Authentic Cyperacee of Linneus. By C. B. Crarxe, M.A., F.LS. [Read 15th March, 1894.] I HAVE, many times, worked through the Cyperacex (other than Carex) in the herbarium of Linneus preserved in this room, and have compared them with the volumes of the two editions of the ‘Species Plantarum’ annotated in Linnezus’s handwriting. The herbarium of Linneus appears to have originally contained a perfect, or very nearly perfect, set of examples (one, two, or rarely more, sheets of each species) representing the Sp. Pl. ed. Each sheet of this set was numbered and named in the hand of Linneus, in ink, on the paper itself, the numbers corresponding always to the species-numbers in the first edition of the Sp. PI. The herbarium of Linnzus has been, as is well known, dis- arranged: a quantity of additional material has been mixed into it; some of the original names have been crossed out; some of the original sheets have been moved, and many have disappeared altogether. There still remain, however, in the Cyperacexe enough of these authentically named sheets to verify more than half the species described by Linneus in his Sp. Pl. From com- paring these so many times, I have come to the conclusion that we possess, in them, in every case, a specimen plant that Linneus himself referred to the name he wrote on them. Besides these (usually numbered) specimens there are a good many other sheets named in Linnzeus’s hand (but without numbers) which refer to species subsequently published under Linnzeus’s names in the Sp. Pl. ed. ii., in the Mantissa, and in Linn. f. Suppl. These authentic examples are not, however, the “types” of Linneus’s species in the sense understood by many modern botanists. Linneus conceived a species as an entity and did not suppose that it could be restricted to one “ type.” A “species,” in his Sp. Pl, is made up of 4 (or fewer) parts, viz. (a) the citations of predecessors ; (6) the citations of pictures; (c) the diagnosis of Linneus himself; (d) the authentic examples in Linneus’s herbarium. Perhaps the most important of these four is the first: Linnzeus meant invariably that his species should be the sp. auct. of his day ; Linnezus generally draws his diagnosis by directly copying from his predecessors, only altering it so far as to make it include his authentic specimens. Where, as is y2 300 MR. C. B. CLARKE ON CERTAIN frequently the case, his authentic specimens were wrong, these alterations of his predecessors’ diagnoses introduce difficulties that can only be dealt with, as best they may, for each individual ease. The pictures cited by Linneus I find of small value, at least in the Cyperacee (and a few other cases where I have tested them). In the first place, the pictures of Linneus’s day were poor, and it 1s often difficult to feel any confidence as to what species they belonged to. They often give no detail, but so very general a view of the plant that it is impossible to be certain to which species, among half a dozen as species are now recognized, the figure exactly belongs. In very many cases, all I can say is that the figure may do for the plant we call it. Secondly, even when the figures were good, Linnzus seems to have allowed for a good deal of variability in the plant and imagination in the artist. It is, in a word, my opinion that a large percentage of the figures cited by Linneus are wrong, or very doubtful. From the foregoing data we have to decide in each case to what plant we will apply each Linnean specific name. I have formerly, as in Journ. Linn. Soe. vol. xxi. [1884] pp. 34-202, retained I believe every Linnean name for the name of some species (now known more or less generally by such name) by the resource of citing Linn. partim; as ‘Cyperus Haspan, Linn. partim”’ : in this case Liuneus meant by Cyperus Haspan the plant that we do, including several which we speci- fically separate, but Linneeus’s authentically-named Cyperus Haspan is a totally different plant, and his diagnosis is therefore unsatisfactory. I can hardly expect systematists to be of one mind how far we ought to strain facts in order to avoid wholesale alteration of present names—the “nomina auctorum.” Thus “ Scirpus supinus, Linn.,” is a widespread and universal species, of which the name has never been doubted; but the authenti- eally-named plant is “ Alisma ranunculoides, Linn.’ I think even here from the citations of his predecessors and the ground- work of his diagnosis that Linneus really meant our Scirpus supinus, and I retain the name accordingly. The case is different where the diagnosis of Linneus (or any other author) is fundamentally wrong, 7. e. states as a character of the species an essential point that is never to be found in the species now so named. In this case also it is usual to cite “ Linneus partim ” or “ Linneus emend.”; but such a citation is only complimentary or traditional and often very misleading. AUTHENTIC CYPERACEE OF LINNEUS. 301 We have to look down into the synonymy and see what species of Kunth or DC. is really meant: it is therefore, I believe, better to cite the species in chief as of Kunth or DC. and to place the “Linn. emend.” among the synonymy. In the reduction of Linneus Sp. Pl., so far as they are repre- sented by authenticated specimens, which reductions form the substance of the present paper, I have been able to retain all the specific names, nearly in their commonly received sense, except two, viz. :—(1) Cyperus elegans, Linn., is altogether (i.e. diagnosis, picture cited, and specimens) the plant which Rottboell calls elegans, and moderns call C. viscosus, Aiton. The plant called Cyperus elegans by Swartz and Kunth is a totally different species, viz. what we call C. diffusus, Vahl. (2) The name Cyperus odoratus, Linn., has been largely employed by botanists: all that can be gathered from the diagnosis is that it was a Jarge species ; from the figures of Sloane cited, some botanists have imagined it to be a species of “ Diclidium” closely allied to Mariscus ferax (Cyperus ferax, L. C. Rich.) ; others have supposed it equal to Cyperus strigosus, Linn., or one of the species closely allied thereto. Linneus has, however, two good specimens on his authentically named sheet: one is Pycreus polystachyus, Beauv., the other is Cyperus malaccensis, Lam. Upon these data, I call Cyperus odoratus, Linn., nomen delendum. There has been much dispute of late regarding the citation and treatment of the Linnwan genera. The first genus of Linneus’s Cyperacee is Schanus, which in the lst edition of the Sp. Pl. contains 9 species. These are referred now as follows :— 2 to Schenus, 1 to Cladium,1 to Cyperus, 1 to Kyllinga, 1 to Scirpus, 2 to Rynchospora, 1 to Crypsis (a Grass). (In the 2nd edition of the Sp. Pl. three new Scheeni sp. are added ; of these 1 is Dulichium, 1 is Tetraria, 1 is Scleria.] The modern genus Schenus contains 60 species, mostly Australian, and in Benth. et Hook. f. Gen. Pl. vol. ii. p. 1062, Bentham cites it as “ Sckenus, Linn., pro parte.” Bentham cites it from the 5th edition of the Genera [1767], whereas some say he ought to have cited it from the 2nd edition [1742]; that makes, no difference whatever. The character of the genus, given by Linneus, opens with “ petals 6,” which character cannot, in Linneus’s idea of petals, be forced to be true either of any one of Linnwus’s 12 species, or any one of the 60 species now in Schanus. 3802 - MR. C. B. CLARKE ON CERTAIN It is frequently said that Linneus’s genera are our Suborders or Tribes, and that when we make a Tribe of 6 or 8 genera out of one of Linneus’s genera it is well to apply the Linnean name to the largest or most characteristic genus, pro parte. But nothing of the kind can be said here; Linneus’s eleven Schent sp. are taken at haphazard throughout the Order (the grass being thrown in) from the most remote and unlike genera. To call the present genus “ Schaenus, Linn.,” in any sense, or even “ pro parva parte,” is wholly complimentary. When I have brought these things to the notice of some experienced botanists, they have attempted to evade any general conclusions thence as to how the genera and species of Linneus are to be treated by observing that the Cyperacee are unlike other orders, and that Linneus’s treatment of other orders will be found different. I have tried several other orders, and it appears to me that Linneus worked in just the same way through- out: he took the diagnoses of his predecessors, he identified the pictures very boldly, he made his diagnosis include somehow his type specimens. Here follows a reduction of all the Cyperacez (except Carex) now in the Linnean Herbarium, which have their names written on the sheets in the hand of Linneus. Where we still keep the species (whole or part) in the same genus in which Linneus placed it, I have further considered in what form the citation of it should be made. Where the species is now to be placed under some genus later than Linneus, it is not requisite to expend thought or time on the citation of Linneus under the synonymy. Scuenvs Maniscvus, Sp. Pl. ed. 1, p. 42, n. 1. There are two sheets marked, the first ‘“‘Mariscus 1” in Linneus’s hand, the second “Schcenus Mariscus” in Linneus’s hand; they are both = CLADIUM JAMAICENSE, Crantz. The diagnosis and citations of Linnexus agree. Screnvs acurzatvs, Sp. Pl. ed. 1, p. 42, n. 2. The plant marked in Linneus’s hand Schenus Mariscus has been ticketed as subsequently Sch. aculeatus of the Sp. Pl.—quite AUTHENTIC CYPERACEE OF LINNEDS. 303 wrongly. ‘The description, the figure of Morison cited by Linneus, and his other references show that Sch. aculeatus, Linn., was altogether Crypsis ACULEATA, Aiton. Scha@nvus Mucronatus, Sp. Pl. ed. 1, p. 42, n. 3. One sheet marked “ Schenus mucronatus” in Linneus’s hand. Schenus mucronatus, Linn., as to the diagnosis, citations, and plant named, is altogether Cyperus capiratus, Vandelli. Scuenvs nrericans, Sp. Pl. ed. 1, p. 43, n. 4. There are two sheets marked, the first ‘4. nigricans” in Linnzeus’s hand, the second “4”? in Linnzeus’s hand. Schenus nigricans, Linn., as to the diagnosis, citations, and examples, is Goon. Scu@nus FERRUGINEDS, Sp. Pl. ed. 1, p. 43, n. 5. One sheet, marked “5. ferrugineus ” in the hand of Linneus. Schenus ferrugineus, Linn., as to diagnosis, citations, and examples, is Goon. Scuenvs cororatvs, Sp. Pl. ed. 1, p. 43, n. 6. Two sheets: the first, marked “Schenus coloratus” in Linneus’s hand, is Kyllinga brevifolia, Rottb.; the second, marked “coloratus” in Linneus’s hand, is Jariscus coloratus, Nees. . The second sheet (i. e. Mariscus coloratus) is supposed to belong to Var. 6 in ed. 2, p. 64; 2. @ it is Scirpus glomeratus, Linn. Sp. Pl. ed. 1, p. 52. A third sheet, marked “glomeratus” in Linneus’s hand, is Kyllinga monocephala, Rottb. The picture cited by Linneus, viz. “Sloane Jamaica, t. 7S. fig. 1,” is Dichromena. Scna@yvs GuomERATs, Sp. Pl. ed. 1, p. 44, n. 8. One sheet, marked “8. glomeratus” in Linneus’s hand, is Rynchospora glomerata, Vabl. Schenus glomeratus, Linn., is, as to diagnosis, citation, and example, . Ryncuospora GLOMERATA, Vahl, 304 MR. C. B. CLARKE ON CERTAIN Scuanvs ausus, Sp. Pl. ed. 1, p. 44, n. 9. One sheet, marked “9. albus” in Linneus’s hand, is Ryncho- spora alba, Vahl, a. [A second sheet, marked “albus,” perhaps not in Linneus’s hand, is also R. alba, Vahl, a. | Schenus albus, Linn., as to diagnosis, citation, and example, is Ryncuospora ALBA, Vahl, a. Scnexts capensis, Sp. Pl. ed. 2, p. 64, n. 7. One sheet, named in Linneus’s hand “ Capensis.” The word “ Capensis”’ has been struck out and “thermalis” substituted by another hand. The plant is Zetraria thermalis: it does not suit Linnzeus’s description at all. Schenus capensis, Linn., as to descr. is referred by Kunth to Thamnochortus dichotomus, R. Br.; but Dr. Masters does not accept this determination. The plant marked by Linneus as “ Capensis ” is according to Riehter the foundation for Schenus thermalis, Linn., Mant. p. 179. _ A sheet, now in the Schenus bundle, and marked in Linneus’s hand “ fuscus,” is Ryncwospvora Fusca, Link. CrpErts articuratvs, Sp. Pl. ed. 1, p. 44, n. 1. One sheet, marked “1. articulatus” in Linneus’s hand, is C. articulatus. Cyperus articulatus, Linn., as to diagnosis, citations, and example, is Goop. . CyPERUS TENELLUS, Linn. f. Suppl. p. 103. One sheet, marked “tenellus” in Linneus’s hand, is CyPERUS TENELLUS. Cyperus monostacuyos, Linn. Mant. p. 180. One sheet, marked “ Cyperus monostachyos’? in Linneus’s hand, is Frupristyiis Monostacnya, Hassk. Cyperus Lxvieatus, Linn. Mant. p. 179. One sheet, marked “ Cyperus levigatus” in Linneus’s hand, is JUNCELLUS L&vIGAtUS, O. B. Clarke. Cyrerus Haspan, Sp. Pl. ed. 1. p. 45, n. 4 One sheet, marked “ 4. haspan ” in Linnzeus’s hand, is Cyperus Tria, Linn, AUTHENTIC CYPERACEE OF LINNEUS. 805 The diagnosis and the figure of Plukenet cited are Cyp. Haspan ; and the species must stand :— Cyperus Haspan, Linn. Sp. Pl. ed. 1, p. 45 et ed. 2, p- 66, neque herb. propr. Cyperus Lonats, Sp. Pl. ed. 1, p. 45, n. 5. One sheet, two examples, marked “5. longus” in Linneus’s hand, are both Juncellus serotinus (Rottb. sub Cypero). The figure of Morison, Hist. sect. 8, t. 11. fig. 13, is, I think, C. esculentus. (Certainly not C. longus, of which species an im- portant character is that the stem is not erect at base.) Rather than change the name of @. longus, Linn., I propose that our English plant shall be called Crrerus Loneus, Linn. Sp. Pl. ed. 1, p. 45 partim (tab. citat. et exemplo in hb. propr. excl.). CyrErus rotunpvs, Sp. Pl. ed. 1, p. 45, n. 6. Two sheets, one marked “ rotundus 6” in Linneus’s hand, is Mariscus, perhaps Mariscus tabularis (C. tabularis, Schrad.). The example is young, but remote from C. rotundus. The other sheet, marked “rotundus” in Linnzus’s hand, is Cyperus esculentus. This species must stand :— Cyrerus rorunpus, Linn. Sp. Pl. ed. 1, p. 45, neque herb. propr. Cyperus squarrosvs, Sp. Pl. ed. 2, p. 66, n. 3. One sheet, marked in Linneus’s hand “ pumilus,” but this name struck out and “ squarrosus ” substituted, all in Linnzus’s hand, contains two things: one is a tuft of many complete plants of Cyperus aristatus, Roxb.; the other is a single head of aris- cus squarrosus. The plate of Plukenet cited by Linneus is probably neither of these. All that can be done with Linnzus’s name is to reduce it :— { partim = Cyperus aristatus, Rottb. partim = Mariscus squarrosus (Linn. sub Cypero). Cyperus pirForMis, Sp. PI. ed. 2, p. 67, n. 8. Three sheets, one marked “Cyperus difformis,” the second “C. diformis,” the third “Scirp. deformis,” all in Linuzus’s hand all are Cyperus difformis. . — The reference to Plukenet, t. 817. fig. 5, is a misprint for t. 417. fig. 5, which is C. difformis. 306 MR. C. B. CLARKE ON CERTAIN Cyperus difformis, Linn., is, as to diagnosis, citation, and examples, Goon. Cypervs Iria, Sp. Pl. ed. 1, p. 45, n. 8. One sheet, originally marked “8. Iria” in Linneeus’s hand, but “Iria” struck out and “haspan’’ substituted, also by Litneus’s hand, contains two good specimens, whereof one is C. Iria, Linn., the other C. diffusus, Vahl. The figure cited of Plukenet (Alm. t. 191. fig. 7) is C. Iria; that cited of Rheede (Hort. Malab. vol. xii. t. 56) is Juncellus tnundatus, Roxb. (sub Cypero). This species may be cited :— Cyrervs Inia, Linn. Sp. Pl. ed. 1, p. 45 (tab. Rheedei citat. excl.), et herb. propr. partim. Cyperus ELATUS, Sp. Pl. ed. 2, p. 67, n. 10. One sheet, marked “ Cyp. elatus” in Linneus’s hand, is the species I have so called in Journ. Linn. Soe. vol. xxi. [1884], p- 189, and is called C. racemosus by Steudel and Boeck. ; but I regard it as very doubtful what C. racemosus, Retz., really was (Arnott considered it to have been O. digitatus, Roxb.). At all events, the description of Linneus in the Amen. Acad. vol. iv. p- 801, altogether agrees with his authentic specimen; and I apply to the plants Wallich 3341 a, Zollinger 350, the name CYPERUS ELATUS, Linn. CYPERUS GLOMERATUS, Sp. Pl. ed. 2, p. 68, n. 12. Two sheets, each marked “ glomeratus ” in Linnzus’s hand. Cyperus glomeratus, Linn., is altogether (i. e. diagnosis, figures cited, and authentic specimens) Goon. CYPERUS ELEGANS, Sp. Pl. ed. 1, p. 45, n. 9. One sheet, marked “9. Cyp. elegans” in the hand of Linnzus, is Cyperus elegans, Linn., i.e. C. viscosus, Aiton, and nor C. elegans of Swartz and Kunth. The description of Linneus fits C. viscosus, Aiton, exactly, and will not fit C. elegans, Kunth. The figure of Sloane cited [t. 75. fig. 1] is also C. elegans, Linn. We have Cyperus ELEGANS, Linn. (and Rottb.), Goon. The Cyperus elegans of Swartz and Kunth will be C. diffusus, Vahl. Cyperus oporatvs, Sp. Pl. ed. 1, p. 46, n. 10. There is one sheet, marked “10. odoratus” in Linneus’s hand, AUTHENTIC CYPERACEX OF LINNEUS. 3807 and containing two good complete specimens; one is Pycreus polystachyus, Beauy., the other is Cyperus malaccensis, Lam. Linneus cites two plates fur his C. odoratus, viz. Sloane, t. 74. fig. 1, and t. 8. fig. 1. As regards the first of these plates, cyperologists are not agreed to what section it refers; some Suppose it to represent a Mariscus near M. strigosus; others suppose it to represent a Diclidium and to be equal to or near Cyperus ferax, L. C. Rich. As regards Sloane’s other plate, t.8. fig. 1, it doubtless represents one of the Papyri near C. gigan- teus, Vahl. As to Sloane, t. 74. fig. 1, the spikelets appear to me much too broad for any Diclidium. It is true that, in the older stages of Diclidium, the glumes stand out very stiffly and the spikelets are then in outline broad, though the outline is very broken ; but Sloane’s plate represents the glumes very closely imbri- cated; in the young state of Diclidium (Cyperus ferax, L. C. Rich.), before the glumes spread, the spikelets are linear. Cyperus odoratus, Linn., is (mihi) nomen omnino delendum. Cyperus compressvs, Sp. PI. ed. 1, p. 46, n. 11. One sheet, marked “ Cyperus compressus ” in Linneeus’s hand, is C. compressus. _ Cyperus compressus, Linn., is altogether (as to diagnosis, picture cited, authentic example) Goon. CYPERUS FLAVESCENS, Sp. Pl. ed. 1, p. 46, n. 12. One sheet, marked “12. flavescens” in Linneus’s hand, is Pycreus rivularis, a North-American sp. nearly allied to the European Pycreus flavescens, Nees. Boott has consequently in MS. in the Mus. Brit. named some of the American plant Cype- rus flavescens, Linn. , From the citations it is clear that Linneus meant the European Pycreus flavescens, Nees, for his species. Under Pycrreus Fra- vescens, Wees, I cite accordingly Cyperus flavescens, Linn. Sp. P). ed. 1, p. 46, neque herb. propr. Cyprrus pumitvs, Sp. Pl. ed. 2, p. 69, n. 17. One sheet, named “ Cyperus humilis” in Linneus’s hand, but the word “humilis” struck out and “ pumilus” substituted, also by Linneus’s hand, contains two specimens. One of these I take to be original and is Pycreus nitens, Nees (= C. pumilus, Linn.). The other specimen is marked “‘ Hort. Fothergill, 177s. 808 MR, C. B. CLARKE ON CERTAIN Fairbairn. J. E. S.;” and I suppose to have been added by Sir J. E. Smith; it is Pycreus flavescens, Nees. Under Pycrevs nirens, Nees, I cite Cyperus pumilus, Linn., Sp. Pl. ed. 2, p. 69. Cyperus TriFLorts, Linn. Mant. p. 180. One sheet, marked “ Cyperus triflorus’’ in Linneus’s hand, is Finbristylis tristachya, Thw. Cyrervs stricosvus, Sp. Pl. ed. 1, p. 14, p. 47. One sheet, marked “ 14. strigosus” in Linneus’s hand, is Mariscus strigosus. The figure cited, viz. Sloane, Jamaica, t. 74. figg. 2, 8, cannot (in my judgment) have been drawn from MM. strigosus [I think t. 74. fig. 1 perhaps was]. Under Mants- cts strigosus I cite C. strigosus, Linn. Sp. Pl. ed. 1, p. 14, n. 47 (tab. Sloane excl.), et herb. propr. Cypervs LiatLaris, Sp. Pl. ed. 2, p. 70, n. 19. One sheet, marked “ Cyp. ligularis’”’ in Linneus’s hand, is Mariscus rufus, H. B. K.—Cyperus ligularis, Linn., is (as to diagnosis, picture cited, and authentic example) Mariscus rufus, H. B. K. Cyrverus Papyrus, Sp. Pl. ed.1, p. 47, n. 15. One sheet, marked “ Papyrus” in Linneus’s hand, is Cyperus Papyrus, Linn. Linneus’s diagnosis of this species is remarkable—“ umbella simplici, pedunculis simplicissimis distiche spicatis.” But I think there is no doubt that Linneus had the right plant; and I make Cyperus Papyrus, Linn., to be Goon. Crrerts arunpInacevs, Sp. Pl. ed. 1, p. 44, n. 8. C. ferrugineus, Sp. Pl. ed. 2, p. 63 in citat. Schenus spathaceus, Sp. Pl. ed. 2, p- 63, n. 4. One sheet, marked originally “3. arundinaceus ” in Linneus’s hand. He then struck out “arundinaceus”’ and substituted “alternifolius;” he then struck out “ alternifolius’’ and sub- stituted “ vaginatus,” which now stands on the sheet. The plant is Dulichium spathaceum, L. C. Rich. ‘ The diagnoses and citations of Linneus agree; and his three names above are all reduced under DuticnHiuM sPATHACEUM, L. C. Rich. AUTHENTIC CYPERACES OF LINNEUS. 309 CYPERUS ALTERNTFOLIUS, Linn. Mant. p- 28. One sheet, marked by Linnzus “ Cyperus alternifolius,” is our commonly cultivated species so named. Cyperus pistans, Linn. f. Suppl. p. 103. One sheet, marked “ distans ” in Linneus’s hand, is C. distans, “auct.” (i.e. C. elatus, Presl, in Oken, Isis, vol. xxi. [1828] p. 271, non Linn.). In the Cyperus bundle is one sheet marked Cyperus panno- nicus; on this Linneus wrote “ Scirpus Michelianus?’” The plant is really Cyperus fuscus, Linn. In the Cyperus bundle are two sheets, named in Linneus’s hand “annuus.” They are Cyperus fuscus, Linn. Scrrevs triernvs, Linn. Mant. p. 180. One sheet, marked “ Scirpus trigynus ” in Linnzus’s hand, is Ficrnra scariosa, Nees. Scrrpvus mutatvs, Sp. Pl. ed. 2, p. 71, n. 4. One sheet, marked “ Scirpus mutatus” in Linneus’s hand, is Eleocharis mutata. The description, citation, and plant of Scirpus mutatus, Linn., are ELrocHaRIS MUTATA, Je. Br. Scrrpus articunatus, Sp. Pl. ed. 1, p. 47, n. 1. One sheet, marked “ articulatus” in Linneus’s hand, is Scirpus articulatus, Linn. The diaguosis, the picture cited, and the authentic example are all ScrrPus ARTICULATUS, Linn. Scirpus patustris, Sp. Pl. ed. 1, p. 47, n. 2. Two sheets, viz. :— A, marked by Linneus “ palustris 2,” is Eleocharis palustris, R. Br. B, marked by Linneus “ palustris,” is Eleocharis multicaulis, Smith. The description of Linneus includes both plants. Scirpus palustris, Linn., therefore, is— partim = Eleocharis palustris, R. Br. partim = E. multicaulis, Smith. Scrrrus cenicutatvs, Sp. Pl. ed. 1, p. 48, n. 3. —— One sheet, marked “Scirpus geniculatus 3” in Linneus’s hand, is Eleocharis interstincta, R. Br. 310 MR. C. B. CLARKE ON CERTAIN Of the two figures of Sloane cited by Linnens, t. 81. fig. 3 is Eleocharis geniculata, R. Br.; t. 75. fig. 2 is not. Under Extrocuaris GENIcULATA, FR. Br., therefore, I cite Scirpus geniculatus, Linn. Sp. Pl. ed. 1, p. 48, partim. Scirpus cxsprrosvs, Sp. Pl. ed. 1, p. 48, n. 4. One sheet, marked “ cxespitosus ” in Linneus’s hand, is Sezrpus cespitosus, Linn. (An added sheet marked “4,” perhaps by Linneus’s hand, is Scirpus paucifiorus, Lightf. The figure cited, Scheuchz. Gram. t. 7. fig. 18, is S. cespitosus.] I make Scirpus cespitosus, Linn., therefore Goon. Scirpus capiratvs, Sp. Pl. ed. 1, p. 48, n. 5. Two sheets, viz. :— A, marked “ Scirpus capitatus 5” in Linneus’s hand, is E7eo- charis capitata, R. Br. B, marked “capitatus” in Linneus’s hand, is Eleocharis ovata, R. Br. Under Exzocuaris capitata, R. Br., I cite Scirpus capitatus, Linn. Scrrpus actcuLaRis, Sp. Pl. ed. 1, p. 48, n. 6. The plant is too young for me to determine; but may be Eleocharis acicularis, R. Br. Scrrpus Lacustris, Sp. Pl. ed. 1, p. 48, n. 8. Two sheets, viz. :— A, marked “8, lacustris’ in Linneus’s hand, is Scirpus lacustris, Linn. B, marked “ lacustris” in Linneus’s hand, is Scirpus pungens, Vahl. Scrreus Horoscuenvs, Sp. Pl. ed. 1, p- 49, n. 9. One sheet, marked by Linneus’s hand “9. Holoschcenus,”’ is Scirpus Holoschenus, Linn. Scirpus romanvs, Sp. Pl. ed. 1, p. 49, n. 10. One sheet, marked by Linneus’s hand “10. Romanus,” is Scirpus Holoschenus, Linn. Scrrpvs seracevs, Sp. Pl. ed. 1, p. 49, n. 11. Two sheets, viz. :— A, marked “ 11. setaceus” in Linneus’s hand, is Scirpus cernuus, Vahl. AUTHENTIC CYPERACEE OF LINNEUS. 311 B, marked “ Scirpus setaceus ” in Linnzus’s hand, is too young for certain determination, but appears to me to be also Scirpus cernuus, Vahl. The figure cited, Morison, sect. 8, t. 10. fig. 23, will do equally well for S. setaceus or S. cernuus. I think it is certain that Linneus would reckon S. setaceus and S. cernuus one species. Now that these are separated, I keep the name Sc. setaceus, Linn., for the plant. auctorum [¢f. Benth. Fl. Austral. vii. p. 327]. Scrreus suprnvs, Sp. Pl. ed. 1, p. 49, n. 12. One sheet, marked in Linnzus’s hand “12. supinus,” is Alisma ranunculoides, Liun. The description of Linneus will not apply at all to this Alisma. The species will therefore stand Scirpus suprnvs, Linn. Sp. Pl. p. 49, ed. 2, p. 73, neque herb. propr. Scrrpus AUTUMNALTIS, Linn. Mant. p. 180. One sheet, marked “ Scirpus autumnalis ” in Linneus’s hand, is FimprisTYLis AUTUMNALIS, Roem. et Sch., var. 3. geminata (sp.), Kunth. ScrrPUS CAPILLARIS, Sp. Pl. ed. 1, p. 49, n. 13. One sheet, marked ‘13. capillaris”’ in Linneus’s hand, is Bulbostylis capillaris, Kunth. Of the two pictures cited by Linneus, Burm. Zeyl. t. 47. fig. 2, is probably Bulbostylis capillaris; Pluk. Alm. t. 300. fig. 5, certainly not. Under Bunpostytis caprtuaris, Kunth, I cite Scirpus ca- pillaris, Linn. Sp. Pl. ed.1, p. 49 (tab. Pluk. citat. excl.). Scrrpvus tristacuyros, Linn. f. Suppl. p. 103. Two sheets, marked “ tristachyos”’ in Linnzus’s hand, are Frornra tristacuya, WVees (vix Kunth, who excludes Linn., but admits Rottb.). { Added is one sheet marked, not by Linneus, Scirpus trispi- catus, Linn. f. Suppl.; this is Fieinia sylvatica, Kunth. | Scrrets mucronatus, Sp. Pl. ed. 1, p. 50, n. 14. One sheet, marked “14. mucronatus” in Linneus’s hand, is Scirpus mucronatus, Liun. 312 MR. C. B. CLARKE ON CERTAIN Of the figures cited by Linneus, Pluk. Alm. t. 40. figg. 1, 3, is right; Morison, Hist. sect. 8, t. 10. fig. 20, is wrong. I cite the species as Scirpus mucronatus, Linn. Sp. Pl. & ed. 1, p. 50 (tab. Morison citat. excl.). Scrrrus Grossvus, Linn. f. Suppl. p. 104. Two sheets, marked “ grossus ” in Linneus’s hand, are Scirpus Grossus, Linn. f. Scirpus Ecutnatvus, Sp. Pl. ed. 1, p. 50, n. 16. One sheet, marked “16. echinatus”’ in Linnzus’s hand, is Mariscus ovutanis, Vahl. Linneus has united under the above name two plants: one, from Virginia, is Mariscus ovularis, Vahl; the other, from Ceylon, is Mariscus paniceus, Vahl. Sctrpus RETROFRACcTUS, Sp. Pl. ed. 1, p. 50, n. 17. One sheet, marked “17. retrofractus” in Linneus’s hand, is MarIscus RETROFRACTUS, Vahl. The figure cited by Linneus (Pluk. Amalth. t. 415. fig. 4) is right; though Plukenet says the plant came from Malabar, whereas Mariscus retrofractus is wholly North-American. ScIRPUS FERRUGINEUS, Sp. Pl. ed. 1, p. 50, n. 18. One sheet, marked “ Scirpus ferrugineus ” in Linneus’s hand, contains two plants, viz. :— 1. Fimbristylis ferruginea, Vahl. 2. Fimbristylis diphylla, Vahl. The figure cited, Sloane, t. 77. fig. 2, is right. I add Scirpus ferrugineus, Linn., as a synonym to Fimbristylis ferruginea, Vahl. Scirpus spapiceus, Sp. Pl. ed. 1, p. 51, n. 19. One sheet, marked “19. spadiceus”’? in Linneus’s hand, is FimMBRIsTYLISs sPADICFA, Vahl. The figure cited by Linnzus, Sloane, t. 76. fig. 2, is right. I make Scirpus spadiceus, Linn., a synonym of Fimbristylis spadicea, Vahl. Scirpus mintacevs, Sp. Pl. ed. 2, p.75, n. 22. One sheet, marked “ miliaceus” in the hand of Linneus, is FIMBRISTYLIS QUINQUANGULARIS, Aunth. Linneus would surely have esteemed 2imbristylis miliacea, Vahl, and F. quinquangularis, Kunth, one species. 1 doubt the AUTHENTIC CYPERACEE OF LINNZEUS. 313 advisability of keeping them distinct. I cite Scirpus miliaceus, Linn., as a synonym under Fimbristylis miliacea, Vahl. Scirpus CYPEROIDES, Linn. Mant. p. 181. One sheet, marked “Scirpus cyperoides ” in Linneus’s hand, is Mariscus Sirserianvs, Wees. Scirpus MaRiTIMvs, Sp. Pl. ed. 1, p- 51, n. 20. Two sheets, viz. :— A, marked “ 20. maritimus”’ in Linneus’s hand, is Scirpus MARITIMUS, Linn. B, marked “ maritimus singularis” in Linneus’s hand, and noted in the interleaf of his second edition, is §. maritimus, Linn., forma spicis elongatis. Scirpus maritimus, Linn., is (altogether) Goon. Scirpus Luzuya, Sp. Pl. ed. 2, p. 75, n. 23. One sheet, marked “Sc. Luzule’’ in Linneus’s hand, is Cyrrerus Luzuba, Retz. Linneus gives the habitat “India,” by which we must under- stand the West Indies. I cite Scirpus Luzule, Linn., as a synonym under Cyperus LuzuLa, Retz. Scrrpus sytvaticus, Sp. Pl. ed. 1, p. 51, n. 22. One sheet, marked “ 22. sylvaticus ” in Linneus’s hand, is Scirpus syLvaTicus, Linn. The figure cited by Linnzus, Loesel Pruss. t. 33, is right. Scirpus sylvaticus, Liun., is (altogether) Goon. Scrreus corymsosus, Linn. Amen. Acad. vol. iv. p. 303, et Sp. Pl. ed. 2, p. 76, n. 25. One sheet, marked “S. corymbosus”’ in Linneus’s hand, is Rynchospora aurea, Vahl. The description agrees, and 8. corymbosus, Linn., has been reduced by all modern authors to RyNcHosPora AUREA, Vahl. Scrrpus sguarrosus, Linn. Mant. p. 181. One sheet, marked “Scirpus squarrosus ” in Linneus’s hand is Scrrpus sguaRRosvus, Linn. Scirpus squarrosus, Linn., is (altogether) Goon. Scrrpus mvrricatos, Linn. Mant. p. 182. One sheet, marked first “Cyperus squarrosus” in Linnus’s LINN. JOURN.—BOTANY, VOL. XXX. Z 314 MR. C. Be CLARKE ON CERTAIN hand, but this struck out and “Scirpus intricatus ” substituted, also in Linnzus’s hand, is CYPERUS ARISTATUS, Jeoftb. Scirpus intricatus, Linn., is accordingly placed as a synonym under Cyprerts ARistatus, fottb. Scrrets ciLraris, Linn. Mant. p. 182. One sheet, two plants, marked “ Scirpus ciliaris ”’ in Linnzus’s hand, are FUIRENA GLOMERATA, Lam. A third plant on this sheet, separately marked in Linneus’s hand as “ Se. intricatus,” is so, i. e. Cyperus aristatus, Rottb. Scirpus ciliaris, Linn., is placed as a synonym under FurRENA GLOMERATA, Lam. Scirpus HoTTENTOTUS, Linn. Mant. p. 182. One sheet, marked “Scirpus hottentotus ” in Linneus’s hand, 18 Fourrena ureta, Vahl. Scirpus hottentotus, Linn., is placed as a synonym under Furrena nieta, Vahl. Scirpus anrarcticus, Linn. Mant. p. 181. There are two sheets, viz. :—A, marked “ Scirpus Antarcticus ” in Linneus’s hand, is [solepis seslerioides, Kunth, which Boeckeler regards (I think rightly) as only the full form of Jsolepis carti- laginea, R. Br.= Scirpus Bergianus, Spreng. The description of Linneus agrees: I therefore take up for this species, so abundant in the Southern Hemisphere, the name ScrRPus ANTARCTICUS, Linn. B, the second sheet, marked “ Scirpus Antarcticus”’ in Linneus’s hand, is Bulbostylis barbata, Kunth. This plant appears anciently to have been distributed as Se. antarcticus. Scirpus cepnatores, Sp. Pl. ed. 2, p. 76, n. 27. One sheet, marked “ Scirpus cephalotes ” in Linneus’s hand, is RyYNcHOSPORA CEPHALOTES, Vahl. The description and citation of P. Browne agree. I cite Scirpus cephalotes, Linn., as a synonym under RyNcHOSPORA CEPHALOTES, Vahl. ERroPHorUM VAGINATUM, Sp. Pl. ed. 1, p. 52, n. 1. One sheet, marked “ vaginatum ” in Linneus’s hand, is ERIOPHORUM VAGINATUM, Linn. The figure cited, Scheuchz. Agrost. t. 7. fig. 1, may do; but avails little. Eriophorum vaginatum, Linn., is (altogether) Goon. AUTHENTIC CYPERAVEE OF LINN EUS. 315 ERtoPHORUM POLYSTACHION, Sp. Pl. ed. 1, p. 52, n. 2. One sheet, marked “ polystachyon ” in Linneeus’s hand, is ER1OPHORUM ANGUSTIFOLIUM, Roth. The species of Linneus, EF. polystachion, has been divided into three very closely allied European species, of which E. angusti- folium, Roth, is the most plentiful. It is very difficult to maintain the distinction between the three on North-American material. Under these data, the proper plan would appear to be to retain the Linnean name £. polystachion for the mass of the species (Z. angustifolium, Roth), and to call the two sub- species E. gracile, Koch, and £. latifolium, Hoppe. Unfortu- nately a considerable number of botanists prefer to apply the name E. polystachion, Linn., to the subspecies EZ. latifolium, Hoppe. This being so, botanists appear agreed, by a very large majority, to drop the Linnean name; it cannot be applied to the prevalent form without the explanatory addition of E. angustifolium, Koch. ERIOPHORUM ViRGTNIcUM, Sp. Pl. ed. 1, p. 52, n. 3. One sheet, marked “ virginicum ” in Linneus’s hand, is ErtIoPHorvuM VIRGINICUM, Linn. The plates cited. Pluk. Alm. t. 299. fig. 4, and Morison, Hist. sect. 8, t. 9. fig. 2, are both right. Eriophorum virginicum, Linn., is (altogether) Goon. ERIOPHORUM ALPInum, Sp. Pl. ed. 1, p. 53, n. 4. One sheet, marked “ 4. alpinum ” in Linneus’s hand, is ERIOPHORUM ALPINUM, Linn. The plate cited, Scheuchz. Agrost. t. 7. fig. 4, may be right, but avails nothing. Eriophorum alpinum, Linvn., is (altogether) Goon. [In the discussion which followed the reading of this paper, I was convinced by Mr. Carruthers and Mr. Daydon Jackson that the common plan of citing ‘““Scnanvs, Linn. partim (or emend.)” must be retained; though the meaning cannot be understood without the addition ‘‘ Benth. Fl. Austral. vii. p. 352.” Also, Mr. Carruthers has since showed me the plant in the British Museum whence Sloane, Jamaica, t. 74. fig. 1, is taken; this plant is Cyperus ferax, L. C. Rich. (a species which, J fear, we are bound to eall Moriscus Pohlianus, Nees, if we place it in the genus Mariscus). Sloane’s plate is so badly engraved (from the drawing prepared for it) that T am not at all surprised that I was confident that it was not taken from Mariscus Pohlianus.— C. B. C., 17th Sept., 1894.] , Z 316 SIGNOR U. MARTELLI ON 'THE CAUSE OF On the Cause of the Fall of the Corolla in Verbascum. By Signor U. Marretut. (Communicated by the President.) [Read 7th December, 1893.] Ir is generally known that if a plant of Verbascum (it is pre- sumably the same with every species) be struck smartly, the corollas already expanded will, after a minute or two, fall like a shower of leaves. This curious phenomenon, though well known, does not seem to have attracted the notice of the older botanists ; and I do not know of any old author who mentions this subject. Even so accurate an investigator as P. A. Micheli has made no mention of it, either in his published work or manuscripts *. Sir J. E. Smith seems to have been the first to write about it, in his ‘ English Flora’ (1824), vol. i. p. 311, where, after having described Verbascum pulverulentum, Vill., he says:—“If the stem of V. pulverulentum be smartly struck three or four times with a stick, all the flowers then open will, in a few minutes, throw off their corolla, the calyx closing round the germen, so that after eight or ten minutes none will remain on the plant.” Charles Darwin, in his classical work, ‘Forms of Flowers’ (p. 78), refers to the phenomenon in the following terms :—‘‘ It is known that many species of Verbascum, when the stem is jarred or struck by a stick, cast off their flowers. This occurs with V. Thapsus, as I have repeatedly observed. The corolla first separates from its attachment, and then the sepals spontaneously bend inwards so as to clasp the ovarium, pushing off the corolla by their movement in the course of two or three minutes. Nothing of this kind takes place with young, barely expanded flowers. With Verbascum Lychnitis and, as I believe, V. pheniceum, the corolla is not cast off, however often and severely the stem may be struck.” My observations do not precisely correspond with those of Mr. Darwin. I have cultivated, and sought out wild in the country, many plants of Verbascum Lychnitis, on which I have continuously made experiments, and have always noticed that if you strike the stem, the corollas fall off as in Verbascum Thapsus and many other species. I asked several friends to make this experiment at different seasons of the year, and their experience has confirmed my opinion. * These lie still unpublished in the Florentine Botanical Museum. THE FALL OF THE COROLLA IN VERBASCUM. 317 Besides the above meagre information, I may mention that M. Maury* attributes the falling of the corollas to traumatic action, and considers it one of the means (only secondary, in my opinion) for assisting fertilization. Observations on this subject, therefore, are by no means numerous; and Darwin’s statement is the only one which illustrates the phenomenon. For this reason I have thought that it would be of interest to publish the observations which I have made during the last two summers in order to satisfy myself as to the cause of the fall of the corollas in these plants. The phenomenon, without doubt, is to be ascribed to irri- tability, which is transmitted by the vascular bundles in an ascending line till it reaches the flower. In fact, if we gently strike one of the branches in the middle part, only the flowers will fall which grow above the place which has received the blow. Sir J. E. Smith and Darwin noted that the perfectly open corollas fall a few minutes after the shock given to the plant ; and if this really happens, we must add that (at least in Ver- bascum Lychnitis, Blattaria, floccosum, sinuatum, garganicum, phlomotdes, crassifolium, and in several of their forms and varieties), if the experiment be made in the early morning—say between 5 a.m. and 8 a.m.—the corollas only just opened (that is to say, when the sun rises) remain on the plant apparently insensible; but if the experiment be repeated in the afternoon, or on the following morning, not one of those corollas will remain on the plant. It is therefore, I think, to be inferred that some time is necessary after the expansion of the flowers before the tissues of the corollas are in a favourable condition to permit the corollas falling off in this manner. I. Of the Tissues of the Corolla. We will first examine Verbascum Blattaria, a common plant which grows in open grassy places, selecting a fully expanded flower which has already been open for some hours; and to prevent the corolla detaching itself from its calyx, we imbed it in glycerine-soap ; we then cut a microscopical section of the part of the corolla which connects it to the thalamus. * Maury, “ Observations sur la Pollinisation et Fécondation des Verbauscum,’ Bull. Soc. Bot. France, 1885, p. 53 . 318 SIGNOR U. MARTELLI ON THE CAUSE OF The tissues of the corolla are formed, in the lower part, of ovate or oblong, large, rather irregularly shaped cells, which contain a few small grains of starch (fig. 1,@). The cells gradually become narrower and longer, and the presence of the starch augments in the cells of the short tube of the corolla. The thalamus is also composed of large round cells, though they are different in shape from those of the corolla (fig. 1,7). But where the corolla is inserted in the thalamus it is easy to observe a zone of a few rows of opaque, slightly yellow-coloured, and minutely granular cells (fig. 1, 6). These rows of cells constitute a stratum which dis- tinetly divides the tissue of the thalamus from the overlying one of the corolla. This stratum forms the separating zone, by the aid of which the corolla easily detaches itself from the thalamus. Also in the leaves an identical structure is found, which in like manner brings about their fall *. In the fully expanded flower the tissue of the corolla and of the thalamus which is nearest to the separating zone, when treated with chloriodide of zine, turns blue, while the separating zone assumes a more or less yellow tint, scarcely visible towards the * Mohl, H., “ Ueber die anatomischen Veriinderungen des Blattgelenkes, welche das Abfallen der Blitter herbeifiihren,” Botanische Zeitung, xviii. pp- 1-9 (1860). Van Tieghem et L. Guignard, ‘Observations sur le Mécanisme de la Chute des Feuilles,” Bull. Soc. Bot. France, 1882, p- 312. THE FALL OF THE COROLLA IN VERBASCUM. 319 centre. If we make use of iodine, the separating zone assumes a dark yellow colour. With fuchsin or iodine-green, even after many washings this tissue retains a beautiful red or green tint. On the other hand, if a section of a flower is treated with chloriodide of zine, the tissue of the corolla assumes a rather deep blue colour, the separating zone a light yellow tending to a green tint, and the underlying tissue of the thalamus remains uncoloured. The severance of the corolla from the thalamus occurs between the last row of cells of the separating zone and those of the corolla (fig. 1, 6). Therefore the stratum of smali cells, of which we have already spoken, remains adherent to the thalamus, as will be easily seen by dividing the two parts—that is, the corolla from the calyx—or by examining the thalamus just when the corolla has fallen off; the adherent tissue, which formed a part of the separating zone, dwindles and gradually disappears almost entirely. We will now refer to its antecedent condition, and take into consideration a flower still in the bud. The separating zone at this epoch is already visible, and the tissue appears opaque and of a yellowish colour: the chemical reactions obtained have already been noticed above. As to size, the cells of the separating zones are very little different from those of the corolla and of the thalamus, though smaller, but they have a pentagonal shape. Little by little, as the flower develops, the tissue of the separating zone assumes more of its character, the cells subdivide, their shape is modified, becoming gradually rounder in form. This variation of shape is, I consider, very important in the phenomenon of the fall of the corolla which we are considering, because, owing to this change, the corolla acquires less power of cohesion to the thalamus, as I will hereafter explain. Il. On the Fall of the Corolla. The power of resistance possessed by the flower when only in the bud-stage, or but just expanded, is, in my opinion, explained by the cohesion of the cells of the tissue of the separating zone. In the flower still in bud these cells are pentagonal, and for this reason they form a united and compact tissue, which offers more resistance; on the other hand, in the flower which has been open some time the cells become smaller and rounder, in consequence 320 SIGNOR U. MARTELLI ON THE CAUSE OF of successive changes, for which reason the continuity and the resistance of this tissue diminishes. Thus, under these circumstances, even a very little effort will occasion the separation of the cells. So that, in conse- quence of the irritating action caused by a stroke given to the plant, the sepals at once rise up, becoming connivent, and by this movement they press against the corolla, which, by the decreased cohesion of the cells of the separating zone, easily separates and is thrown off. But, to explain this phenomenon more clearly, let us divide it into two periods, and consider the sepals as consisting of two parts—the base and the lamina. In the first period it is the base of the sepal which acts, and closes against the corolla in such a way that it compresses the base of the corolla, and obliges it to slide forward, and thereby separates it from the thalamus. By paying close attention to this period, we perceive in the corolla at first a slow forward movement, which probably corresponds to the separation of the cells of the separating zone, and in the second place a slighter, but at the same time abrupt movement of the corolla, which certainly happens when the vascular bundles break. During the second period the lamina of the sepal rises, with a more or less rapid movement, and, leaning against the corolla and compressing it, pushes it continually forward until, owing to its weight, it gradually falls off. Then the sepals continue the con- verging movement till they completely enclose the ovary *. In the same way, the base of the small tube of the corolla seems to me to have a structure which facilitates the expulsive action of the calyx. This is the way that Verbascum Blattaria is formed :—In the lower and external part of the tube of the corolla we notice an unexpected swelling, with a pentagonal outline, almost as if it were a protruding ring (fig. 2, a). From the external border of this ring to the circular orifice of the corolla, the very short part of the tube (fig. 2, 6) is very smooth, * In some species, as for instance in Verbascuim floccosum and Verbascum sinuatum, when the corolla has fallen, the sepals become at once connivent with a rapid movement. In other species, as for instance in Verbascum Blat- taria, this movement is slower, and some time is necessary before the sepals assume their final decisive position—that is to say, perfectly connivent. I have also severed the sepals, so as to ascertain their expulsive action; but I cannot add anything to what I have already said, unless it be that they have confirmed me still more in the opinions expressed. THE FALL OF THE COROLLA IN VERBASCUM. 321 and appears as if varnished. Owing to the above-mentioned phenomenon of irritability produced by a stroke given to the plant, the base of the sepals, contracting itself as we have already stated, presses against the annular projection; and, owing to its inclined position, the corolla is pushed forward and forced from the thalamus. In Verbascum sinuatum, as in some other species, the annular projection does not exist, but instead, towards the base, the tube of the corolla is rather swollen (fig. 3, c), and thence to as far as the orifice it ends in an inclined plane. Fig. 3. Besides this special structure of the base of the corolla, we must not leave unnoticed another kind of structure which is very important—namely, glabrescence or smoothness (almost amounting to polish)—which the lower and external part of the base of the corolla, covered by the sepals, presents to us. All this greatly facilitates the sliding of the sepals on the sides * 322 ON THE FALL OF THE COROLLA IN VERBASCUM. of the corolla, which is generally covered with hairs, often plentiful and sometimes stellate. Having set forth these facts, it is easy to see that my obser- vations and conclusions do not entirely agree with those expressed by Darwin*. He is of opinion that the corolla first separates itself from the thalamus, and that afterwards the sepals push it off. This opinion would admit of a spontaneous action of the corolla in detaching itself from the thalamus; while I believe that the fall of this organ is subject to the effort exercised by the sepals, and especially by the basal part of them. The separating zone, which is not mentioned by Darwin, has not, I believe, at this moment and in this case, any other action than that of diminishing the resistance of the tissue. It may, however, have another function, namely, that of assisting the normal detachment of the corolla from the thalamus. I would further add that, in order that the irritability of the sepals should produce its effect on the corolla, it is necessary that the tissues of the flowers should be in a state of turgescence. In fact, if we place some cut branches of Verbascum in water, during the first and second day we may notice that its flowers are always sensitive to shocks, but that, little by little, this sensitiveness diminishes, and the new flowers do not suffer from any blow aimed at the branch. In conclusion I may state that I should liked to have ex- amined more species of Verbascum, so as to ascertain whether the phenomenon is common to all the species, as I believe to be the case, but have not had opportunities of so doing. The species on which I have experimented are:—Verbascum Blattaria, Cedreti, crassifolium, floccosum, garganicum, Lychnitis, longi- folium, macrurum, montanum, nigrum, niveum, orientale, phant- ceum, phlomoides, pulverulentum, sinuatum, Thapsus, thapsiforme, virgatum, and many of their forms and varieties. In all these I have noticed the same phenomenon. If even other species do not obey this law, the anatomical study of these would certainly be interesting, and might serve as a guide and a comparison for later and more decisive results. * Darwin, ‘ Forms of Flowers,’ p. 78. ON THE HABITS OF SPECIES OF LEMN2. 323 On the Habits of Lemna minor, L. gibba, and L. polyrrhiza. By H. B. Guppy, M.B. (Communicated by W. Borrine Hemstey, F.R.S., A.L.S.) [Read 3rd May, 1894.] AFTER three years’ systematic observations of these plants *, I find that in this branch of the subject my notes add but little to what Hegelmaier recorded about a quarter of a century ago in his general work on the Lemnacew. Though that work is con- cerned chiefly with their structural characters, it cannot easily be supplemented now without a great amount of observation and experiment based on new lines of research. These plants at first sight seem well fitted for an inquiry into their adaptation to the conditions of their existence; and Professor Miall, in an interesting lecture at the Royal Insti- tution in March, 1892, treated of some of their relations to the physical characters of water. Eminently suited as the duck- weeds are for aquatic life, it is a little perplexing that they will thrive just as well on wet mud. For twenty months I have cultivated Lemna minor on wet mud, where it throve and budded all the seasons of the year. To give an instance, two plants set apart in October had increased to twenty by April, and to fifty by the next October. Nor has this long exposure to different habits of life produced any per- manent change in the characters of the plants so far as their external appearance is concerned. After a culture of twenty months on the mud the plants still float on water and retain their root-caps, even in cases where the roots have penetrated a line or two into the soft mud. The winter fronds of Lemna polyrrhiza, also, wheu placed on wet mud in the spring, bud freely, and in the course of a few weeks produce the summer form of the plant. Duckweed stranded on the mud during the drought of the summer of 1893 remained alive most of the season wherever the earth around the ditch or pond remained moist. Lemna minor is to be observed flourishing on the wet soil and oozy ground in the vicinity of springs some feet above the level of the pool beneath; and here it can safely withstand the longest * In the South-east of England, in Epping Forest and around Kingston, and in the rivers Thames and Lea. 324 MR. H. B. GUPPY ON THE drought. It is conceivable that during changes of climate the Lemnw might long find a refuge in the vicinity of perennial springs. Coming to the individual species, I will first refer to Lemna gibba. The plants with gibbous fronds as a rule die long before the winter, and new plants similar in form appear in April or May. Now and then, however, in mid-winter, a few healthy gibbous plants are tv be found in ditches, ponds, and rivers, which bud freely in the spring. A number of plants, kept through the winter of 1892-93 in a room without artificial heat, retained their buoyancy and gibbous form, and were quite healthy when they budded in the spring. The temperature of the water was for some weeks under 40° Fahr., and for ten consecutive days it was under 36°, the plants being inclosed in ice for a day or two. Evidently, therefore, low temperature is not of itself injurious to the gibbous plants; and since, as shewn below, they die in numbers iv the tepid water of August after flowering in July, their death before winter is due to the exhaustion of their vitality and not to the fall of temperature. Assuming that this plant is sometimes able and at other times unable to float through the winter in the gibbous form, it becomes a question, when the plants disappear in the autumn, whether the new plants in the spring are produced from seed, or whether, as in the case of Lemna polyrrhiza, there is a special winter form. Hegelmaier inclined to the latter view in the text of his work (page 64), and although the winter form of the plant had, in a sense, escaped his observation, in his supplement he welcomed the discovery of Dr. van Horen in this respect. The following notes bear on this matter :— In the first week of July, 1893, I found in the almost stagnant water of a ditch, connecting the Diana Pond with the other large ponds in Bushey Park, a truly enormous quantity of Lemna gibba, abundautly in flower, and mixed with a large amount of separate thin flat fronds, just like those of Lemna minor, and also in flower. Mingled with them were a great number of the plants of Lemna polyrrhiza, not in flower, but producing the winter fronds to an extent I have never seen before or since. Beneath the floating mass of duckweed, which was two or three inches thick, flourished Lemna trisulca, but not in flower. The temperature of the surface of the ditch in the afternoon of Juiy 7th was 87° Fahr., and at the bottom, eight inches down, HABITS OF SPECIES OF LEMNZ. 325 it was 82°. From the run of my series of freshwater tempe- ratures, then in progress, it is to be inferred that during the few weeks preceding the flowering, the surface must have been frequently heated to 80° in the daytime, the nocturnal fall of temperature being about half that of the air. Such were the thermal conditions under which Lemna gibba flowered. Under a lens the gibbous plants had the appearance of pro- ducing female flowers, and the separate thin flat fronds male flowers, which was due to the circumstance of the pistil alone protruding in the one case, and of the stamens, usually two in number, alone showing in the other, although in both instances the flowers were hermaphrodite. After several visits to the ditch, only one or two rare exceptions to this rule were found, namely, where the gibbous plants had evolved stamens. It seemed, therefore, as if the flowers were functionally unisexual. However, no plants in the fruiting-stage came subsequently to my notice. Under observation at home, the flat fronds were produced by budding by the gibbous plants, the process being rapid, and their stamens visible through their substance before they floated away. The flowering period covered the greater part of July. During August the gibbous plants produced abundautly thin and flat fronds, which were now flowerless and of a dark green hue. These new fronds presented a great contrast in colour to that of the mother plants, had in each case a single root, and showed the three distinct nerves which, according to Hegelmaier, characterize the flatter forms of Lemna gibba. Such proved to be the winter plants, and, asin the case of Lemna polyrrhiza, they were extensively produced in summer. Their appearance in August was accompanied by the death of a large number of the gibbous mother plants, both in the ditch and in my vessels at home, a result which certainly could not have been the effect ot low temperature. The remainder of the gibbous plants pro- ceeded actively with the development of winter fronds during September and October, and it was not until early in November that the characteristic forms of the plant had disappeared. At this time their dark green flat offspring were budding freely. both in the ditch and at home, some at the surface and some at the bottom; but those afloat could always be sunk by a touch; and, in fact, it was not until March that the winter plants could float independently, though many of them remained at the 326 MR. H. B. GUPPY ON THE surface owing to their power of repelling water after being floated up by bubbles. As the winter advanced, the fronds lost their dark hue and the distinctness of their nerves; and after the beginning of the year the plants could not, in general form and appearance, be distinguished from those of Lemna minor. Budding, which had been suspended during the winter, began briskly early in February, similar fronds being produced; but up to the fourth week of April they exhibited only a slight sponginess in appearance, and would have been taken for vigorous forms of Lemna minor. Up to the time of writing, it was evident from the temperature observations made at the ditch and in my vessels, that for the development of the gibbous character the winter plants require, for a week or more, an average daily maximum temperature at the surface of the water of not much, if at all, under 70° Fahr. It would seem from the above that the form in which Lemna gibba passes the winter in this climate depends a good deal on the thermal conditions of the locality it frequents in summer. Where the water is cool and there is little or no surface-heating, as in a spring-fed brook or in a river, the gibbous plants would never flower and probably also would not develop special winter plants, but would float unimpaired through the winter. In the warm and stagnant water of a shallow ditch the plants would most often flower; but the flowering process, by exhausting their vitality, would result in the intervention of a peculiar winter form. That the intervention of a special winter form depends indirectly on a high summer temperature seems a little paradoxical; but my temperature observations indicate that there is much to learn about the habits of this plant. Lemna polyrrhiza, like Lemna gibba, has a winter form, but the resemblance there ends. Whilst with Lemna gibba the winter fronds, after detachment, bud in the autumn and pass the winter as complete plants, which attain in the spring the peculiar gibbous form, with Lemna polyrrhiza the winter frond remains solitary and produces, by budding, the summer plant without ultimately becoming incorporated with it. When the winter trond of Lemna polyrrhiza has performed its office, it becomes sickly, sinks, and dies. The description of the solitary, dark green, reniform, and rootless fronds that enable Lemna polyrrhiza to pass the winter in this climate without seeding, will be found in Hegelmaier’s HABITS OF SPECIES OF LEMNZ. 327 work, As pointed out there, though destined for the winter, they are largely produced in summer, and, as before remarked, they never came under my notice growing in such numbers from the mother plant as in the first week of July in the tepid water of a stagnant ditch where Lemna gibba was in flower. This budding process goes on until about the end of October, when the summer plants all die. My temperature observations during the autumns of 1892 and 1893 on the Thames above Kingston and on several ponds in that neighbourhood, all plainly indicate that when the surface temperature in the day falls short of 50° Fahr. rapid decay and death of the summer plants ensue. The winter fronds of Lemna polyrrhiza, when first detached, nearly always sink; but occasionally they cannot be sunk, such fronds being rather thicker than the average. Whilst some pass the winter afloat, others spend it at the bottom. Hegelmaier speaks of them as all remaining under water; but they vary much in this respect with the character of the weather at the time of their growth, and with the other conditions under which they are developed. They are to be found constantly in winter amongst the floating seed-drift of the rivers Lea and Thames, and of the ponds around Kingston ; and through the past winter they floated in great numbers on one of the Home Park ponds. In a fine calm autumn with plenty of sunshine, the sunken winter fronds in the shallow water are very liable to be floated up by air-bubbles, and they acquire independent floating powers at the surface. In such weather, also, the mother plants die slowly, and their winter offspring remain sufficiently long attached to be able to float of themselves. On the other hand, in rainy and windy autumns the disturbance of the surface of the water brings about the detachment of the winter fronds before they can float, and they all go to the bottom. During two winters I have kept under observation both the winter fronds that could not be sunk, and those that remained at the bottom. Various agencies assist in floating up the sunken fronds in spring. When exposed in a clean glass vessel to moderate sunshine, they are carried up by bubbles of their own. forming, and, throwing off the water, they emerge at the surface. For the first few days they can always be sunk at a touch, but after that they float of themselves. Without the sun, these fronds usually remain at the bottom, become sickly, and die. Those that bave remained under water all the winter are able to 328 MR. H. B. GUPPY ON THE throw off the water from the first, and do not acquire this power at the surface. If there is any impediment to rising, the sunken fronds begin to bud at the bottom, and through the buoyancy thus obtained they sooner or later appear on the top. Usually, however, the sunken winter frond is independent of its own efforts in floating up. The bubbles, evolved in quantity through the excitement of vegetable activity on the mud of the pond, carry them up in numbers, and once at the surface their power of throwing off the water keeps them afloat, and in calm dry weather they will, in a few days, acquire independent buoyancy. The time when the winter fronds of Lemna polyrrhiza begin active budding is generally in April; but this is mainly deter- mined by the temperature of the water, and though the amount of sunshine received is the same, the fronds in the shallows will always be more advanced in the growth of the summer plants than those in the cooler and deeper water. In considering the temperature at which the winter fronds begin to bud into the summer plants, it is as well to remember that an estimate of the initial temperature of budding or of ger- mination of a water plant, to be of any practical value, should be brought into some kind of relation with the thermal régime of a ditch, pond, or river *. This subject can only be touched upon here. Rivers, on account of their lack of surface-heating, their small daily range of only one or two degrees (Fahr.), and their comparatively uniform rise of temperature with the advance of spring, afford very valuable opportunities of ascertaining the temperature of budding or germination of floating plants or seeds. Here a single forenoon observation gives the day’s mean tempe- rature. With ponds we get the disturbing influence of daily range, not very marked it is true; in the centre of a pond it would be only five or six degrees, but considerable at the margin, where it would amount to some twelve or more degrees. A comparison of the daily mean temperatures of a river and a pond is, therefore, theoretically open to objection; but in practice inconsistent results are not obtained. From my Thames observations I estimated the initial tempe- rature of active budding in the river, in the case of the winter fronds of Lemna polyrrhiza, at about 55° Fahr., sometimes more * Tam dealing with this subject in a series of papers, of which the first was read before the Royal Physical Society of Edinburgh in January. HABITS OF SPECIES OF LEMNE. 329 and sometimes less, this being the mean temperature of the week preceding the budding. Imitating at home the thermal con- ditions of the central parts of a pond with a daily range of five or six degrees, my estimates of two different years were 51° and 53°, Observations in the shallow water of a pond, where the range is more than twice as large, gave me estimates of between 53° and 54° as the mean temperature of the preceding week. It will thus be seen that the results obtained are much the same whether in the shallows or centre of a pond, or in a river. Since, however, it is at the margins of ponds that these winter fronds gather in the spring, I will deal more particularly with this part of a pond, not only as regards the daily mean tempe- ratures, but also with regard to the run of the daily maximum temperatures. When, as is usual at the beginning of April, the temperature of the water rises to 60° in the day, and the daily mean is 53° or 54°, the winter fronds commence to bud briskly. When the daily maximum rises to 65°, the summer plants are half developed. When it rises usually to 70°, as happens in May, the fully-formed plants abound: this represents a daily mean of 62° or 63°, and it is noteworthy that it is not until the temperature of the river Thames reaches 60° that the summer plants are found on its waters, the river lending its testimony to the evidence of the pond. Coming to Lemna minor, my notes contain nothing that in any way supplements Hegelmaier’s description of their habits. During three winters, both at home and in the ponds, they have floated to the spring. In three successive springs in my green- house, they began to bud many weeks before the winter fronds of Lemna polyrrhiza, in the same vessel, showed signs of com- mencing. At the margin of a pond they begin to bud actively when the mean maximum temperature of the water for a week reaches 45°, which represents a mean temperature of about 40° for that period. The seeds came under my notice in February 1891 and 1894, floating amongst the seed-drift of ditches, ponds, and rivers, sub- sequently germinating at the surface, and producing the typical plant. On February 2nd of this year there were a few seeds germinating in a pond where I was recording the temperature, an observation which enables me to say that the thermal con- ditions requisite for germination are the same as above given for the budding of the fronds in early spring. The budding and germination were going on side by side in the same pond. LINN. JOURN.—BOTANY, VOL. XXX. 2a 830 ON THE HABITS OF SPECIES OF LEMN&. Estimates of Pond and Ditch Temperature for the week preceding Germination, etc. (Fahr. scale). Obs. Species. Date. Mean Temp. aioe Germination, | L. minor *. Early in Febr. 40° 45° Budding in | 7. minor, Early in Febr. 40° 45° spring. ” L. gibba. Febr. 45-46° 51° » L. polyrrhiza. | Early in April. 52-53° 60° Full develop-| L. gibba. May. 61-62° 69-70° ment of the summer plant.) L. polyrrhiza, | May. 62-638° 70° Flowering. L, minor, June. 65° 72-73° » L, gibba. July. 73° 80° Death of the - End of October or summer plant. L, polyrrhiza. early in Noy. a Under 50° * In a previous spring the seeds of Lemna minor germinated at the latter part of February, but no temperatures were recorded. Mr. W. Wilson (Hooker’s Botan. Miscell. vol. i. p. 145) regards April as the usual time for the germination of the seeds of Lemna gibba, but he experimented on some which had been kept dry throughout the winter. Nore, Sept. 12th, 1894.—I have since learned that Micheli described the flowering of Lemna gibba in language similar to my own (Journ. de Phys. tome 82, p. 107). This plant has behaved very differently during the cool and unsettled weather of the past summer. The gibbous fronds did not appear until early in July, and very few flowered, showing only the pistil, as before. No winter flat fronds have been developed. A recent experi- ment has proved that Lemna polyrrhiza is able to go through all its stages of growth on wet mud. Winter fronds placed on wet mud in March gave rise to the characteristic summer plauts, which in their turn are now producing the winter fronds.— H. B. G. MR. RICHARD SPRUCE’S HEPATICE ELLIOTTIANE. 301 Hepatic® ELuiott1an”#, insulis Antillanis S% Vincentii et Dominica a clar. W. R. Elliott, aunis 1891-92, lcte, Ricarvo Spruce determinate. [Read 21st December, 1893.] (Puates XX.-XXX.) THE new species in the following list of the Hepaties collected by Mr. W. R. Elliott in the British West Indies during the years 1891 and 1892, on behalf of the Joint Committee of the Royal Society and the British Association for the Exploration of the Natural History of the West Indies, seem not to have been described before. In some of the finest and most con- spicuous species of his gathering he had been anticipated by Eggers, Wallis, and others, who had previously found them in other isles of the Antilles, or on the mainland of S. America, in Venezuela and New Granada. Good descriptions of them have been given by Herr Stephani in the pages of ‘ Hedwigia’ for the years 1888 and 1892. Those here described are most!y minute species and include no novel types, but are some of them interesting as adding new forms to groups of which only a few species were previously known. Since the publication of Swartz’s ‘ Flora Indie Occidentalis,’ vol. iii. (1806), where several West-Indian hepatics were described for the first time, and often very incompletely, with some confusion of allied forms under the same name, there has been no attempt at a complete hepatic flora of the whole group. The only island which has been systematically explored for hepatics is Porto Rico, where the Kunstgirtner C. Schwanecke gathered them during three years (1847-50); and his specimens, comprising 58 species, were published by Hampe and Gottsche in ‘Linnea’ for 1853 (pp. 333-357). Nearly 40 years later (1885-7) Sinteuis added largely to this total, and Stephani’s enumeration of the entire hepatic flora amounts to 111 species (‘ Hedwigia,’ 1888, pp. 276-302). In 1845 Montagne described the few Cuban hepatics known to him in Ramon de la Sagra’s ‘ Historia fisica. ...de la Isla de Cuba,’ vol. ix. 1845, and was able to give for the first time a figure of the perfect plant of the curious Cyathodium cavernarum. I also gathered this plant, but sterile, in a cave at the foot of the volcano Tunguragua in Ecuador, but inadvertently omitted it from my enumeration, ‘ Hepatic Amazonice et Andine,” in Trans. Bot. Soc. Edinb. xv. (1884). [At this point the MS. of Mr. Spruce’s preface abruptly enda, LINN. JOURN.—BOTANY, VOL. XxX. 24 3832 MR. RICHARD SPRUCE’S HEPATICE ELLIOTTIANE. and within two or three days after he wrote these last words he was taken from us. He intended no doubt to cite the chief papers that treat of the Hepatics of the West Indies; and for the benefit of those who are interested in the paper the following notes are added. Montagne’s Cuban list contained 32 species; and some years later additions to it were found among the hepatics gathered by the American traveller Charles Wright, which were named by Gottsche and distributed without numbers by Asa Gray after the collector’s death in 1885. From Jamaica 38 species are recorded in H. Boswell’s list in the ‘ Journal of Botany,’ 1887, p- 49; and 4 more occur in a paper by Lindenberg and Hampe in ‘ Linnea,’ 1851, pp. 300, 640. From San Domingo 34 species gathered by Eggers are given by Stephani in ‘ Hedwigia,’ 1888, p- 299, and in the same paper 6 are mentioned as coming from Dominica, 3 of which do not occur in Mr. Elliott’s collections. The French Antilles—Guadeloupe and Martinique—which are the neighbouring islands to Dominica, have been searched in late years and, in M. Bescherelle’s recent list in ‘ Journal de Botanique,’ 1893, pp. 174, 188, present a total of 148 species ; but they will probably prove more fruitful than this, for their united area—about 1100 sq. miles—is more than double that of Dominica and St. Vincent, which is only 415 sq. miles and yet has produced, as will be seen in the following list, 162 species, of which 80 are new, and in addition to which several other species are likely to be found by future collectors. The results which have been attained in other islands, such as Trinidad, Barbados, and St. Kitts, must be sought in the ‘Synopsis Hepaticarum,’ 1844-7, and in the series of monographs of genera which are being produced with such industry by Herr Stephani. Mr. Spruce’s paper has been recast so as to bring the list of species, the descriptions, and the notes into one series. The author left a number of short notes which have been expanded ; but a few which were so brief as to defy interpretation have been omitted. A few extracts from his letters have been added. The drawing of the plates has been supervised, and an explanation of the figures prepared. The numbers and localities have been carefully collated with Mr. Elliott’s collections and corrected where necessary, and two species which Mr. Spruce had identified but omitted to put in his list have been included. The references to published descriptions have been looked up aud inserted.— A. Gepp. | —— “Ty MR. RICHARD SPRUCE’S HEPATICE ELLIOTTIANA, 333 Subordo I. JUNGERMANNIACE. Tribus 1. JUBULEZ. Genus 1. Frunvanta, Raddi. Four species have been recorded from the French Antilles ; and of these only one appears in Mr. Elliott’s collections. This is F. atrata, a not uncommon plant in the mountains of nearly the whole of tropical America. In the Andes I have seen it growing in huge masses half-a-yard long. 1. F. (§ Cnonanruerta) arretina, Tayl. in Syn. Hep. p. 413. St. Vincent, Richmond Peak (215). Dominica, River Douce Valley (704), Shawford Estate (898). 2. F. ($ Merroriorsts) arrata, ees, in Syn. Hep. p. 463.-— Dominica, Morne Trois Pitons (491), Morne Diablotin (648, 688). Var.— St. Vincent, Richmoud Peak (215). Pendulous from trees. 3. EF. (§ Tuyopste.ta) spATULIFLORA, Spruce,n.sp. (Tab. XX. figs. 1-4.) Monoica, pusilla, prostrata, rufo-badia, caule pollicari vage sat ramoso. olia fragilia, subplana, subdivergentia, inferiora contigua, ramia imbricata, oblique subsemivvata, rotundata, raro abrupte subacuta; cellul@ preminute, subpellucide, equaliter incrassate, trigonis vix ullis; Jobudus folio triplo fere brevior, cauli parallelus et subcontiguus, breviter conico-cylindricus, obtusus. Foliola foliis plus duplo breviora, pallida, tenera, appressa, oblonga, plana, ultra + fissa, segmentis obtusis. Flores 2 in caule et ramis terminales; bractee foliis dimidio longiores, profunde complicato-bilobe, lobo antico oblique oblongo-rotun- dato, postico subminore et angustiore, lanceolato; bracteola magna, cuneato-obovata, ad 3 biloba, complicato-carinata, lobis lanceolatis conniventibus (arte expansis divergentibus). Per:- anthia emersa, compresso-trigona, spatulata, supra basin obconicam fere quadrata, apice lato retuso vel subobcordato brevirostria. Andrecia brevia, paucijuga.— Folia ‘4 x *8—"35, lobulus .15, celiule =l;; foliola 18X12; bractee (lobus) ‘6 x ‘35, bracteola *8 x °65 ; perianthia 1°05 X°8 mm, Hab. Ad cortices in situ Laudat, ins. Dominica (Ettiort, 1892).—Florescentia monoica et stipulis planis, no. 857, a. precipuetamen forma perianthii singular, a plerisque Thyopsiellis distat. oe 2B2 -_ BEY MR. RICHARD SPRUCE’S HEPATICEH ELLIOTTIAN ¥. 4. Frurpanta (§ DrastaLoBa) CAULISEQUA, Nees, in Syn. Hep. p- 448.—St. Vincent, Wallilaboo Valley (177). Dominica, Morne Anglais (492). On bark. 5. F. (§ Drastratopa) supritissima, Lindenb. in Syn. Hep. p. 443.—St. Vincent, Richmond Peak (208). On bark. Genus 2. Lrseunea, Lib. There are 66 species in Mr. Elliott’s collections ; whereas 42 have been recorded from the French Antilles. Subgenus 1. Ompnatantuus, Lindenb. et Nees. 6. LEJEUNEA FILTFORMIS, Mont. in D’ Orb. Voy. Am. Mér., Bot. Crypt. p. 64; Omphalanthus filiformis, Nees, in Syn. Hep. p- 3804.—Dominica, Roseau Valley (62), creeping on Plagio- chila; Morne Micotrin (66), sine loc. (516). Subgenus 2. SricroLEsEUNEA, Spruce. . L. squamata, Nees, in Syn. Hep. p. 322.—St. Vincent, Bon- homme Valley (93). Dominica, Morne Couronne (225). 7 Subgenus 3. Nevroneseunga, Spruce. 8. L. porroricensis, Hampe et Gottsche, in Linnea, xxv. (1852), p. 852.—Frustulum in Plagiochila carpsi. 9. L. Breurenit, Gottsche, in Syn. Hep. p. 324.—Dominica, Roseau Valley (80). Subgenus 4. Brroprerts, Nees. 10. L. tonatsprca, Spruce, in Trans. Bot. Soc. Edinb. xv. p- 114.—Dominica, Morne’ Diablotin (642). Subgenus 5. AcRoLEJEUNEA, Spruce. 11. L. rorutosa, Spruce, in Trans. Bot. Soc. Edinb. xv. p. 117; Phragmicoma torulosa, Lehm. & Lindenb. in Syn. Hep. . p. 294.—St. Vincent, Fancy farm (343).—[Mr. Spruce made a note in his MS. of the four species Acrolejeunea atroviridis, torulosa, juliformis, and polycarpa; but whether because he considered them all closely allied, I am unable to say. He told me in a letter that one of the main differences of A. atro- viridis from its nearest ally, A. torulosa, is that the perianth of the latter has 7 to 9 plice or keels, whereas that of A. atroviridis has only 4—2 marginal, 2 postical—A. G.] «> MR. RICHARD SPRUCE’S HEPATICE ELLIOTTIANE. 335 12. LEsEUNEA (§ ACROLEJEUNEA) ATROVIRIDIS, Spruce, n. sp. (Tab. XX. figs. 5-10.) Dioica, pusilla, late cxspitosa, sordide viridis, nigricans, in etate sepe aterrima; caules {-3-pollicares, e caudice procum- bente ramoso orti, simplices vel parum ramosi, equifoliati. Folia imbricata, patentia, subdivergentia, supra basin saccatam trans- verse oblonga rotundata valde recurvo-concava, medium versus sinuato-complicata, lobulus plus duplo brevior, turgidus, quasi- concheformis, margine tamen planus, apice rectangulus, sepe reclusus; cellule parvule, sabeonformes, trigonis nullis mini- misve. oliola plus triplo brevivra, dissita, late cuneato- orbieulata renifurmiave, tenuissima, cellulis minoribus sub- quadratis rectangulisve. lores in caule, rarius in ramo, terminales ; bractee foliis paulo longiores, plane, oblique ovales obtusata, lobulo perparvo plano liueari vel anguste triangulari basi aucte; bracteola vix brevior late obovata rotundato- truncata, libera. Perianthia emersa, pyriformia, sat compressa, anticesubdepressa, postice altiuscule bicarinata, margine carinisque repando-scaberula.— Folia ‘7 x°6 (bast) x °5 (ad lobuli apicem), lobulus °3, cellule 3; ; foliola *2 x °3-"35 ; bractee ‘8 X°5, bracteola 75X6; perianthia 13x°8 mm. Hab. lus. Sti. Vincentii, Richmond Valley, a1 rivuli exsiceati Japides (Enutorr, no. 219, Febr. 1892)—Ab Acrolejeunea torulosa, L. et L., affini, satis diversa, peria.thio precipue soium 4-(nee 7-9-)plicato. Subgenus 6. LopwoLesevnea, Spruce. 13. L. Sacrwana, Mont. in Syn. Hep. p. 314.—St. Vincent, Richmond Valley (212). Subgenus 7. Pratyitseunna, Spruce. 14. L. araxucata, Nees, in Syn. Hep. p. 311.—St. Vincent, Rich- mond Valley (221). 15. L. vincentina, Gottsche, in Syn. Hep. p. 313.—Dominica, Morne Diablotin (674).—This species has not been noticed among Mr. Elliott’s collection from St. Vincent 5 por has another species—L. Lobsoniana, Lindenb.—which was gathered on Mt. St. Andrews by Hobson in 1822. Stephani indicates L. transversalis among Eggers’s gatherings in Dominica (‘ Hedwigia,’ 1855). 336 MR. RICHARD SPRUCE’S HEPATIC.E ELLIOTTIAN#E. Subgenus 8. ARCHILEJEUNEA, Spruce. 16. LesevNEA xantHocarpa, Lehm. et Lindenb. in Syn. Hep. p- 330.—Dominica, Morne Trois Pitons (70), Morne Couronne (231), River Douce Valley (710).—This plant, which is absent from equatorial America, occurs in Guadeloupe, Bolivia, and Africa. Subgenus 9. BracurioLEseuNnEa, Spruce. 17. L. srconor, Mont. in D’Orb. Voy. Am. Mér., Bot. Crypt. p- 349 ; Phragmicoma bicolor, Nees, in Syn. Hep. p. 294.—-St. Vincent, Fancy farm (343). Subgenus 10. HomatoLEsEunEa, Spruce. 18. L. pracurata, Nees, in Syn. Hep. p. 313; Phragmicoma Guilleminiana, Nees et Mont. in Syn. Hep. p. 293.—St. Vincent, Richmond Peak (210). Dominica, Morne Trois Pitons (764). Subgenus 11. OpontorEsEuNEA, Spruce. 19. L. tununara, Nees, in Syn. Hep. p. 8326.—Dominica, Morne Couronne (233), Morne Trois Pitons (760). 20. L. comRoPHYLLA, Spruce, in Trans. Bot. Soc. Edinb. xv.p.147 ; var. PAROICA.—St. Vincent, Mt. St. Andrews (146), Richmond Peak (356). Dominica, sine loc. (516). In foliis vivis. In the S. American plant described by me (op. ezt. p. 147) the inflorescence is monoicous (autoicous), the male organs being contained in separate branches from the female. But in the St. Vincent plant similar male branches have a female flower in the terminal bud, which is often fertile, thus constituting a paroicous inflorescence. As I can find no other difference wbat- ever, I must refer both forms to a single heteroicous species. Herr Stephani has lately referred this plant to ZL. Sieberiana, Gottsch., a Mauritian species, described in ‘Synopsis Hepatic.’ p- 328, with “foliis margine supero apiceque minute serrato- denticulatis, infero subintegerrimis ;”’ whereas L. cherophylla has the leaves equally and somewhat strongly dentate all round. Moreover, LZ. Siebertana is said to be sterile, and is theretore probably dioicous; whereas I have never seen ZL. cherophylia without perianths ; nor is any mention made of the cordate base of the stipules, so conspicuous in the latter. The plant grows always on living leaves, along with the two following Odonto- lejeunee, and sometimes with species of other subgenera. MR. RICHARD SPRUCE’S HEPATICE: ELLIOTTIAN #. 337 21. Leseunra convexistrpa, Lehm. et Lindenb. in Syn. Hep. p. 328.—St. Vincent, Gibson mtn. (2), Mt. St. Andrews (48, 66, and 145), Morne Garu (54), Richmond Peak (354). Dominica, Morne Couronne (245 and 778), windward road to the Lake (508), and sine loc. (516). 22. L. peruviana, Lehm. et Lindenb. in Syn. Hep. p. 339.—St. Vincent, Morne Garu (55), Richmond Peak (356). Dominica, Morne Couronne (245), and sine loc. (516). Var. PakvtstipuLa, Spruce.—St. Vincent, Richmond Peak (354). 23. L. accepEns, Gottsche, in Syn. Hep. p. 339.—St. Vincent, Mt. St. Andrews (84). Dominica, Morne Couronne (516). 24. L. surrvamensts, Mont. in Ann. Sc. Nat. sér. II. xv. p. 110 ; et in Syn. Hep. p. 329.—St. Vincent, Richmond Peak (356). Subgenus 12. Prrononeseunea, Spruce. The specimens referred to this subgenus are nearly all in small quantity, creeping over other hepatics or on living leaves, especially of ferns. In a final revision of my own plants I found several new species of this group, some of them in sufficient quantity to be distributed in my ‘ Exsiecate’ (Malton, 1892), and three of these reappear in Mr. Elliott’s collection. 25. L. penticunata, Nees, in Syn. Hep. p. 337.—St. Vincent, Richmond Peak (208). 26. L. vutcantca, Spruce,n. sp. (Tab. X XI. figs. 1-3.) Monoica, virescens, flaccida, caule prostrato sesquipollicari inequaliter pinnato et bipinnato. Folia subcontigua vel parum imbricata, late patentia, semicordato-ovata subdimidiata, acuta vel vix subacuminata, toto margine minute crenulata serrulatave, sinuato-complicata, lobulo 4plo breviore, semiovoideo subinflato subapiculato; cellule parve, subconformes, pellucide, trigonis nullis. Foliola distantia, obovato-rotunda, profunde _ bifida, segmentis acutis. Flores 2 in ramulo brevi simplice insidentes ; bractee foliis xquilonge, oblique lanceolate, lobulate ; bracteola anguste lanceolata, ad medium bifida, segmentis erectis acutis. Perianthia emersa, late obcordato-cuneata, compressa, margine supero subdenticulata. Andrecia folio caulino sublongiora, paucijuga.— Folia *6 x 4-45, lobulus +125, cellule q'y-s\y 3 foliola "13; bractee ‘55, bracteola *5 X‘2; perianthia ‘7 X°5 mm. Hab. Ins. Dominica, super alias hepaticas repens juxta craterem yolcanicum Grande Souffri¢re dictum, alt. 3000 ped. 338 MR. RICHARD SPRUCE’S HEPATICE ELLIOTTIAN®. (Ettrott, no. 875, Dee. 1892).—-Prionol. denticulate (Web.) conyenit floresc. monoica, ramulo 2 simplicissimo, differt foliis solum acutis, trigonis cellularum nullis, foliolis minutis, ete.— Pr. emula, G., sat similis, folia habet grosse serrata et perianthia subciliata. 27. LEJEUNEA LACINIATO-ALATA, Spruce, herb.— Dominica, Morne Micotrin (106), on Plagiochila Breuteliana. (Chimborazo, R. Spruce, anno 1860.) 28. L. pissttiFoita, Spruce, n. sp. Dioica, rarius monoica, pusilla, pallida, caule semipollicari, prostrato, fragili, vage ramoso. Fula dissita, patentia vel erecto- patentia, a basi constricta obovata, apiculato-acuta, raro sub- obtusa, subsinuatc-complicata, lobulo vix triplo breviore, ovoideo, inflato, margine toto fere muricato-serrulata ; celiule@ minutulre, trigonis vix ullis, ocello nullo. olzola parva, foliis 4plo fere breviora, orbiculata, ultra 3 bifida, sinu segmentisque acutis. Flores ramulo brevissimo simplice constantes; bractee foliis equilonge, sngustiores, lobulo plano, fortius serrulute. Peri- anthia emersa, obcordato-obconica, compressa, margine supero inequaliter spinulosa. Andrecia terminalia sepe longispica.— Folia 4X +25, lobulus *15, cellule 35; foliola -125 x*125; peri- anthia *6X*+ mm. Hab. Ins. Sti. Vincentii, in monte Richmond, supra Tricho- manis sp. frondem, ZL. Guadalupensi consociata (Et.iorr, no. 355 p. p., Feb. 1892).—Plantam ¢ ejusdem speciei possideo ad fluv. Bombonasa a meipso leectam anno 1857, necnon inter hepaticas Bolivianas a cl. Russy lectas inventam. 29. L. pecora, Tuyl. ?, in Syn. Hep. p. '755.—St. Vincent, Rich- mond Peak (208). 30. L. rracnyopes, Spruce, n. sp. Dioica, pusilla, rufo-badia, caule pollicari vage ramoso, ramis seepe elongatis polyphyllis. Folia dense imbricata, decurvo- concava, sepe subsecunda, oblique semicordato-oblonga, acuta, rarius subapiculata, tota facie externa cellulis prominulis tuber- culata, margine muricato-serrulata, valde sinuato-complicata, lobulo triplo breviore, turgide saccato, margine inflexo subhori- zontali ; cellule minutule, subopace, trigonis conspicuis. Foliola preminata, caulem vix excedentia, orbiculata, ad $ bifida, seg- mentis acutis. Flores 2 ramulo brevi simplice insidentes ; bracte@ foliis sat longivres, patentes, oblique ovales, subacu- > oo» a MR. RICHARD SPRUCE’S HEPATICE ELLIOTTIAN &. 339 minate, integerrime vel paucidentate, late lobulate ; bracteola ovali-lanceolata, ad 3 fere bifida, segmentis subacuminatis.— Folia +375 x °35 (cum lobulo) x°25 (sine lobulo), lubulus +125, cellule 3, ; foliola 07 ; bractee 5 mm. Hab. In monte Diablotin insule Dominice, super alias hepaticas reptans (Extiort, no. 646, Oct. 1892). 31. Leseunea a@uapaLupensis, Lindenb. in Syn. Hep. p. 340. (Tab. XXI. figs. 4, 5.)—St. Vincent, Richmond Peak (355), in Frichomane. Dominica, Roseau Valley (60), “a meagre variety differing in the nearly equal leaf-cells—not wider and laxer about the middle of the leaf .... LZ. guadalupensis is the only Prionolejeunea with quite entire leaves; but the compresso- trigonous perianths with toothed margins, seated on a short proper branch, declare its affinity with Z. denticulata, &c.” [Spruce in litt. Nov. 6th, 1893.] 32. L. vatipruscuna, Spruce, Hepat. Amaz. et And. exsice. anno 1892.—St. Vincent, Souffri¢re (227). Dominica, Morne Micotrin (86). This plaut, which was first found in the Eastern Peruvian Andes, agrees with L. microdonta, a common Amazonian species, in the very short fertile ramulus putting forth a minute subfloral innovation, but differs in its abruptly apiculate (uot verv obtuse) leaves, and its deeply obcordate perianth, which is bordered in the upper half by a narrow incised wing. 33. L. vaaans, Spruce, n. sp. Dioica, virescens, prostrata, flaccida, caule spe bipollicari, crebre sed inwqualiter bipiniato. Folia subimbricata, diver- gentia, decurva, oblique semicordato-ovata, subabrupte acuminu- lata, acuta, e cellulis marginalibus conicis, acutis obtusisve, den- ticulata, sinuato-complicata; lobulus triplo brevior, inflatus, oblique ovato-lageniformis, thargine apiceque involutus; cellule parve, trigonis conspicuis. oliola parva, plus triplo breviora, orbiculata, ad 4 bifida, segmentis acutis. Flores Q in ramo - brevi terminales, innovatione sepe elongata suffulti. Bractee tenere, foliis minores, oblique obovate, lobulo lineari aucte ; bracteola oblonga, bifida. Perianthia alte emersa, obcor- dato-cuneata, compressa, apice lato retuso subalato, setuloso erosulove. Andrecia folio collaterali vix longiora, 3—5-juga, in planta propria.—Folia ‘5 x°35, lobulus ‘18, cellule };3 foliola 15°15; bractee °4, bracteola °3 x°15; perianthia ‘75 X°6 wm, 340 MR, RICUARD SPRUCE’S HEPATICE ELLIOTTIAN E. Hab. Insula Sti. Vincentii, super montem Richmond, in Pla- giochilis repens (Etutort, no. 203, Feb. 1892). L. validiuscula (Hepat. Spruceane exsicc. anno 1892) affinis, distat tloribus monoicis, 2 ramulo perbrevi, innovatione sepius J suffulto ; foliis parvilobulatis, irigonis cellularum vix ullis ; foliolis duplo majoribus, profundius fissis; perianthiis apice profunde emarginatis. 34. LEsEUNEA EFrFusA, Spruce, nD. sp. Dioica, albicans, caule 2-pollicari, flaccido, fragili, vage ramoso, in matrice repens. Folia subimbricata, subdivergentia, decurvo- concava, a basi constricta semicordato-ovata, abrupte apiculata cuspidatave—caulina interdum solum acuta, toto ambitu celluloso- serrulata-crenulatave, valide sinuato-complicata ; lobulus triplo brevior, turgidus, saccatus, margine apiceque incurvis ; cellule parve, subequales, trigonis minutis vel vix ullis, pauce (2-4) sub lobuli margine ceteris majores, vacue, ocellum sistentes. Foliola parva, lobulis subbreviora, subrotunda, ad 3 obtuse bi- fida, segmentis obtusis rarius acutis. Andrecia plerumque brevispica. Flores 2 ramulo brevissimo simplice constantes. Bractee foliis sat minores, serrulate, lobo ovali-lanceolato acuto, lobulo vix duplo breviore obtuso; bracteola obovato- lanceolata, ad 4 fissa, sinu segmentisque acutis. Perianthta alte emersa, anguste obcordato-cuneata, compressa, subtus convexa, margine supra medium sparsim denticulata.— Folia °5 x ‘32, lobulus +16, cellule +, ; foliola ‘15 x°15 mm. Hab. In monte Chimborazo super Plagiochilas repens (R. Spruce, 1860). Situ simili montis Sti. Andree ins. Sti. Vincentii legit cl. Evtiorg, sub no. 83, anno 1892.—A Pr. denticulata distat florese. dioica, foliis haud acuminatis, cellularum trigonis sub- nullis, ete. 35. L. teprocarpta, Spruce, in Trans. Bot. Soc. Edinb. xv. p- 159.—St. Vincent, Mt. St. Andrews (84). Subgenus 13. DrepanoLEseuneEa, Spruce. 36. L. rncnoata, Meissn.in Syn. Hep. p. 343.—St. Vincent, Mariagua (22), Richmond Peak (356). Souffriére, forma foliis secundis (227).—Var. pertssoponta, S.; foliis spina longa margine antico auctis. Dominica, sine loc. (516). 37. L. anoptanrua, Spruce, in Trans. Bot. Soc. Edinb. Xv. p- 189.—Dominiva, Basin Will (505). MR. RICHARD SPRUCE’S HEPATICE ELLIOTTIAN ¥. 341 Subgenus 14. CaEerLoLreseuNnEA, Spruce. 38. Leseunra Ltingata, Lehm. et Lindenb. in Syn. Hep. p. 371.— St. Vincent, Souffri¢re (48), Mt. St. Andrews (69), Richmond Peak (163). Dominica, Morne Micotrin (90), Morne Diablotin (673). Not infrequent in both the islands, but rarely possessing good perianths. It grows among Plagiochila, Sendtnera, &c., and is one of the largest and most conspicuous plants in the genus. It is absent from equatorial America, but is found in Bourbon. Subgenus 15. HarpaLeseunera, Spruce. 39. L. TENUICUSPIS, Spruce, in Trans. Bot. Soc. Edinb. xv. p- 170.—St. Vincent, Richmond Peak (211), Richmond Valley (221), mixed with other hepatics. (Anne eadem ae ZL. oxy- phylla, Nees et Mont., Syn. Hep. p. 357, cui tamen stipule ademptee—forsan pretervise ?).—[ Mr. Spruce wrote L. longi- cuspis, which is unknown; but that he meant tenuicuspis IL have ascertained by the examination of original specimens which were kindly sent to me by Mr. Slater.—A. G. | 40. L. trrpEns, Spruce et Besch., in Bull. Soc. Bot. France (1890), p. elxxxi, pl. 13. (N.B.—The three teeth or spines are often much longer than they are represented in this plate.) — Dominica, Bouhomme mtn. (80) and elsewhere, but always sterile. 41. L. asperrtma, Spruce; Prionolejeunea asperrima, Spruce, in Trans. Bot. Soc. Edinb. xv. p. 160.—St. Vincent, Morne Garu (53). Subgenus 16. SrrepstLEsEUNEA, Spruce. 42. L. rweuexa, Hampe, in Syn. Hep. p. 350.—St. Vincent, Mt. St. Andrews (84). 43. L. invotuta, Gottsche, in Syn. Hep. p. 350.—St. Vincent, Mt. St. Andrews (48, 67). Dominica, Morne Couronne (233). 44. L. Kunrutana, Lindenb. in Syn. Hep. p. 353.—Dominica, Morne Micotrin (84), in Bazzani@ sp. repeus. Subgenus 17. CeratoLesEoNea, Spruce. 45. T.. prumuta, Spruce, in Trans. Bot. Soc. Edinb. xv. p. 201. —St. Vincent, Richmond Peak (217). 342 MR. RICHARD SPRUCE’S HEPATIC ELLIOTTIAN®. 46. LrsEUNEA sPINosA, Gottsche, in Syn. Hep. p. 402.—St. Vincent, Mt. St. Andrews (68 & 85—in